diff options
| author | bors <bors@rust-lang.org> | 2019-10-29 11:28:28 +0000 |
|---|---|---|
| committer | bors <bors@rust-lang.org> | 2019-10-29 11:28:28 +0000 |
| commit | d3d28a49209a21628fda0245b631e5fc3465be1a (patch) | |
| tree | 94730029bbe8da88b9f7fdacf9113748446e4fd7 /src/rustllvm/RustWrapper.cpp | |
| parent | eb5ef813f0d6e3fe8edd3abb046a18f5b1a8cc48 (diff) | |
| parent | beb06aed9af27fbf192d772e39ffe0421c50e02e (diff) | |
| download | rust-d3d28a49209a21628fda0245b631e5fc3465be1a.tar.gz rust-d3d28a49209a21628fda0245b631e5fc3465be1a.zip | |
Auto merge of #65927 - eddyb:eval-always-considered-harmful, r=michaelwoerister
Don't use eval_always for miri queries used from codegen. This should fix the [massive incremental perf regression](https://perf.rust-lang.org/compare.html?start=95f437b3cfb2fec966d7eaf69d7c2e36f9c274d1&end=9285d401a6070094747465962bc49969b93e14c5&stat=instructions:u) introduced in #65664. It seems that `eval_always` was mistakenly(?) added to `const_field` and then it ended up on `const_caller_location` (which is used much more often than `const_field` is). r? @michaelwoerister cc @oli-obk @nnethercote
Diffstat (limited to 'src/rustllvm/RustWrapper.cpp')
0 files changed, 0 insertions, 0 deletions