If there is a gpio range mapping for the pin, then print out the gpio
chip and line index for the pin in the debugfs 'pins' file with the
format: "[gpiochip-label]:line-[index] "
For example, here is a section of 'pins' the PocketBeagle (AM3358):
/sys/kernel/debug/pinctrl/44e10800.pinmux-pinctrl-single/pins
pin 25 (PIN25) gpio-32-63:line-25 44e10864 00000037 pinctrl-single
pin 26 (PIN26) gpio-32-63:line-26 44e10868 00000037 pinctrl-single
pin 27 (PIN27) gpio-32-63:line-27 44e1086c 00000037 pinctrl-single
pin 28 (PIN28) NA 44e10870 00000036 pinctrl-single
pin 29 (PIN29) NA 44e10874 00000006 pinctrl-single
pin 30 (PIN30) gpio-32-63:line-28 44e10878 00000027 pinctrl-single
pin 31 (PIN31) gpio-32-63:line-29 44e1087c 00000037 pinctrl-single
pin 32 (PIN32) gpio-32-63:line-30 44e10880 00000037 pinctrl-single
pin 33 (PIN33) gpio-32-63:line-31 44e10884 00000037 pinctrl-single
pin 34 (PIN34) gpio-64-95:line-0 44e10888 00000037 pinctrl-single
pin 35 (PIN35) gpio-64-95:line-1 44e1088c 00000037 pinctrl-single
Suggested-by: Andy Shevchenko <[email protected]>
Suggested-by: Tony Lindgren <[email protected]>
Signed-off-by: Drew Fustini <[email protected]>
v3 changes:
- gpio column is now gpiochip label and line index
v2 changes:
- print 'NA' if pin does not have a GPIO number
- change gpio_num from unsigned to unsigned int per checkpatch
---
drivers/pinctrl/core.c | 21 +++++++++++++++++++++
1 file changed, 21 insertions(+)
diff --git a/drivers/pinctrl/core.c b/drivers/pinctrl/core.c
index 821242bb4b16..a23dc264cae7 100644
--- a/drivers/pinctrl/core.c
+++ b/drivers/pinctrl/core.c
@@ -27,6 +27,7 @@
#include <linux/pinctrl/machine.h>
#ifdef CONFIG_GPIOLIB
+#include "../gpio/gpiolib.h"
#include <asm-generic/gpio.h>
#endif
@@ -1601,6 +1602,9 @@ static int pinctrl_pins_show(struct seq_file *s, void *what)
struct pinctrl_dev *pctldev = s->private;
const struct pinctrl_ops *ops = pctldev->desc->pctlops;
unsigned i, pin;
+ struct pinctrl_gpio_range *range;
+ unsigned int gpio_num;
+ struct gpio_chip *chip;
seq_printf(s, "registered pins: %d\n", pctldev->desc->npins);
@@ -1618,6 +1622,23 @@ static int pinctrl_pins_show(struct seq_file *s, void *what)
seq_printf(s, "pin %d (%s) ", pin, desc->name);
+#ifdef CONFIG_GPIOLIB
+ gpio_num = 0;
+ list_for_each_entry(range, &pctldev->gpio_ranges, node) {
+ if ((pin >= range->pin_base) &&
+ (pin < (range->pin_base + range->npins))) {
+ gpio_num = range->base + (pin - range->pin_base);
+ break;
+ }
+ }
+ chip = gpio_to_chip(gpio_num);
+ if (chip && chip->gpiodev && chip->gpiodev->base)
+ seq_printf(s, "%s:line-%u ", chip->label,
+ gpio_num - chip->gpiodev->base);
+ else
+ seq_puts(s, "NA ");
+#endif
+
/* Driver-specific info per pin */
if (ops->pin_dbg_show)
ops->pin_dbg_show(pctldev, s, pin);
--
2.25.1
On Tue, Jul 21, 2020 at 04:02:34PM +0200, Drew Fustini wrote:
> If there is a gpio range mapping for the pin, then print out the gpio
> chip and line index for the pin in the debugfs 'pins' file with the
> format: "[gpiochip-label]:line-[index] "
>
> For example, here is a section of 'pins' the PocketBeagle (AM3358):
> /sys/kernel/debug/pinctrl/44e10800.pinmux-pinctrl-single/pins
>
> pin 25 (PIN25) gpio-32-63:line-25 44e10864 00000037 pinctrl-single
> pin 26 (PIN26) gpio-32-63:line-26 44e10868 00000037 pinctrl-single
> pin 27 (PIN27) gpio-32-63:line-27 44e1086c 00000037 pinctrl-single
> pin 28 (PIN28) NA 44e10870 00000036 pinctrl-single
> pin 29 (PIN29) NA 44e10874 00000006 pinctrl-single
> pin 30 (PIN30) gpio-32-63:line-28 44e10878 00000027 pinctrl-single
> pin 31 (PIN31) gpio-32-63:line-29 44e1087c 00000037 pinctrl-single
> pin 32 (PIN32) gpio-32-63:line-30 44e10880 00000037 pinctrl-single
> pin 33 (PIN33) gpio-32-63:line-31 44e10884 00000037 pinctrl-single
> pin 34 (PIN34) gpio-64-95:line-0 44e10888 00000037 pinctrl-single
> pin 35 (PIN35) gpio-64-95:line-1 44e1088c 00000037 pinctrl-single
>
> Suggested-by: Andy Shevchenko <[email protected]>
> Suggested-by: Tony Lindgren <[email protected]>
> Signed-off-by: Drew Fustini <[email protected]>
>
> v3 changes:
> - gpio column is now gpiochip label and line index
>
> v2 changes:
> - print 'NA' if pin does not have a GPIO number
> - change gpio_num from unsigned to unsigned int per checkpatch
> ---
> drivers/pinctrl/core.c | 21 +++++++++++++++++++++
> 1 file changed, 21 insertions(+)
>
> diff --git a/drivers/pinctrl/core.c b/drivers/pinctrl/core.c
> index 821242bb4b16..a23dc264cae7 100644
> --- a/drivers/pinctrl/core.c
> +++ b/drivers/pinctrl/core.c
> @@ -27,6 +27,7 @@
> #include <linux/pinctrl/machine.h>
>
> #ifdef CONFIG_GPIOLIB
> +#include "../gpio/gpiolib.h"
> #include <asm-generic/gpio.h>
> #endif
>
> @@ -1601,6 +1602,9 @@ static int pinctrl_pins_show(struct seq_file *s, void *what)
> struct pinctrl_dev *pctldev = s->private;
> const struct pinctrl_ops *ops = pctldev->desc->pctlops;
> unsigned i, pin;
> + struct pinctrl_gpio_range *range;
> + unsigned int gpio_num;
> + struct gpio_chip *chip;
>
> seq_printf(s, "registered pins: %d\n", pctldev->desc->npins);
>
> @@ -1618,6 +1622,23 @@ static int pinctrl_pins_show(struct seq_file *s, void *what)
>
> seq_printf(s, "pin %d (%s) ", pin, desc->name);
>
> +#ifdef CONFIG_GPIOLIB
> + gpio_num = 0;
> + list_for_each_entry(range, &pctldev->gpio_ranges, node) {
> + if ((pin >= range->pin_base) &&
> + (pin < (range->pin_base + range->npins))) {
> + gpio_num = range->base + (pin - range->pin_base);
> + break;
> + }
> + }
> + chip = gpio_to_chip(gpio_num);
> + if (chip && chip->gpiodev && chip->gpiodev->base)
> + seq_printf(s, "%s:line-%u ", chip->label,
> + gpio_num - chip->gpiodev->base);
> + else
> + seq_puts(s, "NA ");
> +#endif
> +
> /* Driver-specific info per pin */
> if (ops->pin_dbg_show)
> ops->pin_dbg_show(pctldev, s, pin);
> --
> 2.25.1
>
Apologies - I should not have put the change log in the commit message.
Please let me know if I should resubmit.
thanks,
drew
On Tue, Jul 21, 2020 at 5:10 PM Drew Fustini <[email protected]> wrote:
> On Tue, Jul 21, 2020 at 04:02:34PM +0200, Drew Fustini wrote:
> > If there is a gpio range mapping for the pin, then print out the gpio
> > chip and line index for the pin in the debugfs 'pins' file with the
> > format: "[gpiochip-label]:line-[index] "
> >
> > For example, here is a section of 'pins' the PocketBeagle (AM3358):
> > /sys/kernel/debug/pinctrl/44e10800.pinmux-pinctrl-single/pins
> >
> > pin 25 (PIN25) gpio-32-63:line-25 44e10864 00000037 pinctrl-single
> > pin 26 (PIN26) gpio-32-63:line-26 44e10868 00000037 pinctrl-single
> > pin 27 (PIN27) gpio-32-63:line-27 44e1086c 00000037 pinctrl-single
> > pin 28 (PIN28) NA 44e10870 00000036 pinctrl-single
> > pin 29 (PIN29) NA 44e10874 00000006 pinctrl-single
> > pin 30 (PIN30) gpio-32-63:line-28 44e10878 00000027 pinctrl-single
> > pin 31 (PIN31) gpio-32-63:line-29 44e1087c 00000037 pinctrl-single
> > pin 32 (PIN32) gpio-32-63:line-30 44e10880 00000037 pinctrl-single
> > pin 33 (PIN33) gpio-32-63:line-31 44e10884 00000037 pinctrl-single
> > pin 34 (PIN34) gpio-64-95:line-0 44e10888 00000037 pinctrl-single
> > pin 35 (PIN35) gpio-64-95:line-1 44e1088c 00000037 pinctrl-single
Because line is integer and label is string it is better (from parsing
of view) to put it other way around, i.e.
%u:%s, label, line
...
> Apologies - I should not have put the change log in the commit message.
> Please let me know if I should resubmit.
Since the above comment, I guess it would be good.
Thanks for doing this!
--
With Best Regards,
Andy Shevchenko
On Tue, Jul 21, 2020 at 5:17 PM Andy Shevchenko
<[email protected]> wrote:
> On Tue, Jul 21, 2020 at 5:10 PM Drew Fustini <[email protected]> wrote:
> > On Tue, Jul 21, 2020 at 04:02:34PM +0200, Drew Fustini wrote:
> > > If there is a gpio range mapping for the pin, then print out the gpio
> > > chip and line index for the pin in the debugfs 'pins' file with the
> > > format: "[gpiochip-label]:line-[index] "
...
> > > pin 25 (PIN25) gpio-32-63:line-25 44e10864 00000037 pinctrl-single
> > > pin 26 (PIN26) gpio-32-63:line-26 44e10868 00000037 pinctrl-single
> > > pin 27 (PIN27) gpio-32-63:line-27 44e1086c 00000037 pinctrl-single
> > > pin 28 (PIN28) NA 44e10870 00000036 pinctrl-single
> > > pin 29 (PIN29) NA 44e10874 00000006 pinctrl-single
And would be also better to have
0:N/A
to keep both arguments in it.
> Because line is integer and label is string it is better (from parsing
> of view) to put it other way around, i.e.
>
> %u:%s, label, line
>
> ...
>
> > Apologies - I should not have put the change log in the commit message.
> > Please let me know if I should resubmit.
>
> Since the above comment, I guess it would be good.
> Thanks for doing this!
--
With Best Regards,
Andy Shevchenko