ARM: at91: cpuidle: move driver to drivers/cpuidle
Subsystems (4)
cover ARM: at91: cpuidle: move driver to drivers/cpuidle
The number of cpuidle drivers is increasing more and more. Today we have in total 24 drivers. A lot of the code is duplicated, at least for the initialization. A work of consolidation has been done during this year:
* a lot of code cleanup in all the drivers
* time keeping is now part of the framework
* timer broadcast is now part of the framework
* a WFI state function for ARM is defined and used in the drivers
* an init function has been proposed to factor out the initialization across the drivers (patchset pending)
What has been observed is a lot of code duplicationis, modification made in the framework takes awhile before reaching the driver which uses an old API, duplicate routine and bugs, etc ...
It appears the drivers are belonging to different trees, the cpuidle framework is under linux-pm, the drivers are per SoC tree. The communication is made difficult because of the different mailing lists where the cpuidle are submitted.
After this work, it is time to prevent all these problems to occur again.
I propose to move the cpuidle drivers to the drivers/cpuidle directory, hence having one single submission path for cpuidle in order to have the cpuidle framework and the different drivers synced.
This series move the AT91 cpuidle driver under drivers/cpuidle. That does not change the rule to have the patches acked-by the author of the driver.
Note the calxeda and kirkwood drivers are now in drivers/cpuidle.
Daniel Lezcano (2):
ARM: at91: cpuidle: encapsulate the standby code
ARM: at91: cpuidle: move the driver to drivers/cpuidle directory
arch/arm/mach-at91/Makefile | 1 - arch/arm/mach-at91/cpuidle.c | 66 ------------------------------------------ arch/arm/mach-at91/pm.c | 8 ++++- drivers/cpuidle/Makefile | 1 + drivers/cpuidle/at91.c | 55 +++++++++++++++++++++++++++++++++++ 5 files changed, 63 insertions(+), 68 deletions(-) delete mode 100644 arch/arm/mach-at91/cpuidle.c create mode 100644 drivers/cpuidle/at91.c
Signature · 1 line
1.7.9.5
RFC PATCH 1/2 ARM: at91: cpuidle: encapsulate the standby code
In order to split the pm code from the cpuidle driver, add an ops for the standby function which will be initialized by the pm init functions directly, thus no need of the SoC specific headers.
Cleanup also the headers included in this file as they are no longer needed.
--- arch/arm/mach-at91/cpuidle.c | 19 ++++--------------- arch/arm/mach-at91/pm.c | 8 +++++++- 2 files changed, 11 insertions(+), 16 deletions(-)
arch/arm/mach-at91/cpuidle.c +4 -15
@@ -13,32 +13,21 @@* #2 wait-for-interrupt and RAM self refresh */ -#include <linux/kernel.h> +#include <linux/module.h> #include <linux/init.h> -#include <linux/platform_device.h> #include <linux/cpuidle.h> -#include <linux/io.h> -#include <linux/export.h> -#include <asm/proc-fns.h> #include <asm/cpuidle.h> -#include <mach/cpu.h> - -#include "pm.h" #define AT91_MAX_STATES 2 +extern void (*at91_standby_ops)(void ); +/* Actual code that puts the SoC in different idle states */ static int at91_enter_idle (struct cpuidle_device *dev,struct cpuidle_driver *drv,int index) { -if (cpu_is_at91rm9200 ()) -at91rm9200_standby (); -else if (cpu_is_at91sam9g45 ()) -at91sam9g45_standby (); -else -at91sam9_standby (); - +at91_standby_ops ();return index; }
arch/arm/mach-at91/pm.c +7 -1
@@ -39,6 +39,8 @@#include "at91_rstc.h" #include "at91_shdwc.h" +void (*at91_standby_ops)(void ); +static void __init show_reset_status (void ) {static char reset[]__initdata ="reset" ;
@@ -321,8 +323,12 @@ static int __init at91_pm_init(void)pr_info ("AT91: Power Management%s\n" , (slow_clock ?" (with slow clock mode)" :"" ));/* AT91RM9200 SDRAM low-power mode cannot be used with self-refresh. */ -if (cpu_is_at91rm9200 ()) +if (cpu_is_at91rm9200 ()) {at91_ramc_write (0 ,AT91RM9200_SDRAMC_LPR ,0 ); + at91_standby_ops = at91rm9200_standby; + }else if (cpu_is_at91sam9g45 ()) + at91_standby_ops= at91sam9g45_standby; +else at91_standby_ops = at91sam9_standby;suspend_set_ops (&at91_pm_ops);
Signature · 1 line
1.7.9.5
RFC PATCH 1/2 ARM: at91: cpuidle: encapsulate the standby code
Hi Daniel,
On 04/15/2013 03:29 PM, Daniel Lezcano :
Quoted 58 lines
In order to split the pm code from the cpuidle driver, add an ops for the standby function which will be initialized by the pm init functions directly, thus no need of the SoC specific headers.
Cleanup also the headers included in this file as they are no longer needed.
Signed-off-by Daniel Lezcano <[email protected]>--- arch/arm/mach-at91/cpuidle.c | 19 ++++--------------- arch/arm/mach-at91/pm.c | 8 +++++++- 2 files changed, 11 insertions(+), 16 deletions(-)arch/arm/mach-at91/cpuidle.c +4
-15@@ -13,32 +13,21 @@* #2 wait-for-interrupt and RAM self refresh */ -#include <linux/kernel.h> +#include <linux/module.h> #include <linux/init.h> -#include <linux/platform_device.h> #include <linux/cpuidle.h> -#include <linux/io.h> -#include <linux/export.h> -#include <asm/proc-fns.h> #include <asm/cpuidle.h> -#include <mach/cpu.h> - -#include "pm.h" #define AT91_MAX_STATES 2 +extern void (*at91_standby_ops)(void ); +/* Actual code that puts the SoC in different idle states */ static int at91_enter_idle (struct cpuidle_device *dev,struct cpuidle_driver *drv,int index) { -if (cpu_is_at91rm9200 ()) -at91rm9200_standby (); -else if (cpu_is_at91sam9g45 ()) -at91sam9g45_standby (); -else -at91sam9_standby (); - +at91_standby_ops ();return index; }arch/arm/mach-at91/pm.c +0
-0
arch/arm/mach-at91/pm.c line 42
Quoted 4 lines
arch/arm/mach-at91/pm.c +1
-0@@ -39,6 +39,8 @@#include "at91_rstc.h" #include "at91_shdwc.h" +void (*at91_standby_ops)(void );
Is this a common pattern to have such a floating function pointer in the pm code?
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 329
Quoted 13 lines
arch/arm/mach-at91/pm.c +4
-1+static void __init show_reset_status (void ) {static char reset[]__initdata ="reset" ;@@ -321,8 +323,12 @@ static int __init at91_pm_init(void)pr_info ("AT91: Power Management%s\n" , (slow_clock ?" (with slow clock mode)" :"" ));/* AT91RM9200 SDRAM low-power mode cannot be used with self-refresh. */ -if (cpu_is_at91rm9200 ()) +if (cpu_is_at91rm9200 ()) {at91_ramc_write (0 ,AT91RM9200_SDRAMC_LPR ,0 ); + at91_standby_ops = at91rm9200_standby; + }else if (cpu_is_at91sam9g45 ())
CodingStyle: ending "{" is missing.
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 330
Quoted 1 line
arch/arm/mach-at91/pm.c +1
-0+ at91_standby_ops= at91sam9g45_standby;
CodingStyle: " " is missing
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 331
Quoted 1 line
arch/arm/mach-at91/pm.c +1
-0+else at91_standby_ops = at91sam9_standby;
CodingStyle: not on the same line + "{}" missing
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 333
Quoted 2 lines
arch/arm/mach-at91/pm.c +0
-0suspend_set_ops (&at91_pm_ops);
I am in favor for the move. But please rewrite a new series.
Bye,
Signature · 1 line
Nicolas Ferre
RFC PATCH 1/2 ARM: at91: cpuidle: encapsulate the standby code
Quoted 64 lines · Nicolas Ferre
Hi Daniel,
On 04/15/2013 03:29 PM, Daniel Lezcano :
Quoted 61 lines
In order to split the pm code from the cpuidle driver, add an ops for the standby function which will be initialized by the pm init functions directly, thus no need of the SoC specific headers.
Cleanup also the headers included in this file as they are no longer needed.
Signed-off-by Daniel Lezcano <[email protected]>--- arch/arm/mach-at91/cpuidle.c | 19 ++++--------------- arch/arm/mach-at91/pm.c | 8 +++++++- 2 files changed, 11 insertions(+), 16 deletions(-)arch/arm/mach-at91/cpuidle.c +4
-15@@ -13,32 +13,21 @@* #2 wait-for-interrupt and RAM self refresh */ -#include <linux/kernel.h> +#include <linux/module.h> #include <linux/init.h> -#include <linux/platform_device.h> #include <linux/cpuidle.h> -#include <linux/io.h> -#include <linux/export.h> -#include <asm/proc-fns.h> #include <asm/cpuidle.h> -#include <mach/cpu.h> - -#include "pm.h" #define AT91_MAX_STATES 2 +extern void (*at91_standby_ops)(void ); +/* Actual code that puts the SoC in different idle states */ static int at91_enter_idle (struct cpuidle_device *dev,struct cpuidle_driver *drv,int index) { -if (cpu_is_at91rm9200 ()) -at91rm9200_standby (); -else if (cpu_is_at91sam9g45 ()) -at91sam9g45_standby (); -else -at91sam9_standby (); - +at91_standby_ops ();return index; }arch/arm/mach-at91/pm.c +0
-0@@ -39,6 +39,8 @@#include "at91_rstc.h" #include "at91_shdwc.h"
Quoted 4 lines
arch/arm/mach-at91/pm.c line 42
Quoted 1 line
arch/arm/mach-at91/pm.c +1
-0+void (*at91_standby_ops)(void );Is this a common pattern to have such a floating function pointer in the pm code?
Well, already seen but I agree it is not really nice.
The idea I had was to convert little by little all these pm functions into ops, then order, group and use them to initialize the cpuidle drivers through a single ARM driver. The correct order would have been to convert first these to ops then move the driver but there are so many drivers, so many code, I don't know where do I start.
Do you have a suggestion ?
Quoted 20 lines
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 329
Quoted 13 lines
arch/arm/mach-at91/pm.c +4
-1+static void __init show_reset_status (void ) {static char reset[]__initdata ="reset" ;@@ -321,8 +323,12 @@ static int __init at91_pm_init(void)pr_info ("AT91: Power Management%s\n" , (slow_clock ?" (with slow clock mode)" :"" ));/* AT91RM9200 SDRAM low-power mode cannot be used with self-refresh. */ -if (cpu_is_at91rm9200 ()) +if (cpu_is_at91rm9200 ()) {at91_ramc_write (0 ,AT91RM9200_SDRAMC_LPR ,0 ); + at91_standby_ops = at91rm9200_standby; + }else if (cpu_is_at91sam9g45 ())CodingStyle: ending "{" is missing.
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 330
Quoted 1 line
arch/arm/mach-at91/pm.c +1
-0+ at91_standby_ops= at91sam9g45_standby;CodingStyle: " " is missing
Quoted 8 lines
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 331
Quoted 1 line
arch/arm/mach-at91/pm.c +1
-0+else at91_standby_ops = at91sam9_standby;CodingStyle: not on the same line + "{}" missing
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 333
Quoted 2 lines
arch/arm/mach-at91/pm.c +0
-0suspend_set_ops (&at91_pm_ops);I am in favor for the move.
Ok, cool.
Quoted 1 line
But please rewrite a new series.
Yes, sure, that was a draft.
Thanks for the review.
-- Daniel
Signature · 5 lines
<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
RFC PATCH 1/2 ARM: at91: cpuidle: encapsulate the standby code
In reply to Daniel LezcanoQuoted 67 lines · Daniel Lezcano
In order to split the pm code from the cpuidle driver, add an ops for the standby function which will be initialized by the pm init functions directly, thus no need of the SoC specific headers.
Cleanup also the headers included in this file as they are no longer needed.
Signed-off-by Daniel Lezcano <[email protected]>--- arch/arm/mach-at91/cpuidle.c | 19 ++++--------------- arch/arm/mach-at91/pm.c | 8 +++++++- 2 files changed, 11 insertions(+), 16 deletions(-)arch/arm/mach-at91/cpuidle.c +4
-15@@ -13,32 +13,21 @@* #2 wait-for-interrupt and RAM self refresh */ -#include <linux/kernel.h> +#include <linux/module.h> #include <linux/init.h> -#include <linux/platform_device.h> #include <linux/cpuidle.h> -#include <linux/io.h> -#include <linux/export.h> -#include <asm/proc-fns.h> #include <asm/cpuidle.h> -#include <mach/cpu.h> - -#include "pm.h" #define AT91_MAX_STATES 2 +extern void (*at91_standby_ops)(void ); +/* Actual code that puts the SoC in different idle states */ static int at91_enter_idle (struct cpuidle_device *dev,struct cpuidle_driver *drv,int index) { -if (cpu_is_at91rm9200 ()) -at91rm9200_standby (); -else if (cpu_is_at91sam9g45 ()) -at91sam9g45_standby (); -else -at91sam9_standby (); - +at91_standby_ops ();return index; }arch/arm/mach-at91/pm.c +2
-0@@ -39,6 +39,8 @@#include "at91_rstc.h" #include "at91_shdwc.h" +void (*at91_standby_ops)(void ); +static void __init show_reset_status (void ) {static char reset[]__initdata ="reset" ;
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 330
Quoted 9 lines
arch/arm/mach-at91/pm.c +4
-1@@ -321,8 +323,12 @@ static int __init at91_pm_init(void)pr_info ("AT91: Power Management%s\n" , (slow_clock ?" (with slow clock mode)" :"" ));/* AT91RM9200 SDRAM low-power mode cannot be used with self-refresh. */ -if (cpu_is_at91rm9200 ()) +if (cpu_is_at91rm9200 ()) {at91_ramc_write (0 ,AT91RM9200_SDRAMC_LPR ,0 ); + at91_standby_ops = at91rm9200_standby; + }else if (cpu_is_at91sam9g45 ()) + at91_standby_ops= at91sam9g45_standby;
keep the {} for codying stye rules
arch/arm/mach-at91/pm.c static int __init at91_pm_init(void) · line 331
Quoted 1 line
arch/arm/mach-at91/pm.c +1
-0+else at91_standby_ops = at91sam9_standby;
I'm not a fan of global symbol but ok
Best Regards,
J.
RFC PATCH 2/2 ARM: at91: cpuidle: move the driver to drivers/cpuidle directory
In reply to Daniel LezcanoWe don't have any dependency with the SoC specific code.
Move the driver to the drivers/cpuidle directory.
Add Nicolas Ferre as author of the driver, so it will be in copy of the emails.
--- arch/arm/mach-at91/Makefile | 1 - arch/arm/mach-at91/cpuidle.c | 55 ---------------------------------------- drivers/cpuidle/Makefile | 1 + drivers/cpuidle/at91.c | 57 ++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 58 insertions(+), 56 deletions(-) delete mode 100644 arch/arm/mach-at91/cpuidle.c create mode 100644 drivers/cpuidle/at91.c
arch/arm/mach-at91/Makefile +0 -1
@@ -99,7 +99,6 @@ obj-y += leds.o # Power Management obj-$(CONFIG_PM) += pm.o obj-$(CONFIG_AT91_SLOW_CLOCK) += pm_slowclock.o -obj-$(CONFIG_CPU_IDLE) += cpuidle.o ifeq ($(CONFIG_PM_DEBUG),y) CFLAGS_pm.o += -DDEBUG
arch/arm/mach-at91/cpuidle.c deleted +0 -55
55 lines not on this page · raw message
drivers/cpuidle/Makefile +1 -0
@@ -7,3 +7,4 @@ obj-$(CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED) += coupled.o obj-$(CONFIG_CPU_IDLE_CALXEDA) += cpuidle-calxeda.o obj-$(CONFIG_ARCH_KIRKWOOD) += cpuidle-kirkwood.o +obj-$(CONFIG_ARCH_AT91) += at91.o
drivers/cpuidle/at91.c new file +57 -0
57 lines not on this page · raw message
Signature · 1 line
1.7.9.5
RFC PATCH 2/2 ARM: at91: cpuidle: move the driver to drivers/cpuidle directory
On 04/15/2013 03:29 PM, Daniel Lezcano :
Quoted 5 lines
We don't have any dependency with the SoC specific code.
Move the driver to the
drivers/cpuidledirectory.Add Nicolas Ferre as author of the driver, so it will be in copy of the emails.
Unfortunately, I am not the author of this driver. It is the work of Albin Tonnerre <[email protected]>. So, for sure, I (and Jean-Christohpe) can babysit this code but I cannot be awarded for it...
You can certainly add to the header of this file:
Maintained by the AT91 crew: Nicolas Ferre, Jean-Christophe Plagniol-Villard (together with email addresses).
Quoted 6 lines
Signed-off-by Daniel Lezcano <[email protected]>--- arch/arm/mach-at91/Makefile | 1 - arch/arm/mach-at91/cpuidle.c | 55 ---------------------------------------- drivers/cpuidle/Makefile | 1 + drivers/cpuidle/at91.c | 57 ++++++++++++++++++++++++++++++++++++++++++
No, sorry this file name is not appropriate: I prefer cpuidle-at91.c
Quoted 150 lines
4 files changed, 58 insertions(+), 56 deletions(-) delete mode 100644 arch/arm/mach-at91/cpuidle.c create mode 100644 drivers/cpuidle/at91.carch/arm/mach-at91/Makefile +0
-1@@ -99,7 +99,6 @@ obj-y += leds.o # Power Management obj-$(CONFIG_PM) += pm.o obj-$(CONFIG_AT91_SLOW_CLOCK) += pm_slowclock.o -obj-$(CONFIG_CPU_IDLE) += cpuidle.o ifeq ($(CONFIG_PM_DEBUG),y) CFLAGS_pm.o += -DDEBUGarch/arm/mach-at91/cpuidle.c deleted +0
-55@@ -1,55 +0,0 @@ -/* -* based on arch/arm/mach-kirkwood/cpuidle.c -* -* CPU idle support for AT91 SoC -* -* This file is licensed under the terms of the GNU General Public -* License version 2. This program is licensed "as is" without any -* warranty of any kind, whether express or implied. -* -* The cpu idle uses wait-for-interrupt and RAM self refresh in order -* to implement two idle states - -* #1 wait-for-interrupt -* #2 wait-for-interrupt and RAM self refresh -*/ - -#include <linux/module.h> -#include <linux/init.h> -#include <linux/cpuidle.h> -#include <asm/cpuidle.h> - -#define AT91_MAX_STATES 2 - -extern void (*at91_standby_ops)(void ); - -/* Actual code that puts the SoC in different idle states */ -static int at91_enter_idle (struct cpuidle_device *dev, -struct cpuidle_driver *drv, -int index) -{ -at91_standby_ops (); -return index; -} - -static struct cpuidle_driver at91_idle_driver = { - .name ="at91_idle" , - .owner =THIS_MODULE , - .states[0 ] =ARM_CPUIDLE_WFI_STATE , - .states[1 ] = { - .enter = at91_enter_idle, - .exit_latency =10 , - .target_residency =100000 , - .flags =CPUIDLE_FLAG_TIME_VALID , - .name ="RAM_SR" , - .desc ="WFI and DDR Self Refresh" , - }, - .state_count =AT91_MAX_STATES , -}; - -/* Initialize CPU idle by registering the idle states */ -static int __init at91_init_cpuidle (void ) -{ -return cpuidle_register (&at91_idle_driver,NULL ); -} - -device_initcall (at91_init_cpuidle);drivers/cpuidle/Makefile +1
-0@@ -7,3 +7,4 @@ obj-$(CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED) += coupled.o obj-$(CONFIG_CPU_IDLE_CALXEDA) += cpuidle-calxeda.o obj-$(CONFIG_ARCH_KIRKWOOD) += cpuidle-kirkwood.o +obj-$(CONFIG_ARCH_AT91) += at91.odrivers/cpuidle/at91.c new file +57
-0@@ -0,0 +1,57 @@ +/* +* based on arch/arm/mach-kirkwood/cpuidle.c +* +* CPU idle support for AT91 SoC +* +* Author: Nicolas Ferre <[email protected]> +* +* This file is licensed under the terms of the GNU General Public +* License version 2. This program is licensed "as is" without any +* warranty of any kind, whether express or implied. +* +* The cpu idle uses wait-for-interrupt and RAM self refresh in order +* to implement two idle states - +* #1 wait-for-interrupt +* #2 wait-for-interrupt and RAM self refresh +*/ + +#include <linux/module.h> +#include <linux/init.h> +#include <linux/cpuidle.h> +#include <asm/cpuidle.h> + +#define AT91_MAX_STATES 2 + +extern void (*at91_standby_ops)(void ); + +/* Actual code that puts the SoC in different idle states */ +static int at91_enter_idle (struct cpuidle_device *dev, +struct cpuidle_driver *drv, +int index) +{ +at91_standby_ops (); +return index; +} + +static struct cpuidle_driver at91_idle_driver = { + .name ="at91_idle" , + .owner =THIS_MODULE , + .states[0 ] =ARM_CPUIDLE_WFI_STATE , + .states[1 ] = { + .enter = at91_enter_idle, + .exit_latency =10 , + .target_residency =100000 , + .flags =CPUIDLE_FLAG_TIME_VALID , + .name ="RAM_SR" , + .desc ="WFI and DDR Self Refresh" , + }, + .state_count =AT91_MAX_STATES , +}; + +/* Initialize CPU idle by registering the idle states */ +static int __init at91_init_cpuidle (void ) +{ +return cpuidle_register (&at91_idle_driver,NULL ); +} + +device_initcall (at91_init_cpuidle);
Signature · 1 line
Nicolas Ferre
RFC PATCH 2/2 ARM: at91: cpuidle: move the driver to drivers/cpuidle directory
Quoted 5 lines · Nicolas Ferre
On 04/15/2013 03:29 PM, Daniel Lezcano :
Quoted 4 lines
We don't have any dependency with the SoC specific code.
Move the driver to the
drivers/cpuidledirectory.
Quoted 8 lines
Quoted 1 line
Add Nicolas Ferre as author of the driver, so it will be in copy of the emails.
Unfortunately, I am not the author of this driver. It is the work of Albin Tonnerre <[email protected]>. So, for sure, I (and Jean-Christohpe) can babysit this code but I cannot be awarded for it... You can certainly add to the header of this file:
Maintained by the AT91 crew: Nicolas Ferre, Jean-Christophe Plagniol-Villard (together with email addresses).
Ok, will do.
Quoted 8 lines
Quoted 6 lines
Signed-off-by Daniel Lezcano <[email protected]>--- arch/arm/mach-at91/Makefile | 1 - arch/arm/mach-at91/cpuidle.c | 55 ---------------------------------------- drivers/cpuidle/Makefile | 1 + drivers/cpuidle/at91.c | 57 ++++++++++++++++++++++++++++++++++++++++++No, sorry this file name is not appropriate: I prefer
cpuidle-at91.c
Sure ? :)
It is a bit redundant drivers/cpuidle/cpuidle-at91.c, no ? and if all the cpuidle drivers are there, the directory content will look:
cpuidle/cpuidle-tegra114.ccpuidle/cpuidle-tegra20.ccpuidle/cpuidle-tegra30.ccpuidle/cpuidle-imx5.ccpuidle/cpuidle-imx6.ccpuidle/cpuidle-kirkwood.ccpuidle/cpuidle-calxeda.ccpuidle/cpuidle-shmobile.ccpuidle/cpuidle-intel-idle.ccpuidle/cpuidle-at91.ccpuidle/cpuidle-omap34xx.ccpuidle/cpuidle-omap44xx.ccpuidle/cpuidle-ux500.ccpuidle/cpuidle-exynos4.ccpuidle/cpuidle-acpi.ccpuidle/cpuidle-pseries.ccpuidle/cpuidle-sh.ccpuidle/cpuidle-s3c64xx.ccpuidle/cpuidle-davinci.c
etc ...
But I won't argue if you really want cpuidle-at91.c, it is a detail IMO.
Thanks
-- Daniel
Signature · 5 lines
<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
RFC PATCH 2/2 ARM: at91: cpuidle: move the driver to drivers/cpuidle directory
In reply to Daniel LezcanoQuoted 7 lines · Daniel Lezcano
We don't have any dependency with the SoC specific code.
Move the driver to the
drivers/cpuidledirectory.Add Nicolas Ferre as author of the driver, so it will be in copy of the emails.
Signed-off-by Daniel Lezcano <[email protected]>
please use -M when generating the patch
Quoted 112 lines
--- arch/arm/mach-at91/Makefile | 1 - arch/arm/mach-at91/cpuidle.c | 55 ---------------------------------------- drivers/cpuidle/Makefile | 1 + drivers/cpuidle/at91.c | 57 ++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 58 insertions(+), 56 deletions(-) delete mode 100644 arch/arm/mach-at91/cpuidle.c create mode 100644 drivers/cpuidle/at91.carch/arm/mach-at91/Makefile +0
-1@@ -99,7 +99,6 @@ obj-y += leds.o # Power Management obj-$(CONFIG_PM) += pm.o obj-$(CONFIG_AT91_SLOW_CLOCK) += pm_slowclock.o -obj-$(CONFIG_CPU_IDLE) += cpuidle.o ifeq ($(CONFIG_PM_DEBUG),y) CFLAGS_pm.o += -DDEBUGarch/arm/mach-at91/cpuidle.c deleted +0
-55@@ -1,55 +0,0 @@ -/* -* based on arch/arm/mach-kirkwood/cpuidle.c -* -* CPU idle support for AT91 SoC -* -* This file is licensed under the terms of the GNU General Public -* License version 2. This program is licensed "as is" without any -* warranty of any kind, whether express or implied. -* -* The cpu idle uses wait-for-interrupt and RAM self refresh in order -* to implement two idle states - -* #1 wait-for-interrupt -* #2 wait-for-interrupt and RAM self refresh -*/ - -#include <linux/module.h> -#include <linux/init.h> -#include <linux/cpuidle.h> -#include <asm/cpuidle.h> - -#define AT91_MAX_STATES 2 - -extern void (*at91_standby_ops)(void ); - -/* Actual code that puts the SoC in different idle states */ -static int at91_enter_idle (struct cpuidle_device *dev, -struct cpuidle_driver *drv, -int index) -{ -at91_standby_ops (); -return index; -} - -static struct cpuidle_driver at91_idle_driver = { - .name ="at91_idle" , - .owner =THIS_MODULE , - .states[0 ] =ARM_CPUIDLE_WFI_STATE , - .states[1 ] = { - .enter = at91_enter_idle, - .exit_latency =10 , - .target_residency =100000 , - .flags =CPUIDLE_FLAG_TIME_VALID , - .name ="RAM_SR" , - .desc ="WFI and DDR Self Refresh" , - }, - .state_count =AT91_MAX_STATES , -}; - -/* Initialize CPU idle by registering the idle states */ -static int __init at91_init_cpuidle (void ) -{ -return cpuidle_register (&at91_idle_driver,NULL ); -} - -device_initcall (at91_init_cpuidle);drivers/cpuidle/Makefile +1
-0@@ -7,3 +7,4 @@ obj-$(CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED) += coupled.o obj-$(CONFIG_CPU_IDLE_CALXEDA) += cpuidle-calxeda.o obj-$(CONFIG_ARCH_KIRKWOOD) += cpuidle-kirkwood.o +obj-$(CONFIG_ARCH_AT91) += at91.odrivers/cpuidle/at91.c new file +15
-0@@ -0,0 +1,57 @@ +/* +* based on arch/arm/mach-kirkwood/cpuidle.c +* +* CPU idle support for AT91 SoC +* +* Author: Nicolas Ferre <[email protected]> +* +* This file is licensed under the terms of the GNU General Public +* License version 2. This program is licensed "as is" without any +* warranty of any kind, whether express or implied. +* +* The cpu idle uses wait-for-interrupt and RAM self refresh in order +* to implement two idle states - +* #1 wait-for-interrupt +* #2 wait-for-interrupt and RAM self refresh
drivers/cpuidle/at91.c line 25
Quoted 10 lines
drivers/cpuidle/at91.c new file +10
-0+*/ + +#include <linux/module.h> +#include <linux/init.h> +#include <linux/cpuidle.h> +#include <asm/cpuidle.h> + +#define AT91_MAX_STATES 2 + +extern void (*at91_standby_ops)(void );
really don't like can we pass it via the pm?
Quoted 35 lines
drivers/cpuidle/at91.c new file +32
-0+ +/* Actual code that puts the SoC in different idle states */ +static int at91_enter_idle (struct cpuidle_device *dev, +struct cpuidle_driver *drv, +int index) +{ +at91_standby_ops (); +return index; +} + +static struct cpuidle_driver at91_idle_driver = { + .name ="at91_idle" , + .owner =THIS_MODULE , + .states[0 ] =ARM_CPUIDLE_WFI_STATE , + .states[1 ] = { + .enter = at91_enter_idle, + .exit_latency =10 , + .target_residency =100000 , + .flags =CPUIDLE_FLAG_TIME_VALID , + .name ="RAM_SR" , + .desc ="WFI and DDR Self Refresh" , + }, + .state_count =AT91_MAX_STATES , +}; + +/* Initialize CPU idle by registering the idle states */ +static int __init at91_init_cpuidle (void ) +{ +return cpuidle_register (&at91_idle_driver,NULL ); +} + +device_initcall (at91_init_cpuidle);Signature · 1 line
1.7.9.5
RFC PATCH 2/2 ARM: at91: cpuidle: move the driver to drivers/cpuidle directory
Quoted 9 lines · Jean-Christophe PLAGNIOL-VILLARD
Quoted 7 lines · Daniel Lezcano
We don't have any dependency with the SoC specific code.
Move the driver to the
drivers/cpuidledirectory.Add Nicolas Ferre as author of the driver, so it will be in copy of the emails.
Signed-off-by Daniel Lezcano <[email protected]>please use -M when generating the patch
Oh, right. Thanks for reminding me the option.
[ ... ]
Quoted 9 lines
drivers/cpuidle/at91.c line 25
Quoted 8 lines
drivers/cpuidle/at91.c new file +8
-0+#include <linux/module.h> +#include <linux/init.h> +#include <linux/cpuidle.h> +#include <asm/cpuidle.h> + +#define AT91_MAX_STATES 2 + +extern void (*at91_standby_ops)(void );really don't like can we pass it via the pm?
I agree, it is hackish. Can you elaborate when you say "pass it via the pm" ?
Signature · 5 lines
<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
RFC PATCH 2/2 ARM: at91: cpuidle: move the driver to drivers/cpuidle directory
Quoted 25 lines · Daniel Lezcano
Quoted 9 lines · Jean-Christophe PLAGNIOL-VILLARD
Quoted 7 lines · Daniel Lezcano
We don't have any dependency with the SoC specific code.
Move the driver to the
drivers/cpuidledirectory.Add Nicolas Ferre as author of the driver, so it will be in copy of the emails.
Signed-off-by Daniel Lezcano <[email protected]>please use -M when generating the patch
Oh, right. Thanks for reminding me the option.
[ ... ]
Quoted 9 lines
drivers/cpuidle/at91.c line 25
Quoted 8 lines
drivers/cpuidle/at91.c new file +8
-0+#include <linux/module.h> +#include <linux/init.h> +#include <linux/cpuidle.h> +#include <asm/cpuidle.h> + +#define AT91_MAX_STATES 2 + +extern void (*at91_standby_ops)(void );really don't like can we pass it via the pm?
Quoted 2 lines
I agree, it is hackish. Can you elaborate when you say "pass it via the pm" ?
I was thinking about it
I would even prefer to have a platfrom driver that pass it via platform data even in DT
but not something that can touch globally
and for drivers I prefer to do not use cpu_is only for the core
Best Regards,
J.
Quoted 10 lines
Signature · 6 lines
-- <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
RFC PATCH 2/2 ARM: at91: cpuidle: move the driver to drivers/cpuidle directory
Quoted 29 lines · Jean-Christophe PLAGNIOL-VILLARD
Quoted 25 lines · Daniel Lezcano
Quoted 9 lines · Jean-Christophe PLAGNIOL-VILLARD
Quoted 7 lines · Daniel Lezcano
We don't have any dependency with the SoC specific code.
Move the driver to the
drivers/cpuidledirectory.Add Nicolas Ferre as author of the driver, so it will be in copy of the emails.
Signed-off-by Daniel Lezcano <[email protected]>please use -M when generating the patch
Oh, right. Thanks for reminding me the option.
[ ... ]
Quoted 9 lines
drivers/cpuidle/at91.c line 25
Quoted 8 lines
drivers/cpuidle/at91.c new file +8
-0+#include <linux/module.h> +#include <linux/init.h> +#include <linux/cpuidle.h> +#include <asm/cpuidle.h> + +#define AT91_MAX_STATES 2 + +extern void (*at91_standby_ops)(void );really don't like can we pass it via the pm?
Quoted 2 lines
I agree, it is hackish. Can you elaborate when you say "pass it via the pm" ?
Quoted 8 lines
I was thinking about it
I would even prefer to have a platfrom driver that pass it via platform data even in DT
but not something that can touch globally
and for drivers I prefer to do not use
cpu_isonly for the core
Is the following code what you had in mind ?
Please, be lenient, it is the first time I write something based on 'platform' :)
Index: cpuidle-next/arch/arm/mach-at91/cpuidle.c
===================================================================
--- cpuidle-next.orig/arch/arm/mach-at91/cpuidle.c 2013-04-16
10:40:21.058315042 +0200
+++ cpuidle-next/arch/arm/mach-at91/cpuidle.c 2013-04-17
00:35:42.836314931 +0200
@@ -13,32 +13,22 @@
* #2 wait-for-interrupt and RAM self refresh
*/
-#include <linux/kernel.h>
+#include <linux/module.h>
#include <linux/init.h>
-#include <linux/platform_device.h>
#include <linux/cpuidle.h>
-#include <linux/io.h>
-#include <linux/export.h>
-#include <asm/proc-fns.h>
+#include <linux/platform_device.h>
#include <asm/cpuidle.h>
-#include <mach/cpu.h>
-
-#include "pm.h"
#define AT91_MAX_STATES 2
+static void (*at91_standby_ops)(void);
+
/* Actual code that puts the SoC in different idle states */ static int at91_enter_idle(struct cpuidle_device *dev,
struct cpuidle_driver *drv,
int index)
{
- if (cpu_is_at91rm9200())
- at91rm9200_standby();
- else if (cpu_is_at91sam9g45())
- at91sam9g45_standby();
- else
- at91sam9_standby();
-
+ at91_standby_ops();
return index;}
@@ -58,9 +48,16 @@ static struct cpuidle_driver at91_idle_d
};
/* Initialize CPU idle by registering the idle states */
-static int __init at91_init_cpuidle(void)
+int __init at91_init_cpuidle(struct platform_device *pdev)
{
- return cpuidle_register(&at91_idle_driver, NULL); + at91_standby_ops = pdata; + return 0;
}
-device_initcall(at91_init_cpuidle);
+static int __init at91_cpuidle_init(void)
+{
+
+
+ return cpuidle_register(&at91_idle_driver, NULL);
+}
+device_initcall(at91_cpuidle_init);
Index: cpuidle-next/arch/arm/mach-at91/pm.c
===================================================================
--- cpuidle-next.orig/arch/arm/mach-at91/pm.c 2013-04-16
09:52:33.048426618 +0200
+++ cpuidle-next/arch/arm/mach-at91/pm.c 2013-04-16
18:22:03.207420375 +0200
@@ -312,21 +312,40 @@ static const struct platform_suspend_ops
.end = at91_pm_end,
};
-static int __init at91_pm_init(void)
+static int __init at91_pm_probe(struct platform_device *pdev)
{
-#ifdef CONFIG_AT91_SLOW_CLOCK
- slow_clock = (void *) (AT91_IO_VIRT_BASE - at91_slow_clock_sz);
-#endif
-
- pr_info("AT91: Power Management%s\n", (slow_clock ? " (with slow
clock mode)" : ""));
-
/* AT91RM9200 SDRAM low-power mode cannot be used with
self-refresh. */
- if (cpu_is_at91rm9200())
+ if (of_machine_is_compatible("atmel,at91rm9200")) {
at91_ramc_write(0, AT91RM9200_SDRAMC_LPR, 0);+ pdev->dev.platform_data = at91rm9200_standby;
+ } else if (of_machine_is_compatible("atmel,atmel,at91sam9g45")) {
+ pdev->dev.platform_data = at91sam9g45_standby;
+ } else {
+ pdev->dev.platform_data = at91sam9_standby;
+ }
suspend_set_ops(&at91_pm_ops);
show_reset_status();
+
return 0; }
+
+static struct platform_driver at91_pm_driver = {
+ .driver = {
+ .name = "pm-at91", + .owner = THIS_MODULE,
+ },
+};
+
+static int __init at91_pm_init(void)
+{
+#ifdef CONFIG_AT91_SLOW_CLOCK
+ slow_clock = (void *) (AT91_IO_VIRT_BASE - at91_slow_clock_sz);
+#endif
+ pr_info("AT91: Power Management%s\n",
+ (slow_clock ? " (with slow clock mode)" : ""));
+
+ return platform_driver_probe(&at91_pm_driver, at91_pm_probe);
+}
arch_initcall(at91_pm_init);
Index: cpuidle-next/arch/arm/mach-at91/board-usb-a926x.c
===================================================================
--- cpuidle-next.orig/arch/arm/mach-at91/board-usb-a926x.c
2013-04-15 10:31:45.503051053 +0200
+++ cpuidle-next/arch/arm/mach-at91/board-usb-a926x.c 2013-04-17
00:35:38.480314758 +0200
@@ -49,7 +49,6 @@
#include "sam9_smc.h"
#include "generic.h"
-
static void __init ek_init_early(void)
{
/* Initialize processor: 12.00 MHz crystal */
@@ -319,6 +318,19 @@ static void __init ek_add_device_leds(vo
at91_gpio_leds(ek_leds, ARRAY_SIZE(ek_leds));}
+#ifdef CONFIG_PM
+static struct platform_device at91_pm_device = {
+ .name = "pm-at91",
+ .id = -1,
+};
+#endif
+
+static struct platform_device *platform_devices[] __initdata = { +#ifdef CONFIG_PM
+ &at91_pm_device,
+#endif
+}
+
static void __init ek_board_init(void)
{
/* Serial */
@@ -351,6 +363,9 @@ static void __init ek_board_init(void)
| AT91_SHDW_WKMODE0_LOW
| AT91_SHDW_RTTWKEN);
}+
+ /* Platform devices */
+ platform_add_devices(&platform_devices,ARRAY_SIZE(platform_devices));
}
MACHINE_START(USB_A9263, "CALAO USB_A9263")
if (of_machine_is_compatible("atmel,at91rm9200")) {
printk("at91rm9200\n");
} else if (of_machine_is_compatible("atmel,atmel,at91sam9g45")) {
printk(at91sam9g45\n");
} else {
printk(other at91 cpu\n");
}Quoted 13 lines
Best Regards,
J.Quoted 10 lines
Signature · 6 lines
-- <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
Signature · 5 lines
<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
RFC PATCH 2/2 ARM: at91: cpuidle: move the driver to drivers/cpuidle directory
Quoted 32 lines · Daniel Lezcano
Quoted 31 lines · Jean-Christophe PLAGNIOL-VILLARD
Quoted 25 lines · Daniel Lezcano
Quoted 9 lines · Jean-Christophe PLAGNIOL-VILLARD
Quoted 7 lines · Daniel Lezcano
We don't have any dependency with the SoC specific code.
Move the driver to the
drivers/cpuidledirectory.Add Nicolas Ferre as author of the driver, so it will be in copy of the emails.
Signed-off-by Daniel Lezcano <[email protected]>please use -M when generating the patch
Oh, right. Thanks for reminding me the option.
[ ... ]
Quoted 9 lines
drivers/cpuidle/at91.c line 25
Quoted 8 lines
drivers/cpuidle/at91.c new file +8
-0+#include <linux/module.h> +#include <linux/init.h> +#include <linux/cpuidle.h> +#include <asm/cpuidle.h> + +#define AT91_MAX_STATES 2 + +extern void (*at91_standby_ops)(void );really don't like can we pass it via the pm?
Quoted 2 lines
I agree, it is hackish. Can you elaborate when you say "pass it via the pm" ?
I was thinking about it
Quoted 8 lines
Quoted 6 lines
I would even prefer to have a platfrom driver that pass it via platform data even in DT
but not something that can touch globally
and for drivers I prefer to do not use
cpu_isonly for the coreIs the following code what you had in mind ?
Hi Jean-Christophe,
coming back to move the cpuidle driver to drivers/cpuidle.
I was wondering if the code below would be ok for you ?
Thanks
-- Daniel
Quoted 195 lines
Please, be lenient, it is the first time I write something based on 'platform' :)
Index:
cpuidle-next/arch/arm/mach-at91/cpuidle.c
===================================================================
---cpuidle-next.orig/arch/arm/mach-at91/cpuidle.c2013-04-16
10:40:21.058315042 +0200
+++cpuidle-next/arch/arm/mach-at91/cpuidle.c2013-04-17
00:35:42.836314931 +0200(quoted diff) +15
-18@@ -13,32 +13,22 @@ * #2 wait-for-interrupt and RAM self refresh */ -#include <linux/kernel.h> +#include <linux/module.h> #include <linux/init.h> -#include <linux/platform_device.h> #include <linux/cpuidle.h> -#include <linux/io.h> -#include <linux/export.h> -#include <asm/proc-fns.h> +#include <linux/platform_device.h> #include <asm/cpuidle.h> -#include <mach/cpu.h> - -#include "pm.h" #define AT91_MAX_STATES 2 +static void (*at91_standby_ops)(void); + /* Actual code that puts the SoC in different idle states */ static int at91_enter_idle(struct cpuidle_device *dev, struct cpuidle_driver *drv, int index) { - if (cpu_is_at91rm9200()) - at91rm9200_standby(); - else if (cpu_is_at91sam9g45()) - at91sam9g45_standby(); - else - at91sam9_standby(); - + at91_standby_ops(); return index; }@@ -58,9 +48,16 @@ static struct cpuidle_driver at91_idle_d }; /* Initialize CPU idle by registering the idle states */ -static int __init at91_init_cpuidle(void) +int __init at91_init_cpuidle(struct platform_device *pdev) { - return cpuidle_register(&at91_idle_driver, NULL); + at91_standby_ops = pdata; + return 0; } -device_initcall(at91_init_cpuidle); +static int __init at91_cpuidle_init(void) +{ + + + return cpuidle_register(&at91_idle_driver, NULL); +} +device_initcall(at91_cpuidle_init);Index:
cpuidle-next/arch/arm/mach-at91/pm.c
===================================================================
---cpuidle-next.orig/arch/arm/mach-at91/pm.c2013-04-16
09:52:33.048426618 +0200
+++cpuidle-next/arch/arm/mach-at91/pm.c2013-04-16
18:22:03.207420375 +0200(quoted diff) +1
-7@@ -312,21 +312,40 @@ static const struct platform_suspend_ops .end = at91_pm_end, }; -static int __init at91_pm_init(void) +static int __init at91_pm_probe(struct platform_device *pdev) { -#ifdef CONFIG_AT91_SLOW_CLOCK - slow_clock = (void *) (AT91_IO_VIRT_BASE - at91_slow_clock_sz); -#endif - - pr_info("AT91: Power Management%s\n", (slow_clock ? " (with slow clock mode)" : "")); - /* AT91RM9200 SDRAM low-power mode cannot be used withself-refresh. */
(quoted diff) +26
-1- if (cpu_is_at91rm9200()) + if (of_machine_is_compatible("atmel,at91rm9200")) { at91_ramc_write(0, AT91RM9200_SDRAMC_LPR, 0); + pdev->dev.platform_data = at91rm9200_standby; + } else if (of_machine_is_compatible("atmel,atmel,at91sam9g45")) { + pdev->dev.platform_data = at91sam9g45_standby; + } else { + pdev->dev.platform_data = at91sam9_standby; + } suspend_set_ops(&at91_pm_ops); show_reset_status(); + return 0; } + +static struct platform_driver at91_pm_driver = { + .driver = { + .name = "pm-at91", + .owner = THIS_MODULE, + }, +}; + +static int __init at91_pm_init(void) +{ +#ifdef CONFIG_AT91_SLOW_CLOCK + slow_clock = (void *) (AT91_IO_VIRT_BASE - at91_slow_clock_sz); +#endif + pr_info("AT91: Power Management%s\n", + (slow_clock ? " (with slow clock mode)" : "")); + + return platform_driver_probe(&at91_pm_driver, at91_pm_probe); +} arch_initcall(at91_pm_init);Index:
cpuidle-next/arch/arm/mach-at91/board-usb-a926x.c
===================================================================
---cpuidle-next.orig/arch/arm/mach-at91/board-usb-a926x.c2013-04-15 10:31:45.503051053 +0200+++
cpuidle-next/arch/arm/mach-at91/board-usb-a926x.c2013-04-17
00:35:38.480314758 +0200(quoted diff) +16
-1@@ -49,7 +49,6 @@ #include "sam9_smc.h" #include "generic.h" - static void __init ek_init_early(void) { /* Initialize processor: 12.00 MHz crystal */@@ -319,6 +318,19 @@ static void __init ek_add_device_leds(vo at91_gpio_leds(ek_leds, ARRAY_SIZE(ek_leds)); } +#ifdef CONFIG_PM +static struct platform_device at91_pm_device = { + .name = "pm-at91", + .id = -1, +}; +#endif + +static struct platform_device *platform_devices[] __initdata = { +#ifdef CONFIG_PM + &at91_pm_device, +#endif +} + static void __init ek_board_init(void) { /* Serial */@@ -351,6 +363,9 @@ static void __init ek_board_init(void) | AT91_SHDW_WKMODE0_LOW | AT91_SHDW_RTTWKEN); } + + /* Platform devices */ + platform_add_devices(&platform_devices, ARRAY_SIZE(platform_devices)); }
MACHINE_START(USB_A9263, "CALAO USB_A9263")if (of_machine_is_compatible("atmel,at91rm9200")) { printk("at91rm9200\n"); } else if (of_machine_is_compatible("atmel,atmel,at91sam9g45")) { printk(at91sam9g45\n"); } else { printk(other at91 cpu\n"); }Quoted 13 lines
Best Regards,
J.Quoted 10 lines
Signature · 6 lines
-- <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
Signature · 5 lines
<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