- Kernel - mailweb.openeuler.org

[PATCH OLK-5.10] mm/ksm: Remove the ksm_merge_any status
by Longlong Xia 08 Aug '23

08 Aug '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7R48G ---------------------------------------------------------------------- Remove the ksm_merge_any status, as the ksm status of processes within the same cgroup may not be consistent, since the prctl interface can also change the ksm status of a process. Therefore, the ksm_merge_any status may not always represent the ksm status of all processes within a cgroup. Remove it. Signed-off-by: Longlong Xia <xialonglong1(a)huawei.com> --- include/linux/memcontrol.h | 4 ---- mm/memcontrol.c | 9 --------- 2 files changed, 13 deletions(-) diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h index 16b3666393f2..0219e670a436 100644 --- a/include/linux/memcontrol.h +++ b/include/linux/memcontrol.h @@ -402,11 +402,7 @@ struct mem_cgroup { KABI_RESERVE(3) KABI_RESERVE(4) #endif -#ifdef CONFIG_KSM - KABI_USE(5, bool ksm_merge_any) -#else KABI_RESERVE(5) -#endif #if defined(CONFIG_DYNAMIC_HUGETLB) && defined(CONFIG_ARM64) KABI_USE(6, struct dhugetlb_pool *hpool) #else diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 5082d0664adf..c32010692a1b 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -5724,9 +5724,6 @@ static int memcg_set_ksm_for_tasks(struct mem_cgroup *memcg, bool enable) struct css_task_iter it; int ret = 0; - if (enable == READ_ONCE(memcg->ksm_merge_any)) - return 0; - css_task_iter_start(&memcg->css, CSS_TASK_ITER_PROCS, &it); while (!ret && (task = css_task_iter_next(&it))) { if (__task_is_dying(task)) @@ -5781,7 +5778,6 @@ static int memory_ksm_show(struct seq_file *m, void *v) } css_task_iter_end(&it); - seq_printf(m, "merge any state: %d\n", READ_ONCE(memcg->ksm_merge_any)); seq_printf(m, "merge any tasks: %u\n", tasks); seq_printf(m, "ksm_rmap_items %lu\n", ksm_rmap_items); seq_printf(m, "ksm_merging_pages %lu\n", ksm_merging_pages); @@ -5808,8 +5804,6 @@ static ssize_t memory_ksm_write(struct kernfs_open_file *of, char *buf, if (err) return err; - WRITE_ONCE(memcg->ksm_merge_any, enable); - return nbytes; } #endif /* CONFIG_KSM */ @@ -6374,9 +6368,6 @@ mem_cgroup_css_alloc(struct cgroup_subsys_state *parent_css) if (parent != root_mem_cgroup) memory_cgrp_subsys.broken_hierarchy = true; } -#ifdef CONFIG_KSM - memcg->ksm_merge_any = false; -#endif /* The following stuff does not apply to the root */ if (!parent) { root_mem_cgroup = memcg; -- 2.25.1

2 1

[PATCH openEuler-1.0-LTS v3 0/4] Mainline bugfix patches backport 4.19
by Guo Mengqi 07 Aug '23

07 Aug '23

Huang Shijie (1): lib/genalloc.c: change return type to unsigned long for bitmap_set_ll Saravana Kannan (1): driver core: Update device link status properly for device_bind_driver() Suravee Suthikulpanit (2): iommu/amd: Use cmpxchg_double() when updating 128-bit IRTE iommu/amd: Restore IRTE.RemapEn bit after programming IRTE drivers/base/base.h | 1 + drivers/base/core.c | 35 ++++++++++++++++++++++++++++++++++ drivers/base/dd.c | 4 +++- drivers/iommu/Kconfig | 2 +- drivers/iommu/amd_iommu.c | 19 ++++++++++++++---- drivers/iommu/amd_iommu_init.c | 18 +++++++++++++++-- lib/genalloc.c | 3 ++- 7 files changed, 73 insertions(+), 9 deletions(-) -- 2.17.1

2 5

[PATCH OLK-5.10 v3 0/4] Mainline bugfix patches backport 5.10
by Guo Mengqi 07 Aug '23

07 Aug '23

Andy Shevchenko (1): lib/cmdline: fix get_option() for strings starting with hyphen Huang Shijie (1): lib/genalloc.c: change return type to unsigned long for bitmap_set_ll Saravana Kannan (1): driver core: Update device link status properly for device_bind_driver() Sumera Priyadarsini (1): bus: arm-integrator-lm: Add of_node_put() before return statement drivers/base/base.h | 1 + drivers/base/core.c | 35 +++++++++++++++++++++++++++++++++ drivers/base/dd.c | 4 +++- drivers/bus/arm-integrator-lm.c | 1 + lib/cmdline.c | 8 +++++++- lib/genalloc.c | 3 ++- 6 files changed, 49 insertions(+), 3 deletions(-) -- 2.17.1

1 4

[PATCH OLK-5.10 0/4] Mainilne bugfix patches backport
by Guo Mengqi 07 Aug '23

07 Aug '23

Andy Shevchenko (1): lib/cmdline: fix get_option() for strings starting with hyphen Huang Shijie (1): lib/genalloc.c: change return type to unsigned long for bitmap_set_ll Saravana Kannan (1): driver core: Update device link status properly for device_bind_driver() Sumera Priyadarsini (1): bus: arm-integrator-lm: Add of_node_put() before return statement drivers/base/base.h | 1 + drivers/base/core.c | 35 +++++++++++++++++++++++++++++++++ drivers/base/dd.c | 4 +++- drivers/bus/arm-integrator-lm.c | 1 + lib/cmdline.c | 8 +++++++- lib/genalloc.c | 3 ++- 6 files changed, 49 insertions(+), 3 deletions(-) -- 2.17.1

2 5

[PATCH openEuler-1.0-LTS 0/4] Mainline bugfix patches backport
by Guo Mengqi 07 Aug '23

07 Aug '23

Huang Shijie (1): lib/genalloc.c: change return type to unsigned long for bitmap_set_ll Saravana Kannan (1): driver core: Update device link status properly for device_bind_driver() Suravee Suthikulpanit (2): iommu/amd: Use cmpxchg_double() when updating 128-bit IRTE iommu/amd: Restore IRTE.RemapEn bit after programming IRTE drivers/base/base.h | 1 + drivers/base/core.c | 35 ++++++++++++++++++++++++++++++++++ drivers/base/dd.c | 4 +++- drivers/iommu/Kconfig | 2 +- drivers/iommu/amd_iommu.c | 19 ++++++++++++++---- drivers/iommu/amd_iommu_init.c | 18 +++++++++++++++-- lib/genalloc.c | 3 ++- 7 files changed, 73 insertions(+), 9 deletions(-) -- 2.17.1

2 5

[openEuler-1.0-LTS 1/5] workqueue: Rename "delayed" (delayed by active management) to "inactive"
by Zeng Heng 07 Aug '23

07 Aug '23

From: Lai Jiangshan <laijs(a)linux.alibaba.com> mainline inclusion from mainline-v5.15-rc1 commit f97a4a1a3f8769e3452885967955e21c88f3f263 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7LRJF DTS: DTS2023072811781 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- There are two kinds of "delayed" work items in workqueue subsystem. One is for timer-delayed work items which are visible to workqueue users. The other kind is for work items delayed by active management which can not be directly visible to workqueue users. We mixed the word "delayed" for both kinds and caused somewhat ambiguity. This patch renames the later one (delayed by active management) to "inactive", because it is used for workqueue active management and most of its related symbols are named with "active" or "activate". All "delayed" and "DELAYED" are carefully checked and renamed one by one to avoid accidentally changing the name of the other kind for timer-delayed. No functional change intended. Signed-off-by: Lai Jiangshan <laijs(a)linux.alibaba.com> Signed-off-by: Tejun Heo <tj(a)kernel.org> conflict: kernel/workqueue.c Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- include/linux/workqueue.h | 4 +-- kernel/workqueue.c | 58 +++++++++++++++++++-------------------- 2 files changed, 31 insertions(+), 31 deletions(-) diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h index 6f2b042fc44c..3ac0fa0c822d 100644 --- a/include/linux/workqueue.h +++ b/include/linux/workqueue.h @@ -30,7 +30,7 @@ void delayed_work_timer_fn(struct timer_list *t); enum { WORK_STRUCT_PENDING_BIT = 0, /* work item is pending execution */ - WORK_STRUCT_DELAYED_BIT = 1, /* work item is delayed */ + WORK_STRUCT_INACTIVE_BIT= 1, /* work item is inactive */ WORK_STRUCT_PWQ_BIT = 2, /* data points to pwq */ WORK_STRUCT_LINKED_BIT = 3, /* next work is linked to this one */ #ifdef CONFIG_DEBUG_OBJECTS_WORK @@ -43,7 +43,7 @@ enum { WORK_STRUCT_COLOR_BITS = 4, WORK_STRUCT_PENDING = 1 << WORK_STRUCT_PENDING_BIT, - WORK_STRUCT_DELAYED = 1 << WORK_STRUCT_DELAYED_BIT, + WORK_STRUCT_INACTIVE = 1 << WORK_STRUCT_INACTIVE_BIT, WORK_STRUCT_PWQ = 1 << WORK_STRUCT_PWQ_BIT, WORK_STRUCT_LINKED = 1 << WORK_STRUCT_LINKED_BIT, #ifdef CONFIG_DEBUG_OBJECTS_WORK diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 04b558a267ae..fe3c4a292909 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -207,7 +207,7 @@ struct pool_workqueue { /* L: nr of in_flight works */ int nr_active; /* L: nr of active works */ int max_active; /* L: max active works */ - struct list_head delayed_works; /* L: delayed works */ + struct list_head inactive_works; /* L: inactive works */ struct list_head pwqs_node; /* WR: node on wq->pwqs */ struct list_head mayday_node; /* MD: node on wq->maydays */ @@ -1105,7 +1105,7 @@ static void put_pwq_unlocked(struct pool_workqueue *pwq) } } -static void pwq_activate_delayed_work(struct work_struct *work) +static void pwq_activate_inactive_work(struct work_struct *work) { struct pool_workqueue *pwq = get_work_pwq(work); @@ -1113,16 +1113,16 @@ static void pwq_activate_delayed_work(struct work_struct *work) if (list_empty(&pwq->pool->worklist)) pwq->pool->watchdog_ts = jiffies; move_linked_works(work, &pwq->pool->worklist, NULL); - __clear_bit(WORK_STRUCT_DELAYED_BIT, work_data_bits(work)); + __clear_bit(WORK_STRUCT_INACTIVE_BIT, work_data_bits(work)); pwq->nr_active++; } -static void pwq_activate_first_delayed(struct pool_workqueue *pwq) +static void pwq_activate_first_inactive(struct pool_workqueue *pwq) { - struct work_struct *work = list_first_entry(&pwq->delayed_works, + struct work_struct *work = list_first_entry(&pwq->inactive_works, struct work_struct, entry); - pwq_activate_delayed_work(work); + pwq_activate_inactive_work(work); } /** @@ -1145,10 +1145,10 @@ static void pwq_dec_nr_in_flight(struct pool_workqueue *pwq, int color) pwq->nr_in_flight[color]--; pwq->nr_active--; - if (!list_empty(&pwq->delayed_works)) { - /* one down, submit a delayed one */ + if (!list_empty(&pwq->inactive_works)) { + /* one down, submit an inactive one */ if (pwq->nr_active < pwq->max_active) - pwq_activate_first_delayed(pwq); + pwq_activate_first_inactive(pwq); } /* is flush in progress and are we at the flushing tip? */ @@ -1246,14 +1246,14 @@ static int try_to_grab_pending(struct work_struct *work, bool is_dwork, debug_work_deactivate(work); /* - * A delayed work item cannot be grabbed directly because + * An inactive work item cannot be grabbed directly because * it might have linked NO_COLOR work items which, if left - * on the delayed_list, will confuse pwq->nr_active + * on the inactive_works list, will confuse pwq->nr_active * management later on and cause stall. Make sure the work * item is activated before grabbing. */ - if (*work_data_bits(work) & WORK_STRUCT_DELAYED) - pwq_activate_delayed_work(work); + if (*work_data_bits(work) & WORK_STRUCT_INACTIVE) + pwq_activate_inactive_work(work); list_del_init(&work->entry); pwq_dec_nr_in_flight(pwq, get_work_color(work)); @@ -1451,8 +1451,8 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, if (list_empty(worklist)) pwq->pool->watchdog_ts = jiffies; } else { - work_flags |= WORK_STRUCT_DELAYED; - worklist = &pwq->delayed_works; + work_flags |= WORK_STRUCT_INACTIVE; + worklist = &pwq->inactive_works; } debug_work_activate(work); @@ -2496,7 +2496,7 @@ static int rescuer_thread(void *__rescuer) /* * The above execution of rescued work items could * have created more to rescue through - * pwq_activate_first_delayed() or chained + * pwq_activate_first_inactive() or chained * queueing. Let's put @pwq back on mayday list so * that such back-to-back work items, which may be * being used to relieve memory pressure, don't @@ -2922,7 +2922,7 @@ void drain_workqueue(struct workqueue_struct *wq) bool drained; spin_lock_irq(&pwq->pool->lock); - drained = !pwq->nr_active && list_empty(&pwq->delayed_works); + drained = !pwq->nr_active && list_empty(&pwq->inactive_works); spin_unlock_irq(&pwq->pool->lock); if (drained) @@ -3703,7 +3703,7 @@ static void pwq_unbound_release_workfn(struct work_struct *work) * @pwq: target pool_workqueue * * If @pwq isn't freezing, set @pwq->max_active to the associated - * workqueue's saved_max_active and activate delayed work items + * workqueue's saved_max_active and activate inactive work items * accordingly. If @pwq is freezing, clear @pwq->max_active to zero. */ static void pwq_adjust_max_active(struct pool_workqueue *pwq) @@ -3732,9 +3732,9 @@ static void pwq_adjust_max_active(struct pool_workqueue *pwq) pwq->max_active = wq->saved_max_active; - while (!list_empty(&pwq->delayed_works) && + while (!list_empty(&pwq->inactive_works) && pwq->nr_active < pwq->max_active) { - pwq_activate_first_delayed(pwq); + pwq_activate_first_inactive(pwq); kick = true; } @@ -3765,7 +3765,7 @@ static void init_pwq(struct pool_workqueue *pwq, struct workqueue_struct *wq, pwq->wq = wq; pwq->flush_color = -1; pwq->refcnt = 1; - INIT_LIST_HEAD(&pwq->delayed_works); + INIT_LIST_HEAD(&pwq->inactive_works); INIT_LIST_HEAD(&pwq->pwqs_node); INIT_LIST_HEAD(&pwq->mayday_node); INIT_WORK(&pwq->unbound_release_work, pwq_unbound_release_workfn); @@ -4386,7 +4386,7 @@ void destroy_workqueue(struct workqueue_struct *wq) if (WARN_ON((pwq != wq->dfl_pwq) && (pwq->refcnt > 1)) || WARN_ON(pwq->nr_active) || - WARN_ON(!list_empty(&pwq->delayed_works))) { + WARN_ON(!list_empty(&pwq->inactive_works))) { mutex_unlock(&wq->mutex); show_workqueue_state(); return; @@ -4527,7 +4527,7 @@ bool workqueue_congested(int cpu, struct workqueue_struct *wq) else pwq = unbound_pwq_by_node(wq, cpu_to_node(cpu)); - ret = !list_empty(&pwq->delayed_works); + ret = !list_empty(&pwq->inactive_works); rcu_read_unlock_sched(); return ret; @@ -4722,11 +4722,11 @@ static void show_pwq(struct pool_workqueue *pwq) pr_cont("\n"); } - if (!list_empty(&pwq->delayed_works)) { + if (!list_empty(&pwq->inactive_works)) { bool comma = false; - pr_info(" delayed:"); - list_for_each_entry(work, &pwq->delayed_works, entry) { + pr_info(" inactive:"); + list_for_each_entry(work, &pwq->inactive_works, entry) { pr_cont_work(comma, work); comma = !(*work_data_bits(work) & WORK_STRUCT_LINKED); } @@ -4756,7 +4756,7 @@ void show_workqueue_state(void) bool idle = true; for_each_pwq(pwq, wq) { - if (pwq->nr_active || !list_empty(&pwq->delayed_works)) { + if (pwq->nr_active || !list_empty(&pwq->inactive_works)) { idle = false; break; } @@ -4768,7 +4768,7 @@ void show_workqueue_state(void) for_each_pwq(pwq, wq) { spin_lock_irqsave(&pwq->pool->lock, flags); - if (pwq->nr_active || !list_empty(&pwq->delayed_works)) + if (pwq->nr_active || !list_empty(&pwq->inactive_works)) show_pwq(pwq); spin_unlock_irqrestore(&pwq->pool->lock, flags); /* @@ -5143,7 +5143,7 @@ EXPORT_SYMBOL_GPL(work_on_cpu_safe); * freeze_workqueues_begin - begin freezing workqueues * * Start freezing workqueues. After this function returns, all freezable - * workqueues will queue new works to their delayed_works list instead of + * workqueues will queue new works to their inactive_works list instead of * pool->worklist. * * CONTEXT: -- 2.25.1

1 4

[PATCH openEuler-1.0-LTS 0/4] Mainline bugfix patches backport
by Guo Mengqi 07 Aug '23

07 Aug '23

Backported 4 bugfix patches from mainline. Huang Shijie (1): lib/genalloc.c: change return type to unsigned long for bitmap_set_ll Saravana Kannan (1): driver core: Update device link status properly for device_bind_driver() Suravee Suthikulpanit (2): iommu/amd: Use cmpxchg_double() when updating 128-bit IRTE iommu/amd: Restore IRTE.RemapEn bit after programming IRTE drivers/base/base.h | 1 + drivers/base/core.c | 35 ++++++++++++++++++++++++++++++++++ drivers/base/dd.c | 4 +++- drivers/iommu/Kconfig | 2 +- drivers/iommu/amd_iommu.c | 19 ++++++++++++++---- drivers/iommu/amd_iommu_init.c | 18 +++++++++++++++-- lib/genalloc.c | 3 ++- 7 files changed, 73 insertions(+), 9 deletions(-) -- 2.17.1

2 5

[PATCH OLK-5.10] netfilter: nft_set_pipapo: fix improper element removal
by Lu Wei 07 Aug '23

07 Aug '23

From: Florian Westphal <fw(a)strlen.de> stable inclusion from stable-v5.10.188 commit 3a91099ecd59a42d1632fcb152bf7222f268ea2b category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I7P3TK CVE: CVE-2023-4004 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… --------------------------- [ Upstream commit 87b5a5c209405cb6b57424cdfa226a6dbd349232 ] end key should be equal to start unless NFT_SET_EXT_KEY_END is present. Its possible to add elements that only have a start key ("{ 1.0.0.0 . 2.0.0.0 }") without an internval end. Insertion treats this via: if (nft_set_ext_exists(ext, NFT_SET_EXT_KEY_END)) end = (const u8 *)nft_set_ext_key_end(ext)->data; else end = start; but removal side always uses nft_set_ext_key_end(). This is wrong and leads to garbage remaining in the set after removal next lookup/insert attempt will give: BUG: KASAN: slab-use-after-free in pipapo_get+0x8eb/0xb90 Read of size 1 at addr ffff888100d50586 by task nft-pipapo_uaf_/1399 Call Trace: kasan_report+0x105/0x140 pipapo_get+0x8eb/0xb90 nft_pipapo_insert+0x1dc/0x1710 nf_tables_newsetelem+0x31f5/0x4e00 .. Fixes: 3c4287f62044 ("nf_tables: Add set type for arbitrary concatenation of ranges") Reported-by: lonial con <kongln9170(a)gmail.com> Reviewed-by: Stefano Brivio <sbrivio(a)redhat.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Lu Wei <luwei32(a)huawei.com> --- net/netfilter/nft_set_pipapo.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nft_set_pipapo.c b/net/netfilter/nft_set_pipapo.c index 30cf0673d6c1..4781cdb63431 100644 --- a/net/netfilter/nft_set_pipapo.c +++ b/net/netfilter/nft_set_pipapo.c @@ -1904,7 +1904,11 @@ static void nft_pipapo_remove(const struct net *net, const struct nft_set *set, int i, start, rules_fx; match_start = data; - match_end = (const u8 *)nft_set_ext_key_end(&e->ext)->data; + + if (nft_set_ext_exists(&e->ext, NFT_SET_EXT_KEY_END)) + match_end = (const u8 *)nft_set_ext_key_end(&e->ext)->data; + else + match_end = data; start = first_rule; rules_fx = rules_f0; -- 2.34.1

2 1

[PATCH OLK-5.10] nvme-pci: fix page size checks
by Yong Hu 07 Aug '23

07 Aug '23

From: Keith Busch <kbusch(a)kernel.org> stable inclusion from stable-v5.10.188 commit cc512539c44f173c2a5fc967e4feb0043bc5773d category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7PZZC Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit 841734234a28fd5cd0889b84bd4d93a0988fa11e ] The size allocated out of the dma pool is at most NVME_CTRL_PAGE_SIZE, which may be smaller than the PAGE_SIZE. Fixes: c61b82c7b7134 ("nvme-pci: fix PRP pool size") Signed-off-by: Keith Busch <kbusch(a)kernel.org> Signed-off-by: Christoph Hellwig <hch(a)lst.de> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Yong Hu <yong.hu(a)windriver.com> --- drivers/nvme/host/pci.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c index ef2fcd0ed914..f1ac50c7a1d6 100644 --- a/drivers/nvme/host/pci.c +++ b/drivers/nvme/host/pci.c @@ -33,7 +33,7 @@ #define SQ_SIZE(q) ((q)->q_depth << (q)->sqes) #define CQ_SIZE(q) ((q)->q_depth * sizeof(struct nvme_completion)) -#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct nvme_sgl_desc)) +#define SGES_PER_PAGE (NVME_CTRL_PAGE_SIZE / sizeof(struct nvme_sgl_desc)) /* * These can be higher, but we need to ensure that any command doesn't @@ -374,7 +374,7 @@ static int nvme_pci_npages_prp(void) { unsigned max_bytes = (NVME_MAX_KB_SZ * 1024) + NVME_CTRL_PAGE_SIZE; unsigned nprps = DIV_ROUND_UP(max_bytes, NVME_CTRL_PAGE_SIZE); - return DIV_ROUND_UP(8 * nprps, PAGE_SIZE - 8); + return DIV_ROUND_UP(8 * nprps, NVME_CTRL_PAGE_SIZE - 8); } /* @@ -384,7 +384,7 @@ static int nvme_pci_npages_prp(void) static int nvme_pci_npages_sgl(void) { return DIV_ROUND_UP(NVME_MAX_SEGS * sizeof(struct nvme_sgl_desc), - PAGE_SIZE); + NVME_CTRL_PAGE_SIZE); } static size_t nvme_pci_iod_alloc_size(void) @@ -735,7 +735,7 @@ static void nvme_pci_sgl_set_seg(struct nvme_sgl_desc *sge, sge->length = cpu_to_le32(entries * sizeof(*sge)); sge->type = NVME_SGL_FMT_LAST_SEG_DESC << 4; } else { - sge->length = cpu_to_le32(PAGE_SIZE); + sge->length = cpu_to_le32(NVME_CTRL_PAGE_SIZE); sge->type = NVME_SGL_FMT_SEG_DESC << 4; } } -- 2.34.1

2 1

[PATCH OLK-5.10] nvme-pci: fix mempool alloc size
by Yong Hu 07 Aug '23

07 Aug '23

From: Keith Busch <kbusch(a)kernel.org> stable inclusion from stable-v5.10.188 commit dfb6d54893d544151e7f480bc44cfe7823f5ad23 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7PZZC Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=… [ Upstream commit c89a529e823d51dd23c7ec0c047c7a454a428541 ] Convert the max size to bytes to match the units of the divisor that calculates the worst-case number of PRP entries. The result is used to determine how many PRP Lists are required. The code was previously rounding this to 1 list, but we can require 2 in the worst case. In that scenario, the driver would corrupt memory beyond the size provided by the mempool. While unlikely to occur (you'd need a 4MB in exactly 127 phys segments on a queue that doesn't support SGLs), this memory corruption has been observed by kfence. Cc: Jens Axboe <axboe(a)kernel.dk> Fixes: 943e942e6266f ("nvme-pci: limit max IO size and segments to avoid high order allocations") Signed-off-by: Keith Busch <kbusch(a)kernel.org> Reviewed-by: Jens Axboe <axboe(a)kernel.dk> Reviewed-by: Kanchan Joshi <joshi.k(a)samsung.com> Reviewed-by: Chaitanya Kulkarni <kch(a)nvidia.com> Signed-off-by: Christoph Hellwig <hch(a)lst.de> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Yong Hu <yong.hu(a)windriver.com> --- drivers/nvme/host/pci.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c index 510ce0c145c7..ef2fcd0ed914 100644 --- a/drivers/nvme/host/pci.c +++ b/drivers/nvme/host/pci.c @@ -372,8 +372,8 @@ static bool nvme_dbbuf_update_and_check_event(u16 value, __le32 *dbbuf_db, */ static int nvme_pci_npages_prp(void) { - unsigned nprps = DIV_ROUND_UP(NVME_MAX_KB_SZ + NVME_CTRL_PAGE_SIZE, - NVME_CTRL_PAGE_SIZE); + unsigned max_bytes = (NVME_MAX_KB_SZ * 1024) + NVME_CTRL_PAGE_SIZE; + unsigned nprps = DIV_ROUND_UP(max_bytes, NVME_CTRL_PAGE_SIZE); return DIV_ROUND_UP(8 * nprps, PAGE_SIZE - 8); } -- 2.34.1

2 1

2024

2023

2022

2021

2020

2019

Kernel