On Fri, Sep 30, 2022 at 04:43:23PM +0200, Vlastimil Babka wrote: > On 9/30/22 13:51, Feng Tang wrote: > > On Fri, Sep 30, 2022 at 07:25:54PM +0800, Vlastimil Babka wrote: > >> > >> On 9/30/22 12:07, Feng Tang wrote: > >> > When enable kasan and kfence's in-kernel kunit test with slub_debug on, > >> > it caught a problem (in linux-next tree): > >> > > >> > ------------[ cut here ]------------ > >> > kmem_cache_destroy test: Slab cache still has objects when called from test_exit+0x1a/0x30 > >> > WARNING: CPU: 3 PID: 240 at mm/slab_common.c:492 kmem_cache_destroy+0x16c/0x170 > >> > >> Assuming the warning was preceded by some kunit test failures? > >> I don't see how leaving more empty slabs on free list than needed would > >> cause this warning, the shutdown should just drop the empty slab. > > > > The previous code only call remove_partial() to dequeue the slab from > > partial list, and miss to call discard_slab() for it. > > > > From the debug dump, the n->nr_partils stays at 5, while n->nr_slabs > > keeps increasing. And during shutdown, the free_partial() only free > > the 5 slabs on partial list, and n->nr_slabs still has a big numbers > > of orphan slabs > > Thanks, finally I get the exact cause now. I've added the more detailed > explanation to commit log and the result is here: > https://git.kernel.org/pub/scm/linux/kernel/git/vbabka/slab.git/commit/?h=for-6.1/slub_validation_locking&id=b731e3575f7a45a46512708f9fdf953b40c46a53 > Very nice finding Feng, thanks! Yeah, there are some cases where first and second check do not agree, leading unfreed slabs. the latest version looks good to me, Reviewed-by: Hyeonggon Yoo <42.hyeyoo@xxxxxxxxx> Nit: - discard_slab() is not what's actually called, but I get what you mean anyway... - s/Reoganize/Reorganize/g > >> > Modules linked in: > >> > CPU: 3 PID: 240 Comm: kunit_try_catch Tainted: G B N 6.0.0-rc7-next-20220929 #52 > >> > Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.15.0-1 04/01/2014 > >> > RIP: 0010:kmem_cache_destroy+0x16c/0x170 > >> > Code: 41 5c 41 5d e9 a5 04 0b 00 c3 cc cc cc cc 48 8b 55 60 48 8b 4c 24 20 48 c7 c6 40 37 d2 82 48 c7 c7 e8 a0 33 83 e8 4e d7 14 01 <0f> 0b eb a7 41 56 41 89 d6 41 55 49 89 f5 41 54 49 89 fc 55 48 89 > >> > RSP: 0000:ffff88800775fea0 EFLAGS: 00010282 > >> > RAX: 0000000000000000 RBX: ffffffff83bdec48 RCX: 0000000000000000 > >> > RDX: 0000000000000001 RSI: 1ffff11000eebf9e RDI: ffffed1000eebfc6 > >> > RBP: ffff88804362fa00 R08: ffffffff81182e58 R09: ffff88800775fbdf > >> > R10: ffffed1000eebf7b R11: 0000000000000001 R12: 000000008c800d00 > >> > R13: ffff888005e78040 R14: 0000000000000000 R15: ffff888005cdfad0 > >> > FS: 0000000000000000(0000) GS:ffff88807ed00000(0000) knlGS:0000000000000000 > >> > CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > >> > CR2: 0000000000000000 CR3: 000000000360e001 CR4: 0000000000370ee0 > >> > DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 > >> > DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 > >> > Call Trace: > >> > <TASK> > >> > test_exit+0x1a/0x30 > >> > kunit_try_run_case+0xad/0xc0 > >> > kunit_generic_run_threadfn_adapter+0x26/0x50 > >> > kthread+0x17b/0x1b0 > >> > > >> > It was biscted to commit c7323a5ad078 ("mm/slub: restrict sysfs > >> > validation to debug caches and make it safe") > >> > > >> > The problem is inside free_debug_processing(), in one path, the slab > >> > on partial list is missed to be freed when partial list is full. > >> > > >> > Signed-off-by: Feng Tang <feng.tang@xxxxxxxxx> > >> > --- > >> > > >> > Hi reviewers, > >> > > >> > Sorry for the late reporting, but it's curious that this problem didn't > >> > show up in my earlier test (which caught some other problems). > >> > >> I think we can reuse the slab_free and don't need a new bool? > > > > Yes, much simpler! > > > > Thanks, > > Feng > > > >> diff --git a/mm/slub.c b/mm/slub.c > >> index 5c3c31a154ba..a63953f649ed 100644 > >> --- a/mm/slub.c > >> +++ b/mm/slub.c > >> @@ -2886,22 +2886,25 @@ static noinline void free_debug_processing( > >> set_freepointer(s, tail, prior); > >> slab->freelist = head; > >> > >> - /* Do we need to remove the slab from full or partial list? */ > >> + /* > >> + * If the slab is empty, and node's partial list is full, > >> + * it should be discarded anyway no matter it's on full or > >> + * partial list. > >> + */ > >> + if (slab->inuse == 0 && n->nr_partial >= s->min_partial) > >> + slab_free = slab; > >> + > >> if (!prior) { > >> + /* was on full list */ > >> remove_full(s, n, slab); > >> - } else if (slab->inuse == 0 && > >> - n->nr_partial >= s->min_partial) { > >> + if (!slab_free) { > >> + add_partial(n, slab, DEACTIVATE_TO_TAIL); > >> + stat(s, FREE_ADD_PARTIAL); > >> + } > >> + } else if (slab_free) { > >> remove_partial(n, slab); > >> stat(s, FREE_REMOVE_PARTIAL); > >> } > >> - > >> - /* Do we need to discard the slab or add to partial list? */ > >> - if (slab->inuse == 0 && n->nr_partial >= s->min_partial) { > >> - slab_free = slab; > >> - } else if (!prior) { > >> - add_partial(n, slab, DEACTIVATE_TO_TAIL); > >> - stat(s, FREE_ADD_PARTIAL); > >> - } > >> } > >> > >> if (slab_free) { > -- Thanks, Hyeonggon
