Permalink
Browse files

cmd/compile/internal/gc: update comment in plive.go

onebitwalktype1 no longer appears to be a bottleneck for the mentioned
test case. In fact, we appear to compile it significantly faster now
than Go 1.4 did (~1.8s vs ~3s).

Fixes #21951.

Change-Id: I315313e906092a7d6ff4ff60a918d80a4cff7a7f
Reviewed-on: https://go-review.googlesource.com/65110
Reviewed-by: Keith Randall <khr@golang.org>
  • Loading branch information...
mdempsky committed Sep 20, 2017
1 parent 1eacf78 commit 93e97ef066d3aad0a51a4a54212382f04c142255
Showing with 3 additions and 4 deletions.
  1. +3 −4 src/cmd/compile/internal/gc/plive.go
@@ -351,10 +351,9 @@ func (lv *Liveness) blockEffects(b *ssa.Block) *BlockEffects {
return &lv.be[b.ID]
}
// NOTE: The bitmap for a specific type t should be cached in t after the first run
// and then simply copied into bv at the correct offset on future calls with
// the same type t. On https://rsc.googlecode.com/hg/testdata/slow.go, onebitwalktype1
// accounts for 40% of the 6g execution time.
// NOTE: The bitmap for a specific type t could be cached in t after
// the first run and then simply copied into bv at the correct offset
// on future calls with the same type t.
func onebitwalktype1(t *types.Type, off int64, bv bvec) {
if t.Align > 0 && off&int64(t.Align-1) != 0 {
Fatalf("onebitwalktype1: invalid initial alignment, %v", t)

0 comments on commit 93e97ef

Please sign in to comment.