Re: [PATCH] user_ns: use correct check for single-threadedness

[Kees Cook]

On Tue, Jul 28, 2015 at 2:35 PM, Andrew Morton
<akpm@xxxxxxxxxxxxxxxxxxxx> wrote:
> On Tue, 28 Jul 2015 10:15:00 -0700 Kees Cook <keescook@xxxxxxxxxxxx> wrote:
>
>> From: Ricky Zhou <rickyz@xxxxxxxxxxxx>
>>
>> Checking mm_users > 1 does not mean a process is multithreaded. For
>> example, reading /proc/PID/maps temporarily increments mm_users, allowing
>> other processes to (accidentally) interfere with unshare() calls.
>>
>> This fixes observed failures of unshare(CLONE_NEWUSER) incorrectly
>> returning EINVAL if another processes happened to be simultaneously
>> reading the maps file.
>
> Yikes.  current_is_single_threaded() is expensive.  Are we sure this
> isn't going to kill someone's workload?

It _can_ be expensive, but if mm->mm_users == 1 it immediately returns
true, so it's only the cases where there is a race (like what's solved
here), or when it's a legit failure. This doesn't feel to me like it
should hit a real user very hard, since "real" callers of unshare will
normally have mm_users == 1.

-Kees

-- 
Kees Cook
Chrome OS Security
--
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/