Message ID | 1482522460-26613-1-git-send-email-tuliom@linux.vnet.ibm.com |
---|---|
State | New |
Headers | show |
On 12/23/16, 5:47 PM, "Tulio Magno Quites Machado Filho" <libc-alpha-owner@sourceware.org on behalf of tuliom@linux.vnet.ibm.com> wrote: The same error fixed in commit b224637928e9fc04e3cef3e10d02ccf042d01584 happens in the 32-bit implementation of memchr for power7. This patch adopts the same solution, with a minimal change: it implements a saturated addition where overflows sets the maximum pointer size to UINTPTR_MAX. LGTM
diff --git a/sysdeps/powerpc/powerpc32/power7/memchr.S b/sysdeps/powerpc/powerpc32/power7/memchr.S index 318168b..5ce2bb0 100644 --- a/sysdeps/powerpc/powerpc32/power7/memchr.S +++ b/sysdeps/powerpc/powerpc32/power7/memchr.S @@ -26,7 +26,16 @@ ENTRY (__memchr) dcbt 0,r3 clrrwi r8,r3,2 insrwi r4,r4,8,16 /* Replicate byte to word. */ - add r7,r3,r5 /* Calculate the last acceptable address. */ + + /* Calculate the last acceptable address and check for possible + addition overflow by using satured math: + r7 = r3 + r5 + r7 |= -(r7 < x) */ + add r7,r3,r5 + subfc r6,r3,r7 + subfe r9,r9,r9 + or r7,r7,r9 + insrwi r4,r4,16,0 cmplwi r5,16 li r9, -1