time: prevent the loop in timespec_add_ns() from being optimised away
authorSegher Boessenkool <segher@kernel.crashing.org>
Tue, 4 Mar 2008 22:59:54 +0000 (14:59 -0800)
committerThomas Gleixner <tglx@linutronix.de>
Sun, 9 Mar 2008 07:42:57 +0000 (08:42 +0100)
Since some architectures don't support __udivdi3().

Signed-off-by: Segher Boessenkool <segher@kernel.crashing.org>
Cc: john stultz <johnstul@us.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
include/linux/time.h

index 2091a19f1655aab8e5130256f83af4103efa9e3f..d32ef0ad4c0aaf8964d94abe215c5dbf4dde8aaf 100644 (file)
@@ -174,6 +174,10 @@ static inline void timespec_add_ns(struct timespec *a, u64 ns)
 {
        ns += a->tv_nsec;
        while(unlikely(ns >= NSEC_PER_SEC)) {
+               /* The following asm() prevents the compiler from
+                * optimising this loop into a modulo operation.  */
+               asm("" : "+r"(ns));
+
                ns -= NSEC_PER_SEC;
                a->tv_sec++;
        }