On Tue, 11 Oct 2016, Martin Storsjö wrote:
On Mon, 10 Oct 2016, Luca Barbato wrote:
On 10/10/2016 22:02, Martin Storsjö wrote:
Would the benchmark numbers be more readable/usable if presented as
relative speedup vs the C version? That'd cut the number of rows
in half, and making it easier to
On Mon, 10 Oct 2016, Luca Barbato wrote:
On 10/10/2016 22:02, Martin Storsjö wrote:
Would the benchmark numbers be more readable/usable if presented as
relative speedup vs the C version? That'd cut the number of rows
in half, and making it easier to spot outliers or areas needing
more tuning.
On 10/10/2016 22:02, Martin Storsjö wrote:
> This makes it match the pattern already used for VP8 MC functions.
>
> This also makes the signature match ffmpeg's version of these
> functions, easing porting of code in both directions.
> ---
> libavcodec/vp9.h | 5 ++--
>
On 10/10/2016 22:02, Martin Storsjö wrote:
> Would the benchmark numbers be more readable/usable if presented as
> relative speedup vs the C version? That'd cut the number of rows
> in half, and making it easier to spot outliers or areas needing
> more tuning.
>
> The relative speedups for the
On 10/10/2016 22:22, Mark Thompson wrote:
> ---
> On 10/10/16 21:00, Luca Barbato wrote:
>> On 10/10/2016 21:43, Mark Thompson wrote:
>>> No, constrained high is high|set4|set5. Maybe it needs some new
>>> FF_PROFILE_H264 flags - *_PROGRESSIVE and *_NO_B_SLICES, or something?
>>
>> I guess, do
On 10/10/2016 22:35, Luca Barbato wrote:
> +} else if (c->srcBpc == 12 && 0) { \
Meh, disregard that =P
___
libav-devel mailing list
libav-devel@libav.org
https://lists.libav.org/mailman/listinfo/libav-devel
[PATCH 1/4] Add GBRP12 pixel format support
Unchanged
[PATCH 2/4] swscale: Add input support for 12-bit formats
Unchanged
PATCH 3/4] swscale: x86: Add some forgotten 12-bit planar YUV cases
The scalers wiring was missing from Diego's cherry pick.
[PATCH 4/4] swscale: Enable GBRP12 output
---
libswscale/output.c | 2 ++
libswscale/utils.c | 4 ++--
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/libswscale/output.c b/libswscale/output.c
index 78ebbc6..a8845f5 100644
--- a/libswscale/output.c
+++ b/libswscale/output.c
@@ -1455,6 +1455,8 @@ av_cold void
From: Michael Niedermayer
Signed-off-by: Diego Biurrun
---
libswscale/x86/scale.asm | 2 ++
libswscale/x86/swscale.c | 8 +---
2 files changed, 7 insertions(+), 3 deletions(-)
diff --git a/libswscale/x86/scale.asm b/libswscale/x86/scale.asm
index
From: Michael Niedermayer
Implemented for AV_PIX_FMT_GBRP12.
Signed-off-by: Vittorio Giovara
---
libswscale/input.c| 34 ++
libswscale/ppc/swscale_altivec.c | 2 +-
libswscale/utils.c
From: Michael Niedermayer
Signed-off-by: Vittorio Giovara
---
doc/APIchanges | 3 +++
libavutil/pixdesc.c | 25 +
libavutil/pixfmt.h | 4
libavutil/version.h | 2 +-
4 files changed, 33 insertions(+), 1
---
libavcodec/x86/hevc_idct.asm | 787 +++--
libavcodec/x86/hevcdsp_init.c | 54 ++-
2 files changed, 811 insertions(+), 30 deletions(-)
diff --git a/libavcodec/x86/hevc_idct.asm b/libavcodec/x86/hevc_idct.asm
index d662aa9..f397cc1 100644
---
---
On 10/10/16 21:00, Luca Barbato wrote:
> On 10/10/2016 21:43, Mark Thompson wrote:
>> No, constrained high is high|set4|set5. Maybe it needs some new
>> FF_PROFILE_H264 flags - *_PROGRESSIVE and *_NO_B_SLICES, or something?
>
> I guess, do you have time to get the right combination in?
This makes it match the pattern already used for VP8 MC functions.
This also makes the signature match ffmpeg's version of these
functions, easing porting of code in both directions.
---
libavcodec/vp9.h | 5 ++--
libavcodec/vp9block.c| 16 +-
libavcodec/vp9dsp.c
This work is sponsored by, and copyright, Google.
The filter coefficients are signed values, where
the product of the multiplication with one individual
filter coefficient doesn't overflow a 16 bit signed
value (the largest filter coefficient is 127). But
when the products are accumulated, the
On 10/10/2016 21:43, Mark Thompson wrote:
> No, constrained high is high|set4|set5. Maybe it needs some new
> FF_PROFILE_H264 flags - *_PROGRESSIVE and *_NO_B_SLICES, or something?
I guess, do you have time to get the right combination in?
lu
___
Only works if packed headers are supported, where we can know the
output before generating the first frame.
---
Added padding; fail harder; informative comment in header.
Not sure how to do this in the non-packed-header case - we could just invoke
this anyway, but the result is unlikely to
On 07/10/16 09:47, Luca Barbato wrote:
$subject =~ s/VE/EV/
> ---
>
> Not totally sure about constrained high. If it is correct I'd add it
> to the profiles mapped.
>
> libavcodec/nvenc.c | 8 +---
> 1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/libavcodec/nvenc.c
On 07/10/2016 10:47, Luca Barbato wrote:
> ---
>
> Not totally sure about constrained high. If it is correct I'd add it
> to the profiles mapped.
>
> libavcodec/nvenc.c | 8 +---
> 1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/libavcodec/nvenc.c b/libavcodec/nvenc.c
>
Quoting Vittorio Giovara (2016-10-04 17:59:53)
> ---
> libavcodec/avcodec.h | 7 +--
> 1 file changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
> index 167525d..02f391c 100644
> --- a/libavcodec/avcodec.h
> +++ b/libavcodec/avcodec.h
> @@
Quoting Diego Biurrun (2016-10-05 22:41:38)
> Its single forward declaration can be moved to the only place
> it is used, like is done for all other dsp init files.
> ---
> libavcodec/x86/audiodsp.h | 25 -
> libavcodec/x86/audiodsp_init.c | 4 +++-
> 2 files
Quoting James Almer (2016-10-03 21:40:06)
> Arguments for bitstream filters weren't being parsed at all, and instead
> checked by av_bsf_get_by_name() as if they were part of the filter's name
>
> Signed-off-by: James Almer
> ---
> This should also go into Libav 12, but i'm
On 10/3/2016 4:40 PM, James Almer wrote:
> Arguments for bitstream filters weren't being parsed at all, and instead
> checked by av_bsf_get_by_name() as if they were part of the filter's name
>
> Signed-off-by: James Almer
> ---
> This should also go into Libav 12, but i'm not
On Sat, Oct 08, 2016 at 12:51:25PM +0200, Alexandra Hájková wrote:
> --- a/libavcodec/x86/hevc_idct.asm
> +++ b/libavcodec/x86/hevc_idct.asm
> @@ -22,6 +23,217 @@
> @@ -74,34 +286,572 @@ cglobal hevc_idct_%1x%1_dc_%2, 1, 2, 1, coeff, tmp
> -; 8-bit
> -INIT_MMX mmxext
> -IDCT_DC_NL 4, 8
>
24 matches
Mail list logo