diff options
| author | bors <bors@rust-lang.org> | 2019-05-28 01:56:44 +0000 |
|---|---|---|
| committer | bors <bors@rust-lang.org> | 2019-05-28 01:56:44 +0000 |
| commit | 87ed0b421de4fada7cf68c7a6b80c5bf918c131a (patch) | |
| tree | 0fa0c42dad268e9eec46f9bf4518f1e6525970fe /src/test/ui/thinlto | |
| parent | b71117947ffdb87c7615ae846c94a4b3fbd90827 (diff) | |
| parent | 577ea539dc4a264b480404700a2463e657c09c87 (diff) | |
| download | rust-87ed0b421de4fada7cf68c7a6b80c5bf918c131a.tar.gz rust-87ed0b421de4fada7cf68c7a6b80c5bf918c131a.zip | |
Auto merge of #61036 - michaelwoerister:pgo-xlto-test, r=alexcrichton
PGO - Add a smoketest for combining PGO with cross-language LTO. This PR - Adds a test making sure that PGO can be combined with cross-language LTO. - Does a little cleanup on how the `pgo-use` flag is handled internally. - Makes the compiler error if the `pgo-use` file given to `rustc` doesn't actually exist. LLVM only gives a warning and then just doesn't do PGO. Clang, on the other hand, does give an error in this case. - Makes the build system also build `compiler-rt` when building LLDB. This way the Clang compiler that we get from building LLDB can perform PGO, which is something that the new test case wants to do. CI compile times shouldn't be affected too much.
Diffstat (limited to 'src/test/ui/thinlto')
0 files changed, 0 insertions, 0 deletions