linux-kernel - Re: [PATCH 5.17 404/772] media: rkvdec: Stop overclocking the decoder

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:   Tue, 7 Jun 2022 14:20:51 -0700
From:   Brian Norris <briannorris@...omium.org>
To:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Cc:     linux-kernel@...r.kernel.org, stable@...r.kernel.org,
        Nicolas Dufresne <nicolas.dufresne@...labora.com>,
        Sebastian Fricke <sebastian.fricke@...labora.com>,
        Ezequiel Garcia <ezequiel@...guardiasur.com.ar>,
        Hans Verkuil <hverkuil-cisco@...all.nl>,
        Mauro Carvalho Chehab <mchehab@...nel.org>,
        Sasha Levin <sashal@...nel.org>
Subject: Re: [PATCH 5.17 404/772] media: rkvdec: Stop overclocking the decoder

Hi,

On Tue, Jun 07, 2022 at 06:59:56PM +0200, Greg Kroah-Hartman wrote:
> From: Nicolas Dufresne <nicolas.dufresne@...labora.com>
> 
> [ Upstream commit 9998943f6dfc5d5472bfab2e38527fb6ba5e9da7 ]
> 
> While this overclock hack seems to work on some implementations
> (some ChromeBooks, RockPi4) it also causes instability on other
> implementations (notably LibreComputer Renegade, but there were more
> reports in the LibreELEC project, where this has been removed). While
> performance is indeed affected (tested with GStreamer), 4K playback
> still works as long as you don't operate in lock step and keep at
> least 1 frame ahead of time in the decode queue.
> 
> After discussion with ChromeOS members, it would seem that their
> implementation indeed used to synchronously decode each frame, so
> this hack was simply compensating for their code being less
> efficient. In my opinion, this hack should not have been included
> upstream.
> 
> Fixes: cd33c830448ba ("media: rkvdec: Add the rkvdec driver")
> Signed-off-by: Nicolas Dufresne <nicolas.dufresne@...labora.com>
> Reviewed-by: Sebastian Fricke <sebastian.fricke@...labora.com>
> Reviewed-by: Ezequiel Garcia <ezequiel@...guardiasur.com.ar>
> Signed-off-by: Hans Verkuil <hverkuil-cisco@...all.nl>
> Signed-off-by: Mauro Carvalho Chehab <mchehab@...nel.org>
> Signed-off-by: Sasha Levin <sashal@...nel.org>

FWIW, I've noticed a problem that is uncovered by this patch, because
the default clock rate is not currently acceptable all the time. See my
fix here:

https://lore.kernel.org/all/20220607141535.1.Idafe043ffc94756a69426ec68872db0645c5d6e2@changeid/
[PATCH] arm64: dts: rockchip: Assign RK3399 VDU clock rate

It might be nice if $subject patch could be delayed until the fix is in
too. The 5.19 cycle is only in -rc1, after all.

(The same seems to apply for the 5.{18,15,10}.y series too.)

Brian