Message ID | 8f1c9645-26ac-e10b-5321-213034c36d70@gmail.com |
---|---|
State | New |
Headers | show |
On 31/01/18 00:35, Jun Zhao wrote: > Now FFmpeg-VAAPI work with iHD open source driver > (https://github.com/intel/media-driver) initiatory. > > From 069611945880643aad6ffb5ea70b732f40bbb510 Mon Sep 17 00:00:00 2001 > From: Jun Zhao <jun.zhao@intel.com> > Date: Wed, 31 Jan 2018 08:21:30 +0800 > Subject: [PATCH] hwcontext_vaapi: add iHD open source driver. > > Signed-off-by: Jun Zhao <jun.zhao@intel.com> > --- > libavutil/hwcontext_vaapi.c | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/libavutil/hwcontext_vaapi.c b/libavutil/hwcontext_vaapi.c > index 29698d1b27..34c5a2c264 100644 > --- a/libavutil/hwcontext_vaapi.c > +++ b/libavutil/hwcontext_vaapi.c > @@ -285,6 +285,10 @@ static const struct { > "ubit", > AV_VAAPI_DRIVER_QUIRK_ATTRIB_MEMTYPE, > }, > + { > + "Intel iHD open source driver", > + "Intel iHD driver", > + }, > { > "VDPAU wrapper", > "Splitted-Desktop Systems VDPAU backend for VA-API", > -- > 2.14.1 > Why? This is the "list of shame" for drivers with bad or incomplete implementations which need explicit workarounds. Since the iHD driver is new (in some sense), hopefully it will be a correct implementation which does not require any of these. - Mark
On 2018/1/31 19:13, Mark Thompson wrote: > On 31/01/18 00:35, Jun Zhao wrote: >> Now FFmpeg-VAAPI work with iHD open source driver >> (https://github.com/intel/media-driver) initiatory. >> >> From 069611945880643aad6ffb5ea70b732f40bbb510 Mon Sep 17 00:00:00 2001 >> From: Jun Zhao <jun.zhao@intel.com> >> Date: Wed, 31 Jan 2018 08:21:30 +0800 >> Subject: [PATCH] hwcontext_vaapi: add iHD open source driver. >> >> Signed-off-by: Jun Zhao <jun.zhao@intel.com> >> --- >> libavutil/hwcontext_vaapi.c | 4 ++++ >> 1 file changed, 4 insertions(+) >> >> diff --git a/libavutil/hwcontext_vaapi.c b/libavutil/hwcontext_vaapi.c >> index 29698d1b27..34c5a2c264 100644 >> --- a/libavutil/hwcontext_vaapi.c >> +++ b/libavutil/hwcontext_vaapi.c >> @@ -285,6 +285,10 @@ static const struct { >> "ubit", >> AV_VAAPI_DRIVER_QUIRK_ATTRIB_MEMTYPE, >> }, >> + { >> + "Intel iHD open source driver", >> + "Intel iHD driver", >> + }, >> { >> "VDPAU wrapper", >> "Splitted-Desktop Systems VDPAU backend for VA-API", >> -- >> 2.14.1 >> > Why? This is the "list of shame" for drivers with bad or incomplete implementations which need explicit workarounds. Since the iHD driver is new (in some sense), hopefully it will be a correct implementation which does not require any of these. > > - Mark I leave the driver_quick with zero and just want to give a log message to identify FFmpeg VA-API can co-work with iHD open source driver, so if iHD fix the related issue, you don't want to give any log message in FFmpeg VA-API? > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
On 2018/2/1 14:45, Jun Zhao wrote: > > On 2018/1/31 19:13, Mark Thompson wrote: >> On 31/01/18 00:35, Jun Zhao wrote: >>> Now FFmpeg-VAAPI work with iHD open source driver >>> (https://github.com/intel/media-driver) initiatory. >>> >>> From 069611945880643aad6ffb5ea70b732f40bbb510 Mon Sep 17 00:00:00 2001 >>> From: Jun Zhao <jun.zhao@intel.com> >>> Date: Wed, 31 Jan 2018 08:21:30 +0800 >>> Subject: [PATCH] hwcontext_vaapi: add iHD open source driver. >>> >>> Signed-off-by: Jun Zhao <jun.zhao@intel.com> >>> --- >>> libavutil/hwcontext_vaapi.c | 4 ++++ >>> 1 file changed, 4 insertions(+) >>> >>> diff --git a/libavutil/hwcontext_vaapi.c b/libavutil/hwcontext_vaapi.c >>> index 29698d1b27..34c5a2c264 100644 >>> --- a/libavutil/hwcontext_vaapi.c >>> +++ b/libavutil/hwcontext_vaapi.c >>> @@ -285,6 +285,10 @@ static const struct { >>> "ubit", >>> AV_VAAPI_DRIVER_QUIRK_ATTRIB_MEMTYPE, >>> }, >>> + { >>> + "Intel iHD open source driver", >>> + "Intel iHD driver", >>> + }, >>> { >>> "VDPAU wrapper", >>> "Splitted-Desktop Systems VDPAU backend for VA-API", >>> -- >>> 2.14.1 >>> >> Why? This is the "list of shame" for drivers with bad or incomplete implementations which need explicit workarounds. Since the iHD driver is new (in some sense), hopefully it will be a correct implementation which does not require any of these. >> >> - Mark > I leave the driver_quick with zero and just want to give a log message to > identify FFmpeg VA-API can co-work with iHD open source driver, so if > iHD fix the related issue, you don't want to give any log message in > FFmpeg VA-API? Forget this commits, the code have handle this case, my fault. >> _______________________________________________ >> ffmpeg-devel mailing list >> ffmpeg-devel@ffmpeg.org >> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
diff --git a/libavutil/hwcontext_vaapi.c b/libavutil/hwcontext_vaapi.c index 29698d1b27..34c5a2c264 100644 --- a/libavutil/hwcontext_vaapi.c +++ b/libavutil/hwcontext_vaapi.c @@ -285,6 +285,10 @@ static const struct { "ubit", AV_VAAPI_DRIVER_QUIRK_ATTRIB_MEMTYPE, }, + { + "Intel iHD open source driver", + "Intel iHD driver", + }, { "VDPAU wrapper", "Splitted-Desktop Systems VDPAU backend for VA-API",