Re: [libav-devel] [PATCH] avconv: add avconv_d3d11va

Just a few cursory remarks as I only have a short moment right now. On Tue, Dec 13, 2016 at 02:19:03PM +0100, Steve Lhomme wrote: > --- /dev/null > +++ b/avconv_d3d11va.c > @@ -0,0 +1,410 @@ > + > +/* define all the GUIDs used directly here, > + to avoid problems with inconsistent dxva2api.h

Re: [libav-devel] [PATCH] configure: use -O1 with MSVC by default

On Tue, Dec 13, 2016 at 04:23:02PM +0100, Steve Lhomme wrote: > On Tue, Dec 13, 2016 at 3:33 PM, Hendrik Leppkes wrote: > > On Tue, Dec 13, 2016 at 2:24 PM, Steve Lhomme wrote: > >> From: Steve Lhomme > >> > >> Otherwise some ARM and

Re: [libav-devel] [PATCH] avconv: add avconv_d3d11va

On 13/12/2016 14:19, Steve Lhomme wrote: > +DEFINE_GUID(DXVA2_ModeMPEG2_VLD, 0xee27417f, > 0x5e28,0x4e65,0xbe,0xea,0x1d,0x26,0xb5,0x08,0xad,0xc9); > +DEFINE_GUID(DXVA2_ModeMPEG2and1_VLD, 0x86695f12, > 0x340e,0x4f04,0x9f,0xd3,0x92,0x53,0xdd,0x32,0x74,0x60); > +DEFINE_GUID(DXVA2_ModeH264_E,

Re: [libav-devel] [PATCH] configure: use -O1 with MSVC by default

On Tue, Dec 13, 2016 at 10:23 AM, Steve Lhomme wrote: > On Tue, Dec 13, 2016 at 3:33 PM, Hendrik Leppkes wrote: >> On Tue, Dec 13, 2016 at 2:24 PM, Steve Lhomme wrote: >>> From: Steve Lhomme >>> >>> Otherwise some ARM

Re: [libav-devel] [RFC] [PATCH 2/2] ppc: Merge types_altivec.h into util_altivec.h

On Tue, Dec 13, 2016 at 8:02 AM, Diego Biurrun wrote: > There is no point in keeping the two separate. > --- > > Maybe there should be a better name than util_altivec.h. altivecutils.h? -- Vittorio ___ libav-devel mailing list

Re: [libav-devel] [PATCH 2/2] vaapi_h264: Fix POC on IDR frames

On 12/12/2016 22:30, Mark Thompson wrote: > From: Jun Zhao > > In H.264 section 8.2.1, we have that "The bitstream shall not contain > data that result in Min(TopFieldOrderCnt, BottomFieldOrderCnt) not > equal to 0 for a coded IDR frame". This fixes the encoder to always >

Re: [libav-devel] [RFC] [PATCH 1/2] ppc: Centralize compiler-specific altivec.h #include handling in one place

On 13/12/2016 14:02, Diego Biurrun wrote: > Also move #includes into canonical order where appropriate. > --- > > It seems saner not to spread this ugliness all over the place. > Set ok if it does build. lu ___ libav-devel mailing list

Re: [libav-devel] [RFC] [PATCH 2/2] ppc: Merge types_altivec.h into util_altivec.h

On 13/12/2016 14:02, Diego Biurrun wrote: > +/*** > + * Null vector > + **/ > +#define LOAD_ZERO const vec_u8 zerov = vec_splat_u8( 0 ) > + > +#define zero_u8v

Re: [libav-devel] [PATCH] configure: use -O1 with MSVC by default

On Tue, Dec 13, 2016 at 3:33 PM, Hendrik Leppkes wrote: > On Tue, Dec 13, 2016 at 2:24 PM, Steve Lhomme wrote: >> From: Steve Lhomme >> >> Otherwise some ARM and other unsupported CPU/OS is linked with -O0 >> --- >> configure | 1 + >> 1

Re: [libav-devel] [PATCH] configure: use -O1 with MSVC by default

On Tue, Dec 13, 2016 at 2:24 PM, Steve Lhomme wrote: > From: Steve Lhomme > > Otherwise some ARM and other unsupported CPU/OS is linked with -O0 > --- > configure | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/configure b/configure > index

Re: [libav-devel] [PATCH] avconv: add avconv_d3d11va

On Tue, 13 Dec 2016 14:19:03 +0100 Steve Lhomme wrote: > From: Steve Lhomme > > The code is similar to avconv_dxva2. The decoded output needs to be copied > into > a staging texture that can be accessed by the CPU as the decoder texture can't > be

[libav-devel] [PATCH] configure: use -O1 with MSVC by default

From: Steve Lhomme Otherwise some ARM and other unsupported CPU/OS is linked with -O0 --- configure | 1 + 1 file changed, 1 insertion(+) diff --git a/configure b/configure index 6d809f7..251b7c4 100755 --- a/configure +++ b/configure @@ -3278,6 +3278,7 @@ probe_cc(){

[libav-devel] [PATCH] avconv: add avconv_d3d11va

From: Steve Lhomme The code is similar to avconv_dxva2. The decoded output needs to be copied into a staging texture that can be accessed by the CPU as the decoder texture can't be accessed by the CPU. --- Makefile | 1 + avconv.h

[libav-devel] [PATCH] avconv: add avconv_d3d11va

From: Steve Lhomme The code is similar to avconv_dxva2. The decoded output needs to be copied into a staging texture that can be accessed by the CPU as the decoder texture can't be accessed by the CPU. --- Makefile | 1 + avconv.h

[libav-devel] [RFC] [PATCH 2/2] ppc: Merge types_altivec.h into util_altivec.h

There is no point in keeping the two separate. --- Maybe there should be a better name than util_altivec.h. libavutil/ppc/types_altivec.h | 47 --- libavutil/ppc/util_altivec.h | 23 - 2 files changed, 22 insertions(+), 48

[libav-devel] [RFC] [PATCH 1/2] ppc: Centralize compiler-specific altivec.h #include handling in one place

Also move #includes into canonical order where appropriate. --- It seems saner not to spread this ugliness all over the place. libavcodec/ppc/apedsp_altivec.c | 6 ++ libavcodec/ppc/audiodsp.c | 5 + libavcodec/ppc/blockdsp.c | 7 +++ libavcodec/ppc/fdctdsp.c

Re: [libav-devel] [PATCH 1/2] vaapi_encode: Support forcing IDR frames via AVFrame.pict_type

On 12/12/2016 22:25, Mark Thompson wrote: > --- > Makes it easier to hit the problem in > > (previously we only truncated at the end), so that should be applied first. > Probably Ok.