Hello,
This small patchset adds the reset functionality to the I2C and SPI
peripherals on HI6220 SoC from HiSilicon. Those peripherals needs to be
taken out of reset before being used. But earlier we were depending on the
bootloader to do the job but as suggested by Daniel Thompson, a more
cleaner approach would be to handle the reset in corresponding drivers.
Hence, one of the patch adds reset properties to the I2C and SPI nodes
and the other one adds missing reset functionality to the AMBA Primecell
bus driver. Because the AMBA devices are being accessed before the driver
probe (reading pid and cid values), we need to deassert the reset line in
the bus driver itself just before the read.
I'd expect the patch 1 to go via arm tree and patch 2 via hisi tree.
Thanks,
Mani
Manivannan Sadhasivam (2):
amba: Take device out of reset before reading pid and cid values
arm64: dts: hisilicon: Add reset properties for HI6220 I2C and SPI
arch/arm64/boot/dts/hisilicon/hi6220.dtsi | 8 ++++++++
drivers/amba/bus.c | 9 +++++++++
2 files changed, 17 insertions(+)
--
2.17.1
For the AMBA Primecell devices having the reset lines wired, it is
necessary to take them out of reset before reading the pid and cid values.
Earlier we were dependent on the bootloader to do this but a more cleaner
approach would be to do it in the kernel itself. Hence, this commit
deasserts the reset line just before reading the pid and cid values.
Suggested-by: Daniel Thompson <[email protected]>
Signed-off-by: Manivannan Sadhasivam <[email protected]>
---
drivers/amba/bus.c | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/drivers/amba/bus.c b/drivers/amba/bus.c
index 41b706403ef7..da8f1aac5315 100644
--- a/drivers/amba/bus.c
+++ b/drivers/amba/bus.c
@@ -21,6 +21,7 @@
#include <linux/limits.h>
#include <linux/clk/clk-conf.h>
#include <linux/platform_device.h>
+#include <linux/reset.h>
#include <asm/irq.h>
@@ -352,6 +353,7 @@ static void amba_device_release(struct device *dev)
static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
{
+ struct reset_control *rst;
u32 size;
void __iomem *tmp;
int i, ret;
@@ -388,6 +390,13 @@ static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
if (ret == 0) {
u32 pid, cid;
+ /* De-assert the reset line to take the device out of reset */
+ rst = reset_control_get_optional_exclusive(&dev->dev, NULL);
+ if (IS_ERR(rst))
+ return PTR_ERR(rst);
+
+ reset_control_deassert(rst);
+
/*
* Read pid and cid based on size of resource
* they are located at end of region
--
2.17.1
Both I2C and SPI needs to be taken out of reset before being used. Earlier,
we relied on the bootloader to do the job but a more cleaner approach would
be to handle the reset in kernel. Hence, add the reset properties to
the nodes to let the corresponding drivers take the peripherals out of
reset.
Suggested-by: Daniel Thompson <[email protected]>
Signed-off-by: Manivannan Sadhasivam <[email protected]>
---
arch/arm64/boot/dts/hisilicon/hi6220.dtsi | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/arch/arm64/boot/dts/hisilicon/hi6220.dtsi b/arch/arm64/boot/dts/hisilicon/hi6220.dtsi
index 108e2a4227f6..b071dc466374 100644
--- a/arch/arm64/boot/dts/hisilicon/hi6220.dtsi
+++ b/arch/arm64/boot/dts/hisilicon/hi6220.dtsi
@@ -730,6 +730,8 @@
pinctrl-0 = <&spi0_pmx_func &spi0_cfg_func>;
num-cs = <1>;
cs-gpios = <&gpio6 2 0>;
+ resets = <&sys_ctrl PERIPH_RSTEN3_SSP>;
+ reset-names = "reset";
status = "disabled";
};
@@ -741,6 +743,8 @@
i2c-sda-hold-time-ns = <300>;
pinctrl-names = "default";
pinctrl-0 = <&i2c0_pmx_func &i2c0_cfg_func>;
+ resets = <&sys_ctrl PERIPH_RSTEN3_I2C0>;
+ reset-names = "reset";
status = "disabled";
};
@@ -752,6 +756,8 @@
i2c-sda-hold-time-ns = <300>;
pinctrl-names = "default";
pinctrl-0 = <&i2c1_pmx_func &i2c1_cfg_func>;
+ resets = <&sys_ctrl PERIPH_RSTEN3_I2C1>;
+ reset-names = "reset";
status = "disabled";
};
@@ -763,6 +769,8 @@
i2c-sda-hold-time-ns = <300>;
pinctrl-names = "default";
pinctrl-0 = <&i2c2_pmx_func &i2c2_cfg_func>;
+ resets = <&sys_ctrl PERIPH_RSTEN3_I2C2>;
+ reset-names = "reset";
status = "disabled";
};
--
2.17.1
On Sat, Mar 9, 2019 at 2:59 AM Manivannan Sadhasivam
<[email protected]> wrote:
> For the AMBA Primecell devices having the reset lines wired, it is
> necessary to take them out of reset before reading the pid and cid values.
> Earlier we were dependent on the bootloader to do this but a more cleaner
> approach would be to do it in the kernel itself. Hence, this commit
> deasserts the reset line just before reading the pid and cid values.
>
> Suggested-by: Daniel Thompson <[email protected]>
> Signed-off-by: Manivannan Sadhasivam <[email protected]>
This looks correct to me.
Reviewed-by: Linus Walleij <[email protected]>
I think you can put it in Russell's patch tracker in some week.
Yours,
Linus Walleij
On Sat, Mar 09, 2019 at 07:26:34AM +0530, Manivannan Sadhasivam wrote:
> For the AMBA Primecell devices having the reset lines wired, it is
> necessary to take them out of reset before reading the pid and cid values.
> Earlier we were dependent on the bootloader to do this but a more cleaner
> approach would be to do it in the kernel itself. Hence, this commit
> deasserts the reset line just before reading the pid and cid values.
>
> Suggested-by: Daniel Thompson <[email protected]>
> Signed-off-by: Manivannan Sadhasivam <[email protected]>
> ---
> drivers/amba/bus.c | 9 +++++++++
> 1 file changed, 9 insertions(+)
>
> diff --git a/drivers/amba/bus.c b/drivers/amba/bus.c
> index 41b706403ef7..da8f1aac5315 100644
> --- a/drivers/amba/bus.c
> +++ b/drivers/amba/bus.c
> @@ -21,6 +21,7 @@
> #include <linux/limits.h>
> #include <linux/clk/clk-conf.h>
> #include <linux/platform_device.h>
> +#include <linux/reset.h>
>
> #include <asm/irq.h>
>
> @@ -352,6 +353,7 @@ static void amba_device_release(struct device *dev)
>
> static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> {
> + struct reset_control *rst;
> u32 size;
> void __iomem *tmp;
> int i, ret;
> @@ -388,6 +390,13 @@ static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> if (ret == 0) {
> u32 pid, cid;
>
> + /* De-assert the reset line to take the device out of reset */
> + rst = reset_control_get_optional_exclusive(&dev->dev, NULL);
> + if (IS_ERR(rst))
> + return PTR_ERR(rst);
It is really correct to propagate an error if we cannot get exclusive
ownership of the reset line.
With drivers for vendor specific cells it is ok to "just know" that the
reset line is never shared but we cannot know this for generic cells and
we certainly can't know this for the bus.
I think it *might* be OK to propagate an error if you used
reset_control_get_optional_shared() instead because if that reports an
error than arguably we have either a mistake in the DT or a bug in the
driver we are sharing a reset with.
> +
> + reset_control_deassert(rst);
Perhaps we might also need to explain why we can ignore -ENOTSUPP
here. Perhaps something like the following based on the comment
found in in reset_control_deassert():
/*
* -ENOTSUPP means occurs when the reset controller
* does not implement .deassert(), in which case the
* the reset lines should be self-deasserting (and
* deasserted by default).
*/
WARN_ON(deassert did not return 0 or -ENOTSUPP);
> +
> /*
> * Read pid and cid based on size of resource
> * they are located at end of region
This looks like it will leak the control reference... shouldn't there
be a put after you have read the pid/cid?
Daniel.
Hi Daniel,
On Tue, Mar 12, 2019 at 12:27:11PM +0000, Daniel Thompson wrote:
> On Sat, Mar 09, 2019 at 07:26:34AM +0530, Manivannan Sadhasivam wrote:
> > For the AMBA Primecell devices having the reset lines wired, it is
> > necessary to take them out of reset before reading the pid and cid values.
> > Earlier we were dependent on the bootloader to do this but a more cleaner
> > approach would be to do it in the kernel itself. Hence, this commit
> > deasserts the reset line just before reading the pid and cid values.
> >
> > Suggested-by: Daniel Thompson <[email protected]>
> > Signed-off-by: Manivannan Sadhasivam <[email protected]>
> > ---
> > drivers/amba/bus.c | 9 +++++++++
> > 1 file changed, 9 insertions(+)
> >
> > diff --git a/drivers/amba/bus.c b/drivers/amba/bus.c
> > index 41b706403ef7..da8f1aac5315 100644
> > --- a/drivers/amba/bus.c
> > +++ b/drivers/amba/bus.c
> > @@ -21,6 +21,7 @@
> > #include <linux/limits.h>
> > #include <linux/clk/clk-conf.h>
> > #include <linux/platform_device.h>
> > +#include <linux/reset.h>
> >
> > #include <asm/irq.h>
> >
> > @@ -352,6 +353,7 @@ static void amba_device_release(struct device *dev)
> >
> > static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> > {
> > + struct reset_control *rst;
> > u32 size;
> > void __iomem *tmp;
> > int i, ret;
> > @@ -388,6 +390,13 @@ static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> > if (ret == 0) {
> > u32 pid, cid;
> >
> > + /* De-assert the reset line to take the device out of reset */
> > + rst = reset_control_get_optional_exclusive(&dev->dev, NULL);
> > + if (IS_ERR(rst))
> > + return PTR_ERR(rst);
>
> It is really correct to propagate an error if we cannot get exclusive
> ownership of the reset line.
>
> With drivers for vendor specific cells it is ok to "just know" that the
> reset line is never shared but we cannot know this for generic cells and
> we certainly can't know this for the bus.
>
> I think it *might* be OK to propagate an error if you used
> reset_control_get_optional_shared() instead because if that reports an
> error than arguably we have either a mistake in the DT or a bug in the
> driver we are sharing a reset with.
>
Hmm. I'm not sure whether we can assume shared reset lines here or not! Maybe
Russell can share his opinion here.
>
> > +
> > + reset_control_deassert(rst);
>
> Perhaps we might also need to explain why we can ignore -ENOTSUPP
> here. Perhaps something like the following based on the comment
> found in in reset_control_deassert():
>
> /*
> * -ENOTSUPP means occurs when the reset controller
> * does not implement .deassert(), in which case the
> * the reset lines should be self-deasserting (and
> * deasserted by default).
> */
> WARN_ON(deassert did not return 0 or -ENOTSUPP);
Ack.
> > +
> > /*
> > * Read pid and cid based on size of resource
> > * they are located at end of region
>
> This looks like it will leak the control reference... shouldn't there
> be a put after you have read the pid/cid?
>
Yes, but it can come before this as well. Right after deassert.
Thanks,
Mani
>
> Daniel.
On Wed, Mar 20, 2019 at 12:26:58PM +0530, Manivannan Sadhasivam wrote:
> Hi Daniel,
>
> On Tue, Mar 12, 2019 at 12:27:11PM +0000, Daniel Thompson wrote:
> > On Sat, Mar 09, 2019 at 07:26:34AM +0530, Manivannan Sadhasivam wrote:
> > > For the AMBA Primecell devices having the reset lines wired, it is
> > > necessary to take them out of reset before reading the pid and cid values.
> > > Earlier we were dependent on the bootloader to do this but a more cleaner
> > > approach would be to do it in the kernel itself. Hence, this commit
> > > deasserts the reset line just before reading the pid and cid values.
> > >
> > > Suggested-by: Daniel Thompson <[email protected]>
> > > Signed-off-by: Manivannan Sadhasivam <[email protected]>
> > > ---
> > > drivers/amba/bus.c | 9 +++++++++
> > > 1 file changed, 9 insertions(+)
> > >
> > > diff --git a/drivers/amba/bus.c b/drivers/amba/bus.c
> > > index 41b706403ef7..da8f1aac5315 100644
> > > --- a/drivers/amba/bus.c
> > > +++ b/drivers/amba/bus.c
> > > @@ -21,6 +21,7 @@
> > > #include <linux/limits.h>
> > > #include <linux/clk/clk-conf.h>
> > > #include <linux/platform_device.h>
> > > +#include <linux/reset.h>
> > >
> > > #include <asm/irq.h>
> > >
> > > @@ -352,6 +353,7 @@ static void amba_device_release(struct device *dev)
> > >
> > > static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> > > {
> > > + struct reset_control *rst;
> > > u32 size;
> > > void __iomem *tmp;
> > > int i, ret;
> > > @@ -388,6 +390,13 @@ static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> > > if (ret == 0) {
> > > u32 pid, cid;
> > >
> > > + /* De-assert the reset line to take the device out of reset */
> > > + rst = reset_control_get_optional_exclusive(&dev->dev, NULL);
> > > + if (IS_ERR(rst))
> > > + return PTR_ERR(rst);
> >
> > It is really correct to propagate an error if we cannot get exclusive
> > ownership of the reset line.
> >
> > With drivers for vendor specific cells it is ok to "just know" that the
> > reset line is never shared but we cannot know this for generic cells and
> > we certainly can't know this for the bus.
> >
> > I think it *might* be OK to propagate an error if you used
> > reset_control_get_optional_shared() instead because if that reports an
> > error than arguably we have either a mistake in the DT or a bug in the
> > driver we are sharing a reset with.
> >
>
> Hmm. I'm not sure whether we can assume shared reset lines here or not! Maybe
> Russell can share his opinion here.
I didn't *assume* anything. I read about the difference between what a
shared reset line does and an exclusive reset line. Providing the
get/put is properly balanced it is both safe and correct to request
shared access to the reset line.
The question is whether it is OK to propagate an error code if another
device has already requested exclusive access to the reset line (which
would cause our request for shared access to fail) since this causes us
to give up on adding the device. Here I concluded that it OK because a
failure does indicates a bug in DT or other driver but added the "I
think" because just-WARN_ON()-and-carry-on" is also a defensible
recovery approach (and *much* less likely to provoke boot sequence
regressions for existing systems).
> > > +
> > > + reset_control_deassert(rst);
> >
> > Perhaps we might also need to explain why we can ignore -ENOTSUPP
> > here. Perhaps something like the following based on the comment
> > found in in reset_control_deassert():
> >
> > /*
> > * -ENOTSUPP means occurs when the reset controller
> > * does not implement .deassert(), in which case the
> > * the reset lines should be self-deasserting (and
> > * deasserted by default).
> > */
> > WARN_ON(deassert did not return 0 or -ENOTSUPP);
>
> Ack.
>
> > > +
> > > /*
> > > * Read pid and cid based on size of resource
> > > * they are located at end of region
> >
> > This looks like it will leak the control reference... shouldn't there
> > be a put after you have read the pid/cid?
> >
>
> Yes, but it can come before this as well. Right after deassert.
No it can't.
If the reset line really really is shared then another driver could
but the cell back into the reset state before you have read the pid/cid.
Daniel.
On Wed, Mar 20, 2019 at 12:26:58PM +0530, Manivannan Sadhasivam wrote:
> Hi Daniel,
>
> On Tue, Mar 12, 2019 at 12:27:11PM +0000, Daniel Thompson wrote:
> > On Sat, Mar 09, 2019 at 07:26:34AM +0530, Manivannan Sadhasivam wrote:
> > > For the AMBA Primecell devices having the reset lines wired, it is
> > > necessary to take them out of reset before reading the pid and cid values.
> > > Earlier we were dependent on the bootloader to do this but a more cleaner
> > > approach would be to do it in the kernel itself. Hence, this commit
> > > deasserts the reset line just before reading the pid and cid values.
> > >
> > > Suggested-by: Daniel Thompson <[email protected]>
> > > Signed-off-by: Manivannan Sadhasivam <[email protected]>
> > > ---
> > > drivers/amba/bus.c | 9 +++++++++
> > > 1 file changed, 9 insertions(+)
> > >
> > > diff --git a/drivers/amba/bus.c b/drivers/amba/bus.c
> > > index 41b706403ef7..da8f1aac5315 100644
> > > --- a/drivers/amba/bus.c
> > > +++ b/drivers/amba/bus.c
> > > @@ -21,6 +21,7 @@
> > > #include <linux/limits.h>
> > > #include <linux/clk/clk-conf.h>
> > > #include <linux/platform_device.h>
> > > +#include <linux/reset.h>
> > >
> > > #include <asm/irq.h>
> > >
> > > @@ -352,6 +353,7 @@ static void amba_device_release(struct device *dev)
> > >
> > > static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> > > {
> > > + struct reset_control *rst;
> > > u32 size;
> > > void __iomem *tmp;
> > > int i, ret;
> > > @@ -388,6 +390,13 @@ static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> > > if (ret == 0) {
> > > u32 pid, cid;
> > >
> > > + /* De-assert the reset line to take the device out of reset */
> > > + rst = reset_control_get_optional_exclusive(&dev->dev, NULL);
> > > + if (IS_ERR(rst))
> > > + return PTR_ERR(rst);
> >
> > It is really correct to propagate an error if we cannot get exclusive
> > ownership of the reset line.
> >
> > With drivers for vendor specific cells it is ok to "just know" that the
> > reset line is never shared but we cannot know this for generic cells and
> > we certainly can't know this for the bus.
> >
> > I think it *might* be OK to propagate an error if you used
> > reset_control_get_optional_shared() instead because if that reports an
> > error than arguably we have either a mistake in the DT or a bug in the
> > driver we are sharing a reset with.
> >
>
> Hmm. I'm not sure whether we can assume shared reset lines here or not! Maybe
> Russell can share his opinion here.
I've no reference to base an opinion on as I have no hardware that
would use this facility.
However, it seems obvious to me that if a reset line is shared between
several devices, and when the reset line is asserted, it blocks reading
the ID, then we do need a way for the AMBA primecell code to be able
to lift the reset to read the device ID, and we need to do it in a way
that is capable of being done even when another device/driver has
already bound and taken control of the reset line.
That said, if a reset line is shared between multiple devices, and a
driver wants to assert the reset line, it would disrupt the operation
of all those devices, so there would need to be some kind of
synchronisation between the drivers.
A different way to look at it though is that such a reset line is not
a property of the individual devices, but of the bus - in which case
it should be specified at bus level and controlled at bus level, not
at individual device level.
What is appropriate is entirely dependent on the hardware setup, and
as I said above, with no view of the hardware I am unable to give a
meaningful opinion.
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up
According to speedtest.net: 11.9Mbps down 500kbps up
On Wed, Mar 20, 2019 at 05:29:56PM +0000, Russell King - ARM Linux admin wrote:
> On Wed, Mar 20, 2019 at 12:26:58PM +0530, Manivannan Sadhasivam wrote:
> > Hi Daniel,
> >
> > On Tue, Mar 12, 2019 at 12:27:11PM +0000, Daniel Thompson wrote:
> > > On Sat, Mar 09, 2019 at 07:26:34AM +0530, Manivannan Sadhasivam wrote:
> > > > For the AMBA Primecell devices having the reset lines wired, it is
> > > > necessary to take them out of reset before reading the pid and cid values.
> > > > Earlier we were dependent on the bootloader to do this but a more cleaner
> > > > approach would be to do it in the kernel itself. Hence, this commit
> > > > deasserts the reset line just before reading the pid and cid values.
> > > >
> > > > Suggested-by: Daniel Thompson <[email protected]>
> > > > Signed-off-by: Manivannan Sadhasivam <[email protected]>
> > > > ---
> > > > drivers/amba/bus.c | 9 +++++++++
> > > > 1 file changed, 9 insertions(+)
> > > >
> > > > diff --git a/drivers/amba/bus.c b/drivers/amba/bus.c
> > > > index 41b706403ef7..da8f1aac5315 100644
> > > > --- a/drivers/amba/bus.c
> > > > +++ b/drivers/amba/bus.c
> > > > @@ -21,6 +21,7 @@
> > > > #include <linux/limits.h>
> > > > #include <linux/clk/clk-conf.h>
> > > > #include <linux/platform_device.h>
> > > > +#include <linux/reset.h>
> > > >
> > > > #include <asm/irq.h>
> > > >
> > > > @@ -352,6 +353,7 @@ static void amba_device_release(struct device *dev)
> > > >
> > > > static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> > > > {
> > > > + struct reset_control *rst;
> > > > u32 size;
> > > > void __iomem *tmp;
> > > > int i, ret;
> > > > @@ -388,6 +390,13 @@ static int amba_device_try_add(struct amba_device *dev, struct resource *parent)
> > > > if (ret == 0) {
> > > > u32 pid, cid;
> > > >
> > > > + /* De-assert the reset line to take the device out of reset */
> > > > + rst = reset_control_get_optional_exclusive(&dev->dev, NULL);
> > > > + if (IS_ERR(rst))
> > > > + return PTR_ERR(rst);
> > >
> > > It is really correct to propagate an error if we cannot get exclusive
> > > ownership of the reset line.
> > >
> > > With drivers for vendor specific cells it is ok to "just know" that the
> > > reset line is never shared but we cannot know this for generic cells and
> > > we certainly can't know this for the bus.
> > >
> > > I think it *might* be OK to propagate an error if you used
> > > reset_control_get_optional_shared() instead because if that reports an
> > > error than arguably we have either a mistake in the DT or a bug in the
> > > driver we are sharing a reset with.
> > >
> >
> > Hmm. I'm not sure whether we can assume shared reset lines here or not! Maybe
> > Russell can share his opinion here.
>
> I've no reference to base an opinion on as I have no hardware that
> would use this facility.
>
> However, it seems obvious to me that if a reset line is shared between
> several devices, and when the reset line is asserted, it blocks reading
> the ID, then we do need a way for the AMBA primecell code to be able
> to lift the reset to read the device ID, and we need to do it in a way
> that is capable of being done even when another device/driver has
> already bound and taken control of the reset line.
>
> That said, if a reset line is shared between multiple devices, and a
> driver wants to assert the reset line, it would disrupt the operation
> of all those devices, so there would need to be some kind of
> synchronisation between the drivers.
That is what shared ownership of the reset line provides. When a line is
shared a single driver does not have the authority to unilaterally
assert reset because deasserts and asserts are counted and the line only
goes high again when they balance.
> A different way to look at it though is that such a reset line is not
> a property of the individual devices, but of the bus - in which case
> it should be specified at bus level and controlled at bus level, not
> at individual device level.
>
> What is appropriate is entirely dependent on the hardware setup, and
> as I said above, with no view of the hardware I am unable to give a
> meaningful opinion.
>
> --
> RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
> FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up
> According to speedtest.net: 11.9Mbps down 500kbps up
On Thu, Mar 21, 2019 at 12:17 PM Daniel Thompson
<[email protected]> wrote:
> On Wed, Mar 20, 2019 at 05:29:56PM +0000, Russell King - ARM Linux admin wrote:
> > On Wed, Mar 20, 2019 at 12:26:58PM +0530, Manivannan Sadhasivam wrote:
> > That said, if a reset line is shared between multiple devices, and a
> > driver wants to assert the reset line, it would disrupt the operation
> > of all those devices, so there would need to be some kind of
> > synchronisation between the drivers.
>
> That is what shared ownership of the reset line provides. When a line is
> shared a single driver does not have the authority to unilaterally
> assert reset because deasserts and asserts are counted and the line only
> goes high again when they balance.
This is what we want for this I'm pretty sure. Regulators have the same
thing with internal reference counting so that if you turn a shared regulator
on from several clients it just increases the refcount and it will not really
be turned off until the last consumer is disabling it.
Yours,
Linus Walleij