* [PATCH] pinctrl: at91-pio4: handle suspend to ram
@ 2017-04-06 13:36 Alexandre Belloni
2017-04-07 10:16 ` Linus Walleij
2017-04-07 12:45 ` Ludovic Desroches
0 siblings, 2 replies; 5+ messages in thread
From: Alexandre Belloni @ 2017-04-06 13:36 UTC (permalink / raw)
To: Linus Walleij, Ludovic Desroches
Cc: linux-arm-kernel, linux-gpio, linux-kernel, Alexandre Belloni
When suspending to RAM, the power to the core is cut and the register
values are lost. Save and restore more registers than just IMR.
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
---
drivers/pinctrl/pinctrl-at91-pio4.c | 34 ++++++++++++++++++++++++++++------
1 file changed, 28 insertions(+), 6 deletions(-)
diff --git a/drivers/pinctrl/pinctrl-at91-pio4.c b/drivers/pinctrl/pinctrl-at91-pio4.c
index 28bbc1bb9e6c..dc8591543dee 100644
--- a/drivers/pinctrl/pinctrl-at91-pio4.c
+++ b/drivers/pinctrl/pinctrl-at91-pio4.c
@@ -126,7 +126,11 @@ struct atmel_pioctrl {
struct irq_domain *irq_domain;
int *irqs;
unsigned *pm_wakeup_sources;
- unsigned *pm_suspend_backup;
+ struct {
+ u32 imr;
+ u32 odsr;
+ u32 cfgr[ATMEL_PIO_NPINS_PER_BANK];
+ } *pm_suspend_backup;
struct device *dev;
struct device_node *node;
};
@@ -830,17 +834,26 @@ static int __maybe_unused atmel_pctrl_suspend(struct device *dev)
{
struct platform_device *pdev = to_platform_device(dev);
struct atmel_pioctrl *atmel_pioctrl = platform_get_drvdata(pdev);
- int i;
+ int i, j;
/*
* For each bank, save IMR to restore it later and disable all GPIO
* interrupts excepting the ones marked as wakeup sources.
*/
for (i = 0; i < atmel_pioctrl->nbanks; i++) {
- atmel_pioctrl->pm_suspend_backup[i] =
+ atmel_pioctrl->pm_suspend_backup[i].imr =
atmel_gpio_read(atmel_pioctrl, i, ATMEL_PIO_IMR);
atmel_gpio_write(atmel_pioctrl, i, ATMEL_PIO_IDR,
~atmel_pioctrl->pm_wakeup_sources[i]);
+ atmel_pioctrl->pm_suspend_backup[i].odsr =
+ atmel_gpio_read(atmel_pioctrl, i, ATMEL_PIO_ODSR);
+ for (j = 0; j < ATMEL_PIO_NPINS_PER_BANK; j++) {
+ atmel_gpio_write(atmel_pioctrl, i,
+ ATMEL_PIO_MSKR, BIT(j));
+ atmel_pioctrl->pm_suspend_backup[i].cfgr[j] =
+ atmel_gpio_read(atmel_pioctrl, i,
+ ATMEL_PIO_CFGR);
+ }
}
return 0;
@@ -850,11 +863,20 @@ static int __maybe_unused atmel_pctrl_resume(struct device *dev)
{
struct platform_device *pdev = to_platform_device(dev);
struct atmel_pioctrl *atmel_pioctrl = platform_get_drvdata(pdev);
- int i;
+ int i, j;
- for (i = 0; i < atmel_pioctrl->nbanks; i++)
+ for (i = 0; i < atmel_pioctrl->nbanks; i++) {
atmel_gpio_write(atmel_pioctrl, i, ATMEL_PIO_IER,
- atmel_pioctrl->pm_suspend_backup[i]);
+ atmel_pioctrl->pm_suspend_backup[i].imr);
+ atmel_gpio_write(atmel_pioctrl, i, ATMEL_PIO_SODR,
+ atmel_pioctrl->pm_suspend_backup[i].odsr);
+ for (j = 0; j < ATMEL_PIO_NPINS_PER_BANK; j++) {
+ atmel_gpio_write(atmel_pioctrl, i,
+ ATMEL_PIO_MSKR, BIT(j));
+ atmel_gpio_write(atmel_pioctrl, i, ATMEL_PIO_CFGR,
+ atmel_pioctrl->pm_suspend_backup[i].cfgr[j]);
+ }
+ }
return 0;
}
--
2.11.0
^ permalink raw reply related [flat|nested] 5+ messages in thread* Re: [PATCH] pinctrl: at91-pio4: handle suspend to ram
2017-04-06 13:36 [PATCH] pinctrl: at91-pio4: handle suspend to ram Alexandre Belloni
@ 2017-04-07 10:16 ` Linus Walleij
2017-04-07 10:48 ` Alexandre Belloni
2017-04-07 12:45 ` Ludovic Desroches
1 sibling, 1 reply; 5+ messages in thread
From: Linus Walleij @ 2017-04-07 10:16 UTC (permalink / raw)
To: Alexandre Belloni
Cc: Ludovic Desroches, linux-arm-kernel, linux-gpio, linux-kernel
On Thu, Apr 6, 2017 at 3:36 PM, Alexandre Belloni
<alexandre.belloni@free-electrons.com> wrote:
> When suspending to RAM, the power to the core is cut and the register
> values are lost. Save and restore more registers than just IMR.
>
> Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Should this be tagged for stable?
Waiting for input from the AT91 maintainers here.
Yours,
Linus Walleij
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] pinctrl: at91-pio4: handle suspend to ram
2017-04-07 10:16 ` Linus Walleij
@ 2017-04-07 10:48 ` Alexandre Belloni
2017-04-07 12:34 ` Linus Walleij
0 siblings, 1 reply; 5+ messages in thread
From: Alexandre Belloni @ 2017-04-07 10:48 UTC (permalink / raw)
To: Linus Walleij
Cc: Ludovic Desroches, linux-arm-kernel, linux-gpio, linux-kernel
On 07/04/2017 at 12:16:22 +0200, Linus Walleij wrote:
> On Thu, Apr 6, 2017 at 3:36 PM, Alexandre Belloni
> <alexandre.belloni@free-electrons.com> wrote:
>
> > When suspending to RAM, the power to the core is cut and the register
> > values are lost. Save and restore more registers than just IMR.
> >
> > Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
>
> Should this be tagged for stable?
>
No because that particular suspend to ram mode is not yet implemented
upstream.
> Waiting for input from the AT91 maintainers here.
>
Well, I've been an AT91 maintainer for 2 years now and I have done all
the AT91 pull requests since 3.16. I hope this qualifies me here.
--
Alexandre Belloni, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] pinctrl: at91-pio4: handle suspend to ram
2017-04-07 10:48 ` Alexandre Belloni
@ 2017-04-07 12:34 ` Linus Walleij
0 siblings, 0 replies; 5+ messages in thread
From: Linus Walleij @ 2017-04-07 12:34 UTC (permalink / raw)
To: Alexandre Belloni
Cc: Ludovic Desroches, linux-arm-kernel, linux-gpio, linux-kernel
On Fri, Apr 7, 2017 at 12:48 PM, Alexandre Belloni
<alexandre.belloni@free-electrons.com> wrote:
> On 07/04/2017 at 12:16:22 +0200, Linus Walleij wrote:
>> On Thu, Apr 6, 2017 at 3:36 PM, Alexandre Belloni
>> <alexandre.belloni@free-electrons.com> wrote:
>>
>> > When suspending to RAM, the power to the core is cut and the register
>> > values are lost. Save and restore more registers than just IMR.
>> >
>> > Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
>>
>> Should this be tagged for stable?
>>
>
> No because that particular suspend to ram mode is not yet implemented
> upstream.
>
>> Waiting for input from the AT91 maintainers here.
>>
>
> Well, I've been an AT91 maintainer for 2 years now and I have done all
> the AT91 pull requests since 3.16. I hope this qualifies me here.
So much for trying to keep MAINTAINERS in my head :(
I wish there was gmail integration so I could annotate people
with roles...
Patch applied.
Yours,
Linus Walleij
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] pinctrl: at91-pio4: handle suspend to ram
2017-04-06 13:36 [PATCH] pinctrl: at91-pio4: handle suspend to ram Alexandre Belloni
2017-04-07 10:16 ` Linus Walleij
@ 2017-04-07 12:45 ` Ludovic Desroches
1 sibling, 0 replies; 5+ messages in thread
From: Ludovic Desroches @ 2017-04-07 12:45 UTC (permalink / raw)
To: Alexandre Belloni
Cc: Linus Walleij, Ludovic Desroches, linux-arm-kernel, linux-gpio,
linux-kernel
On Thu, Apr 06, 2017 at 03:36:23PM +0200, Alexandre Belloni wrote:
> When suspending to RAM, the power to the core is cut and the register
> values are lost. Save and restore more registers than just IMR.
>
> Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
> ---
> drivers/pinctrl/pinctrl-at91-pio4.c | 34 ++++++++++++++++++++++++++++------
> 1 file changed, 28 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/pinctrl/pinctrl-at91-pio4.c b/drivers/pinctrl/pinctrl-at91-pio4.c
> index 28bbc1bb9e6c..dc8591543dee 100644
> --- a/drivers/pinctrl/pinctrl-at91-pio4.c
> +++ b/drivers/pinctrl/pinctrl-at91-pio4.c
> @@ -126,7 +126,11 @@ struct atmel_pioctrl {
> struct irq_domain *irq_domain;
> int *irqs;
> unsigned *pm_wakeup_sources;
> - unsigned *pm_suspend_backup;
> + struct {
> + u32 imr;
> + u32 odsr;
> + u32 cfgr[ATMEL_PIO_NPINS_PER_BANK];
> + } *pm_suspend_backup;
> struct device *dev;
> struct device_node *node;
> };
> @@ -830,17 +834,26 @@ static int __maybe_unused atmel_pctrl_suspend(struct device *dev)
> {
> struct platform_device *pdev = to_platform_device(dev);
> struct atmel_pioctrl *atmel_pioctrl = platform_get_drvdata(pdev);
> - int i;
> + int i, j;
>
> /*
> * For each bank, save IMR to restore it later and disable all GPIO
> * interrupts excepting the ones marked as wakeup sources.
> */
> for (i = 0; i < atmel_pioctrl->nbanks; i++) {
> - atmel_pioctrl->pm_suspend_backup[i] =
> + atmel_pioctrl->pm_suspend_backup[i].imr =
> atmel_gpio_read(atmel_pioctrl, i, ATMEL_PIO_IMR);
> atmel_gpio_write(atmel_pioctrl, i, ATMEL_PIO_IDR,
> ~atmel_pioctrl->pm_wakeup_sources[i]);
> + atmel_pioctrl->pm_suspend_backup[i].odsr =
> + atmel_gpio_read(atmel_pioctrl, i, ATMEL_PIO_ODSR);
> + for (j = 0; j < ATMEL_PIO_NPINS_PER_BANK; j++) {
> + atmel_gpio_write(atmel_pioctrl, i,
> + ATMEL_PIO_MSKR, BIT(j));
> + atmel_pioctrl->pm_suspend_backup[i].cfgr[j] =
> + atmel_gpio_read(atmel_pioctrl, i,
> + ATMEL_PIO_CFGR);
> + }
> }
>
> return 0;
> @@ -850,11 +863,20 @@ static int __maybe_unused atmel_pctrl_resume(struct device *dev)
> {
> struct platform_device *pdev = to_platform_device(dev);
> struct atmel_pioctrl *atmel_pioctrl = platform_get_drvdata(pdev);
> - int i;
> + int i, j;
>
> - for (i = 0; i < atmel_pioctrl->nbanks; i++)
> + for (i = 0; i < atmel_pioctrl->nbanks; i++) {
> atmel_gpio_write(atmel_pioctrl, i, ATMEL_PIO_IER,
> - atmel_pioctrl->pm_suspend_backup[i]);
> + atmel_pioctrl->pm_suspend_backup[i].imr);
> + atmel_gpio_write(atmel_pioctrl, i, ATMEL_PIO_SODR,
> + atmel_pioctrl->pm_suspend_backup[i].odsr);
I would also configure CODR to be sure to be in the same state.
Otherwise: Acked-by: Ludovic Desroches <ludovic.desroches@microchip.com>
> + for (j = 0; j < ATMEL_PIO_NPINS_PER_BANK; j++) {
> + atmel_gpio_write(atmel_pioctrl, i,
> + ATMEL_PIO_MSKR, BIT(j));
> + atmel_gpio_write(atmel_pioctrl, i, ATMEL_PIO_CFGR,
> + atmel_pioctrl->pm_suspend_backup[i].cfgr[j]);
> + }
> + }
>
> return 0;
> }
> --
> 2.11.0
>
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2017-04-07 12:46 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-04-06 13:36 [PATCH] pinctrl: at91-pio4: handle suspend to ram Alexandre Belloni
2017-04-07 10:16 ` Linus Walleij
2017-04-07 10:48 ` Alexandre Belloni
2017-04-07 12:34 ` Linus Walleij
2017-04-07 12:45 ` Ludovic Desroches
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).