diff options
| author | bors[bot] <26634292+bors[bot]@users.noreply.github.com> | 2020-10-25 12:54:18 +0000 |
|---|---|---|
| committer | GitHub <[email protected]> | 2020-10-25 12:54:18 +0000 |
| commit | 91c1af36120ed9b5ad1da4972ab18d0d042e991f (patch) | |
| tree | 06fc97459be2a8a210d03a43542c9297a9d8d4f6 /crates/hir_def/src/item_tree | |
| parent | 34034e4a64cd49e82171601e55ab6dd1d5b52cd4 (diff) | |
| parent | 39dfca23f1634c1e21819806bdd696c8096d0f0f (diff) | |
Merge #6354
6354: Add tracing to main rust-analyzer binary r=flodiebold a=flodiebold
This makes `CHALK_DEBUG` logging work again e.g. when running `analysis-stats`, which is very helpful for debugging.
This change shouldn't regress compile times at all. The reason for that is that chalk-solve already pulls in these crates, and while that's behind a feature (mostly for our benefit, I think) we never actually disabled that feature :sweat_smile: So alternatively, we could disable the feature and maybe get an improvement in compile times. In my test I just did to see the impact of that, this PR actually compiled faster than the one just removing tracing though, so it's probably not a big deal.
Co-authored-by: Florian Diebold <[email protected]>
Diffstat (limited to 'crates/hir_def/src/item_tree')
0 files changed, 0 insertions, 0 deletions