diff mbox

[v3,5/5] pinctrl: sunxi: Use pin number when calling sunxi_pmx_set

Message ID 1454542430-16572-6-git-send-email-k@japko.eu (mailing list archive)
State New, archived
Headers show

Commit Message

Krzysztof Adamski Feb. 3, 2016, 11:33 p.m. UTC
sunxi_pmx_set accepts pin number and then calculates offset by
subtracting pin_base from it. sunxi_pinctrl_gpio_get, on the other hand,
gets offset so we have to convert it to pin number so we won't get
negative value in sunxi_pmx_set.

This was only used on A10 so far, where there is only one GPIO chip with
pin_base set to 0 so it didn't matter. However H3 also requires this
workaround but have two pinmux sections, triggering problem for PL port.

Signed-off-by: Krzysztof Adamski <k@japko.eu>
---
 drivers/pinctrl/sunxi/pinctrl-sunxi.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

Comments

Chen-Yu Tsai Feb. 5, 2016, 9:45 a.m. UTC | #1
On Thu, Feb 4, 2016 at 7:33 AM, Krzysztof Adamski <k@japko.eu> wrote:
> sunxi_pmx_set accepts pin number and then calculates offset by
> subtracting pin_base from it. sunxi_pinctrl_gpio_get, on the other hand,
> gets offset so we have to convert it to pin number so we won't get
> negative value in sunxi_pmx_set.
>
> This was only used on A10 so far, where there is only one GPIO chip with
> pin_base set to 0 so it didn't matter. However H3 also requires this
> workaround but have two pinmux sections, triggering problem for PL port.
>
> Signed-off-by: Krzysztof Adamski <k@japko.eu>

Acked-by: Chen-Yu Tsai <wens@csie.org>

(resent as my mail setup failed to deliver)
Maxime Ripard Feb. 5, 2016, 11:15 a.m. UTC | #2
On Thu, Feb 04, 2016 at 12:33:50AM +0100, Krzysztof Adamski wrote:
> sunxi_pmx_set accepts pin number and then calculates offset by
> subtracting pin_base from it. sunxi_pinctrl_gpio_get, on the other hand,
> gets offset so we have to convert it to pin number so we won't get
> negative value in sunxi_pmx_set.
> 
> This was only used on A10 so far, where there is only one GPIO chip with
> pin_base set to 0 so it didn't matter. However H3 also requires this
> workaround but have two pinmux sections, triggering problem for PL port.
> 
> Signed-off-by: Krzysztof Adamski <k@japko.eu>

Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>

Thanks!
Maxime
diff mbox

Patch

diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.c b/drivers/pinctrl/sunxi/pinctrl-sunxi.c
index 7a2465f..96f64a1 100644
--- a/drivers/pinctrl/sunxi/pinctrl-sunxi.c
+++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.c
@@ -459,15 +459,16 @@  static int sunxi_pinctrl_gpio_get(struct gpio_chip *chip, unsigned offset)
 	u8 index = sunxi_data_offset(offset);
 	u32 set_mux = pctl->desc->irq_read_needs_mux &&
 			test_bit(FLAG_USED_AS_IRQ, &chip->desc[offset].flags);
+	u32 pin = offset + chip->base;
 	u32 val;
 
 	if (set_mux)
-		sunxi_pmx_set(pctl->pctl_dev, offset, SUN4I_FUNC_INPUT);
+		sunxi_pmx_set(pctl->pctl_dev, pin, 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);
+		sunxi_pmx_set(pctl->pctl_dev, pin, SUN4I_FUNC_IRQ);
 
 	return !!val;
 }