Closed smarr closed 5 months ago
I was a bit too eager and didn't notice that benchmarking was broken for #190, which fixed memory leaks as noticed in #188.
This PR fixes the initialization of ChartJS, doing it now unconditionally when even the relevant code is loaded.
This PR also fixes the Dockerfile so that the patching from #190 also works.
And with that, we see that results are generated now in 22%-40% less time:
https://rebench.dev/ReBenchDB/compare/b0b149a77371ca2e685cf0aab786a32ec26ca958..55c5c5d7ba7d1f6cd4ebbef5febae9bd3bd40539
I was a bit too eager and didn't notice that benchmarking was broken for #190, which fixed memory leaks as noticed in #188.
This PR fixes the initialization of ChartJS, doing it now unconditionally when even the relevant code is loaded.
This PR also fixes the Dockerfile so that the patching from #190 also works.
And with that, we see that results are generated now in 22%-40% less time:
https://rebench.dev/ReBenchDB/compare/b0b149a77371ca2e685cf0aab786a32ec26ca958..55c5c5d7ba7d1f6cd4ebbef5febae9bd3bd40539