---
libavcodec/exr.c | 2 +-
libavcodec/exrenc.c| 2 +-
libavcodec/pnmdec.c| 3 ++-
libavcodec/pnmenc.c| 2 +-
{libavcodec => libavutil}/float2half.h | 6 +++---
{libavcodec => libavutil}/half2float.h | 6 +++---
6 f
IEEE-754 differentiates two different kind of NaNs.
Quiet and Signaling ones. They are differentiated by the MSB of the
mantissa.
For whatever reason, actual hardware conversion of half to single always
sets the signaling bit to 1 if the mantissa is != 0, and to 0 if it's 0.
So our code has to fol
Having to put the knowledge of the size of those arrays into a multitude
of places is rather smelly.
---
libavcodec/exr.c | 27 --
libavcodec/exrenc.c| 11 +
libavcodec/pnm.h | 5 ++---
libavcodec/pnmdec.c| 42 --
lib
---
libavcodec/Makefile | 8 +++---
libavcodec/exr.c| 2 +-
libavcodec/exrenc.c | 2 +-
libavcodec/float2half.c | 19 +
libavcodec/half2float.c | 19 +
libavcodec/pnmdec.c | 2 +-
libavcodec/pnmenc.c | 2 +-
libavutil/float2half.c | 53
_Float16 support was available on arm/aarch64 for a while, and with gcc
12 was enabled on x86 as long as SSE2 is supported.
If the target arch supports f16c, gcc emits fairly efficient assembly,
taking advantage of it. This is the case on x86-64-v3 or higher.
Same goes on arm, which has native flo
This is by no means perfect, since at least ddagrab will return scRGB
data with values outside of 0.0f to 1.0f for HDR values.
Its primary purpose is to be able to work with the format at all.
---
libavutil/Makefile| 1 +
libswscale/half2float.c | 19 +
libswscale/input.c
---
libswscale/hscale.c | 12 +--
libswscale/input.c| 149 ++
libswscale/swscale_internal.h | 17 ++--
libswscale/x86/swscale.c | 13 +--
4 files changed, 106 insertions(+), 85 deletions(-)
diff --git a/libswscale/hscale.c b/libswscale
On Sat, Aug 13, 2022 at 09:06:50PM +0200, Timo Rothenpieler wrote:
> On 13.08.2022 18:29, Michael Niedermayer wrote:
> > I fully support better IPFS support
> > what iam a bit "upset" about is that running a IPFS node is presented as
> > if that was more private than using a gateway.
>
> That's no
On Thu, Aug 11, 2022 at 11:31:32PM +0200, Andreas Rheinhardt wrote:
> Michael Niedermayer:
> > On Thu, Aug 11, 2022 at 10:50:10PM +0200, Andreas Rheinhardt wrote:
> >> Michael Niedermayer:
> >>> On Wed, Aug 10, 2022 at 10:47:09PM +0200, Timo Rothenpieler wrote:
> ---
> libavcodec/Makefil
On 11.08.2022 23:31, Andreas Rheinhardt wrote:
Michael Niedermayer:
On Thu, Aug 11, 2022 at 10:50:10PM +0200, Andreas Rheinhardt wrote:
Michael Niedermayer:
On Wed, Aug 10, 2022 at 10:47:09PM +0200, Timo Rothenpieler wrote:
---
libavcodec/Makefile | 8 +++---
libavcodec/exr.c|
On 8/14/22 12:48, Timo Rothenpieler wrote:
This is by no means perfect, since at least ddagrab will return scRGB
data with values outside of 0.0f to 1.0f for HDR values.
Its primary purpose is to be able to work with the format at all.
---
libavutil/Makefile| 1 +
libswscale/half
Some code in alimiter assumes that there are 2 channels, resulting in
clipping if the loudest channel is 3 or above and an out-of-bounds read if
the input is monophonic. Fix that in 2 places.
Signed-off-by: David Flater
---
libavfilter/af_alimiter.c | 23 +--
1 file changed,
Leo Izen:
> On 8/14/22 12:48, Timo Rothenpieler wrote:
>> This is by no means perfect, since at least ddagrab will return scRGB
>> data with values outside of 0.0f to 1.0f for HDR values.
>> Its primary purpose is to be able to work with the format at all.
>> ---
>> libavutil/Makefile
Michael Niedermayer:
> On Thu, Aug 11, 2022 at 11:31:32PM +0200, Andreas Rheinhardt wrote:
>> Michael Niedermayer:
>>> On Thu, Aug 11, 2022 at 10:50:10PM +0200, Andreas Rheinhardt wrote:
Michael Niedermayer:
> On Wed, Aug 10, 2022 at 10:47:09PM +0200, Timo Rothenpieler wrote:
>> ---
>>
Andreas Rheinhardt:
> Signed-off-by: Andreas Rheinhardt
> ---
> libavcodec/avcodec.c | 6 --
> libavcodec/encode.c | 6 ++
> 2 files changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/libavcodec/avcodec.c b/libavcodec/avcodec.c
> index f82d9e9f74..0451f57f82 100644
> --- a/libavc
From: bwang30
This commit enabled assembly code with intel AVX512 VNNI and added unit test
for sobel filter
sobel_c: 4537
sobel_avx512icl 2470
Signed-off-by: bwang30
---
libavfilter/convolution.h | 2 +
libavfilter/vf_convolution.c | 23
libavfilter/x86/vf_convol
On Sun, 2022-08-14 at 14:33 -0700, Philip Langdale wrote:
> These are the formats returned by the Intel VAAPI decoder for 12bit
> 4:2:2, 10bit 4:4:4, and 12bit 4:4:4 respectively. As with the already
> supported Y210 and YUVA (AVUY) formats, they are the formats Microsoft
> picked as their preferre
-Original Message-
From: ffmpeg-devel On Behalf Of Andreas
Rheinhardt
Sent: Friday, August 12, 2022 5:11 PM
To: ffmpeg-devel@ffmpeg.org
Subject: Re: [FFmpeg-devel] [PATCH] libavfilter/x86/vf_convolution: add sobel
filter optimization and unit test with intel AVX512 VNNI
bin.wang-at-inte
18 matches
Mail list logo
