Re: [PATCH 1/1] UBIFS: fix incorrect unlocking handling

[Taesoo Kim]

Yes. The last commit that modifies 'ubifs/dir.c' (sorry for missing
the context). All error handling routines introduced by
'd7f0b70d30ffb9bbe6b8a3e1035cf0b79965ef53' 1) incorrectly 'unlock' and
2) incorrectly restore 'i_size'.

Thanks,
Taesoo

On 03/17/15 at 08:18pm, Brian Norris wrote:
> On Tue, Mar 17, 2015 at 10:09:26PM -0400, Taesoo Kim wrote:
> > When ubifs_init_security() fails, 'ui_mutex' is incorrectly
> > unlocked and incorrectly restores 'i_size'. There are four
> > such places that were introduce by the last commit.
> 
> "The last commit" is not very descriptive. Are you speaking of this
> commit?
> 
> commit d7f0b70d30ffb9bbe6b8a3e1035cf0b79965ef53
> Author: Subodh Nijsure <snijsure@xxxxxxxxxxxx>
> Date:   Fri Oct 31 13:50:30 2014 -0500
> 
>     UBIFS: Add security.* XATTR support for the UBIFS
> 
> CC'ing authors/reviewers.
> 
> Brian
> 
> > Signed-off-by: Taesoo Kim <tsgatesv@xxxxxxxxx>
> > ---
> >  fs/ubifs/dir.c | 11 +++++++----
> >  1 file changed, 7 insertions(+), 4 deletions(-)
> > 
> > diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c
> > index 0fa6c80..5b24bc4 100644
> > --- a/fs/ubifs/dir.c
> > +++ b/fs/ubifs/dir.c
> > @@ -272,7 +272,7 @@ static int ubifs_create(struct inode *dir, struct dentry *dentry, umode_t mode,
> > 
> >  	err = ubifs_init_security(dir, inode, &dentry->d_name);
> >  	if (err)
> > -		goto out_cancel;
> > +		goto out_inode;
> > 
> >  	mutex_lock(&dir_ui->ui_mutex);
> >  	dir->i_size += sz_change;
> > @@ -292,6 +292,7 @@ out_cancel:
> >  	dir->i_size -= sz_change;
> >  	dir_ui->ui_size = dir->i_size;
> >  	mutex_unlock(&dir_ui->ui_mutex);
> > +out_inode:
> >  	make_bad_inode(inode);
> >  	iput(inode);
> >  out_budg:
> > @@ -732,7 +733,7 @@ static int ubifs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
> > 
> >  	err = ubifs_init_security(dir, inode, &dentry->d_name);
> >  	if (err)
> > -		goto out_cancel;
> > +		goto out_inode;
> > 
> >  	mutex_lock(&dir_ui->ui_mutex);
> >  	insert_inode_hash(inode);
> > @@ -757,6 +758,7 @@ out_cancel:
> >  	dir_ui->ui_size = dir->i_size;
> >  	drop_nlink(dir);
> >  	mutex_unlock(&dir_ui->ui_mutex);
> > +out_inode:
> >  	make_bad_inode(inode);
> >  	iput(inode);
> >  out_budg:
> > @@ -816,7 +818,7 @@ static int ubifs_mknod(struct inode *dir, struct dentry *dentry,
> > 
> >  	err = ubifs_init_security(dir, inode, &dentry->d_name);
> >  	if (err)
> > -		goto out_cancel;
> > +		goto out_inode;
> > 
> >  	mutex_lock(&dir_ui->ui_mutex);
> >  	dir->i_size += sz_change;
> > @@ -836,6 +838,7 @@ out_cancel:
> >  	dir->i_size -= sz_change;
> >  	dir_ui->ui_size = dir->i_size;
> >  	mutex_unlock(&dir_ui->ui_mutex);
> > +out_inode:
> >  	make_bad_inode(inode);
> >  	iput(inode);
> >  out_budg:
> > @@ -896,7 +899,7 @@ static int ubifs_symlink(struct inode *dir, struct dentry *dentry,
> > 
> >  	err = ubifs_init_security(dir, inode, &dentry->d_name);
> >  	if (err)
> > -		goto out_cancel;
> > +		goto out_inode;
> > 
> >  	mutex_lock(&dir_ui->ui_mutex);
> >  	dir->i_size += sz_change;
> > --
> > 2.3.3
> > 
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/