[PATCH v4 14/15] ARM: local timers: convert MSM to runtime registration interface

On 27/01/12 16:08, Stephen Boyd wrote:

On 01/20/12 04:04, Marc Zyngier wrote:

quoted

Convert the MSM timers to the runtime registration interface.

Acked-by: Stephen Boyd <redacted>

Thanks Stephen.

quoted

Tested-by: David Brown <redacted>
Acked-by: David Brown <redacted>
Signed-off-by: Marc Zyngier <redacted>
---
 arch/arm/mach-msm/timer.c |   79 +++++++++++++++++++++++++--------------------
 1 files changed, 44 insertions(+), 35 deletions(-)

diff --git a/arch/arm/mach-msm/timer.c b/arch/arm/mach-msm/timer.c
index 11d0d8f..75f4be4 100644
--- a/arch/arm/mach-msm/timer.c
+++ b/arch/arm/mach-msm/timer.c

[snip]

quoted

+
+static struct local_timer_ops msm_local_timer_ops __cpuinitdata = {
+	.setup	= msm_local_timer_setup,
+	.stop	= msm_local_timer_stop,

Would this be a good time to rename these to start/stop? setup sounds
like it's done once, but its actually done every cpu up from what I recall.

I thought about that. But you could also argue that next_event() is also
a kind of start for the timer. Actually, if I was to change anything
about the naming, I'd turn stop to teardown.

And not changing the naming makes the patch series smaller and more
easily verified.

quoted

+};
+#endif /* CONFIG_LOCAL_TIMERS */
+
 static void __init msm_timer_init(void)
 {
 	struct clock_event_device *ce = &msm_clockevent;

@@ -173,8 +212,12 @@ static void __init msm_timer_init(void)
 		*__this_cpu_ptr(msm_evt.percpu_evt) = ce;
 		res = request_percpu_irq(ce->irq, msm_timer_interrupt,
 					 ce->name, msm_evt.percpu_evt);
-		if (!res)
+		if (!res) {
 			enable_percpu_irq(ce->irq, 0);
+#ifdef CONFIG_LOCAL_TIMERS
+			local_timer_register(&msm_local_timer_ops);
+#endif
+		}

It's unfortunate we can't avoid this ifdef here. O well.

I tried. It's actually uglier without the ifdef...

	M.
-- 
Jazz is not dead. It just smells funny...