linux-kernel - Re: [PATCH v2 061/101] fbdev/ps3fb: Duplicate video-mode option string

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]

Date:   Fri, 10 Mar 2023 13:21:51 +0100
From:   Thomas Zimmermann <tzimmermann@...e.de>
To:     Geert Uytterhoeven <geert@...ux-m68k.org>
Cc:     deller@....de, timur@...nel.org, rdunlap@...radead.org,
        paulus@...ba.org, benh@...nel.crashing.org, linux@...linux.org.uk,
        pjones@...hat.com, adaplas@...il.com, s.hauer@...gutronix.de,
        shawnguo@...nel.org, mbroemme@...mpq.org, thomas@...ischhofer.net,
        James.Bottomley@...senpartnership.com, sudipm.mukherjee@...il.com,
        teddy.wang@...iconmotion.com, corbet@....net,
        linux-fbdev@...r.kernel.org, dri-devel@...ts.freedesktop.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 061/101] fbdev/ps3fb: Duplicate video-mode option
 string

Hi Geert

Am 10.03.23 um 09:18 schrieb Geert Uytterhoeven:
> Hi Thomas,
> 
> On Thu, Mar 9, 2023 at 5:02 PM Thomas Zimmermann <tzimmermann@...e.de> wrote:
>> Assume that the driver does not own the option string or its substrings
>> and hence duplicate the option string for the video mode. Allocate the
>> copy's memory with kstrdup() and free it in the module's exit function.
>>
>> Done in preparation of switching the driver to struct option_iter and
>> constifying the option string.
>>
>> v2:
>>          * replace static memory with kstrdup()/kfree() (Geert)
>>
>> Signed-off-by: Thomas Zimmermann <tzimmermann@...e.de>
> 
> Thanks for the upodate!
> 
>> --- a/drivers/video/fbdev/ps3fb.c
>> +++ b/drivers/video/fbdev/ps3fb.c
>> @@ -260,6 +260,7 @@ static const struct fb_videomode ps3fb_modedb[] = {
>>   static int ps3fb_mode;
>>   module_param(ps3fb_mode, int, 0);
>>
>> +static char *mode_option_buf;
> 
> Do you really need this variable? It contains the same value
> as mode_option below.
> This is a common pattern in several patches.

In most cases, the extra variable is necessary. mode_option, or a 
similar variable, is often also used to store module parameters or is 
initialized from a static string. In those cases we obviously cannot 
free the memory. Hence the additional variable for the freeing the 
allocation.

I reviewed the patches and it looks like kyrofb, ps3fb, and sm712fb 
might be able to work without the _buf variable. I'll try to update those.

Best regards
Thomas

> 
>>   static char *mode_option;
>>
>>   static int ps3fb_cmp_mode(const struct fb_videomode *vmode,
>> @@ -1276,8 +1277,11 @@ static int __init ps3fb_setup(void)
>>                          continue;
>>                  if (!strncmp(this_opt, "mode:", 5))
>>                          ps3fb_mode = simple_strtoul(this_opt + 5, NULL, 0);
>> -               else
>> -                       mode_option = this_opt;
>> +               else {
>> +                       kfree(mode_option_buf);
>> +                       mode_option_buf = kstrdup(this_opt, GFP_KERNEL); // ignore errors
>> +                       mode_option = mode_option_buf;
>> +               }
>>          }
>>          return 0;
>>   }
>> @@ -1294,6 +1298,7 @@ static void __exit ps3fb_exit(void)
>>   {
>>          pr_debug(" -> %s:%d\n", __func__, __LINE__);
>>          ps3_system_bus_driver_unregister(&ps3fb_driver);
>> +       kfree(mode_option_buf);
>>          pr_debug(" <- %s:%d\n", __func__, __LINE__);
>>   }
> 
> Gr{oetje,eeting}s,
> 
>                          Geert
> 

-- 
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Maxfeldstr. 5, 90409 Nürnberg, Germany
(HRB 36809, AG Nürnberg)
Geschäftsführer: Ivo Totev

Download attachment "OpenPGP_signature" of type "application/pgp-signature" (841 bytes)