lkml.org 
[lkml]   [2015]   [Sep]   [3]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[RFC PATCH] clocksource: exynos_mct: use container_of() instead of this_cpu_ptr()
Date
Since evt structure is embedded in per-CPU mevt structure it's
definitely faster to use container_of() to get access to mevt
if we have evt (for example as incoming function argument) instead
of more expensive approach with this_cpu_ptr(&percpu_mct_tick).
this_cpu_ptr() on per-CPU mevt structure leads to access to cp15
to get cpu id and arithmetic operations.
Container_of() is cheaper since it's just one asm instruction.
This should work if used evt pointer is correct and owned by
local mevt structure.

For example, before this patch set_state_shutdown() looks like:

4a4: e92d4010 push {r4, lr}
4a8: e3004000 movw r4, #0
4ac: ebfffffe bl 0 <debug_smp_processor_id>
4b0: e3003000 movw r3, #0
4b4: e3404000 movt r4, #0
4b8: e3403000 movt r3, #0
4bc: e7933100 ldr r3, [r3, r0, lsl #2]
4c0: e0844003 add r4, r4, r3
4c4: e59400c0 ldr r0, [r4, #192] ; 0xc0
4c8: ebffffd4 bl 420 <exynos4_mct_tick_stop.isra.1>
4cc: e3a00000 mov r0, #0
4d0: e8bd8010 pop {r4, pc}

With this patch:

4a4: e92d4010 push {r4, lr}
4a8: e59000c0 ldr r0, [r0, #192] ; 0xc0
4ac: ebffffdb bl 420 <exynos4_mct_tick_stop.isra.1>
4b0: e3a00000 mov r0, #0
4b4: e8bd8010 pop {r4, pc}

Also, for me size of exynos_mct.o decreased from 84588 bytes
to 83956.

Signed-off-by: Alexey Klimov <alexey.klimov@linaro.org>
---
drivers/clocksource/exynos_mct.c | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/drivers/clocksource/exynos_mct.c b/drivers/clocksource/exynos_mct.c
index 029f96a..ff44082 100644
--- a/drivers/clocksource/exynos_mct.c
+++ b/drivers/clocksource/exynos_mct.c
@@ -382,24 +382,28 @@ static void exynos4_mct_tick_start(unsigned long cycles,
static int exynos4_tick_set_next_event(unsigned long cycles,
struct clock_event_device *evt)
{
- struct mct_clock_event_device *mevt = this_cpu_ptr(&percpu_mct_tick);
+ struct mct_clock_event_device *mevt;

+ mevt = container_of(evt, struct mct_clock_event_device, evt);
exynos4_mct_tick_start(cycles, mevt);
-
return 0;
}

static int set_state_shutdown(struct clock_event_device *evt)
{
- exynos4_mct_tick_stop(this_cpu_ptr(&percpu_mct_tick));
+ struct mct_clock_event_device *mevt;
+
+ mevt = container_of(evt, struct mct_clock_event_device, evt);
+ exynos4_mct_tick_stop(mevt);
return 0;
}

static int set_state_periodic(struct clock_event_device *evt)
{
- struct mct_clock_event_device *mevt = this_cpu_ptr(&percpu_mct_tick);
+ struct mct_clock_event_device *mevt;
unsigned long cycles_per_jiffy;

+ mevt = container_of(evt, struct mct_clock_event_device, evt);
cycles_per_jiffy = (((unsigned long long)NSEC_PER_SEC / HZ * evt->mult)
>> evt->shift);
exynos4_mct_tick_stop(mevt);
--
2.1.4


\
 
 \ /
  Last update: 2015-09-04 02:01    [W:0.036 / U:6.304 seconds]
©2003-2017 Jasper Spaans. hosted at Digital OceanAdvertise on this site