about summary refs log tree commit diff
path: root/compiler/rustc_middle/src
diff options
context:
space:
mode:
authorNicholas Nethercote <n.nethercote@gmail.com>2022-10-20 18:59:07 +1100
committerNicholas Nethercote <n.nethercote@gmail.com>2022-10-31 10:16:39 +1100
commit003a3f8cd3ea4594e2c2bc89c05562a202868ce0 (patch)
treeeded7bafa2d9a51eff572a3510d175c3a494191b /compiler/rustc_middle/src
parentf42b6fa7cad0d221b0c5407dca70b085784b1b5e (diff)
downloadrust-003a3f8cd3ea4594e2c2bc89c05562a202868ce0.tar.gz
rust-003a3f8cd3ea4594e2c2bc89c05562a202868ce0.zip
Use `br` instead of `switch` in more cases.
`codegen_switchint_terminator` already uses `br` instead of `switch`
when there is one normal target plus the `otherwise` target. But there's
another common case with two normal targets and an `otherwise` target
that points to an empty unreachable BB. This comes up a lot when
switching on the tags of enums that use niches.

The pattern looks like this:
```
bb1:                                              ; preds = %bb6
  %3 = load i8, ptr %_2, align 1, !range !9, !noundef !4
  %4 = sub i8 %3, 2
  %5 = icmp eq i8 %4, 0
  %_6 = select i1 %5, i64 0, i64 1
  switch i64 %_6, label %bb3 [
    i64 0, label %bb4
    i64 1, label %bb2
  ]

bb3:                                              ; preds = %bb1
  unreachable
```
This commit adds code to convert the `switch` to a `br`:
```
bb1:                                              ; preds = %bb6
  %3 = load i8, ptr %_2, align 1, !range !9, !noundef !4
  %4 = sub i8 %3, 2
  %5 = icmp eq i8 %4, 0
  %_6 = select i1 %5, i64 0, i64 1
  %6 = icmp eq i64 %_6, 0
  br i1 %6, label %bb4, label %bb2

bb3:                                              ; No predecessors!
  unreachable
```
This has a surprisingly large effect on compile times, with reductions
of 5% on debug builds of some crates. The reduction is all due to LLVM
taking less time. Maybe LLVM is just much better at handling `br` than
`switch`.

The resulting code is still suboptimal.
- The `icmp`, `select`, `icmp` sequence is silly, converting an `i1` to an `i64`
  and back to an `i1`. But with the current code structure it's hard to avoid,
  and LLVM will easily clean it up, in opt builds at least.
- `bb3` is usually now truly dead code (though not always, so it can't
  be removed universally).
Diffstat (limited to 'compiler/rustc_middle/src')
-rw-r--r--compiler/rustc_middle/src/mir/mod.rs5
1 files changed, 5 insertions, 0 deletions
diff --git a/compiler/rustc_middle/src/mir/mod.rs b/compiler/rustc_middle/src/mir/mod.rs
index 79db35a764a..068daaadbda 100644
--- a/compiler/rustc_middle/src/mir/mod.rs
+++ b/compiler/rustc_middle/src/mir/mod.rs
@@ -1186,6 +1186,11 @@ impl<'tcx> BasicBlockData<'tcx> {
     pub fn visitable(&self, index: usize) -> &dyn MirVisitable<'tcx> {
         if index < self.statements.len() { &self.statements[index] } else { &self.terminator }
     }
+
+    /// Does the block have no statements and an unreachable terminator?
+    pub fn is_empty_unreachable(&self) -> bool {
+        self.statements.is_empty() && matches!(self.terminator().kind, TerminatorKind::Unreachable)
+    }
 }
 
 impl<O> AssertKind<O> {