[FFmpeg-devel] [PATCH] avcodec: only warn about hwaccel with frame threads

Hendrik Leppkes h.leppkes at gmail.com
Sat Jan 23 17:14:25 CET 2016


On Sat, Jan 23, 2016 at 4:36 PM, Ronald S. Bultje <rsbultje at gmail.com> wrote:
> Hi,
>
> On Sat, Jan 23, 2016 at 9:52 AM, Andreas Cadhalpun <
> andreas.cadhalpun at googlemail.com> wrote:
>
>> On 23.01.2016 15:10, Hendrik Leppkes wrote:
>> > On Sat, Jan 23, 2016 at 2:45 PM, Ronald S. Bultje <rsbultje at gmail.com>
>> wrote:
>> >> Both of you keep shouting from one side of the room to the other without
>> >> trying to actually converge to a point that might somehow be mutually
>> >> acceptable. I'm getting very grumbly here.
>> >>
>> >
>> > I'm not trying to be unreasonable here, all I'm asking for at this
>> > point is to make sure the new vp9 hwaccel doesn't break spectacularly
>> > when MT is turned on before its allowed back.
>>
>> OK, that would be fine with me.
>>
>> Do you agree that the attached patch is sufficient for this?
>
>
> I'm not against this patch, but I repeatedly hear claims that it might
> crash in other codecs also. I also heard some claim in this thread that VLC
> works around that with extensive locking.
>
> How true is all of that? Does that mean non-VLC applications will still
> crash with other codecs?

Those crashes are rare, but yes they could happen. Due to the
threading nature its unfortunately not quite predictable, and also
depends on the driver.
I've had the most issues with Intel drivers on Windows.

>
> Also, this patch disables hwaccel if threading is enabled for vp9 (i.e. by
> default). From a power usage perspective, the inverse probably makes more
> sense. Disabling ffvp9hw by default seems like a bad idea from a user
> friendliness perspective.
>
> Ronald
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel


More information about the ffmpeg-devel mailing list