I'd like to use NvPipe with the x264 FFMPEG backend to support visualization clients with pre-Kepler hardware. Unfortunately, NvPipe master fails to compile as soon as I enable USE_FFMPEG:
NvPipe/build-pristine$ cmake -DUSE_FFMPEG=ON -DCUDA_TOOLKIT_ROOT_DIR=/opt/cuda/7.5 ..
-- The C compiler identification is GNU 4.9.2
-- The CXX compiler identification is GNU 4.9.2
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found CUDA: /opt/cuda/7.5 (found version "7.5")
-- Found PkgConfig: /usr/bin/pkg-config (found version "0.28")
-- Checking for module 'libavformat'
-- Found libavformat, version 57.56.101
-- Checking for module 'libavcodec'
-- Found libavcodec, version 57.64.101
-- Checking for module 'libavutil'
-- Found libavutil, version 55.34.101
-- Could NOT find EGL (missing: EGL_LIBRARY EGL_opengl_LIBRARY EGL_INCLUDE_DIR)
-- Found GLEW: /usr/include
-- Configuring done
-- Generating done
-- Build files have been written to: NvPipe/build-pristine
NvPipe/build-pristine$ make
[ 7%] Building NVCC (Device) object CMakeFiles/nvpipe.dir/convert.o
[ 15%] Building NVCC (Device) object CMakeFiles/nvpipe.dir/util/kernels.o
Scanning dependencies of target nvpipe
[ 23%] Building CXX object CMakeFiles/nvpipe.dir/ffmpeg.cxx.o
/home/r/build/nvpipe/NvPipe/ffmpeg.cxx: In function ‘nvp_impl_t* nvp_create_ffmpeg(bool, uint64_t)’:
/home/r/build/nvpipe/NvPipe/ffmpeg.cxx:135:21: error: invalid conversion from ‘nvp_err_t (*)(nvpipe*, const void*, size_t, void*, size_t*, size_t, size_t, nvp_fmt_t) {aka nvpipe_error_code (*)(void*, const void*, long unsigned int, void*, long unsigned int*, long unsigned int, long unsigned int, nvpipe_format)}’ to ‘nvp_err_t (*)(nvpipe*, const void*, size_t, void*, size_t*, uint32_t, uint32_t, nvp_fmt_t) {aka nvpipe_error_code (*)(void*, const void*, long unsigned int, void*, long unsigned int*, unsigned int, unsigned int, nvpipe_format)}’ [-fpermissive]
rv->impl.encode = nvp_ffmpeg_encode;
^
/home/r/build/nvpipe/NvPipe/ffmpeg.cxx:137:21: error: invalid conversion from ‘nvp_err_t (*)(nvpipe*, const void*, size_t, void*, size_t, size_t) {aka nvpipe_error_code (*)(void*, const void*, long unsigned int, void*, long unsigned int, long unsigned int)}’ to ‘nvp_err_t (*)(nvpipe*, const void*, size_t, void*, uint32_t, uint32_t, nvp_fmt_t) {aka nvpipe_error_code (*)(void*, const void*, long unsigned int, void*, unsigned int, unsigned int, nvpipe_format)}’ [-fpermissive]
rv->impl.decode = nvp_ffmpeg_decode;
^
CMakeFiles/nvpipe.dir/build.make:76: recipe for target 'CMakeFiles/nvpipe.dir/ffmpeg.cxx.o' failed
make[2]: *** [CMakeFiles/nvpipe.dir/ffmpeg.cxx.o] Error 1
CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/nvpipe.dir/all' failed
make[1]: *** [CMakeFiles/nvpipe.dir/all] Error 2
Makefile:127: recipe for target 'all' failed
make: *** [all] Error 2
diff --git a/decode.c b/decode.c
index c5949fc..94f5133 100644
--- a/decode.c
+++ b/decode.c
@@ -351,7 +351,7 @@ nvp_cuvid_decode(nvpipe* const cdc,
const void* const __restrict ibuf,
const size_t ibuf_sz,
void* const __restrict obuf,
- const uint32_t width, const uint32_t height,
+ const size_t width, const size_t height,
nvp_fmt_t format) {
struct nvp_decoder* nvp = (struct nvp_decoder*)cdc;
if(nvp->impl.type != DECODER) {
@@ -485,7 +485,7 @@ nvp_cuvid_encode(nvpipe * const __restrict codec,
const size_t ibuf_sz,
void *const __restrict obuf,
size_t* const __restrict obuf_sz,
- const uint32_t width, const uint32_t height,
+ const size_t width, const size_t height,
nvp_fmt_t format) {
(void)codec; (void)ibuf; (void)ibuf_sz;
(void)obuf; (void)obuf_sz;
diff --git a/ffmpeg.cxx b/ffmpeg.cxx
index 8d40b17..dc38bca 100644
--- a/ffmpeg.cxx
+++ b/ffmpeg.cxx
@@ -94,7 +94,8 @@ nvp_ffmpeg_decode(nvpipe* const __restrict cdc,
const void* const __restrict input_buffer,
const size_t input_buffer_size,
void* const __restrict output_buffer,
- size_t width, size_t height) {
+ size_t width, size_t height,
+ const nvp_fmt_t format) {
assert(cdc);
/* input images must be a power of two */
if(((width | height) & 1) != 0) {
@@ -104,6 +105,12 @@ nvp_ffmpeg_decode(nvpipe* const __restrict cdc,
if(input_buffer_size == 0) {
return NVPIPE_EINVAL;
}
+
+ /* only RGB input is supported */
+ if(format != NVPIPE_RGB) {
+ return NVPIPE_EINVAL;
+ }
+
nvpipe_* codec = static_cast<nvpipe_*>(cdc);
NvPipeCodec *codec_ptr = static_cast<NvPipeCodec*>(codec->codec_ptr_);
diff --git a/internal-api.h b/internal-api.h
index d0af9df..00c6397 100644
--- a/internal-api.h
+++ b/internal-api.h
@@ -40,7 +40,7 @@ typedef nvp_err_t (fqn_encode)(
const size_t ibuf_sz,
void *const __restrict obuf,
size_t* const __restrict obuf_sz,
- const uint32_t width, const uint32_t height,
+ const size_t width, const size_t height,
nvp_fmt_t format
);
typedef nvp_err_t (fqn_bitrate)(nvpipe* codec, uint64_t);
@@ -48,8 +48,8 @@ typedef nvp_err_t (fqn_decode)(
nvpipe* const __restrict codec,
const void* const __restrict ibuf, const size_t ibuf_sz,
void* const __restrict obuf,
- const uint32_t width, const uint32_t height,
- nvp_fmt_t format
+ const size_t width, const size_t height,
+ const nvp_fmt_t format
);
typedef void (fqn_destroy)(nvpipe* const __restrict);
After the above changes NvPipe compiles but an NVPIPE_H264_FFMPEG decoder fails to decode frames encoded by NVPIPE_H264_NV. I believe that's a separate issue (which I'll submit shortly) but I'd still appreciate it if the compilation issue could be fixed properly in NvPipe upstream.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent