linux-kernel - Re: [PATCH v3] drm/framebuffer: Acquire internal references on GEM handles

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>] [day] [month] [year] [list]

Message-ID: <a421623f-e93f-4423-bf30-c7388c6a3508@suse.de>
Date: Tue, 8 Jul 2025 09:38:52 +0200
From: Thomas Zimmermann <tzimmermann@...e.de>
To: Satadru Pramanik <satadru@...il.com>
Cc: Christian König <christian.koenig@....com>,
 asrivats@...hat.com, maarten.lankhorst@...ux.intel.com, mripard@...nel.org,
 airlied@...il.com, simona@...ll.ch, jean-christophe@...llain.net,
 superm1@...nel.org, bp@...en8.de, dri-devel@...ts.freedesktop.org,
 linux-kernel@...r.kernel.org, Bert Karwatzki <spasswolf@....de>,
 Sumit Semwal <sumit.semwal@...aro.org>, linux-media@...r.kernel.org,
 linaro-mm-sig@...ts.linaro.org, stable@...r.kernel.org
Subject: Re: [PATCH v3] drm/framebuffer: Acquire internal references on GEM
 handles

Hi

Am 07.07.25 um 18:14 schrieb Satadru Pramanik:
> Applying this patch to 6.16-rc5 resolves the sleep issue regression 
> from 6.16-rc4 I was having on my MacBookPro11,3 (Mid-2014 15" 
> MacBookPro), which has the NVIDIA GK107M GPU enabled via the Nouveau 
> driver.

Thanks for testing. I think the sleep regression was just a side effect 
of the broken reference counting.

Best regards
Thomas

>
> Many thanks,
>
> Satadru
>
> On Mon, Jul 7, 2025 at 9:33 AM Thomas Zimmermann <tzimmermann@...e.de> 
> wrote:
>
>     Hi
>
>     Am 07.07.25 um 15:21 schrieb Christian König:
>
>     >>
>     >> +#define DRM_FRAMEBUFFER_HAS_HANDLE_REF(_i)  BIT(0u + (_i))
>     > Why the "0u + (_i)" here? An macro trick?
>
>     You mean why not just BIT(_i)? internal_flags could possibly contain
>     additional flags. Just using BIT(_i) would make it look as if it's
>     only
>     for those handle refs.
>
>     Best regards
>     Thomas
>
>     >
>     > Regards,
>     > Christian.
>     >
>     >> +
>     >>   /**
>     >>    * struct drm_framebuffer - frame buffer object
>     >>    *
>     >> @@ -188,6 +191,10 @@ struct drm_framebuffer {
>     >>       * DRM_MODE_FB_MODIFIERS.
>     >>       */
>     >>      int flags;
>     >> +    /**
>     >> +     * @internal_flags: Framebuffer flags like
>     DRM_FRAMEBUFFER_HAS_HANDLE_REF.
>     >> +     */
>     >> +    unsigned int internal_flags;
>     >>      /**
>     >>       * @filp_head: Placed on &drm_file.fbs, protected by
>     &drm_file.fbs_lock.
>     >>       */
>
>     -- 
>     --
>     Thomas Zimmermann
>     Graphics Driver Developer
>     SUSE Software Solutions Germany GmbH
>     Frankenstrasse 146, 90461 Nuernberg, Germany
>     GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman
>     HRB 36809 (AG Nuernberg)
>

-- 
--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Frankenstrasse 146, 90461 Nuernberg, Germany
GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman
HRB 36809 (AG Nuernberg)