---
Updated to try to add movsxdifnidn for all occurrances
of "int stride" in libavcodec/x86/h264_idct*.
Using plain movsxd instead of movsxdifnidn within
ifdef ARCH_X86_64.
---
libavcodec/x86/h264_idct.asm | 21 +
libavcodec/x86/h264_idct_10bit.asm | 8
2
On Sat, 11 Jun 2016, Martin Storsjö wrote:
The forward dcts are based partially on x264. The idct tests
themselves are inspired by similar tests for vp9 by Ronald
Bultje.
---
Should we mark this as GPL, or are x264 people ok with picking
this part (which is more or less straight from the spec)?
On 12/06/16 20:45, Anton Khirnov wrote:
> ---
> Now also use the FrameInfo from the frames context, to get the correct
> allocated
> dimensions.
> ---
> libavcodec/qsvenc.c | 101
>
> libavcodec/qsvenc.h | 2 ++
> 2 files changed, 80
On Sun, Jun 12, 2016 at 08:43:39PM +0200, Anton Khirnov wrote:
> ---
> Now aligning the framesctx (allocated) dimensions to 32.
> ---
> Changelog | 1 +
> configure | 1 +
> libavfilter/Makefile | 1 +
> libavfilter/allfilters.c | 1 +
>
---
Now also use the FrameInfo from the frames context, to get the correct allocated
dimensions.
---
libavcodec/qsvenc.c | 101
libavcodec/qsvenc.h | 2 ++
2 files changed, 80 insertions(+), 23 deletions(-)
diff --git a/libavcodec/qsvenc.c
---
Now aligning the framesctx (allocated) dimensions to 32.
---
Changelog | 1 +
configure | 1 +
libavfilter/Makefile | 1 +
libavfilter/allfilters.c | 1 +
libavfilter/vf_scale_qsv.c | 633 +
5 files
On 09/06/16 17:13, Diego Biurrun wrote:
> ---
> libavcodec/eamad.c | 82 +
> libavcodec/mdec.c | 74
> libavcodec/mjpegdec.c | 56 +++---
> libavcodec/mpeg12.c| 78 +
> libavcodec/mpeg12dec.c | 417
>
On 09/06/16 17:13, Diego Biurrun wrote:
> ---
Ok.
___
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel
On 09/06/16 17:13, Diego Biurrun wrote:
> From: Alexandra Hájková
>
Seems fine.
___
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel
On Sat, Jun 11, 2016 at 11:38:27PM +0300, Martin Storsjö wrote:
> On Sat, 11 Jun 2016, Martin Storsjö wrote:
> > ---
> > libavcodec/x86/h264_idct.asm | 5 +
> > libavcodec/x86/h264_idct_10bit.asm | 4
> > 2 files changed, 9 insertions(+)
>
> I guess I should add, that I guess there's
No longer leaks memory when used with a driver with the "render does
not destroy param buffers" quirk (i.e. Intel i965).
---
libavcodec/vaapi_encode.c | 20 +++-
1 file changed, 19 insertions(+), 1 deletion(-)
diff --git a/libavcodec/vaapi_encode.c b/libavcodec/vaapi_encode.c
The driver being used is detected inside av_hwdevice_ctx_init(), and
the quirks field then set from a table of known drivers.
Also adds the Intel i965 driver quirk (it does not destroy parameter
buffers used in a call to vaRenderPicture()) and detects that driver
to set it.
---
Another
On 12/06/16 12:28, Anton Khirnov wrote:
> Ok, it seems the improvement is quite dependent on the compiler version.
> I did some more measurements and the results are quite ambiguous.
> The tests all decode a 1080p 175M dnxhd file on an i7-4770K. The
> following table shows the change in overall
On Sun, Jun 12, 2016 at 12:28 PM, Anton Khirnov wrote:
> Ok, it seems the improvement is quite dependent on the compiler version.
> I did some more measurements and the results are quite ambiguous.
> The tests all decode a 1080p 175M dnxhd file on an i7-4770K. The
> following
Ok, it seems the improvement is quite dependent on the compiler version.
I did some more measurements and the results are quite ambiguous.
The tests all decode a 1080p 175M dnxhd file on an i7-4770K. The
following table shows the change in overall decoding performance (as
measured by perf stat
15 matches
Mail list logo