Use msleep() instead of schedule_timeout() to guarantee the task
delays as expected. Neither signals nor wait-queue events are important at this
point in the code, I believe.
Signed-off-by: Nishanth Aravamudan <[email protected]>
Signed-off-by: Domen Puncer <[email protected]>
---
kj-domen/arch/arm/mach-sa1100/cpu-sa1110.c | 3 +--
1 files changed, 1 insertion(+), 2 deletions(-)
diff -puN arch/arm/mach-sa1100/cpu-sa1110.c~msleep-arch_arm_mach-sa1100_cpu-sa1110 arch/arm/mach-sa1100/cpu-sa1110.c
--- kj/arch/arm/mach-sa1100/cpu-sa1110.c~msleep-arch_arm_mach-sa1100_cpu-sa1110 2005-03-05 16:10:43.000000000 +0100
+++ kj-domen/arch/arm/mach-sa1100/cpu-sa1110.c 2005-03-05 16:10:43.000000000 +0100
@@ -271,8 +271,7 @@ static int sa1110_target(struct cpufreq_
*/
sdram_set_refresh(2);
if (!irqs_disabled()) {
- set_current_state(TASK_UNINTERRUPTIBLE);
- schedule_timeout(20 * HZ / 1000);
+ msleep(20);
} else {
mdelay(20);
}
_