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


Groups > linux.kernel > #1441408 > unrolled thread

[PATCH RT 11/20] ARM: imx: always use TWD on IMX6Q

Started bySteven Rostedt <rostedt@goodmis.org>
First post2016-07-12 16:30 +0200
Last post2016-07-18 20:20 +0200
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

  [PATCH RT 11/20] ARM: imx: always use TWD on IMX6Q Steven Rostedt <rostedt@goodmis.org> - 2016-07-12 16:30 +0200
    Re: [PATCH RT 11/20] ARM: imx: always use TWD on IMX6Q Sebastian Andrzej Siewior <bigeasy@linutronix.de> - 2016-07-12 22:10 +0200
      Re: [PATCH RT 11/20] ARM: imx: always use TWD on IMX6Q Steven Rostedt <rostedt@goodmis.org> - 2016-07-12 22:30 +0200
      Re: [PATCH RT 11/20] ARM: imx: always use TWD on IMX6Q Grygorii Strashko <grygorii.strashko@ti.com> - 2016-07-18 20:20 +0200

#1441408 — [PATCH RT 11/20] ARM: imx: always use TWD on IMX6Q

FromSteven Rostedt <rostedt@goodmis.org>
Date2016-07-12 16:30 +0200
Subject[PATCH RT 11/20] ARM: imx: always use TWD on IMX6Q
Message-ID<rU6Rc-66R-43@gated-at.bofh.it>
4.1.27-rt31-rc1 stable review patch.
If anyone has any objections, please let me know.

------------------

From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>

There is no reason to limit the TWD to be used on SMP kernels only if the
hardware has it available.
On Wandboard i.MX6SOLO, running PREEMPT-RT and cyclictest I see as max
immediately after start in idle:
UP : ~90us
SMP: ~50us
UP + TWD: ~20us.
Based on this numbers I prefer the TWD over the slightly slower MXC
timer.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
 arch/arm/mach-imx/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/mach-imx/Kconfig b/arch/arm/mach-imx/Kconfig
index 3a3d3e9d7bfd..92c0cf1276fe 100644
--- a/arch/arm/mach-imx/Kconfig
+++ b/arch/arm/mach-imx/Kconfig
@@ -558,7 +558,7 @@ config SOC_IMX6Q
 	bool "i.MX6 Quad/DualLite support"
 	select ARM_ERRATA_764369 if SMP
 	select HAVE_ARM_SCU if SMP
-	select HAVE_ARM_TWD if SMP
+	select HAVE_ARM_TWD
 	select PCI_DOMAINS if PCI
 	select PINCTRL_IMX6Q
 	select SOC_IMX6
-- 
2.8.1

[toc] | [next] | [standalone]


#1441701

FromSebastian Andrzej Siewior <bigeasy@linutronix.de>
Date2016-07-12 22:10 +0200
Message-ID<rUcae-1jn-25@gated-at.bofh.it>
In reply to#1441408
* Steven Rostedt | 2016-07-12 10:24:16 [-0400]:

>4.1.27-rt31-rc1 stable review patch.
>If anyone has any objections, please let me know.

I don't recommend to cherry-pick this. The symbol it selects
(HAVE_ARM_TWD) has "depends on SMP" in v4.1. In v4.4 there should be
AM43xx which is UP and also uses the TWD so the change is simple here.

Sebastian

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


#1441723

FromSteven Rostedt <rostedt@goodmis.org>
Date2016-07-12 22:30 +0200
Message-ID<rUctA-1qV-13@gated-at.bofh.it>
In reply to#1441701
On Tue, 12 Jul 2016 22:04:29 +0200
Sebastian Andrzej Siewior <bigeasy@linutronix.de> wrote:

> * Steven Rostedt | 2016-07-12 10:24:16 [-0400]:
> 
> >4.1.27-rt31-rc1 stable review patch.
> >If anyone has any objections, please let me know.  
> 
> I don't recommend to cherry-pick this. The symbol it selects
> (HAVE_ARM_TWD) has "depends on SMP" in v4.1. In v4.4 there should be
> AM43xx which is UP and also uses the TWD so the change is simple here.

So remove it from 4.1 and older?

I can do that.

-- Steve

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


#1445689

FromGrygorii Strashko <grygorii.strashko@ti.com>
Date2016-07-18 20:20 +0200
Message-ID<rWlj4-1gf-57@gated-at.bofh.it>
In reply to#1441701
On 07/12/2016 11:04 PM, Sebastian Andrzej Siewior wrote:
> * Steven Rostedt | 2016-07-12 10:24:16 [-0400]:
> 
>> 4.1.27-rt31-rc1 stable review patch.
>> If anyone has any objections, please let me know.
> 
> I don't recommend to cherry-pick this. The symbol it selects
> (HAVE_ARM_TWD) has "depends on SMP" in v4.1. In v4.4 there should be
> AM43xx which is UP and also uses the TWD so the change is simple here.
> 

yes. AM43xx is UP and supports ARM TWD, but there could be some 'issues' in !RT case :(

ARM TWD might be powered off in deep CPU Idle states and lose its context if !RT & CPUIDLE,
so there are should be backup broadcast timer configured. And that can cause build issues
related to tick_broadcast()->smp_cross_call().

I've solved it locally for OMAP2 only: commit 0b3e6fca4d1af4aa150d32506220f4241323a00c

^Just FYI. But may be it can be done in more generic way.

regards,
-grygorii

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web