summaryrefslogtreecommitdiffstats
path: root/configure
diff options
context:
space:
mode:
authorRodger Combs <rodger.combs@gmail.com>2019-07-30 02:51:42 -0500
committerTimo Rothenpieler <timo@rothenpieler.org>2019-08-04 19:08:08 +0200
commit01994c93db43e50c01349203a76c0b7111d1d2f3 (patch)
tree341dfdb3daff9907d2d0c7087c4ba43ae7bbd9a2 /configure
parentad97be9fa30d98c0946ee97ffbff3dfdb7831bf9 (diff)
downloadffmpeg-streaming-01994c93db43e50c01349203a76c0b7111d1d2f3.zip
ffmpeg-streaming-01994c93db43e50c01349203a76c0b7111d1d2f3.tar.gz
build: add support for building CUDA files with clang
This avoids using the CUDA SDK at all; instead, we provide a minimal reimplementation of the basic functionality that lavfi actually uses. It generates very similar code to what NVCC produces. The header contains no implementation code derived from the SDK. The function and type declarations are derived from the SDK only to the extent required to build a compatible implementation. This is generally accepted to qualify as fair use. Because this option does not require the proprietary SDK, it does not require the "--enable-nonfree" flag in configure. Signed-off-by: Timo Rothenpieler <timo@rothenpieler.org>
Diffstat (limited to 'configure')
-rwxr-xr-xconfigure67
1 files changed, 45 insertions, 22 deletions
diff --git a/configure b/configure
index 5a4f507..bddc05b 100755
--- a/configure
+++ b/configure
@@ -322,6 +322,7 @@ External library support:
--disable-amf disable AMF video encoding code [autodetect]
--disable-audiotoolbox disable Apple AudioToolbox code [autodetect]
--enable-cuda-nvcc enable Nvidia CUDA compiler [no]
+ --disable-cuda-llvm disable CUDA compilation using clang [autodetect]
--disable-cuvid disable Nvidia CUVID support [autodetect]
--disable-d3d11va disable Microsoft Direct3D 11 video acceleration code [autodetect]
--disable-dxva2 disable Microsoft DirectX 9 video acceleration code [autodetect]
@@ -370,7 +371,7 @@ Toolchain options:
--cxx=CXX use C compiler CXX [$cxx_default]
--objcc=OCC use ObjC compiler OCC [$cc_default]
--dep-cc=DEPCC use dependency generator DEPCC [$cc_default]
- --nvcc=NVCC use Nvidia CUDA compiler NVCC [$nvcc_default]
+ --nvcc=NVCC use Nvidia CUDA compiler NVCC or clang [$nvcc_default]
--ld=LD use linker LD [$ld_default]
--pkg-config=PKGCONFIG use pkg-config tool PKGCONFIG [$pkg_config_default]
--pkg-config-flags=FLAGS pass additional flags to pkgconf []
@@ -1038,12 +1039,16 @@ test_nvcc(){
tmpcu_=$TMPCU
tmpo_=$TMPO
[ -x "$(command -v cygpath)" ] && tmpcu_=$(cygpath -m $tmpcu_) && tmpo_=$(cygpath -m $tmpo_)
- test_cmd $nvcc -ptx $NVCCFLAGS "$@" $NVCC_C $(nvcc_o $tmpo_) $tmpcu_
+ test_cmd $nvcc $nvccflags "$@" $NVCC_C $(nvcc_o $tmpo_) $tmpcu_
}
check_nvcc() {
log check_nvcc "$@"
- test_nvcc <<EOF
+ name=$1
+ shift 1
+ disabled $name && return
+ disable $name
+ test_nvcc "$@" <<EOF && enable $name
extern "C" {
__global__ void hello(unsigned char *data) {}
}
@@ -1814,6 +1819,7 @@ HWACCEL_AUTODETECT_LIBRARY_LIST="
audiotoolbox
crystalhd
cuda
+ cuda_llvm
cuvid
d3d11va
dxva2
@@ -2987,8 +2993,10 @@ v4l2_m2m_deps="linux_videodev2_h sem_timedwait"
hwupload_cuda_filter_deps="ffnvcodec"
scale_npp_filter_deps="ffnvcodec libnpp"
-scale_cuda_filter_deps="ffnvcodec cuda_nvcc"
-thumbnail_cuda_filter_deps="ffnvcodec cuda_nvcc"
+scale_cuda_filter_deps="ffnvcodec"
+scale_cuda_filter_deps_any="cuda_nvcc cuda_llvm"
+thumbnail_cuda_filter_deps="ffnvcodec"
+thumbnail_cuda_filter_deps_any="cuda_nvcc cuda_llvm"
transpose_npp_filter_deps="ffnvcodec libnpp"
amf_deps_any="libdl LoadLibrary"
@@ -3547,7 +3555,8 @@ zscale_filter_deps="libzimg const_nan"
scale_vaapi_filter_deps="vaapi"
vpp_qsv_filter_deps="libmfx"
vpp_qsv_filter_select="qsvvpp"
-yadif_cuda_filter_deps="ffnvcodec cuda_nvcc"
+yadif_cuda_filter_deps="ffnvcodec"
+yadif_cuda_filter_deps_any="cuda_nvcc cuda_llvm"
# examples
avio_dir_cmd_deps="avformat avutil"
@@ -3651,8 +3660,6 @@ version_script='--version-script'
objformat="elf32"
x86asmexe_default="nasm"
windres_default="windres"
-nvcc_default="nvcc"
-nvccflags_default="-gencode arch=compute_30,code=sm_30 -O2"
striptype="direct"
# OS
@@ -4220,6 +4227,20 @@ windres_default="${cross_prefix}${windres_default}"
sysinclude_default="${sysroot}/usr/include"
+if enabled cuda_sdk; then
+ warn "Option --enable-cuda-sdk is deprecated. Use --enable-cuda-nvcc instead."
+ enable cuda_nvcc
+fi
+
+if enabled cuda_nvcc; then
+ nvcc_default="nvcc"
+ nvccflags_default="-gencode arch=compute_30,code=sm_30 -O2"
+else
+ nvcc_default="clang"
+ nvccflags_default="--cuda-gpu-arch=sm_30 -O2"
+ NVCC_C=""
+fi
+
set_default arch cc cxx doxygen pkg_config ranlib strip sysinclude \
target_exec x86asmexe nvcc
enabled cross_compile || host_cc_default=$cc
@@ -6060,9 +6081,21 @@ check_type "d3d9.h dxva2api.h" DXVA2_ConfigPictureDecode -D_WIN32_WINNT=0x0602
check_type "vdpau/vdpau.h" "VdpPictureInfoHEVC"
-if enabled cuda_sdk; then
- warn "Option --enable-cuda-sdk is deprecated. Use --enable-cuda-nvcc instead."
- enable cuda_nvcc
+if [ -z "$nvccflags" ]; then
+ nvccflags=$nvccflags_default
+fi
+
+if enabled x86_64 || enabled ppc64 || enabled aarch64; then
+ nvccflags="$nvccflags -m64"
+else
+ nvccflags="$nvccflags -m32"
+fi
+
+if enabled cuda_nvcc; then
+ nvccflags="$nvccflags -ptx"
+else
+ nvccflags="$nvccflags -S -nocudalib -nocudainc --cuda-device-only -include${source_link}/compat/cuda/cuda_runtime.h"
+ check_nvcc cuda_llvm
fi
if ! disabled ffnvcodec; then
@@ -6140,7 +6173,7 @@ for func in $COMPLEX_FUNCS; do
done
# these are off by default, so fail if requested and not available
-enabled cuda_nvcc && { check_nvcc || die "ERROR: failed checking for nvcc."; }
+enabled cuda_nvcc && { check_nvcc cuda_nvcc || die "ERROR: failed checking for nvcc."; }
enabled chromaprint && require chromaprint chromaprint.h chromaprint_get_version -lchromaprint
enabled decklink && { require_headers DeckLinkAPI.h &&
{ test_cpp_condition DeckLinkAPIVersion.h "BLACKMAGIC_DECKLINK_API_VERSION >= 0x0a090500" || die "ERROR: Decklink API version must be >= 10.9.5."; } }
@@ -6701,16 +6734,6 @@ if [ -z "$optflags" ]; then
fi
fi
-if [ -z "$nvccflags" ]; then
- nvccflags=$nvccflags_default
-fi
-
-if enabled x86_64 || enabled ppc64 || enabled aarch64; then
- nvccflags="$nvccflags -m64"
-else
- nvccflags="$nvccflags -m32"
-fi
-
check_optflags(){
check_cflags "$@"
enabled lto && check_ldflags "$@"
OpenPOWER on IntegriCloud