linux-kernel - Re: [External] Re: [PATCH V3 2/3] mm: add max swappiness arg to lru

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]

Message-ID: <CACSyD1M8EZ-qEhSGsLTAs+i+YMGPMw-+39Es1SLg-OxEbgg9yA@mail.gmail.com>
Date: Thu, 1 May 2025 09:56:57 +0800
From: Zhongkun He <hezhongkun.hzk@...edance.com>
To: Dan Carpenter <dan.carpenter@...aro.org>
Cc: akpm@...ux-foundation.org, hannes@...xchg.org, mhocko@...e.com, 
	yosry.ahmed@...ux.dev, muchun.song@...ux.dev, yuzhao@...gle.com, 
	linux-mm@...ck.org, linux-kernel@...r.kernel.org
Subject: Re: [External] Re: [PATCH V3 2/3] mm: add max swappiness arg to
 lru_gen for anonymous memory only

Hi Dan

On Wed, Apr 30, 2025 at 3:59 PM Dan Carpenter <dan.carpenter@...aro.org> wrote:
>
> On Wed, Apr 09, 2025 at 03:06:19PM +0800, Zhongkun He wrote:
> > +             /* set by userspace for anonymous memory only */
> > +             if (!strncmp("max", swap_string, sizeof("max"))) {
>
> This pattern of strncmp("foo", str, sizeof("foo")) is exactly the same
> as strcmp().  It doesn't provide any additional security.  The strncmp()
> function is meant for matching string prefixes and it's a relatively
> common bug to do this:
>
> intended: if (strcmp(string, "prefix", sizeof("prefix") - 1) == 0) {
>   actual: if (strcmp(string, "prefix", sizeof("prefix")) == 0) {
>

Yes, I understand the difference.

> I have a static checker warning for these:
> https://lore.kernel.org/all/30210ed77b40b4b6629de659cb56b9ec7832c447.1744452787.git.dan.carpenter@linaro.org/
>
> If people deliberately misuse the function then it makes it trickier
> to tell accidental mistakes from deliberate mistakes.
>

if (!strncmp("max", swap_string, sizeof("max"))) {

The length of swap_string is 5 because it's read using sscanf, which
will add the null terminator \0
at the end of the string. If we input max into the interface,
swap_string will contain max\0, which is
equivalent to the string "max". Since we only need to compare the
first few characters(There are other
possible inputs as well.) — effectively treating it as a prefix match
— I used strncmp.

Thank you for the reminder， Dan.

> regards,
> dan carpenter
>