November 2023 - Kernel - mailweb.openeuler.org

[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