Re: [PATCH 2/2] ASoC: max98927: Add reset-gpio support

Re: [PATCH 2/2] ASoC: max98927: Add reset-gpio support

[Philipp Zabel]

Hi Cheng-yi,

[adding Maxime, devicetree to Cc:, the old discussion about GPIO resets
in [4] has never been resolved]

On Tue, 2018-10-09 at 21:46 +0800, Cheng-yi Chiang wrote:
> +reset controller maintainer Philipp
> 
> Hi Mark,
>   Sorry for the late reply. It took me a while to investigate reset
> controller and its possible usage. I would like to figure out the
> proper way of reset handling because it is common to have a shared
> reset line for two max98927 codecs for left and right channels.
> Without supporting this usage, a simple reset-gpios change for single
> codec would not be useful, and perhaps will be duplicated if reset
> controller is a better way.
> 
> Hi Philipp,
>   I would like to seek your advice about whether we can use reset
> controller to support the use case where multiple devices share the
> same reset line.
>
> Let me summarize the use case:
> There are two max98927 codecs sharing the same reset line.
> The reset line is controlled by a GPIO.
> The goal is to toggle reset line once and only once.
> 
> There was a similar scenario in tlv320aic3x codec driver [1].
> A static list is used in the codec driver so probe function can know
> whether it is needed to toggle reset line.
> Mark pointed out that it is not suitable to handle the shared reset
> line inside codec driver.
> A point is that this only works for multiple devices using the same
> device driver.
> He suggested to look into reset controller so I searched through the
> usage for common reset line.
>
> Here I found a shared reset line use case [2].
> With the patch [2], reset_control_reset can support this “reset once
> and for all” use case.

This patch was applied as 7da33a37b48f. So the reset framework has
support for shared reset controls where only the first reset request
actually causes a reset pulse, and all others are no-ops.

> However, I found that there are some missing pieces:
> 
> Let’s assume there are two codecs c1 and c2 sharing a reset line
> controlled by GPIO.
> 
> c1’s spec:
> Hold time: The minimum time to assert the reset line is t_a1.
> Release time: The minimum time to access the chip after deassert of
> the reset line is t_d1.
> 
> c2’s spec:
> Hold time: The minimum time to assert the reset line is t_a2.
> Release time: The minimum time to access the chip after deassert of
> the reset line is t_d2.
> 
> For both c1 and c2 to work properly, we need a reset controller that
> can assert the reset line for
> T = max(t_a1, t_a2).
> 
> 1. We need reset controller to support GPIO.

I still don't like the idea of describing a separate gpio reset
controller "device" in DT very much, as this is really just a software
abstraction, not actual hardware. For example:

	gpio_reset: reset-controller {
		compatible = "gpio-reset";
		reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>,
			      <&gpio1
16 GPIO_ACTIVE_HIGH>;
		reset-delays-us = <10000>, <500>;
	};

	c1 {
		resets = <&gpio_reset 0>; /* maps to <&gpio0 15 ...> */
	};

	c2 {
		resets = <&gpio_reset 0>;
	};

What I would like better would be to let the consumers keep their reset-
gpios bindings, but add an optional hold time override in the DT:

	c1 {
		reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>;
		reset-delays-us = <10000>;
	};

	c2 {
		reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>;
		re
set-delays-us = <10000>;
	};

The reset framework could create a reset_control backed by a gpio
instead of a rcdev. I have an unfinished patch for this, but without the
sharing requirement adding the reset framework abstraction between gpiod
and drivers never seemed really worth it.

> 2. We need to specify hold time T to reset controller in device tree
> so it knows that it needs hold reset line for that long in its
> implementation of reset_control_reset.

Agreed. Ideally, I'd like to make this optional, and have the drivers
continue to provide the hold time if they think they know it.

> 3. Assuming we have 1 and 2 in place. In codec driver of c1, it can
> call reset_control_reset and wait for t_a1 + t_d1. In codec driver of
> c2, it can call reset_control_reset and wait for t_a2 + t_d2.

The reset framework should wait for the configured assertion time,
max(t_a1, t_a2). The drivers only should only have to wait for
t_d1 / t_d2 after reset_control_reset returns.

>     We need to wait for hold time + release time because
> triggered_count is increased before reset ops is called. When the
> second driver finds that triggered_count is 1 and skip the real reset
> operation, reset ops might just begin doing its work a short time ago.

That is a good point. Maybe the reset framework should just wait for the
hold time even for the second reset. Another possibility would be to
serialize them with a mutex.

>     I am not sure whether we would need a flag in reset controller to
> mark that "reset is done". When driver sees this flag is done, it can
> just wait for release time instead of hold time + release time.

Let's not complicate the drivers with this too much. I think
reset_control_reset should guarantee that the reset line is not asserted
anymore upon return.

> And I found that you already solved 1 and mentioned the
> possible usage of 2 in [3].
> There were discussion about letting device driver to deal with
> reset-gpios by itself in [4], but it seems that reset controller can
> better deal with the shared reset line situation.
> Maybe we could revive the patch of GPIO support for reset controller ?

The discussion with Maxime in [4] hasn't really been resolved. I still
think the reset-gpios property should be kept, and the reset framework
could adapt to it. This is something the device tree maintainers' input
would be welcome on.

> Please let me know what direction I should look into.
> Thanks a lot!
> 
> [1] http://mailman.alsa-project.org/pipermail/alsa-devel/2010-November/033246.html
> https://patchwork.kernel.org/patch/9424123/
> 
> [2] https://patchwork.kernel.org/patch/9424123/
> 
> [3] https://lore.kernel.org/patchwork/patch/455839/
> 
> [4] https://patchwork.kernel.org/patch/3978121/

regards
Philipp

Re: [PATCH 2/2] ASoC: max98927: Add reset-gpio support

[Maxime Ripard]

On Fri, Oct 12, 2018 at 12:05:16PM +0200, Philipp Zabel wrote:
> Hi Cheng-yi,
> 
> [adding Maxime, devicetree to Cc:, the old discussion about GPIO resets
> in [4] has never been resolved]
> 
> On Tue, 2018-10-09 at 21:46 +0800, Cheng-yi Chiang wrote:
> > +reset controller maintainer Philipp
> > 
> > Hi Mark,
> >   Sorry for the late reply. It took me a while to investigate reset
> > controller and its possible usage. I would like to figure out the
> > proper way of reset handling because it is common to have a shared
> > reset line for two max98927 codecs for left and right channels.
> > Without supporting this usage, a simple reset-gpios change for single
> > codec would not be useful, and perhaps will be duplicated if reset
> > controller is a better way.
> > 
> > Hi Philipp,
> >   I would like to seek your advice about whether we can use reset
> > controller to support the use case where multiple devices share the
> > same reset line.
> >
> > Let me summarize the use case:
> > There are two max98927 codecs sharing the same reset line.
> > The reset line is controlled by a GPIO.
> > The goal is to toggle reset line once and only once.
> > 
> > There was a similar scenario in tlv320aic3x codec driver [1].
> > A static list is used in the codec driver so probe function can know
> > whether it is needed to toggle reset line.
> > Mark pointed out that it is not suitable to handle the shared reset
> > line inside codec driver.
> > A point is that this only works for multiple devices using the same
> > device driver.
> > He suggested to look into reset controller so I searched through the
> > usage for common reset line.
> >
> > Here I found a shared reset line use case [2].
> > With the patch [2], reset_control_reset can support this “reset once
> > and for all” use case.
> 
> This patch was applied as 7da33a37b48f. So the reset framework has
> support for shared reset controls where only the first reset request
> actually causes a reset pulse, and all others are no-ops.
> 
> > However, I found that there are some missing pieces:
> > 
> > Let’s assume there are two codecs c1 and c2 sharing a reset line
> > controlled by GPIO.
> > 
> > c1’s spec:
> > Hold time: The minimum time to assert the reset line is t_a1.
> > Release time: The minimum time to access the chip after deassert of
> > the reset line is t_d1.
> > 
> > c2’s spec:
> > Hold time: The minimum time to assert the reset line is t_a2.
> > Release time: The minimum time to access the chip after deassert of
> > the reset line is t_d2.
> > 
> > For both c1 and c2 to work properly, we need a reset controller that
> > can assert the reset line for
> > T = max(t_a1, t_a2).
> > 
> > 1. We need reset controller to support GPIO.
> 
> I still don't like the idea of describing a separate gpio reset
> controller "device" in DT very much, as this is really just a software
> abstraction, not actual hardware. For example:
> 
> 	gpio_reset: reset-controller {
> 		compatible = "gpio-reset";
> 		reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>,
> 			      <&gpio1
> 16 GPIO_ACTIVE_HIGH>;
> 		reset-delays-us = <10000>, <500>;
> 	};
> 
> 	c1 {
> 		resets = <&gpio_reset 0>; /* maps to <&gpio0 15 ...> */
> 	};
> 
> 	c2 {
> 		resets = <&gpio_reset 0>;
> 	};
> 
> What I would like better would be to let the consumers keep their reset-
> gpios bindings, but add an optional hold time override in the DT:
> 
> 	c1 {
> 		reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>;
> 		reset-delays-us = <10000>;
> 	};
> 
> 	c2 {
> 		reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>;
> 		reset-delays-us = <10000>;
> 	};
> 
> The reset framework could create a reset_control backed by a gpio
> instead of a rcdev. I have an unfinished patch for this, but without the
> sharing requirement adding the reset framework abstraction between gpiod
> and drivers never seemed really worth it.

I don't remember the exact details of our past discussion, but I
(still) don't really think that this would work well. I see two main
shortcomings with that approach:

  - I guess that the main reason you want to do that would be to have
    easy DT backward compatibility. Yet, the addition of the
    reset-delay-us would break that compatibility, since drivers that
    would have been converted now need to have it somehow, but older
    DTs wouldn't have it

  - There's so many variations of the reset-gpios property name that
    you wouldn't be able to cover all the cases anyhow, at least
    without breaking the compatibility (again).

But I also see your point, and you're right that converting everyone
to a gpio-reset node will not happen (even though I'd still really
like to have that binding).

What about having a function that would be called by the consumer to
instantiate that reset controller from a GPIO for us, instead of
trying to do it automatically? That function could take the property
name that holds the GPIO, which would cover the second drawback, and
the delay, which would cover the first.

And in order to cover shared GPIO reset lines, we could just look at
the one already created by other drivers, and if one has been created,
we just give the reference to that one instead of creating it.

Does that make sense?

Maxime

-- 
Maxime Ripard, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

Re: [PATCH 2/2] ASoC: max98927: Add reset-gpio support

[Philipp Zabel]

Hi Maxime,

On Fri, 2018-10-12 at 15:46 +0200, Maxime Ripard wrote:
> On Fri, Oct 12, 2018 at 12:05:16PM +0200, Philipp Zabel wrote:
[...]
> > What I would like better would be to let the consumers keep their reset-
> > gpios bindings, but add an optional hold time override in the DT:
> > 
> > 	c1 {
> > 		reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>;
> > 		reset-delays-us = <10000>;
> > 	};
> > 
> > 	c2 {
> > 		reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>;
> > 		reset-delays-us = <10000>;
> > 	};
> > 
> > The reset framework could create a reset_control backed by a gpio
> > instead of a rcdev. I have an unfinished patch for this, but without the
> > sharing requirement adding the reset framework abstraction between gpiod
> > and drivers never seemed really worth it.
> 
> I don't remember the exact details of our past discussion, but I
> (still) don't really think that this would work well.

It has been a while :) Thanks for jumping back in.

> I see two main shortcomings with that approach:
> 
>   - I guess that the main reason you want to do that would be to have
>     easy DT backward compatibility.

Yes, that is true. The other reason is that I'd like devices to have a
single binding, regardless of whether somebody decided to put them onto
a board with shared reset lines.

I'd find it hard to advocate for changing the thankfully common case
of device-exclusive reset gpios from:

	some-device {
		reset-gpios = <&gpiox y>;
	};

to:

	rstc: reset-controller {
		compatible = "gpio-reset";
		reset-gpios <&gpiox y>;
	};

	some-device {
		resets = <&rstc 0>;
	};

even for new bindings.

If the reset framework only supports the latter, and not the former,
drivers for devices which already have reset-gpios would have to handle
both reset_control and gpiod functionality, which I think should not be
necessary.

Note that this is not really an argument against a "gpio-reset"
controller driver, but an argument for reset-gpios support integrated
into the reset framework.
My argument against a "gpio-reset" device node in the DT is only that it
is basically a virtual device that would only be used to work around
missing reset-gpios support in the reset framework. Physically, this
"device" consists of no more than a few PCB traces.

>     Yet, the addition of the
>     reset-delay-us would break that compatibility, since drivers that
>     would have been converted now need to have it somehow, but older
>     DTs wouldn't have it

That is why such a property would have to be optional, and the drivers
would have to keep providing the reset delay themselves, as they
currently do when handling GPIOs directly.
It would only serve to override the driver default in case of additional
delay requirements due to board specifics (such as delay elements, or
shared resets).

>   - There's so many variations of the reset-gpios property name that
>     you wouldn't be able to cover all the cases anyhow, at least
>     without breaking the compatibility (again).

This is true, and I do not have a solution for this. Especially for
those cases that don't use gpiod yet.

	of_get_named_gpio(node, "reset-gpio")

is still quite common, but I'd really like on gpiolib for polarity
support.

> But I also see your point, and you're right that converting everyone
> to a gpio-reset node will not happen (even though I'd still really
> like to have that binding).

See above. I'd be a lot less reluctant to support this binding if
somebody could demonstrate a real gpio controlled reset controller
device of some kind. And even then I would not want to have to use
this device just to connect a single GPIO with a single reset input.

> What about having a function that would be called by the consumer to
> instantiate that reset controller from a GPIO for us, instead of
> trying to do it automatically? That function could take the property
> name that holds the GPIO, which would cover the second drawback, and
> the delay, which would cover the first.

I like this idea.

I'd like to avoid having to fall back from gpiod to of_get_named_gpio if
possible, so maybe we'd have to extend gpiolib-of.c with an
of_find_reset_gpio function, as is done for SPI and regulators already.

We probably would have to support delay ranges. I see a lot of
usleep_range calls between reset GPIO toggles in the drivers.

> And in order to cover shared GPIO reset lines, we could just look at
> the one already created by other drivers, and if one has been created,
> we just give the reference to that one instead of creating it.
> 
> Does that make sense?

I'm not quite sure how to match an already requested reset control from
the list given of_node and gpio property name at the moment - this might
involve exporting gpiod_find from gpiolib, but API wise I think this is
a sane proposal.

I would not want to add reset controller devices for each of these GPIO
reset controls, but rather store the gpio_desc reference in the
reset_control instead of the rcdev reference.

regards
Philipp

Re: [PATCH 2/2] ASoC: max98927: Add reset-gpio support

[Cheng-yi Chiang]

Hi Philipp,

On Thu, Oct 18, 2018 at 1:02 AM Philipp Zabel <p.zabel@pengutronix.de> wrote:
>
> Hi Maxime,
>
> On Fri, 2018-10-12 at 15:46 +0200, Maxime Ripard wrote:
> > On Fri, Oct 12, 2018 at 12:05:16PM +0200, Philipp Zabel wrote:
> [...]
> > > What I would like better would be to let the consumers keep their reset-
> > > gpios bindings, but add an optional hold time override in the DT:
> > >
> > >     c1 {
> > >             reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>;
> > >             reset-delays-us = <10000>;
> > >     };
> > >
> > >     c2 {
> > >             reset-gpios = <&gpio0 15 GPIO_ACTIVE_LOW>;
> > >             reset-delays-us = <10000>;
> > >     };
> > >
> > > The reset framework could create a reset_control backed by a gpio
> > > instead of a rcdev. I have an unfinished patch for this, but without the
> > > sharing requirement adding the reset framework abstraction between gpiod
> > > and drivers never seemed really worth it.
> >
> > I don't remember the exact details of our past discussion, but I
> > (still) don't really think that this would work well.
>
> It has been a while :) Thanks for jumping back in.
>
> > I see two main shortcomings with that approach:
> >
> >   - I guess that the main reason you want to do that would be to have
> >     easy DT backward compatibility.
>
> Yes, that is true. The other reason is that I'd like devices to have a
> single binding, regardless of whether somebody decided to put them onto
> a board with shared reset lines.
>
> I'd find it hard to advocate for changing the thankfully common case
> of device-exclusive reset gpios from:
>
>         some-device {
>                 reset-gpios = <&gpiox y>;
>         };
>
> to:
>
>         rstc: reset-controller {
>                 compatible = "gpio-reset";
>                 reset-gpios <&gpiox y>;
>         };
>
>         some-device {
>                 resets = <&rstc 0>;
>         };
>
> even for new bindings.
>
> If the reset framework only supports the latter, and not the former,
> drivers for devices which already have reset-gpios would have to handle
> both reset_control and gpiod functionality, which I think should not be
> necessary.
>
> Note that this is not really an argument against a "gpio-reset"
> controller driver, but an argument for reset-gpios support integrated
> into the reset framework.
> My argument against a "gpio-reset" device node in the DT is only that it
> is basically a virtual device that would only be used to work around
> missing reset-gpios support in the reset framework. Physically, this
> "device" consists of no more than a few PCB traces.
>
> >     Yet, the addition of the
> >     reset-delay-us would break that compatibility, since drivers that
> >     would have been converted now need to have it somehow, but older
> >     DTs wouldn't have it
>
> That is why such a property would have to be optional, and the drivers
> would have to keep providing the reset delay themselves, as they
> currently do when handling GPIOs directly.
> It would only serve to override the driver default in case of additional
> delay requirements due to board specifics (such as delay elements, or
> shared resets).
>
> >   - There's so many variations of the reset-gpios property name that
> >     you wouldn't be able to cover all the cases anyhow, at least
> >     without breaking the compatibility (again).
>
> This is true, and I do not have a solution for this. Especially for
> those cases that don't use gpiod yet.
>
>         of_get_named_gpio(node, "reset-gpio")
>
> is still quite common, but I'd really like on gpiolib for polarity
> support.
>
> > But I also see your point, and you're right that converting everyone
> > to a gpio-reset node will not happen (even though I'd still really
> > like to have that binding).
>
> See above. I'd be a lot less reluctant to support this binding if
> somebody could demonstrate a real gpio controlled reset controller
> device of some kind. And even then I would not want to have to use
> this device just to connect a single GPIO with a single reset input.
>
> > What about having a function that would be called by the consumer to
> > instantiate that reset controller from a GPIO for us, instead of
> > trying to do it automatically? That function could take the property
> > name that holds the GPIO, which would cover the second drawback, and
> > the delay, which would cover the first.
>
> I like this idea.
>
> I'd like to avoid having to fall back from gpiod to of_get_named_gpio if
> possible, so maybe we'd have to extend gpiolib-of.c with an
> of_find_reset_gpio function, as is done for SPI and regulators already.
>
> We probably would have to support delay ranges. I see a lot of
> usleep_range calls between reset GPIO toggles in the drivers.
>

This looks great.

I am not sure what is the plan proceeding from here.
Are you going to implement this feature with the unfinished patch you
mentioned to create reset_control backed by a GPIO?
I am willing to help.
If you have patch I can help testing it.
It will be easier to test the shared reset line on my side since the
board I am working now has this use case.

If you don't have time for this, I can work on it based on your WIP patch.
In that case could you please post the patch ?
I just want to make sure I am on the right direction solving this
without making duplicate efforts..

> > And in order to cover shared GPIO reset lines, we could just look at
> > the one already created by other drivers, and if one has been created,
> > we just give the reference to that one instead of creating it.
> >
> > Does that make sense?
>
> I'm not quite sure how to match an already requested reset control from
> the list given of_node and gpio property name at the moment - this might
> involve exporting gpiod_find from gpiolib, but API wise I think this is
> a sane proposal.
>
> I would not want to add reset controller devices for each of these GPIO
> reset controls, but rather store the gpio_desc reference in the
> reset_control instead of the rcdev reference.

This plan to support shared reset line makes sense.

Thanks a lot!

>
> regards
> Philipp