dev-util/Tensile: bump version to 5.0.2

Package-Manager: Portage-3.0.30, Repoman-3.0.3
Signed-off-by: Yiyang Wu <xgreenlandforwyy@gmail.com>
Signed-off-by: Benda Xu <heroxbd@gentoo.org>

6 files changed, 297 insertions, 0 deletions

diff --git a/dev-util/Tensile/Manifest b/dev-util/Tensile/Manifest
index fd4cac044af8..c3e8f48bab20 100644
--- a/dev-util/Tensile/Manifest
+++ b/dev-util/Tensile/Manifest
@@ -1 +1,3 @@
+DIST Tensile-5.0.2-PR1419.patch.gz 43754 BLAKE2B 3b1d7333ed0115f77816806e7735e25d897dbae3f7a40ece61d655c8f016d9d1c804b586b4e0e75661e20c517f5937a61d5ead25a545f0ce2e4a178bcbb3e7be SHA512 0154d8e14d335f4ecd86e5db335289f8662cede7e5d4b93bf6c56d8e48c1d7d0457271c078c9abbbca1f4aa91f028f6ce6a89efb783ba0207d0f39fd5de8a42c
 DIST rocm-Tensile-4.3.0.tar.gz 14250149 BLAKE2B 030138eaca2a0aadd96801e6dbd72e510716dd90553ef3795c5e04e00a34a05ecae82b24f755e4033a4acfcdb1cf26291da1e7902bb090f89a010d403e832beb SHA512 126db0b413c716fba8d5be9bff7a44fd1badacbf32f3db8d0db649819177db37ebd56fd22dd3c809655f5d29675be115e698cd10bc3d0b4b23878ae3726fce47
+DIST rocm-Tensile-5.0.2.tar.gz 14362090 BLAKE2B d44c3135496db41c4eb2545f6054c84493e89ef6357da881dae6859763338417e7a9e142cdce7e2de0536684bb641049ad0440964adbe68d03ca329f01efabc9 SHA512 f5ab7d7cbc5ba25498a02a6b7c07d5a9a548edcd44b439b92c8718072382adc18ae78274db1a7174e06436ac1396f3b4edafc7f491a2c87086c368e7270e64e3
diff --git a/dev-util/Tensile/Tensile-5.0.2.ebuild b/dev-util/Tensile/Tensile-5.0.2.ebuild
new file mode 100644
index 000000000000..1210cd769aa5
--- /dev/null
+++ b/dev-util/Tensile/Tensile-5.0.2.ebuild
@@ -0,0 +1,83 @@
+# Copyright 1999-2022 Gentoo Authors
+# Distributed under the terms of the GNU General Public License v2
+
+EAPI=8
+
+PYTHON_COMPAT=( python3_{8..10} )
+DISTUTILS_USE_PEP517=setuptools
+inherit distutils-r1 prefix
+
+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
+		https://github.com/littlewu2508/littlewu2508.github.io/raw/main/gentoo-distfiles/${P}-PR1419.patch.gz"
+S="${WORKDIR}/${PN}-rocm-${PV}"
+
+LICENSE="MIT"
+KEYWORDS="~amd64"
+SLOT="0/$(ver_cut 1-2)"
+
+# Not compatible with recent versions of pytest
+RESTRICT="test"
+
+RDEPEND="${PYTHON_DEPS}
+	dev-python/pyyaml[${PYTHON_USEDEP}]
+	dev-python/msgpack[${PYTHON_USEDEP}]
+	>=dev-util/rocm-smi-4.3.0
+"
+DEPEND="${RDEPEND}
+	dev-util/hip:${SLOT}
+"
+
+PATCHES=( "${FILESDIR}"/${PN}-4.3.0-output-commands.patch
+		  "${FILESDIR}"/${PN}-5.0.2-gfx1031.patch
+		  "${FILESDIR}"/${PN}-5.0.2-fix-arch-parse.patch
+		  "${FILESDIR}"/${PN}-5.0.2-use-ninja.patch
+		  "${WORKDIR}"/${PN}-5.0.2-PR1419.patch
+	  )
+
+CMAKE_USE_DIR="${WORKDIR}/Source"
+
+src_prepare() {
+	distutils-r1_src_prepare
+	eapply $(prefixify_ro "${FILESDIR}"/${PN}-5.0.2-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
+	sed -e "s,\${Tensile_ROOT}/bin/,,g" -i Source/TensileCreateLibrary.cmake cmake/TensileConfig.cmake || die # ${Tensile_ROOT}/bin does not exists; call command directly
+
+	local Tensile_share_dir="\"${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
+}
+
+python_install() {
+	distutils-r1_python_install
+
+	python_moduleinto Tensile
+	pushd Tensile
+	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
+}
diff --git a/dev-util/Tensile/files/Tensile-5.0.2-fix-arch-parse.patch b/dev-util/Tensile/files/Tensile-5.0.2-fix-arch-parse.patch
new file mode 100644
index 000000000000..47bee8819c44
--- /dev/null
+++ b/dev-util/Tensile/files/Tensile-5.0.2-fix-arch-parse.patch
@@ -0,0 +1,12 @@
+Index: Tensile-rocm-5.0.1/Tensile/Tensile.py
+===================================================================
+--- Tensile-rocm-5.0.1.orig/Tensile/Tensile.py
++++ Tensile-rocm-5.0.1/Tensile/Tensile.py
+@@ -97,7 +97,6 @@ def addCommonArguments(argParser):
+     """
+ 
+     (key, value) = par.split("=")
+-    value = eval(value)
+     return (key, value)
+ 
+   argParser.add_argument("-d", "--device", dest="device", type=int, \
diff --git a/dev-util/Tensile/files/Tensile-5.0.2-gentoopath.patch b/dev-util/Tensile/files/Tensile-5.0.2-gentoopath.patch
new file mode 100644
index 000000000000..b8b59af5184e
--- /dev/null
+++ b/dev-util/Tensile/files/Tensile-5.0.2-gentoopath.patch
@@ -0,0 +1,75 @@
+Fix incorrect path in scripts and CMakeLists, in order to build tensile_client 
+===================================================================
+--- Tensile-rocm-5.0.2.orig/Tensile/Source/client/CMakeLists.txt
++++ Tensile-rocm-5.0.2/Tensile/Source/client/CMakeLists.txt
+@@ -56,3 +56,5 @@ endif()
+ foreach(arch IN LISTS TENSILE_GPU_ARCHS)
+     target_link_libraries(tensile_client PRIVATE "--amdgpu-target=${arch}")
+ endforeach(arch)
++
++target_link_libraries(tensile_client PRIVATE LLVMSupport)
+Index: Tensile-rocm-5.0.2/Tensile/Source/lib/CMakeLists.txt
+===================================================================
+--- Tensile-rocm-5.0.2.orig/Tensile/Source/lib/CMakeLists.txt
++++ Tensile-rocm-5.0.2/Tensile/Source/lib/CMakeLists.txt
+@@ -47,7 +47,7 @@ set(tensile_sources  ${tensile_sources}
+     )
+ 
+ if(TENSILE_USE_LLVM)
+-    find_package(LLVM 13.0 QUIET CONFIG)
++    find_package(LLVM PATHS @GENTOO_PORTAGE_EPREFIX@/usr/lib/llvm/roc QUIET CONFIG)
+     if(NOT LLVM_FOUND)
+         find_package(LLVM 12.0 QUIET CONFIG)
+         if(NOT LLVM_FOUND)
+Index: Tensile-rocm-5.0.2/Tensile/Common.py
+===================================================================
+--- Tensile-rocm-5.0.2.orig/Tensile/Common.py
++++ Tensile-rocm-5.0.2/Tensile/Common.py
+@@ -1873,7 +1873,7 @@ def assignGlobalParameters( config ):
+     else:
+       print2(" %24s: %8s (unspecified)" % (key, defaultValue))
+ 
+-  globalParameters["ROCmPath"] = "/opt/rocm"
++  globalParameters["ROCmPath"] = "@GENTOO_PORTAGE_EPREFIX@/usr"
+   if "ROCM_PATH" in os.environ:
+     globalParameters["ROCmPath"] = os.environ.get("ROCM_PATH")
+   if "TENSILE_ROCM_PATH" in os.environ:
+Index: Tensile-rocm-5.0.2/Tensile/ClientWriter.py
+===================================================================
+--- Tensile-rocm-5.0.2.orig/Tensile/ClientWriter.py
++++ Tensile-rocm-5.0.2/Tensile/ClientWriter.py
+@@ -205,7 +205,7 @@ def getBuildClientLibraryScript(buildPat
+   import io
+   runScriptFile = io.StringIO()
+ 
+-  callCreateLibraryCmd = globalParameters["ScriptPath"] + "/bin/TensileCreateLibrary"
++  callCreateLibraryCmd = "TensileCreateLibrary"
+ 
+ 
+   if globalParameters["MergeFiles"]:
+Index: Tensile-rocm-5.0.2/Tensile/GenerateSummations.py
+===================================================================
+--- Tensile-rocm-5.0.2.orig/Tensile/GenerateSummations.py
++++ Tensile-rocm-5.0.2/Tensile/GenerateSummations.py
+@@ -57,7 +57,7 @@ def createLibraryForBenchmark(logicPath,
+     Selection.
+     """
+ 
+-    pythonExePath = os.path.join(os.path.dirname(os.path.realpath(__file__)), "bin", "TensileCreateLibrary")
++    pythonExePath = "TensileCreateLibrary"
+     args = [pythonExePath, \
+         "--merge-files", "--new-client-only", "--no-short-file-names", "--no-library-print-debug", \
+         "--architecture=all", "--code-object-version=V3", "--cxx-compiler=hipcc", "--library-format=yaml", \
+Index: Tensile-rocm-5.0.2/HostLibraryTests/CMakeLists.txt
+===================================================================
+--- Tensile-rocm-5.0.2.orig/HostLibraryTests/CMakeLists.txt
++++ Tensile-rocm-5.0.2/HostLibraryTests/CMakeLists.txt
+@@ -135,7 +135,7 @@ set(test_sources ${test_sources}
+ )
+ 
+ if(TENSILE_USE_LLVM)
+-    find_package(LLVM 13.0 QUIET CONFIG)
++    find_package(LLVM PATHS @GENTOO_PORTAGE_EPREFIX@/usr/lib/llvm/roc QUIET CONFIG)
+     if(NOT LLVM_FOUND)
+         find_package(LLVM 12.0 QUIET CONFIG)
+         if(NOT LLVM_FOUND)
diff --git a/dev-util/Tensile/files/Tensile-5.0.2-gfx1031.patch b/dev-util/Tensile/files/Tensile-5.0.2-gfx1031.patch
new file mode 100644
index 000000000000..17925cfd49e0
--- /dev/null
+++ b/dev-util/Tensile/files/Tensile-5.0.2-gfx1031.patch
@@ -0,0 +1,104 @@
+Index: Tensile-rocm-5.0.1/pytest.ini
+===================================================================
+--- Tensile-rocm-5.0.1.orig/pytest.ini
++++ Tensile-rocm-5.0.1/pytest.ini
+@@ -87,6 +87,7 @@ markers =
+  xfail-gfx1011: architecture
+  xfail-gfx1012: architecture
+  xfail-gfx1030: architecture
++ xfail-gfx1031: architecture
+  skip-gfx000:  architecture
+  skip-gfx900:  architecture
+  skip-gfx906:  architecture
+@@ -96,3 +97,4 @@ markers =
+  skip-gfx1011: architecture
+  skip-gfx1012: architecture
+  skip-gfx1030: architecture
++ skip-gfx1031: architecture
+Index: Tensile-rocm-5.0.1/Tensile/Common.py
+===================================================================
+--- Tensile-rocm-5.0.1.orig/Tensile/Common.py
++++ Tensile-rocm-5.0.1/Tensile/Common.py
+@@ -198,7 +198,7 @@ globalParameters["MergeFiles"] = True
+ globalParameters["NumMergedFiles"] = 1            # The number of files that kernels should be split between when merging
+ 
+ globalParameters["MaxFileName"] = 64              # If a file name would be longer than this, shorten it with a hash.
+-globalParameters["SupportedISA"] = [(8,0,3), (9,0,0), (9,0,6), (9,0,8), (9,0,10), (10,1,0), (10,1,1), (10,1,2), (10,3,0)] # assembly kernels writer supports these architectures
++globalParameters["SupportedISA"] = [(8,0,3), (9,0,0), (9,0,6), (9,0,8), (9,0,10), (10,1,0), (10,1,1), (10,1,2), (10,3,0),(10,3,1)] # assembly kernels writer supports these architectures
+ 
+ globalParameters["GenerateManifestAndExit"] = False               # Output manifest file with list of expected library objects and exit
+ globalParameters["NewClient"] = 2                                 # Old client deprecated: NewClient must be set to 2.
+@@ -264,7 +264,7 @@ architectureMap = {
+   'gfx906':'vega20', 'gfx906:xnack+':'vega20', 'gfx906:xnack-':'vega20',
+   'gfx908':'arcturus','gfx908:xnack+':'arcturus', 'gfx908:xnack-':'arcturus',
+   'gfx90a':'aldebaran', 'gfx90a:xnack+':'aldebaran', 'gfx90a:xnack-':'aldebaran',
+-  'gfx1010':'navi10', 'gfx1011':'navi12', 'gfx1012':'navi14', 'gfx1030':'navi21'
++  'gfx1010':'navi10', 'gfx1011':'navi12', 'gfx1012':'navi14', 'gfx1030':'navi21', 'gfx1031':'navi22'
+ }
+ 
+ def getArchitectureName(gfxName):
+Index: Tensile-rocm-5.0.1/Tensile/KernelWriterAssembly.py
+===================================================================
+--- Tensile-rocm-5.0.1.orig/Tensile/KernelWriterAssembly.py
++++ Tensile-rocm-5.0.1/Tensile/KernelWriterAssembly.py
+@@ -2364,7 +2364,7 @@ class KernelWriterAssembly(KernelWriter)
+ 
+     # integer i8x4
+     elif kernel["ProblemType"]["DataType"].isInt8x4():
+-      if self.version == (9,0,6) or self.version == (9,0,8) or self.version == (9,0,10) or self.version == (10,3,0):
++      if self.version == (9,0,6) or self.version == (9,0,8) or self.version == (9,0,10) or self.version == (10,3,0) or self.version == (10,3,1):
+         for b in range(0, kernel["ThreadTile1"]):
+           for a in range(0, kernel["ThreadTile0"]):
+             for iui in range(0, innerUnroll):
+Index: Tensile-rocm-5.0.1/Tensile/Source/CMakeLists.txt
+===================================================================
+--- Tensile-rocm-5.0.1.orig/Tensile/Source/CMakeLists.txt
++++ Tensile-rocm-5.0.1/Tensile/Source/CMakeLists.txt
+@@ -47,7 +47,7 @@ if(NOT DEFINED CXX_VERSION_STRING)
+ endif()
+ 
+ if(CMAKE_CXX_COMPILER STREQUAL "hipcc")
+-  set(TENSILE_GPU_ARCHS gfx803 gfx900 gfx906:xnack- gfx908:xnack- gfx90a:xnack- gfx1010 gfx1011 gfx1012 gfx1030 CACHE STRING "GPU architectures")
++  set(TENSILE_GPU_ARCHS gfx803 gfx900 gfx906:xnack- gfx908:xnack- gfx90a:xnack- gfx1010 gfx1011 gfx1012 gfx1030 gfx1031 CACHE STRING "GPU architectures")
+ else()
+   set(TENSILE_GPU_ARCHS gfx803 gfx900 gfx906 gfx908 gfx90a CACHE STRING "GPU architectures")
+ endif()
+Index: Tensile-rocm-5.0.1/Tensile/Source/lib/include/Tensile/AMDGPU.hpp
+===================================================================
+--- Tensile-rocm-5.0.1.orig/Tensile/Source/lib/include/Tensile/AMDGPU.hpp
++++ Tensile-rocm-5.0.1/Tensile/Source/lib/include/Tensile/AMDGPU.hpp
+@@ -62,7 +62,8 @@ namespace Tensile
+             gfx1010 = 1010,
+             gfx1011 = 1011,
+             gfx1012 = 1012,
+-            gfx1030 = 1030
++            gfx1030 = 1030,
++            gfx1031 = 1031
+         };
+ 
+         AMDGPU();
+Index: Tensile-rocm-5.0.1/Tensile/Source/lib/include/Tensile/Serialization/Predicates.hpp
+===================================================================
+--- Tensile-rocm-5.0.1.orig/Tensile/Source/lib/include/Tensile/Serialization/Predicates.hpp
++++ Tensile-rocm-5.0.1/Tensile/Source/lib/include/Tensile/Serialization/Predicates.hpp
+@@ -215,6 +215,7 @@ namespace Tensile
+                 iot::enumCase(io, value, "gfx1011", AMDGPU::Processor::gfx1011);
+                 iot::enumCase(io, value, "gfx1012", AMDGPU::Processor::gfx1012);
+                 iot::enumCase(io, value, "gfx1030", AMDGPU::Processor::gfx1030);
++                iot::enumCase(io, value, "gfx1031", AMDGPU::Processor::gfx1031);
+             }
+         };
+     } // namespace Serialization
+Index: Tensile-rocm-5.0.1/Tensile/Source/lib/source/AMDGPU.cpp
+===================================================================
+--- Tensile-rocm-5.0.1.orig/Tensile/Source/lib/source/AMDGPU.cpp
++++ Tensile-rocm-5.0.1/Tensile/Source/lib/source/AMDGPU.cpp
+@@ -82,6 +82,8 @@ namespace Tensile
+             return stream << "gfx1012";
+         case AMDGPU::Processor::gfx1030:
+             return stream << "gfx1030";
++        case AMDGPU::Processor::gfx1031:
++            return stream << "gfx1031";
+         }
+         return stream;
+     }
diff --git a/dev-util/Tensile/files/Tensile-5.0.2-use-ninja.patch b/dev-util/Tensile/files/Tensile-5.0.2-use-ninja.patch
new file mode 100644
index 000000000000..485ec992d808
--- /dev/null
+++ b/dev-util/Tensile/files/Tensile-5.0.2-use-ninja.patch
@@ -0,0 +1,21 @@
+diff --git a/Tensile/ClientExecutable.py b/Tensile/ClientExecutable.py
+index ca3809f..2d1d83b 100644
+--- a/Tensile/ClientExecutable.py
++++ b/Tensile/ClientExecutable.py
+@@ -36,6 +36,7 @@ class CMakeEnvironment:
+ 
+         args = ['cmake']
+         args += itertools.chain.from_iterable([ ['-D', '{}={}'.format(key, value)] for key,value in self.options.items()])
++        args += itertools.chain.from_iterable([ ['-G', 'Ninja'], [ '-D', 'CMAKE_EXPORT_COMPILE_COMMANDS=ON' ] ])
+         args += [self.sourceDir]
+ 
+         Common.print2(' '.join(args))
+@@ -43,7 +44,7 @@ class CMakeEnvironment:
+             subprocess.check_call(args, cwd=Common.ensurePath(self.buildDir))
+ 
+     def build(self):
+-        args = ['make', '-j']
++        args = ['ninja']
+         Common.print2(' '.join(args))
+         with Common.ClientExecutionLock():
+             subprocess.check_call(args, cwd=self.buildDir)