linux-kernel - Re: [PATCH] drm/kmb: fix dereference before NULL check in kmb_plane_atomic

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: <fef74b28-80f1-9184-efa5-25f3343ace40@suse.de>
Date:   Fri, 29 Jul 2022 16:25:46 +0200
From:   Thomas Zimmermann <tzimmermann@...e.de>
To:     Zeng Jingxiang <zengjx95@...il.com>, anitha.chrisanthus@...el.com,
        edmund.j.dea@...el.com, airlied@...ux.ie, daniel@...ll.ch,
        laurent.pinchart@...asonboard.com, maxime@...no.tech,
        ville.syrjala@...ux.intel.com
Cc:     Zeng Jingxiang <linuszeng@...cent.com>,
        linux-kernel@...r.kernel.org, dri-devel@...ts.freedesktop.org
Subject: Re: [PATCH] drm/kmb: fix dereference before NULL check in
 kmb_plane_atomic_update()

Hi

Am 29.07.22 um 16:23 schrieb Thomas Zimmermann:
> Hi
> 
> Am 29.07.22 um 05:07 schrieb Zeng Jingxiang:
>> From: Zeng Jingxiang <linuszeng@...cent.com>
>>
>> The "plane" pointer was access before checking if it was NULL.
>>
>> The drm_atomic_get_old_plane_state() function will dereference
>> the pointer "plane".
>> 345    struct drm_plane_state *old_plane_state =
>>         drm_atomic_get_old_plane_state(state, plane);
>> 346    struct drm_plane_state *new_plane_state =
>>         drm_atomic_get_new_plane_state(state, plane);
>>
>> A NULL check for "plane" indicates that it may be NULL
>> 363    if (!plane || !new_plane_state || !old_plane_state)
> 
> Is this an actual bug that happens?
> 
> All planes should always have a state. Therefore the tests for 
> !new_plane_state and !old_plane_state can be removed, I'd say.

Just to clarify: moving the test for !plane before using one of the 
get_plane_state functions is a correct bugfix.

Best regards
Thomas

> 
> Best regards
> Thomas
> 
>>
>> Fixes: 977697e20b3d ("drm/atomic: Pass the full state to planes atomic 
>> disable and update")
>> Fixes: 37418bf14c13 ("drm: Use state helper instead of the plane state 
>> pointer")
>> Signed-off-by: Zeng Jingxiang <linuszeng@...cent.com>
>> ---
>>   drivers/gpu/drm/kmb/kmb_plane.c | 13 ++++++++-----
>>   1 file changed, 8 insertions(+), 5 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/kmb/kmb_plane.c 
>> b/drivers/gpu/drm/kmb/kmb_plane.c
>> index 2735b8eb3537..d2bc998b65ce 100644
>> --- a/drivers/gpu/drm/kmb/kmb_plane.c
>> +++ b/drivers/gpu/drm/kmb/kmb_plane.c
>> @@ -342,10 +342,7 @@ static void kmb_plane_set_alpha(struct 
>> kmb_drm_private *kmb,
>>   static void kmb_plane_atomic_update(struct drm_plane *plane,
>>                       struct drm_atomic_state *state)
>>   {
>> -    struct drm_plane_state *old_plane_state = 
>> drm_atomic_get_old_plane_state(state,
>> -                                         plane);
>> -    struct drm_plane_state *new_plane_state = 
>> drm_atomic_get_new_plane_state(state,
>> -                                         plane);
>> +    struct drm_plane_state *old_plane_state, *new_plane_state;
>>       struct drm_framebuffer *fb;
>>       struct kmb_drm_private *kmb;
>>       unsigned int width;
>> @@ -360,7 +357,13 @@ static void kmb_plane_atomic_update(struct 
>> drm_plane *plane,
>>       static dma_addr_t addr[MAX_SUB_PLANES];
>>       struct disp_cfg *init_disp_cfg;
>> -    if (!plane || !new_plane_state || !old_plane_state)
>> +    if (!plane)
>> +        return;
>> +
>> +    old_plane_state = drm_atomic_get_old_plane_state(state, plane);
>> +    new_plane_state = drm_atomic_get_new_plane_state(state, plane);
>> +
>> +    if (!new_plane_state || !old_plane_state)
>>           return;
>>       fb = new_plane_state->fb;
> 

-- 
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)