[PATCH 1/4] regulator: pfuze: remove unsued define

[PATCH 1/4] regulator: pfuze: remove unsued define

[Michael Grzeschik]

Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
---
 drivers/regulator/pfuze.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/regulator/pfuze.c b/drivers/regulator/pfuze.c
index 1950ffb04c..55f7eb5d4c 100644
--- a/drivers/regulator/pfuze.c
+++ b/drivers/regulator/pfuze.c
@@ -31,8 +31,6 @@
 
 #define DRIVERNAME		"pfuze"
 
-#define MC13XXX_NUMREGS		0x3f
-
 #define PFUZE100_SW1ABMODE	0x23
 #define PFUZE100_SW1CMODE	0x31
 #define PFUZE100_SW2MODE	0x38
-- 
2.25.0


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

[PATCH 2/4] regulator: pfuze: add support to other architectures

[Michael Grzeschik]

Currently the pfuze driver is build dependent to ARCH_IMX6. To make it
possible to work with ARCH_IMX8 we move the imx6_poweroff call to an own
poweroff handler.

Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
---
 arch/arm/mach-imx/imx6.c  | 8 ++++++++
 drivers/regulator/Kconfig | 2 +-
 drivers/regulator/pfuze.c | 2 --
 3 files changed, 9 insertions(+), 3 deletions(-)

diff --git a/arch/arm/mach-imx/imx6.c b/arch/arm/mach-imx/imx6.c
index 41e0066add..8f1638c1ea 100644
--- a/arch/arm/mach-imx/imx6.c
+++ b/arch/arm/mach-imx/imx6.c
@@ -401,3 +401,11 @@ void __noreturn imx6_pm_stby_poweroff(void)
 
 	while(1);
 }
+
+static int imx6_init(void)
+{
+	poweroff_handler_register_fn(imx6_pm_stby_poweroff);
+
+	return 0;
+}
+coredevice_initcall(imx6_init);
diff --git a/drivers/regulator/Kconfig b/drivers/regulator/Kconfig
index 28bd69a2a5..f47a115da2 100644
--- a/drivers/regulator/Kconfig
+++ b/drivers/regulator/Kconfig
@@ -19,7 +19,7 @@ config REGULATOR_BCM283X
 config REGULATOR_PFUZE
 	bool "Freescale PFUZE100/200/3000 regulator driver"
 	depends on I2C
-	depends on ARCH_IMX6
+	depends on ARCH_IMX6 || ARCH_IMX8MQ
 
 config REGULATOR_STPMIC1
 	tristate "STMicroelectronics STPMIC1 PMIC Regulators"
diff --git a/drivers/regulator/pfuze.c b/drivers/regulator/pfuze.c
index 55f7eb5d4c..91aaec0e7e 100644
--- a/drivers/regulator/pfuze.c
+++ b/drivers/regulator/pfuze.c
@@ -142,8 +142,6 @@ static void pfuze_power_off_prepare(struct poweroff_handler *handler)
 	regmap_write_bits(pfuze_dev->map, PFUZE100_VGEN6VOL,
 			  PFUZE100_VGENxLPWR | PFUZE100_VGENxSTBY,
 			  PFUZE100_VGENxSTBY);
-
-	imx6_pm_stby_poweroff();
 }
 
 static struct regmap_bus regmap_pfuze_i2c_bus = {
-- 
2.25.0


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

Re: [PATCH 2/4] regulator: pfuze: add support to other architectures

[Sascha Hauer]

Hi Michael,

On Mon, Jan 27, 2020 at 04:50:11PM +0100, Michael Grzeschik wrote:
> Currently the pfuze driver is build dependent to ARCH_IMX6. To make it
> possible to work with ARCH_IMX8 we move the imx6_poweroff call to an own
> poweroff handler.
> 
> Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
> ---
>  arch/arm/mach-imx/imx6.c  | 8 ++++++++
>  drivers/regulator/Kconfig | 2 +-
>  drivers/regulator/pfuze.c | 2 --
>  3 files changed, 9 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/arm/mach-imx/imx6.c b/arch/arm/mach-imx/imx6.c
> index 41e0066add..8f1638c1ea 100644
> --- a/arch/arm/mach-imx/imx6.c
> +++ b/arch/arm/mach-imx/imx6.c
> @@ -401,3 +401,11 @@ void __noreturn imx6_pm_stby_poweroff(void)
>  
>  	while(1);
>  }
> +
> +static int imx6_init(void)
> +{
> +	poweroff_handler_register_fn(imx6_pm_stby_poweroff);
> +
> +	return 0;
> +}

imx6_pm_stby_poweroff() only works when a pfuze is present and is
configured correctly. Calling this on every i.MX board is not nice.

What you can do is to call pfuze_register_init_callback() from here.
You'll have to add a static inline wrapper for this in case pfuze
support is not compiled in.

I wonder how this compiles as my version of this file already has a
function named imx6_init(). Using this is preferred btw as it's already
made sure that it will run only on i.MX6.

> +coredevice_initcall(imx6_init);

Note that this doesn't necessarily means that we are running an i.MX6.
It could be whatever else is compiled in.

Sascha

-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

Re: [PATCH 2/4] regulator: pfuze: add support to other architectures

[Michael Grzeschik]

[-- Attachment #1.1: Type: text/plain, Size: 2497 bytes --]

Hi Sascha,

On Tue, Jan 28, 2020 at 09:45:50AM +0100, Sascha Hauer wrote:
> Hi Michael,
> 
> On Mon, Jan 27, 2020 at 04:50:11PM +0100, Michael Grzeschik wrote:
> > Currently the pfuze driver is build dependent to ARCH_IMX6. To make it
> > possible to work with ARCH_IMX8 we move the imx6_poweroff call to an own
> > poweroff handler.
> > 
> > Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
> > ---
> >  arch/arm/mach-imx/imx6.c  | 8 ++++++++
> >  drivers/regulator/Kconfig | 2 +-
> >  drivers/regulator/pfuze.c | 2 --
> >  3 files changed, 9 insertions(+), 3 deletions(-)
> > 
> > diff --git a/arch/arm/mach-imx/imx6.c b/arch/arm/mach-imx/imx6.c
> > index 41e0066add..8f1638c1ea 100644
> > --- a/arch/arm/mach-imx/imx6.c
> > +++ b/arch/arm/mach-imx/imx6.c
> > @@ -401,3 +401,11 @@ void __noreturn imx6_pm_stby_poweroff(void)
> >  
> >  	while(1);
> >  }
> > +
> > +static int imx6_init(void)
> > +{
> > +	poweroff_handler_register_fn(imx6_pm_stby_poweroff);
> > +
> > +	return 0;
> > +}
> 
> imx6_pm_stby_poweroff() only works when a pfuze is present and is
> configured correctly. Calling this on every i.MX board is not nice.

Right

> What you can do is to call pfuze_register_init_callback() from here.
> You'll have to add a static inline wrapper for this in case pfuze
> support is not compiled in.

So you mean, call pfuze_register_init_callback to register a callback
function which then itself registers the poweroff_handler as above?

I think that could work, but then the init_callback is already used for
every imx6. The callback would also be overwritten by users in board
code trying to preconfigure the pfuze into a sane state, right?

> I wonder how this compiles as my version of this file already has a
> function named imx6_init(). Using this is preferred btw as it's already
> made sure that it will run only on i.MX6.

Duh, it wont. I tried it with imx6 disabled.

> > +coredevice_initcall(imx6_init);
> 
> Note that this doesn't necessarily means that we are running an i.MX6.
> It could be whatever else is compiled in.

Yes, therefor the callback approach is to be prefered.

Michael

-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[-- Attachment #2: Type: text/plain, Size: 149 bytes --]

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

Re: [PATCH 2/4] regulator: pfuze: add support to other architectures

[Sascha Hauer]

On Wed, Jan 29, 2020 at 11:19:20PM +0100, Michael Grzeschik wrote:
> Hi Sascha,
> 
> On Tue, Jan 28, 2020 at 09:45:50AM +0100, Sascha Hauer wrote:
> > Hi Michael,
> > 
> > On Mon, Jan 27, 2020 at 04:50:11PM +0100, Michael Grzeschik wrote:
> > > Currently the pfuze driver is build dependent to ARCH_IMX6. To make it
> > > possible to work with ARCH_IMX8 we move the imx6_poweroff call to an own
> > > poweroff handler.
> > > 
> > > Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
> > > ---
> > >  arch/arm/mach-imx/imx6.c  | 8 ++++++++
> > >  drivers/regulator/Kconfig | 2 +-
> > >  drivers/regulator/pfuze.c | 2 --
> > >  3 files changed, 9 insertions(+), 3 deletions(-)
> > > 
> > > diff --git a/arch/arm/mach-imx/imx6.c b/arch/arm/mach-imx/imx6.c
> > > index 41e0066add..8f1638c1ea 100644
> > > --- a/arch/arm/mach-imx/imx6.c
> > > +++ b/arch/arm/mach-imx/imx6.c
> > > @@ -401,3 +401,11 @@ void __noreturn imx6_pm_stby_poweroff(void)
> > >  
> > >  	while(1);
> > >  }
> > > +
> > > +static int imx6_init(void)
> > > +{
> > > +	poweroff_handler_register_fn(imx6_pm_stby_poweroff);
> > > +
> > > +	return 0;
> > > +}
> > 
> > imx6_pm_stby_poweroff() only works when a pfuze is present and is
> > configured correctly. Calling this on every i.MX board is not nice.
> 
> Right
> 
> > What you can do is to call pfuze_register_init_callback() from here.
> > You'll have to add a static inline wrapper for this in case pfuze
> > support is not compiled in.
> 
> So you mean, call pfuze_register_init_callback to register a callback
> function which then itself registers the poweroff_handler as above?
> 
> I think that could work, but then the init_callback is already used for
> every imx6. The callback would also be overwritten by users in board
> code trying to preconfigure the pfuze into a sane state, right?

Yes. Board code would then have to call imx6_pm_stby_poweroff()
additionally to its own fixups.

Sascha



-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

[PATCH 3/4] regulator: pfuze: add include regmap.h so struct regmap is known

[Michael Grzeschik]

Without this the user of the function will get a compile warning
about the undeclared struct regmap.

Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
---
 include/mfd/pfuze.h | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/include/mfd/pfuze.h b/include/mfd/pfuze.h
index 6045ceec0a..5427b21005 100644
--- a/include/mfd/pfuze.h
+++ b/include/mfd/pfuze.h
@@ -1,6 +1,8 @@
 #ifndef __INCLUDE_PFUZE_H
 #define __INCLUDE_PFUZE_H
 
+#include <regmap.h>
+
 int pfuze_register_init_callback(void(*callback)(struct regmap *map));
 
 #endif /* __INCLUDE_PFUZE_H */
-- 
2.25.0


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

[PATCH 4/4] ARM: phyCORE-i.MX8M SOM: add pmic initialisation for power good

[Michael Grzeschik]

This patch is adding the same PMIC handling as the u-boot spl does for
this Board. It ensures sane defaults.

Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
---
 arch/arm/boards/phytec-som-imx8mq/board.c | 38 +++++++++++++++++++++++
 1 file changed, 38 insertions(+)

diff --git a/arch/arm/boards/phytec-som-imx8mq/board.c b/arch/arm/boards/phytec-som-imx8mq/board.c
index 4fd098c5f6..7df53a7cfb 100644
--- a/arch/arm/boards/phytec-som-imx8mq/board.c
+++ b/arch/arm/boards/phytec-som-imx8mq/board.c
@@ -10,9 +10,45 @@
 #include <init.h>
 #include <linux/sizes.h>
 #include <mach/bbu.h>
+#include <mfd/pfuze.h>
 
 #include <envfs.h>
 
+#define PFUZE100_DEVICEID	0x0
+#define PFUZE100_REVID		0x3
+
+#define PFUZE100_SW1ABMODE	0x23
+#define PFUZE100_SW2MODE	0x38
+#define PFUZE100_SW1CMODE	0x31
+#define PFUZE100_SW3AVOL	0x3c
+
+#define APS_PFM			0xc
+
+static void imx8mq_setup_pmic_voltages(struct regmap *map)
+{
+	int offset = PFUZE100_SW1CMODE;
+	int switch_num = 6;
+	int val, i;
+
+	regmap_read(map, PFUZE100_SW3AVOL, &val);
+
+	/* ensure the correct VDD_DRAM_0V9 output voltage */
+	regmap_write_bits(map, PFUZE100_SW3AVOL, 0x3f, 0x18);
+
+	/* pfuze200 */
+	regmap_read(map, PFUZE100_DEVICEID, &val);
+	if (val & 0xf) {
+		offset = PFUZE100_SW2MODE;
+		switch_num = 4;
+	}
+
+	/* set all switches APS in normal and PFM mode in standby */
+	regmap_write(map, PFUZE100_SW1ABMODE, APS_PFM);
+
+	for (i = 0; i < switch_num - 1; i++)
+		regmap_write(map, offset + i * 7, APS_PFM);
+}
+
 static int physom_imx8mq_devices_init(void)
 {
 	int flag_emmc = 0;
@@ -23,6 +59,8 @@ static int physom_imx8mq_devices_init(void)
 
 	barebox_set_hostname("phycore-imx8mq");
 
+	pfuze_register_init_callback(imx8mq_setup_pmic_voltages);
+
 	switch (bootsource_get_instance()) {
 	case 0:
 		flag_emmc = BBU_HANDLER_FLAG_DEFAULT;
-- 
2.25.0


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

Re: [PATCH 4/4] ARM: phyCORE-i.MX8M SOM: add pmic initialisation for power good

[Sascha Hauer]

On Mon, Jan 27, 2020 at 04:50:13PM +0100, Michael Grzeschik wrote:
> This patch is adding the same PMIC handling as the u-boot spl does for
> this Board. It ensures sane defaults.
> 
> Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
> ---
>  arch/arm/boards/phytec-som-imx8mq/board.c | 38 +++++++++++++++++++++++
>  1 file changed, 38 insertions(+)
> 
> diff --git a/arch/arm/boards/phytec-som-imx8mq/board.c b/arch/arm/boards/phytec-som-imx8mq/board.c
> index 4fd098c5f6..7df53a7cfb 100644
> --- a/arch/arm/boards/phytec-som-imx8mq/board.c
> +++ b/arch/arm/boards/phytec-som-imx8mq/board.c
> @@ -10,9 +10,45 @@
>  #include <init.h>
>  #include <linux/sizes.h>
>  #include <mach/bbu.h>
> +#include <mfd/pfuze.h>
>  
>  #include <envfs.h>
>  
> +#define PFUZE100_DEVICEID	0x0
> +#define PFUZE100_REVID		0x3
> +
> +#define PFUZE100_SW1ABMODE	0x23
> +#define PFUZE100_SW2MODE	0x38
> +#define PFUZE100_SW1CMODE	0x31
> +#define PFUZE100_SW3AVOL	0x3c
> +
> +#define APS_PFM			0xc
> +
> +static void imx8mq_setup_pmic_voltages(struct regmap *map)
> +{
> +	int offset = PFUZE100_SW1CMODE;
> +	int switch_num = 6;
> +	int val, i;
> +
> +	regmap_read(map, PFUZE100_SW3AVOL, &val);
> +
> +	/* ensure the correct VDD_DRAM_0V9 output voltage */
> +	regmap_write_bits(map, PFUZE100_SW3AVOL, 0x3f, 0x18);
> +
> +	/* pfuze200 */
> +	regmap_read(map, PFUZE100_DEVICEID, &val);
> +	if (val & 0xf) {
> +		offset = PFUZE100_SW2MODE;
> +		switch_num = 4;
> +	}
> +
> +	/* set all switches APS in normal and PFM mode in standby */
> +	regmap_write(map, PFUZE100_SW1ABMODE, APS_PFM);
> +
> +	for (i = 0; i < switch_num - 1; i++)
> +		regmap_write(map, offset + i * 7, APS_PFM);
> +}
> +
>  static int physom_imx8mq_devices_init(void)
>  {
>  	int flag_emmc = 0;
> @@ -23,6 +59,8 @@ static int physom_imx8mq_devices_init(void)
>  
>  	barebox_set_hostname("phycore-imx8mq");
>  
> +	pfuze_register_init_callback(imx8mq_setup_pmic_voltages);

Yes, providing a static inline wrapper function for this is a good idea.

Sascha

-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox