Merge branch 'develop' into lwpck-726

.github/dependabot.yml

@@ -6,7 +6,7 @@
 version: 2
 updates:
   - package-ecosystem: "pip" # See documentation for possible values
-    directory: "/docs/.sphinx" # Location of package manifests
+    directory: "/docs/sphinx" # Location of package manifests
     open-pull-requests-limit: 10
     schedule:
       interval: "daily"

.gitignore

@@ -49,10 +49,10 @@ build*
 install.dir*
 
 # documentation artifacts
-build/
 _build/
 _images/
 _static/
 _templates/
 _toc.yml
 docBin/
+_doxygen/

.readthedocs.yaml

@@ -11,8 +11,8 @@ build:
 sphinx:
    configuration: docs/conf.py
 
-formats: [htmlzip]
+formats: [htmlzip, pdf, epub]
 
 python:
    install:
-   - requirements: docs/.sphinx/requirements.txt
+   - requirements: docs/sphinx/requirements.txt

Jenkinsfile

@@ -493,10 +493,11 @@ def Build_CK(Map conf=[:]){
                 {
                     cmake_build(conf)
                     dir("build"){
+                        //run tests and examples 	
+                        sh 'make -j\$(( \$(nproc) / 2 )) check'
                         if (navi_node == 0 ){
-                           //run tests and examples on all nodes except Navi
-                           sh 'make -j check'
-                           //we only need the ckProfiler to run the performance tests, so we pack and stash it
+                            //we only need the ckProfiler to run the performance tests, so we pack and stash it
+                            //do not stash profiler on Navi nodes
                            sh 'tar -zcvf ckProfiler.tar.gz bin/ckProfiler'
                            stash "ckProfiler.tar.gz"
                         }
@@ -686,12 +687,31 @@ pipeline {
         {
             parallel
             {
+                stage("Build CK and run Tests on MI100/MI200/MI300")
+                {
+                    when {
+                        beforeAgent true
+                        expression { params.RUN_FULL_QA.toBoolean() }
+                    }
+                    agent{ label rocmnode("gfx908 || gfx90a") }
+                    environment{
+                        setup_args = """ -DCMAKE_INSTALL_PREFIX=../install -DGPU_TARGETS="gfx908;gfx90a;gfx940" """
+                        execute_args = """ cd ../client_example && rm -rf build && mkdir build && cd build && cmake -D CMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" -DGPU_TARGETS="gfx908;gfx90a;gfx940" -D CMAKE_CXX_COMPILER="${build_compiler()}" .. && 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')
+                    }
+                }
                 stage("Build CK and run Tests on MI100/MI200")
                 {
+                    when {
+                        beforeAgent true
+                        expression { !params.RUN_FULL_QA.toBoolean() }
+                    }
                     agent{ label rocmnode("gfx908 || gfx90a") }
                     environment{
                         setup_args = """ -DCMAKE_INSTALL_PREFIX=../install -DGPU_TARGETS="gfx908;gfx90a" """
-                        execute_args = """ cd ../client_example && rm -rf build && mkdir build && cd build && cmake -D CMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" -DGPU_TARGETS="gfx908,gfx90a" -D CMAKE_CXX_COMPILER="${build_compiler()}" .. && make -j """ 
+                        execute_args = """ cd ../client_example && rm -rf build && mkdir build && cd build && cmake -D CMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" -DGPU_TARGETS="gfx908;gfx90a" -D CMAKE_CXX_COMPILER="${build_compiler()}" .. && 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')
@@ -705,8 +725,8 @@ pipeline {
                     }
                     agent{ label rocmnode("navi21") }
                     environment{
-                        setup_args = """ -DCMAKE_INSTALL_PREFIX=../install """ 
-                        execute_args = """ cd ../client_example && rm -rf build && mkdir build && cd build && cmake -D CMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" -DGPU_TARGETS="gfx1030;gfx1100;gfx1101;gfx1102" -D CMAKE_CXX_COMPILER="${build_compiler()}" .. && make -j """
+                        setup_args = """ -DCMAKE_INSTALL_PREFIX=../install -DGPU_TARGETS="gfx1030" """ 
+                        execute_args = """ cd ../client_example && rm -rf build && mkdir build && cd build && cmake -D CMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" -DGPU_TARGETS="gfx1030" -D CMAKE_CXX_COMPILER="${build_compiler()}" .. && make -j """
 
                     }
                     steps{

README.md

 # Composable Kernel
 
 ## Methodology
+
 Composable Kernel (CK) library aims to provide a programming model for writing performance critical kernels for machine learning workloads across multiple architectures including GPUs, CPUs, etc, through general purpose kernel languages, like HIP C++.
 
 CK utilizes two concepts to achieve performance portability and code maintainability:
@@ -10,6 +11,7 @@ CK utilizes two concepts to achieve performance portability and code maintainabi
 ![ALT](/docs/data/ck_component.png "CK Components")
 
 ## Code Structure
+
 Current CK library are structured into 4 layers:
 * "Templated Tile Operators" layer
 * "Templated Kernel and Invoker" layer
@@ -24,30 +26,35 @@ Run the steps below to build documentation locally.
 
 ```
 cd docs
-pip3 install -r .sphinx/requirements.txt
+pip3 install -r sphinx/requirements.txt
 python3 -m sphinx -T -E -b html -d _build/doctrees -D language=en . _build/html
 ```
 
 ## Contributors
+
 The list of developers and contributors is here: [Contributors](/CONTRIBUTORS.md)
 
 ## Citation
+
 If you use CK, please use following citations:
 * CK paper will be freely available on arXiv soon: [Realizing Tensor Operators Using Coordinate Transformations and Tile Based Programming](???)
 * [CITATION.cff](/CITATION.cff)
 
 ## License
+
 CK is released under the MIT license. [License File](/LICENSE)
 
 
 # Build CK
 
 ## Build docker image
+
 ```bash
 DOCKER_BUILDKIT=1 docker build -t ck:latest -f Dockerfile .
 ```
 
 ## Launch docker
+
 ```bash
 docker run                                     \
 -it                                            \
@@ -60,10 +67,12 @@ ck:latest                                      \
 ```
 
 ## Build CK
+
 ```bash
 mkdir build && cd build
 
 # Need to specify target ID, example below is for gfx908 and gfx90a
+
 cmake                                                                                             \
 -D CMAKE_PREFIX_PATH=/opt/rocm                                                                    \
 -D CMAKE_CXX_COMPILER=/opt/rocm/bin/hipcc                                                         \
@@ -74,6 +83,7 @@ cmake
 ```
 
 ### Build examples and tests
+
 ```bash
  make -j examples tests
  make test
@@ -83,21 +93,25 @@ Instructions for running each individual examples are under [example](/example)
 
 
 ## Build ckProfiler
+
 ```bash
  make -j ckProfiler
 ```
 Instructions for running ckProfiler are under [profiler](/profiler)
 
 ## Install CK
+
 ```bash
 make install
 ```
 
 ## Using CK as pre-built kernel library
+
 Instructions for using CK as a pre-built kernel library are under [client_example](/client_example)
 
 ## Caveat
 ### Kernel Timing and Verification
+
 CK's own kernel timer will warn up kernel once, and then run it multiple times
 to get average kernel time. For some kernels that use atomic add, this will cause
 output buffer to be accumulated multiple times, causing verification failure.

docs/.sphinx/_toc.yml.in

-root: index

docs/conf.py

@@ -4,10 +4,21 @@
 # list see the documentation:
 # https://www.sphinx-doc.org/en/master/usage/configuration.html
 
+import subprocess
+
 from rocm_docs import ROCmDocs
 
-docs_core = ROCmDocs("Composable Kernel Documentation")
-docs_core.run_doxygen()
+
+name = "Composable Kernel"
+get_version = r'sed -n -e "s/^rocm_setup_version(.* \([0-9\.]\{1,\}\).*/\1/p" ../CMakeLists.txt'
+version = subprocess.getoutput(get_version)
+if len(version) > 0:
+    name = f"{name} {version}"
+
+external_toc_path = "./sphinx/_toc.yml"
+
+docs_core = ROCmDocs(f"{name} Documentation")
+docs_core.run_doxygen(doxygen_root="doxygen", doxygen_path="doxygen/docBin/xml")
 docs_core.setup()
 
 mathjax3_config = {

docs/license.rst

+=======
+License
+=======
+
+.. include:: ../LICENSE
+   :literal:

docs/sphinx/_toc.yml.in

+# Anywhere {branch} is used, the branch name will be substituted.
+# These comments will also be removed.
+defaults:
+  numbered: False
+  maxdepth: 6
+root: index
+subtrees:
+  - caption: About
+    entries:
+      - file: license

docs/.sphinx/requirements.in → docs/sphinx/requirements.in

-rocm-docs-core==0.2.0
+rocm-docs-core==0.10.3
 sphinxcontrib-bibtex==2.5.0

docs/.sphinx/requirements.txt → docs/sphinx/requirements.txt

 #
-# This file is autogenerated by pip-compile with Python 3.10
+# This file is autogenerated by pip-compile with Python 3.8
 # by the following command:
 #
-#    pip-compile .sphinx/requirements.in
+#    pip-compile requirements.in
 #
 accessible-pygments==0.0.3
     # via pydata-sphinx-theme
 alabaster==0.7.13
     # via sphinx
-asttokens==2.2.1
-    # via stack-data
-attrs==22.2.0
-    # via
-    #   jsonschema
-    #   jupyter-cache
 babel==2.12.1
     # via
     #   pydata-sphinx-theme
     #   sphinx
-backcall==0.2.0
-    # via ipython
 beautifulsoup4==4.11.2
     # via pydata-sphinx-theme
 breathe==4.34.0
@@ -27,19 +19,15 @@ breathe==4.34.0
 certifi==2022.12.7
     # via requests
 cffi==1.15.1
-    # via pynacl
+    # via
+    #   cryptography
+    #   pynacl
 charset-normalizer==3.1.0
     # via requests
 click==8.1.3
-    # via
-    #   jupyter-cache
-    #   sphinx-external-toc
-comm==0.1.2
-    # via ipykernel
-debugpy==1.6.6
-    # via ipykernel
-decorator==5.1.1
-    # via ipython
+    # via sphinx-external-toc
+cryptography==40.0.2
+    # via pyjwt
 deprecated==1.2.13
     # via pygithub
 docutils==0.16
@@ -48,52 +36,26 @@ docutils==0.16
     #   myst-parser
     #   pybtex-docutils
     #   pydata-sphinx-theme
-    #   rocm-docs-core
     #   sphinx
     #   sphinxcontrib-bibtex
-executing==1.2.0
-    # via stack-data
-fastjsonschema==2.16.3
-    # via nbformat
 gitdb==4.0.10
     # via gitpython
 gitpython==3.1.31
     # via rocm-docs-core
-greenlet==2.0.2
-    # via sqlalchemy
 idna==3.4
     # via requests
 imagesize==1.4.1
     # via sphinx
 importlib-metadata==6.0.0
     # via
-    #   jupyter-cache
-    #   myst-nb
-ipykernel==6.21.3
-    # via myst-nb
-ipython==8.11.0
-    # via
-    #   ipykernel
-    #   myst-nb
-jedi==0.18.2
-    # via ipython
+    #   sphinx
+    #   sphinxcontrib-bibtex
+importlib-resources==5.12.0
+    # via rocm-docs-core
 jinja2==3.1.2
     # via
     #   myst-parser
     #   sphinx
-jsonschema==4.17.3
-    # via nbformat
-jupyter-cache==0.5.0
-    # via myst-nb
-jupyter-client==8.0.3
-    # via
-    #   ipykernel
-    #   nbclient
-jupyter-core==5.3.0
-    # via
-    #   ipykernel
-    #   jupyter-client
-    #   nbformat
 latexcodec==2.0.1
     # via pybtex
 linkify-it-py==1.0.3
@@ -104,54 +66,16 @@ markdown-it-py==2.2.0
     #   myst-parser
 markupsafe==2.1.2
     # via jinja2
-matplotlib-inline==0.1.6
-    # via
-    #   ipykernel
-    #   ipython
 mdit-py-plugins==0.3.5
     # via myst-parser
 mdurl==0.1.2
     # via markdown-it-py
-myst-nb==0.17.1
+myst-parser[linkify]==1.0.0
     # via rocm-docs-core
-myst-parser[linkify]==0.18.1
-    # via
-    #   myst-nb
-    #   rocm-docs-core
-nbclient==0.5.13
-    # via
-    #   jupyter-cache
-    #   myst-nb
-nbformat==5.7.3
-    # via
-    #   jupyter-cache
-    #   myst-nb
-    #   nbclient
-nest-asyncio==1.5.6
-    # via
-    #   ipykernel
-    #   nbclient
 packaging==23.0
     # via
-    #   ipykernel
     #   pydata-sphinx-theme
     #   sphinx
-parso==0.8.3
-    # via jedi
-pexpect==4.8.0
-    # via ipython
-pickleshare==0.7.5
-    # via ipython
-platformdirs==3.1.1
-    # via jupyter-core
-prompt-toolkit==3.0.38
-    # via ipython
-psutil==5.9.4
-    # via ipykernel
-ptyprocess==0.7.0
-    # via pexpect
-pure-eval==0.2.2
-    # via stack-data
 pybtex==0.24.0
     # via
     #   pybtex-docutils
@@ -160,57 +84,47 @@ pybtex-docutils==1.0.2
     # via sphinxcontrib-bibtex
 pycparser==2.21
     # via cffi
-pydata-sphinx-theme==0.13.1
-    # via sphinx-book-theme
-pygithub==1.57
+pydata-sphinx-theme==0.13.3
+    # via
+    #   rocm-docs-core
+    #   sphinx-book-theme
+pygithub==1.58.2
     # via rocm-docs-core
 pygments==2.14.0
     # via
     #   accessible-pygments
-    #   ipython
     #   pydata-sphinx-theme
     #   sphinx
-pyjwt==2.6.0
+pyjwt[crypto]==2.6.0
     # via pygithub
 pynacl==1.5.0
     # via pygithub
-pyrsistent==0.19.3
-    # via jsonschema
-python-dateutil==2.8.2
-    # via jupyter-client
+pytz==2023.3
+    # via babel
 pyyaml==6.0
     # via
-    #   jupyter-cache
-    #   myst-nb
     #   myst-parser
     #   pybtex
     #   sphinx-external-toc
-pyzmq==25.0.1
-    # via
-    #   ipykernel
-    #   jupyter-client
 requests==2.28.2
     # via
     #   pygithub
     #   sphinx
-rocm-docs-core==0.2.0
-    # via -r .sphinx/requirements.in
+rocm-docs-core==0.10.3
+    # via -r requirements.in
 six==1.16.0
     # via
-    #   asttokens
     #   latexcodec
     #   pybtex
-    #   python-dateutil
 smmap==5.0.0
     # via gitdb
 snowballstemmer==2.2.0
     # via sphinx
 soupsieve==2.4
     # via beautifulsoup4
-sphinx==4.3.1
+sphinx==5.3.0
     # via
     #   breathe
-    #   myst-nb
     #   myst-parser
     #   pydata-sphinx-theme
     #   rocm-docs-core
@@ -220,7 +134,7 @@ sphinx==4.3.1
     #   sphinx-external-toc
     #   sphinx-notfound-page
     #   sphinxcontrib-bibtex
-sphinx-book-theme==1.0.0rc2
+sphinx-book-theme==1.0.1
     # via rocm-docs-core
 sphinx-copybutton==0.5.1
     # via rocm-docs-core
@@ -233,7 +147,7 @@ sphinx-notfound-page==0.8.3
 sphinxcontrib-applehelp==1.0.4
     # via sphinx
 sphinxcontrib-bibtex==2.5.0
-    # via -r .sphinx/requirements.in
+    # via -r requirements.in
 sphinxcontrib-devhelp==1.0.2
     # via sphinx
 sphinxcontrib-htmlhelp==2.0.1
@@ -244,40 +158,15 @@ sphinxcontrib-qthelp==1.0.3
     # via sphinx
 sphinxcontrib-serializinghtml==1.1.5
     # via sphinx
-sqlalchemy==1.4.46
-    # via jupyter-cache
-stack-data==0.6.2
-    # via ipython
-tabulate==0.9.0
-    # via jupyter-cache
-tornado==6.2
-    # via
-    #   ipykernel
-    #   jupyter-client
-traitlets==5.9.0
-    # via
-    #   comm
-    #   ipykernel
-    #   ipython
-    #   jupyter-client
-    #   jupyter-core
-    #   matplotlib-inline
-    #   nbclient
-    #   nbformat
 typing-extensions==4.5.0
-    # via
-    #   myst-nb
-    #   myst-parser
+    # via pydata-sphinx-theme
 uc-micro-py==1.0.1
     # via linkify-it-py
 urllib3==1.26.15
     # via requests
-wcwidth==0.2.6
-    # via prompt-toolkit
 wrapt==1.15.0
     # via deprecated
 zipp==3.15.0
-    # via importlib-metadata
-
-# The following packages are considered to be unsafe in a requirements file:
-# setuptools
+    # via
+    #   importlib-metadata
+    #   importlib-resources

example/02_gemm_bilinear/CMakeLists.txt

-add_example_executable(example_gemm_bilinear_xdl_fp16 gemm_bilinear_xdl_fp16.cpp)
 if(GPU_TARGETS MATCHES "gfx1100" OR GPU_TARGETS MATCHES "gfx1101" OR GPU_TARGETS MATCHES "gfx1102")
     add_example_executable(example_gemm_bilinear_wmma_fp16 gemm_bilinear_wmma_fp16.cpp)
 endif()
+if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR GPU_TARGETS MATCHES "gfx940")
+    add_example_executable(example_gemm_bilinear_xdl_fp16 gemm_bilinear_xdl_fp16.cpp)
+endif()

example/03_gemm_bias_relu/CMakeLists.txt

-add_example_executable(example_gemm_bias_relu_xdl_fp16 gemm_bias_relu_xdl_fp16.cpp)
+if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR GPU_TARGETS MATCHES "gfx940")
+    add_example_executable(example_gemm_bias_relu_xdl_fp16 gemm_bias_relu_xdl_fp16.cpp)
+endif()
\ No newline at end of file

example/04_gemm_add_add_fastgelu/CMakeLists.txt

-add_custom_target(example_gemm_add_add_fastgelu_xdl)
+if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR GPU_TARGETS MATCHES "gfx940")
+    add_custom_target(example_gemm_add_add_fastgelu_xdl)
 
-add_example_executable(example_gemm_add_add_fastgelu_xdl_bf16 gemm_add_add_fastgelu_xdl_bf16.cpp)
-add_example_executable(example_gemm_add_add_fastgelu_xdl_fp16 gemm_add_add_fastgelu_xdl_fp16.cpp)
-add_example_executable(example_gemm_add_add_fastgelu_xdl_fp32 gemm_add_add_fastgelu_xdl_fp32.cpp)
-if(USE_BITINT_EXTENSION_INT4)
-  add_example_executable(example_gemm_add_add_fastgelu_xdl_int4 gemm_add_add_fastgelu_xdl_int4.cpp)
-endif(USE_BITINT_EXTENSION_INT4)
-add_example_executable(example_gemm_add_add_fastgelu_xdl_int8 gemm_add_add_fastgelu_xdl_int8.cpp)
+    add_example_executable(example_gemm_add_add_fastgelu_xdl_bf16 gemm_add_add_fastgelu_xdl_bf16.cpp)
+    add_example_executable(example_gemm_add_add_fastgelu_xdl_fp16 gemm_add_add_fastgelu_xdl_fp16.cpp)
+    add_example_executable(example_gemm_add_add_fastgelu_xdl_fp32 gemm_add_add_fastgelu_xdl_fp32.cpp)
+    if(USE_BITINT_EXTENSION_INT4)
+       add_example_executable(example_gemm_add_add_fastgelu_xdl_int4 gemm_add_add_fastgelu_xdl_int4.cpp)
+    endif(USE_BITINT_EXTENSION_INT4)
+    add_example_executable(example_gemm_add_add_fastgelu_xdl_int8 gemm_add_add_fastgelu_xdl_int8.cpp)
 
-add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_bf16)
-add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_fp16)
-add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_fp32)
-if(USE_BITINT_EXTENSION_INT4)
-  add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_int4)
-endif(USE_BITINT_EXTENSION_INT4)
-add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_int8)
+    add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_bf16)
+    add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_fp16)
+    add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_fp32)
+    if(USE_BITINT_EXTENSION_INT4)
+        add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_int4)
+    endif(USE_BITINT_EXTENSION_INT4)
+    add_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_int8)
+endif()
\ No newline at end of file

example/09_convnd_fwd/CMakeLists.txt

-add_example_executable(example_convnd_fwd_xdl_fp32 convnd_fwd_xdl_fp32.cpp)
-add_example_executable(example_convnd_fwd_xdl_fp16 convnd_fwd_xdl_fp16.cpp)
-add_example_executable(example_convnd_fwd_xdl_bf16 convnd_fwd_xdl_bf16.cpp)
-add_example_executable(example_convnd_fwd_xdl_int8 convnd_fwd_xdl_int8.cpp)
-# FIXME: re-enable this exampe as test when SWDEV-335738 is fixed
-add_example_executable_no_testing(example_convnd_fwd_xdl_fp64 convnd_fwd_xdl_fp64.cpp)
-
+if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR GPU_TARGETS MATCHES "gfx940")
+    add_example_executable(example_convnd_fwd_xdl_fp32 convnd_fwd_xdl_fp32.cpp)
+    add_example_executable(example_convnd_fwd_xdl_fp16 convnd_fwd_xdl_fp16.cpp)
+    add_example_executable(example_convnd_fwd_xdl_bf16 convnd_fwd_xdl_bf16.cpp)
+    add_example_executable(example_convnd_fwd_xdl_int8 convnd_fwd_xdl_int8.cpp)
+    # FIXME: re-enable this exampe as test when SWDEV-335738 is fixed
+    add_example_executable_no_testing(example_convnd_fwd_xdl_fp64 convnd_fwd_xdl_fp64.cpp)
+endif()
 add_example_executable(example_convnd_fwd_dl_fp16 convnd_fwd_dl_fp16.cpp)
 add_example_executable(example_convnd_fwd_dl_fp32 convnd_fwd_dl_fp32.cpp)
 add_example_executable(example_convnd_fwd_dl_int8 convnd_fwd_dl_int8.cpp)

example/10_convnd_fwd_multiple_d_multiple_reduce/CMakeLists.txt

-add_custom_target(example_convnd_fwd_reduce_xdl)
-
-add_example_executable(example_convnd_fwd_max_xdl_int8 convnd_fwd_max_xdl_int8.cpp)
-add_example_executable_no_testing(example_convnd_fwd_max_xdl_bf16 convnd_fwd_max_xdl_bf16.cpp)
-add_example_executable_no_testing(example_convnd_fwd_max_xdl_fp16 convnd_fwd_max_xdl_fp16.cpp)
-add_example_executable(example_convnd_fwd_max_xdl_fp32 convnd_fwd_max_xdl_fp32.cpp)
-
-add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int8)
-add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_bf16)
-add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp16)
-add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp32)
-
-if(USE_BITINT_EXTENSION_INT4)
-  add_example_executable(example_convnd_fwd_max_xdl_int4 convnd_fwd_max_xdl_int4.cpp)
-  add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int4)
-endif(USE_BITINT_EXTENSION_INT4)
+if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR GPU_TARGETS MATCHES "gfx940")
+   add_custom_target(example_convnd_fwd_reduce_xdl)
+   add_example_executable(example_convnd_fwd_max_xdl_int8 convnd_fwd_max_xdl_int8.cpp)
+   add_example_executable_no_testing(example_convnd_fwd_max_xdl_bf16 convnd_fwd_max_xdl_bf16.cpp)
+   add_example_executable_no_testing(example_convnd_fwd_max_xdl_fp16 convnd_fwd_max_xdl_fp16.cpp)
+   add_example_executable(example_convnd_fwd_max_xdl_fp32 convnd_fwd_max_xdl_fp32.cpp)
+   add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int8)
+   add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_bf16)
+   add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp16)
+   add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_fp32)
+   if(USE_BITINT_EXTENSION_INT4)
+      add_example_executable(example_convnd_fwd_max_xdl_int4 convnd_fwd_max_xdl_int4.cpp)
+      add_dependencies(example_convnd_fwd_reduce_xdl example_convnd_fwd_max_xdl_int4)
+   endif(USE_BITINT_EXTENSION_INT4)
+endif()
\ No newline at end of file

example/14_gemm_quantization/CMakeLists.txt

@@ -2,5 +2,7 @@
 add_example_executable(example_gemm_dl_quantization_int8 gemm_dl_quantization_int8.cpp)
 
 # xdlops
-add_example_executable(example_gemm_xdl_bias_relu_quantization_int8 gemm_xdl_bias_relu_quantization_int8.cpp)
-add_example_executable(example_gemm_xdl_quantization_int8 gemm_xdl_quantization_int8.cpp)
\ No newline at end of file
+if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR GPU_TARGETS MATCHES "gfx940")
+   add_example_executable(example_gemm_xdl_bias_relu_quantization_int8 gemm_xdl_bias_relu_quantization_int8.cpp)
+   add_example_executable(example_gemm_xdl_quantization_int8 gemm_xdl_quantization_int8.cpp)
+endif()
\ No newline at end of file

example/16_gemm_multi_d_multi_reduces/CMakeLists.txt

-add_custom_target(example_gemm_reduce_xdl)
-add_custom_target(example_gemm_reduce_xdl_max)
-add_custom_target(example_gemm_reduce_xdl_mean_meansquare)
-add_custom_target(example_gemm_add_add_mean_meansquare_xdl)
+if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR GPU_TARGETS MATCHES "gfx940")
+   add_custom_target(example_gemm_reduce_xdl)
+   add_custom_target(example_gemm_reduce_xdl_max)
+   add_custom_target(example_gemm_reduce_xdl_mean_meansquare)
+   add_custom_target(example_gemm_add_add_mean_meansquare_xdl)
 
-add_example_executable(example_gemm_max_xdl_fp16 gemm_max_xdl_fp16.cpp)
-add_example_executable(example_gemm_max_xdl_int8 gemm_max_xdl_int8.cpp)
-add_example_executable(example_gemm_max_xdl_fp32 gemm_max_xdl_fp32.cpp)
-add_example_executable(example_gemm_max_xdl_bf16 gemm_max_xdl_bf16.cpp)
+   add_example_executable(example_gemm_max_xdl_fp16 gemm_max_xdl_fp16.cpp)
+   add_example_executable(example_gemm_max_xdl_int8 gemm_max_xdl_int8.cpp)
+   add_example_executable(example_gemm_max_xdl_fp32 gemm_max_xdl_fp32.cpp)
+   add_example_executable(example_gemm_max_xdl_bf16 gemm_max_xdl_bf16.cpp)
 
-add_example_executable(example_gemm_add_add_mean_meansquare_xdl_fp16 gemm_add_add_mean_meansquare_xdl_fp16.cpp)
+   add_example_executable(example_gemm_add_add_mean_meansquare_xdl_fp16 gemm_add_add_mean_meansquare_xdl_fp16.cpp)
 
-add_example_executable(example_gemm_mean_meansquare_xdl_fp16 gemm_mean_meansquare_xdl_fp16.cpp)
-add_example_executable(example_gemm_mean_meansquare_xdl_fp32 gemm_mean_meansquare_xdl_fp32.cpp)
-add_example_executable(example_gemm_mean_meansquare_xdl_bf16 gemm_mean_meansquare_xdl_bf16.cpp)
-add_example_executable(example_gemm_add_addsquare_xdl_int8 gemm_add_addsquare_xdl_int8.cpp)
+   add_example_executable(example_gemm_mean_meansquare_xdl_fp16 gemm_mean_meansquare_xdl_fp16.cpp)
+   add_example_executable(example_gemm_mean_meansquare_xdl_fp32 gemm_mean_meansquare_xdl_fp32.cpp)
+   add_example_executable(example_gemm_mean_meansquare_xdl_bf16 gemm_mean_meansquare_xdl_bf16.cpp)
+   add_example_executable(example_gemm_add_addsquare_xdl_int8 gemm_add_addsquare_xdl_int8.cpp)
 
-add_dependencies(example_gemm_reduce_xdl_max
+   add_dependencies(example_gemm_reduce_xdl_max
                  example_gemm_max_xdl_bf16
                  example_gemm_max_xdl_fp16
                  example_gemm_max_xdl_fp32
                  example_gemm_max_xdl_int8)
 
-add_dependencies(example_gemm_reduce_xdl_mean_meansquare
+   add_dependencies(example_gemm_reduce_xdl_mean_meansquare
                  example_gemm_mean_meansquare_xdl_fp16
                  example_gemm_mean_meansquare_xdl_fp32
                  example_gemm_mean_meansquare_xdl_bf16
                  example_gemm_add_addsquare_xdl_int8)
 
-add_dependencies(example_gemm_add_add_mean_meansquare_xdl example_gemm_add_add_mean_meansquare_xdl_fp16)
+   add_dependencies(example_gemm_add_add_mean_meansquare_xdl example_gemm_add_add_mean_meansquare_xdl_fp16)
 
-add_dependencies(example_gemm_reduce_xdl
+   add_dependencies(example_gemm_reduce_xdl
                  example_gemm_reduce_xdl_mean_meansquare
                  example_gemm_reduce_xdl_max
                  example_gemm_add_add_mean_meansquare_xdl)
 
-if(USE_BITINT_EXTENSION_INT4)
-  add_example_executable(example_gemm_max_xdl_int4 gemm_max_xdl_int4.cpp)
-  add_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_int4)
+   if(USE_BITINT_EXTENSION_INT4)
+      add_example_executable(example_gemm_max_xdl_int4 gemm_max_xdl_int4.cpp)
+      add_dependencies(example_gemm_reduce_xdl_max example_gemm_max_xdl_int4)
+   endif()
 endif()

example/17_convnd_bwd_data/CMakeLists.txt

-add_example_executable(example_convnd_bwd_data_xdl_fp16 convnd_bwd_data_xdl_fp16.cpp)
-target_link_libraries(example_convnd_bwd_data_xdl_fp16 PRIVATE utility)
-
+if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR GPU_TARGETS MATCHES "gfx940")
+   add_example_executable(example_convnd_bwd_data_xdl_fp16 convnd_bwd_data_xdl_fp16.cpp)
+   target_link_libraries(example_convnd_bwd_data_xdl_fp16 PRIVATE utility)
+endif()
 add_example_executable(example_convnd_bwd_data_dl_fp16 convnd_bwd_data_dl_fp16.cpp)
 target_link_libraries(example_convnd_bwd_data_dl_fp16 PRIVATE utility)