linux-kernel - Re: [PATCH] fork: fix incorrect fput of ->exe

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

Message-ID: <20170824165935.GA21624@gmail.com>
Date:   Thu, 24 Aug 2017 09:59:35 -0700
From:   Eric Biggers <ebiggers3@...il.com>
To:     Oleg Nesterov <oleg@...hat.com>
Cc:     linux-mm@...ck.org, Andrew Morton <akpm@...ux-foundation.org>,
        linux-kernel@...r.kernel.org, Dmitry Vyukov <dvyukov@...gle.com>,
        Ingo Molnar <mingo@...nel.org>,
        Konstantin Khlebnikov <koct9i@...il.com>,
        Michal Hocko <mhocko@...e.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Vlastimil Babka <vbabka@...e.cz>, stable@...r.kernel.org,
        Eric Biggers <ebiggers@...gle.com>
Subject: Re: [PATCH] fork: fix incorrect fput of ->exe_file causing
 use-after-free

On Thu, Aug 24, 2017 at 03:20:41PM +0200, Oleg Nesterov wrote:
> On 08/23, Eric Biggers wrote:
> >
> > From: Eric Biggers <ebiggers@...gle.com>
> >
> > Commit 7c051267931a ("mm, fork: make dup_mmap wait for mmap_sem for
> > write killable") made it possible to kill a forking task while it is
> > waiting to acquire its ->mmap_sem for write, in dup_mmap().  However, it
> > was overlooked that this introduced an new error path before a reference
> > is taken on the mm_struct's ->exe_file.
> 
> Hmm. Unless I am totally confused, the same problem with mm->exol_area?
> I'll recheck....

I'm not sure what you mean by ->exol_area.

> 
> > --- a/kernel/fork.c
> > +++ b/kernel/fork.c
> > @@ -806,6 +806,7 @@ static struct mm_struct *mm_init(struct mm_struct *mm, struct task_struct *p,
> >  	mm_init_cpumask(mm);
> >  	mm_init_aio(mm);
> >  	mm_init_owner(mm, p);
> > +	RCU_INIT_POINTER(mm->exe_file, NULL);
> 
> Can't we simply move
> 
> 	RCU_INIT_POINTER(mm->exe_file, get_mm_exe_file(oldmm));
> 
> from dup_mmap() here? Afaics this doesn't need mmap_sem.
> 

Two problems, even assuming that get_mm_exe_file() doesn't require mmap_sem:

- If mm_alloc_pgd() or init_new_context() in mm_init() fails, mm_init() doesn't
  do the full mmput(), so the file reference would not be dropped.  So it would
  need to be changed to drop the file reference too.

- The file would also be set when called from mm_alloc() which is used when
  exec'ing a new task.  *Maybe* it would be safe to do temporarily, but it's
  pointless because ->exe_file will be set later by flush_old_exec().

Eric