[PATCH] xtensa: Switch to sched_clock_register()

[PATCH] xtensa: Switch to sched_clock_register()

[Stephen Boyd]

The 32 bit sched_clock interface now supports 64 bits. Upgrade
to the 64 bit function to allow us to remove the 32 bit
registration interface.

Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
---

Patch is based on Linus' tip. Not event compile tested.

 arch/xtensa/kernel/time.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/xtensa/kernel/time.c b/arch/xtensa/kernel/time.c
index 9af3dd8..415c3f6 100644
--- a/arch/xtensa/kernel/time.c
+++ b/arch/xtensa/kernel/time.c
@@ -36,7 +36,7 @@ static cycle_t ccount_read(struct clocksource *cs)
 	return (cycle_t)get_ccount();
 }
 
-static u32 notrace ccount_sched_clock_read(void)
+static u64 notrace ccount_sched_clock_read(void)
 {
 	return get_ccount();
 }
@@ -141,7 +141,7 @@ void __init time_init(void)
 	setup_irq(ccount_timer.evt.irq, &timer_irqaction);
 	ccount_timer.irq_enabled = 1;
 
-	setup_sched_clock(ccount_sched_clock_read, 32, ccount_freq);
+	sched_clock_register(ccount_sched_clock_read, 32, ccount_freq);
 }
 
 /*
-- 
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
hosted by The Linux Foundation

Re: [PATCH] xtensa: Switch to sched_clock_register()

[Max Filippov]

On Sat, Nov 16, 2013 at 3:31 AM, Stephen Boyd <sboyd@codeaurora.org> wrote:
> The 32 bit sched_clock interface now supports 64 bits. Upgrade
> to the 64 bit function to allow us to remove the 32 bit
> registration interface.
>
> Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
> ---
>
> Patch is based on Linus' tip. Not event compile tested.
>
>  arch/xtensa/kernel/time.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)

Build and run-tested, appears to work.
Acked-by: Max Filippov <jcmvbkbc@gmail.com>

Should it go through xtensa tree, or through some other tree as a part
of a bigger conversion series?

> diff --git a/arch/xtensa/kernel/time.c b/arch/xtensa/kernel/time.c
> index 9af3dd8..415c3f6 100644
> --- a/arch/xtensa/kernel/time.c
> +++ b/arch/xtensa/kernel/time.c
> @@ -36,7 +36,7 @@ static cycle_t ccount_read(struct clocksource *cs)
>         return (cycle_t)get_ccount();
>  }
>
> -static u32 notrace ccount_sched_clock_read(void)
> +static u64 notrace ccount_sched_clock_read(void)
>  {
>         return get_ccount();
>  }
> @@ -141,7 +141,7 @@ void __init time_init(void)
>         setup_irq(ccount_timer.evt.irq, &timer_irqaction);
>         ccount_timer.irq_enabled = 1;
>
> -       setup_sched_clock(ccount_sched_clock_read, 32, ccount_freq);
> +       sched_clock_register(ccount_sched_clock_read, 32, ccount_freq);
>  }
>
>  /*
> --
> The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
> hosted by The Linux Foundation
>



-- 
Thanks.
-- Max

Re: [PATCH] xtensa: Switch to sched_clock_register()

[Stephen Boyd]

On 11/17/13 12:27, Max Filippov wrote:
> On Sat, Nov 16, 2013 at 3:31 AM, Stephen Boyd <sboyd@codeaurora.org> wrote:
>> The 32 bit sched_clock interface now supports 64 bits. Upgrade
>> to the 64 bit function to allow us to remove the 32 bit
>> registration interface.
>>
>> Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
>> ---
>>
>> Patch is based on Linus' tip. Not event compile tested.
>>
>>  arch/xtensa/kernel/time.c | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
> Build and run-tested, appears to work.
> Acked-by: Max Filippov <jcmvbkbc@gmail.com>
>
> Should it go through xtensa tree, or through some other tree as a part
> of a bigger conversion series?

If it can go through xtensa that sounds like the best option. Otherwise
we can send it through the tip/timers tree.

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by The Linux Foundation

Re: [PATCH] xtensa: Switch to sched_clock_register()

[Stephen Boyd]

On 11/20/13 10:26, Stephen Boyd wrote:
> On 11/17/13 12:27, Max Filippov wrote:
>> On Sat, Nov 16, 2013 at 3:31 AM, Stephen Boyd <sboyd@codeaurora.org> wrote:
>>> The 32 bit sched_clock interface now supports 64 bits. Upgrade
>>> to the 64 bit function to allow us to remove the 32 bit
>>> registration interface.
>>>
>>> Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
>>> ---
>>>
>>> Patch is based on Linus' tip. Not event compile tested.
>>>
>>>  arch/xtensa/kernel/time.c | 4 ++--
>>>  1 file changed, 2 insertions(+), 2 deletions(-)
>> Build and run-tested, appears to work.
>> Acked-by: Max Filippov <jcmvbkbc@gmail.com>
>>
>> Should it go through xtensa tree, or through some other tree as a part
>> of a bigger conversion series?
> If it can go through xtensa that sounds like the best option. Otherwise
> we can send it through the tip/timers tree.
>

Ping? I don't see this patch in linux-next yet so should I send this
through the tip/timers tree?

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by The Linux Foundation

Re: [PATCH] xtensa: Switch to sched_clock_register()

[czankel]

Hi Stephen,

Sorry for the delay. Will add it later today.

Thanks,
-Chris

On 12/12/13 3:54 PM, Stephen Boyd wrote:
> On 11/20/13 10:26, Stephen Boyd wrote:
>> On 11/17/13 12:27, Max Filippov wrote:
>>> On Sat, Nov 16, 2013 at 3:31 AM, Stephen Boyd <sboyd@codeaurora.org> wrote:
>>>> The 32 bit sched_clock interface now supports 64 bits. Upgrade
>>>> to the 64 bit function to allow us to remove the 32 bit
>>>> registration interface.
>>>>
>>>> Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
>>>> ---
>>>>
>>>> Patch is based on Linus' tip. Not event compile tested.
>>>>
>>>>   arch/xtensa/kernel/time.c | 4 ++--
>>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>> Build and run-tested, appears to work.
>>> Acked-by: Max Filippov <jcmvbkbc@gmail.com>
>>>
>>> Should it go through xtensa tree, or through some other tree as a part
>>> of a bigger conversion series?
>> If it can go through xtensa that sounds like the best option. Otherwise
>> we can send it through the tip/timers tree.
>>
> Ping? I don't see this patch in linux-next yet so should I send this
> through the tip/timers tree?
>