linux-kernel - Re: [PATCH] f2fs: fix ifolio memory leak in f2fs_move_inline

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <aN8_05vg6Lz1eAkF@google.com>
Date: Fri, 3 Oct 2025 03:15:31 +0000
From: Jaegeuk Kim <jaegeuk@...nel.org>
To: Chao Yu <chao@...nel.org>
Cc: Haofeng Li <920484857@...com>, linux-f2fs-devel@...ts.sourceforge.net,
	linux-kernel@...r.kernel.org, Haofeng Li <13266079573@....com>,
	Haofeng Li <lihaofeng@...inos.cn>
Subject: Re: [PATCH] f2fs: fix ifolio memory leak in f2fs_move_inline_dirents
 error path

On 10/03, Chao Yu wrote:
> On 2025/9/30 17:16, Haofeng Li wrote:
> > From: Haofeng Li <lihaofeng@...inos.cn>
> > 
> > Fixes a memory leak issue in f2fs_move_inline_dirents() where
> > the ifolio is not properly released in certain error paths.
> > 
> > Problem Analysis:
> > - In f2fs_try_convert_inline_dir(), ifolio is acquired via f2fs_get_inode_folio()
> > - When do_convert_inline_dir() fails, the caller expects ifolio to be released
> > - However, in f2fs_move_inline_dirents(), two specific error paths don't release ifolio
> > 
> > Fixes: 201a05be9628a ("f2fs: add key function to handle inline dir")
> > Signed-off-by: Haofeng Li <lihaofeng@...inos.cn>
> > ---
> >   fs/f2fs/inline.c | 8 ++++++--
> >   1 file changed, 6 insertions(+), 2 deletions(-)
> > 
> > diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c
> > index 58ac831ef704..2496866fc45d 100644
> > --- a/fs/f2fs/inline.c
> > +++ b/fs/f2fs/inline.c
> > @@ -425,7 +425,7 @@ static int f2fs_move_inline_dirents(struct inode *dir, struct folio *ifolio,
> >   	set_new_dnode(&dn, dir, ifolio, NULL, 0);
> >   	err = f2fs_reserve_block(&dn, 0);
> 
> f2fs_reserve_block() will call f2fs_put_dnode() in its error path, it has
> unlocked & released inode folio?
> 
> >   	if (err)
> > -		goto out;
> > +		goto out_put_ifolio;
> >   	if (unlikely(dn.data_blkaddr != NEW_ADDR)) {
> >   		f2fs_put_dnode(&dn);
> 
> Ditto, or am I missing something?

It seems you're right. Let me drop this patch.

> 
> Thanks,
> 
> > @@ -434,7 +434,7 @@ static int f2fs_move_inline_dirents(struct inode *dir, struct folio *ifolio,
> >   			  __func__, dir->i_ino, dn.data_blkaddr);
> >   		f2fs_handle_error(F2FS_F_SB(folio), ERROR_INVALID_BLKADDR);
> >   		err = -EFSCORRUPTED;
> > -		goto out;
> > +		goto out_put_ifolio;
> >   	}
> >   	f2fs_folio_wait_writeback(folio, DATA, true, true);
> > @@ -479,6 +479,10 @@ static int f2fs_move_inline_dirents(struct inode *dir, struct folio *ifolio,
> >   out:
> >   	f2fs_folio_put(folio, true);
> >   	return err;
> > +
> > +out_put_ifolio:
> > +	f2fs_folio_put(ifolio, true);
> > +	goto out;
> >   }
> >   static int f2fs_add_inline_entries(struct inode *dir, void *inline_dentry)