Merge remote-tracking branch 'origin/develop' into ggemm_multid_two_stage

.github/CODEOWNERS

-* @zjing14 @asroy @junliume @illsilin @carlushuang @aosewski
+* @zjing14 @junliume @illsilin @carlushuang @aosewski
 # Documentation files
-docs/* @saadrahim @LisaDelaney
-*.md  @saadrahim @LisaDelaney
-*.rst  @saadrahim @LisaDelaney
-# Header directory
-library/include/*  @saadrahim @LisaDelaney
+docs/* @ROCm/rocm-documentation
+*.md @ROCm/rocm-documentation
+*.rst @ROCm/rocm-documentation
+# Header directory for Doxygen documentation
+library/include/* @ROCm/rocm-documentation

CHANGELOG.md

@@ -2,20 +2,27 @@
 
 Full documentation for Composable Kernel is not yet available.
 
-## (Unreleased) CK
+## CK for ROCm 6.1.0
 
-### Fixes
-None
+### Additions
+* Added generic instances for GEMM XDL operations (#1161)
+* Added gamma and beta parameters for the layernorm and groupnorm bwd operations (#1133)
+* Introduced wrapper sublibrary (limited functionality). (#1071, #1098, #1108, #1126)
+* Added an option to vary the number of warm-up cycles and iterations for ckProfiler (#1124)
 
 ### Optimizations
-None
+* New performance optimizations for GEMM operations on MI200 and MI300 architectures (#1135)
 
-### Additions
-* Introduced wrapper sublibrary (limited functionality). (#1071, #1098, #1108, #1126, #1139)
+### Fixes
+* Reduced the build time for most GPU architectures (#1084)
+* Fixed some conversion issues for fp8 data type (#1099)
 
 ### Changes
 None
 
+### Known issues
+None
+
 ## CK for ROCm 6.0.0
 
 ### Fixes
@@ -32,7 +39,7 @@ None
 * Grouped convolution support for small K and C (#822 #879 #897)
 * Support for NHWGC (2D and 3D) grouped convolution backward weight (#769 #804)
 * Support for bf16/f32/f16 and NHWGC (2D and 3D) grouped convolution backward data (#757 #799)
-* Support for Batched Gemm DL (#732)
+* Support for Batched GEMM DL (#732)
 
 ### Changes
  * Changed the grouped convolution API to maintain consistency with other convolution kernels (#817)
@@ -48,7 +55,7 @@ None
 
 ### Additions
 * New CMake flags:
-  * "DL_KERNELS"-* Must be set to "ON" in order to build the gemm_dl and batched_gemm_multi_d_dl instances
+  * "DL_KERNELS"-* Must be set to "ON" in order to build the GEMM DL and batched_gemm_multi_d_dl instances
   * "DTYPES" -- Can be set to any subset of "fp64;fp32;fp16;fp8;bf16;int8" to build an instance of the specified data types
   * "INSTANCES_ONLY" -- Only builds CK library and instances without tests, examples, or profiler
 * New feature: if GPU_TARGETS is not set in the CMake command line, CK will be built for all targets supported by the compiler

Dockerfile

@@ -16,17 +16,17 @@ RUN apt-get install -y --allow-unauthenticated apt-utils wget gnupg2 curl
 ENV APT_KEY_DONT_WARN_ON_DANGEROUS_USAGE=DontWarn
 RUN curl -fsSL https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor -o /etc/apt/trusted.gpg.d/rocm-keyring.gpg
 
-RUN if [ "$ROCMVERSION" != "6.0.1" ]; then \
+RUN if [ "$ROCMVERSION" != "6.1" ]; then \
         sh -c "wget https://repo.radeon.com/amdgpu-install/6.0/ubuntu/focal/amdgpu-install_6.0.60000-1_all.deb  --no-check-certificate" && \
         apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --allow-unauthenticated ./amdgpu-install_6.0.60000-1_all.deb && \
         wget -qO - http://repo.radeon.com/rocm/rocm.gpg.key | apt-key add - && \
         sh -c "echo deb [arch=amd64 signed-by=/etc/apt/trusted.gpg.d/rocm-keyring.gpg] $DEB_ROCM_REPO focal main > /etc/apt/sources.list.d/rocm.list" && \
         sh -c 'echo deb [arch=amd64 signed-by=/etc/apt/trusted.gpg.d/rocm-keyring.gpg] https://repo.radeon.com/amdgpu/$ROCMVERSION/ubuntu focal main > /etc/apt/sources.list.d/amdgpu.list'; \
-    elif [ "$ROCMVERSION" = "6.0.1" ] && [ "$compiler_version" = "rc1" ]; then \
-        sh -c "wget http://artifactory-cdn.amd.com/artifactory/list/amdgpu-deb/amdgpu-install-internal_6.0-20.04-1_all.deb --no-check-certificate" && \
-        apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install dialog && DEBIAN_FRONTEND=noninteractive apt-get install ./amdgpu-install-internal_6.0-20.04-1_all.deb && \
-        sh -c 'echo deb [arch=amd64 trusted=yes] http://compute-artifactory.amd.com/artifactory/list/rocm-release-archive-20.04-deb/ 6.0.1 rel-95 > /etc/apt/sources.list.d/rocm-build.list' && \
-        amdgpu-repo --amdgpu-build=1704947; \
+    elif [ "$ROCMVERSION" = "6.1" ] && [ "$compiler_version" = "rc2" ]; then \
+        sh -c "wget http://artifactory-cdn.amd.com/artifactory/list/amdgpu-deb/amdgpu-install-internal_6.1-20.04-1_all.deb --no-check-certificate" && \
+        apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install dialog && DEBIAN_FRONTEND=noninteractive apt-get install ./amdgpu-install-internal_6.1-20.04-1_all.deb && \
+        sh -c 'echo deb [arch=amd64 trusted=yes] http://compute-artifactory.amd.com/artifactory/list/rocm-release-archive-20.04-deb/ 6.1 rel-48 > /etc/apt/sources.list.d/rocm-build.list' && \
+        amdgpu-repo --amdgpu-build=1736298; \
     fi
 
 RUN sh -c "echo deb http://mirrors.kernel.org/ubuntu focal main universe | tee -a /etc/apt/sources.list"
@@ -41,6 +41,7 @@ chmod +x ${SCCACHE_INSTALL_LOCATION}/sccache
 ENV PATH=$PATH:${SCCACHE_INSTALL_LOCATION}
 
 # Install dependencies
+# hipTensor requires rocm-llvm-dev for rocm versions > 6.0.1
 RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --allow-unauthenticated \
     build-essential \
     cmake \
@@ -60,6 +61,7 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --allow-
     python3-dev \
     python3-pip \
     redis \
+    rocm-llvm-dev \
     sshpass \
     stunnel \
     software-properties-common \
@@ -73,6 +75,9 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --allow-
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
 
+# Update the cmake to version 3.27.5
+RUN pip install --upgrade cmake==3.27.5
+
 #Install latest ccache
 RUN git clone https://github.com/ccache/ccache.git && \
     cd ccache && mkdir build && cd build && cmake .. && make install
@@ -82,8 +87,6 @@ RUN wget -qO /usr/local/bin/ninja.gz https://github.com/ninja-build/ninja/releas
 RUN gunzip /usr/local/bin/ninja.gz
 RUN chmod a+x /usr/local/bin/ninja
 RUN git clone https://github.com/nico/ninjatracing.git
-# Update the cmake to the latest version
-RUN pip install --upgrade cmake==3.27.5
 
 #Install latest cppcheck
 RUN git clone https://github.com/danmar/cppcheck.git && \

Jenkinsfile

 def rocmnode(name) {
-    return '(rocmtest || miopen) && ' + name
+    return '(rocmtest || miopen) && (' + name + ')'
 }
 
 def show_node_info() {
@@ -7,6 +7,7 @@ def show_node_info() {
         echo "NODE_NAME = \$NODE_NAME"
         lsb_release -sd
         uname -r
+        cat /sys/module/amdgpu/version
         ls /opt/ -la
     """
 }
@@ -33,7 +34,11 @@ def runShell(String command){
 
 def getDockerImageName(){
     def img
-    if (params.ROCMVERSION != "6.0.1"){
+    if (params.USE_CUSTOM_DOCKER != ""){
+        img = "${params.USE_CUSTOM_DOCKER}"
+    }
+    else{
+    if (params.ROCMVERSION != "6.1"){
        if (params.COMPILER_VERSION == "") {
            img = "${env.CK_DOCKERHUB}:ck_ub20.04_rocm${params.ROCMVERSION}"
        }
@@ -61,6 +66,7 @@ def getDockerImageName(){
           }
        }
     }
+    }
     return img
 }
 
@@ -111,7 +117,9 @@ def getDockerImage(Map conf=[:]){
     {
         echo "Pulling down image: ${image}"
         retimage = docker.image("${image}")
-        retimage.pull()
+        withDockerRegistry([ credentialsId: "docker_test_cred", url: "" ]) {
+            retimage.pull()
+        }
     }
     catch(Exception ex)
     {
@@ -258,18 +266,24 @@ def cmake_build(Map conf=[:]){
             """)
         sh cmd3
     }
-
-    def setup_cmd = conf.get("setup_cmd", "${cmake_envs} cmake ${setup_args}   .. ")
     // reduce parallelism when compiling, clang uses too much memory
     def nt = nthreads()
-    def build_cmd = conf.get("build_cmd", "${build_envs} dumb-init make  -j${nt} ${config_targets}")
+    def cmd
     def execute_cmd = conf.get("execute_cmd", "")
-
-    def cmd = conf.get("cmd", """
+    if(!setup_args.contains("NO_CK_BUILD")){
+        def setup_cmd = conf.get("setup_cmd", "${cmake_envs} cmake ${setup_args}   .. ")
+        def build_cmd = conf.get("build_cmd", "${build_envs} dumb-init make  -j${nt} ${config_targets}")
+        cmd = conf.get("cmd", """
             ${setup_cmd}
             ${build_cmd}
             ${execute_cmd}
         """)
+    }
+    else{
+        cmd = conf.get("cmd", """
+            ${execute_cmd}
+        """)
+    }
 
     echo cmd
 
@@ -365,8 +379,8 @@ def runCKProfiler(Map conf=[:]){
                 (retimage, image) = getDockerImage(conf)
                 withDockerContainer(image: image, args: dockerOpts) {
                     timeout(time: 5, unit: 'MINUTES'){
-                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
-                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
+                        sh 'rocminfo | tee rocminfo.log'
+                        if ( !runShell('grep -n "gfx" rocminfo.log') ){
                             throw new Exception ("GPU not found")
                         }
                         else{
@@ -379,20 +393,6 @@ def runCKProfiler(Map conf=[:]){
                 echo "The job was cancelled or aborted"
                 throw e
             }
-            catch(Exception ex) {
-                retimage = docker.build("${image}", dockerArgs + " --no-cache .")
-                withDockerContainer(image: image, args: dockerOpts) {
-                    timeout(time: 5, unit: 'MINUTES'){
-                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
-                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
-                            throw new Exception ("GPU not found")
-                        }
-                        else{
-                            echo "GPU is OK"
-                        }
-                    }
-                }
-            }
 
             withDockerContainer(image: image, args: dockerOpts + ' -v=/var/jenkins/:/var/jenkins') {
                 timeout(time: 24, unit: 'HOURS')
@@ -408,7 +408,7 @@ def runCKProfiler(Map conf=[:]){
 
 					dir("script"){
                         if (params.RUN_FULL_QA){
-                            sh "./run_full_performance_tests.sh 1 QA_${params.COMPILER_VERSION} ${env.BRANCH_NAME} ${NODE_NAME}"
+                            sh "./run_full_performance_tests.sh 0 QA_${params.COMPILER_VERSION} ${env.BRANCH_NAME} ${NODE_NAME}"
                             archiveArtifacts "perf_gemm.log"
                             archiveArtifacts "perf_resnet50_N256.log"
                             archiveArtifacts "perf_resnet50_N4.log"
@@ -418,9 +418,9 @@ def runCKProfiler(Map conf=[:]){
                             archiveArtifacts "perf_conv_bwd_data.log"
                             archiveArtifacts "perf_gemm_bilinear.log"
                             archiveArtifacts "perf_reduction.log"
-                            archiveArtifacts "perf_splitK_gemm_verify.log"
                             archiveArtifacts "perf_splitK_gemm.log"
                             archiveArtifacts "perf_onnx_gemm.log"
+                            archiveArtifacts "perf_mixed_gemm.log"
                            // stash perf files to master
                             stash name: "perf_gemm.log"
                             stash name: "perf_resnet50_N256.log"
@@ -433,6 +433,7 @@ def runCKProfiler(Map conf=[:]){
                             stash name: "perf_reduction.log"
                             stash name: "perf_splitK_gemm.log"
                             stash name: "perf_onnx_gemm.log"
+                            stash name: "perf_mixed_gemm.log"
                             //we will process results on the master node
                         }
                         else{
@@ -473,6 +474,7 @@ def Build_CK(Map conf=[:]){
         show_node_info()
 
         env.HSA_ENABLE_SDMA=0
+        env.DOCKER_BUILDKIT=1
         checkout scm
 
         def image = getDockerImageName() 
@@ -487,26 +489,25 @@ def Build_CK(Map conf=[:]){
         if (params.COMPILER_VERSION == "amd-staging" || params.COMPILER_VERSION == "amd-mainline-open" || params.COMPILER_COMMIT != ""){
             dockerOpts = dockerOpts + " --env HIP_CLANG_PATH='/llvm-project/build/bin' "
         }
+        def video_id = sh(returnStdout: true, script: 'getent group video | cut -d: -f3')
+        def render_id = sh(returnStdout: true, script: 'getent group render | cut -d: -f3')
+        dockerOpts = dockerOpts + " --group-add=${video_id} --group-add=${render_id} "
+        echo "Docker flags: ${dockerOpts}"
 
         def variant = env.STAGE_NAME
         def retimage
-        def navi_node = 0
-
-        gitStatusWrapper(credentialsId: "${status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCm', repo: 'composable_kernel') {
+        gitStatusWrapper(credentialsId: "${env.status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCm', repo: 'composable_kernel') {
             try {
                 (retimage, image) = getDockerImage(conf)
                 withDockerContainer(image: image, args: dockerOpts) {
                     timeout(time: 5, unit: 'MINUTES'){
-                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
-                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
+                        sh 'rocminfo | tee rocminfo.log'
+                        if ( !runShell('grep -n "gfx" rocminfo.log') ){
                             throw new Exception ("GPU not found")
                         }
                         else{
                             echo "GPU is OK"
                         }
-                        if ( runShell('grep -n "gfx1030" clinfo.log') || runShell('grep -n "gfx1101" clinfo.log') ){
-                            navi_node = 1
-                        }
                     }
                 }
             }
@@ -514,43 +515,38 @@ def Build_CK(Map conf=[:]){
                 echo "The job was cancelled or aborted"
                 throw e
             }
-            catch(Exception ex) {
-                retimage = docker.build("${image}", dockerArgs + " --no-cache .")
-                withDockerContainer(image: image, args: dockerOpts) {
-                    timeout(time: 5, unit: 'MINUTES'){
-                        sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo |tee clinfo.log'
-                        if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
-                            throw new Exception ("GPU not found")
-                        }
-                        else{
-                            echo "GPU is OK"
-                        }
-                        if ( runShell('grep -n "gfx1030" clinfo.log') || runShell('grep -n "gfx1101" clinfo.log') ){
-                            navi_node = 1
-                        }
-                    }
-                }
-            }
             withDockerContainer(image: image, args: dockerOpts + ' -v=/var/jenkins/:/var/jenkins') {
                 timeout(time: 24, unit: 'HOURS')
                 {
+                    //check whether running on Navi or MI300 node
+                    def navi_node = 0
+                    def mi300_node = 0
+                    sh 'rocminfo | tee rocminfo.log'
+                    if ( runShell('grep -n "gfx1030" rocminfo.log') || runShell('grep -n "gfx1101" rocminfo.log') ){
+                        navi_node = 1
+                        echo "This is a Navi node"
+                    }
+                    if ( runShell('grep -n "gfx942" rocminfo.log') ){
+                        mi300_node = 1
+                        echo "This is MI300 node"
+                    }
                     cmake_build(conf)
                     dir("build"){
                         //run tests and examples
                         sh 'make -j check'
-                        if (navi_node == 0 ){
+                        if (params.RUN_PERFORMANCE_TESTS && navi_node == 0 && mi300_node == 0 ){
                             //we only need the ckProfiler to run the performance tests, so we pack and stash it
-                            //do not stash profiler on Navi nodes
+                            //do not stash profiler on Navi or MI300 nodes
                            sh 'tar -zcvf ckProfiler.tar.gz bin/ckProfiler'
-                           stash "ckProfiler.tar.gz"
+                           stash name: "ckProfiler.tar.gz"
                         }
-                        if (params.RUN_FULL_QA){
-                           // build deb packages
+                        if (params.RUN_FULL_QA && mi300_node == 0 ){
+                           // build deb packages for all MI100/200/300 targets and prepare to export
                            sh 'make -j package'
                            archiveArtifacts artifacts: 'composablekernel-ckprofiler_*.deb'
                            archiveArtifacts artifacts: 'composablekernel-tests_*.deb'
                            sh 'mv composablekernel-ckprofiler_*.deb ckprofiler_0.2.0_amd64.deb'
-                           stash "ckprofiler_0.2.0_amd64.deb"
+                           stash name: "ckprofiler_0.2.0_amd64.deb"
                         }
                     }
                     if (params.hipTensor_test && navi_node == 0 ){
@@ -610,7 +606,7 @@ def process_results(Map conf=[:]){
     def variant = env.STAGE_NAME
     def retimage
 
-    gitStatusWrapper(credentialsId: "${status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCm', repo: 'composable_kernel') {
+    gitStatusWrapper(credentialsId: "${env.status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCm', repo: 'composable_kernel') {
         try {
             (retimage, image) = getDockerImage(conf)
         }
@@ -637,6 +633,7 @@ def process_results(Map conf=[:]){
                         unstash "perf_reduction.log"
                         unstash "perf_splitK_gemm.log"
                         unstash "perf_onnx_gemm.log"
+                        unstash "perf_mixed_gemm.log"
                         sh "./process_qa_data.sh"
                         unstash "ckprofiler_0.2.0_amd64.deb"
                         sh "sshpass -p ${env.ck_deb_pw} scp -o StrictHostKeyChecking=no ckprofiler_0.2.0_amd64.deb ${env.ck_deb_user}@${env.ck_deb_ip}:/var/www/html/composable_kernel/"
@@ -678,6 +675,10 @@ pipeline {
             name: "BUILD_DOCKER",
             defaultValue: false,
             description: "Force building docker image (default: false), set to true if docker image needs to be updated.")
+        string(
+            name: 'USE_CUSTOM_DOCKER',
+            defaultValue: '',
+            description: 'If you want to use a custom docker image, please specify it here (default: leave blank).')
         string(
             name: 'ROCMVERSION', 
             defaultValue: '6.0', 
@@ -720,8 +721,12 @@ pipeline {
             description: "Run the cppcheck static analysis (default: OFF)")
         booleanParam(
             name: "RUN_PERFORMANCE_TESTS",
-            defaultValue: false,
-            description: "Run the performance tests (default: OFF)")
+            defaultValue: true,
+            description: "Run the performance tests (default: ON)")
+        booleanParam(
+            name: "RUN_CODEGEN_TESTS",
+            defaultValue: true,
+            description: "Run the codegen tests (default: ON)")
     }
     environment{
         dbuser = "${dbuser}"
@@ -800,7 +805,34 @@ pipeline {
                 }
             }
         }
-    
+        stage("Run Codegen Tests")
+        {
+            parallel
+            {
+                stage("Run Codegen Tests on MI100/MI200")
+                {
+                    when {
+                        beforeAgent true
+                        expression { params.RUN_CODEGEN_TESTS.toBoolean() }
+                    }
+                    options { retry(2) }
+                    agent{ label rocmnode("gfx908 || gfx90a")}
+                    environment{
+                        setup_args = "NO_CK_BUILD"
+                        execute_args = """ cd ../codegen && rm -rf build && mkdir build && cd build && \
+                                           cmake -D CMAKE_PREFIX_PATH=/opt/rocm \
+                                           -D CMAKE_CXX_COMPILER=/opt/rocm/llvm/bin/clang++ \
+                                           -D CMAKE_BUILD_TYPE=Release \
+                                           -D GPU_TARGETS="gfx908;gfx90a" \
+                                           -DCMAKE_CXX_FLAGS=" -O3 " .. && make -j check"""
+                   }
+                    steps{
+                        buildHipClangJobAndReboot(setup_args:setup_args, no_reboot:true, build_type: 'Release', execute_cmd: execute_args)
+                        cleanWs()
+                    }
+                }
+            }
+        }
 		stage("Build CK and run Tests")
         {
             parallel
@@ -828,6 +860,26 @@ pipeline {
                         cleanWs()
                     }
                 }
+                stage("Build CK and run Tests on MI300")
+                {
+                    when {
+                        beforeAgent true
+                        expression { params.RUN_FULL_QA.toBoolean() }
+                    }
+                    agent{ label rocmnode("gfx942") }
+                    environment{
+                        setup_args = """ -DCMAKE_INSTALL_PREFIX=../install -DGPU_TARGETS="gfx942" -DCMAKE_CXX_FLAGS=" -O3 " """
+                        execute_args = """ cd ../client_example && rm -rf build && mkdir build && cd build && \
+                                           cmake -DCMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" \
+                                           -DGPU_TARGETS="gfx942" \
+                                           -DCMAKE_CXX_COMPILER="${build_compiler()}" \
+                                           -DCMAKE_CXX_FLAGS=" -O3 " .. && make -j """
+                    }
+                    steps{
+                        Build_CK_and_Reboot(setup_args: setup_args, config_targets: "install", no_reboot:true, build_type: 'Release', execute_cmd: execute_args, prefixpath: '/usr/local')
+                        cleanWs()
+                    }
+                }
                 stage("Build CK and run Tests on MI100/MI200")
                 {
                     when {

client_example/01_gemm/gemm.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>
@@ -83,7 +83,7 @@ int main(int argc, char* argv[])
         [](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
             using Layout = decltype(layout);
 
-            if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
+            if constexpr(std::is_same<Layout, Row>::value)
             {
                 return (nRow - 1) * stride + nCol;
             }

client_example/02_gemm_add_add_fastgelu/gemm_add_add_fastgelu.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>
@@ -92,7 +92,7 @@ int main(int argc, char* argv[])
         [](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
             using Layout = decltype(layout);
 
-            if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
+            if constexpr(std::is_same<Layout, Row>::value)
             {
                 return (nRow - 1) * stride + nCol;
             }

client_example/02_gemm_add_add_fastgelu/gemm_add_add_fastgelu_generic.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>
@@ -93,7 +93,7 @@ int main(int argc, char* argv[])
         [](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
             using Layout = decltype(layout);
 
-            if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
+            if constexpr(std::is_same<Layout, Row>::value)
             {
                 return (nRow - 1) * stride + nCol;
             }

client_example/02_gemm_add_add_fastgelu/gemm_add_fastgelu.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>
@@ -88,7 +88,7 @@ int main(int argc, char* argv[])
         [](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
             using Layout = decltype(layout);
 
-            if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
+            if constexpr(std::is_same<Layout, Row>::value)
             {
                 return (nRow - 1) * stride + nCol;
             }

client_example/02_gemm_add_add_fastgelu/gemm_add_fastgelu_generic.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>
@@ -89,7 +89,7 @@ int main(int argc, char* argv[])
         [](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
             using Layout = decltype(layout);
 
-            if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
+            if constexpr(std::is_same<Layout, Row>::value)
             {
                 return (nRow - 1) * stride + nCol;
             }

client_example/02_gemm_add_add_fastgelu/gemm_fastgelu.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>
@@ -84,7 +84,7 @@ int main(int argc, char* argv[])
         [](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
             using Layout = decltype(layout);
 
-            if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
+            if constexpr(std::is_same<Layout, Row>::value)
             {
                 return (nRow - 1) * stride + nCol;
             }

client_example/02_gemm_add_add_fastgelu/gemm_fastgelu_generic.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>
@@ -85,7 +85,7 @@ int main(int argc, char* argv[])
         [](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
             using Layout = decltype(layout);
 
-            if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
+            if constexpr(std::is_same<Layout, Row>::value)
             {
                 return (nRow - 1) * stride + nCol;
             }

client_example/03_gemm_layernorm/gemm_add_add_layernorm_naive.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>
@@ -17,6 +17,8 @@
 using F16 = ck::half_t;
 using F32 = float;
 
+using Row = ck::tensor_layout::gemm::RowMajor;
+
 using ADataType            = F16;
 using BDataType            = F16;
 using BiasDataType         = F32;
@@ -191,7 +193,7 @@ int main()
         [](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
             using Layout = decltype(layout);
 
-            if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
+            if constexpr(std::is_same<Layout, Row>::value)
             {
                 return (nRow - 1) * stride + nCol;
             }

client_example/03_gemm_layernorm/gemm_add_relu_add_layernorm_welford.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <iostream>
@@ -78,7 +78,7 @@ int main(int argc, char* argv[])
         [](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
             using Layout = decltype(layout);
 
-            if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
+            if constexpr(std::is_same<Layout, Row>::value)
             {
                 return (nRow - 1) * stride + nCol;
             }

client_example/04_contraction/contraction_bilinear_fp32.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <numeric>

client_example/04_contraction/contraction_bilinear_fp64.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <numeric>

client_example/04_contraction/contraction_g1m2n3k1_add_xdl_fp16.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <numeric>

client_example/04_contraction/contraction_scale_fp32.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <numeric>

client_example/04_contraction/contraction_scale_fp64.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <numeric>

client_example/05_layernorm/layernorm2d_bwd_data.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>

client_example/05_layernorm/layernorm2d_bwd_gamma_beta.cpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #include <iomanip>
 #include <vector>