[Sbcl-commits] master: Microoptimisation for block headers on x86-64

[Paul K. <pk...@us...>]

The branch "master" has been updated in SBCL:
       via  ce118eb9bca784396335f6f10d59219362545afa (commit)
      from  6596b18b95780986ff4eb511ab384da138adbb58 (commit)

- Log -----------------------------------------------------------------
commit ce118eb9bca784396335f6f10d59219362545afa
Author: Paul Khuong <pv...@pv...>
Date:   Thu Jun 23 11:05:54 2011 -0400

    Microoptimisation for block headers on x86-64
    
     Only emit a jmp over the header if a predecessor falls through to
     the current block, and it there is a trampoline.  Otherwise, it's
     not worth the cost of a branch to skip a few NOPs.
---
 src/compiler/x86-64/call.lisp |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/src/compiler/x86-64/call.lisp b/src/compiler/x86-64/call.lisp
index b19771d..4071fb7 100644
--- a/src/compiler/x86-64/call.lisp
+++ b/src/compiler/x86-64/call.lisp
@@ -561,7 +561,7 @@
 ;;; calls. Emit a header for local calls to pop the return address
 ;;; in the right place.
 (defun emit-block-header (start-label trampoline-label fall-thru-p alignp)
-  (when (and fall-thru-p (or trampoline-label alignp))
+  (when (and fall-thru-p trampoline-label)
     (inst jmp start-label))
   (when alignp
     (emit-alignment n-lowtag-bits #x90))

-----------------------------------------------------------------------


hooks/post-receive
-- 
SBCL