diff options
| author | bors <bors@rust-lang.org> | 2022-04-30 05:06:47 +0000 |
|---|---|---|
| committer | bors <bors@rust-lang.org> | 2022-04-30 05:06:47 +0000 |
| commit | 9a98c63b30bd8419be662b8fb3f3b22ba33d72c5 (patch) | |
| tree | 1603f7b9476f6194d86bb9db61a7b8b1f3aea1a8 /compiler/rustc_codegen_ssa/src | |
| parent | 0c8e520b3e0d9fbdafb8311f1b18a1f847531051 (diff) | |
| parent | 7149bbcdc5739e4ed1fa3697c1b4d441f1d4441c (diff) | |
| download | rust-9a98c63b30bd8419be662b8fb3f3b22ba33d72c5.tar.gz rust-9a98c63b30bd8419be662b8fb3f3b22ba33d72c5.zip | |
Auto merge of #96500 - SparrowLii:rpo, r=tmiasko
Reduce duplication of RPO calculation of mir Computing the RPO of mir is not a low-cost thing, but it is duplicate in many places. In particular the `iterate_to_fixpoint` method which is called multiple times when computing the data flow. This PR reduces the number of times the RPO is recalculated as much as possible, which should save some compile time.
Diffstat (limited to 'compiler/rustc_codegen_ssa/src')
| -rw-r--r-- | compiler/rustc_codegen_ssa/src/mir/mod.rs | 1 |
1 files changed, 0 insertions, 1 deletions
diff --git a/compiler/rustc_codegen_ssa/src/mir/mod.rs b/compiler/rustc_codegen_ssa/src/mir/mod.rs index 6c139df0a85..0c958de64fa 100644 --- a/compiler/rustc_codegen_ssa/src/mir/mod.rs +++ b/compiler/rustc_codegen_ssa/src/mir/mod.rs @@ -244,7 +244,6 @@ pub fn codegen_mir<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>>( fx.debug_introduce_locals(&mut bx); // Codegen the body of each block using reverse postorder - // FIXME(eddyb) reuse RPO iterator between `analysis` and this. for (bb, _) in traversal::reverse_postorder(&mir) { fx.codegen_block(bb); } |