Re: [PATCH v3] selinux: lsm: fix a missing-check bug in selinux_sb_eat_lsm_opts()
- Date: Sat, 1 Jun 2019 10:47:53 +0800
- From: Gen Zhang <blackgod016574@xxxxxxxxx>
- Subject: Re: [PATCH v3] selinux: lsm: fix a missing-check bug in selinux_sb_eat_lsm_opts()
On Sat, Jun 01, 2019 at 03:34:49AM +0100, Al Viro wrote:
> On Sat, Jun 01, 2019 at 03:25:27AM +0100, Al Viro wrote:
> > On Sat, Jun 01, 2019 at 10:15:26AM +0800, Gen Zhang wrote:
> > > In selinux_sb_eat_lsm_opts(), 'arg' is allocated by kmemdup_nul(). It
> > > returns NULL when fails. So 'arg' should be checked. And 'mnt_opts'
> > > should be freed when error.
> > What's the latter one for? On failure we'll get to put_fs_context()
> > pretty soon, so
> > security_free_mnt_opts(&fc->security);
> > will be called just fine. Leaving it allocated on failure is fine...
> Actually, right now in mainline it is not (btrfs_mount_root() has
> an odd call of security_sb_eat_lsm_opts()); eventually we will be
> down to just the callers in ->parse_monolithic() instances, at which
> point the above will become correct. At the moment it is not, so
> consider the objection withdrawn (and I really need to get some sleep,
> seeing how long did it take me to recall the context... ;-/)
Thanks for your comments. And have a good dream.