Minor optimization: movl and andl are slightly faster than movb and andb,

respectively.
This commit is contained in:
junyoung 2002-10-23 03:28:34 +00:00
parent f64a8c827d
commit 06ad574db3
1 changed files with 7 additions and 7 deletions

View File

@ -1,4 +1,4 @@
/* $NetBSD: locore.s,v 1.266 2002/10/23 00:58:28 christos Exp $ */
/* $NetBSD: locore.s,v 1.267 2002/10/23 03:28:34 junyoung Exp $ */
/*-
* Copyright (c) 1998, 2000 The NetBSD Foundation, Inc.
@ -1054,8 +1054,8 @@ ENTRY(i386_copyout)
shrl $2,%ecx
rep
movsl
movb %al,%cl
andb $3,%cl
movl %eax,%ecx
andl $3,%ecx
rep
movsb
@ -1096,8 +1096,8 @@ ENTRY(i486_copyout)
shrl $2,%ecx
rep
movsl
movb %al,%cl
andb $3,%cl
movl %eax,%ecx
andl $3,%ecx
rep
movsb
@ -1148,8 +1148,8 @@ ENTRY(i386_copyin)
shrl $2,%ecx
rep
movsl
movb %al,%cl
andb $3,%cl
movl %eax,%ecx
andl $3,%ecx
rep
movsb