The patch titled
Subject: slub: fix failure when we delete and create a slab cache
has been added to the -mm tree. Its filename is
slub-fix-failure-when-we-delete-and-create-a-slab-cache.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/slub-fix-failure-when-we-delete-and-create-a-slab-cache.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/slub-fix-failure-when-we-delete-and-create-a-slab-cache.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Mikulas Patocka <mpatocka@xxxxxxxxxx>
Subject: slub: fix failure when we delete and create a slab cache
In kernel 4.17 I removed some code from dm-bufio that did slab cache
merging (21bb13276768) - both slab and slub support merging caches with
identical attributes, so dm-bufio now just calls kmem_cache_create and
relies on implicit merging.
This uncovered a bug in the slub subsystem - if we delete a cache and
immediatelly create another cache with the same attributes, it fails
because of duplicate filename in /sys/kernel/slab/. The slub subsystem
offloads freeing the cache to a workqueue - and if we create the new cache
before the workqueue runs, it complains because of duplicate filename in
sysfs.
This patch fixes the bug by moving the call of kobject_del from
sysfs_slab_remove_workfn to shutdown_cache. kobject_del must be called
while we hold slab_mutex - so that the sysfs entry is deleted before a
cache with the same attributes could be created.
Running device-mapper-test-suite with:
dmtest run --suite thin-provisioning -n /commit_failure_causes_fallback/
triggers:
[ 119.618958] Buffer I/O error on dev dm-0, logical block 1572848, async page read
[ 119.686224] device-mapper: thin: 253:1: metadata operation 'dm_pool_alloc_data_block' failed: error = -5
[ 119.695821] device-mapper: thin: 253:1: aborting current metadata transaction
[ 119.703255] sysfs: cannot create duplicate filename '/kernel/slab/:a-0000144'
[ 119.710394] CPU: 2 PID: 1037 Comm: kworker/u48:1 Not tainted 4.17.0.snitm+ #25
[ 119.717608] Hardware name: Supermicro SYS-1029P-WTR/X11DDW-L, BIOS 2.0a 12/06/2017
[ 119.725177] Workqueue: dm-thin do_worker [dm_thin_pool]
[ 119.730401] Call Trace:
[ 119.732856] dump_stack+0x5a/0x73
[ 119.736173] sysfs_warn_dup+0x58/0x70
[ 119.739839] sysfs_create_dir_ns+0x77/0x80
[ 119.743939] kobject_add_internal+0xba/0x2e0
[ 119.748210] kobject_init_and_add+0x70/0xb0
[ 119.752399] ? sysfs_slab_add+0x101/0x250
[ 119.756409] sysfs_slab_add+0xb1/0x250
[ 119.760161] __kmem_cache_create+0x116/0x150
[ 119.764436] ? number+0x2fb/0x340
[ 119.767755] ? _cond_resched+0x15/0x30
[ 119.771508] create_cache+0xd9/0x1f0
[ 119.775085] kmem_cache_create_usercopy+0x1c1/0x250
[ 119.779965] kmem_cache_create+0x18/0x20
[ 119.783894] dm_bufio_client_create+0x1ae/0x410 [dm_bufio]
[ 119.789380] ? dm_block_manager_alloc_callback+0x20/0x20 [dm_persistent_data]
[ 119.796509] ? kmem_cache_alloc_trace+0xae/0x1d0
[ 119.801131] dm_block_manager_create+0x5e/0x90 [dm_persistent_data]
[ 119.807397] __create_persistent_data_objects+0x38/0x940 [dm_thin_pool]
[ 119.814008] dm_pool_abort_metadata+0x64/0x90 [dm_thin_pool]
[ 119.819669] metadata_operation_failed+0x59/0x100 [dm_thin_pool]
[ 119.825673] alloc_data_block.isra.53+0x86/0x180 [dm_thin_pool]
[ 119.831592] process_cell+0x2a3/0x550 [dm_thin_pool]
[ 119.836558] ? mempool_alloc+0x6f/0x180
[ 119.840400] ? u32_swap+0x10/0x10
[ 119.843717] ? sort+0x17b/0x270
[ 119.846863] ? u32_swap+0x10/0x10
[ 119.850181] do_worker+0x28d/0x8f0 [dm_thin_pool]
[ 119.854890] ? move_linked_works+0x6f/0xa0
[ 119.858989] process_one_work+0x171/0x370
[ 119.862999] worker_thread+0x49/0x3f0
[ 119.866669] kthread+0xf8/0x130
[ 119.869813] ? max_active_store+0x80/0x80
[ 119.873827] ? kthread_bind+0x10/0x10
[ 119.877493] ret_from_fork+0x35/0x40
[ 119.881076] kobject_add_internal failed for :a-0000144 with -EEXIST, don't try to register things with the same name in the same directory.
[ 119.893580] kmem_cache_create(dm_bufio_buffer-16) failed with error -17
Link: http://lkml.kernel.org/r/alpine.LRH.2.02.1806151817130.6333@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Signed-off-by: Mikulas Patocka <mpatocka@xxxxxxxxxx>
Reported-by: Mike Snitzer <snitzer@xxxxxxxxxx>
Tested-by: Mike Snitzer <snitzer@xxxxxxxxxx>
Cc: Christoph Lameter <cl@xxxxxxxxx>
Cc: Pekka Enberg <penberg@xxxxxxxxxx>
Cc: David Rientjes <rientjes@xxxxxxxxxx>
Cc: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
Cc: <stable@xxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/slub_def.h | 4 ++++
mm/slab_common.c | 4 ++++
mm/slub.c | 7 ++++++-
3 files changed, 14 insertions(+), 1 deletion(-)
diff -puN include/linux/slub_def.h~slub-fix-failure-when-we-delete-and-create-a-slab-cache include/linux/slub_def.h
--- a/include/linux/slub_def.h~slub-fix-failure-when-we-delete-and-create-a-slab-cache
+++ a/include/linux/slub_def.h
@@ -155,8 +155,12 @@ struct kmem_cache {
#ifdef CONFIG_SYSFS
#define SLAB_SUPPORTS_SYSFS
+void sysfs_slab_unlink(struct kmem_cache *);
void sysfs_slab_release(struct kmem_cache *);
#else
+static inline void sysfs_slab_unlink(struct kmem_cache *s)
+{
+}
static inline void sysfs_slab_release(struct kmem_cache *s)
{
}
diff -puN mm/slab_common.c~slub-fix-failure-when-we-delete-and-create-a-slab-cache mm/slab_common.c
--- a/mm/slab_common.c~slub-fix-failure-when-we-delete-and-create-a-slab-cache
+++ a/mm/slab_common.c
@@ -567,10 +567,14 @@ static int shutdown_cache(struct kmem_ca
list_del(&s->list);
if (s->flags & SLAB_TYPESAFE_BY_RCU) {
+#ifdef SLAB_SUPPORTS_SYSFS
+ sysfs_slab_unlink(s);
+#endif
list_add_tail(&s->list, &slab_caches_to_rcu_destroy);
schedule_work(&slab_caches_to_rcu_destroy_work);
} else {
#ifdef SLAB_SUPPORTS_SYSFS
+ sysfs_slab_unlink(s);
sysfs_slab_release(s);
#else
slab_kmem_cache_release(s);
diff -puN mm/slub.c~slub-fix-failure-when-we-delete-and-create-a-slab-cache mm/slub.c
--- a/mm/slub.c~slub-fix-failure-when-we-delete-and-create-a-slab-cache
+++ a/mm/slub.c
@@ -5667,7 +5667,6 @@ static void sysfs_slab_remove_workfn(str
kset_unregister(s->memcg_kset);
#endif
kobject_uevent(&s->kobj, KOBJ_REMOVE);
- kobject_del(&s->kobj);
out:
kobject_put(&s->kobj);
}
@@ -5752,6 +5751,12 @@ static void sysfs_slab_remove(struct kme
schedule_work(&s->kobj_remove_work);
}
+void sysfs_slab_unlink(struct kmem_cache *s)
+{
+ if (slab_state >= FULL)
+ kobject_del(&s->kobj);
+}
+
void sysfs_slab_release(struct kmem_cache *s)
{
if (slab_state >= FULL)
_
Patches currently in -mm which might be from mpatocka@xxxxxxxxxx are
slub-fix-failure-when-we-delete-and-create-a-slab-cache.patch
