kill whole process group when retrying

pytorch · May 9, 2024 · 21d0482 · 21d0482
1 parent b20d82a
commit 21d0482
Showing 1 changed file with 3 additions and 2 deletions.
diff --git a/.github/workflows/unit-tests.yml b/.github/workflows/unit-tests.yml
@@ -127,9 +127,10 @@ jobs:
           shell: bash
           command: SKIP_DISTRIB_TESTS=${{ matrix.skip-distrib-tests }} bash tests/run_cpu_tests.sh
           on_retry_command: |
+            echo sending kill signal until process group no longer exists...
             tests_pid=$(cat /tmp/ignite_testing.pid)
-            kill -0 $tests_pid && kill -INT $tests_pid && sleep 5
-            while kill -0 $tests_pid > /dev/null 2>&1; do kill -9 $tests_pid; sleep 3; done
+            kill -0 -- -$tests_pid && kill -INT -- -$tests_pid && sleep 5
+            while kill -0 -- -$tests_pid > /dev/null 2>&1; do kill -9 -- -$tests_pid; sleep 3; done
           new_command_on_retry: USE_LAST_FAILED=1  SKIP_DISTRIB_TESTS=${{ matrix.skip-distrib-tests }} bash tests/run_cpu_tests.sh
 
       - name: Upload coverage to Codecov