linux-kernel - Re: [PATCH] f2fs: fix to avoid atomicity corruption of atomic file

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

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <Z9Sr9cjNGHkOM74n@google.com>
Date: Fri, 14 Mar 2025 22:21:41 +0000
From: Jaegeuk Kim <jaegeuk@...nel.org>
To: Yeongjin Gil <youngjin.gil@...sung.com>
Cc: chao@...nel.org, daehojeong@...gle.com,
	linux-f2fs-devel@...ts.sourceforge.net,
	linux-kernel@...r.kernel.org, sj1557.seo@...sung.com,
	s_min.jeong@...sung.com
Subject: Re: [PATCH] f2fs: fix to avoid atomicity corruption of atomic file

On 03/14, Jaegeuk Kim wrote:
> On 03/14, Yeongjin Gil wrote:
> > In the case of the following call stack for an atomic file,
> > FI_DIRTY_INODE is set, but FI_ATOMIC_DIRTIED is not subsequently set.
> > 
> > f2fs_file_write_iter
> >   f2fs_map_blocks
> >     f2fs_reserve_new_blocks
> >       inc_valid_block_count
> >         __mark_inode_dirty(dquot)
> >           f2fs_dirty_inode
> > 
> > If FI_ATOMIC_DIRTIED is not set, atomic file can encounter corruption
> > due to a mismatch between old file size and new data.
> > 
> > To resolve this issue, I changed to set FI_ATOMIC_DIRTIED when
> > FI_DIRTY_INODE is set. This ensures that FI_DIRTY_INODE, which was
> > previously cleared by the Writeback thread during the commit atomic, is
> > set and i_size is updated.
> > 
> > Fixes: fccaa81de87e ("f2fs: prevent atomic file from being dirtied before commit")
> > Reviewed-by: Sungjong Seo <sj1557.seo@...sung.com>
> > Reviewed-by: Sunmin Jeong <s_min.jeong@...sung.com>
> > Signed-off-by: Yeongjin Gil <youngjin.gil@...sung.com>
> > ---
> >  fs/f2fs/inode.c | 4 +---
> >  fs/f2fs/super.c | 4 ++++
> >  2 files changed, 5 insertions(+), 3 deletions(-)
> > 
> > diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c
> > index aa2f41696a88..83f862578fc8 100644
> > --- a/fs/f2fs/inode.c
> > +++ b/fs/f2fs/inode.c
> > @@ -34,10 +34,8 @@ void f2fs_mark_inode_dirty_sync(struct inode *inode, bool sync)
> >  	if (f2fs_inode_dirtied(inode, sync))
> >  		return;
> >  
> > -	if (f2fs_is_atomic_file(inode)) {
> > -		set_inode_flag(inode, FI_ATOMIC_DIRTIED);
> > +	if (f2fs_is_atomic_file(inode))
> >  		return;
> > -	}
> >  
> >  	mark_inode_dirty_sync(inode);
> >  }
> > diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
> > index 397df271885c..c08d52c6467a 100644
> > --- a/fs/f2fs/super.c
> > +++ b/fs/f2fs/super.c
> > @@ -1534,6 +1534,10 @@ int f2fs_inode_dirtied(struct inode *inode, bool sync)
> >  		inc_page_count(sbi, F2FS_DIRTY_IMETA);
> >  	}
> >  	spin_unlock(&sbi->inode_lock[DIRTY_META]);
> > +
> > +	if (!ret && f2fs_is_atomic_file(inode))
> > +		set_inode_flag(inode, FI_ATOMIC_DIRTIED);
> > +
> 
> I'm not sure what's different here.
> 
> Before and after this patch, set_inode_flag(inode, FI_ATOMIC_DIRTIED) is called
> only if f2fs_inode_dirtied() returns zero and f2fs_is_atomic_file(inode).
> 
> Note, f2fs_mark_inode_dirty_sync() looks the single caller of f2fs_inode_dirtied.

Ok, I missed another caller, f2fs_dirty_inode, per discussion with Daeho.
Let me apply this.

> 
> 
> >  	return ret;
> >  }
> >  
> > -- 
> > 2.34.1