diff options
| author | Daniel Bermond | 2022-05-14 10:43:13 -0300 |
|---|---|---|
| committer | Daniel Bermond | 2022-05-14 10:43:13 -0300 |
| commit | 58ff1bc45bd85ec7e0967e8e3a66607602301bff (patch) | |
| tree | bb2fae1add80ed3811df172f134d5093cd14c16e | |
| parent | d31d186ce364e9e50b9319f63fb7d816b92f9e76 (diff) | |
| download | aur-58ff1bc45bd85ec7e0967e8e3a66607602301bff.tar.gz | |
Update to version 8.2.5.1
| -rw-r--r-- | .SRCINFO | 10 | ||||
| -rw-r--r-- | PKGBUILD | 11 |
2 files changed, 10 insertions, 11 deletions
@@ -1,7 +1,7 @@ pkgbase = tensorrt pkgdesc = A platform for high-performance deep learning inference on NVIDIA hardware - pkgver = 8.2.4.2 - pkgrel = 2 + pkgver = 8.2.5.1 + pkgrel = 1 url = https://developer.nvidia.com/tensorrt/ arch = x86_64 license = custom:NVIDIA-SLA @@ -18,8 +18,8 @@ pkgbase = tensorrt makedepends = python-setuptools makedepends = python-wheel noextract = protobuf-cpp-3.17.3.tar.gz - source = local://TensorRT-8.2.4.2.Linux.x86_64-gnu.cuda-11.4.cudnn8.2.tar.gz - source = git+https://github.com/NVIDIA/TensorRT.git#tag=22.04 + source = local://TensorRT-8.2.5.1.Linux.x86_64-gnu.cuda-11.4.cudnn8.2.tar.gz + source = git+https://github.com/NVIDIA/TensorRT.git#tag=22.05 source = protobuf-protocolbuffers::git+https://github.com/protocolbuffers/protobuf.git source = cub-nvlabs::git+https://github.com/NVlabs/cub.git source = git+https://github.com/onnx/onnx-tensorrt.git @@ -29,7 +29,7 @@ pkgbase = tensorrt source = https://github.com/google/protobuf/releases/download/v3.17.3/protobuf-cpp-3.17.3.tar.gz source = 010-tensorrt-use-local-protobuf-sources.patch source = 020-tensorrt-fix-python.patch - sha256sums = 826180eaaecdf9a7e76116855b9f1f3400ea9b06e66b06a3f6a0747ba6f863ad + sha256sums = 2dc36962f0e24fb01abeaccad4c4658a3a36a8337b8f39e281e4bbed44bbd771 sha256sums = SKIP sha256sums = SKIP sha256sums = SKIP @@ -9,15 +9,15 @@ pkgbase=tensorrt pkgname=('tensorrt' 'python-tensorrt' 'tensorrt-doc') -pkgver=8.2.4.2 -_ossver=22.04 +pkgver=8.2.5.1 +_ossver=22.05 _cudaver=11.4 _cudnnver=8.2 _protobuf_ver=3.17.3 _pybind11_ver=2.9.2 _graphsurgeon_ver=0.4.5 _uffver=0.6.9 -pkgrel=2 +pkgrel=1 pkgdesc='A platform for high-performance deep learning inference on NVIDIA hardware' arch=('x86_64') url='https://developer.nvidia.com/tensorrt/' @@ -36,7 +36,7 @@ source=("local://TensorRT-${pkgver}.Linux.${CARCH}-gnu.cuda-${_cudaver}.cudnn${_ '010-tensorrt-use-local-protobuf-sources.patch' '020-tensorrt-fix-python.patch') noextract=("protobuf-cpp-${_protobuf_ver}.tar.gz") -sha256sums=('826180eaaecdf9a7e76116855b9f1f3400ea9b06e66b06a3f6a0747ba6f863ad' +sha256sums=('2dc36962f0e24fb01abeaccad4c4658a3a36a8337b8f39e281e4bbed44bbd771' 'SKIP' 'SKIP' 'SKIP' @@ -79,14 +79,13 @@ prepare() { build() { export CXXFLAGS+=' -ffat-lto-objects' - export LDFLAGS+=' -L/usr/lib' cmake -B build -S TensorRT \ -DBUILD_ONNX_PYTHON:BOOL='ON' \ -DBUILD_SAMPLES:BOOL='OFF' \ -DCMAKE_BUILD_TYPE:STRING='None' \ -DCMAKE_INSTALL_PREFIX:PATH='/usr' \ -DTRT_LIB_DIR="${srcdir}/TensorRT-${pkgver}/lib" \ - -DGPU_ARCHS='52 53 60 61 62 70 72 75 80 86' \ + -DGPU_ARCHS='53 60 61 62 70 72 75 80 86' \ -DPROTOBUF_VERSION="$_protobuf_ver" \ -DCUDA_VERSION="$_cudaver" \ -DCUDNN_VERSION="$_cudnnver" \ |