linux-kernel - Re: [PATCH v7 4/6] drm: Refuse to async flip with atomic prop changes

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: <OmiJ2t3KUJGjN5EM8mo1cX52UpRgtMZx0oT11igYlRjfeqXtkeQlFdV3OdHO3eA4KkyWKuvSSVX8Ico1zFkioaQQObjLgp0dTu15YE-RWc8=@emersion.fr>
Date:   Mon, 23 Oct 2023 08:44:46 +0000
From:   Simon Ser <contact@...rsion.fr>
To:     Michel Dänzer <michel.daenzer@...lbox.org>
Cc:     pierre-eric.pelloux-prayer@....com,
        André Almeida <andrealmeid@...lia.com>,
        'Marek Olšák' <maraeo@...il.com>,
        Randy Dunlap <rdunlap@...radead.org>,
        linux-kernel@...r.kernel.org, amd-gfx@...ts.freedesktop.org,
        Pekka Paalanen <ppaalanen@...il.com>,
        dri-devel@...ts.freedesktop.org, kernel-dev@...lia.com,
        alexander.deucher@....com, hwentlan@....com,
        christian.koenig@....com, joshua@...ggi.es
Subject: Re: [PATCH v7 4/6] drm: Refuse to async flip with atomic prop changes

On Monday, October 23rd, 2023 at 10:42, Michel Dänzer <michel.daenzer@...lbox.org> wrote:

> On 10/23/23 10:27, Simon Ser wrote:
> 
> > On Sunday, October 22nd, 2023 at 12:12, Michel Dänzer michel.daenzer@...lbox.org wrote:
> > 
> > > On 10/17/23 14:16, Simon Ser wrote:
> > > 
> > > > After discussing with André it seems like we missed a plane type check
> > > > here. We need to make sure FB_ID changes are only allowed on primary
> > > > planes.
> > > 
> > > Can you elaborate why that's needed?
> > 
> > Current drivers are in general not prepared to perform async page-flips
> > on planes other than primary. For instance I don't think i915 has logic
> > to perform async page-flip on an overlay plane FB_ID change.
> 
> 
> That should be handled in the driver's atomic_check then?
> 
> Async flips of overlay planes would be useful e.g. for presenting a windowed application with tearing, while the rest of the desktop is tear-free.

Yes, that would be useful, but requires more work. Small steps: first
expose what the legacy uAPI can do in atomic, then later extend that in
some drivers.