Re: [FFmpeg-devel] [PATCH 7/7] h264dec: add a CUVID hwaccel

2017-10-03 Thread Timo Rothenpieler
Am 03.10.2017 um 15:15 schrieb wm4: From: Anton Khirnov <an...@khirnov.net> Some parts of the code are based on a patch by Timo Rothenpieler <t...@rothenpieler.org> Merges Libav commit b9129ec4668c511e0a79e25c6f25d748cee172c9. As a complication, all the names conflict. Add a _hw

Re: [FFmpeg-devel] [PATCH 03/11] libavutil/opencl: fix potentiall nul dereference

2017-08-27 Thread Timo Rothenpieler
Am 14.08.2017 um 03:47 schrieb Wei Gao: 2017-06-11 22:05 GMT+08:00 Timo Rothenpieler <t...@rothenpieler.org>: Fixes CID 1396840 --- libavutil/opencl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavutil/opencl.c b/libavutil/opencl.c index af35770e06..2027

Re: [FFmpeg-devel] [PATCH] compat/cuda/ptx2c: strip CR from each line

2017-08-29 Thread Timo Rothenpieler
Am 29.08.2017 um 02:46 schrieb Ricardo Constantino: Windows nvcc + cl.exe produce a .ctx file with CR+LF newlines which need to be stripped to work with gcc. --- compat/cuda/ptx2c.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/compat/cuda/ptx2c.sh

Re: [FFmpeg-devel] [Patch] CUDA Thumbnail Filter

2017-09-01 Thread Timo Rothenpieler
Also missing a dep on cuda_sdk in configure. smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [Patch] CUDA Thumbnail Filter

2017-09-04 Thread Timo Rothenpieler
We have av_pixelutils_sad_fn which does SAD and has SIMD, there's no point in reinventing the wheel. I also don't see why this needs to be implemented with CUDA. You're not even doing the SAD in CUDA. I bet it'll be just as fast if not faster in C (unless you cheat somehow). The point is to do

[FFmpeg-devel] [RFC] ffmpeg: reverse logic for interlaced field ordering heuristic

2017-09-01 Thread Timo Rothenpieler
When re-encoding for example interlaced h264 from mkv into mkv, the field ordering type switches from TT to TB, confusing some players. Same happens on a lot of other cases as well. I have no idea if this is the correct fix for it, but something is definitely going wrong. --- ffmpeg.c | 4 ++--

Re: [FFmpeg-devel] [PATCH] nvenc: Don't segfault on close if no cuda is available

2017-09-02 Thread Timo Rothenpieler
This should already be fixed in master and release/3.3 smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

[FFmpeg-devel] nvenc: missing slice offsets in NV_ENC_LOCK_BITSTREAM::sliceOffsets

2017-09-02 Thread Timo Rothenpieler
, as is required for field mode interlaced encoding to work properly. Is this indeed a driver bug, or is there some other way to get the required offset? Regards, Timo Rothenpieler smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing

[FFmpeg-devel] [PATCH] avcodec/nvenc: migrate to new encode API

2017-09-02 Thread Timo Rothenpieler
Signed-off-by: Timo Rothenpieler <t...@rothenpieler.org> --- libavcodec/nvenc.c | 65 - libavcodec/nvenc.h | 6 + libavcodec/nvenc_h264.c | 6 + libavcodec/nvenc_hevc.c | 4 +++ 4 files changed, 70 insertions(+), 11 del

[FFmpeg-devel] [PATCH] configure: add support for libnpp* from cuda sdk 9

2017-08-29 Thread Timo Rothenpieler
Signed-off-by: Timo Rothenpieler <t...@rothenpieler.org> --- configure | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/configure b/configure index 61d8160491..172ff3fc07 100755 --- a/configure +++ b/configure @@ -5875,7 +5875,9 @@ enabled

Re: [FFmpeg-devel] [PATCH v17] avformat/dashdec: add dash demuxer base version

2017-08-30 Thread Timo Rothenpieler
Just wanted to point out, as stated on IRC, that dashenc already has a function that seems to safely solve the filename template issue, and looks like it can be moved into a common file + header, and reused as is:

Re: [FFmpeg-devel] [Patch] CUDA Thumbnail Filter

2017-09-01 Thread Timo Rothenpieler
Am 30.08.2017 um 06:19 schrieb Yogender Gupta: Attached is a CUDA version of the thumbnail filter, this helps accelerate thumbnails generations significantly, when using the GPU pipeline. Regards, Yogender After having a look at the code: The filter is using a global "CUdeviceptr data;"

Re: [FFmpeg-devel] [PATCH 2/7] decode: add a method for attaching lavc-internal data to frames

2017-10-17 Thread Timo Rothenpieler
Am 17.10.2017 um 20:37 schrieb Thilo Borgmann: Am 17.10.17 um 20:27 schrieb wm4: On Tue, 17 Oct 2017 20:23:25 +0200 Thilo Borgmann wrote: Am 17.10.17 um 06:49 schrieb wm4: I have realized that your veto is actually not valid: - it's a Libav merge - it has been for

[FFmpeg-devel] [PATCH] hwcontext_d3d11va: add missing stdint.h include

2017-11-15 Thread Timo Rothenpieler
This caused checkheaders to fail on cygwin on this file. --- libavutil/hwcontext_d3d11va.h | 1 + 1 file changed, 1 insertion(+) diff --git a/libavutil/hwcontext_d3d11va.h b/libavutil/hwcontext_d3d11va.h index 98db7ce343..9f91e9b1b6 100644 --- a/libavutil/hwcontext_d3d11va.h +++

[FFmpeg-devel] [PATCH] movenc-test: fix potential uninitialized read

2017-11-13 Thread Timo Rothenpieler
Fixes CID #1413023 --- libavformat/tests/movenc.c | 1 + 1 file changed, 1 insertion(+) diff --git a/libavformat/tests/movenc.c b/libavformat/tests/movenc.c index 8e59b74259..ad6c92a94c 100644 --- a/libavformat/tests/movenc.c +++ b/libavformat/tests/movenc.c @@ -115,6 +115,7 @@ static int

[FFmpeg-devel] [PATCH] avformat/fitsenc: validate input pixel format

2017-11-13 Thread Timo Rothenpieler
Fixes CID #1416961 and #1416962 --- libavformat/fitsenc.c | 6 +- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/libavformat/fitsenc.c b/libavformat/fitsenc.c index 7cb171596c..cc3999aa8a 100644 --- a/libavformat/fitsenc.c +++ b/libavformat/fitsenc.c @@ -106,6 +106,8 @@ static

Re: [FFmpeg-devel] [PATCH 9/9] avfilter/signature_lookup: fix potential uninitialized reads

2017-11-13 Thread Timo Rothenpieler
Am 08.11.2017 um 19:17 schrieb Timo Rothenpieler: Fixes CID #1403238 and #1403239 --- libavfilter/signature_lookup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) ping I'm not fully certain if this is the right fix, or if there is a logic error in the quite complex code. Gonna

Re: [FFmpeg-devel] I just ran FATE. Did the tests pass?

2017-11-13 Thread Timo Rothenpieler
If there's no error, it passed. smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH 2/2] avcodec: Implement vc1 nvdec hwaccel

2017-11-14 Thread Timo Rothenpieler
looks straight forward enough to me, struct naming matches up as well. add a Changelog and bump lavc version, and it's good to go. smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org

Re: [FFmpeg-devel] AMD external header

2017-11-27 Thread Timo Rothenpieler
Am 27.11.2017 um 03:15 schrieb Mironov, Mikhail: Hi, I would like to summarize thoughts on several threads on this forum related to the issue of including AMD/AMF header file into FFmpeg source tree. It looks like they reflect some policies formal or informal. Mark tried to create some policy

Re: [FFmpeg-devel] AMD external header

2017-11-27 Thread Timo Rothenpieler
Personally, I would prefer if the bundled external headers were installed together with ffmpeg public headers (so nvenc/cuda/etc... weren't simply private headers within ffmpeg). There are some nvenc APIs I need to query hardware capabilities to avoid setting nvenc codec parameters that would

Re: [FFmpeg-devel] [PATCH 02/17] lavc: Add hardware config metadata for decoders supporting hardware output

2017-11-24 Thread Timo Rothenpieler
diff --git a/libavcodec/cuviddec.c b/libavcodec/cuviddec.c index 806dab2074..3bd2409ea2 100644 --- a/libavcodec/cuviddec.c +++ b/libavcodec/cuviddec.c @@ -32,6 +32,7 @@ #include "avcodec.h" #include "decode.h" +#include "hwaccel.h" #include "internal.h" typedef struct CuvidContext

Re: [FFmpeg-devel] Added HW H.264 and HEVC encoding for AMD GPUs based on AMF SDK

2017-11-22 Thread Timo Rothenpieler
Am 17.11.2017 um 06:07 schrieb mmironov: From 454aad17fea28e8e4c5abb904341181271971bfc Mon Sep 17 00:00:00 2001 From: mmironov Date: Tue, 14 Nov 2017 17:54:24 -0500 Subject: [PATCH] Added HW H.264 and HEVC encoding for AMD GPUs based on AMF SDK Signed-off-by:

Re: [FFmpeg-devel] Added HW H.264 and HEVC encoding for AMD GPUs based on AMF SDK

2017-11-22 Thread Timo Rothenpieler
I'd like to look through it again and test a bit more (will try to do so tomorrow, certainly by the end of the week), but I think it should be ready to commit with the external header removed. Are you planning to remove Nvidia headers as well? No, I am very much against this. And others have

Re: [FFmpeg-devel] Added HW H.264 and HEVC encoding for AMD GPUs based on AMF SDK

2017-11-22 Thread Timo Rothenpieler
Am 17.11.2017 um 06:07 schrieb mmironov: From 454aad17fea28e8e4c5abb904341181271971bfc Mon Sep 17 00:00:00 2001 From: mmironov Date: Tue, 14 Nov 2017 17:54:24 -0500 Subject: [PATCH] Added HW H.264 and HEVC encoding for AMD GPUs based on AMF SDK Signed-off-by:

Re: [FFmpeg-devel] [PATCH] examples/hw_decode: Use hw-config information to find pixfmt

2017-11-29 Thread Timo Rothenpieler
Am 29.11.2017 um 13:22 schrieb Moritz Barsnick: On Wed, Nov 29, 2017 at 00:34:03 +, Mark Thompson wrote: +if (type == AV_HWDEVICE_TYPE_NONE) { +fprintf(stderr, "Device type %s is not supported.\n", argv[1]); +fprintf(stderr, "Available device types:"); +type =

Re: [FFmpeg-devel] AMD external header

2017-11-29 Thread Timo Rothenpieler
Yeah, I understand that. But the function loading should make only a small part of it, and could just be a normal source file. (We do ad-hoc function loading for a lot of other stuff too.) Then we could depend on external, unmodified SDK headers, which someone could extract to a git repo, to

Re: [FFmpeg-devel] AMD external header

2017-11-29 Thread Timo Rothenpieler
Am 29.11.2017 um 17:40 schrieb wm4: On Wed, 29 Nov 2017 17:27:01 +0100 Hendrik Leppkes wrote: On Wed, Nov 29, 2017 at 5:24 PM, wm4 wrote: What really irks me is that nvidia is not giving us any support for supporting their stuff. AFAIK the

Re: [FFmpeg-devel] AMD external header

2017-11-27 Thread Timo Rothenpieler
I also noticed that recently there is a lot of activity to add full Nvidia decoders to FFmpeg (VC1, MPEG4, VP8, VP9 etc.). I am guessing this is to overcome DXVA or VAAPI shortcoming. What about AMD? We also have them. The primary motivation here is that nvidia abandoned vdpau, so to access

Re: [FFmpeg-devel] avcodec/nvenc: set correct error code

2017-11-27 Thread Timo Rothenpieler
lgtm smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] AMD external header

2017-11-27 Thread Timo Rothenpieler
Am 27.11.2017 um 03:15 schrieb Mironov, Mikhail: Hi, I would like to summarize thoughts on several threads on this forum related to the issue of including AMD/AMF header file into FFmpeg source tree. It looks like they reflect some policies formal or informal. Mark tried to create some policy

Re: [FFmpeg-devel] AMD external header

2017-11-28 Thread Timo Rothenpieler
Your use-case looks like an argument for moving the external headers into some separate repository. Not that I personally care much about bundling or not bundling. To me the more important question seems to be whether to auto-enable the encoders/decoders that depend on the external headers and

Re: [FFmpeg-devel] [PATCH] avcodec/nvdec: Round up odd width/height values

2017-11-24 Thread Timo Rothenpieler
Am 24.11.2017 um 19:24 schrieb Philip Langdale: nvdec will not produce odd width/height output, and while this is basically never an issue with most codecs, due to internal alignment requirements, you can get odd sized jpegs. If an odd-sized jpb is encountered, nvdec will actually round down

[FFmpeg-devel] [PATCH 1/4] configure: require external ffnvcodec headers

2017-12-04 Thread Timo Rothenpieler
The external headers can be found at https://git.videolan.org/?p=ffmpeg/nv-codec-headers.git --- configure | 6 +- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/configure b/configure index d05388638d..b4d82f1df6 100755 --- a/configure +++ b/configure @@ -2662,6 +2662,7 @@

[FFmpeg-devel] [PATCH 4/4] Changelog: mention nvidia headers move

2017-12-04 Thread Timo Rothenpieler
--- Changelog | 2 ++ 1 file changed, 2 insertions(+) diff --git a/Changelog b/Changelog index b00996157c..0b02a9aee7 100644 --- a/Changelog +++ b/Changelog @@ -24,6 +24,8 @@ version : - VAAPI VP8 decoding - AMD AMF H.264 and HEVC encoders - video fillborders filter +- Moved nvidia codec

[FFmpeg-devel] [PATCH 2/4] compat: use external nvidia headers

2017-12-04 Thread Timo Rothenpieler
--- compat/cuda/dynlink_loader.h | 7 --- libavcodec/nvenc.h | 2 +- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/compat/cuda/dynlink_loader.h b/compat/cuda/dynlink_loader.h index fa43782c9a..6e2a912a6e 100644 --- a/compat/cuda/dynlink_loader.h +++

Re: [FFmpeg-devel] [PATCH 1/4] configure: require external ffnvcodec headers

2017-12-04 Thread Timo Rothenpieler
Am 04.12.2017 um 20:58 schrieb Carl Eugen Hoyos: 2017-12-04 20:41 GMT+01:00 Timo Rothenpieler <t...@rothenpieler.org>: The external headers can be found at https://git.videolan.org/?p=ffmpeg/nv-codec-headers.git Afaict, 1, 2 and four should be merged. Could you explain (repea

Re: [FFmpeg-devel] order T-shirts

2017-12-18 Thread Timo Rothenpieler
Forgot to mention it here, but I got my shirt a while ago and it fits perfectly. Thank you very much! smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org

Re: [FFmpeg-devel] AMD external header

2017-12-15 Thread Timo Rothenpieler
Am 15.12.2017 um 15:27 schrieb Compn: On Fri, 15 Dec 2017 00:31:27 +, "Mironov, Mikhail" wrote: Do you think that two and half weeks is long enough to wait? So far except creating a repo for NVidia there is no activity. did you resend your patchset without the

Re: [FFmpeg-devel] AMD external header

2017-12-15 Thread Timo Rothenpieler
Do you think that two and half weeks is long enough to wait? So far except creating a repo for NVidia there is no activity. I'm not yet happy with how the external repository works, and I'm quite busy with other stuff at the moment. But it will get used eventually. smime.p7s Description:

Re: [FFmpeg-devel] [PATCH] compat: use external nvidia headers

2017-12-15 Thread Timo Rothenpieler
Am 15.12.2017 um 21:02 schrieb Michael Niedermayer: On Fri, Dec 15, 2017 at 02:13:56PM +0100, Timo Rothenpieler wrote: The external headers can be found at https://git.videolan.org/?p=ffmpeg/nv-codec-headers.git This installs headers without read permission i think normally install is used

Re: [FFmpeg-devel] [PATCH] configure: call flatten_extralibs in a subshell

2017-11-16 Thread Timo Rothenpieler
applied smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] avcodec: Implement mpeg2 nvdec hwaccel

2017-11-18 Thread Timo Rothenpieler
Am 16.11.2017 um 17:36 schrieb Philip Langdale: This is mostly straight-forward. The weird part is that it should just work for mpeg1, but I see corruption in my test cases, so I'm going to try and fix that separately. Signed-off-by: Philip Langdale --- Changelog

[FFmpeg-devel] [PATCH 1/4] avutil/frame: Add private_ref to AVFrame

2017-11-10 Thread Timo Rothenpieler
From: Michael Niedermayer This gives FFmpeg libs a field that they can freely and safely use. Avoiding the need of wrapping of a users opaque_ref field and its issues. Signed-off-by: Michael Niedermayer --- libavutil/frame.c | 8 +++-

[FFmpeg-devel] [PATCH 3/4] decode: add a mechanism for performing delayed processing on the decoded frames

2017-11-10 Thread Timo Rothenpieler
From: Anton Khirnov This will be useful in the CUVID hwaccel. Merges Libav commit badf0951f54c1332e77455dc40398f3512540c1b. --- libavcodec/decode.c | 15 +++ libavcodec/decode.h | 14 ++ 2 files changed, 29 insertions(+) diff --git

[FFmpeg-devel] [PATCH 2/4] decode: add a method for attaching lavc-internal data to frames

2017-11-10 Thread Timo Rothenpieler
From: Anton Khirnov Use the AVFrame.private_ref field. This new struct will be useful in the following commits. Merges Libav commit 359a8a3e2d1194b52b6c386f94fd0929567dfb67. --- libavcodec/decode.c | 51 ++--

[FFmpeg-devel] [PATCH 4/4] decode: add a per-frame private data for hwaccel use

2017-11-10 Thread Timo Rothenpieler
From: Anton Khirnov This will be useful in the CUVID hwaccel. It should also eventually replace current decoder-specific mechanisms used by various other hwaccels. Merges Libav commit 704311b2946d74a80f65906961cd9baaa18683a3. --- libavcodec/decode.c | 3 +++

[FFmpeg-devel] [PATCH 1/4] MAINTAINERS: update for cuvid/nvdec changes

2017-11-11 Thread Timo Rothenpieler
Philip Langdale cscd.cReimar Doeffinger - cuvid.c Timo Rothenpieler + cuviddec.cTimo Rothenpieler dca* foo86 dirac

[FFmpeg-devel] [PATCH 2/4] avcodec/nvdec: correctly set output frame size

2017-11-11 Thread Timo Rothenpieler
--- libavcodec/nvdec.c | 15 +-- 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/libavcodec/nvdec.c b/libavcodec/nvdec.c index db338accfa..780c2e67cf 100644 --- a/libavcodec/nvdec.c +++ b/libavcodec/nvdec.c @@ -232,8 +232,8 @@ int ff_nvdec_decode_init(AVCodecContext

[FFmpeg-devel] [PATCH 2/2] wmavoice: free frame before ff_get_buffer

2017-11-11 Thread Timo Rothenpieler
synth_superframe can be called twice per call to decode_packet. It is not fully clear if calling ff_get_buffer on the same frame twice is supported, so unref the frame first to be save. --- libavcodec/wmavoice.c | 4 1 file changed, 4 insertions(+) diff --git a/libavcodec/wmavoice.c

[FFmpeg-devel] [PATCH 1/2] avcodec/decode: always free private_ref

2017-11-11 Thread Timo Rothenpieler
There is no reason to keep this intact when decoding failed, specially as private_ref is supposed to always be NULL when a frame is returned to the user. --- libavcodec/decode.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/libavcodec/decode.c b/libavcodec/decode.c

[FFmpeg-devel] [PATCH 2/4] avcodec/nvdec: correctly set output frame size

2017-11-11 Thread Timo Rothenpieler
--- libavcodec/nvdec.c | 9 + 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/libavcodec/nvdec.c b/libavcodec/nvdec.c index db338accfa..894112a144 100644 --- a/libavcodec/nvdec.c +++ b/libavcodec/nvdec.c @@ -232,8 +232,8 @@ int ff_nvdec_decode_init(AVCodecContext *avctx,

Re: [FFmpeg-devel] [PATCH] ffmpeg: use explicitly requested hwaccel only

2017-11-11 Thread Timo Rothenpieler
Am 11.11.2017 um 13:02 schrieb Michael Niedermayer: On Fri, Nov 10, 2017 at 10:20:46PM +0100, Timo Rothenpieler wrote: With there being two hwaccels that use the CUDA pix_fmt now, just relying on the pix_fmt to identify the selected hwaccel is not enough anymore. So this checks if the user

Re: [FFmpeg-devel] [PATCH 1/2] avcodec/decode: always free private_ref

2017-11-11 Thread Timo Rothenpieler
series applied 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 3/4] avcodec/nvdec: add support for 12 bit formats

2017-11-11 Thread Timo Rothenpieler
--- libavcodec/nvdec.c | 18 +++--- 1 file changed, 15 insertions(+), 3 deletions(-) diff --git a/libavcodec/nvdec.c b/libavcodec/nvdec.c index 780c2e67cf..d3620f0d1f 100644 --- a/libavcodec/nvdec.c +++ b/libavcodec/nvdec.c @@ -234,9 +234,21 @@ int ff_nvdec_decode_init(AVCodecContext

[FFmpeg-devel] [PATCH 4/4] avcodec/nvdec: check hardware capabilities

2017-11-11 Thread Timo Rothenpieler
--- libavcodec/nvdec.c | 56 ++ 1 file changed, 56 insertions(+) diff --git a/libavcodec/nvdec.c b/libavcodec/nvdec.c index d3620f0d1f..d706540db0 100644 --- a/libavcodec/nvdec.c +++ b/libavcodec/nvdec.c @@ -74,6 +74,56 @@ static int

Re: [FFmpeg-devel] [PATCH] avcodec: implement vp9 nvdec hwaccel

2017-11-11 Thread Timo Rothenpieler
actual push will have update Changelog entry and bump minor version 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] avcodec: implement vp9 nvdec hwaccel

2017-11-11 Thread Timo Rothenpieler
/libavcodec/nvdec_vp9.c @@ -0,0 +1,225 @@ +/* + * VP9 HW decode acceleration through NVDEC + * + * Copyright (c) 2016 Timo Rothenpieler + * + * 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 + * L

[FFmpeg-devel] [PATCH] ffmpeg: use explicitly requested hwaccel only

2017-11-10 Thread Timo Rothenpieler
With there being two hwaccels that use the CUDA pix_fmt now, just relying on the pix_fmt to identify the selected hwaccel is not enough anymore. So this checks if the user explicitly selected a hwaccel, and only accepts that one. --- fftools/ffmpeg.c | 7 --- 1 file changed, 4 insertions(+),

Re: [FFmpeg-devel] [PATCH 2/4] avcodec/nvdec: correctly set output frame size

2017-11-12 Thread Timo Rothenpieler
This is actually not correct, as the classic hwaccels all set the cropping information properly on the frame. So not using the coded_width/height as frame size results in too much cropping. So just dropping this one. smime.p7s Description: S/MIME Cryptographic Signature

Re: [FFmpeg-devel] [PATCH 1/2] nvenc: factor context push/pop into functions

2017-11-12 Thread Timo Rothenpieler
Am 12.11.2017 um 10:30 schrieb Hendrik Leppkes: This reduces code repetition, and will allow adding further push/pop refinement for D3D11 devices in future commits. --- libavcodec/nvenc.c | 171 - 1 file changed, 76 insertions(+), 95

Re: [FFmpeg-devel] [PATCH 2/2] nvenc: support d3d11 surface input

2017-11-12 Thread Timo Rothenpieler
Am 12.11.2017 um 10:30 schrieb Hendrik Leppkes: --- libavcodec/nvenc.c | 106 ++--- libavcodec/nvenc.h | 11 +- 2 files changed, 95 insertions(+), 22 deletions(-) Don't have a setup to test D3D11VA with right now, but it does not break

Re: [FFmpeg-devel] [PATCH 3/4] avcodec/nvdec: add support for 12 bit formats

2017-11-12 Thread Timo Rothenpieler
Ideally, this should also set avctx->bits_per_raw_sample or an equivalent, if it exists. I would assume that the surrounding codec parser already sets that? A hwaccel does not usually set fields in avctx. smime.p7s Description: S/MIME Cryptographic Signature

[FFmpeg-devel] [PATCH] avcodec/decode: don't increase hw frame size

2017-11-12 Thread Timo Rothenpieler
In case the hw frames context does not return way too large frames, but slightly smaller ones(for example height being 1080, while coded_height is 1088), this causes failures because various code will try to read/write beyond the frame size. --- libavcodec/decode.c | 4 ++-- 1 file changed, 2

[FFmpeg-devel] [PATCH] configure: call flatten_extralibs in a subshell

2017-11-14 Thread Timo Rothenpieler
By putting the call in a subshell, the problem of it spilling cleanup-decision from a previous library to other libraries is avoided. For example, it could have already cleaned up cuda_extralibs in a previous library that depended on cuda. Then when it gets to avutil, it will never pick up the

Re: [FFmpeg-devel] [PATCH] avformat/fitsenc: validate input pixel format

2017-11-13 Thread Timo Rothenpieler
applied smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH 1/9] lavfi/paletteuse: check get_color return value

2017-11-13 Thread Timo Rothenpieler
applied smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] avcodec: implement vp9 nvdec hwaccel

2017-11-13 Thread Timo Rothenpieler
applied smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH 7/9] avfilter/vf_premultiply: fix memory-leak on failure

2017-11-13 Thread Timo Rothenpieler
applied smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] movenc-test: fix potential uninitialized read

2017-11-13 Thread Timo Rothenpieler
applied smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] configure: call flatten_extralibs in a subshell

2017-11-14 Thread Timo Rothenpieler
Am 14.11.2017 um 14:50 schrieb Timo Rothenpieler: By putting the call in a subshell, the problem of it spilling cleanup-decision from a previous library to other libraries is avoided. For example, it could have already cleaned up cuda_extralibs in a previous library that depended on cuda

Re: [FFmpeg-devel] [PATCH 1/3] avcodec: Refactor common nvdec hwaccel logic

2017-11-20 Thread Timo Rothenpieler
this version is good as well smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH 2/3] avcodec: Implement mpeg1 nvdec hwaccel

2017-11-20 Thread Timo Rothenpieler
nvdec part looks good again smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH 3/3] avcodec: Implement mpeg4 nvdec hwaccel

2017-11-20 Thread Timo Rothenpieler
Duplicate. Will remove. Can't really tell if the parameters are in any way correct, but I guess time will tell if some sample comes up that breaks it. lgtm with that duplicate removed smime.p7s Description: S/MIME Cryptographic Signature ___

Re: [FFmpeg-devel] [PATCH] avcodec: Implement mpeg1 nvdec hwaccel

2017-11-20 Thread Timo Rothenpieler
simple enough, lgtm smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] avcodec: Refactor common nvdec hwaccel logic

2017-11-20 Thread Timo Rothenpieler
lgtm smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH 09/13] compat/cuda: Pass a logging context to load functions

2017-11-18 Thread Timo Rothenpieler
lgtm smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] avcodec: Implement mpeg2 nvdec hwaccel

2017-11-18 Thread Timo Rothenpieler
Am 18.11.2017 um 17:38 schrieb Carl Eugen Hoyos: 2017-11-16 17:36 GMT+01:00 Philip Langdale : +AVHWAccel ff_mpeg2_nvdec_hwaccel = { +.name = "mpeg2_nvdec", +.type = AVMEDIA_TYPE_VIDEO, +.id =

Re: [FFmpeg-devel] Subscribe to ffmpeg-cvslog for patch discussion? Really?

2017-11-19 Thread Timo Rothenpieler
Usually if a discussion comes up the mail from cvslog is replied to on this list, so no actual discussion happens on the automatic cvslog list. smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list

Re: [FFmpeg-devel] Added HW H.264 and HEVC encoding for AMD GPUs based on AMF SDK

2017-11-21 Thread Timo Rothenpieler
Am 21.11.2017 um 16:32 schrieb Mironov, Mikhail: ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel Are you all busy right now? Any hint on timing? Thanks, Mikhail I cannot test this patch due to lack of hardware, but by now the code has been

Re: [FFmpeg-devel] [PATCH] configure: call flatten_extralibs in a subshell

2017-11-16 Thread Timo Rothenpieler
will push soon if nobody objects smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] [PATCH] hwcontext_d3d11va: add missing stdint.h include

2017-11-17 Thread Timo Rothenpieler
applied smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] On in-tree external headers

2017-11-05 Thread Timo Rothenpieler
Am 05.11.2017 um 19:59 schrieb Mark Thompson: On 05/11/17 18:42, Carl Eugen Hoyos wrote: 2017-11-05 19:35 GMT+01:00 Mark Thompson : On 05/11/17 18:28, Carl Eugen Hoyos wrote: 2017-11-05 15:24 GMT+01:00 Mark Thompson : On 30/10/17 19:51, Mark Thompson wrote:

Re: [FFmpeg-devel] On in-tree external headers

2017-11-05 Thread Timo Rothenpieler
How about: "No external headers may be added to the ffmpeg tree, unless they are for AviSynth or Nvidia." I don't think a strict "no external headers" rule makes sense or is a good idea at all. Specially if there are seemingly arbitrary exceptions. If such a rule is to be added at all, it

Re: [FFmpeg-devel] [PATCH] avutil/frame: Add avcodec_private_ref to AVFrame

2017-11-06 Thread Timo Rothenpieler
Am 05.11.2017 um 14:35 schrieb James Almer: On 11/5/2017 9:34 AM, Michael Niedermayer wrote: This gives libavcodec a field that it can freely and safely use. Avoiding the need of wraping of a users opaque_ref field and its issues. Could this perhaps be in an opaque internal struct instead,

Re: [FFmpeg-devel] [PATCH]lavf/tls_openssl: Fix compilation with current libressl

2017-11-07 Thread Timo Rothenpieler
Am 07.11.2017 um 13:28 schrieb Carl Eugen Hoyos: Hi! Attached patch fixes compilation with current libressl, related to ticket #6801. Please comment, Carl Eugen Wouldn't it be better to move that check to configure, so it does not build the openssl backend with libressl, and then merge the

Re: [FFmpeg-devel] [PATCH]lavf/tls_openssl: Fix compilation with current libressl

2017-11-07 Thread Timo Rothenpieler
I would very much rather have a way to get libressl to compile with tls_openssl.c, or just reject it altogether, than adding a duplicate module just for a fork that pretends to be compatible with a version of openssl but not providing the required API for it. I refuse to give special treatment

Re: [FFmpeg-devel] [PATCH]lavf/tls_openssl: Fix compilation with current libressl

2017-11-07 Thread Timo Rothenpieler
Am 07.11.2017 um 14:37 schrieb James Almer: On 11/7/2017 10:28 AM, Timo Rothenpieler wrote: I would very much rather have a way to get libressl to compile with tls_openssl.c, or just reject it altogether, than adding a duplicate module just for a fork that pretends to be compatible

[FFmpeg-devel] [PATCH 2/9] avformat/hlsenc: allocate space for terminating null

2017-11-08 Thread Timo Rothenpieler
Fixes CID #1420394 --- libavformat/hlsenc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavformat/hlsenc.c b/libavformat/hlsenc.c index 5ea9d216a4..b571772f60 100644 --- a/libavformat/hlsenc.c +++ b/libavformat/hlsenc.c @@ -1455,7 +1455,7 @@ static int

[FFmpeg-devel] [PATCH 7/9] avfilter/vf_premultiply: fix memory-leak on failure

2017-11-08 Thread Timo Rothenpieler
Fixes CID #1416352 --- libavfilter/vf_premultiply.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/libavfilter/vf_premultiply.c b/libavfilter/vf_premultiply.c index 1af6433bc9..4b6abc6e73 100644 --- a/libavfilter/vf_premultiply.c +++ b/libavfilter/vf_premultiply.c @@

[FFmpeg-devel] [PATCH 9/9] avfilter/signature_lookup: fix potential uninitialized reads

2017-11-08 Thread Timo Rothenpieler
Fixes CID #1403238 and #1403239 --- libavfilter/signature_lookup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavfilter/signature_lookup.c b/libavfilter/signature_lookup.c index 272c717c77..7de4a88ca2 100644 --- a/libavfilter/signature_lookup.c +++

[FFmpeg-devel] [PATCH 6/9] avfilter/vf_premultiply: make sure plane count is less than 4

2017-11-08 Thread Timo Rothenpieler
Fixes CID #1416350, #1416351, #1416353 and #1416354 --- libavfilter/vf_premultiply.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/libavfilter/vf_premultiply.c b/libavfilter/vf_premultiply.c index 5120adc476..1af6433bc9 100644 --- a/libavfilter/vf_premultiply.c +++

[FFmpeg-devel] [PATCH 3/9] lavc/vaapi_decode: fix uninitialized use of variables

2017-11-08 Thread Timo Rothenpieler
Fixes CID #1419524 and #1412855 --- libavcodec/vaapi_decode.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/libavcodec/vaapi_decode.c b/libavcodec/vaapi_decode.c index 27ef33837c..a782cfd8da 100644 --- a/libavcodec/vaapi_decode.c +++ b/libavcodec/vaapi_decode.c @@

[FFmpeg-devel] [PATCH 4/9] avfilter/af_haas: validate par_m_source parameter

2017-11-08 Thread Timo Rothenpieler
Fixes CID #1417663 --- libavfilter/af_haas.c | 1 + 1 file changed, 1 insertion(+) diff --git a/libavfilter/af_haas.c b/libavfilter/af_haas.c index 691c251f54..fac4b6cf2f 100644 --- a/libavfilter/af_haas.c +++ b/libavfilter/af_haas.c @@ -162,6 +162,7 @@ static int filter_frame(AVFilterLink

[FFmpeg-devel] [PATCH 0/9] Fix various issues found by coverity

2017-11-08 Thread Timo Rothenpieler
Timo Rothenpieler (9): lavfi/paletteuse: check get_color return value avformat/hlsenc: allocate space for terminating null lavc/vaapi_decode: fix uninitialized use of variables avfilter/af_haas: validate par_m_source parameter avformat/fitsenc: validate input pixel format avfilter

[FFmpeg-devel] [PATCH 1/9] lavfi/paletteuse: check get_color return value

2017-11-08 Thread Timo Rothenpieler
Fixes CID #1420396 --- libavfilter/vf_paletteuse.c | 5 - 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/libavfilter/vf_paletteuse.c b/libavfilter/vf_paletteuse.c index ed80ab04d5..1980907e70 100644 --- a/libavfilter/vf_paletteuse.c +++ b/libavfilter/vf_paletteuse.c @@ -380,8

[FFmpeg-devel] [PATCH 8/9] movenc-test: fix potential uninitialized read

2017-11-08 Thread Timo Rothenpieler
Fixes CID #1413023 --- libavformat/tests/movenc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavformat/tests/movenc.c b/libavformat/tests/movenc.c index 8e59b74259..a210aec7ba 100644 --- a/libavformat/tests/movenc.c +++ b/libavformat/tests/movenc.c @@ -108,7 +108,7 @@

[FFmpeg-devel] [PATCH 5/9] avformat/fitsenc: validate input pixel format

2017-11-08 Thread Timo Rothenpieler
Fixes CID #1416961 and #1416962 --- libavformat/fitsenc.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/libavformat/fitsenc.c b/libavformat/fitsenc.c index 7cb171596c..91411f2606 100644 --- a/libavformat/fitsenc.c +++ b/libavformat/fitsenc.c @@ -106,6 +106,8 @@ static int

Re: [FFmpeg-devel] [PATCH] avutil/frame: Add avcodec_private_ref to AVFrame

2017-11-06 Thread Timo Rothenpieler
I would prefer if this field would not be library-specific, but perhaps just "private_ref" which is not allowed to be touched by users, and documented to only be valid while within one library - ie. if you pass a frame from avcodec to avfilter, avfilter could take over the field (and just free

Re: [FFmpeg-devel] [PATCH] avcodec/nvdec: Fix capability check with old drivers.

2017-12-08 Thread Timo Rothenpieler
lgtm, feel free to push or remind me to do so if I forget smime.p7s Description: S/MIME Cryptographic Signature ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

<    1   2   3   4   5   6   7   >