From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-wr0-f200.google.com (mail-wr0-f200.google.com [209.85.128.200]) by kanga.kvack.org (Postfix) with ESMTP id E7F396B0003 for ; Sat, 17 Mar 2018 11:55:05 -0400 (EDT) Received: by mail-wr0-f200.google.com with SMTP id 7so7180255wrp.2 for ; Sat, 17 Mar 2018 08:55:05 -0700 (PDT) Received: from mail-sor-f65.google.com (mail-sor-f65.google.com. [209.85.220.65]) by mx.google.com with SMTPS id i22sor4765576eda.24.2018.03.17.08.55.03 for (Google Transport Security); Sat, 17 Mar 2018 08:55:03 -0700 (PDT) Date: Sat, 17 Mar 2018 18:54:37 +0300 From: "Kirill A. Shutemov" Subject: Re: [PATCH] mm: Warn on lock_page() from reclaim context. Message-ID: <20180317155437.pcbeigeivn4a23gt@node.shutemov.name> References: <1521295866-9670-1-git-send-email-penguin-kernel@I-love.SAKURA.ne.jp> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1521295866-9670-1-git-send-email-penguin-kernel@I-love.SAKURA.ne.jp> Sender: owner-linux-mm@kvack.org List-ID: To: Tetsuo Handa Cc: akpm@linux-foundation.org, linux-mm@kvack.org, "Kirill A. Shutemov" , Michal Hocko On Sat, Mar 17, 2018 at 11:11:06PM +0900, Tetsuo Handa wrote: > Kirill A. Shutemov noticed that calling lock_page[_killable]() from > reclaim context might cause deadlock. In order to help finding such > lock_page[_killable]() users (including out of tree users), this patch > emits warning messages when CONFIG_PROVE_LOCKING is enabled. > > [ 81.532721] ------------[ cut here ]------------ > [ 81.534253] lock_page() from reclaim context might deadlock > [ 81.534264] WARNING: CPU: 3 PID: 56 at mm/filemap.c:3340 __warn_lock_page_from_reclaim_context+0x1e/0x30 > [ 81.539982] Modules linked in: pcspkr sg vmw_vmci shpchp i2c_piix4 sd_mod ata_generic pata_acpi vmwgfx drm_kms_helper syscopyarea sysfillrect mptspi sysimgblt fb_sys_fops ahci ttm libahci e1000 scsi_transport_spi mptscsih drm ata_piix mptbase i2c_core libata serio_raw ipv6 > [ 81.546750] CPU: 3 PID: 56 Comm: kswapd0 Kdump: loaded Not tainted 4.16.0-rc5-next-20180316+ #698 > [ 81.549333] Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 05/19/2017 > [ 81.552256] RIP: 0010:__warn_lock_page_from_reclaim_context+0x1e/0x30 > [ 81.554164] RSP: 0018:ffffc9000085bbf8 EFLAGS: 00010282 > [ 81.555833] RAX: 0000000000000000 RBX: ffffea0004b2b880 RCX: 0000000000000007 > [ 81.558283] RDX: 0000000000000b6d RSI: ffff88013aa0b700 RDI: ffff88013aa0ae80 > [ 81.562130] RBP: 0000000000000000 R08: 0000000000000001 R09: 0000000000000000 > [ 81.564291] R10: 0000000000000040 R11: 0000000000000000 R12: ffff8801357b4958 > [ 81.566551] R13: 0000000000000000 R14: ffffffff82068220 R15: 0000000000000002 > [ 81.568956] FS: 0000000000000000(0000) GS:ffff88013bcc0000(0000) knlGS:0000000000000000 > [ 81.571310] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 > [ 81.573193] CR2: 00000000006020bf CR3: 000000000200f005 CR4: 00000000001606e0 > [ 81.575453] Call Trace: > [ 81.577001] pagecache_get_page+0x22a/0x230 > [ 81.578497] ? iput+0x52/0x2f0 > [ 81.579762] shmem_unused_huge_shrink+0x2e9/0x380 > [ 81.581465] super_cache_scan+0x17a/0x180 > [ 81.582999] shrink_slab+0x218/0x590 > [ 81.584398] shrink_node+0x346/0x350 > [ 81.585777] kswapd+0x322/0x930 > [ 81.587090] kthread+0xf0/0x130 > [ 81.588370] ? mem_cgroup_shrink_node+0x320/0x320 > [ 81.589996] ? kthread_create_on_node+0x60/0x60 > [ 81.591670] ret_from_fork+0x3a/0x50 > [ 81.593716] Code: f0 41 80 8c 24 08 01 00 00 02 eb b3 90 80 3d b8 54 f8 00 00 74 02 f3 c3 48 c7 c7 b0 87 df 81 c6 05 a6 54 f8 00 01 e8 d2 89 ee ff <0f> 0b c3 0f 1f 44 00 00 66 2e 0f 1f 84 00 00 00 00 00 41 55 41 > [ 81.599356] ---[ end trace d9238c3d53557ed5 ]--- > > Signed-off-by: Tetsuo Handa > Cc: Kirill A. Shutemov > Cc: Michal Hocko > --- > include/linux/pagemap.h | 12 ++++++++++++ > mm/filemap.c | 8 ++++++++ > 2 files changed, 20 insertions(+) > > diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h > index 34ce3ebf..8e38f4a 100644 > --- a/include/linux/pagemap.h > +++ b/include/linux/pagemap.h > @@ -467,6 +467,10 @@ extern int __lock_page_or_retry(struct page *page, struct mm_struct *mm, > unsigned int flags); > extern void unlock_page(struct page *page); > > +#ifdef CONFIG_PROVE_LOCKING > +extern void __warn_lock_page_from_reclaim_context(void); > +#endif > + No need in ifdeffery around declarations. They are harmless. > static inline int trylock_page(struct page *page) > { > page = compound_head(page); > @@ -479,6 +483,10 @@ static inline int trylock_page(struct page *page) > static inline void lock_page(struct page *page) > { > might_sleep(); > +#ifdef CONFIG_PROVE_LOCKING > + if (unlikely(current->flags & PF_MEMALLOC)) IS_ENABLED(CONFIG_PROVE_LOCKING) > + __warn_lock_page_from_reclaim_context(); > +#endif > if (!trylock_page(page)) > __lock_page(page); > } > @@ -491,6 +499,10 @@ static inline void lock_page(struct page *page) > static inline int lock_page_killable(struct page *page) > { > might_sleep(); > +#ifdef CONFIG_PROVE_LOCKING > + if (unlikely(current->flags & PF_MEMALLOC)) Ditto. > + __warn_lock_page_from_reclaim_context(); > +#endif > if (!trylock_page(page)) > return __lock_page_killable(page); > return 0; > diff --git a/mm/filemap.c b/mm/filemap.c > index 693f622..de2eb21 100644 > --- a/mm/filemap.c > +++ b/mm/filemap.c > @@ -3328,3 +3328,11 @@ int try_to_release_page(struct page *page, gfp_t gfp_mask) > } > > EXPORT_SYMBOL(try_to_release_page); > + > +#ifdef CONFIG_PROVE_LOCKING > +void __warn_lock_page_from_reclaim_context(void) > +{ > + WARN_ONCE(1, "lock_page() from reclaim context might deadlock"); > +} > +EXPORT_SYMBOL(__warn_lock_page_from_reclaim_context); > +#endif > -- > 1.8.3.1 > -- Kirill A. Shutemov