The "rebuild benchmarks" CI workflow found an issue rebuilding the SpiderMonkey benchmark: https://github.com/bytecodealliance/sightglass/actions/runs/4453084060/jobs/7821309649#step:4:5101. This CI workflow only rebuilds a subset of the benchmarks each CI run (for time's sake) to check that they can be rebuilt deterministically, so the failure will only reappear occasionally. cc: @cfallin