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

Groups > linux.kernel > #1260312 > unrolled thread

Re: [PATCH] clocksource: dw_apb_timer_of: support timer-based delay

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: dw_apb_timer_of: support timer-based delay Jisheng Zhang <jszhang@marvell.com> - 2015-11-02 04:10 +0100
    Re: [PATCH] clocksource: dw_apb_timer_of: support timer-based delay Arnd Bergmann <arnd@arndb.de> - 2015-11-02 23:00 +0100
      Re: [PATCH] clocksource: dw_apb_timer_of: support timer-based delay Jisheng Zhang <jszhang@marvell.com> - 2015-11-03 08:10 +0100
        Re: [PATCH] clocksource: dw_apb_timer_of: support timer-based delay Arnd Bergmann <arnd@arndb.de> - 2015-11-03 10:00 +0100
          Re: [PATCH] clocksource: dw_apb_timer_of: support timer-based delay Jisheng Zhang <jszhang@marvell.com> - 2015-11-03 11:00 +0100

#1260312 — Re: [PATCH] clocksource: dw_apb_timer_of: support timer-based delay

Dear Arnd and Daniel,

On Fri, 30 Oct 2015 13:42:01 +0100
Arnd Bergmann wrote:

> On Friday 30 October 2015 19:09:29 Jisheng Zhang wrote:
> > > > diff --git a/drivers/clocksource/Kconfig b/drivers/clocksource/Kconfig
> > > > index a7726db..7b081805 100644
> > > > --- a/drivers/clocksource/Kconfig
> > > > +++ b/drivers/clocksource/Kconfig
> > > > @@ -29,6 +29,16 @@ config DW_APB_TIMER_OF
> > > >     select DW_APB_TIMER
> > > >     select CLKSRC_OF
> > > >
> > > > +config DW_APB_TIMER_BASED_DELAY
> > > > +   bool "DW APB timer based delay"
> > > > +   depends on ARM && DW_APB_TIMER_OF
> > > > +   default n
> > > > +   help
> > > > +     This option enables support for using the DW APB timer to
> > > > +     implement timer-based delay. It is useful for skiping the
> > > > +     delay loop calibration at boot on some platforms. And the
> > > > +     udelay() will be unaffected by CPU frequency changes.
> > > > +    
> > > 
> > > Why do you want it to be optional ?
> > >   
> > 
> > Because in some platforms which has arm arch timer, this dw apb timer
> > delay isn't needed, the arch timer is better. So we want it be optional
> > so that the platforms which need this feature select it manually when config
> > the kernel.
> >   
> 
> This is not ideal from an overall maintenance perspective. We want to
> be able to have a kernel with all drivers enabled that gives us the
> best behavior on all platforms.
> 
> The current behavior appears to be that we override all previous
> registrations as long as the new one is higher resolution. Is that
> the case here? I.e. does the arch timer have a lower resultion than
> the dw-apb timer but have some other advantages?

Take one Marvell Berlin platform for example, the arch timer freq is 25MHZ,
whose resolution is lower than the dw apb timer at 100MHZ. But dw apb timer
is on the APB bus while arch timer sits in CPU, so I guess the cost of
accessing the apb timer is higher than arch timer. 

I have a solution for this case: in platforms with arch timer, I can mark
the dw apb timer as "disabled" in the dts even though the timer sits there.
Then I could make DW_APB_TIMER_BASED_DELAY non-optional but selected by the
the ARCH_XYZ. Is this acceptable?

Thanks in advance,
Jisheng
--
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]

#1261013

On Monday 02 November 2015 11:03:34 Jisheng Zhang wrote:
> On Fri, 30 Oct 2015 13:42:01 +0100 Arnd Bergmann wrote:
> > 
> > This is not ideal from an overall maintenance perspective. We want to
> > be able to have a kernel with all drivers enabled that gives us the
> > best behavior on all platforms.
> > 
> > The current behavior appears to be that we override all previous
> > registrations as long as the new one is higher resolution. Is that
> > the case here? I.e. does the arch timer have a lower resultion than
> > the dw-apb timer but have some other advantages?
> 
> Take one Marvell Berlin platform for example, the arch timer freq is 25MHZ,
> whose resolution is lower than the dw apb timer at 100MHZ. But dw apb timer
> is on the APB bus while arch timer sits in CPU, so I guess the cost of
> accessing the apb timer is higher than arch timer. 

Ok, I see.

> I have a solution for this case: in platforms with arch timer, I can mark
> the dw apb timer as "disabled" in the dts even though the timer sits there.
> Then I could make DW_APB_TIMER_BASED_DELAY non-optional but selected by the
> the ARCH_XYZ. Is this acceptable?

That would do the right thing, but doesn't look ideal: The DW_APB timer
on those platforms is fully functional, and a future Linux version or
another OS might decide to use both timers for one reason or another.

I'd be happier with a solution that keeps the DT describing the hardware
and not the way we expect Linux to use it, and instead has some heuristic
in the selection of the delay timer. At the moment, we purely base this
on the frequency, which as you say is suboptimal.

One possible way to improve this would be to add an optional 'latency'
property to the DT nodes (or the driver), and use a combination of latency
and resolution to make the decision.
A simpler way would be to always prefer the arch timer on ARM if that
is present, even if another timer has a higher resolution. This should
be only a few additional lines in register_current_timer_delay(), or
possibly an additional function argument.

	Arnd
--
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]

#1261230

Dear Arnd,

On Mon, 2 Nov 2015 22:56:02 +0100
Arnd Bergmann wrote:

> On Monday 02 November 2015 11:03:34 Jisheng Zhang wrote:
> > On Fri, 30 Oct 2015 13:42:01 +0100 Arnd Bergmann wrote:  
> > > 
> > > This is not ideal from an overall maintenance perspective. We want to
> > > be able to have a kernel with all drivers enabled that gives us the
> > > best behavior on all platforms.
> > > 
> > > The current behavior appears to be that we override all previous
> > > registrations as long as the new one is higher resolution. Is that
> > > the case here? I.e. does the arch timer have a lower resultion than
> > > the dw-apb timer but have some other advantages?  
> > 
> > Take one Marvell Berlin platform for example, the arch timer freq is 25MHZ,
> > whose resolution is lower than the dw apb timer at 100MHZ. But dw apb timer
> > is on the APB bus while arch timer sits in CPU, so I guess the cost of
> > accessing the apb timer is higher than arch timer.   
> 
> Ok, I see.
> 
> > I have a solution for this case: in platforms with arch timer, I can mark
> > the dw apb timer as "disabled" in the dts even though the timer sits there.
> > Then I could make DW_APB_TIMER_BASED_DELAY non-optional but selected by the
> > the ARCH_XYZ. Is this acceptable?  
> 
> That would do the right thing, but doesn't look ideal: The DW_APB timer
> on those platforms is fully functional, and a future Linux version or
> another OS might decide to use both timers for one reason or another.
> 
> I'd be happier with a solution that keeps the DT describing the hardware
> and not the way we expect Linux to use it, and instead has some heuristic
> in the selection of the delay timer. At the moment, we purely base this
> on the frequency, which as you say is suboptimal.
> 
> One possible way to improve this would be to add an optional 'latency'
> property to the DT nodes (or the driver), and use a combination of latency
> and resolution to make the decision.

Got it. Thanks for the suggestions. The 'latency' here seems a 'rating'
similar as the one in clocksource. I will cook a series for review:

patch 1 to make register_current_timer_delay() aware of 'rating'

patch 2 to set rating of arch timer as 400

patch 3 to add timer based delay support to dw_apb_timer whose rating is 300

Thanks a lot,
Jisheng

> A simpler way would be to always prefer the arch timer on ARM if that
> is present, even if another timer has a higher resolution. This should
> be only a few additional lines in register_current_timer_delay(), or
> possibly an additional function argument.
> 



--
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]

#1261313

On Tuesday 03 November 2015 14:59:40 Jisheng Zhang wrote:
> > On Monday 02 November 2015 11:03:34 Jisheng Zhang wrote:
> > > On Fri, 30 Oct 2015 13:42:01 +0100 Arnd Bergmann wrote:  
> > I'd be happier with a solution that keeps the DT describing the hardware
> > and not the way we expect Linux to use it, and instead has some heuristic
> > in the selection of the delay timer. At the moment, we purely base this
> > on the frequency, which as you say is suboptimal.
> > 
> > One possible way to improve this would be to add an optional 'latency'
> > property to the DT nodes (or the driver), and use a combination of latency
> > and resolution to make the decision.
> 
> Got it. Thanks for the suggestions. The 'latency' here seems a 'rating'
> similar as the one in clocksource. I will cook a series for review:
> 
> patch 1 to make register_current_timer_delay() aware of 'rating'
> 
> patch 2 to set rating of arch timer as 400
> 
> patch 3 to add timer based delay support to dw_apb_timer whose rating is 300

Ok. Just to make sure I got this right: your plan is to use the existing
'rating' setting as a primary indication, and fall back to comparing the
frequency if the rating is the same?

	Arnd
--
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]

#1261351

Dear Arnd

On Tue, 3 Nov 2015 09:49:32 +0100
Arnd Bergmann wrote:

> On Tuesday 03 November 2015 14:59:40 Jisheng Zhang wrote:
> > > On Monday 02 November 2015 11:03:34 Jisheng Zhang wrote:  
> > > > On Fri, 30 Oct 2015 13:42:01 +0100 Arnd Bergmann wrote:    
> > > I'd be happier with a solution that keeps the DT describing the hardware
> > > and not the way we expect Linux to use it, and instead has some heuristic
> > > in the selection of the delay timer. At the moment, we purely base this
> > > on the frequency, which as you say is suboptimal.
> > > 
> > > One possible way to improve this would be to add an optional 'latency'
> > > property to the DT nodes (or the driver), and use a combination of latency
> > > and resolution to make the decision.  
> > 
> > Got it. Thanks for the suggestions. The 'latency' here seems a 'rating'
> > similar as the one in clocksource. I will cook a series for review:
> > 
> > patch 1 to make register_current_timer_delay() aware of 'rating'
> > 
> > patch 2 to set rating of arch timer as 400
> > 
> > patch 3 to add timer based delay support to dw_apb_timer whose rating is 300  
> 
> Ok. Just to make sure I got this right: your plan is to use the existing
> 'rating' setting as a primary indication, and fall back to comparing the
> frequency if the rating is the same?

Yes, this is my plan.

Thanks,
Jisheng
--
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