rustc-perf
rustc-perf copied to clipboard
rust-lang
Website for graphing performance of rustc
Original problem: > I still can't manage to properly use rustc-perf locally. This time I got this obscure error when trying to look at per-query results when comparing a bench...
The benchmark suite includes two benchmarks that stress match checking (and I assume match codegen too): `unicode_normalization` and `match-stress-enum`. The problem with these two is that they really are not...
Maybe `--force`? My use case is that I made changes to `collector` because it was buggy, but now it refuses to collect the new benchmark because it has the same...
The coercions benchmark has been consistently quite noisy (over 1% regressions/improvements) for the last several months, at least, and may have recently gotten worse. The debug profile seems noisiest. We...
We currently only offer minimal assistance in analysis on the comparison page: * Noisy benchmarks are marked as such (though confusingly, it doesn't matter if one benchmark/incr-cache-state is noisy, the...
Currently, we only record the total build time for each crate in the 'bootstrap timings' section. Would it be possible to build the compiler while capturing the same level of...
Hi. There should be the way to cancel queued perf run - for example, if we accidentally queued one task few times, or if build fails. For example: https://github.com/rust-lang/rust/pull/83207 here...
To track down regressions due to LLVM's optimizations changing due to subtle code changes can be hard. One useful tool in that box is dumping the symbols in `librustc_driver.so` and...
cc https://github.com/rust-lang/rust/issues/83150 https://github.com/rust-lang/rust/pull/83406 fixed this bug but didn't actually show major performance changes, so we can regress and not notice.
In the ideal case the recompilation of the dependent crate would be exactly as fast as a completely unchanged incremental compilation if the dependency didn't change any declarations, generic function...
