Optimize memcpy(9) by always doing a forward copy; it should never be used

for overlapping copies.  Attempt to be instruction cache friendly by turning
things around and make memcpy(9) use the forward copy branch of memmove(9),
and implement bcopy(9) by swapping its arguments and dropping into memmove(9).

Same change as the one just made to for hppa.

1 files changed, 9 insertions, 9 deletions

diff --git a/sys/lib/libkern/arch/hppa64/bcopy.m4 b/sys/lib/libkern/arch/hppa64/bcopy.m4
index 977ece174c5..c2608e993d1 100644
--- a/sys/lib/libkern/arch/hppa64/bcopy.m4
+++ b/sys/lib/libkern/arch/hppa64/bcopy.m4
@@ -1,4 +1,4 @@
-define(_rcsid,``$OpenBSD: bcopy.m4,v 1.9 2011/11/27 21:02:10 kettenis Exp $'')dnl
+define(_rcsid,``$OpenBSD: bcopy.m4,v 1.10 2013/06/11 20:05:53 kettenis Exp $'')dnl
 dnl
 dnl
 dnl  This is the source file for bcopy.S, spcopy.S
@@ -183,22 +183,22 @@ L($1, done)
 '
 ifelse(NAME, `bcopy',
 `
-LEAF_ENTRY(memcpy)
-ALTENTRY(memmove)
+LEAF_ENTRY(bcopy)
+ALTENTRY(ovbcopy)
 	copy	%arg0, %ret0
 	copy	%arg1, %arg0
 	copy	%ret0, %arg1
-ALTENTRY(ovbcopy)
-ALTENTRY(bcopy)
-	cmpb,*>,n %arg1, %arg0, L(bcopy, reverse)
-	hppa_copy(bcopy_f, %sr0, %arg0, %sr0, %arg1, %arg2, `+')
+ALTENTRY(memmove)
+	cmpb,*>,n %arg0, %arg1, L(bcopy, reverse)
+ALTENTRY(memcpy)
+	hppa_copy(bcopy_f, %sr0, %arg1, %sr0, %arg0, %arg2, `+')
 	bv	%r0(%rp)
 	nop
 L(bcopy, reverse)
-	hppa_copy(bcopy_r, %sr0, %arg0, %sr0, %arg1, %arg2, `-')
+	hppa_copy(bcopy_r, %sr0, %arg1, %sr0, %arg0, %arg2, `-')
 	bv	%r0(%rp)
 	nop
-EXIT(memcpy)
+EXIT(bcopy)
 ')dnl
 dnl
 ifelse(NAME, `spcopy',