Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1271137 > unrolled thread

Re: [PATCH] clocksource/drivers/tango-xtal: Replace code by clocksource_mmio_init

Started byDaniel Lezcano <daniel.lezcano@linaro.org>
First post2015-11-17 13:30 +0100
Last post2015-11-17 14:20 +0100
Articles 4 — 3 participants

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Re: [PATCH] clocksource/drivers/tango-xtal: Replace code by  clocksource_mmio_init Daniel Lezcano <daniel.lezcano@linaro.org> - 2015-11-17 13:30 +0100
    Re: [PATCH] clocksource/drivers/tango-xtal: Replace code by  clocksource_mmio_init Marc Gonzalez <marc_gonzalez@sigmadesigns.com> - 2015-11-17 13:50 +0100
      Re: [PATCH] clocksource/drivers/tango-xtal: Replace code by clocksource_mmio_init Måns Rullgård <mans@mansr.com> - 2015-11-17 14:10 +0100
      Re: [PATCH] clocksource/drivers/tango-xtal: Replace code by  clocksource_mmio_init Daniel Lezcano <daniel.lezcano@linaro.org> - 2015-11-17 14:20 +0100

#1271137 — Re: [PATCH] clocksource/drivers/tango-xtal: Replace code by clocksource_mmio_init

FromDaniel Lezcano <daniel.lezcano@linaro.org>
Date2015-11-17 13:30 +0100
SubjectRe: [PATCH] clocksource/drivers/tango-xtal: Replace code by clocksource_mmio_init
Message-ID<qvNyy-4XK-21@gated-at.bofh.it>
On 11/13/2015 01:20 PM, Marc Gonzalez wrote:
> On 13/11/2015 11:58, Daniel Lezcano wrote:
>
>> The current code to initialize, register and read the clocksource is
>> already factored out in mmio.c via the clocksource_mmio_init function.
>>
>> Factor out the code with the clocksource_mmio_init function.
>
> The reason I didn't like clocksource_mmio_init() is because it exports
> 4 generic accessors.
>
> I guess this function makes more sense when all platforms are using it,
> in an ARCH_MULTIPLATFORM kernel. (Also the accessors are probably quite
> small, so the waste is probably minimal.)

Hi Marc,

it is not clear for me if you agree with this patch or not. Can you 
clarify ?

Thanks

   -- Daniel

-- 
  <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs

Follow Linaro:  <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [next] | [standalone]


#1271145

FromMarc Gonzalez <marc_gonzalez@sigmadesigns.com>
Date2015-11-17 13:50 +0100
Message-ID<qvNRT-54p-3@gated-at.bofh.it>
In reply to#1271137
On 17/11/2015 13:22, Daniel Lezcano wrote:
> On 11/13/2015 01:20 PM, Marc Gonzalez wrote:
>> On 13/11/2015 11:58, Daniel Lezcano wrote:
>>
>>> The current code to initialize, register and read the clocksource is
>>> already factored out in mmio.c via the clocksource_mmio_init function.
>>>
>>> Factor out the code with the clocksource_mmio_init function.
>>
>> The reason I didn't like clocksource_mmio_init() is because it exports
>> 4 generic accessors.
>>
>> I guess this function makes more sense when all platforms are using it,
>> in an ARCH_MULTIPLATFORM kernel. (Also the accessors are probably quite
>> small, so the waste is probably minimal.)
> 
> Hi Marc,
> 
> it is not clear for me if you agree with this patch or not. Can you 
> clarify ?

[ Adding rmk as the original mmio.c author ]

Hello Daniel,

It's hard to give a straight answer. From my limited perspective (building
a kernel for Sigma boards only) I feel that mmio.c brings nothing. (As you
mentioned, it would be interesting to measure whether having reg_base in
the ctx, rather than as a global is better or worse for perf, though the
actual difference may be lost in the noise.)

On the other hand, I can see how a different perspective, such as yours,
may see benefits in having all drivers use the same APIs; and there may
be other savings for ARCH_MULTIPLATFORM builds with lots of platforms.

I guess if you think it is a good patch, I will defer to your experience.

I just wish Thomas would take a look at my mmio patch. I will make an
official submission, so that it can be properly shot down :-)

Regards.

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]


#1271166 — Re: [PATCH] clocksource/drivers/tango-xtal: Replace code by clocksource_mmio_init

FromMåns Rullgård <mans@mansr.com>
Date2015-11-17 14:10 +0100
SubjectRe: [PATCH] clocksource/drivers/tango-xtal: Replace code by clocksource_mmio_init
Message-ID<qvObg-5rX-1@gated-at.bofh.it>
In reply to#1271145
Marc Gonzalez <marc_gonzalez@sigmadesigns.com> writes:

> On 17/11/2015 13:22, Daniel Lezcano wrote:
>> On 11/13/2015 01:20 PM, Marc Gonzalez wrote:
>>> On 13/11/2015 11:58, Daniel Lezcano wrote:
>>>
>>>> The current code to initialize, register and read the clocksource is
>>>> already factored out in mmio.c via the clocksource_mmio_init function.
>>>>
>>>> Factor out the code with the clocksource_mmio_init function.
>>>
>>> The reason I didn't like clocksource_mmio_init() is because it exports
>>> 4 generic accessors.
>>>
>>> I guess this function makes more sense when all platforms are using it,
>>> in an ARCH_MULTIPLATFORM kernel. (Also the accessors are probably quite
>>> small, so the waste is probably minimal.)
>> 
>> Hi Marc,
>> 
>> it is not clear for me if you agree with this patch or not. Can you 
>> clarify ?
>
> [ Adding rmk as the original mmio.c author ]
>
> Hello Daniel,
>
> It's hard to give a straight answer. From my limited perspective (building
> a kernel for Sigma boards only) I feel that mmio.c brings nothing. (As you
> mentioned, it would be interesting to measure whether having reg_base in
> the ctx, rather than as a global is better or worse for perf, though the
> actual difference may be lost in the noise.)
>
> On the other hand, I can see how a different perspective, such as yours,
> may see benefits in having all drivers use the same APIs; and there may
> be other savings for ARCH_MULTIPLATFORM builds with lots of platforms.

I think the patch is good.  If there are issues with clocksource_mmio
those should addressed separately.

-- 
Måns Rullgård
mans@mansr.com
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]


#1271171

FromDaniel Lezcano <daniel.lezcano@linaro.org>
Date2015-11-17 14:20 +0100
Message-ID<qvOkV-5vY-3@gated-at.bofh.it>
In reply to#1271145
On 11/17/2015 01:48 PM, Marc Gonzalez wrote:
> On 17/11/2015 13:22, Daniel Lezcano wrote:
>> On 11/13/2015 01:20 PM, Marc Gonzalez wrote:
>>> On 13/11/2015 11:58, Daniel Lezcano wrote:
>>>
>>>> The current code to initialize, register and read the clocksource is
>>>> already factored out in mmio.c via the clocksource_mmio_init function.
>>>>
>>>> Factor out the code with the clocksource_mmio_init function.
>>>
>>> The reason I didn't like clocksource_mmio_init() is because it exports
>>> 4 generic accessors.
>>>
>>> I guess this function makes more sense when all platforms are using it,
>>> in an ARCH_MULTIPLATFORM kernel. (Also the accessors are probably quite
>>> small, so the waste is probably minimal.)
>>
>> Hi Marc,
>>
>> it is not clear for me if you agree with this patch or not. Can you
>> clarify ?
>
> [ Adding rmk as the original mmio.c author ]

Good idea. Russell, the patch is https://lkml.org/lkml/2015/11/13/261

> Hello Daniel,
>
> It's hard to give a straight answer. From my limited perspective (building
> a kernel for Sigma boards only) I feel that mmio.c brings nothing. (As you
> mentioned, it would be interesting to measure whether having reg_base in
> the ctx, rather than as a global is better or worse for perf, though the
> actual difference may be lost in the noise.)
>
> On the other hand, I can see how a different perspective, such as yours,
> may see benefits in having all drivers use the same APIs; and there may
> be other savings for ARCH_MULTIPLATFORM builds with lots of platforms.

IMO, if all the drivers are using the same code, then optimizing it will 
benefit to all drivers at the same time.

> I guess if you think it is a good patch, I will defer to your experience.

Ok, I will wait for Russell comments then.

Thanks!

   -- Daniel

> I just wish Thomas would take a look at my mmio patch. I will make an
> official submission, so that it can be properly shot down :-)



-- 
  <http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs

Follow Linaro:  <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web