[PATCH] pinctrl: sun4i: GPIOs configured as irq must be set to input before reading

[PATCH] pinctrl: sun4i: GPIOs configured as irq must be set to input before reading

[Hans de Goede]

On sun4i-a10, when GPIOs are configured as external interrupt the value for
them in the data register does not seem to get updated, so set their mux to
input (and restore afterwards) when reading the pin.

Missed edges seem to be buffered, so this does not introduce a race
condition.

I've also tested this on sun5i-a13 and sun7i-a20 and those do not seem to
be affected, the input value representation in the data register does seem
to correctly get updated to the actual pin value while in irq mode there.

Signed-off-by: Hans de Goede <hdegoede-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
---
 drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c |  1 +
 drivers/pinctrl/sunxi/pinctrl-sunxi.c     | 14 ++++++++++++--
 drivers/pinctrl/sunxi/pinctrl-sunxi.h     |  4 ++++
 3 files changed, 17 insertions(+), 2 deletions(-)

diff --git a/drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c b/drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c
index 24c5d88..3c68a8e 100644
--- a/drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c
+++ b/drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c
@@ -1011,6 +1011,7 @@ static const struct sunxi_pinctrl_desc sun4i_a10_pinctrl_data = {
 	.pins = sun4i_a10_pins,
 	.npins = ARRAY_SIZE(sun4i_a10_pins),
 	.irq_banks = 1,
+	.irq_read_needs_mux = true,
 };
 
 static int sun4i_a10_pinctrl_probe(struct platform_device *pdev)
diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.c b/drivers/pinctrl/sunxi/pinctrl-sunxi.c
index 3d074433..f8e171b 100644
--- a/drivers/pinctrl/sunxi/pinctrl-sunxi.c
+++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.c
@@ -29,6 +29,7 @@
 #include <linux/slab.h>
 
 #include "../core.h"
+#include "../../gpio/gpiolib.h"
 #include "pinctrl-sunxi.h"
 
 static struct irq_chip sunxi_pinctrl_edge_irq_chip;
@@ -464,10 +465,19 @@ static int sunxi_pinctrl_gpio_direction_input(struct gpio_chip *chip,
 static int sunxi_pinctrl_gpio_get(struct gpio_chip *chip, unsigned offset)
 {
 	struct sunxi_pinctrl *pctl = dev_get_drvdata(chip->dev);
-
 	u32 reg = sunxi_data_reg(offset);
 	u8 index = sunxi_data_offset(offset);
-	u32 val = (readl(pctl->membase + reg) >> index) & DATA_PINS_MASK;
+	u32 set_mux = pctl->desc->irq_read_needs_mux &&
+			test_bit(FLAG_USED_AS_IRQ, &chip->desc[offset].flags);
+	u32 val;
+
+	if (set_mux)
+		sunxi_pmx_set(pctl->pctl_dev, offset, SUN4I_FUNC_INPUT);
+
+	val = (readl(pctl->membase + reg) >> index) & DATA_PINS_MASK;
+
+	if (set_mux)
+		sunxi_pmx_set(pctl->pctl_dev, offset, SUN4I_FUNC_IRQ);
 
 	return val;
 }
diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.h b/drivers/pinctrl/sunxi/pinctrl-sunxi.h
index 5a51523..e248e81 100644
--- a/drivers/pinctrl/sunxi/pinctrl-sunxi.h
+++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.h
@@ -77,6 +77,9 @@
 #define IRQ_LEVEL_LOW		0x03
 #define IRQ_EDGE_BOTH		0x04
 
+#define SUN4I_FUNC_INPUT	0
+#define SUN4I_FUNC_IRQ		6
+
 struct sunxi_desc_function {
 	const char	*name;
 	u8		muxval;
@@ -94,6 +97,7 @@ struct sunxi_pinctrl_desc {
 	int				npins;
 	unsigned			pin_base;
 	unsigned			irq_banks;
+	bool				irq_read_needs_mux;
 };
 
 struct sunxi_pinctrl_function {
-- 
2.3.1

Re: [PATCH] pinctrl: sun4i: GPIOs configured as irq must be set to input before reading

[Maxime Ripard]

[-- Attachment #1: Type: text/plain, Size: 891 bytes --]

On Sun, Mar 08, 2015 at 10:13:57PM +0100, Hans de Goede wrote:
> On sun4i-a10, when GPIOs are configured as external interrupt the value for
> them in the data register does not seem to get updated, so set their mux to
> input (and restore afterwards) when reading the pin.
> 
> Missed edges seem to be buffered, so this does not introduce a race
> condition.
> 
> I've also tested this on sun5i-a13 and sun7i-a20 and those do not seem to
> be affected, the input value representation in the data register does seem
> to correctly get updated to the actual pin value while in irq mode there.
> 
> Signed-off-by: Hans de Goede <hdegoede-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>

Acked-by: Maxime Ripard <maxime.ripard-wi1+55ScJUtKEb57/3fJTNBPR1lH4CV8@public.gmane.org>

Thanks!
Maxime

-- 
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

Re: [PATCH] pinctrl: sun4i: GPIOs configured as irq must be set to input before reading

[Linus Walleij]

On Sun, Mar 8, 2015 at 10:13 PM, Hans de Goede <hdegoede-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> wrote:

> On sun4i-a10, when GPIOs are configured as external interrupt the value for
> them in the data register does not seem to get updated, so set their mux to
> input (and restore afterwards) when reading the pin.
>
> Missed edges seem to be buffered, so this does not introduce a race
> condition.
>
> I've also tested this on sun5i-a13 and sun7i-a20 and those do not seem to
> be affected, the input value representation in the data register does seem
> to correctly get updated to the actual pin value while in irq mode there.
>
> Signed-off-by: Hans de Goede <hdegoede-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>

Patch applied for fixes with Maxime's ACK.

Yours,
Linus Walleij