summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorMarek Szuba <marecki@gentoo.org>2023-03-30 23:37:21 +0100
committerMarek Szuba <marecki@gentoo.org>2023-03-30 23:41:42 +0100
commita0035c9a825799ec5df594b8b7e3184fbf8cdd4f (patch)
treedc858e67e5617327ebf02fdc2eadd566ee1fb609 /dev-util/Tensile/Tensile-5.4.2-r2.ebuild
parentdev-cpp/msgpack-cxx: add 6.0.0 (diff)
downloadgentoo-a0035c9a825799ec5df594b8b7e3184fbf8cdd4f.tar.gz
gentoo-a0035c9a825799ec5df594b8b7e3184fbf8cdd4f.tar.bz2
gentoo-a0035c9a825799ec5df594b8b7e3184fbf8cdd4f.zip
dev-util/Tensile: adapt for msgpack-cxx-6.0.0
The name of the cmake module has changed again. Signed-off-by: Marek Szuba <marecki@gentoo.org>
Diffstat (limited to 'dev-util/Tensile/Tensile-5.4.2-r2.ebuild')
-rw-r--r--dev-util/Tensile/Tensile-5.4.2-r2.ebuild119
1 files changed, 119 insertions, 0 deletions
diff --git a/dev-util/Tensile/Tensile-5.4.2-r2.ebuild b/dev-util/Tensile/Tensile-5.4.2-r2.ebuild
new file mode 100644
index 000000000000..c846dce60997
--- /dev/null
+++ b/dev-util/Tensile/Tensile-5.4.2-r2.ebuild
@@ -0,0 +1,119 @@
+# Copyright 1999-2023 Gentoo Authors
+# Distributed under the terms of the GNU General Public License v2
+
+EAPI=8
+
+PYTHON_COMPAT=( python3_{10..11} )
+DISTUTILS_USE_PEP517=setuptools
+ROCM_VERSION=${PV}
+inherit cmake distutils-r1 llvm prefix rocm
+
+LLVM_MAX_SLOT=15
+
+DESCRIPTION="Stretching GPU performance for GEMMs and tensor contractions"
+HOMEPAGE="https://github.com/ROCmSoftwarePlatform/Tensile"
+SRC_URI="https://github.com/ROCmSoftwarePlatform/Tensile/archive/rocm-${PV}.tar.gz -> rocm-Tensile-${PV}.tar.gz"
+S="${WORKDIR}/${PN}-rocm-${PV}"
+
+LICENSE="MIT"
+KEYWORDS="~amd64"
+SLOT="0/$(ver_cut 1-2)"
+IUSE="client"
+REQUIRED_USE="client? ( ${ROCM_REQUIRED_USE} )"
+
+# Not compatible with recent versions of pytest
+RESTRICT="test"
+
+RDEPEND="${PYTHON_DEPS}
+ >=dev-cpp/msgpack-cxx-6.0.0
+ dev-python/pyyaml[${PYTHON_USEDEP}]
+ dev-python/msgpack[${PYTHON_USEDEP}]
+ dev-util/hip
+ >=dev-util/rocm-smi-4.3.0
+ sys-devel/clang:${LLVM_MAX_SLOT}
+"
+DEPEND="${RDEPEND}"
+
+PATCHES=(
+ "${FILESDIR}"/${PN}-change-cmake-name-for-msgpack-cxx-6-release.patch
+ "${FILESDIR}"/${PN}-4.3.0-output-commands.patch
+ "${FILESDIR}"/${PN}-5.4.2-gfx1031.patch
+ "${FILESDIR}"/${PN}-5.4.2-fix-arch-parse.patch
+ "${FILESDIR}"/${PN}-5.4.2-use-ninja.patch
+)
+
+CMAKE_USE_DIR="${S}/${PN}/Source"
+
+src_prepare() {
+ distutils-r1_src_prepare
+ sed -e "s,\@LLVM_PATH\@,$(get_llvm_prefix ${LLVM_MAX_SLOT}),g" \
+ "${FILESDIR}"/${PN}-5.4.2-gentoopath.patch > "${S}"/gentoopath.patch || die
+ eapply $(prefixify_ro "${S}"/gentoopath.patch)
+
+ pushd ${PN} || die
+
+ sed -e "/ROCM_SMI_ROOT/s,lib,$(get_libdir)," \
+ -i Source/cmake/FindROCmSMI.cmake || die
+ sed -r -e "/TENSILE_USE_LLVM/s/ON/OFF/" \
+ -i Source/CMakeLists.txt || die
+ sed -e "/chmod 755/d" -i Source/TensileCreateLibrary.cmake || die # remove chmod 755 on
+
+ # ${Tensile_ROOT}/bin does not exists; call command directly
+ sed -e "s,\${Tensile_ROOT}/bin/,,g" -i Source/TensileCreateLibrary.cmake cmake/TensileConfig.cmake || die
+
+ local Tensile_share_dir="\"${EPREFIX}/usr/share/${PN}\""
+ sed -e "/HipClangVersion/s/0.0.0/$(hipconfig -v)/" -i Common.py || die
+
+ sed -e "s,os.path.dirname(os.path.realpath(__file__)),${Tensile_share_dir},g" \
+ -i ReplacementKernels.py Common.py ${PN}.py || die
+
+ sed -e "s|os\.path\.dirname.*$|\"${EPREFIX}/usr/share/Tensile/Source\", end='')|" -i __init__.py || die
+
+ popd || die
+
+ sed -e "/package_data/d" -e "/data_files/d" -i setup.py || die
+ use client && PATCHES= cmake_src_prepare # do not apply patches again in cmake_src_prepare
+}
+
+src_configure() {
+ distutils-r1_src_configure
+ if use client; then
+ local mycmakeargs=(
+ -DCMAKE_SKIP_RPATH=ON
+ -DTENSILE_USE_MSGPACK=ON
+ -DTENSILE_USE_LLVM=ON
+ -DTensile_LIBRARY_FORMAT=msgpack
+ -DAMDGPU_TARGETS="$(get_amdgpu_flags)"
+ )
+ CXX=hipcc cmake_src_configure
+ fi
+}
+
+src_compile() {
+ distutils-r1_src_compile
+ use client && cmake_src_compile
+}
+
+python_install() {
+ distutils-r1_python_install
+
+ python_moduleinto Tensile
+ pushd Tensile || die
+ python_domodule Components
+ python_newexe Utilities/merge.py ${PN}-merge
+}
+
+src_install() {
+ distutils-r1_src_install
+
+ pushd ${PN} || die
+ insinto /usr/share/${PN}
+ doins -r Configs Perf ReplacementKernels ReplacementKernels-cov3 Source CustomKernels
+ insinto /usr/$(get_libdir)/cmake/${PN}
+ doins cmake/*.cmake
+
+ if use client; then
+ pushd "${BUILD_DIR}" || die
+ dobin client/tensile_client
+ fi
+}