| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <lfc5sqvxmxho5q7hyjflrcvx7jb4gjtzb4zf6fsrsafpmgplyn@e2gr4hmamaqo>
Date: Tue, 24 Jun 2025 10:03:51 +1000
From: Alistair Popple <apopple@...dia.com>
To: Haiyue Wang <haiyuewa@....com>
Cc: Andrew Morton <akpm@...ux-foundation.org>,
Miklos Szeredi <miklos@...redi.hu>, Dan Williams <dan.j.williams@...el.com>,
"open list:FUSE: FILESYSTEM IN USERSPACE" <linux-fsdevel@...r.kernel.org>, open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3] fuse: Fix runtime warning on
truncate_folio_batch_exceptionals()
On Sun, Jun 22, 2025 at 01:13:51AM +0800, Haiyue Wang wrote:
> The WARN_ON_ONCE is introduced on truncate_folio_batch_exceptionals() to
> capture whether the filesystem has removed all DAX entries or not.
>
> And the fix has been applied on the filesystem xfs and ext4 by the
> commit 0e2f80afcfa6 ("fs/dax: ensure all pages are idle prior to filesystem unmount").
>
> Apply the missed fix on filesystem fuse to fix the runtime warning:
>
> [ 2.011450] ------------[ cut here ]------------
> [ 2.011873] WARNING: CPU: 0 PID: 145 at mm/truncate.c:89 truncate_folio_batch_exceptionals+0x272/0x2b0
> [ 2.012468] Modules linked in:
> [ 2.012718] CPU: 0 UID: 1000 PID: 145 Comm: weston Not tainted 6.16.0-rc2-WSL2-STABLE #2 PREEMPT(undef)
> [ 2.013292] RIP: 0010:truncate_folio_batch_exceptionals+0x272/0x2b0
> [ 2.013704] Code: 48 63 d0 41 29 c5 48 8d 1c d5 00 00 00 00 4e 8d 6c 2a 01 49 c1 e5 03 eb 09 48 83 c3 08 49 39 dd 74 83 41 f6 44 1c 08 01 74 ef <0f> 0b 49 8b 34 1e 48 89 ef e8 10 a2 17 00 eb df 48 8b 7d 00 e8 35
> [ 2.014845] RSP: 0018:ffffa47ec33f3b10 EFLAGS: 00010202
> [ 2.015279] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
> [ 2.015884] RDX: 0000000000000000 RSI: ffffa47ec33f3ca0 RDI: ffff98aa44f3fa80
> [ 2.016377] RBP: ffff98aa44f3fbf0 R08: ffffa47ec33f3ba8 R09: 0000000000000000
> [ 2.016942] R10: 0000000000000001 R11: 0000000000000000 R12: ffffa47ec33f3ca0
> [ 2.017437] R13: 0000000000000008 R14: ffffa47ec33f3ba8 R15: 0000000000000000
> [ 2.017972] FS: 000079ce006afa40(0000) GS:ffff98aade441000(0000) knlGS:0000000000000000
> [ 2.018510] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [ 2.018987] CR2: 000079ce03e74000 CR3: 000000010784f006 CR4: 0000000000372eb0
> [ 2.019518] Call Trace:
> [ 2.019729] <TASK>
> [ 2.019901] truncate_inode_pages_range+0xd8/0x400
> [ 2.020280] ? timerqueue_add+0x66/0xb0
> [ 2.020574] ? get_nohz_timer_target+0x2a/0x140
> [ 2.020904] ? timerqueue_add+0x66/0xb0
> [ 2.021231] ? timerqueue_del+0x2e/0x50
> [ 2.021646] ? __remove_hrtimer+0x39/0x90
> [ 2.022017] ? srso_alias_untrain_ret+0x1/0x10
> [ 2.022497] ? psi_group_change+0x136/0x350
> [ 2.023046] ? _raw_spin_unlock+0xe/0x30
> [ 2.023514] ? finish_task_switch.isra.0+0x8d/0x280
> [ 2.024068] ? __schedule+0x532/0xbd0
> [ 2.024551] fuse_evict_inode+0x29/0x190
> [ 2.025131] evict+0x100/0x270
> [ 2.025641] ? _atomic_dec_and_lock+0x39/0x50
> [ 2.026316] ? __pfx_generic_delete_inode+0x10/0x10
> [ 2.026843] __dentry_kill+0x71/0x180
> [ 2.027335] dput+0xeb/0x1b0
> [ 2.027725] __fput+0x136/0x2b0
> [ 2.028054] __x64_sys_close+0x3d/0x80
> [ 2.028469] do_syscall_64+0x6d/0x1b0
> [ 2.028832] ? clear_bhb_loop+0x30/0x80
> [ 2.029182] ? clear_bhb_loop+0x30/0x80
> [ 2.029533] ? clear_bhb_loop+0x30/0x80
> [ 2.029902] entry_SYSCALL_64_after_hwframe+0x76/0x7e
> [ 2.030423] RIP: 0033:0x79ce03d0d067
> [ 2.030820] Code: b8 ff ff ff ff e9 3e ff ff ff 66 0f 1f 84 00 00 00 00 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 b8 03 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 41 c3 48 83 ec 18 89 7c 24 0c e8 c3 a7 f8 ff
> [ 2.032354] RSP: 002b:00007ffef0498948 EFLAGS: 00000246 ORIG_RAX: 0000000000000003
> [ 2.032939] RAX: ffffffffffffffda RBX: 00007ffef0498960 RCX: 000079ce03d0d067
> [ 2.033612] RDX: 0000000000000003 RSI: 0000000000001000 RDI: 000000000000000d
> [ 2.034289] RBP: 00007ffef0498a30 R08: 000000000000000d R09: 0000000000000000
> [ 2.034944] R10: 00007ffef0498978 R11: 0000000000000246 R12: 0000000000000001
> [ 2.035610] R13: 00007ffef0498960 R14: 000079ce03e09ce0 R15: 0000000000000003
> [ 2.036301] </TASK>
> [ 2.036532] ---[ end trace 0000000000000000 ]---
>
> Fixes: bde708f1a65d ("fs/dax: always remove DAX page-cache entries when breaking layouts")
Thanks Haiyue, and sorry for not noticing this in my inbox sooner. Fix looks good though, so please add:
Reviewed-by: Alistair Popple <apopple@...dia.com>
> Signed-off-by: Haiyue Wang <haiyuewa@....com>
> ---
> v3:
> - Rewrite the commit message to make things clean
> v2: https://lore.kernel.org/linux-fsdevel/20250608042418.358-1-haiyuewa@163.com/
> - Use 'FUSE_IS_DAX()' to control DAX checking
> v1: https://lore.kernel.org/linux-fsdevel/20250604142251.2426-1-haiyuewa@163.com/
> ---
> fs/fuse/inode.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
> index bfe8d8af46f3..9572bdef49ee 100644
> --- a/fs/fuse/inode.c
> +++ b/fs/fuse/inode.c
> @@ -9,6 +9,7 @@
> #include "fuse_i.h"
> #include "dev_uring_i.h"
>
> +#include <linux/dax.h>
> #include <linux/pagemap.h>
> #include <linux/slab.h>
> #include <linux/file.h>
> @@ -162,6 +163,9 @@ static void fuse_evict_inode(struct inode *inode)
> /* Will write inode on close/munmap and in all other dirtiers */
> WARN_ON(inode->i_state & I_DIRTY_INODE);
>
> + if (FUSE_IS_DAX(inode))
> + dax_break_layout_final(inode);
> +
> truncate_inode_pages_final(&inode->i_data);
> clear_inode(inode);
> if (inode->i_sb->s_flags & SB_ACTIVE) {
> --
> 2.49.0
>
Powered by blists - more mailing lists