linux-kernel - Re: [PATCH] fix: Prevent memory leak by checking for NULL buffer before calling css

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]

Date: Mon, 3 Jun 2024 10:42:07 -0700
From: Yosry Ahmed <yosryahmed@...gle.com>
To: Axel Rasmussen <axelrasmussen@...gle.com>
Cc: Geunsik Lim <geunsik.lim@...il.com>, Vlastimil Babka <vbabka@...e.cz>, Ingo Molnar <mingo@...e.hu>, 
	Andrew Morton <akpm@...ux-foundation.org>, Thomas Gleixner <tglx@...utronix.de>, 
	Peter Zijlstra <a.p.zijlstra@...llo.nl>, Hugh Dickins <hughd@...gle.com>, 
	"H. Peter Anvin" <hpa@...or.com>, Steven Rostedt <rostedt@...dmis.org>, 
	linux-kernel <linux-kernel@...r.kernel.org>, linux-mm <linux-mm@...ck.org>
Subject: Re: [PATCH] fix: Prevent memory leak by checking for NULL buffer
 before calling css_put()

On Mon, Jun 3, 2024 at 10:32 AM Axel Rasmussen <axelrasmussen@...gle.com> wrote:
>
> On Mon, Jun 3, 2024 at 5:33 AM Geunsik Lim <geunsik.lim@...il.com> wrote:
> >
> > This commit addresses a potential memory leak in the
> > `get_mm_memcg_path()` function
> > by explicitly checking if the allocated buffer (`buf`) is NULL before
> > calling the
> > `css_put()` function. The prefix 'css' means abbreviation of cgroup_subsys_state
> >
> > Previously, the code would directly call `css_put()` without checking
> > the value of
> > `buf`, which could lead to a memory leak if the buffer allocation failed.
> > This commit introduces a conditional check to ensure that `css_put()`
> > is only called
> > if `buf` is not NULL.
> >
> > This change enhances the code's robustness and prevents memory leaks, improving
> > overall system stability.
> >
> > **Specific Changes:**
> >
> > * In the `out_put` label, an `if` statement is added to check
> >   if `buf` is not NULL before calling `css_put()`.
> >
> > **Benefits:**
> >
> > * Prevents potential memory leaks
> > * Enhances code robustness
> > * Improves system stability
> >
> > Signed-off-by: Geunsik Lim <leemgs@...il.com>
> > Signed-off-by: Geunsik Lim <geunsik.lim@...sung.com>
> > ---
> >  mm/mmap_lock.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/mm/mmap_lock.c b/mm/mmap_lock.c
> > index 1854850b4b89..7314045b0e3b 100644
> > --- a/mm/mmap_lock.c
> > +++ b/mm/mmap_lock.c
> > @@ -213,7 +213,8 @@ static const char *get_mm_memcg_path(struct mm_struct *mm)
> >         cgroup_path(memcg->css.cgroup, buf, MEMCG_PATH_BUF_SIZE);
> >
> >  out_put:
> > -       css_put(&memcg->css);
> > +        if (buf != NULL)
> > +                css_put(&memcg->css);
> >  out:
> >         return buf;
> >  }
>
> I think the existing code is correct, and this change actually
> introduces a memory leak where there was none before.
>
> In the case where get_memcg_path_buf() returns NULL, we *still* need
> to css_put() what we got from get_mem_cgroup_from_mm() before.
>
> NAK, unless I'm missing something.

+1

We already skip css_put() if get_mem_cgroup_from_mm() returns NULL.
Whether or not get_memcg_path_buf() succeeds is irrelevant here.