Currently, the GPIO interface is said to Open Drain if it is Single
Ended and active LOW. Similarly, it is said as Open Source if it is
Single Ended and active HIGH.
The active HIGH/LOW is used in the interface for setting the pin
state to HIGH or LOW when enabling/disabling the interface.
In Open Drain interface, pin is set to HIGH by putting pin in
high impedance and LOW by driving to the LOW.
In Open Source interface, pin is set to HIGH by driving pin to
HIGH and set to LOW by putting pin in high impedance.
With above, the Open Drain/Source is unrelated to the active LOW/HIGH
in interface. There is interface where the enable/disable of interface
is ether active LOW or HIGH but it is Open Drain type.
Hence decouple the Open Drain with Single Ended + Active LOW and
Open Source with Single Ended + Active HIGH.
Adding different flag for the Open Drain/Open Source which is valid
only when Single ended flag is enabled.
Signed-off-by: Laxman Dewangan <[email protected]>
---
drivers/gpio/gpiolib-of.c | 2 +-
drivers/gpio/gpiolib.c | 4 +++-
include/dt-bindings/gpio/gpio.h | 12 ++++++++----
include/linux/of_gpio.h | 1 +
4 files changed, 13 insertions(+), 6 deletions(-)
diff --git a/drivers/gpio/gpiolib-of.c b/drivers/gpio/gpiolib-of.c
index 975b9f6..b13b7c7 100644
--- a/drivers/gpio/gpiolib-of.c
+++ b/drivers/gpio/gpiolib-of.c
@@ -147,7 +147,7 @@ struct gpio_desc *of_find_gpio(struct device *dev, const char *con_id,
*flags |= GPIO_ACTIVE_LOW;
if (of_flags & OF_GPIO_SINGLE_ENDED) {
- if (of_flags & OF_GPIO_ACTIVE_LOW)
+ if (of_flags & OF_GPIO_OPEN_DRAIN)
*flags |= GPIO_OPEN_DRAIN;
else
*flags |= GPIO_OPEN_SOURCE;
diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index 4aa1e78..c22e572 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -3333,6 +3333,7 @@ struct gpio_desc *fwnode_get_named_gpiod(struct fwnode_handle *fwnode,
unsigned long lflags = 0;
bool active_low = false;
bool single_ended = false;
+ bool open_drain = false;
int ret;
if (!fwnode)
@@ -3346,6 +3347,7 @@ struct gpio_desc *fwnode_get_named_gpiod(struct fwnode_handle *fwnode,
if (!IS_ERR(desc)) {
active_low = flags & OF_GPIO_ACTIVE_LOW;
single_ended = flags & OF_GPIO_SINGLE_ENDED;
+ open_drain = flags & OF_GPIO_OPEN_DRAIN;
}
} else if (is_acpi_node(fwnode)) {
struct acpi_gpio_info info;
@@ -3366,7 +3368,7 @@ struct gpio_desc *fwnode_get_named_gpiod(struct fwnode_handle *fwnode,
lflags |= GPIO_ACTIVE_LOW;
if (single_ended) {
- if (active_low)
+ if (open_drain)
lflags |= GPIO_OPEN_DRAIN;
else
lflags |= GPIO_OPEN_SOURCE;
diff --git a/include/dt-bindings/gpio/gpio.h b/include/dt-bindings/gpio/gpio.h
index c673d2c..b4f54da 100644
--- a/include/dt-bindings/gpio/gpio.h
+++ b/include/dt-bindings/gpio/gpio.h
@@ -17,11 +17,15 @@
#define GPIO_PUSH_PULL 0
#define GPIO_SINGLE_ENDED 2
+/* Bit 2 express Open drain or open source */
+#define GPIO_LINE_OPEN_SOURCE 0
+#define GPIO_LINE_OPEN_DRAIN 4
+
/*
- * Open Drain/Collector is the combination of single-ended active low,
- * Open Source/Emitter is the combination of single-ended active high.
+ * Open Drain/Collector is the combination of single-ended open drain interface.
+ * Open Source/Emitter is the combination of single-ended open source interface.
*/
-#define GPIO_OPEN_DRAIN (GPIO_SINGLE_ENDED | GPIO_ACTIVE_LOW)
-#define GPIO_OPEN_SOURCE (GPIO_SINGLE_ENDED | GPIO_ACTIVE_HIGH)
+#define GPIO_OPEN_DRAIN (GPIO_SINGLE_ENDED | GPIO_LINE_OPEN_DRAIN)
+#define GPIO_OPEN_SOURCE (GPIO_SINGLE_ENDED | GPIO_LINE_OPEN_SOURCE)
#endif
diff --git a/include/linux/of_gpio.h b/include/linux/of_gpio.h
index 3f87ea5..1e089d5 100644
--- a/include/linux/of_gpio.h
+++ b/include/linux/of_gpio.h
@@ -30,6 +30,7 @@ struct device_node;
enum of_gpio_flags {
OF_GPIO_ACTIVE_LOW = 0x1,
OF_GPIO_SINGLE_ENDED = 0x2,
+ OF_GPIO_OPEN_DRAIN = 0x4,
};
#ifdef CONFIG_OF_GPIO
--
2.1.4
On Thu, Apr 6, 2017 at 4:35 PM, Laxman Dewangan <[email protected]> wrote:
> Currently, the GPIO interface is said to Open Drain if it is Single
> Ended and active LOW. Similarly, it is said as Open Source if it is
> Single Ended and active HIGH.
>
> The active HIGH/LOW is used in the interface for setting the pin
> state to HIGH or LOW when enabling/disabling the interface.
>
> In Open Drain interface, pin is set to HIGH by putting pin in
> high impedance and LOW by driving to the LOW.
>
> In Open Source interface, pin is set to HIGH by driving pin to
> HIGH and set to LOW by putting pin in high impedance.
>
> With above, the Open Drain/Source is unrelated to the active LOW/HIGH
> in interface. There is interface where the enable/disable of interface
> is ether active LOW or HIGH but it is Open Drain type.
>
> Hence decouple the Open Drain with Single Ended + Active LOW and
> Open Source with Single Ended + Active HIGH.
>
> Adding different flag for the Open Drain/Open Source which is valid
> only when Single ended flag is enabled.
> if (single_ended) {
> - if (active_low)
> + if (open_drain)
This breaks ACPI case, right?
--
With Best Regards,
Andy Shevchenko
On Thursday 06 April 2017 09:40 PM, Andy Shevchenko wrote:
> On Thu, Apr 6, 2017 at 4:35 PM, Laxman Dewangan <[email protected]> wrote:
>> Currently, the GPIO interface is said to Open Drain if it is Single
>> Ended and active LOW. Similarly, it is said as Open Source if it is
>> Single Ended and active HIGH.
>>
>> The active HIGH/LOW is used in the interface for setting the pin
>> state to HIGH or LOW when enabling/disabling the interface.
>>
>> In Open Drain interface, pin is set to HIGH by putting pin in
>> high impedance and LOW by driving to the LOW.
>>
>> In Open Source interface, pin is set to HIGH by driving pin to
>> HIGH and set to LOW by putting pin in high impedance.
>>
>> With above, the Open Drain/Source is unrelated to the active LOW/HIGH
>> in interface. There is interface where the enable/disable of interface
>> is ether active LOW or HIGH but it is Open Drain type.
>>
>> Hence decouple the Open Drain with Single Ended + Active LOW and
>> Open Source with Single Ended + Active HIGH.
>>
>> Adding different flag for the Open Drain/Open Source which is valid
>> only when Single ended flag is enabled.
>> if (single_ended) {
>> - if (active_low)
>> + if (open_drain)
> This breaks ACPI case, right?
>
In acpi case, single_ended is not handled. It only handles the active LOW.
From code:
bool active_low = false;
bool single_ended = false;
int ret;
if (!fwnode)
return ERR_PTR(-EINVAL);
if (is_of_node(fwnode)) {
enum of_gpio_flags flags;
desc = of_get_named_gpiod_flags(to_of_node(fwnode),
propname,
index, &flags);
if (!IS_ERR(desc)) {
active_low = flags & OF_GPIO_ACTIVE_LOW;
single_ended = flags & OF_GPIO_SINGLE_ENDED;
}
} else if (is_acpi_node(fwnode)) {
struct acpi_gpio_info info;
desc = acpi_node_get_gpiod(fwnode, propname, index, &info);
if (!IS_ERR(desc))
active_low = info.polarity == GPIO_ACTIVE_LOW;
}
On Thu, Apr 6, 2017 at 7:56 PM, Laxman Dewangan <[email protected]> wrote:
> On Thursday 06 April 2017 09:40 PM, Andy Shevchenko wrote:
>> On Thu, Apr 6, 2017 at 4:35 PM, Laxman Dewangan <[email protected]>
>> wrote:
>>> Adding different flag for the Open Drain/Open Source which is valid
>>> only when Single ended flag is enabled.
>>> if (single_ended) {
>>> - if (active_low)
>>> + if (open_drain)
>>
>> This breaks ACPI case, right?
> In acpi case, single_ended is not handled. It only handles the active LOW.
>
> From code:
Fair enough.
I would look into ACPI spec for possibility to add this kind of functionality.
Thanks.
--
With Best Regards,
Andy Shevchenko
On Thu, Apr 6, 2017 at 3:35 PM, Laxman Dewangan <[email protected]> wrote:
> Currently, the GPIO interface is said to Open Drain if it is Single
> Ended and active LOW. Similarly, it is said as Open Source if it is
> Single Ended and active HIGH.
>
> The active HIGH/LOW is used in the interface for setting the pin
> state to HIGH or LOW when enabling/disabling the interface.
>
> In Open Drain interface, pin is set to HIGH by putting pin in
> high impedance and LOW by driving to the LOW.
>
> In Open Source interface, pin is set to HIGH by driving pin to
> HIGH and set to LOW by putting pin in high impedance.
>
> With above, the Open Drain/Source is unrelated to the active LOW/HIGH
> in interface. There is interface where the enable/disable of interface
> is ether active LOW or HIGH but it is Open Drain type.
>
> Hence decouple the Open Drain with Single Ended + Active LOW and
> Open Source with Single Ended + Active HIGH.
>
> Adding different flag for the Open Drain/Open Source which is valid
> only when Single ended flag is enabled.
>
> Signed-off-by: Laxman Dewangan <[email protected]>
Patch applied.
Good that you found this and fixed it before someone git hurt.
Sorry for screwing it up, my fault.
Yours,
Linus Walleij
On Friday 07 April 2017 03:55 PM, Linus Walleij wrote:
> On Thu, Apr 6, 2017 at 3:35 PM, Laxman Dewangan <[email protected]> wrote:
>
>> Currently, the GPIO interface is said to Open Drain if it is Single
>> Ended and active LOW. Similarly, it is said as Open Source if it is
>> Single Ended and active HIGH.
>>
>> The active HIGH/LOW is used in the interface for setting the pin
>> state to HIGH or LOW when enabling/disabling the interface.
>>
>> In Open Drain interface, pin is set to HIGH by putting pin in
>> high impedance and LOW by driving to the LOW.
>>
>> In Open Source interface, pin is set to HIGH by driving pin to
>> HIGH and set to LOW by putting pin in high impedance.
>>
>> With above, the Open Drain/Source is unrelated to the active LOW/HIGH
>> in interface. There is interface where the enable/disable of interface
>> is ether active LOW or HIGH but it is Open Drain type.
>>
>> Hence decouple the Open Drain with Single Ended + Active LOW and
>> Open Source with Single Ended + Active HIGH.
>>
>> Adding different flag for the Open Drain/Open Source which is valid
>> only when Single ended flag is enabled.
>>
>> Signed-off-by: Laxman Dewangan <[email protected]>
> Patch applied.
>
> Good that you found this and fixed it before someone git hurt.
Thanks you very much.
Now we are going to use the descriptor based GPIO APIs as this is very
powerful and passing this info from DT is very useful.
On Fri, Apr 07, 2017 at 12:25:49PM +0200, Linus Walleij wrote:
> On Thu, Apr 6, 2017 at 3:35 PM, Laxman Dewangan <[email protected]> wrote:
>
> > Currently, the GPIO interface is said to Open Drain if it is Single
> > Ended and active LOW. Similarly, it is said as Open Source if it is
> > Single Ended and active HIGH.
> >
> > The active HIGH/LOW is used in the interface for setting the pin
> > state to HIGH or LOW when enabling/disabling the interface.
> >
> > In Open Drain interface, pin is set to HIGH by putting pin in
> > high impedance and LOW by driving to the LOW.
> >
> > In Open Source interface, pin is set to HIGH by driving pin to
> > HIGH and set to LOW by putting pin in high impedance.
> >
> > With above, the Open Drain/Source is unrelated to the active LOW/HIGH
> > in interface. There is interface where the enable/disable of interface
> > is ether active LOW or HIGH but it is Open Drain type.
> >
> > Hence decouple the Open Drain with Single Ended + Active LOW and
> > Open Source with Single Ended + Active HIGH.
> >
> > Adding different flag for the Open Drain/Open Source which is valid
> > only when Single ended flag is enabled.
> >
> > Signed-off-by: Laxman Dewangan <[email protected]>
>
> Patch applied.
>
> Good that you found this and fixed it before someone git hurt.
Well, while decoupling single-endedness from polarity was the right
thing to do, this change did actually break the DT binary interface.
If you have an old compiled dtb whose source used GPIO_OPEN_DRAIN, you
now instead get *open-source* behaviour on 4.12:
GPIO_OPEN_DRAIN = GPIO_SINGLE_ENDED | GPIO_ACTIVE_LOW
=> active-low, but *open source*
while if you recompile that source against 4.12 you do get the expected
open-drain behaviour, but now with inverted polarity:
GPIO_OPEN_DRAIN = GPIO_SINGLE_ENDED | GPIO_LINE_OPEN_DRAIN
=> open drain, but *active high*
requiring the device tree to be updated by specifying
(GPIO_OPEN_DRAIN | GPIO_ACTIVE_LOW)
I guess the latter is fine, even if it is likely to amount to a fair bit
of debugging world wide.
Perhaps all this can still be avoided by adding further flags and
deprecating others before people start migrating to 4.12 (after all,
GPIO_OPEN_DRAIN has been around since 4.4 even if there are no in-kernel
users).
Or we accept the binary interface breakage -- it probably is pretty rare
that people update the kernel without updating the dtb. I can just
update the dts on the system that broke for me, and hopefully anyone
debugging this issue while updating to 4.12 will find this mail quickly.
Johan
On Wednesday 19 July 2017 06:55 PM, Johan Hovold wrote:
> On Fri, Apr 07, 2017 at 12:25:49PM +0200, Linus Walleij wrote:
>> On Thu, Apr 6, 2017 at 3:35 PM, Laxman Dewangan <[email protected]> wrote:
>>
>>> Currently, the GPIO interface is said to Open Drain if it is Single
>>> Ended and active LOW. Similarly, it is said as Open Source if it is
>>> Single Ended and active HIGH.
>>>
>>> The active HIGH/LOW is used in the interface for setting the pin
>>> state to HIGH or LOW when enabling/disabling the interface.
>>>
>>> In Open Drain interface, pin is set to HIGH by putting pin in
>>> high impedance and LOW by driving to the LOW.
>>>
>>> In Open Source interface, pin is set to HIGH by driving pin to
>>> HIGH and set to LOW by putting pin in high impedance.
>>>
>>> With above, the Open Drain/Source is unrelated to the active LOW/HIGH
>>> in interface. There is interface where the enable/disable of interface
>>> is ether active LOW or HIGH but it is Open Drain type.
>>>
>>> Hence decouple the Open Drain with Single Ended + Active LOW and
>>> Open Source with Single Ended + Active HIGH.
>>>
>>> Adding different flag for the Open Drain/Open Source which is valid
>>> only when Single ended flag is enabled.
>>>
>>> Signed-off-by: Laxman Dewangan <[email protected]>
>> Patch applied.
>>
>> Good that you found this and fixed it before someone git hurt.
> Well, while decoupling single-endedness from polarity was the right
> thing to do, this change did actually break the DT binary interface.
>
> If you have an old compiled dtb whose source used GPIO_OPEN_DRAIN, you
> now instead get *open-source* behaviour on 4.12:
>
> GPIO_OPEN_DRAIN = GPIO_SINGLE_ENDED | GPIO_ACTIVE_LOW
>
> => active-low, but *open source*
>
> while if you recompile that source against 4.12 you do get the expected
> open-drain behaviour, but now with inverted polarity:
>
> GPIO_OPEN_DRAIN = GPIO_SINGLE_ENDED | GPIO_LINE_OPEN_DRAIN
>
> => open drain, but *active high*
>
> requiring the device tree to be updated by specifying
>
> (GPIO_OPEN_DRAIN | GPIO_ACTIVE_LOW)
>
> I guess the latter is fine, even if it is likely to amount to a fair bit
> of debugging world wide.
>
> Perhaps all this can still be avoided by adding further flags and
> deprecating others before people start migrating to 4.12 (after all,
> GPIO_OPEN_DRAIN has been around since 4.4 even if there are no in-kernel
> users).
>
> Or we accept the binary interface breakage -- it probably is pretty rare
> that people update the kernel without updating the dtb. I can just
> update the dts on the system that broke for me, and hopefully anyone
> debugging this issue while updating to 4.12 will find this mail quickly.
>
Yes, it breaks the older DTS with new kernel. However, this point was
discussed before sending patch. As there was no user in the mainline DTs
for these macros, we made change.
On Wed, Jul 19, 2017 at 08:29:08PM +0530, Laxman Dewangan wrote:
> >> Good that you found this and fixed it before someone git hurt.
> > Well, while decoupling single-endedness from polarity was the right
> > thing to do, this change did actually break the DT binary interface.
> >
> > If you have an old compiled dtb whose source used GPIO_OPEN_DRAIN, you
> > now instead get *open-source* behaviour on 4.12:
> >
> > GPIO_OPEN_DRAIN = GPIO_SINGLE_ENDED | GPIO_ACTIVE_LOW
> >
> > => active-low, but *open source*
> >
> > while if you recompile that source against 4.12 you do get the expected
> > open-drain behaviour, but now with inverted polarity:
> >
> > GPIO_OPEN_DRAIN = GPIO_SINGLE_ENDED | GPIO_LINE_OPEN_DRAIN
> >
> > => open drain, but *active high*
> >
> > requiring the device tree to be updated by specifying
> >
> > (GPIO_OPEN_DRAIN | GPIO_ACTIVE_LOW)
> >
> > I guess the latter is fine, even if it is likely to amount to a fair bit
> > of debugging world wide.
> >
> > Perhaps all this can still be avoided by adding further flags and
> > deprecating others before people start migrating to 4.12 (after all,
> > GPIO_OPEN_DRAIN has been around since 4.4 even if there are no in-kernel
> > users).
> >
> > Or we accept the binary interface breakage -- it probably is pretty rare
> > that people update the kernel without updating the dtb. I can just
> > update the dts on the system that broke for me, and hopefully anyone
> > debugging this issue while updating to 4.12 will find this mail quickly.
> >
>
> Yes, it breaks the older DTS with new kernel. However, this point was
> discussed before sending patch. As there was no user in the mainline DTs
> for these macros, we made change.
These are generic gpio flags that can be used with a multitude of
devices (e.g. regulators, reset-signals for all sorts of ICs, etc.), and
whether there are any in-kernel users (dts) should probably not carry
much weight.
Some people are still stuck with 4.4 or 4.9 and it may still be a while
before they update to, say, the next LTS kernel (4.13) and get bitten by
this.
Thanks,
Johan
On Wed, Jul 19, 2017 at 4:59 PM, Laxman Dewangan <[email protected]> wrote:
> On Wednesday 19 July 2017 06:55 PM, Johan Hovold wrote:
>> I guess the latter is fine, even if it is likely to amount to a fair bit
>> of debugging world wide.
>>
>> Perhaps all this can still be avoided by adding further flags and
>> deprecating others before people start migrating to 4.12 (after all,
>> GPIO_OPEN_DRAIN has been around since 4.4 even if there are no in-kernel
>> users).
>>
>> Or we accept the binary interface breakage -- it probably is pretty rare
>> that people update the kernel without updating the dtb. I can just
>> update the dts on the system that broke for me, and hopefully anyone
>> debugging this issue while updating to 4.12 will find this mail quickly.
>
> Yes, it breaks the older DTS with new kernel. However, this point was
> discussed before sending patch. As there was no user in the mainline DTs for
> these macros, we made change.
My operating assumption is usually "rough consensus and running code".
If the code doesn't run, i.e. if there are regressions then we need to fix
them.
I am not aware of any systems having picked up the flags as they were
set before this patch, but if they exist we need to patch it of course.
On the other hand we don't patch theoretical compatibility issues either,
only those that occure in practice.
It falls back on when a DT binding is really standardized. When it is
starting to get discussed? When the final binding is merged to the kernel
tree? When the devicetree.org people publish it? When a big enough
vendor ships more than 100 devices using it no matter what happened
discussion-wise?
All these are a bit fluid concepts, so as a result, the handling of bindings
is a bit fluid.
If we really want DT bindings to have a point when they are "set in stone"
then we need to talk to devicetree.org about that. But in reality I think all
standards ever are a bit fuzzy around the edges.
Yours,
Linus Walleij
On Thu, Aug 03, 2017 at 10:14:32AM +0200, Linus Walleij wrote:
> On Wed, Jul 19, 2017 at 4:59 PM, Laxman Dewangan <[email protected]> wrote:
> > On Wednesday 19 July 2017 06:55 PM, Johan Hovold wrote:
>
> >> I guess the latter is fine, even if it is likely to amount to a fair bit
> >> of debugging world wide.
> >>
> >> Perhaps all this can still be avoided by adding further flags and
> >> deprecating others before people start migrating to 4.12 (after all,
> >> GPIO_OPEN_DRAIN has been around since 4.4 even if there are no in-kernel
> >> users).
> >>
> >> Or we accept the binary interface breakage -- it probably is pretty rare
> >> that people update the kernel without updating the dtb. I can just
> >> update the dts on the system that broke for me, and hopefully anyone
> >> debugging this issue while updating to 4.12 will find this mail quickly.
> >
> > Yes, it breaks the older DTS with new kernel. However, this point was
> > discussed before sending patch. As there was no user in the mainline DTs for
> > these macros, we made change.
>
> My operating assumption is usually "rough consensus and running code".
>
> If the code doesn't run, i.e. if there are regressions then we need to fix
> them.
>
> I am not aware of any systems having picked up the flags as they were
> set before this patch, but if they exist we need to patch it of course.
Eight kernel releases, including two long-term stable kernels (4.4 and
4.9), is a long time for people start using these flags and without us
finding out about it for a while still.
> On the other hand we don't patch theoretical compatibility issues either,
> only those that occure in practice.
I have a customer who was bitten by this after updating to 4.12, and I'm
sure there will be others.
So I wouldn't simply dismiss it as theoretical.
> It falls back on when a DT binding is really standardized. When it is
> starting to get discussed? When the final binding is merged to the kernel
> tree? When the devicetree.org people publish it? When a big enough
> vendor ships more than 100 devices using it no matter what happened
> discussion-wise?
>
> All these are a bit fluid concepts, so as a result, the handling of bindings
> is a bit fluid.
>
> If we really want DT bindings to have a point when they are "set in stone"
> then we need to talk to devicetree.org about that. But in reality I think all
> standards ever are a bit fuzzy around the edges.
I agree. And the interface needed to be fixed. But for next time,
perhaps being more explicit about the implications for current users
would help people notice such changes when they do happen. Backporting a
fix for a broken interface could also be considered (at least if
discovered in the next kernel release cycle or so).
And in this particular case, it looks like it might have been possible
to pick flags in such a way that the breakage could have be avoided in
the first place.
Johan