powerpc: Fix string library functions

PATCH In review Andreas Schwab 2 participants 2 messages May 18, 2010 – May 20, 2010 PowerPC

Share & tools

Andreas Schwab

author

2010-05-18 18:15 UTC #0 lore raw

PATCH 1/1 powerpc: Fix string library functions

The powerpc strncmp implementation does not correctly handle a zero length, despite the claim in 0119536cd314ef95553604208c25bc35581f7f0a (Add hand-coded assembly strcmp).

Additionally, all the length arguments are size_t, not int, so use PPC_LCMPI and eq instead of cmpwi and le throughout.

Signed-off-by Andreas Schwab <[email protected]>

---
 arch/powerpc/lib/string.S |   18 ++++++++++--------
 1 files changed, 10 insertions(+), 8 deletions(-)

arch/powerpc/lib/string.S +10 -8

index 64e2e49..455881a 100644
@@ -28,7 +28,7 @@ _GLOBAL(strcpy)
 /* This clears out any unused part of the destination buffer,
    just as the libc version does.  -- paulus */
 _GLOBAL(strncpy)
-	cmpwi	0,r5,0
+	PPC_LCMPI 0,r5,0
 	beqlr
 	mtctr	r5
 	addi	r6,r3,-1

@@ -39,7 +39,7 @@ _GLOBAL(strncpy)
 	bdnzf	2,1b		/* dec ctr, branch if ctr != 0 && !cr0.eq */
 	bnelr			/* if we didn't hit a null char, we're done */
 	mfctr	r5
-	cmpwi	0,r5,0		/* any space left in destination buffer? */
+	PPC_LCMPI 0,r5,0	/* any space left in destination buffer? */
 	beqlr			/* we know r0 == 0 here */
 2:	stbu	r0,1(r6)	/* clear it out if so */
 	bdnz	2b

@@ -70,8 +70,8 @@ _GLOBAL(strcmp)
 	blr
 
 _GLOBAL(strncmp)
-	PPC_LCMPI r5,0
-	beqlr
+	PPC_LCMPI 0,r5,0
+	beq-	2f
 	mtctr	r5
 	addi	r5,r3,-1
 	addi	r4,r4,-1

@@ -82,6 +82,8 @@ _GLOBAL(strncmp)
 	beqlr	1
 	bdnzt	eq,1b
 	blr
+2:	li	r3,0
+	blr
 
 _GLOBAL(strlen)
 	addi	r4,r3,-1

@@ -92,8 +94,8 @@ _GLOBAL(strlen)
 	blr
 
 _GLOBAL(memcmp)
-	cmpwi	0,r5,0
-	ble-	2f
+	PPC_LCMPI 0,r5,0
+	beq-	2f
 	mtctr	r5
 	addi	r6,r3,-1
 	addi	r4,r4,-1

@@ -106,8 +108,8 @@ _GLOBAL(memcmp)
 	blr
 
 _GLOBAL(memchr)
-	cmpwi	0,r5,0
-	ble-	2f
+	PPC_LCMPI 0,r5,0
+	beq-	2f
 	mtctr	r5
 	addi	r3,r3,-1
 1:	lbzu	r0,1(r3)

Signature · 6 lines

1.7.1

-- 
Andreas Schwab, [email protected]
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Paul Mackerras

Acked-by on v1 1/1

2010-05-20 03:03 UTC #1 lore raw

Quoted 8 lines · Andreas Schwab

The powerpc strncmp implementation does not correctly handle a zero length, despite the claim in 0119536cd314ef95553604208c25bc35581f7f0a (Add hand-coded assembly strcmp).
Additionally, all the length arguments are size_t, not int, so use PPC_LCMPI and eq instead of cmpwi and le throughout.
Signed-off-by Andreas Schwab <[email protected]>

Looks fine to me.

Acked-by Paul Mackerras <[email protected]>