* [PATCH] gpio: dwapb: Rework support for 1 interrupt per port A GPIO
@ 2018-05-23 8:52 Phil Edworthy
2018-05-23 9:12 ` Simon Horman
2018-05-23 9:30 ` Linus Walleij
0 siblings, 2 replies; 4+ messages in thread
From: Phil Edworthy @ 2018-05-23 8:52 UTC (permalink / raw)
To: Andy Shevchenko, Hoan Tran, Linus Walleij
Cc: Lee Jones, Michel Pollet, linux-gpio, linux-renesas-soc,
linux-kernel, Phil Edworthy
Treat DT and ACPI the same as much as possible. Note that we can't use
platform_get_irq() to get the DT interrupts as they are in the port
sub-node and hence do not have an associated platform device.
This also fixes a problem introduced with error checking when calling
platform_get_irq().
Signed-off-by: Phil Edworthy <phil.edworthy@renesas.com>
---
drivers/gpio/gpio-dwapb.c | 53 ++++++++++++++++++++---------------------------
1 file changed, 22 insertions(+), 31 deletions(-)
diff --git a/drivers/gpio/gpio-dwapb.c b/drivers/gpio/gpio-dwapb.c
index 7dcd06b..15b4154 100644
--- a/drivers/gpio/gpio-dwapb.c
+++ b/drivers/gpio/gpio-dwapb.c
@@ -444,7 +444,7 @@ static void dwapb_configure_irqs(struct dwapb_gpio *gpio,
int i;
for (i = 0; i < pp->ngpio; i++) {
- if (pp->irq[i])
+ if (pp->irq[i] >= 0)
irq_set_chained_handler_and_data(pp->irq[i],
dwapb_irq_handler, gpio);
}
@@ -562,7 +562,7 @@ dwapb_gpio_get_pdata(struct device *dev)
struct dwapb_platform_data *pdata;
struct dwapb_port_property *pp;
int nports;
- int i;
+ int i, j;
nports = device_get_child_node_count(dev);
if (nports == 0)
@@ -580,6 +580,8 @@ dwapb_gpio_get_pdata(struct device *dev)
i = 0;
device_for_each_child_node(dev, fwnode) {
+ struct device_node *np = NULL;
+
pp = &pdata->properties[i++];
pp->fwnode = fwnode;
@@ -599,46 +601,35 @@ dwapb_gpio_get_pdata(struct device *dev)
pp->ngpio = 32;
}
+ pp->irq_shared = false;
+ pp->gpio_base = -1;
+
/*
* Only port A can provide interrupts in all configurations of
* the IP.
*/
- if (dev->of_node && pp->idx == 0 &&
- fwnode_property_read_bool(fwnode,
- "interrupt-controller")) {
- struct device_node *np = to_of_node(fwnode);
- unsigned int j;
-
- /*
- * The IP has configuration options to allow a single
- * combined interrupt or one per gpio. If one per gpio,
- * some might not be used.
- */
- for (j = 0; j < pp->ngpio; j++) {
- int irq = of_irq_get(np, j);
- if (irq < 0)
- continue;
-
- pp->irq[j] = irq;
- pp->has_irq = true;
- }
+ if (pp->idx != 0)
+ continue;
- if (!pp->has_irq)
- dev_warn(dev, "no irq for port%d\n", pp->idx);
+ if (dev->of_node && fwnode_property_read_bool(fwnode,
+ "interrupt-controller")) {
+ np = to_of_node(fwnode);
}
- if (has_acpi_companion(dev) && pp->idx == 0) {
- unsigned int j;
+ for (j = 0; j < pp->ngpio; j++) {
+ pp->irq[j] = -ENXIO;
- for (j = 0; j < pp->ngpio; j++) {
+ if (np)
+ pp->irq[j] = of_irq_get(np, j);
+ else if (has_acpi_companion(dev))
pp->irq[j] = platform_get_irq(to_platform_device(dev), j);
- if (pp->irq[j])
- pp->has_irq = true;
- }
+
+ if (pp->irq[j] >= 0)
+ pp->has_irq = true;
}
- pp->irq_shared = false;
- pp->gpio_base = -1;
+ if (!pp->has_irq)
+ dev_warn(dev, "no irq for port%d\n", pp->idx);
}
return pdata;
--
2.7.4
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH] gpio: dwapb: Rework support for 1 interrupt per port A GPIO
2018-05-23 8:52 [PATCH] gpio: dwapb: Rework support for 1 interrupt per port A GPIO Phil Edworthy
@ 2018-05-23 9:12 ` Simon Horman
2018-05-23 9:18 ` Phil Edworthy
2018-05-23 9:30 ` Linus Walleij
1 sibling, 1 reply; 4+ messages in thread
From: Simon Horman @ 2018-05-23 9:12 UTC (permalink / raw)
To: Phil Edworthy
Cc: Andy Shevchenko, Hoan Tran, Linus Walleij, Lee Jones,
Michel Pollet, linux-gpio, linux-renesas-soc, linux-kernel
On Wed, May 23, 2018 at 09:52:44AM +0100, Phil Edworthy wrote:
> Treat DT and ACPI the same as much as possible. Note that we can't use
> platform_get_irq() to get the DT interrupts as they are in the port
> sub-node and hence do not have an associated platform device.
>
> This also fixes a problem introduced with error checking when calling
> platform_get_irq().
What is the problem? In general I think fixes should be in separate
patches.
>
> Signed-off-by: Phil Edworthy <phil.edworthy@renesas.com>
> ---
> drivers/gpio/gpio-dwapb.c | 53 ++++++++++++++++++++---------------------------
> 1 file changed, 22 insertions(+), 31 deletions(-)
>
> diff --git a/drivers/gpio/gpio-dwapb.c b/drivers/gpio/gpio-dwapb.c
> index 7dcd06b..15b4154 100644
> --- a/drivers/gpio/gpio-dwapb.c
> +++ b/drivers/gpio/gpio-dwapb.c
> @@ -444,7 +444,7 @@ static void dwapb_configure_irqs(struct dwapb_gpio *gpio,
> int i;
>
> for (i = 0; i < pp->ngpio; i++) {
> - if (pp->irq[i])
> + if (pp->irq[i] >= 0)
> irq_set_chained_handler_and_data(pp->irq[i],
> dwapb_irq_handler, gpio);
> }
> @@ -562,7 +562,7 @@ dwapb_gpio_get_pdata(struct device *dev)
> struct dwapb_platform_data *pdata;
> struct dwapb_port_property *pp;
> int nports;
> - int i;
> + int i, j;
>
> nports = device_get_child_node_count(dev);
> if (nports == 0)
> @@ -580,6 +580,8 @@ dwapb_gpio_get_pdata(struct device *dev)
>
> i = 0;
> device_for_each_child_node(dev, fwnode) {
> + struct device_node *np = NULL;
> +
> pp = &pdata->properties[i++];
> pp->fwnode = fwnode;
>
> @@ -599,46 +601,35 @@ dwapb_gpio_get_pdata(struct device *dev)
> pp->ngpio = 32;
> }
>
> + pp->irq_shared = false;
> + pp->gpio_base = -1;
> +
> /*
> * Only port A can provide interrupts in all configurations of
> * the IP.
> */
> - if (dev->of_node && pp->idx == 0 &&
> - fwnode_property_read_bool(fwnode,
> - "interrupt-controller")) {
> - struct device_node *np = to_of_node(fwnode);
> - unsigned int j;
> -
> - /*
> - * The IP has configuration options to allow a single
> - * combined interrupt or one per gpio. If one per gpio,
> - * some might not be used.
> - */
> - for (j = 0; j < pp->ngpio; j++) {
> - int irq = of_irq_get(np, j);
> - if (irq < 0)
> - continue;
> -
> - pp->irq[j] = irq;
> - pp->has_irq = true;
> - }
> + if (pp->idx != 0)
> + continue;
>
> - if (!pp->has_irq)
> - dev_warn(dev, "no irq for port%d\n", pp->idx);
> + if (dev->of_node && fwnode_property_read_bool(fwnode,
> + "interrupt-controller")) {
> + np = to_of_node(fwnode);
> }
>
> - if (has_acpi_companion(dev) && pp->idx == 0) {
> - unsigned int j;
> + for (j = 0; j < pp->ngpio; j++) {
> + pp->irq[j] = -ENXIO;
>
> - for (j = 0; j < pp->ngpio; j++) {
> + if (np)
> + pp->irq[j] = of_irq_get(np, j);
> + else if (has_acpi_companion(dev))
> pp->irq[j] = platform_get_irq(to_platform_device(dev), j);
> - if (pp->irq[j])
> - pp->has_irq = true;
> - }
> +
> + if (pp->irq[j] >= 0)
> + pp->has_irq = true;
> }
>
> - pp->irq_shared = false;
> - pp->gpio_base = -1;
> + if (!pp->has_irq)
> + dev_warn(dev, "no irq for port%d\n", pp->idx);
> }
>
> return pdata;
> --
> 2.7.4
>
^ permalink raw reply [flat|nested] 4+ messages in thread
* RE: [PATCH] gpio: dwapb: Rework support for 1 interrupt per port A GPIO
2018-05-23 9:12 ` Simon Horman
@ 2018-05-23 9:18 ` Phil Edworthy
0 siblings, 0 replies; 4+ messages in thread
From: Phil Edworthy @ 2018-05-23 9:18 UTC (permalink / raw)
To: Simon Horman
Cc: Andy Shevchenko, Hoan Tran, Linus Walleij, Lee Jones,
Michel Pollet, linux-gpio@vger.kernel.org,
linux-renesas-soc@vger.kernel.org, linux-kernel@vger.kernel.org
Hi Simon,
On 23 May 2018 10:12 Simon Horman wrote:
> On Wed, May 23, 2018 at 09:52:44AM +0100, Phil Edworthy wrote:
> > Treat DT and ACPI the same as much as possible. Note that we can't use
> > platform_get_irq() to get the DT interrupts as they are in the port
> > sub-node and hence do not have an associated platform device.
> >
> > This also fixes a problem introduced with error checking when calling
> > platform_get_irq().
>
> What is the problem? In general I think fixes should be in separate patches.
The ACPI code would ignore errors returned by platform_get_irq(), instead
treating the error as an interrupt number.
Andy Shevchenko provided some late feedback to the v5 patch, but v5 was
already applied by Linus W. This patch just has the incremental changes that
were made in v6 of the patch.
BR
Phil
> >
> > Signed-off-by: Phil Edworthy <phil.edworthy@renesas.com>
> > ---
> > drivers/gpio/gpio-dwapb.c | 53
> > ++++++++++++++++++++---------------------------
> > 1 file changed, 22 insertions(+), 31 deletions(-)
> >
> > diff --git a/drivers/gpio/gpio-dwapb.c b/drivers/gpio/gpio-dwapb.c
> > index 7dcd06b..15b4154 100644
> > --- a/drivers/gpio/gpio-dwapb.c
> > +++ b/drivers/gpio/gpio-dwapb.c
> > @@ -444,7 +444,7 @@ static void dwapb_configure_irqs(struct
> dwapb_gpio *gpio,
> > int i;
> >
> > for (i = 0; i < pp->ngpio; i++) {
> > - if (pp->irq[i])
> > + if (pp->irq[i] >= 0)
> > irq_set_chained_handler_and_data(pp-
> >irq[i],
> > dwapb_irq_handler, gpio);
> > }
> > @@ -562,7 +562,7 @@ dwapb_gpio_get_pdata(struct device *dev)
> > struct dwapb_platform_data *pdata;
> > struct dwapb_port_property *pp;
> > int nports;
> > - int i;
> > + int i, j;
> >
> > nports = device_get_child_node_count(dev);
> > if (nports == 0)
> > @@ -580,6 +580,8 @@ dwapb_gpio_get_pdata(struct device *dev)
> >
> > i = 0;
> > device_for_each_child_node(dev, fwnode) {
> > + struct device_node *np = NULL;
> > +
> > pp = &pdata->properties[i++];
> > pp->fwnode = fwnode;
> >
> > @@ -599,46 +601,35 @@ dwapb_gpio_get_pdata(struct device *dev)
> > pp->ngpio = 32;
> > }
> >
> > + pp->irq_shared = false;
> > + pp->gpio_base = -1;
> > +
> > /*
> > * Only port A can provide interrupts in all configurations of
> > * the IP.
> > */
> > - if (dev->of_node && pp->idx == 0 &&
> > - fwnode_property_read_bool(fwnode,
> > - "interrupt-controller")) {
> > - struct device_node *np = to_of_node(fwnode);
> > - unsigned int j;
> > -
> > - /*
> > - * The IP has configuration options to allow a single
> > - * combined interrupt or one per gpio. If one per
> gpio,
> > - * some might not be used.
> > - */
> > - for (j = 0; j < pp->ngpio; j++) {
> > - int irq = of_irq_get(np, j);
> > - if (irq < 0)
> > - continue;
> > -
> > - pp->irq[j] = irq;
> > - pp->has_irq = true;
> > - }
> > + if (pp->idx != 0)
> > + continue;
> >
> > - if (!pp->has_irq)
> > - dev_warn(dev, "no irq for port%d\n", pp-
> >idx);
> > + if (dev->of_node && fwnode_property_read_bool(fwnode,
> > + "interrupt-controller")) {
> > + np = to_of_node(fwnode);
> > }
> >
> > - if (has_acpi_companion(dev) && pp->idx == 0) {
> > - unsigned int j;
> > + for (j = 0; j < pp->ngpio; j++) {
> > + pp->irq[j] = -ENXIO;
> >
> > - for (j = 0; j < pp->ngpio; j++) {
> > + if (np)
> > + pp->irq[j] = of_irq_get(np, j);
> > + else if (has_acpi_companion(dev))
> > pp->irq[j] =
> platform_get_irq(to_platform_device(dev), j);
> > - if (pp->irq[j])
> > - pp->has_irq = true;
> > - }
> > +
> > + if (pp->irq[j] >= 0)
> > + pp->has_irq = true;
> > }
> >
> > - pp->irq_shared = false;
> > - pp->gpio_base = -1;
> > + if (!pp->has_irq)
> > + dev_warn(dev, "no irq for port%d\n", pp->idx);
> > }
> >
> > return pdata;
> > --
> > 2.7.4
> >
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] gpio: dwapb: Rework support for 1 interrupt per port A GPIO
2018-05-23 8:52 [PATCH] gpio: dwapb: Rework support for 1 interrupt per port A GPIO Phil Edworthy
2018-05-23 9:12 ` Simon Horman
@ 2018-05-23 9:30 ` Linus Walleij
1 sibling, 0 replies; 4+ messages in thread
From: Linus Walleij @ 2018-05-23 9:30 UTC (permalink / raw)
To: Phil Edworthy
Cc: Andy Shevchenko, Hoan Tran, Lee Jones, Michel Pollet,
open list:GPIO SUBSYSTEM, Linux-Renesas,
linux-kernel@vger.kernel.org
On Wed, May 23, 2018 at 10:52 AM, Phil Edworthy
<phil.edworthy@renesas.com> wrote:
> Treat DT and ACPI the same as much as possible. Note that we can't use
> platform_get_irq() to get the DT interrupts as they are in the port
> sub-node and hence do not have an associated platform device.
>
> This also fixes a problem introduced with error checking when calling
> platform_get_irq().
>
> Signed-off-by: Phil Edworthy <phil.edworthy@renesas.com>
OK this patch applied!
Yours,
Linus Walleij
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2018-05-23 9:30 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-05-23 8:52 [PATCH] gpio: dwapb: Rework support for 1 interrupt per port A GPIO Phil Edworthy
2018-05-23 9:12 ` Simon Horman
2018-05-23 9:18 ` Phil Edworthy
2018-05-23 9:30 ` Linus Walleij
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).