MFlowCode · sbryngelson · Mar 2, 2026
@@ -36,3 +36,6 @@ checkall: &checkall
   - *tests
   - *scripts
   - *yml
+
+cases_py:
+  - 'toolchain/mfc/test/cases.py'
@@ -18,12 +18,18 @@ if [ -n "$job_shard" ]; then
     shard_opts="--shard $job_shard"
 fi
 
+# Only prune tests on PRs; master pushes must run the full suite.
+prune_flag=""
+if [ "$GITHUB_EVENT_NAME" = "pull_request" ]; then
+    prune_flag="--only-changes"
+fi
+
 if [ "$job_device" = "gpu" ]; then
     rdma_opts=""
     if [ "$job_cluster" = "frontier" ]; then
         rdma_opts="--rdma-mpi"
     fi
-    ./mfc.sh test -v -a $rdma_opts --max-attempts 3 -j $ngpus $device_opts $shard_opts -- -c $job_cluster
+    ./mfc.sh test -v -a $rdma_opts --max-attempts 3 $prune_flag -j $ngpus $device_opts $shard_opts -- -c $job_cluster
 else
-    ./mfc.sh test -v -a --max-attempts 3 -j 32 --no-gpu $shard_opts -- -c $job_cluster
+    ./mfc.sh test -v -a --max-attempts 3 $prune_flag -j 32 --no-gpu $shard_opts -- -c $job_cluster
 fi
@@ -18,12 +18,18 @@ if [ -n "$job_shard" ]; then
     shard_opts="--shard $job_shard"
 fi
 
+# Only prune tests on PRs; master pushes must run the full suite.
+prune_flag=""
+if [ "$GITHUB_EVENT_NAME" = "pull_request" ]; then
+    prune_flag="--only-changes"
+fi
+
 if [ "$job_device" = "gpu" ]; then
     rdma_opts=""
     if [ "$job_cluster" = "frontier" ]; then
         rdma_opts="--rdma-mpi"
     fi
-    ./mfc.sh test -v -a $rdma_opts --max-attempts 3 -j $ngpus $device_opts $shard_opts -- -c $job_cluster
+    ./mfc.sh test -v -a $rdma_opts --max-attempts 3 $prune_flag -j $ngpus $device_opts $shard_opts -- -c $job_cluster
 else
-    ./mfc.sh test -v -a --max-attempts 3 -j 32 --no-gpu $shard_opts -- -c $job_cluster
+    ./mfc.sh test -v -a --max-attempts 3 $prune_flag -j 32 --no-gpu $shard_opts -- -c $job_cluster
 fi
@@ -0,0 +1,21 @@
+#!/bin/bash
+set -e
+
+# Number of parallel jobs: use SLURM allocation or default to 24.
+# Cap at 64 to avoid overwhelming MPI's ORTE daemons with concurrent launches.
+NJOBS="${SLURM_CPUS_ON_NODE:-24}"
+if [ "$NJOBS" -gt 64 ]; then NJOBS=64; fi
+
+# Clean stale build artifacts: the self-hosted runner may have a cached
+# GPU build (e.g. --gpu mp) whose CMake flags are incompatible with gcov.
+./mfc.sh clean
+
+# Build MFC with gcov coverage instrumentation (CPU-only, gfortran).
+# -j 8 for compilation (memory-heavy, more cores doesn't help much).
+./mfc.sh build --gcov -j 8
+
+# Run all tests in parallel, collecting per-test coverage data.
+# Each test gets an isolated GCOV_PREFIX directory so .gcda files
+# don't collide. Coverage is collected per-test after all tests finish.
+# --gcov is required so the internal build step preserves instrumentation.
+./mfc.sh test --build-coverage-cache --gcov -j "$NJOBS"
@@ -24,9 +24,9 @@ case "$script_basename" in
 esac
 
 sbatch_cpu_opts="\
-#SBATCH -p cpu-small               # partition
-#SBATCH --ntasks-per-node=24       # Number of cores per node required
-#SBATCH --mem-per-cpu=2G           # Memory per core\
+#SBATCH -p cpu-gnr                 # partition (full Granite Rapids node)
+#SBATCH --exclusive                # exclusive access to all cores
+#SBATCH -C graniterapids           # constrain to GNR architecture\
 "
 
 if [ "$job_type" = "bench" ]; then

@@ -51,7 +51,9 @@ while [ $attempt -le $max_attempts ]; do
     attempt=$((attempt + 1))
 done
 
-n_test_threads=8
+# Use up to 64 parallel test threads on CPU (GNR nodes have 192 cores).
+# Cap at 64 to avoid overwhelming MPI's ORTE daemons with concurrent launches.
+n_test_threads=$(( SLURM_CPUS_ON_NODE > 64 ? 64 : ${SLURM_CPUS_ON_NODE:-8} ))
 
 if [ "$job_device" = "gpu" ]; then
     gpu_count=$(nvidia-smi -L | wc -l)        # number of GPUs on node
@@ -60,4 +62,10 @@ if [ "$job_device" = "gpu" ]; then
     n_test_threads=`expr $gpu_count \* 2`
 fi
 
-./mfc.sh test -v --max-attempts 3 -a -j $n_test_threads $device_opts -- -c phoenix
+# Only prune tests on PRs; master pushes must run the full suite.
+prune_flag=""
+if [ "$GITHUB_EVENT_NAME" = "pull_request" ]; then
+    prune_flag="--only-changes"
+fi
+
+./mfc.sh test -v --max-attempts 3 $prune_flag -a -j $n_test_threads $device_opts -- -c phoenix
@@ -56,22 +56,106 @@ jobs:
   file-changes:
     name: Detect File Changes
     runs-on: 'ubuntu-latest'
-    outputs: 
+    outputs:
       checkall: ${{ steps.changes.outputs.checkall }}
+      cases_py: ${{ steps.changes.outputs.cases_py }}
+      dep_changed: ${{ steps.dep-check.outputs.dep_changed }}
     steps:
       - name: Clone
         uses: actions/checkout@v4
 
       - name: Detect Changes
         uses: dorny/paths-filter@v3
         id: changes
-        with: 
+        with:
           filters: ".github/file-filter.yml"
 
+      - name: Check for Fortran dependency changes
+        id: dep-check
+        env:
+          GH_TOKEN: ${{ github.token }}
+        run: |
+          # Detect added/removed use/include statements that change the
+          # Fortran dependency graph, which would make the coverage cache stale.
+          if [ "${{ github.event_name }}" = "pull_request" ]; then
+            DIFF=$(gh pr diff ${{ github.event.pull_request.number }})
+          elif [ "${{ github.event_name }}" = "push" ]; then
+            DIFF=$(git diff ${{ github.event.before }}..${{ github.event.after }} 2>/dev/null || echo "")
+          else
+            DIFF=""
+          fi
+          if echo "$DIFF" | \
+             grep -qP '^\+\s*(use[\s,]+\w|#:include\s|include\s+['"'"'"])'; then
+            echo "dep_changed=true" >> "$GITHUB_OUTPUT"
+            echo "Fortran dependency change detected — will rebuild coverage cache."
+          else
+            echo "dep_changed=false" >> "$GITHUB_OUTPUT"
+          fi
+
+  rebuild-cache:
+    name: Rebuild Coverage Cache
+    needs: [lint-gate, file-changes]
+    if: >-
+      github.repository == 'MFlowCode/MFC' &&
+      (
+        (github.event_name == 'pull_request' &&
+         (needs.file-changes.outputs.cases_py == 'true' ||
+          needs.file-changes.outputs.dep_changed == 'true')) ||
+        (github.event_name == 'push' &&
+         (needs.file-changes.outputs.cases_py == 'true' ||
+          needs.file-changes.outputs.dep_changed == 'true')) ||
+        github.event_name == 'workflow_dispatch'
+      )
+    timeout-minutes: 240
+    runs-on:
+      group:  phoenix
+      labels: gt
+    permissions:
+      contents: write
+    steps:
+      - name: Clone
+        uses: actions/checkout@v4
+        with:
+          ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
+          clean: false
+
+      - name: Rebuild Cache via SLURM
+        run: bash .github/workflows/phoenix/submit.sh .github/workflows/phoenix/rebuild-cache.sh cpu none
+
+      - name: Print Logs
+        if:   always()
+        run:  cat rebuild-cache-cpu-none.out
+
+      - name: Upload Cache Artifact
+        if: github.event_name == 'pull_request'
+        uses: actions/upload-artifact@v4
+        with:
+          name: coverage-cache
+          path: toolchain/mfc/test/test_coverage_cache.json.gz
+          retention-days: 1
+
+      - name: Commit Cache to Master
+        if: github.event_name == 'push' || github.event_name == 'workflow_dispatch'
+        run: |
+          git config user.name "github-actions[bot]"
+          git config user.email "github-actions[bot]@users.noreply.github.com"
+          git add toolchain/mfc/test/test_coverage_cache.json.gz
+          if git diff --cached --quiet; then
+            echo "Coverage cache unchanged."
+          else
+            git commit -m "Regenerate gcov coverage cache [skip ci]"
+            git push
+          fi
+
   github:
     name: Github
-    if: needs.file-changes.outputs.checkall == 'true'
-    needs: [lint-gate, file-changes]
+    needs: [lint-gate, file-changes, rebuild-cache]
+    if: >-
+      always() &&
+      needs.lint-gate.result == 'success' &&
+      needs.file-changes.result == 'success' &&
+      needs.rebuild-cache.result != 'cancelled' &&
+      needs.file-changes.outputs.checkall == 'true'
     strategy:
       matrix:
         os:    ['ubuntu', 'macos']
@@ -91,13 +175,26 @@ jobs:
             intel: false
 
       fail-fast: false
-    continue-on-error: true
     runs-on: ${{ matrix.os }}-latest
 
     steps:
       - name: Clone
         uses: actions/checkout@v4
 
+      - name: Fetch master for coverage diff
+        run: |
+          git fetch origin master:master --depth=1
+          git fetch --deepen=200
+        continue-on-error: true
+
+      - name: Download Coverage Cache
+        if: needs.rebuild-cache.result == 'success'
+        uses: actions/download-artifact@v4
+        with:
+          name: coverage-cache
+          path: toolchain/mfc/test
+        continue-on-error: true
+
       - name: Setup MacOS
         if:   matrix.os == 'macos'
         run:  |
@@ -159,7 +256,7 @@ jobs:
         run: |
           rm -f tests/failed_uuids.txt
           TEST_EXIT=0
-          /bin/bash mfc.sh test -v --max-attempts 3 -j "$(nproc)" $TEST_ALL $TEST_PCT || TEST_EXIT=$?
+          /bin/bash mfc.sh test -v --max-attempts 3 -j "$(nproc)" $ONLY_CHANGES $TEST_ALL $TEST_PCT || TEST_EXIT=$?
 
           # Retry only if a small number of tests failed (sporadic failures)
           if [ -s tests/failed_uuids.txt ]; then
@@ -180,11 +277,19 @@ jobs:
         env:
           TEST_ALL: ${{ matrix.mpi == 'mpi' && '--test-all' || '' }}
           TEST_PCT: ${{ matrix.debug == 'debug' && '-% 20' || '' }}
+          ONLY_CHANGES: ${{ github.event_name == 'pull_request' && '--only-changes' || '' }}
 
   self:
     name: "${{ matrix.cluster_name }} (${{ matrix.device }}${{ matrix.interface != 'none' && format('-{0}', matrix.interface) || '' }}${{ matrix.shard != '' && format(' [{0}]', matrix.shard) || '' }})"
-    if: github.repository == 'MFlowCode/MFC' && needs.file-changes.outputs.checkall == 'true' && github.event.pull_request.draft != true
-    needs: [lint-gate, file-changes]
+    needs: [lint-gate, file-changes, rebuild-cache]
+    if: >-
+      always() &&
+      needs.lint-gate.result == 'success' &&
+      needs.file-changes.result == 'success' &&
+      needs.rebuild-cache.result != 'cancelled' &&
+      github.repository == 'MFlowCode/MFC' &&
+      needs.file-changes.outputs.checkall == 'true' &&
+      github.event.pull_request.draft != true
     continue-on-error: false
     timeout-minutes: 480
     strategy:
@@ -265,6 +370,14 @@ jobs:
         with:
           clean: false
 
+      - name: Download Coverage Cache
+        if: needs.rebuild-cache.result == 'success'
+        uses: actions/download-artifact@v4
+        with:
+          name: coverage-cache
+          path: toolchain/mfc/test
+        continue-on-error: true
+
       - name: Build
         if:   matrix.cluster != 'phoenix'
         uses: nick-fields/retry@ce71cc2ab81d554ebbe88c79ab5975992d79ba08 # v3
@@ -299,3 +412,4 @@ jobs:
         with:
           name: logs-${{ strategy.job-index }}-${{ steps.log.outputs.slug }}
           path: ${{ steps.log.outputs.slug }}.out
+
@@ -22,6 +22,9 @@ __pycache__
 # Auto-generated version file
 toolchain/mfc/_version.py
 
+# Raw coverage cache — legacy, not tracked (the .json.gz version IS committed)
+toolchain/mfc/test/test_coverage_cache.json
+
 # Auto-generated toolchain files (regenerate with: ./mfc.sh generate)
 toolchain/completions/mfc.bash
 toolchain/completions/_mfc

diff --git a/CMakeLists.txt b/CMakeLists.txt
@@ -131,13 +131,20 @@ if (CMAKE_Fortran_COMPILER_ID STREQUAL "GNU")
         add_compile_options(
             $<$<COMPILE_LANGUAGE:Fortran>:-fprofile-arcs>
             $<$<COMPILE_LANGUAGE:Fortran>:-ftest-coverage>
-            $<$<COMPILE_LANGUAGE:Fortran>:-O1>
-	    )
+        )
 
         add_link_options(
             $<$<COMPILE_LANGUAGE:Fortran>:-lgcov>
             $<$<COMPILE_LANGUAGE:Fortran>:--coverage>
         )
+
+        # Override Release -O3 with -O1 for gcov: coverage instrumentation is
+        # inaccurate at -O3, and aggressive codegen (e.g. AVX-512 FP16 on
+        # Granite Rapids) can emit instructions that older assemblers reject.
+        set(CMAKE_Fortran_FLAGS_RELEASE "-O1 -DNDEBUG" CACHE STRING "" FORCE)
+
+        # Use gfortran5 line markers so gcov can map coverage to .fpp sources.
+        set(FYPP_GCOV_OPTS "--line-marker-format=gfortran5")
     endif()
 
     if (CMAKE_BUILD_TYPE STREQUAL "Debug")
@@ -224,18 +231,25 @@ endif()
 
 if (CMAKE_BUILD_TYPE STREQUAL "Release")
     # Processor tuning: Check if we can target the host's native CPU's ISA.
-    CHECK_FORTRAN_COMPILER_FLAG("-march=native" SUPPORTS_MARCH_NATIVE)
-    if (SUPPORTS_MARCH_NATIVE)
-        add_compile_options($<$<COMPILE_LANGUAGE:Fortran>:-march=native>)
-    else()
-    	CHECK_FORTRAN_COMPILER_FLAG("-mcpu=native" SUPPORTS_MCPU_NATIVE)
-        if (SUPPORTS_MCPU_NATIVE)
-            add_compile_options($<$<COMPILE_LANGUAGE:Fortran>:-mcpu=native>)
+    # Skip for gcov builds — -march=native on newer CPUs (e.g. Granite Rapids)
+    # can emit instructions the system assembler doesn't support.
+    if (NOT MFC_GCov)
+        CHECK_FORTRAN_COMPILER_FLAG("-march=native" SUPPORTS_MARCH_NATIVE)
+        if (SUPPORTS_MARCH_NATIVE)
+            add_compile_options($<$<COMPILE_LANGUAGE:Fortran>:-march=native>)
+        else()
+            CHECK_FORTRAN_COMPILER_FLAG("-mcpu=native" SUPPORTS_MCPU_NATIVE)
+            if (SUPPORTS_MCPU_NATIVE)
+                add_compile_options($<$<COMPILE_LANGUAGE:Fortran>:-mcpu=native>)
+            endif()
         endif()
     endif()
 
-    # Enable LTO/IPO if supported
-    if (CMAKE_Fortran_COMPILER_ID STREQUAL "NVHPC")
+    # Enable LTO/IPO if supported (skip for gcov — LTO interferes with coverage
+    # instrumentation and can trigger assembler errors on newer architectures).
+    if (MFC_GCov)
+        message(STATUS "LTO/IPO disabled for gcov build")
+    elseif (CMAKE_Fortran_COMPILER_ID STREQUAL "NVHPC")
         if (MFC_Unified)
             message(STATUS "LTO/IPO is not available with NVHPC using Unified Memory")
         elseif (CMAKE_Fortran_COMPILER_VERSION VERSION_GREATER "24.11" AND CMAKE_Fortran_COMPILER_VERSION VERSION_LESS "25.9")
@@ -381,6 +395,7 @@ macro(HANDLE_SOURCES target useCommon)
                                  --no-folding
 								 --line-length=999
 		 						 --line-numbering-mode=nocontlines
+                                 ${FYPP_GCOV_OPTS}
                                  "${fpp}" "${f90}"
             DEPENDS  "${fpp};${${target}_incs}"
             COMMENT  "Preprocessing (Fypp) ${fpp_filename}"

diff --git a/src/simulation/m_bubbles.fpp b/src/simulation/m_bubbles.fpp
@@ -16,6 +16,7 @@ module m_bubbles
     use m_variables_conversion !< State variables type conversion procedures
 
     use m_helper_basic         !< Functions to compare floating point numbers
+    use m_helper_basic
 
     implicit none