Re: [FFmpeg-devel] [PATCH]configure: Add -Wno-main

2017-09-17 Thread Clément Bœsch 
On Mon, Sep 18, 2017 at 03:55:12AM +0200, Carl Eugen Hoyos wrote:
> 2017-09-18 3:47 GMT+02:00 James Almer :
> > On 9/17/2017 10:37 PM, Carl Eugen Hoyos wrote:
> >> Hi!
> >>
> >> Attached patch fixes several warnings when compiling libavfilter with
> >> current gcc.
> >>
> >> Please comment, Carl Eugen
> >
> > IMO, it would be better if we instead rename all the cases of "main"
> > used across the codebase.
> 
> No strong opinion here, I would prefer if the warning were silenced.
> 

I'm with James on this one, it's easy and harmless to fix, so I think
we should do that instead.

-- 
Clément B.


signature.asc
Description: PGP signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH 3/3] avfilter/interlace: add support for 10 and 12 bit

2017-09-17 Thread James Almer 
On 9/14/2017 5:58 PM, Thomas Mundt wrote:
> Patch attached
> 

Applied the addition of the two lowpass filtering tests only for now.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] Configure Script change request to support libcdio on macOS

2017-09-17 Thread 桃源老師 
Hello, 

> 2017/09/17 午後11:58、桃源老師 のメール:
> 
> I have tried to use libcdio function with ffmpeg on macOS. Then I get an 
> information from libcdio team.  That is ffmpeg configure script should read 
> libcdio's pkgconfig file, libcdio.pc.

I attach my request as patch file.  Please review it.



ffmpeg_configure.patch
Description: Binary data



Best Regards,


// Miya aka. TougenRoushi





smime.p7s
Description: S/MIME cryptographic signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

[FFmpeg-devel] [PATCH 1/2] avformat/hlsenc: fix CID 1418106

2017-09-17 Thread Steven Liu 
fix the "Uninitialized scalar variable (UNINIT)" problem.

Signed-off-by: Steven Liu 
---
 libavformat/hlsenc.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/libavformat/hlsenc.c b/libavformat/hlsenc.c
index 3a9a235514..d70a2f7ff6 100644
--- a/libavformat/hlsenc.c
+++ b/libavformat/hlsenc.c
@@ -1327,13 +1327,13 @@ static int hls_write_header(AVFormatContext *s)
 {
 HLSContext *hls = s->priv_data;
 int ret, i;
-char *p;
+char *p = NULL;
 const char *pattern = "%d.ts";
 const char *pattern_localtime_fmt = get_default_pattern_localtime_fmt(s);
 const char *vtt_pattern = "%d.vtt";
 AVDictionary *options = NULL;
-int basename_size;
-int vtt_basename_size;
+int basename_size = 0;
+int vtt_basename_size = 0;
 
 if (hls->segment_type == SEGMENT_TYPE_FMP4) {
 pattern = "%d.m4s";
-- 
2.11.0 (Apple Git-81)



___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

[FFmpeg-devel] [PATCH 2/2] MAINTAINERS: modify the hlsenc description

2017-09-17 Thread Steven Liu 
change the hlsenc from hls encryption to hlsenc

Suggested-by: Aman Gupta 
Signed-off-by: Steven Liu 
---
 MAINTAINERS | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 0b0f7fa1e4..81e1e4916a 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -409,7 +409,7 @@ Muxers/Demuxers:
   gxf.c Reimar Doeffinger
   gxfenc.c  Baptiste Coudurier
   hls.c Anssi Hannula
-  hls encryption (hlsenc.c) Christian Suloway, Steven Liu
+  hlsenc.c  Christian Suloway, Steven Liu
   idcin.c   Mike Melanson
   idroqdec.cMike Melanson
   iff.c Jaikrishnan Menon
-- 
2.11.0 (Apple Git-81)



___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH]configure: Add -Wno-main

2017-09-17 Thread Carl Eugen Hoyos 
2017-09-18 3:47 GMT+02:00 James Almer :
> On 9/17/2017 10:37 PM, Carl Eugen Hoyos wrote:
>> Hi!
>>
>> Attached patch fixes several warnings when compiling libavfilter with
>> current gcc.
>>
>> Please comment, Carl Eugen
>
> IMO, it would be better if we instead rename all the cases of "main"
> used across the codebase.

No strong opinion here, I would prefer if the warning were silenced.

Carl Eugen
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

[FFmpeg-devel] [PATCH] x86/exrdsp: optimize ff_reorder_pixels_avx2()

2017-09-17 Thread James Almer 
From: Henrik Gramner 

Tested with "checkasm --test=exrdsp -bench"

Before:
reorder_pixels_c: 5187.8
reorder_pixels_sse2: 377.0
reorder_pixels_avx2: 331.3

After:
reorder_pixels_c: 5181.5
reorder_pixels_sse2: 377.0
reorder_pixels_avx2: 313.8

Signed-off-by: James Almer 
---
 libavcodec/x86/exrdsp.asm | 13 ++---
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/libavcodec/x86/exrdsp.asm b/libavcodec/x86/exrdsp.asm
index b91a7be20d..06c629e59e 100644
--- a/libavcodec/x86/exrdsp.asm
+++ b/libavcodec/x86/exrdsp.asm
@@ -39,16 +39,15 @@ cglobal reorder_pixels, 3,4,3, dst, src1, size, src2
 neg  sizeq; size = offset for 
dst, src1, src2
 .loop:
 
-%if cpuflag(avx2)
-vpermq  m0, [src1q + sizeq], 0xd8; load first 
part
-vpermq  m1, [src2q + sizeq], 0xd8; load second 
part
-%else
 movam0, [src1q+sizeq]; load first 
part
 movum1, [src2q+sizeq]; load second 
part
-%endif
 SBUTTERFLY bw, 0, 1, 2   ; interleaved
-mova [dstq+2*sizeq   ], m0   ; copy to dst
-mova [dstq+2*sizeq+mmsize], m1
+mova [dstq+2*sizeq   ], xm0  ; copy to dst
+mova [dstq+2*sizeq+16], xm1
+%if cpuflag(avx2)
+vperm2i128  m0, m0, m1, q0301
+mova [dstq+2*sizeq+32], m0
+%endif
 add sizeq, mmsize
 jl .loop
 RET
-- 
2.14.1

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH]configure: Add -Wno-main

2017-09-17 Thread James Almer 
On 9/17/2017 10:37 PM, Carl Eugen Hoyos wrote:
> Hi!
> 
> Attached patch fixes several warnings when compiling libavfilter with
> current gcc.
> 
> Please comment, Carl Eugen

IMO, it would be better if we instead rename all the cases of "main"
used across the codebase.
I'm surprised for that matter that SDL, with its "-Dmain=SDL_main"
cflag, didn't make a mess out of this.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

[FFmpeg-devel] [PATCH]configure: Add -Wno-main

2017-09-17 Thread Carl Eugen Hoyos 
Hi!

Attached patch fixes several warnings when compiling libavfilter with
current gcc.

Please comment, Carl Eugen
From 2d2893e8f7406043f970ae347509c39832c97b50 Mon Sep 17 00:00:00 2001
From: Carl Eugen Hoyos 
Date: Mon, 18 Sep 2017 03:35:37 +0200
Subject: [PATCH] configure: Add -Wno-main.

Silences several warnings:
main is usually a function
---
 configure |1 +
 1 file changed, 1 insertion(+)

diff --git a/configure b/configure
index 2de20a0..952cb16 100755
--- a/configure
+++ b/configure
@@ -6248,6 +6248,7 @@ check_cflags -Wmissing-prototypes
 check_cflags -Wno-pointer-to-int-cast
 check_cflags -Wstrict-prototypes
 check_cflags -Wempty-body
+check_cflags -Wno-main
 
 if enabled extra_warnings; then
 check_cflags -Wcast-qual
-- 
1.7.10.4

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] avfilter: add vmafmotion filter

2017-09-17 Thread James Almer 
On 9/15/2017 5:47 PM, Ashish Pratap Singh wrote:
> From: Ashish Singh 
> 
> Hi, this patch addresses the previous issues and changes it to a single
> input filter.
> 
> Signed-off-by: Ashish Singh 
> ---
>  Changelog   |   1 +
>  doc/filters.texi|  14 ++
>  libavfilter/Makefile|   1 +
>  libavfilter/allfilters.c|   1 +
>  libavfilter/vf_vmafmotion.c | 325 
> 
>  libavfilter/vmaf_motion.h   |  58 
>  6 files changed, 400 insertions(+)
>  create mode 100644 libavfilter/vf_vmafmotion.c
>  create mode 100644 libavfilter/vmaf_motion.h
> 
> diff --git a/Changelog b/Changelog
> index ea48e81..574f46e 100644
> --- a/Changelog
> +++ b/Changelog
> @@ -48,6 +48,7 @@ version :
>  - convolve video filter
>  - VP9 tile threading support
>  - KMS screen grabber
> +- vmafmotion video filter
>  
>  version 3.3:
>  - CrystalHD decoder moved to new decode API
> diff --git a/doc/filters.texi b/doc/filters.texi
> index 830de54..d996357 100644
> --- a/doc/filters.texi
> +++ b/doc/filters.texi
> @@ -15570,6 +15570,20 @@ vignette='PI/4+random(1)*PI/50':eval=frame
>  
>  @end itemize
>  
> +@section vmafmotion
> +
> +Obtain the average vmaf motion score of a video.
> +It is one of the component filters of VMAF.
> +
> +The obtained average motion score is printed through the logging system.
> +
> +In the below example the input file @file{ref.mpg} is being processed and 
> score
> +is computed.
> +
> +@example
> +ffmpeg -i ref.mpg -lavfi vmafmotion -f null -
> +@end example
> +
>  @section vstack
>  Stack input videos vertically.
>  
> diff --git a/libavfilter/Makefile b/libavfilter/Makefile
> index 8aa974e..4289ee0 100644
> --- a/libavfilter/Makefile
> +++ b/libavfilter/Makefile
> @@ -330,6 +330,7 @@ OBJS-$(CONFIG_VFLIP_FILTER)  += vf_vflip.o
>  OBJS-$(CONFIG_VIDSTABDETECT_FILTER)  += vidstabutils.o 
> vf_vidstabdetect.o
>  OBJS-$(CONFIG_VIDSTABTRANSFORM_FILTER)   += vidstabutils.o 
> vf_vidstabtransform.o
>  OBJS-$(CONFIG_VIGNETTE_FILTER)   += vf_vignette.o
> +OBJS-$(CONFIG_VMAFMOTION_FILTER) += vf_vmafmotion.o framesync.o
>  OBJS-$(CONFIG_VSTACK_FILTER) += vf_stack.o framesync.o
>  OBJS-$(CONFIG_W3FDIF_FILTER) += vf_w3fdif.o
>  OBJS-$(CONFIG_WAVEFORM_FILTER)   += vf_waveform.o
> diff --git a/libavfilter/allfilters.c b/libavfilter/allfilters.c
> index 63e8672..8ec54be 100644
> --- a/libavfilter/allfilters.c
> +++ b/libavfilter/allfilters.c
> @@ -341,6 +341,7 @@ static void register_all(void)
>  REGISTER_FILTER(VIDSTABDETECT,  vidstabdetect,  vf);
>  REGISTER_FILTER(VIDSTABTRANSFORM, vidstabtransform, vf);
>  REGISTER_FILTER(VIGNETTE,   vignette,   vf);
> +REGISTER_FILTER(VMAFMOTION, vmafmotion, vf);
>  REGISTER_FILTER(VSTACK, vstack, vf);
>  REGISTER_FILTER(W3FDIF, w3fdif, vf);
>  REGISTER_FILTER(WAVEFORM,   waveform,   vf);
> diff --git a/libavfilter/vf_vmafmotion.c b/libavfilter/vf_vmafmotion.c
> new file mode 100644
> index 000..c31c37c
> --- /dev/null
> +++ b/libavfilter/vf_vmafmotion.c
> @@ -0,0 +1,325 @@
> +/*
> + * Copyright (c) 2017 Ronald S. Bultje 
> + * Copyright (c) 2017 Ashish Pratap Singh 
> + *
> + * This file is part of FFmpeg.
> + *
> + * FFmpeg is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2.1 of the License, or (at your option) any later version.
> + *
> + * FFmpeg is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with FFmpeg; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 
> USA
> + */
> +
> +/**
> + * @file
> + * Calculate VMAF Motion score.
> + */
> +
> +#include "libavutil/opt.h"
> +#include "libavutil/pixdesc.h"
> +#include "avfilter.h"
> +#include "drawutils.h"
> +#include "formats.h"
> +#include "internal.h"
> +#include "vmaf_motion.h"
> +
> +#define vmafmotion_options NULL

This is unused.

> +#define BIT_SHIFT 10
> +
> +static const float FILTER_5[5] = {
> +0.054488685,
> +0.244201342,
> +0.402619947,
> +0.244201342,
> +0.054488685
> +};
> +
> +typedef struct VMAFMotionContext {
> +const AVClass *class;
> +VMAFMotionData data;
> +} VMAFMotionContext;
> +
> +AVFILTER_DEFINE_CLASS(vmafmotion);
> +
> +static uint64_t image_sad(const uint16_t *img1, const uint16_t *img2, int w,
> +  int h, ptrdiff_t

[FFmpeg-devel] [PATCH] avcodec/dirac_dwt: Fix integer overflow in COMPOSE_FIDELITYi*()

2017-09-17 Thread Michael Niedermayer 
Fixes: runtime error: signed integer overflow: 161 * 13872281 cannot be 
represented in type 'int'

Fixes: 3295/clusterfuzz-testcase-minimized-4738998142500864

Found-by: continuous fuzzing process 
https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer 
---
 libavcodec/dirac_dwt.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/libavcodec/dirac_dwt.h b/libavcodec/dirac_dwt.h
index adf5178714..755d5e5d2d 100644
--- a/libavcodec/dirac_dwt.h
+++ b/libavcodec/dirac_dwt.h
@@ -111,10 +111,10 @@ void ff_spatial_idwt_slice2(DWTContext *d, int y);
 (b0 + b1)
 
 #define COMPOSE_FIDELITYiL0(b0, b1, b2, b3, b4, b5, b6, b7, b8)\
-(b4 - ((-8*(b0+b8) + 21*(b1+b7) - 46*(b2+b6) + 161*(b3+b5) + 128) >> 8))
+(b4 - ((int)(-8*(b0+(unsigned)b8) + 21*(b1+(unsigned)b7) - 
46*(b2+(unsigned)b6) + 161*(b3+(unsigned)b5) + 128) >> 8))
 
 #define COMPOSE_FIDELITYiH0(b0, b1, b2, b3, b4, b5, b6, b7, b8)\
-(b4 + ((-2*(b0+b8) + 10*(b1+b7) - 25*(b2+b6) + 81*(b3+b5) + 128) >> 8))
+(b4 + ((int)(-2*(b0+(unsigned)b8) + 10*(b1+(unsigned)b7) - 
25*(b2+(unsigned)b6) +  81*(b3+(unsigned)b5) + 128) >> 8))
 
 #define COMPOSE_DAUB97iL1(b0, b1, b2)\
 (b1 - ((int)(1817*(b0 + (unsigned)b2) + 2048) >> 12))
-- 
2.14.1

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH 1/2] avcodec/wmv2dec: Check end of bitstream in parse_mb_skip() and ff_wmv2_decode_mb()

2017-09-17 Thread Michael Niedermayer 
Hi

On Sun, Sep 17, 2017 at 04:16:13PM -0400, Ronald S. Bultje wrote:
> Hi,
> 
> On Sat, Sep 16, 2017 at 8:42 PM, Michael Niedermayer  > wrote:
> 
> > +if (get_bits_left(>gb) < 0) {
> > +av_log(s->avctx, AV_LOG_ERROR,
> > +"Insufficient bits left at %d %d\n", s->mb_x, s->mb_y);
> > +return AVERROR_INVALIDDATA;
> > +}
> 
> 
> We've talked about this before, av_log(AV_LOG_ERROR) is not appropriate for
> such terse and unhelpful messages that really only apply to fuzz-broken
> files...

We talked about this before. And we disagree on almost everything
certainly the implied assumptions you repeat here.

Also i must say i really would prefer to work on the code without
these debates. I maintain the code in question here, can people not
just let me maintain the code  ...

IIRC, previously you wanted "All" these error messages removed from
the binary. Your statement above now leaves it more wide open how to
resolve it
But if you still want them removed from the binary, then my reply is
still the same

Iam happy to follow what the community prefers.

I have seen no evidence that theres a majority preferring to remove
all error messages for errors that have been found by help of automated
fuzzers. Ive seen 3 or 4 people complaining about error messages and
ive seen people working on fixing security issues (not just me)
disagreeing when they where told to remove messages.

I may be wrong but i suspect the "hardline" total removal has
few people supporting it. Especially outside FFmpeg and the projects
surrounding it. Ive not seen this radical viewpoint elsewhere.
Detailed error messages are valuable

If OTOH what you suggest now (you did after all not state it clearly
in your reply) is litterally just changing away from a
terse av_log(AV_LOG_ERROR) message. I certainly can try to improve
the error message, make it clearer and more verbose.

Thanks

[...]
-- 
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Never trust a computer, one day, it may think you are the virus. -- Compn


signature.asc
Description: Digital signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

[FFmpeg-devel] [PATCH]lavf/utils: Do not force the chapter end time before chapter start

2017-09-17 Thread Carl Eugen Hoyos 
Hi!

Attached patch fixes ticket #6671.

Please comment, Carl Eugen
From 87d7914d0f60c26ac89fde3cf2a21c6c95c64b0f Mon Sep 17 00:00:00 2001
From: Carl Eugen Hoyos 
Date: Sun, 17 Sep 2017 23:34:58 +0200
Subject: [PATCH] lavf/utils: Do not force chapter end time before chapter
 start.

Fixes ticket #6671.
---
 libavformat/utils.c |2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/libavformat/utils.c b/libavformat/utils.c
index 23865c8..7abca63 100644
--- a/libavformat/utils.c
+++ b/libavformat/utils.c
@@ -3167,7 +3167,7 @@ static void compute_chapters_end(AVFormatContext *s)
 if (j != i && next_start > ch->start && next_start < end)
 end = next_start;
 }
-ch->end = (end == INT64_MAX) ? ch->start : end;
+ch->end = (end == INT64_MAX || end < ch->start) ? ch->start : end;
 }
 }
 
-- 
1.7.10.4

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] lavd: add audio packets adopting example of NewTek NDI

2017-09-17 Thread Moritz Barsnick 
On Mon, Sep 04, 2017 at 17:12:45 +0300, Maksym Veremeyenko wrote:
> Subject: [PATCH 2/3] lavd: add audio packets adopting example of NewTek NDI
   ^ doc/outdevs:

> +Some NDI receiver want audio packets with 1920 samples per audio packet and 
> interleaved closer to video packets:
^ receivers
 ^ require

Assuming they actually require this and don't "prefer" it.

Moritz
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] libavcodec/exr : add X86 64 SIMD for reorder pixels (SSE and AVX2) (v4)

2017-09-17 Thread James Almer 
On 9/17/2017 5:05 PM, Martin Vignali wrote:
>>
>> fate-exr passes on mingw-w64 as well.
>>
> 
> Thanks for testing
> 
> Following your comments, new patch in attach
> 
> Martin

Pushed, thanks.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH 1/2] avcodec/wmv2dec: Check end of bitstream in parse_mb_skip() and ff_wmv2_decode_mb()

2017-09-17 Thread Ronald S. Bultje 
Hi,

On Sat, Sep 16, 2017 at 8:42 PM, Michael Niedermayer  wrote:

> +if (get_bits_left(>gb) < 0) {
> +av_log(s->avctx, AV_LOG_ERROR,
> +"Insufficient bits left at %d %d\n", s->mb_x, s->mb_y);
> +return AVERROR_INVALIDDATA;
> +}


We've talked about this before, av_log(AV_LOG_ERROR) is not appropriate for
such terse and unhelpful messages that really only apply to fuzz-broken
files...

Ronald
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] libavcodec/exr : add X86 64 SIMD for reorder pixels (SSE and AVX2) (v4)

2017-09-17 Thread Martin Vignali 
>
> fate-exr passes on mingw-w64 as well.
>

Thanks for testing

Following your comments, new patch in attach

Martin


0001-libavcodec-exr-add-X86-64-SIMD-for-reorder_pixels.patch
Description: Binary data
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] libavcodec/exr : add X86 64 SIMD for reorder pixels (SSE and AVX2) (v4)

2017-09-17 Thread James Almer 
On 9/17/2017 3:22 PM, Martin Vignali wrote:
> From 338f96a7f3c0f97cfafc0deda2322695a4006b5a Mon Sep 17 00:00:00 2001
> From: Martin Vignali 
> Date: Sun, 17 Sep 2017 20:05:16 +0200
> Subject: [PATCH] libavcodec/exr : add X86 64 SIMD for reorder_pixels
> 
> ---
>  libavcodec/Makefile  |  2 +-
>  libavcodec/exr.c | 44 ++--
>  libavcodec/exrdsp.c  | 45 +
>  libavcodec/exrdsp.h  | 32 
>  libavcodec/x86/Makefile  |  2 ++
>  libavcodec/x86/exrdsp.asm| 69 
> 
>  libavcodec/x86/exrdsp_init.c | 43 +++
>  7 files changed, 213 insertions(+), 24 deletions(-)
>  create mode 100644 libavcodec/exrdsp.c
>  create mode 100644 libavcodec/exrdsp.h
>  create mode 100644 libavcodec/x86/exrdsp.asm
>  create mode 100644 libavcodec/x86/exrdsp_init.c
> 
> diff --git a/libavcodec/Makefile b/libavcodec/Makefile
> index 943e5db511..fad56129a3 100644
> --- a/libavcodec/Makefile
> +++ b/libavcodec/Makefile
> @@ -286,7 +286,7 @@ OBJS-$(CONFIG_EIGHTSVX_FIB_DECODER)+= 8svx.o
>  OBJS-$(CONFIG_ESCAPE124_DECODER)   += escape124.o
>  OBJS-$(CONFIG_ESCAPE130_DECODER)   += escape130.o
>  OBJS-$(CONFIG_EVRC_DECODER)+= evrcdec.o acelp_vectors.o lsp.o
> -OBJS-$(CONFIG_EXR_DECODER) += exr.o
> +OBJS-$(CONFIG_EXR_DECODER) += exr.o exrdsp.o
>  OBJS-$(CONFIG_FFV1_DECODER)+= ffv1dec.o ffv1.o
>  OBJS-$(CONFIG_FFV1_ENCODER)+= ffv1enc.o ffv1.o
>  OBJS-$(CONFIG_FFWAVESYNTH_DECODER) += ffwavesynth.o
> diff --git a/libavcodec/exr.c b/libavcodec/exr.c
> index 759880756d..478c127ebe 100644
> --- a/libavcodec/exr.c
> +++ b/libavcodec/exr.c
> @@ -40,6 +40,7 @@
>  #include "libavutil/avassert.h"
>  #include "libavutil/common.h"
>  #include "libavutil/imgutils.h"
> +#include "libavutil/timer.h"

Not needed.

>  #include "libavutil/intfloat.h"
>  #include "libavutil/opt.h"
>  #include "libavutil/color_utils.h"
> @@ -55,6 +56,7 @@
>  #include "internal.h"
>  #include "mathops.h"
>  #include "thread.h"
> +#include "exrdsp.h"

Add this one above of get_bits.h, to keep the alphabetical order.

>  
>  enum ExrCompr {
>  EXR_RAW,
> @@ -121,6 +123,7 @@ typedef struct EXRContext {
>  AVClass *class;
>  AVFrame *picture;
>  AVCodecContext *avctx;
> +ExrDSPContext dsp;
>  
>  #if HAVE_BIGENDIAN
>  BswapDSPContext bbdsp;
> @@ -275,23 +278,7 @@ static void predictor(uint8_t *src, int size)
>  }
>  }
>  
> -static void reorder_pixels(uint8_t *src, uint8_t *dst, int size)
> -{
> -const uint8_t *t1 = src;
> -int half_size = size / 2;
> -const uint8_t *t2 = src + half_size;
> -uint8_t *s= dst;
> -int i;
> -
> -av_assert1(size % 2 == 0);
> -
> -for (i = 0; i < half_size; i++) {
> -*(s++) = *(t1++);
> -*(s++) = *(t2++);
> -}
> -}
> -
> -static int zip_uncompress(const uint8_t *src, int compressed_size,
> +static int zip_uncompress(EXRContext *s, const uint8_t *src, int 
> compressed_size,
>int uncompressed_size, EXRThreadData *td)
>  {
>  unsigned long dest_len = uncompressed_size;
> @@ -300,13 +287,18 @@ static int zip_uncompress(const uint8_t *src, int 
> compressed_size,
>  dest_len != uncompressed_size)
>  return AVERROR_INVALIDDATA;
>  
> +av_assert1(uncompressed_size % 2 == 0);
> +
>  predictor(td->tmp, uncompressed_size);
> -reorder_pixels(td->tmp, td->uncompressed_data, uncompressed_size);
> +
> +//START_TIMER;

Don't add dead benchmarking/debug code.

> +s->dsp.reorder_pixels(td->tmp, td->uncompressed_data, uncompressed_size);
> +//STOP_TIMER("reorder_pixels_zip");
>  
>  return 0;
>  }
>  
> -static int rle_uncompress(const uint8_t *src, int compressed_size,
> +static int rle_uncompress(EXRContext *ctx, const uint8_t *src, int 
> compressed_size,
>int uncompressed_size, EXRThreadData *td)
>  {
>  uint8_t *d  = td->tmp;
> @@ -345,8 +337,10 @@ static int rle_uncompress(const uint8_t *src, int 
> compressed_size,
>  if (dend != d)
>  return AVERROR_INVALIDDATA;
>  
> +av_assert1(uncompressed_size % 2 == 0);
> +
>  predictor(td->tmp, uncompressed_size);
> -reorder_pixels(td->tmp, td->uncompressed_data, uncompressed_size);
> +ctx->dsp.reorder_pixels(td->tmp, td->uncompressed_data, 
> uncompressed_size);
>  
>  return 0;
>  }
> @@ -954,6 +948,7 @@ static void unpack_14(const uint8_t b[14], uint16_t s[16])
>  }
>  }
>  
> +

Stray new line.

>  static void unpack_3(const uint8_t b[3], uint16_t s[16])
>  {
>  int i;
> @@ -1000,6 +995,7 @@ static int b44_uncompress(EXRContext *s, const uint8_t 
> *src, int compressed_size
>  
>  if (src[compressed_size - stay_to_uncompress + 2] == 
> 0xfc) { /* B44A block */
>

[FFmpeg-devel] libavcodec/exr : add X86 64 SIMD for reorder pixels (SSE and AVX2) (v4)

2017-09-17 Thread Martin Vignali 
Hello,

Following Henrik Grammer comments
new patch in attach

replace int size by ptrdiff_t size

I simplify the code, keeping only 1 loop (more easy to read, and doesn't
have a real impact on speed)
i use the SBUTTERFLY Macro for sse
for avx2 i keep my previous approach

Pass fate-exr tests for me (os X)

Current benchmark
AVX2
239920 decicycles in reorder_pixels_zip,  130958 runs,114 skips
bench: utime=101.367s

SSE
283768 decicycles in reorder_pixels_zip,  130948 runs,124 skips
bench: utime=101.424s

Scalar
3119101 decicycles in reorder_pixels_zip,  130429 runs,643 skips
bench: utime=114.414s


The result of the suggested asm by Henrik
AVX2 :
258602 decicycles in reorder_pixels_zip,  130853 runs,219 skips

SSE :
285167 decicycles in reorder_pixels_zip,  130863 runs,209 skips

In term of speed using -benchmark, the difference with the current patch is
hard to see.


Martin


0001-libavcodec-exr-add-X86-64-SIMD-for-reorder_pixels.patch
Description: Binary data
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] avformat/libopenmpt: Query duration and metadata after selecting subsong

2017-09-17 Thread Reino Wijnsma 
On 17-9-2017 15:35, Jörn Heusipp  wrote:
> Duration depends on the selected subsong and thus must be queried after
> selecting the subsong. There is no compelling reason to query other
> metadata earlier either.
>
> Signed-off-by: Jörn Heusipp 
> ---
>  libavformat/libopenmpt.c | 17 +
>  1 file changed, 9 insertions(+), 8 deletions(-)
>
> diff --git a/libavformat/libopenmpt.c b/libavformat/libopenmpt.c
> index a7e3859..af6eb1a 100644
> --- a/libavformat/libopenmpt.c
> +++ b/libavformat/libopenmpt.c
> @@ -93,14 +93,7 @@ static int read_header_openmpt(AVFormatContext *s)
>  if (!openmpt->module)
>  return AVERROR_INVALIDDATA;
>  
> -openmpt->channels   = av_get_channel_layout_nb_channels(openmpt->layout);
> -openmpt->duration   = 
> openmpt_module_get_duration_seconds(openmpt->module);
> -
> -add_meta(s, "artist",  openmpt_module_get_metadata(openmpt->module, 
> "artist"));
> -add_meta(s, "title",   openmpt_module_get_metadata(openmpt->module, 
> "title"));
> -add_meta(s, "encoder", openmpt_module_get_metadata(openmpt->module, 
> "tracker"));
> -add_meta(s, "comment", openmpt_module_get_metadata(openmpt->module, 
> "message"));
> -add_meta(s, "date",openmpt_module_get_metadata(openmpt->module, 
> "date"));
> +openmpt->channels = av_get_channel_layout_nb_channels(openmpt->layout);
>  
>  if (openmpt->subsong >= 
> openmpt_module_get_num_subsongs(openmpt->module)) {
>  openmpt_module_destroy(openmpt->module);
> @@ -120,6 +113,14 @@ static int read_header_openmpt(AVFormatContext *s)
>  }
>  }
>  
> +openmpt->duration = openmpt_module_get_duration_seconds(openmpt->module);
> +
> +add_meta(s, "artist",  openmpt_module_get_metadata(openmpt->module, 
> "artist"));
> +add_meta(s, "title",   openmpt_module_get_metadata(openmpt->module, 
> "title"));
> +add_meta(s, "encoder", openmpt_module_get_metadata(openmpt->module, 
> "tracker"));
> +add_meta(s, "comment", openmpt_module_get_metadata(openmpt->module, 
> "message"));
> +add_meta(s, "date",openmpt_module_get_metadata(openmpt->module, 
> "date"));
> +
>  st = avformat_new_stream(s, NULL);
>  if (!st) {
>  openmpt_module_destroy(openmpt->module);
I've tested your patch and I can confirm FFmpeg now correctly shows the
subsong duration.
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

[FFmpeg-devel] Configure Script change request to support libcdio on macOS

2017-09-17 Thread 桃源老師 
Hello,

In fact, I'm not a kind of ffmpeg developer at all.  While trying to link 
libcdio library with ffmpeg, I just happened to find out the way...

So I'm not sure I can answer any question about this configure script change...


I have tried to use libcdio function with ffmpeg on macOS. Then I get an 
information from libcdio team.  That is ffmpeg configure script should read 
libcdio's pkgconfig file, libcdio.pc.

So I tried and error, then found that the below change should be done to link 
libcdio library with ffmpeg on macOS.


$ diff -u a/configure b/configure 
--- a/configure 2017-09-16 15:02:06.0 +0900
+++ b/configure 2017-09-17 04:53:34.0 +0900
@@ -6151,7 +6151,7 @@

if enabled libcdio; then
check_lib libcdio "cdio/cdda.h cdio/paranoia.h" cdio_cddap_open 
-lcdio_paranoia -lcdio_cdda -lcdio ||
-check_lib libcdio "cdio/paranoia/cdda.h cdio/paranoia/paranoia.h" 
cdio_cddap_open -lcdio_paranoia -lcdio_cdda -lcdio ||
+require_pkg_config libcdio "cdio/paranoia/cdda.h cdio/paranoia/paranoia.h" 
cdio_cddap_open -lcdio_paranoia -lcdio_cdda -lcdio && check_lib libcdio 
"cdio/paranoia/cdda.h cdio/paranoia/paranoia.h" cdio_cddap_open -lcdio_paranoia 
-lcdio_cdda -lcdio ||
die "ERROR: No usable libcdio/cdparanoia found"
fi

Please review it and make it acceptable for everyone...


Best Regards,


// Miya aka. TougenRoushi



smime.p7s
Description: S/MIME cryptographic signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

[FFmpeg-devel] [PATCH] avformat/libopenmpt: Query duration and metadata after selecting subsong

2017-09-17 Thread Jörn Heusipp 
Duration depends on the selected subsong and thus must be queried after
selecting the subsong. There is no compelling reason to query other
metadata earlier either.

Signed-off-by: Jörn Heusipp 
---
 libavformat/libopenmpt.c | 17 +
 1 file changed, 9 insertions(+), 8 deletions(-)

diff --git a/libavformat/libopenmpt.c b/libavformat/libopenmpt.c
index a7e3859..af6eb1a 100644
--- a/libavformat/libopenmpt.c
+++ b/libavformat/libopenmpt.c
@@ -93,14 +93,7 @@ static int read_header_openmpt(AVFormatContext *s)
 if (!openmpt->module)
 return AVERROR_INVALIDDATA;
 
-openmpt->channels   = av_get_channel_layout_nb_channels(openmpt->layout);
-openmpt->duration   = openmpt_module_get_duration_seconds(openmpt->module);
-
-add_meta(s, "artist",  openmpt_module_get_metadata(openmpt->module, 
"artist"));
-add_meta(s, "title",   openmpt_module_get_metadata(openmpt->module, 
"title"));
-add_meta(s, "encoder", openmpt_module_get_metadata(openmpt->module, 
"tracker"));
-add_meta(s, "comment", openmpt_module_get_metadata(openmpt->module, 
"message"));
-add_meta(s, "date",openmpt_module_get_metadata(openmpt->module, 
"date"));
+openmpt->channels = av_get_channel_layout_nb_channels(openmpt->layout);
 
 if (openmpt->subsong >= openmpt_module_get_num_subsongs(openmpt->module)) {
 openmpt_module_destroy(openmpt->module);
@@ -120,6 +113,14 @@ static int read_header_openmpt(AVFormatContext *s)
 }
 }
 
+openmpt->duration = openmpt_module_get_duration_seconds(openmpt->module);
+
+add_meta(s, "artist",  openmpt_module_get_metadata(openmpt->module, 
"artist"));
+add_meta(s, "title",   openmpt_module_get_metadata(openmpt->module, 
"title"));
+add_meta(s, "encoder", openmpt_module_get_metadata(openmpt->module, 
"tracker"));
+add_meta(s, "comment", openmpt_module_get_metadata(openmpt->module, 
"message"));
+add_meta(s, "date",openmpt_module_get_metadata(openmpt->module, 
"date"));
+
 st = avformat_new_stream(s, NULL);
 if (!st) {
 openmpt_module_destroy(openmpt->module);
-- 
1.9.1

___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel