[PATCH] powerpc: small fix to POWER7 32-bit memset
Luis Machado
luisgpm@linux.vnet.ibm.com
Mon May 24 19:04:00 GMT 2010
Hi,
This is a little gotcha that slipped through on my previous memset
patch. It just makes things simpler by not having an additional load
immediate instruction.
Works just the same as the other version.
Regards,
Luis
2010-05-24 Luis Machado <luisgpm@br.ibm.com>
* sysdeps/powerpc/powerpc32/power7/memset.S: POWER7 32-bit memset fix.
diff --git a/sysdeps/powerpc/powerpc32/power7/memset.S b/sysdeps/powerpc/powerpc32/power7/memset.S
index 99d07ec..8aabb49 100644
--- a/sysdeps/powerpc/powerpc32/power7/memset.S
+++ b/sysdeps/powerpc/powerpc32/power7/memset.S
@@ -155,7 +155,6 @@ L(big_loop):
/* Now that we're probably past the LHS window, use the VSX to
speed up the loop. */
L(big_loop_fast_setup):
- li 0,0
li 11,24
li 6,16
lxvdsx 4,1,11
@@ -163,12 +162,12 @@ L(big_loop_fast_setup):
.align 4
L(big_loop_fast):
addi 12,10,32
- stxvd2x 4,10,0
+ stxvd2x 4,0,10
stxvd2x 4,10,6
bdz L(tail_bytes)
addi 10,10,64
- stxvd2x 4,12,0
+ stxvd2x 4,0,12
stxvd2x 4,12,6
bdnz L(big_loop_fast)
More information about the Libc-alpha
mailing list