- Kernel - mailweb.openeuler.org

[PATCH openEuler-1.0-LTS] IB/hfi1: Restore allocated resources on failed copyout
by Zeng Heng 13 Jun '24

13 Jun '24

From: Dean Luick <dean.luick(a)cornelisnetworks.com> stable inclusion from stable-v4.19.273 commit 00d9e212b8a39e6ffcf31b9d2e503d2bf6009d45 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9R4LR CVE: CVE-2023-52747 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… ---------------------------------------------------- [ Upstream commit 6601fc0d15ffc20654e39486f9bef35567106d68 ] Fix a resource leak if an error occurs. Fixes: f404ca4c7ea8 ("IB/hfi1: Refactor hfi_user_exp_rcv_setup() IOCTL") Signed-off-by: Dean Luick <dean.luick(a)cornelisnetworks.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro(a)cornelisnetworks.com> Link: https://lore.kernel.org/r/167354736291.2132367.10894218740150168180.stgit@a… Signed-off-by: Leon Romanovsky <leon(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- drivers/infiniband/hw/hfi1/file_ops.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/file_ops.c b/drivers/infiniband/hw/hfi1/file_ops.c index 34ffca618427..499bfae098c2 100644 --- a/drivers/infiniband/hw/hfi1/file_ops.c +++ b/drivers/infiniband/hw/hfi1/file_ops.c @@ -1345,12 +1345,15 @@ static int user_exp_rcv_setup(struct hfi1_filedata *fd, unsigned long arg, addr = arg + offsetof(struct hfi1_tid_info, tidcnt); if (copy_to_user((void __user *)addr, &tinfo.tidcnt, sizeof(tinfo.tidcnt))) - return -EFAULT; + ret = -EFAULT; addr = arg + offsetof(struct hfi1_tid_info, length); - if (copy_to_user((void __user *)addr, &tinfo.length, + if (!ret && copy_to_user((void __user *)addr, &tinfo.length, sizeof(tinfo.length))) ret = -EFAULT; + + if (ret) + hfi1_user_exp_rcv_invalid(fd, &tinfo); } return ret; -- 2.25.1

2 1

[PATCH OLK-5.10] keys: Fix overwrite of key expiration on instantiation
by Xiang Yang 13 Jun '24

13 Jun '24

From: Silvio Gissi <sifonsec(a)amazon.com> stable inclusion from stable-v5.10.217 commit ad2011ea787928b2accb5134f1e423b11fe80a8a category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9TM8D CVE: CVE-2024-36031 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 9da27fb65a14c18efd4473e2e82b76b53ba60252 upstream. The expiry time of a key is unconditionally overwritten during instantiation, defaulting to turn it permanent. This causes a problem for DNS resolution as the expiration set by user-space is overwritten to TIME64_MAX, disabling further DNS updates. Fix this by restoring the condition that key_set_expiry is only called when the pre-parser sets a specific expiry. Fixes: 39299bdd2546 ("keys, dns: Allow key types (eg. DNS) to be reclaimed immediately on expiry") Signed-off-by: Silvio Gissi <sifonsec(a)amazon.com> cc: David Howells <dhowells(a)redhat.com> cc: Hazem Mohamed Abuelfotoh <abuehaze(a)amazon.com> cc: linux-afs(a)lists.infradead.org cc: linux-cifs(a)vger.kernel.org cc: keyrings(a)vger.kernel.org cc: netdev(a)vger.kernel.org cc: stable(a)vger.kernel.org Reviewed-by: Jarkko Sakkinen <jarkko(a)kernel.org> Signed-off-by: Jarkko Sakkinen <jarkko(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Xiang Yang <xiangyang3(a)huawei.com> --- security/keys/key.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/security/keys/key.c b/security/keys/key.c index 67ad0826e385..e5111ce17e25 100644 --- a/security/keys/key.c +++ b/security/keys/key.c @@ -464,7 +464,8 @@ static int __key_instantiate_and_link(struct key *key, if (authkey) key_invalidate(authkey); - key_set_expiry(key, prep->expiry); + if (prep->expiry != TIME64_MAX) + key_set_expiry(key, prep->expiry); } } -- 2.34.1

2 1

[PATCH openEuler-1.0-LTS] IB/hfi1: Restore allocated resources on failed copyout
by Zeng Heng 13 Jun '24

13 Jun '24

From: Dean Luick <dean.luick(a)cornelisnetworks.com> stable inclusion from stable-v4.19.273 commit 00d9e212b8a39e6ffcf31b9d2e503d2bf6009d45 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9R4LR CVE: CVE-2023-52747 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… ---------------------------------------------------- [ Upstream commit 6601fc0d15ffc20654e39486f9bef35567106d68 ] Fix a resource leak if an error occurs. Fixes: f404ca4c7ea8 ("IB/hfi1: Refactor hfi_user_exp_rcv_setup() IOCTL") Signed-off-by: Dean Luick <dean.luick(a)cornelisnetworks.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro(a)cornelisnetworks.com> Link: https://lore.kernel.org/r/167354736291.2132367.10894218740150168180.stgit@a… Signed-off-by: Leon Romanovsky <leon(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- drivers/infiniband/hw/hfi1/file_ops.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/file_ops.c b/drivers/infiniband/hw/hfi1/file_ops.c index 34ffca618427..499bfae098c2 100644 --- a/drivers/infiniband/hw/hfi1/file_ops.c +++ b/drivers/infiniband/hw/hfi1/file_ops.c @@ -1345,12 +1345,15 @@ static int user_exp_rcv_setup(struct hfi1_filedata *fd, unsigned long arg, addr = arg + offsetof(struct hfi1_tid_info, tidcnt); if (copy_to_user((void __user *)addr, &tinfo.tidcnt, sizeof(tinfo.tidcnt))) - return -EFAULT; + ret = -EFAULT; addr = arg + offsetof(struct hfi1_tid_info, length); - if (copy_to_user((void __user *)addr, &tinfo.length, + if (!ret && copy_to_user((void __user *)addr, &tinfo.length, sizeof(tinfo.length))) ret = -EFAULT; + + if (ret) + hfi1_user_exp_rcv_invalid(fd, &tinfo); } return ret; -- 2.25.1

2 1

[PATCH openEuler-1.0-LTS] IB/hfi1: Restore allocated resources on failed copyout
by Zeng Heng 13 Jun '24

13 Jun '24

From: Dean Luick <dean.luick(a)cornelisnetworks.com> stable inclusion from stable-v5.10.168 commit 79b595d9591426156a9e0635a5b5115508a36fef category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9R4LR CVE: CVE-2023-52747 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… ---------------------------------------------------- [ Upstream commit 6601fc0d15ffc20654e39486f9bef35567106d68 ] Fix a resource leak if an error occurs. Fixes: f404ca4c7ea8 ("IB/hfi1: Refactor hfi_user_exp_rcv_setup() IOCTL") Signed-off-by: Dean Luick <dean.luick(a)cornelisnetworks.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro(a)cornelisnetworks.com> Link: https://lore.kernel.org/r/167354736291.2132367.10894218740150168180.stgit@a… Signed-off-by: Leon Romanovsky <leon(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- drivers/infiniband/hw/hfi1/file_ops.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/file_ops.c b/drivers/infiniband/hw/hfi1/file_ops.c index 34ffca618427..499bfae098c2 100644 --- a/drivers/infiniband/hw/hfi1/file_ops.c +++ b/drivers/infiniband/hw/hfi1/file_ops.c @@ -1345,12 +1345,15 @@ static int user_exp_rcv_setup(struct hfi1_filedata *fd, unsigned long arg, addr = arg + offsetof(struct hfi1_tid_info, tidcnt); if (copy_to_user((void __user *)addr, &tinfo.tidcnt, sizeof(tinfo.tidcnt))) - return -EFAULT; + ret = -EFAULT; addr = arg + offsetof(struct hfi1_tid_info, length); - if (copy_to_user((void __user *)addr, &tinfo.length, + if (!ret && copy_to_user((void __user *)addr, &tinfo.length, sizeof(tinfo.length))) ret = -EFAULT; + + if (ret) + hfi1_user_exp_rcv_invalid(fd, &tinfo); } return ret; -- 2.25.1

2 1

[PATCH OLK-5.10] block: support to dispatch bio asynchronously
by Li Nan 13 Jun '24

13 Jun '24

From: Yu Kuai <yukuai3(a)huawei.com> In certain environments, specific CPUs handle a large number of tasks and become bottlenecks, affecting overall system performance. This commit introduces a new feature that enables asynchronous I/O dispatch to designated CPUs, thereby relieving the pressure on the busy CPUs. Signed-off-by: Yu Kuai <yukuai3(a)huawei.com> Signed-off-by: Li Nan <linan122(a)huawei.com> --- block/blk.h | 7 ++ include/linux/blk_types.h | 1 + include/linux/blkdev.h | 8 ++ block/blk-core.c | 229 +++++++++++++++++++++++++++++++++++++- block/blk-mq-debugfs.c | 1 + block/blk-sysfs.c | 58 ++++++++++ block/Kconfig | 12 ++ 7 files changed, 315 insertions(+), 1 deletion(-) diff --git a/block/blk.h b/block/blk.h index 4bbcc971d4f7..5e7c00356ddc 100644 --- a/block/blk.h +++ b/block/blk.h @@ -450,4 +450,11 @@ int bio_add_hw_page(struct request_queue *q, struct bio *bio, struct page *page, unsigned int len, unsigned int offset, unsigned int max_sectors, bool *same_page); +#ifdef CONFIG_BLK_BIO_DISPATCH_ASYNC +void blk_free_queue_dispatch_async(struct request_queue *q); +#else +static inline void blk_free_queue_dispatch_async(struct request_queue *q) +{ +} +#endif #endif /* BLK_INTERNAL_H */ diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h index 1853ec569b72..865e1198e837 100644 --- a/include/linux/blk_types.h +++ b/include/linux/blk_types.h @@ -301,6 +301,7 @@ enum { * of this bio. */ BIO_CGROUP_ACCT, /* has been accounted to a cgroup */ BIO_TRACKED, /* set if bio goes through the rq_qos path */ + BIO_ASYNC, BIO_FLAG_LAST }; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 50b4fd0a0687..453bae093b65 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -601,6 +601,12 @@ struct request_queue { #define BLK_MAX_WRITE_HINTS 5 u64 write_hints[BLK_MAX_WRITE_HINTS]; + +#ifdef CONFIG_BLK_BIO_DISPATCH_ASYNC + /* used when QUEUE_FLAG_DISPATCH_ASYNC is set */ + struct cpumask dispatch_async_cpus; + int __percpu *last_dispatch_cpu; +#endif KABI_REPLACE(unsigned long dtag_wait_time, struct blk_mq_tags *shared_sbitmap_tags) KABI_RESERVE(1) @@ -643,6 +649,8 @@ struct request_queue { #define QUEUE_FLAG_NOWAIT 29 /* device supports NOWAIT */ /*at least one blk-mq hctx can't get driver tag */ #define QUEUE_FLAG_HCTX_WAIT 30 +/* support to dispatch bio asynchronously */ +#define QUEUE_FLAG_DISPATCH_ASYNC 31 #define QUEUE_FLAG_MQ_DEFAULT ((1 << QUEUE_FLAG_IO_STAT) | \ (1 << QUEUE_FLAG_SAME_COMP) | \ diff --git a/block/blk-core.c b/block/blk-core.c index f91f8e8be482..03375d28d664 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -87,6 +87,223 @@ struct kmem_cache *blk_requestq_cachep; */ static struct workqueue_struct *kblockd_workqueue; +#ifdef CONFIG_BLK_BIO_DISPATCH_ASYNC + +#define BIO_DISPATCH_MAX_LOOP 16 + +struct async_bio { + struct bio_list list; + spinlock_t lock; +} ____cacheline_aligned_in_smp; + +struct bio_dispatch_async_ctl { + /* + * Vector size is nr_cpu_ids, list stores bio dispatched from other cpu, + * such bio will be dispatched asynchronously to the cpu this structure + * is serviced. + */ + struct async_bio *bios; + /* kthread to handle bio dispatched from other cpu. */ + struct task_struct *thread; + wait_queue_head_t wait; +}; + +static struct bio_dispatch_async_ctl __percpu *bio_dispatch_async_ctl; + +static int blk_alloc_queue_dispatch_async(struct request_queue *q) +{ + int cpu; + + q->last_dispatch_cpu = alloc_percpu(int); + if (!q->last_dispatch_cpu) + return -ENOMEM; + + cpumask_setall(&q->dispatch_async_cpus); + for_each_possible_cpu(cpu) + *per_cpu_ptr(q->last_dispatch_cpu, cpu) = cpu; + + return 0; +} + +void blk_free_queue_dispatch_async(struct request_queue *q) +{ + free_percpu(q->last_dispatch_cpu); + q->last_dispatch_cpu = NULL; +} + +static int get_dispatch_cpu(struct request_queue *q, int cpu) +{ + int *last_dispatch_cpu = + per_cpu_ptr(q->last_dispatch_cpu, cpu); + + cpu = cpumask_next(*last_dispatch_cpu, &q->dispatch_async_cpus); + if (cpu >= nr_cpu_ids) + cpu = cpumask_first(&q->dispatch_async_cpus); + + *last_dispatch_cpu = cpu; + + return cpu; +} + +static bool __submit_bio_noacct_async(struct bio *bio) +{ + int cpu = smp_processor_id(); + int dispatch_cpu = get_dispatch_cpu(bio->bi_disk->queue, cpu); + struct bio_dispatch_async_ctl *ctl; + + if (dispatch_cpu >= nr_cpu_ids || dispatch_cpu == cpu) + return false; + + ctl = per_cpu_ptr(bio_dispatch_async_ctl, dispatch_cpu); + spin_lock_irq(&ctl->bios[cpu].lock); + bio_list_add(&ctl->bios[cpu].list, bio); + spin_unlock_irq(&ctl->bios[cpu].lock); + + if (wq_has_sleeper(&ctl->wait)) + wake_up(&ctl->wait); + + return true; +} + +static bool submit_bio_noacct_async(struct bio *bio) +{ + struct request_queue *q; + + if (bio_flagged(bio, BIO_ASYNC)) + return false; + + bio_set_flag(bio, BIO_ASYNC); + /* + * Don't dispatch bio asynchronously in following cases: + * + * - QUEUE_FLAG_DISPATCH_ASYNC is not set; + * - current cpu is the target cpu; + * - bio is flagged no wait; + * - io polling is enabled; + */ + q = bio->bi_disk->queue; + if (!test_bit(QUEUE_FLAG_DISPATCH_ASYNC, &q->queue_flags) || + test_bit(QUEUE_FLAG_POLL, &q->queue_flags) || + cpumask_test_cpu(smp_processor_id(), &q->dispatch_async_cpus) || + bio->bi_opf & REQ_NOWAIT) + return false; + + return __submit_bio_noacct_async(bio); +} + +static bool collect_bio(struct bio_dispatch_async_ctl *ctl, + struct bio_list *list) +{ + bool has_bio = false; + int cpu; + + for_each_possible_cpu(cpu) { + struct async_bio *abio = &ctl->bios[cpu]; + + if (bio_list_empty(&abio->list)) + continue; + + has_bio = true; + + spin_lock_irq(&abio->lock); + bio_list_merge(list, &abio->list); + bio_list_init(&abio->list); + spin_unlock_irq(&abio->lock); + } + + return has_bio; +} + +static int bio_dispatch_work(void *data) +{ + int loop_count = 0; + struct bio_list bio_list_on_stack; + struct blk_plug plug; + struct bio_dispatch_async_ctl *ctl; + + bio_list_init(&bio_list_on_stack); + ctl = this_cpu_ptr(bio_dispatch_async_ctl); + + for (;; loop_count++) { + struct bio *bio; + bool has_bio = collect_bio(ctl, &bio_list_on_stack); + + if (!has_bio) { + DEFINE_WAIT(wait); + + for (;;) { + prepare_to_wait(&ctl->wait, &wait, + TASK_INTERRUPTIBLE); + has_bio = collect_bio(ctl, &bio_list_on_stack); + if (has_bio) + break; + schedule(); + loop_count = 0; + } + finish_wait(&ctl->wait, &wait); + } + + blk_start_plug(&plug); + while ((bio = bio_list_pop(&bio_list_on_stack))) + submit_bio_noacct(bio); + blk_finish_plug(&plug); + + /* prevent soft lockup. */ + if (loop_count >= BIO_DISPATCH_MAX_LOOP) { + loop_count = 0; + cond_resched(); + } + } + + return 0; +} + +static void init_blk_queue_async_dispatch(void) +{ + int cpu; + + bio_dispatch_async_ctl = alloc_percpu(struct bio_dispatch_async_ctl); + if (!bio_dispatch_async_ctl) + panic("Failed to alloc bio_dispatch_async_ctl\n"); + + for_each_possible_cpu(cpu) { + int i; + struct bio_dispatch_async_ctl *ctl = + per_cpu_ptr(bio_dispatch_async_ctl, cpu); + + init_waitqueue_head(&ctl->wait); + ctl->bios = kmalloc_array(nr_cpu_ids, sizeof(struct async_bio), + GFP_KERNEL | __GFP_NOFAIL); + for (i = 0; i < nr_cpu_ids; ++i) { + bio_list_init(&ctl->bios[i].list); + spin_lock_init(&ctl->bios[i].lock); + } + + ctl->thread = + kthread_create_on_cpu(bio_dispatch_work, NULL, cpu, + "bio_dispatch_work_%u"); + if (IS_ERR_OR_NULL(ctl->thread)) + panic("Failed to create bio dispatch thread\n"); + + wake_up_process(ctl->thread); + } +} +#else +static int blk_alloc_queue_dispatch_async(struct request_queue *q) +{ + return 0; +} + +static bool submit_bio_noacct_async(struct bio *bio) +{ + return false; +} + +static void init_blk_queue_async_dispatch(void) +{ +} +#endif + /** * blk_queue_flag_set - atomically set a queue flag * @flag: flag to be set @@ -539,9 +756,12 @@ struct request_queue *blk_alloc_queue(int node_id) q->last_merge = NULL; + if (blk_alloc_queue_dispatch_async(q)) + goto fail_q; + q->id = ida_simple_get(&blk_queue_ida, 0, 0, GFP_KERNEL); if (q->id < 0) - goto fail_q; + goto fail_dispatch_async; ret = bioset_init(&q->bio_split, BIO_POOL_SIZE, 0, BIOSET_NEED_BVECS); if (ret) @@ -606,6 +826,8 @@ struct request_queue *blk_alloc_queue(int node_id) bioset_exit(&q->bio_split); fail_id: ida_simple_remove(&blk_queue_ida, q->id); +fail_dispatch_async: + blk_free_queue_dispatch_async(q); fail_q: kmem_cache_free(blk_requestq_cachep, q); return NULL; @@ -1055,6 +1277,9 @@ static blk_qc_t __submit_bio_noacct_mq(struct bio *bio) */ blk_qc_t submit_bio_noacct(struct bio *bio) { + if (submit_bio_noacct_async(bio)) + return BLK_QC_T_NONE; + if (!submit_bio_checks(bio)) return BLK_QC_T_NONE; @@ -1905,5 +2130,7 @@ int __init blk_dev_init(void) blk_debugfs_root = debugfs_create_dir("block", NULL); + init_blk_queue_async_dispatch(); + return 0; } diff --git a/block/blk-mq-debugfs.c b/block/blk-mq-debugfs.c index a879f94782e4..b5b17c6ee650 100644 --- a/block/blk-mq-debugfs.c +++ b/block/blk-mq-debugfs.c @@ -131,6 +131,7 @@ static const char *const blk_queue_flag_name[] = { QUEUE_FLAG_NAME(RQ_ALLOC_TIME), QUEUE_FLAG_NAME(HCTX_ACTIVE), QUEUE_FLAG_NAME(NOWAIT), + QUEUE_FLAG_NAME(DISPATCH_ASYNC), }; #undef QUEUE_FLAG_NAME diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index c95be9626a09..e363222858b4 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -288,6 +288,9 @@ QUEUE_SYSFS_BIT_FNS(nonrot, NONROT, 1); QUEUE_SYSFS_BIT_FNS(random, ADD_RANDOM, 0); QUEUE_SYSFS_BIT_FNS(iostats, IO_STAT, 0); QUEUE_SYSFS_BIT_FNS(stable_writes, STABLE_WRITES, 0); +#ifdef CONFIG_BLK_BIO_DISPATCH_ASYNC +QUEUE_SYSFS_BIT_FNS(dispatch_async, DISPATCH_ASYNC, 0); +#endif #undef QUEUE_SYSFS_BIT_FNS static ssize_t queue_zoned_show(struct request_queue *q, char *page) @@ -619,6 +622,56 @@ QUEUE_RW_ENTRY(queue_iostats, "iostats"); QUEUE_RW_ENTRY(queue_random, "add_random"); QUEUE_RW_ENTRY(queue_stable_writes, "stable_writes"); +#ifdef CONFIG_BLK_BIO_DISPATCH_ASYNC + +static ssize_t queue_dispatch_async_cpus_show(struct request_queue *q, + char *page) +{ + + return sprintf(page, "%*pb\n", nr_cpu_ids, + cpumask_bits(&q->dispatch_async_cpus)); +} + +static ssize_t queue_dispatch_async_cpus_store(struct request_queue *q, + const char *page, size_t count) +{ + cpumask_var_t cpumask; + ssize_t ret; + + if (!alloc_cpumask_var(&cpumask, GFP_KERNEL)) + return -ENOMEM; + + ret = bitmap_parse(page, count, cpumask_bits(cpumask), + nr_cpumask_bits); + if (ret < 0) + goto out; + + if (cpumask_empty(cpumask) || + !cpumask_subset(cpumask, cpu_online_mask)) + return -EINVAL; + + blk_mq_freeze_queue(q); + blk_mq_quiesce_queue(q); + + cpumask_copy(&q->dispatch_async_cpus, cpumask); + + blk_mq_unquiesce_queue(q); + blk_mq_unfreeze_queue(q); + ret = count; +out: + free_cpumask_var(cpumask); + return ret; +} + +static struct queue_sysfs_entry queue_dispatch_async_cpus_entry = { + .attr = {.name = "dispatch_async_cpus", .mode = 0644 }, + .show = queue_dispatch_async_cpus_show, + .store = queue_dispatch_async_cpus_store, +}; + +QUEUE_RW_ENTRY(queue_dispatch_async, "dispatch_async"); +#endif + static struct attribute *queue_attrs[] = { &queue_requests_entry.attr, &queue_ra_entry.attr, @@ -659,6 +712,10 @@ static struct attribute *queue_attrs[] = { &queue_wb_lat_entry.attr, &queue_poll_delay_entry.attr, &queue_io_timeout_entry.attr, +#ifdef CONFIG_BLK_BIO_DISPATCH_ASYNC + &queue_dispatch_async_cpus_entry.attr, + &queue_dispatch_async_entry.attr, +#endif #ifdef CONFIG_BLK_DEV_THROTTLING_LOW &blk_throtl_sample_time_entry.attr, #endif @@ -795,6 +852,7 @@ static void blk_release_queue(struct kobject *kobj) blk_stat_remove_callback(q, q->poll_cb); blk_stat_free_callback(q->poll_cb); + blk_free_queue_dispatch_async(q); blk_free_queue_stats(q->stats); blk_exit_queue(q); diff --git a/block/Kconfig b/block/Kconfig index e5c965f1ea25..bda77b006352 100644 --- a/block/Kconfig +++ b/block/Kconfig @@ -209,6 +209,18 @@ config BLK_DEV_DUMPINFO Dump info when open an write opened block device exclusively or open an exclusive opened device for write +config BLK_BIO_DISPATCH_ASYNC + bool "Dispatch bios asynchronously on specific cpus" + default n + help + If there are multiple nodes, memory access across nodes is rather bad + compare to local node. And if some drivers are using internal spin + locks, io performance will be bad if bios are issued concurrently from + different nodes. This feature will dispatch bio asynchronously to the + specific CPUs to avoid across nodes memory access in driver, noted this + feature will require special care in the driver to work. If unsure, + say N here. + menu "Partition Types" source "block/partitions/Kconfig" -- 2.39.2

2 1

virtcca patch
by Lishengjie 12 Jun '24

12 Jun '24

1 0

[PATCH openEuler-22.03-LTS-SP1] bpf, skmsg: Fix NULL pointer dereference in sk_psock_skb_ingress_enqueue
by Liu Jian 12 Jun '24

12 Jun '24

From: Jason Xing <kernelxing(a)tencent.com> mainline inclusion from mainline-v6.9-rc7 commit 6648e613226e18897231ab5e42ffc29e63fa3365 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9U1UZ CVE: CVE-2024-36938 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- Fix NULL pointer data-races in sk_psock_skb_ingress_enqueue() which syzbot reported [1]. [1] BUG: KCSAN: data-race in sk_psock_drop / sk_psock_skb_ingress_enqueue write to 0xffff88814b3278b8 of 8 bytes by task 10724 on cpu 1: sk_psock_stop_verdict net/core/skmsg.c:1257 [inline] sk_psock_drop+0x13e/0x1f0 net/core/skmsg.c:843 sk_psock_put include/linux/skmsg.h:459 [inline] sock_map_close+0x1a7/0x260 net/core/sock_map.c:1648 unix_release+0x4b/0x80 net/unix/af_unix.c:1048 __sock_release net/socket.c:659 [inline] sock_close+0x68/0x150 net/socket.c:1421 __fput+0x2c1/0x660 fs/file_table.c:422 __fput_sync+0x44/0x60 fs/file_table.c:507 __do_sys_close fs/open.c:1556 [inline] __se_sys_close+0x101/0x1b0 fs/open.c:1541 __x64_sys_close+0x1f/0x30 fs/open.c:1541 do_syscall_64+0xd3/0x1d0 entry_SYSCALL_64_after_hwframe+0x6d/0x75 read to 0xffff88814b3278b8 of 8 bytes by task 10713 on cpu 0: sk_psock_data_ready include/linux/skmsg.h:464 [inline] sk_psock_skb_ingress_enqueue+0x32d/0x390 net/core/skmsg.c:555 sk_psock_skb_ingress_self+0x185/0x1e0 net/core/skmsg.c:606 sk_psock_verdict_apply net/core/skmsg.c:1008 [inline] sk_psock_verdict_recv+0x3e4/0x4a0 net/core/skmsg.c:1202 unix_read_skb net/unix/af_unix.c:2546 [inline] unix_stream_read_skb+0x9e/0xf0 net/unix/af_unix.c:2682 sk_psock_verdict_data_ready+0x77/0x220 net/core/skmsg.c:1223 unix_stream_sendmsg+0x527/0x860 net/unix/af_unix.c:2339 sock_sendmsg_nosec net/socket.c:730 [inline] __sock_sendmsg+0x140/0x180 net/socket.c:745 ____sys_sendmsg+0x312/0x410 net/socket.c:2584 ___sys_sendmsg net/socket.c:2638 [inline] __sys_sendmsg+0x1e9/0x280 net/socket.c:2667 __do_sys_sendmsg net/socket.c:2676 [inline] __se_sys_sendmsg net/socket.c:2674 [inline] __x64_sys_sendmsg+0x46/0x50 net/socket.c:2674 do_syscall_64+0xd3/0x1d0 entry_SYSCALL_64_after_hwframe+0x6d/0x75 value changed: 0xffffffff83d7feb0 -> 0x0000000000000000 Reported by Kernel Concurrency Sanitizer on: CPU: 0 PID: 10713 Comm: syz-executor.4 Tainted: G W 6.8.0-syzkaller-08951-gfe46a7dd189e #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 02/29/2024 Prior to this, commit 4cd12c6065df ("bpf, sockmap: Fix NULL pointer dereference in sk_psock_verdict_data_ready()") fixed one NULL pointer similarly due to no protection of saved_data_ready. Here is another different caller causing the same issue because of the same reason. So we should protect it with sk_callback_lock read lock because the writer side in the sk_psock_drop() uses "write_lock_bh(&sk->sk_callback_lock);". To avoid errors that could happen in future, I move those two pairs of lock into the sk_psock_data_ready(), which is suggested by John Fastabend. Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface") Reported-by: syzbot+aa8c8ec2538929f18f2d(a)syzkaller.appspotmail.com Signed-off-by: Jason Xing <kernelxing(a)tencent.com> Signed-off-by: Daniel Borkmann <daniel(a)iogearbox.net> Reviewed-by: John Fastabend <john.fastabend(a)gmail.com> Closes: https://syzkaller.appspot.com/bug?extid=aa8c8ec2538929f18f2d Link: https://lore.kernel.org/all/20240329134037.92124-1-kerneljasonxing@gmail.com Link: https://lore.kernel.org/bpf/20240404021001.94815-1-kerneljasonxing@gmail.com Conflicts: include/linux/skmsg.h net/core/skmsg.c [This is because we didnot backport 5a685cd94b21 and 6df7f764cd3c.] Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- include/linux/skmsg.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/include/linux/skmsg.h b/include/linux/skmsg.h index a83885c5bb86..9bd49d9539e9 100644 --- a/include/linux/skmsg.h +++ b/include/linux/skmsg.h @@ -476,10 +476,12 @@ static inline void sk_psock_put(struct sock *sk, struct sk_psock *psock) static inline void sk_psock_data_ready(struct sock *sk, struct sk_psock *psock) { + read_lock_bh(&sk->sk_callback_lock); if (psock->parser.enabled) psock->parser.saved_data_ready(sk); else sk->sk_data_ready(sk); + read_unlock_bh(&sk->sk_callback_lock); } static inline void psock_set_prog(struct bpf_prog **pprog, -- 2.34.1

2 1

[PATCH OLK-5.10] bpf, skmsg: Fix NULL pointer dereference in sk_psock_skb_ingress_enqueue
by Liu Jian 12 Jun '24

12 Jun '24

From: Jason Xing <kernelxing(a)tencent.com> mainline inclusion from mainline-v6.9-rc7 commit 6648e613226e18897231ab5e42ffc29e63fa3365 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9U1UZ CVE: CVE-2024-36938 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- Fix NULL pointer data-races in sk_psock_skb_ingress_enqueue() which syzbot reported [1]. [1] BUG: KCSAN: data-race in sk_psock_drop / sk_psock_skb_ingress_enqueue write to 0xffff88814b3278b8 of 8 bytes by task 10724 on cpu 1: sk_psock_stop_verdict net/core/skmsg.c:1257 [inline] sk_psock_drop+0x13e/0x1f0 net/core/skmsg.c:843 sk_psock_put include/linux/skmsg.h:459 [inline] sock_map_close+0x1a7/0x260 net/core/sock_map.c:1648 unix_release+0x4b/0x80 net/unix/af_unix.c:1048 __sock_release net/socket.c:659 [inline] sock_close+0x68/0x150 net/socket.c:1421 __fput+0x2c1/0x660 fs/file_table.c:422 __fput_sync+0x44/0x60 fs/file_table.c:507 __do_sys_close fs/open.c:1556 [inline] __se_sys_close+0x101/0x1b0 fs/open.c:1541 __x64_sys_close+0x1f/0x30 fs/open.c:1541 do_syscall_64+0xd3/0x1d0 entry_SYSCALL_64_after_hwframe+0x6d/0x75 read to 0xffff88814b3278b8 of 8 bytes by task 10713 on cpu 0: sk_psock_data_ready include/linux/skmsg.h:464 [inline] sk_psock_skb_ingress_enqueue+0x32d/0x390 net/core/skmsg.c:555 sk_psock_skb_ingress_self+0x185/0x1e0 net/core/skmsg.c:606 sk_psock_verdict_apply net/core/skmsg.c:1008 [inline] sk_psock_verdict_recv+0x3e4/0x4a0 net/core/skmsg.c:1202 unix_read_skb net/unix/af_unix.c:2546 [inline] unix_stream_read_skb+0x9e/0xf0 net/unix/af_unix.c:2682 sk_psock_verdict_data_ready+0x77/0x220 net/core/skmsg.c:1223 unix_stream_sendmsg+0x527/0x860 net/unix/af_unix.c:2339 sock_sendmsg_nosec net/socket.c:730 [inline] __sock_sendmsg+0x140/0x180 net/socket.c:745 ____sys_sendmsg+0x312/0x410 net/socket.c:2584 ___sys_sendmsg net/socket.c:2638 [inline] __sys_sendmsg+0x1e9/0x280 net/socket.c:2667 __do_sys_sendmsg net/socket.c:2676 [inline] __se_sys_sendmsg net/socket.c:2674 [inline] __x64_sys_sendmsg+0x46/0x50 net/socket.c:2674 do_syscall_64+0xd3/0x1d0 entry_SYSCALL_64_after_hwframe+0x6d/0x75 value changed: 0xffffffff83d7feb0 -> 0x0000000000000000 Reported by Kernel Concurrency Sanitizer on: CPU: 0 PID: 10713 Comm: syz-executor.4 Tainted: G W 6.8.0-syzkaller-08951-gfe46a7dd189e #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 02/29/2024 Prior to this, commit 4cd12c6065df ("bpf, sockmap: Fix NULL pointer dereference in sk_psock_verdict_data_ready()") fixed one NULL pointer similarly due to no protection of saved_data_ready. Here is another different caller causing the same issue because of the same reason. So we should protect it with sk_callback_lock read lock because the writer side in the sk_psock_drop() uses "write_lock_bh(&sk->sk_callback_lock);". To avoid errors that could happen in future, I move those two pairs of lock into the sk_psock_data_ready(), which is suggested by John Fastabend. Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface") Reported-by: syzbot+aa8c8ec2538929f18f2d(a)syzkaller.appspotmail.com Signed-off-by: Jason Xing <kernelxing(a)tencent.com> Signed-off-by: Daniel Borkmann <daniel(a)iogearbox.net> Reviewed-by: John Fastabend <john.fastabend(a)gmail.com> Closes: https://syzkaller.appspot.com/bug?extid=aa8c8ec2538929f18f2d Link: https://lore.kernel.org/all/20240329134037.92124-1-kerneljasonxing@gmail.com Link: https://lore.kernel.org/bpf/20240404021001.94815-1-kerneljasonxing@gmail.com Conflicts: include/linux/skmsg.h net/core/skmsg.c [This is because we didnot backport 5a685cd94b21 and 6df7f764cd3c.] Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- include/linux/skmsg.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/include/linux/skmsg.h b/include/linux/skmsg.h index 9c14c5fbb8c0..f9d8e8b1a4e8 100644 --- a/include/linux/skmsg.h +++ b/include/linux/skmsg.h @@ -477,10 +477,12 @@ static inline void sk_psock_put(struct sock *sk, struct sk_psock *psock) static inline void sk_psock_data_ready(struct sock *sk, struct sk_psock *psock) { + read_lock_bh(&sk->sk_callback_lock); if (psock->parser.enabled) psock->parser.saved_data_ready(sk); else sk->sk_data_ready(sk); + read_unlock_bh(&sk->sk_callback_lock); } static inline void psock_set_prog(struct bpf_prog **pprog, -- 2.34.1

2 1

[PATCH OLK-6.6] tty: n_gsm: fix possible out-of-bounds in gsm0_receive()
by Cui GaoSheng 12 Jun '24

12 Jun '24

From: Daniel Starke <daniel.starke(a)siemens.com> mainline inclusion from mainline-v6.10-rc1 commit 47388e807f85948eefc403a8a5fdc5b406a65d5a category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9TAHR CVE: CVE-2024-36016 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- Assuming the following: - side A configures the n_gsm in basic option mode - side B sends the header of a basic option mode frame with data length 1 - side A switches to advanced option mode - side B sends 2 data bytes which exceeds gsm->len Reason: gsm->len is not used in advanced option mode. - side A switches to basic option mode - side B keeps sending until gsm0_receive() writes past gsm->buf Reason: Neither gsm->state nor gsm->len have been reset after reconfiguration. Fix this by changing gsm->count to gsm->len comparison from equal to less than. Also add upper limit checks against the constant MAX_MRU in gsm0_receive() and gsm1_receive() to harden against memory corruption of gsm->len and gsm->mru. All other checks remain as we still need to limit the data according to the user configuration and actual payload size. Reported-by: j51569436(a)gmail.com Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218708 Tested-by: j51569436(a)gmail.com Fixes: e1eaea46bb40 ("tty: n_gsm line discipline") Cc: stable(a)vger.kernel.org Signed-off-by: Daniel Starke <daniel.starke(a)siemens.com> Link: https://lore.kernel.org/r/20240424054842.7741-1-daniel.starke@siemens.com Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Cui GaoSheng <cuigaosheng1(a)huawei.com> --- drivers/tty/n_gsm.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/tty/n_gsm.c b/drivers/tty/n_gsm.c index 0ee7531c9201..f8858865c5bb 100644 --- a/drivers/tty/n_gsm.c +++ b/drivers/tty/n_gsm.c @@ -2912,7 +2912,10 @@ static void gsm0_receive(struct gsm_mux *gsm, unsigned char c) break; case GSM_DATA: /* Data */ gsm->buf[gsm->count++] = c; - if (gsm->count == gsm->len) { + if (gsm->count >= MAX_MRU) { + gsm->bad_size++; + gsm->state = GSM_SEARCH; + } else if (gsm->count >= gsm->len) { /* Calculate final FCS for UI frames over all data */ if ((gsm->control & ~PF) != UIH) { gsm->fcs = gsm_fcs_add_block(gsm->fcs, gsm->buf, @@ -3025,7 +3028,7 @@ static void gsm1_receive(struct gsm_mux *gsm, unsigned char c) gsm->state = GSM_DATA; break; case GSM_DATA: /* Data */ - if (gsm->count > gsm->mru) { /* Allow one for the FCS */ + if (gsm->count > gsm->mru || gsm->count > MAX_MRU) { /* Allow one for the FCS */ gsm->state = GSM_OVERRUN; gsm->bad_size++; } else -- 2.34.1

2 1

[PATCH openEuler-22.03-LTS-SP1] tty: n_gsm: fix possible out-of-bounds in gsm0_receive()
by Cui GaoSheng 12 Jun '24

12 Jun '24

From: Daniel Starke <daniel.starke(a)siemens.com> mainline inclusion from mainline-v6.10-rc1 commit 47388e807f85948eefc403a8a5fdc5b406a65d5a category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9TAHR CVE: CVE-2024-36016 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- Assuming the following: - side A configures the n_gsm in basic option mode - side B sends the header of a basic option mode frame with data length 1 - side A switches to advanced option mode - side B sends 2 data bytes which exceeds gsm->len Reason: gsm->len is not used in advanced option mode. - side A switches to basic option mode - side B keeps sending until gsm0_receive() writes past gsm->buf Reason: Neither gsm->state nor gsm->len have been reset after reconfiguration. Fix this by changing gsm->count to gsm->len comparison from equal to less than. Also add upper limit checks against the constant MAX_MRU in gsm0_receive() and gsm1_receive() to harden against memory corruption of gsm->len and gsm->mru. All other checks remain as we still need to limit the data according to the user configuration and actual payload size. Reported-by: j51569436(a)gmail.com Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218708 Tested-by: j51569436(a)gmail.com Fixes: e1eaea46bb40 ("tty: n_gsm line discipline") Cc: stable(a)vger.kernel.org Signed-off-by: Daniel Starke <daniel.starke(a)siemens.com> Link: https://lore.kernel.org/r/20240424054842.7741-1-daniel.starke@siemens.com Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Conflicts: drivers/tty/n_gsm.c [7a0e4b1733b6 ("tty: n_gsm: fix frame reception handling") has not been merged, context conflict] Signed-off-by: Cui GaoSheng <cuigaosheng1(a)huawei.com> --- drivers/tty/n_gsm.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/drivers/tty/n_gsm.c b/drivers/tty/n_gsm.c index 6cd87e71fecc..594be57f44ec 100644 --- a/drivers/tty/n_gsm.c +++ b/drivers/tty/n_gsm.c @@ -2015,8 +2015,12 @@ static void gsm0_receive(struct gsm_mux *gsm, unsigned char c) break; case GSM_DATA: /* Data */ gsm->buf[gsm->count++] = c; - if (gsm->count == gsm->len) + if (gsm->count >= MAX_MRU) { + gsm->bad_size++; + gsm->state = GSM_SEARCH; + } else if (gsm->count >= gsm->len) { gsm->state = GSM_FCS; + } break; case GSM_FCS: /* FCS follows the packet */ gsm->received_fcs = c; @@ -2109,7 +2113,7 @@ static void gsm1_receive(struct gsm_mux *gsm, unsigned char c) gsm->state = GSM_DATA; break; case GSM_DATA: /* Data */ - if (gsm->count > gsm->mru) { /* Allow one for the FCS */ + if (gsm->count > gsm->mru || gsm->count > MAX_MRU) { /* Allow one for the FCS */ gsm->state = GSM_OVERRUN; gsm->bad_size++; } else -- 2.34.1

2 1

2024

2023

2022

2021

2020

2019

Kernel