CI: Run js-benchmarks with --continue-on-failure for manual builds

If we ever want to rerun benchmarks for an older commit, e.g. if we
added new benchmarks that we would want historical performance data on,
we need to consider that some benchmarks might fail because of a bug or
missing feature. For manual runs of the upstream build, pass
`--continue-on-failure` so we simply skip these failing benchmarks.

.github/workflows/js-and-wasm-benchmarks.yml

@@ -72,14 +72,25 @@ jobs:
           cd wasm-repl
           tar -xvzf ladybird-wasm-${{ matrix.os_name }}-${{ matrix.arch }}.tar.gz
 
-      - name: 'Run the JS and Wasm benchmarks'
+      - name: 'Run the benchmarks'
         shell: bash
         run: |
           cd js-benchmarks
           python3 -m venv .venv
           source .venv/bin/activate
           python3 -m pip install -r requirements.txt
-          ./run.py --executable=${{ github.workspace }}/js-repl/bin/js --wasm-executable=${{ github.workspace }}/wasm-repl/bin/wasm --iterations=5
+
+          run_options="--iterations=5"
+
+          if [ "${{ github.event.workflow_run.event }}" = "workflow_dispatch" ]; then
+              # Upstream was run manually; we might run into failing benchmarks as a result of older builds.
+              run_options="${run_options} --continue-on-failure"
+          fi
+
+          ./run.py \
+              --executable=${{ github.workspace }}/js-repl/bin/js \
+              --wasm-executable=${{ github.workspace }}/wasm-repl/bin/wasm \
+              ${run_options}
 
       - name: 'Save results as an artifact'
         uses: actions/upload-artifact@v4