Updating GitHub CI.

.github/workflows/ci.sh

@@ -17,8 +17,10 @@ module load $COMPILER
 
 cd src
 
-if [ "$COMPONENTS" = "cuda nvml lmsensors io net powercap appio lustre coretemp rapl sde stealtime" ]; then
-  # Set necessary environment variables for lmsensors
+MATCH_COMPONENTS="cuda nvml rocm rocm_smi powercap powercap_ppc rapl sensors_ppc infiniband net appio io lustre stealtime coretemp lmsensors mx sde"
+# load necessary environment variables for the above set of MATCH_COMPONENTS
+if [ "$COMPONENTS" = "$MATCH_COMPONENTS" ]; then
+  # lmsensors
   wget https://github.com/groeck/lm-sensors/archive/V3-4-0.tar.gz
   tar -zxf V3-4-0.tar.gz 
   cd lm-sensors-3-4-0
@@ -27,45 +29,52 @@ if [ "$COMPONENTS" = "cuda nvml lmsensors io net powercap appio lustre coretemp
   export PAPI_LMSENSORS_ROOT=lm
   export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$PAPI_LMSENSORS_ROOT/lib
 
-  # Set necessary environment variables for cuda and nvml
+  # rocm and rocm_smi
+  export PAPI_ROCM_ROOT=`ls -d /opt/rocm-*`
+  export PAPI_ROCMSMI_ROOT=$PAPI_ROCM_ROOT/rocm_smi
+
+  # cuda and nvml
   module load cuda
   export PAPI_CUDA_ROOT=$ICL_CUDA_ROOT
   export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$PAPI_CUDA_ROOT/extras/CUPTI/lib64
 
 fi
 
-if [ "$COMPONENTS" = "rocm rocm_smi" ]; then
-  export PAPI_ROCM_ROOT=`ls -d /opt/rocm-*`
-  export PAPI_ROCMSMI_ROOT=$PAPI_ROCM_ROOT/rocm_smi
-fi
+#MATCH_COMPONENTS="rocm rocm_smi"
+# load necessary environment variables for testing of the rocm and rocm_smi components
+#if [ "$COMPONENTS" = "$MATCH_COMPONENTS" ]; then
+#  export PAPI_ROCM_ROOT=`ls -d /opt/rocm-*`
+#  export PAPI_ROCMSMI_ROOT=$PAPI_ROCM_ROOT/rocm_smi
+#fi
 
+# test linking with or without --with-shlib-tools 
 if [ "$SHLIB" = "with" ]; then
     ./configure --with-debug=$DEBUG --enable-warnings --with-components="$COMPONENTS" --with-shlib-tools
 else
     ./configure --with-debug=$DEBUG --enable-warnings --with-components="$COMPONENTS"
 fi
 
-#touch /tmp/component_names.txt
-#for component in $COMPONENTS; do
-#    echo "Name:   $component" >> /tmp/component_names.txt
-#done
-
-CHECK_COMPONENTS=$(echo $COMPONENTS | sed 's/ /|/g')
-echo $CHECK_COMPONENTS
-
 make -j4
 
+# run PAPI utilities
+echo "Running papi_component_avail: "
 utils/papi_component_avail
 
-# Make sure the list of components are active
-utils/papi_component_avail | grep -A1000 'Active components' | grep -E $CHECK_COMPONENTS
-
-#if [ "$COMPONENTS" != "cuda nvml" ]; then
-#   echo Testing
-#   ./run_tests.sh
-#fi
+# check list of active components
+ACTIVE_COMPONENTS=$(utils/papi_component_avail | grep -A1000 'Active components' | grep "Name:" | sed 's/Name: //' | awk '{print $1}' | paste -sd ' ' -)
+EXPECTED_ACTIVE_COMPONENTS="perf_event perf_event_uncore cuda nvml powercap net appio io stealtime coretemp lmsensors sde sysdetect"
+if [ "$ACTIVE_COMPONENTS" != "perf_event perf_event_uncore cuda nvml powercap net appio io stealtime coretemp lmsensors" ]; then
+    DISABLED_COMPONENTS=$(diff --side-by-side --suppress-common-lines <(echo "$EXPECTED_ACTIVE_COMPONENTS" | sed 's/ /\n/g') <(echo "$ACTIVE_COMPONENTS" | sed 's/ /\n/g') | awk '{print $2}')
+    echo -e "Components are disabled that should be active, these are:\n$DISABLED_COMPONENTS"
+    exit 1
+fi
 
-if [ "$SHLIB" = "with" ]; then
-    echo "Running single component test for active components"
-    ./run_tests_shlib.sh TESTS_QUIET
+# without '--with-shlib-tools' in ./configure
+if [ "$SHLIB" = "without" ]; then
+   echo "Running full test suite for active components"
+   ./run_tests.sh
+# with '--with-shlib-tools' in ./configure
+else
+   echo "Running single component test for active components"
+   ./run_tests_shlib.sh TESTS_QUIET
 fi

.github/workflows/main.yml

@@ -13,42 +13,18 @@ jobs:
   papi_component:
     strategy:
       matrix:
-        components: [cuda nvml lmsensors io net powercap appio lustre coretemp rapl sde stealtime]
+        components: [cuda nvml rocm rocm_smi powercap powercap_ppc rapl sensors_ppc infiniband net appio io lustre stealtime coretemp lmsensors mx sde]
         debug: [yes, no]
         shlib: [with, without]
+    #    machine: [cpu_intel, gpu_amd, gpu_intel]
       fail-fast: false
+    #runs-on: ${{matrix.machine}}
     runs-on: [self-hosted, cpu_intel, gpu_nvidia]
     timeout-minutes: 60
     steps:
       - uses: actions/checkout@v4
       - name: Test
         run: .github/workflows/ci.sh "${{matrix.components}}" ${{matrix.debug}} ${{matrix.shlib}}
-  papi_component_amd:
-    strategy:
-      matrix:
-        components: [rocm rocm_smi]
-        debug: [yes, no]
-        shlib: [with, without]
-      fail-fast: false
-    runs-on: [self-hosted, gpu_amd]
-    timeout-minutes: 60
-    steps:
-      - uses: actions/checkout@v4
-      - name: Test
-        run: .github/workflows/ci.sh "${{matrix.components}}" ${{matrix.debug}} ${{matrix.shlib}}
-  papi_component_intel_gpu:
-    strategy:
-      matrix:
-        components: [intel_gpu]
-        debug: [yes, no] 
-        shlib: [with, without]
-      fail-fast: false
-    runs-on: [self-hosted, gpu_intel]
-    timeout-minutes: 60
-    steps:
-      - uses: actions/checkout@v4
-      - name: Test
-        run: .github/workflows/ci.sh "${{matrix.components}}" ${{matrix.debug}} ${{matrix.shlib}}
   papi_spack:
     runs-on: cpu
     timeout-minutes: 60

src/ctests/all_native_events.c

@@ -80,13 +80,20 @@ main( int argc, char **argv )
 	PAPI_event_info_t info, info1;
 	const PAPI_hw_info_t *hwinfo = NULL;
 	const PAPI_component_info_t* cmpinfo;
+        char enableCuda[PAPI_MIN_STR_LEN] = "no";
 	int event_code;
 	int numcmp, cid;
 	int quiet;
 
 	/* Set quiet variable */
 	quiet=tests_quiet( argc, argv );
 
+        for (i = 0; i < argc; i++) {
+            if (strncmp(argv[i], "--enable_cuda_evts=", 19) == 0) {
+                strcpy(enableCuda, argv[i] + 19);
+            }
+        }
+
 	/* Init PAPI library */
 	retval = PAPI_library_init( PAPI_VER_CURRENT );
 	if ( retval != PAPI_VER_CURRENT ) {
@@ -107,7 +114,7 @@ main( int argc, char **argv )
 	numcmp = PAPI_num_components(  );
 
     int rocm_id = PAPI_get_component_index("rocm");
-
+        int cuda_id = PAPI_get_component_index("cuda");
 	/* Loop through all components */
 	for( cid = 0; cid < numcmp; cid++ ) {
 
@@ -117,6 +124,15 @@ main( int argc, char **argv )
             continue;
         }
 
+          /* possible option to fix the long run time of this file for ci, by completing skipping cuda tests */
+//        if (cid == cuda_id)
+//            continue;
+
+          /* possible option to fix the long run time of this file for CI, by using an option */
+          if (strcmp(enableCuda, "no") == 0 && cid == cuda_id) {
+              continue;
+          }
+
 		cmpinfo = PAPI_get_component_info( cid );
 		if (cmpinfo  == NULL) {
 			test_fail( __FILE__, __LINE__, "PAPI_get_component_info", 2 );