s390/lib: improve memmove, memset and memcpy

Improve the memmove implementation to save one instruction and use
better label names. Also use better label names for the memset and
memcpy implementations so everything looks consistent.
Suggested-by: Jens Remus <jremus@linux.vnet.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

arch/s390/lib/mem.S

@@ -14,31 +14,29 @@ ENTRY(memmove)
 	ltgr	%r4,%r4
 	lgr	%r1,%r2
 	bzr	%r14
+	aghi	%r4,-1
 	clgr	%r2,%r3
 	jnh	.Lmemmove_forward
-	la	%r5,0(%r4,%r3)
+	la	%r5,1(%r4,%r3)
 	clgr	%r2,%r5
 	jl	.Lmemmove_reverse
 .Lmemmove_forward:
-	aghi	%r4,-1
 	srlg	%r0,%r4,8
 	ltgr	%r0,%r0
-	jz	.Lmemmove_rest
-.Lmemmove_loop:
+	jz	.Lmemmove_forward_remainder
+.Lmemmove_forward_loop:
 	mvc	0(256,%r1),0(%r3)
 	la	%r1,256(%r1)
 	la	%r3,256(%r3)
-	brctg	%r0,.Lmemmove_loop
-.Lmemmove_rest:
+	brctg	%r0,.Lmemmove_forward_loop
+.Lmemmove_forward_remainder:
 	larl	%r5,.Lmemmove_mvc
 	ex	%r4,0(%r5)
 	br	%r14
 .Lmemmove_reverse:
-	aghi	%r4,-1
-.Lmemmove_reverse_loop:
 	ic	%r0,0(%r4,%r3)
 	stc	%r0,0(%r4,%r1)
-	brctg	%r4,.Lmemmove_reverse_loop
+	brctg	%r4,.Lmemmove_reverse
 	ic	%r0,0(%r4,%r3)
 	stc	%r0,0(%r4,%r1)
 	br	%r14
@@ -70,12 +68,12 @@ ENTRY(memset)
 	srlg	%r3,%r4,8
 	ltgr	%r3,%r3
 	lgr	%r1,%r2
-	jz	.Lmemset_clear_rest
+	jz	.Lmemset_clear_remainder
 .Lmemset_clear_loop:
 	xc	0(256,%r1),0(%r1)
 	la	%r1,256(%r1)
 	brctg	%r3,.Lmemset_clear_loop
-.Lmemset_clear_rest:
+.Lmemset_clear_remainder:
 	larl	%r3,.Lmemset_xc
 	ex	%r4,0(%r3)
 	br	%r14
@@ -87,12 +85,12 @@ ENTRY(memset)
 	aghi	%r4,-2
 	srlg	%r3,%r4,8
 	ltgr	%r3,%r3
-	jz	.Lmemset_fill_rest
+	jz	.Lmemset_fill_remainder
 .Lmemset_fill_loop:
 	mvc	1(256,%r1),0(%r1)
 	la	%r1,256(%r1)
 	brctg	%r3,.Lmemset_fill_loop
-.Lmemset_fill_rest:
+.Lmemset_fill_remainder:
 	larl	%r3,.Lmemset_mvc
 	ex	%r4,0(%r3)
 	br	%r14
@@ -115,7 +113,7 @@ ENTRY(memcpy)
 	ltgr	%r5,%r5
 	lgr	%r1,%r2
 	jnz	.Lmemcpy_loop
-.Lmemcpy_rest:
+.Lmemcpy_remainder:
 	larl	%r5,.Lmemcpy_mvc
 	ex	%r4,0(%r5)
 	br	%r14
@@ -124,7 +122,7 @@ ENTRY(memcpy)
 	la	%r1,256(%r1)
 	la	%r3,256(%r3)
 	brctg	%r5,.Lmemcpy_loop
-	j	.Lmemcpy_rest
+	j	.Lmemcpy_remainder
 .Lmemcpy_mvc:
 	mvc	0(1,%r1),0(%r3)
 EXPORT_SYMBOL(memcpy)