- Kernel - mailweb.openeuler.org

[PATCH 1/4] btrfs: delayed-inode: Kill the BUG_ON() in btrfs_delete_delayed_dir_index()
by Yang Yingliang 17 Apr '20

17 Apr '20

From: Qu Wenruo <wqu(a)suse.com> mainline inclusion from mainline-v5.4-rc1 commit 933c22a7512c5c09b1fdc46b557384efe8d03233 category: bugfix bugzilla: 13690 CVE: CVE-2019-19813 ------------------------------------------------- There is one report of fuzzed image which leads to BUG_ON() in btrfs_delete_delayed_dir_index(). Although that fuzzed image can already be addressed by enhanced extent-tree error handler, it's still better to hunt down more BUG_ON(). This patch will hunt down two BUG_ON()s in btrfs_delete_delayed_dir_index(): - One for error from btrfs_delayed_item_reserve_metadata() Instead of BUG_ON(), we output an error message and free the item. And return the error. All callers of this function handles the error by aborting current trasaction. - One for possible EEXIST from __btrfs_add_delayed_deletion_item() That function can return -EEXIST. We already have a good enough error message for that, only need to clean up the reserved metadata space and allocated item. To help above cleanup, also modifiy __btrfs_remove_delayed_item() called in btrfs_release_delayed_item(), to skip unassociated item. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=203253 Signed-off-by: Qu Wenruo <wqu(a)suse.com> Reviewed-by: David Sterba <dsterba(a)suse.com> Signed-off-by: David Sterba <dsterba(a)suse.com> Conflicts: fs/btrfs/delayed-inode.c [yyl: adjust context] Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/btrfs/delayed-inode.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index e9522f2..5dc6141 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -471,6 +471,9 @@ static void __btrfs_remove_delayed_item(struct btrfs_delayed_item *delayed_item) struct rb_root *root; struct btrfs_delayed_root *delayed_root; + /* Not associated with any delayed_node */ + if (!delayed_item->delayed_node) + return; delayed_root = delayed_item->delayed_node->root->fs_info->delayed_root; BUG_ON(!delayed_root); @@ -1526,7 +1529,12 @@ int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans, * we have reserved enough space when we start a new transaction, * so reserving metadata failure is impossible. */ - BUG_ON(ret); + if (ret < 0) { + btrfs_err(trans->fs_info, +"metadata reservation failed for delayed dir item deltiona, should have been reserved"); + btrfs_release_delayed_item(item); + goto end; + } mutex_lock(&node->mutex); ret = __btrfs_add_delayed_deletion_item(node, item); @@ -1534,7 +1542,8 @@ int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans, btrfs_err(trans->fs_info, "err add delayed dir index item(index: %llu) into the deletion tree of the delayed node(root id: %llu, inode id: %llu, errno: %d)", index, node->root->objectid, node->inode_id, ret); - BUG(); + btrfs_delayed_item_release_metadata(dir->root, item); + btrfs_release_delayed_item(item); } mutex_unlock(&node->mutex); end: -- 1.8.3

1 3

[PATCH 01/30] net: hns3: add one printing information in hnae3_unregister_client() function
by Yang Yingliang 17 Apr '20

17 Apr '20

From: shenhao <shenhao21(a)huawei.com> driver inclusion category: bugfix bugzilla: NA CVE: NA ----------------------------------------------------- This patch adds one printing information to let user know the client does not exist when unregister a noneexistent client. Signed-off-by: Guangbin Huang <huangguangbin2(a)huawei.com> Signed-off-by: shenhao <shenhao21(a)huawei.com> Reviewed-by: Zhong Zhaohui <zhongzhaohui(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/ethernet/hisilicon/hns3/hnae3.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.c b/drivers/net/ethernet/hisilicon/hns3/hnae3.c index 4afa509..53a87b3 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.c +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.c @@ -156,6 +156,7 @@ void hnae3_unregister_client(struct hnae3_client *client) if (!existed) { mutex_unlock(&hnae3_common_lock); + pr_err("client %s does not exist!\n", client->name); return; } -- 1.8.3

1 29

[PATCH 001/193] net: hns3: update hns3 version to 1.9.37.4
by Yang Yingliang 17 Apr '20

17 Apr '20

From: youshengzui <youshengzui(a)huawei.com> driver inclusion category: bugfix bugzilla: NA CVE: NA -------------------------- This patch is used to modify the hns3 driver version to 1.9.37.4 Signed-off-by: youshengzui <youshengzui(a)huawei.com> Reviewed-by: Weiwei Deng <dengweiwei(a)huawei.com> Reviewed-by: Zhaohui Zhong <zhongzhaohui(a)huawei.com> Reviewed-by: Junxin Chen <chenjunxin1(a)huawei.com> Reviewed-by: Zhong Zhaohui <zhongzhaohui(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/ethernet/hisilicon/hns3/hnae3.h | 2 +- drivers/net/ethernet/hisilicon/hns3/hns3_cae/hns3_cae_version.h | 2 +- drivers/net/ethernet/hisilicon/hns3/hns3_enet.h | 2 +- drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h | 2 +- drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h | 2 +- 5 files changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index f795bfd..98dfa7c 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -30,7 +30,7 @@ #include <linux/pci.h> #include <linux/types.h> -#define HNAE3_MOD_VERSION "1.9.37.3" +#define HNAE3_MOD_VERSION "1.9.37.4" #define HNAE3_MIN_VECTOR_NUM 2 /* one for msi-x, another for IO */ diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_cae/hns3_cae_version.h b/drivers/net/ethernet/hisilicon/hns3/hns3_cae/hns3_cae_version.h index 3977883..630f642 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_cae/hns3_cae_version.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_cae/hns3_cae_version.h @@ -4,7 +4,7 @@ #ifndef __HNS3_CAE_VERSION_H__ #define __HNS3_CAE_VERSION_H__ -#define HNS3_CAE_MOD_VERSION "1.9.37.3" +#define HNS3_CAE_MOD_VERSION "1.9.37.4" #define CMT_ID_LEN 8 #define RESV_LEN 3 diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h index 5f1d5a3..9e11ec3 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h @@ -8,7 +8,7 @@ #include "hnae3.h" -#define HNS3_MOD_VERSION "1.9.37.3" +#define HNS3_MOD_VERSION "1.9.37.4" extern char hns3_driver_version[]; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h index 0146470..5e64d2a 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h @@ -12,7 +12,7 @@ #include "hclge_cmd.h" #include "hnae3.h" -#define HCLGE_MOD_VERSION "1.9.37.3" +#define HCLGE_MOD_VERSION "1.9.37.4" #define HCLGE_DRIVER_NAME "hclge" #define HCLGE_MAX_PF_NUM 8 diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h index 51af1050..596618e 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h @@ -10,7 +10,7 @@ #include "hclgevf_cmd.h" #include "hnae3.h" -#define HCLGEVF_MOD_VERSION "1.9.37.3" +#define HCLGEVF_MOD_VERSION "1.9.37.4" #define HCLGEVF_DRIVER_NAME "hclgevf" #define HCLGEVF_MAX_VLAN_ID 4095 -- 1.8.3

1 192

[PATCH 01/40] perf/amd/uncore: Replace manual sampling check with CAP_NO_INTERRUPT flag
by Yang Yingliang 17 Apr '20

17 Apr '20

From: Kim Phillips <kim.phillips(a)amd.com> [ Upstream commit f967140dfb7442e2db0868b03b961f9c59418a1b ] Enable the sampling check in kernel/events/core.c::perf_event_open(), which returns the more appropriate -EOPNOTSUPP. BEFORE: $ sudo perf record -a -e instructions,l3_request_g1.caching_l3_cache_accesses true Error: The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (l3_request_g1.caching_l3_cache_accesses). /bin/dmesg | grep -i perf may provide additional information. With nothing relevant in dmesg. AFTER: $ sudo perf record -a -e instructions,l3_request_g1.caching_l3_cache_accesses true Error: l3_request_g1.caching_l3_cache_accesses: PMU Hardware doesn't support sampling/overflow-interrupts. Try 'perf stat' Fixes: c43ca5091a37 ("perf/x86/amd: Add support for AMD NB and L2I "uncore" counters") Signed-off-by: Kim Phillips <kim.phillips(a)amd.com> Signed-off-by: Borislav Petkov <bp(a)suse.de> Acked-by: Peter Zijlstra <peterz(a)infradead.org> Cc: stable(a)vger.kernel.org Link: https://lkml.kernel.org/r/20200311191323.13124-1-kim.phillips@amd.com Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/x86/events/amd/uncore.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/arch/x86/events/amd/uncore.c b/arch/x86/events/amd/uncore.c index baa7e36..604a855 100644 --- a/arch/x86/events/amd/uncore.c +++ b/arch/x86/events/amd/uncore.c @@ -193,20 +193,18 @@ static int amd_uncore_event_init(struct perf_event *event) /* * NB and Last level cache counters (MSRs) are shared across all cores - * that share the same NB / Last level cache. Interrupts can be directed - * to a single target core, however, event counts generated by processes - * running on other cores cannot be masked out. So we do not support - * sampling and per-thread events. + * that share the same NB / Last level cache. On family 16h and below, + * Interrupts can be directed to a single target core, however, event + * counts generated by processes running on other cores cannot be masked + * out. So we do not support sampling and per-thread events via + * CAP_NO_INTERRUPT, and we do not enable counter overflow interrupts: */ - if (is_sampling_event(event) || event->attach_state & PERF_ATTACH_TASK) - return -EINVAL; /* NB and Last level cache counters do not have usr/os/guest/host bits */ if (event->attr.exclude_user || event->attr.exclude_kernel || event->attr.exclude_host || event->attr.exclude_guest) return -EINVAL; - /* and we do not enable counter overflow interrupts */ hwc->config = event->attr.config & AMD64_RAW_EVENT_MASK_NB; hwc->idx = -1; @@ -314,6 +312,7 @@ static ssize_t amd_uncore_attr_show_cpumask(struct device *dev, .start = amd_uncore_start, .stop = amd_uncore_stop, .read = amd_uncore_read, + .capabilities = PERF_PMU_CAP_NO_INTERRUPT, }; static struct pmu amd_llc_pmu = { @@ -324,6 +323,7 @@ static ssize_t amd_uncore_attr_show_cpumask(struct device *dev, .start = amd_uncore_start, .stop = amd_uncore_stop, .read = amd_uncore_read, + .capabilities = PERF_PMU_CAP_NO_INTERRUPT, }; static struct amd_uncore *amd_uncore_alloc(unsigned int cpu) -- 1.8.3

1 39

[PATCH 1/9] hpre: dfx for IO operation and delay
by Yang Yingliang 17 Apr '20

17 Apr '20

From: Yu'an Wang <wangyuan46(a)huawei.com> driver inclusion category: feature bugzilla: NA CVE: NA In this patch, we try to add dfx for io operation, including send/ recv/send_fail/send_busy. We also can define overtime_threshold to judge timeout task. Signed-off-by: Yu'an Wang <wangyuan46(a)huawei.com> Reviewed-by: Mingqiang Ling <lingmingqiang(a)huawei.com> Reviewed-by: Guangwei Zhou <zhouguangwei5(a)huawei.com> Reviewed-by: Ye Kai <yekai13(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/crypto/hisilicon/hpre/hpre.h | 16 ++++++ drivers/crypto/hisilicon/hpre/hpre_crypto.c | 89 ++++++++++++++++++++++++----- drivers/crypto/hisilicon/hpre/hpre_main.c | 55 ++++++++++++++++++ 3 files changed, 146 insertions(+), 14 deletions(-) diff --git a/drivers/crypto/hisilicon/hpre/hpre.h b/drivers/crypto/hisilicon/hpre/hpre.h index 42b2f2a..203eb2a 100644 --- a/drivers/crypto/hisilicon/hpre/hpre.h +++ b/drivers/crypto/hisilicon/hpre/hpre.h @@ -25,6 +25,16 @@ enum hpre_ctrl_dbgfs_file { HPRE_DEBUG_FILE_NUM, }; +enum hpre_dfx_dbgfs_file { + HPRE_SEND_CNT, + HPRE_RECV_CNT, + HPRE_SEND_FAIL_CNT, + HPRE_SEND_BUSY_CNT, + HPRE_OVER_THRHLD_CNT, + HPRE_OVERTIME_THRHLD, + HPRE_DFX_FILE_NUM +}; + #define HPRE_DEBUGFS_FILE_NUM (HPRE_DEBUG_FILE_NUM + HPRE_CLUSTERS_NUM - 1) struct hpre_debugfs_file { @@ -34,12 +44,18 @@ struct hpre_debugfs_file { struct hpre_debug *debug; }; +struct hpre_dfx { + atomic64_t value; + enum hpre_dfx_dbgfs_file type; +}; + /* * One HPRE controller has one PF and multiple VFs, some global configurations * which PF has need this structure. * Just relevant for PF. */ struct hpre_debug { + struct hpre_dfx dfx[HPRE_DFX_FILE_NUM]; struct hpre_debugfs_file files[HPRE_DEBUGFS_FILE_NUM]; }; diff --git a/drivers/crypto/hisilicon/hpre/hpre_crypto.c b/drivers/crypto/hisilicon/hpre/hpre_crypto.c index 7610e13..b68b30c 100644 --- a/drivers/crypto/hisilicon/hpre/hpre_crypto.c +++ b/drivers/crypto/hisilicon/hpre/hpre_crypto.c @@ -10,6 +10,7 @@ #include <linux/dma-mapping.h> #include <linux/fips.h> #include <linux/module.h> +#include <linux/time.h> #include "hpre.h" struct hpre_ctx; @@ -68,6 +69,7 @@ struct hpre_dh_ctx { struct hpre_ctx { struct hisi_qp *qp; struct hpre_asym_request **req_list; + struct hpre *hpre; spinlock_t req_lock; unsigned int key_sz; bool crt_g2_mode; @@ -90,6 +92,7 @@ struct hpre_asym_request { int err; int req_id; hpre_cb cb; + struct timespec64 req_time; }; static DEFINE_MUTEX(hpre_alg_lock); @@ -119,6 +122,7 @@ static void hpre_free_req_id(struct hpre_ctx *ctx, int req_id) static int hpre_add_req_to_ctx(struct hpre_asym_request *hpre_req) { struct hpre_ctx *ctx; + struct hpre_dfx *dfx; int id; ctx = hpre_req->ctx; @@ -129,6 +133,10 @@ static int hpre_add_req_to_ctx(struct hpre_asym_request *hpre_req) ctx->req_list[id] = hpre_req; hpre_req->req_id = id; + dfx = ctx->hpre->debug.dfx; + if (atomic64_read(&dfx[HPRE_OVERTIME_THRHLD].value)) + ktime_get_ts64(&hpre_req->req_time); + return id; } @@ -308,12 +316,16 @@ static int hpre_alg_res_post_hf(struct hpre_ctx *ctx, struct hpre_sqe *sqe, static int hpre_ctx_set(struct hpre_ctx *ctx, struct hisi_qp *qp, int qlen) { + struct hpre *hpre; + if (!ctx || !qp || qlen < 0) return -EINVAL; spin_lock_init(&ctx->req_lock); ctx->qp = qp; + hpre = container_of(ctx->qp->qm, struct hpre, qm); + ctx->hpre = hpre; ctx->req_list = kcalloc(qlen, sizeof(void *), GFP_KERNEL); if (!ctx->req_list) return -ENOMEM; @@ -336,30 +348,67 @@ static void hpre_ctx_clear(struct hpre_ctx *ctx, bool is_clear_all) ctx->key_sz = 0; } +static bool hpre_is_bd_timeout(struct hpre_asym_request *req, + u64 overtime_thrhld) +{ + struct timespec64 reply_time; + u64 time_use_us; + +#define HPRE_DFX_SEC_TO_US 1000000 +#define HPRE_DFX_US_TO_NS 1000 + + ktime_get_ts64(&reply_time); + time_use_us = (reply_time.tv_sec - req->req_time.tv_sec) * + HPRE_DFX_SEC_TO_US + + (reply_time.tv_nsec - req->req_time.tv_nsec) / + HPRE_DFX_US_TO_NS; + + if (time_use_us <= overtime_thrhld) + return false; + + return true; +} + static void hpre_dh_cb(struct hpre_ctx *ctx, void *resp) { + struct hpre_dfx *dfx = ctx->hpre->debug.dfx; struct hpre_asym_request *req; struct kpp_request *areq; + u64 overtime_thrhld; int ret; ret = hpre_alg_res_post_hf(ctx, resp, (void **)&req); areq = req->areq.dh; areq->dst_len = ctx->key_sz; + + overtime_thrhld = atomic64_read(&dfx[HPRE_OVERTIME_THRHLD].value); + if (overtime_thrhld && hpre_is_bd_timeout(req, overtime_thrhld)) + atomic64_inc(&dfx[HPRE_OVER_THRHLD_CNT].value); + hpre_hw_data_clr_all(ctx, req, areq->dst, areq->src); kpp_request_complete(areq, ret); + atomic64_inc(&dfx[HPRE_RECV_CNT].value); } static void hpre_rsa_cb(struct hpre_ctx *ctx, void *resp) { + struct hpre_dfx *dfx = ctx->hpre->debug.dfx; struct hpre_asym_request *req; struct akcipher_request *areq; + u64 overtime_thrhld; int ret; ret = hpre_alg_res_post_hf(ctx, resp, (void **)&req); + + overtime_thrhld = atomic64_read(&dfx[HPRE_OVERTIME_THRHLD].value); + if (overtime_thrhld && hpre_is_bd_timeout(req, overtime_thrhld)) + atomic64_inc(&dfx[HPRE_OVER_THRHLD_CNT].value); + areq = req->areq.rsa; areq->dst_len = ctx->key_sz; hpre_hw_data_clr_all(ctx, req, areq->dst, areq->src); akcipher_request_complete(areq, ret); + atomic64_inc(&dfx[HPRE_RECV_CNT].value); } static void hpre_alg_cb(struct hisi_qp *qp, void *resp) @@ -435,6 +484,29 @@ static int hpre_msg_request_set(struct hpre_ctx *ctx, void *req, bool is_rsa) return 0; } +static int hpre_send(struct hpre_ctx *ctx, struct hpre_sqe *msg) +{ + struct hpre_dfx *dfx = ctx->hpre->debug.dfx; + int ctr = 0; + int ret; + + do { + atomic64_inc(&dfx[HPRE_SEND_CNT].value); + ret = hisi_qp_send(ctx->qp, msg); + if (ret != -EBUSY) + break; + atomic64_inc(&dfx[HPRE_SEND_BUSY_CNT].value); + } while (ctr++ < HPRE_TRY_SEND_TIMES); + + if (likely(!ret)) + return ret; + + if (ret != -EBUSY) + atomic64_inc(&dfx[HPRE_SEND_FAIL_CNT].value); + + return ret; +} + #ifdef CONFIG_CRYPTO_DH static int hpre_dh_compute_value(struct kpp_request *req) { @@ -443,7 +515,6 @@ static int hpre_dh_compute_value(struct kpp_request *req) void *tmp = kpp_request_ctx(req); struct hpre_asym_request *hpre_req = PTR_ALIGN(tmp, HPRE_ALIGN_SZ); struct hpre_sqe *msg = &hpre_req->req; - int ctr = 0; int ret; ret = hpre_msg_request_set(ctx, req, false); @@ -464,11 +535,9 @@ static int hpre_dh_compute_value(struct kpp_request *req) msg->dw0 = cpu_to_le32(le32_to_cpu(msg->dw0) | HPRE_ALG_DH_G2); else msg->dw0 = cpu_to_le32(le32_to_cpu(msg->dw0) | HPRE_ALG_DH); - do { - ret = hisi_qp_send(ctx->qp, msg); - } while (ret == -EBUSY && ctr++ < HPRE_TRY_SEND_TIMES); /* success */ + ret = hpre_send(ctx, msg); if (likely(!ret)) return -EINPROGRESS; @@ -646,7 +715,6 @@ static int hpre_rsa_enc(struct akcipher_request *req) void *tmp = akcipher_request_ctx(req); struct hpre_asym_request *hpre_req = PTR_ALIGN(tmp, HPRE_ALIGN_SZ); struct hpre_sqe *msg = &hpre_req->req; - int ctr = 0; int ret; /* For 512 and 1536 bits key size, use soft tfm instead */ @@ -676,11 +744,8 @@ static int hpre_rsa_enc(struct akcipher_request *req) if (unlikely(ret)) goto clear_all; - do { - ret = hisi_qp_send(ctx->qp, msg); - } while (ret == -EBUSY && ctr++ < HPRE_TRY_SEND_TIMES); - /* success */ + ret = hpre_send(ctx, msg); if (likely(!ret)) return -EINPROGRESS; @@ -698,7 +763,6 @@ static int hpre_rsa_dec(struct akcipher_request *req) void *tmp = akcipher_request_ctx(req); struct hpre_asym_request *hpre_req = PTR_ALIGN(tmp, HPRE_ALIGN_SZ); struct hpre_sqe *msg = &hpre_req->req; - int ctr = 0; int ret; /* For 512 and 1536 bits key size, use soft tfm instead */ @@ -735,11 +799,8 @@ static int hpre_rsa_dec(struct akcipher_request *req) if (unlikely(ret)) goto clear_all; - do { - ret = hisi_qp_send(ctx->qp, msg); - } while (ret == -EBUSY && ctr++ < HPRE_TRY_SEND_TIMES); - /* success */ + ret = hpre_send(ctx, msg); if (likely(!ret)) return -EINPROGRESS; diff --git a/drivers/crypto/hisilicon/hpre/hpre_main.c b/drivers/crypto/hisilicon/hpre/hpre_main.c index f727158..2ede8d78 100644 --- a/drivers/crypto/hisilicon/hpre/hpre_main.c +++ b/drivers/crypto/hisilicon/hpre/hpre_main.c @@ -169,6 +169,15 @@ struct hpre_hw_error { {"INT_STATUS ", HPRE_INT_STATUS}, }; +static const char *hpre_dfx_files[HPRE_DFX_FILE_NUM] = { + "send_cnt", + "recv_cnt", + "send_fail_cnt", + "send_busy_cnt", + "over_thrhld_cnt", + "overtime_thrhld" +}; + #ifdef CONFIG_CRYPTO_QM_UACCE static int uacce_mode_set(const char *val, const struct kernel_param *kp) { @@ -588,6 +597,33 @@ static ssize_t hpre_ctrl_debug_write(struct file *filp, const char __user *buf, .write = hpre_ctrl_debug_write, }; +static int hpre_debugfs_atomic64_get(void *data, u64 *val) +{ + struct hpre_dfx *dfx_item = data; + + *val = atomic64_read(&dfx_item->value); + return 0; +} + +static int hpre_debugfs_atomic64_set(void *data, u64 val) +{ + struct hpre_dfx *dfx_item = data; + + if (dfx_item->type == HPRE_OVERTIME_THRHLD) { + struct hpre_dfx *hpre_dfx = dfx_item - HPRE_OVERTIME_THRHLD; + + atomic64_set(&hpre_dfx[HPRE_OVER_THRHLD_CNT].value, 0); + } else if (val) { + return -EINVAL; + } + + atomic64_set(&dfx_item->value, val); + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(hpre_atomic64_ops, hpre_debugfs_atomic64_get, + hpre_debugfs_atomic64_set, "%llu\n"); + static int hpre_create_debugfs_file(struct hpre_debug *dbg, struct dentry *dir, enum hpre_ctrl_dbgfs_file type, int indx) { @@ -691,6 +727,22 @@ static int hpre_ctrl_debug_init(struct hpre_debug *debug) return hpre_cluster_debugfs_init(debug); } +static void hpre_dfx_debug_init(struct hpre_debug *debug) +{ + struct hpre *hpre = container_of(debug, struct hpre, debug); + struct hpre_dfx *dfx = hpre->debug.dfx; + struct hisi_qm *qm = &hpre->qm; + struct dentry *parent; + int i; + + parent = debugfs_create_dir("hpre_dfx", qm->debug.debug_root); + for (i = 0; i < HPRE_DFX_FILE_NUM; i++) { + dfx[i].type = i; + debugfs_create_file(hpre_dfx_files[i], 0644, parent, &dfx[i], + &hpre_atomic64_ops); + } +} + static int hpre_debugfs_init(struct hisi_qm *qm) { struct hpre *hpre = container_of(qm, struct hpre, qm); @@ -709,6 +761,9 @@ static int hpre_debugfs_init(struct hisi_qm *qm) if (ret) goto failed_to_create; } + + hpre_dfx_debug_init(&hpre->debug); + return 0; failed_to_create: -- 1.8.3

1 8

[PATCH 01/92] KVM: SVM: fix up incorrect backport
by Yang Yingliang 17 Apr '20

17 Apr '20

From: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> When I backported 52918ed5fcf0 ("KVM: SVM: Override default MMIO mask if memory encryption is enabled") to 4.19 (which resulted in commit a4e761c9f63a ("KVM: SVM: Override default MMIO mask if memory encryption is enabled")), I messed up the call to kvm_mmu_set_mmio_spte_mask() Fix that here now. Reported-by: Tom Lendacky <thomas.lendacky(a)amd.com> Cc: Sean Christopherson <sean.j.christopherson(a)intel.com> Cc: Paolo Bonzini <pbonzini(a)redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/x86/kvm/svm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 3f0565e..cc8f3b41 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -1336,7 +1336,7 @@ static __init void svm_adjust_mmio_mask(void) */ mask = (mask_bit < 52) ? rsvd_bits(mask_bit, 51) | PT_PRESENT_MASK : 0; - kvm_mmu_set_mmio_spte_mask(mask, PT_WRITABLE_MASK | PT_USER_MASK); + kvm_mmu_set_mmio_spte_mask(mask, mask); } static __init int svm_hardware_setup(void) -- 1.8.3

1 91

[PATCH 1/5] Revert "dm crypt: fix benbi IV constructor crash if used in authenticated mode"
by Yang Yingliang 17 Apr '20

17 Apr '20

From: Xiongfeng Wang <wangxiongfeng2(a)huawei.com> hulk inclusion category: bugfix bugzilla: 31797 CVE: NA ------------------------- The next two patches to be reverted will conflict with this patch. Let's revert this patch and merge the original patch. This reverts commit 808f3768deec5e1a5c9d2a4a2d8593fbbaf3e4cc. Signed-off-by: Xiongfeng Wang <wangxiongfeng2(a)huawei.com> Reviewed-by: ZhangXiaoxu <zhangxiaoxu5(a)huawei.com> Reviewed-by: Hou Tao <houtao1(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/md/dm-crypt.c | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-) diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index d451f98..aa7f741 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -563,14 +563,8 @@ static int crypt_iv_essiv_gen(struct geniv_ctx *ctx, static int crypt_iv_benbi_ctr(struct geniv_ctx *ctx) { - unsigned bs; - int log; - - if (test_bit(CRYPT_MODE_INTEGRITY_AEAD, &ctx->cipher_flags)) - bs = crypto_aead_blocksize(ctx->tfms.tfms_aead[0]); - else - bs = crypto_skcipher_blocksize(ctx->tfms.tfms[0]); - log = ilog2(bs); + unsigned int bs = crypto_skcipher_blocksize(ctx->tfms.tfms[0]); + int log = ilog2(bs); /* we need to calculate how far we must shift the sector count * to get the cipher block count, we use this shift in _gen */ -- 1.8.3

1 4

[PATCH] arm64: clear_page: Add new implementation of clear_page() by STNP
by Yang Yingliang 17 Apr '20

17 Apr '20

From: Wei Li <liwei391(a)huawei.com> hulk inclusion category: feature bugzilla: 31400 CVE: NA --------------------------- Currently, clear_page() clear the page through 'dc zva', while the page may not be used immediately mostly, so the cache flush is in vain. Add an optimized implementation of clear_page() by 'stnp' for performance promotion. It can be switched by the boot cmdline 'mm.use_clearpage_stnp'. In the hugetlb clear test, we gained about 53.7% performance improvement: Set mm.use_clearpage_stnp = 0 | Set mm.use_clearpage_stnp = 1 [root@localhost liwei]# ./a.out 50 20 | [root@localhost liwei]# ./a.out 50 20 size is 50 Gib, test times is 20 | size is 50 Gib, test times is 20 test_time[0] : use 8.438046 sec | test_time[0] : use 3.722682 sec test_time[1] : use 8.028493 sec | test_time[1] : use 3.640274 sec test_time[2] : use 8.646547 sec | test_time[2] : use 4.095052 sec test_time[3] : use 8.122490 sec | test_time[3] : use 3.998446 sec test_time[4] : use 8.053038 sec | test_time[4] : use 4.084259 sec test_time[5] : use 8.843512 sec | test_time[5] : use 3.933871 sec test_time[6] : use 8.308906 sec | test_time[6] : use 3.934334 sec test_time[7] : use 8.093817 sec | test_time[7] : use 3.869142 sec test_time[8] : use 8.303504 sec | test_time[8] : use 3.902916 sec test_time[9] : use 8.178336 sec | test_time[9] : use 3.541885 sec test_time[10] : use 8.003625 sec | test_time[10] : use 3.595554 sec test_time[11] : use 8.163807 sec | test_time[11] : use 3.583813 sec test_time[12] : use 8.267464 sec | test_time[12] : use 3.863033 sec test_time[13] : use 8.055326 sec | test_time[13] : use 3.770953 sec test_time[14] : use 8.246986 sec | test_time[14] : use 3.808006 sec test_time[15] : use 8.546992 sec | test_time[15] : use 3.653194 sec test_time[16] : use 8.727256 sec | test_time[16] : use 3.722395 sec test_time[17] : use 8.288951 sec | test_time[17] : use 3.683508 sec test_time[18] : use 8.019322 sec | test_time[18] : use 4.253087 sec test_time[19] : use 8.250685 sec | test_time[19] : use 4.082845 sec hugetlb test end! | hugetlb test end! Signed-off-by: Wei Li <liwei391(a)huawei.com> Reviewed-by: Hanjun Guo <guohanjun(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/arm64/include/asm/cpucaps.h | 3 ++- arch/arm64/kernel/cpufeature.c | 34 ++++++++++++++++++++++++++++++++++ arch/arm64/lib/clear_page.S | 21 +++++++++++++++++++++ 3 files changed, 57 insertions(+), 1 deletion(-) diff --git a/arch/arm64/include/asm/cpucaps.h b/arch/arm64/include/asm/cpucaps.h index a9090f2..3cd169f 100644 --- a/arch/arm64/include/asm/cpucaps.h +++ b/arch/arm64/include/asm/cpucaps.h @@ -56,7 +56,8 @@ #define ARM64_WORKAROUND_1463225 35 #define ARM64_HAS_CRC32 36 #define ARM64_SSBS 37 +#define ARM64_CLEARPAGE_STNP 38 -#define ARM64_NCAPS 38 +#define ARM64_NCAPS 39 #endif /* __ASM_CPUCAPS_H */ diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c index b1f621c..8b84a47 100644 --- a/arch/arm64/kernel/cpufeature.c +++ b/arch/arm64/kernel/cpufeature.c @@ -1232,6 +1232,34 @@ static bool can_use_gic_priorities(const struct arm64_cpu_capabilities *entry, } #endif +static bool use_clearpage_stnp; + +static int __init early_use_clearpage_stnp(char *p) +{ + return strtobool(p, &use_clearpage_stnp); +} +early_param("mm.use_clearpage_stnp", early_use_clearpage_stnp); + +static bool has_mor_nontemporal(const struct arm64_cpu_capabilities *entry) +{ + /* + * List of CPUs which have memory ordering ruled non-temporal + * load and store. + */ + static const struct midr_range cpus[] = { + MIDR_ALL_VERSIONS(MIDR_HISI_TSV110), + {}, + }; + + return is_midr_in_range_list(read_cpuid_id(), cpus); +} + +static bool can_clearpage_use_stnp(const struct arm64_cpu_capabilities *entry, + int scope) +{ + return use_clearpage_stnp && has_mor_nontemporal(entry); +} + static const struct arm64_cpu_capabilities arm64_features[] = { { .desc = "GIC system register CPU interface", @@ -1467,6 +1495,12 @@ static bool can_use_gic_priorities(const struct arm64_cpu_capabilities *entry, .cpu_enable = cpu_enable_ssbs, }, #endif + { + .desc = "Clear Page by STNP", + .capability = ARM64_CLEARPAGE_STNP, + .type = ARM64_CPUCAP_SYSTEM_FEATURE, + .matches = can_clearpage_use_stnp, + }, {}, }; diff --git a/arch/arm64/lib/clear_page.S b/arch/arm64/lib/clear_page.S index ef08e90..9aa1de1 100644 --- a/arch/arm64/lib/clear_page.S +++ b/arch/arm64/lib/clear_page.S @@ -18,6 +18,25 @@ #include <linux/const.h> #include <asm/assembler.h> #include <asm/page.h> +#include <asm/alternative.h> + +/* + * Clear page @dest + * + * Parameters: + * x0 - dest + */ +ENTRY(clear_page_stnp) + .align 6 +1: stnp xzr, xzr, [x0] + stnp xzr, xzr, [x0, #0x10] + stnp xzr, xzr, [x0, #0x20] + stnp xzr, xzr, [x0, #0x30] + add x0, x0, #0x40 + tst x0, #(PAGE_SIZE - 1) + b.ne 1b + ret +ENDPROC(clear_page_stnp) /* * Clear page @dest @@ -26,6 +45,8 @@ * x0 - dest */ ENTRY(clear_page) + ALTERNATIVE("nop", "b clear_page_stnp", ARM64_CLEARPAGE_STNP) + mrs x1, dczid_el0 and w1, w1, #0xf mov x2, #4 -- 1.8.3

1 0

[PATCH] RDMA/hns: optimize mtr management and fix mtr addressing bug
by Yang Yingliang 17 Apr '20

17 Apr '20

From: Tao Jihua <taojihua4(a)huawei.com> driver inclusion category: Bugfix bugzilla: NA This modification is mainly to optimize mtr management and solve mtr addressing bug:When mtt_ba_pg_sz = 0, hem-> start / step = 1, which eventually results in an additional BA_BYTE_LEN added to the offset Signed-off-by: Tao Jihua <taojihua4(a)huawei.com> Reviewed-by: Hu Chunzhi <huchunzhi(a)huawei.com> Reviewed-by: Wang Lin <wanglin137(a)huawei.com> Reviewed-by: Zhao Weibo <zhaoweibo3(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/infiniband/hw/hns/hns_roce_hem.c | 43 ++++++++++++++++---------------- 1 file changed, 21 insertions(+), 22 deletions(-) diff --git a/drivers/infiniband/hw/hns/hns_roce_hem.c b/drivers/infiniband/hw/hns/hns_roce_hem.c index 510c008..1911470 100644 --- a/drivers/infiniband/hw/hns/hns_roce_hem.c +++ b/drivers/infiniband/hw/hns/hns_roce_hem.c @@ -1165,13 +1165,6 @@ struct roce_hem_item { int end; /* end buf offset in this hem */ }; -#define hem_list_for_each(pos, n, head) \ - list_for_each_entry_safe(pos, n, head, list) - -#define hem_list_del_item(hem) list_del(&hem->list) -#define hem_list_add_item(hem, head) list_add(&hem->list, head) -#define hem_list_link_item(hem, head) list_add(&hem->sibling, head) - static struct roce_hem_item *hem_list_alloc_item(struct hns_roce_dev *hr_dev, int start, int end, int count, bool exist_bt, @@ -1216,8 +1209,8 @@ static void hem_list_free_all(struct hns_roce_dev *hr_dev, { struct roce_hem_item *hem, *temp_hem; - hem_list_for_each(hem, temp_hem, head) { - hem_list_del_item(hem); + list_for_each_entry_safe(hem, temp_hem, head, list) { + list_del(&hem->list); hem_list_free_item(hr_dev, hem, exist_bt); } } @@ -1249,7 +1242,7 @@ static struct roce_hem_item *hem_list_search_item(struct list_head *ba_list, struct roce_hem_item *hem, *temp_hem; struct roce_hem_item *found = NULL; - hem_list_for_each(hem, temp_hem, ba_list) { + list_for_each_entry_safe(hem, temp_hem, ba_list, list) { if (hem_list_page_is_in_range(hem, page_offset)) { found = hem; break; @@ -1391,9 +1384,9 @@ static int hem_list_alloc_mid_bt(struct hns_roce_dev *hr_dev, goto err_exit; } hem_ptrs[level] = cur; - hem_list_add_item(cur, &temp_list[level]); + list_add(&cur->list, &temp_list[level]); if (hem_list_is_bottom_bt(hopnum, level)) - hem_list_link_item(cur, &temp_list[0]); + list_add(&cur->sibling, &temp_list[0]); /* link bt to parent bt */ if (level > 1) { @@ -1430,6 +1423,7 @@ static int hem_list_alloc_root_bt(struct hns_roce_dev *hr_dev, void *cpu_base; u64 phy_base; int ret = 0; + int ba_num; int offset; int total; int step; @@ -1440,15 +1434,19 @@ static int hem_list_alloc_root_bt(struct hns_roce_dev *hr_dev, if (root_hem) return 0; + ba_num = hns_roce_hem_list_calc_root_ba(regions, region_cnt, unit); + if (ba_num < 1) + return -ENOMEM; + INIT_LIST_HEAD(&temp_root); - total = r->offset; + offset = r->offset; /* indicate to last region */ r = &regions[region_cnt - 1]; - root_hem = hem_list_alloc_item(hr_dev, total, r->offset + r->count - 1, - unit, true, 0); + root_hem = hem_list_alloc_item(hr_dev, offset, r->offset + r->count - 1, + ba_num, true, 0); if (!root_hem) return -ENOMEM; - hem_list_add_item(root_hem, &temp_root); + list_add(&root_hem->list, &temp_root); hem_list->root_ba = root_hem->dma_addr; @@ -1457,7 +1455,7 @@ static int hem_list_alloc_root_bt(struct hns_roce_dev *hr_dev, INIT_LIST_HEAD(&temp_list[i]); total = 0; - for (i = 0; i < region_cnt && total < unit; i++) { + for (i = 0; i < region_cnt && total < ba_num; i++) { r = &regions[i]; if (!r->count) continue; @@ -1478,8 +1476,8 @@ static int hem_list_alloc_root_bt(struct hns_roce_dev *hr_dev, goto err_exit; } hem_list_assign_bt(hr_dev, hem, cpu_base, phy_base); - hem_list_add_item(hem, &temp_list[i]); - hem_list_link_item(hem, &temp_btm); + list_add(&hem->list, &temp_list[i]); + list_add(&hem->sibling, &temp_btm); total += r->count; } else { step = hem_list_calc_ba_range(r->hopnum, 1, unit); @@ -1488,9 +1486,10 @@ static int hem_list_alloc_root_bt(struct hns_roce_dev *hr_dev, goto err_exit; } /* if exist mid bt, link L1 to L0 */ - hem_list_for_each(hem, temp_hem, - &hem_list->mid_bt[i][1]) { - offset = hem->start / step * BA_BYTE_LEN; + list_for_each_entry_safe(hem, temp_hem, + &hem_list->mid_bt[i][1], list) { + offset = ((hem->start - r->offset) / step) * + BA_BYTE_LEN; hem_list_link_bt(hr_dev, cpu_base + offset, hem->dma_addr); total++; -- 1.8.3

1 0

[PATCH 001/160] irqchip/gic-v3-its: Fix misuse of GENMASK macro
by Yang Yingliang 17 Apr '20

17 Apr '20

From: Joe Perches <joe(a)perches.com> [ Upstream commit 20faba848752901de23a4d45a1174d64d2069dde ] Arguments are supposed to be ordered high then low. Signed-off-by: Joe Perches <joe(a)perches.com> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Acked-by: Marc Zyngier <marc.zyngier(a)arm.com> Link: https://lkml.kernel.org/r/ab5deb4fc3cd604cb620054770b7d00016d736bc.15627348… Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/irqchip/irq-gic-v3-its.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/irqchip/irq-gic-v3-its.c b/drivers/irqchip/irq-gic-v3-its.c index 4a9c14f..860f3ef 100644 --- a/drivers/irqchip/irq-gic-v3-its.c +++ b/drivers/irqchip/irq-gic-v3-its.c @@ -215,7 +215,7 @@ static struct its_collection *dev_event_to_col(struct its_device *its_dev, static struct its_collection *valid_col(struct its_collection *col) { - if (WARN_ON_ONCE(col->target_address & GENMASK_ULL(0, 15))) + if (WARN_ON_ONCE(col->target_address & GENMASK_ULL(15, 0))) return NULL; return col; -- 1.8.3

1 159