[PATCH 1/2] gpio: tegra: create irq mapping in gpio_to_irq

[Laxman Dewangan]

The gpio interrupts get mapped linearly and hence the mapping
of irq need to be created by irq_create_mapping().

The function gpio_to_irq() returns the irq by irq_find_mapping()
and so returns 0 as there is no mapping created. Fix the function
to create mapping when gpio_to_irq() get called.

Signed-off-by: Laxman Dewangan <ldew...@nvidia.com>
---
drivers/gpio/gpio-tegra.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/gpio/gpio-tegra.c b/drivers/gpio/gpio-tegra.c
index c7c175a..6c08613 100644
--- a/drivers/gpio/gpio-tegra.c
+++ b/drivers/gpio/gpio-tegra.c
@@ -156,7 +156,7 @@ static int tegra_gpio_direction_output(struct gpio_chip *chip, unsigned offset,

static int tegra_gpio_to_irq(struct gpio_chip *chip, unsigned offset)
{
- return irq_find_mapping(irq_domain, offset);
+ return irq_create_mapping(irq_domain, offset);
}

static struct gpio_chip tegra_gpio_chip = {
--
1.7.1.1

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

[Laxman Dewangan]

Following are changes done to fix the suspend/resume
functionality of tegra gpio driver:
- Protect suspend/resume callbacks with CONFIG_PM_SLEEP
because CONFIG_PM doesn't actually enable any of the PM callbacks, it
only allows to enable CONFIG_PM_SLEEP and CONFIG_PM_RUNTIME.
This means if CONFIG_PM is used to protect system sleep callbacks
then it may end up unreferenced if only runtime PM is enabled.

- Fix the suspend/resume APIs declaration as per callback prototype.

Signed-off-by: Laxman Dewangan <ldew...@nvidia.com>
---

drivers/gpio/gpio-tegra.c | 18 +++++++++++++-----
1 files changed, 13 insertions(+), 5 deletions(-)

diff --git a/drivers/gpio/gpio-tegra.c b/drivers/gpio/gpio-tegra.c
index 6c08613..9c1884b 100644
--- a/drivers/gpio/gpio-tegra.c
+++ b/drivers/gpio/gpio-tegra.c
@@ -27,6 +27,7 @@
#include <linux/module.h>
#include <linux/irqdomain.h>
#include <linux/pinctrl/consumer.h>
+#include <linux/pm.h>

#include <asm/mach/irq.h>

@@ -64,7 +65,7 @@ struct tegra_gpio_bank {
int bank;
int irq;
spinlock_t lvl_lock[4];
-#ifdef CONFIG_PM
+#ifdef CONFIG_PM_SLEEP
u32 cnf[4];
u32 out[4];
u32 oe[4];
@@ -285,8 +286,8 @@ static void tegra_gpio_irq_handler(unsigned int irq, struct irq_desc *desc)

}

-#ifdef CONFIG_PM
-void tegra_gpio_resume(void)
+#ifdef CONFIG_PM_SLEEP
+static int tegra_gpio_resume(struct device *dev)
{
unsigned long flags;
int b;
@@ -308,9 +309,10 @@ void tegra_gpio_resume(void)
}

local_irq_restore(flags);
+ return 0;
}

-void tegra_gpio_suspend(void)
+static int tegra_gpio_suspend(struct device *dev)
{
unsigned long flags;
int b;
@@ -330,6 +332,7 @@ void tegra_gpio_suspend(void)
}
}
local_irq_restore(flags);
+ return 0;
}

static int tegra_gpio_wake_enable(struct irq_data *d, unsigned int enable)
@@ -345,11 +348,15 @@ static struct irq_chip tegra_gpio_irq_chip = {
.irq_mask = tegra_gpio_irq_mask,
.irq_unmask = tegra_gpio_irq_unmask,
.irq_set_type = tegra_gpio_irq_set_type,
-#ifdef CONFIG_PM
+#ifdef CONFIG_PM_SLEEP
.irq_set_wake = tegra_gpio_wake_enable,
#endif
};

+static const struct dev_pm_ops tegra_gpio_pm_ops = {
+ SET_SYSTEM_SLEEP_PM_OPS(tegra_gpio_suspend, tegra_gpio_resume)
+};
+
struct tegra_gpio_soc_config {
u32 bank_stride;
u32 upper_offset;
@@ -489,6 +496,7 @@ static struct platform_driver tegra_gpio_driver = {
.driver = {
.name = "tegra-gpio",
.owner = THIS_MODULE,
+ .pm = &tegra_gpio_pm_ops,
.of_match_table = tegra_gpio_of_match,
},
.probe = tegra_gpio_probe,

[Stephen Warren]

On 11/07/2012 08:01 AM, Laxman Dewangan wrote:
> The gpio interrupts get mapped linearly and hence the mapping
> of irq need to be created by irq_create_mapping().
>
> The function gpio_to_irq() returns the irq by irq_find_mapping()
> and so returns 0 as there is no mapping created. Fix the function
> to create mapping when gpio_to_irq() get called.

I'm not convinced this should be needed. tegra_gpio_probe() contains:

> for (gpio = 0; gpio < tegra_gpio_chip.ngpio; gpio++) {
> int irq = irq_create_mapping(irq_domain, gpio);

which should create the mapping for every IRQ.

(although I do think the gpiochip_add() should be moved to the very end
of probe(), I doubt that impacts this issue much)

[Stephen Warren]

On 11/07/2012 08:01 AM, Laxman Dewangan wrote:

> Following are changes done to fix the suspend/resume
> functionality of tegra gpio driver:
> - Protect suspend/resume callbacks with CONFIG_PM_SLEEP
> because CONFIG_PM doesn't actually enable any of the PM callbacks, it
> only allows to enable CONFIG_PM_SLEEP and CONFIG_PM_RUNTIME.
> This means if CONFIG_PM is used to protect system sleep callbacks
> then it may end up unreferenced if only runtime PM is enabled.
>
> - Fix the suspend/resume APIs declaration as per callback prototype.

Seems plausible, so this patch,
Acked-by: Stephen Warren <swa...@nvidia.com>

I wonder why the suspend/resume functions weren't actually hooked up
anywhere before. Oh well.

Just one very minor nit below:

> diff --git a/drivers/gpio/gpio-tegra.c b/drivers/gpio/gpio-tegra.c

> +#ifdef CONFIG_PM_SLEEP
> +static int tegra_gpio_resume(struct device *dev)

There are two spaces after "int" there.

[Laxman Dewangan]

On Wednesday 07 November 2012 10:38 PM, Stephen Warren wrote:
> On 11/07/2012 08:01 AM, Laxman Dewangan wrote:
>> The gpio interrupts get mapped linearly and hence the mapping
>> of irq need to be created by irq_create_mapping().
>>
>> The function gpio_to_irq() returns the irq by irq_find_mapping()
>> and so returns 0 as there is no mapping created. Fix the function
>> to create mapping when gpio_to_irq() get called.
> I'm not convinced this should be needed. tegra_gpio_probe() contains:

Oh, yes, I did not observe this code in my review. So this change is not
resolving any issue.


I think we should move the mapping to gpio_to_irq() rather than doing
this in probe.

[Linus Walleij]

On Wed, Nov 7, 2012 at 4:01 PM, Laxman Dewangan <ldew...@nvidia.com> wrote:

> Following are changes done to fix the suspend/resume
> functionality of tegra gpio driver:
> - Protect suspend/resume callbacks with CONFIG_PM_SLEEP
> because CONFIG_PM doesn't actually enable any of the PM callbacks, it
> only allows to enable CONFIG_PM_SLEEP and CONFIG_PM_RUNTIME.
> This means if CONFIG_PM is used to protect system sleep callbacks
> then it may end up unreferenced if only runtime PM is enabled.
>
> - Fix the suspend/resume APIs declaration as per callback prototype.
>
> Signed-off-by: Laxman Dewangan <ldew...@nvidia.com>

OK patch applied to my devel branch, with Stephens ACK
and I also fixed the whitespace issue he pointed out.

Yours,
Linus Walleij