Reviewers: Erik Corry,
Description:
Lower waste from alignment of deferred code blocks.
Some ARM chips load instructions 8 byte at a time.
Please review this at http://codereview.chromium.org/2809029/show
Affected files:
M src/arm/assembler-arm.cc
Index: src/arm/assembler-arm.cc
diff --git a/src/arm/assembler-arm.cc b/src/arm/assembler-arm.cc
index
c16d6ec0aa06bfde455bd60eb0ae1d7de775800f..cc8cec457186178e18fb2582a6505572ca35e081
100644
--- a/src/arm/assembler-arm.cc
+++ b/src/arm/assembler-arm.cc
@@ -395,7 +395,10 @@ void Assembler::Align(int m) {
void Assembler::CodeTargetAlign() {
- Align(16); // Tentative value.
+ // Preferred alignment of jump targets on some ARM chips.
+ // Remove it if makes no difference in performance, since it costs
+ // 2 bytes on average.
+ Align(8);
}
--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
