Re: [PATCH] backlight: corgi: Convert to use GPIO descriptors

2019-12-17 Thread Linus Walleij
On Mon, Dec 16, 2019 at 5:27 PM Lee Jones  wrote:
> On Tue, 03 Dec 2019, Linus Walleij wrote:
>
> > The code in the Corgi backlight driver can be considerably
> > simplified by moving to GPIO descriptors and lookup tables
> > from the board files instead of passing GPIO numbers using
> > the old API.
> >
> > Make sure to encode inversion semantics for the Akita and
> > Spitz platforms inside the GPIO lookup table and drop the
> > custom inversion semantics from the driver.
> >
> > All in-tree users are converted in this patch.
> >
> > Cc: Andrea Adami 
> > Cc: Robert Jarzmik 
> > Signed-off-by: Linus Walleij 
> > ---
> >  arch/arm/mach-pxa/corgi.c   | 12 -
> >  arch/arm/mach-pxa/spitz.c   | 34 +++
>
> Does this have an Arm Ack yet?

Yes Robert Jarzmik is the maintainer of arch/arm/mach-pxa,
or do you mean I should ask for the ARM SoC maintainers'
ACK explicitly (then I'd just resend, NP).

> >  drivers/video/backlight/corgi_lcd.c | 68 -
> >  include/linux/spi/corgi_lcd.h   |  3 --
>
> What about SPI?

That include file should technically be in the backlight
include path or platform_data (yeah someone should
fix...) but let's ask Mark explicitly if this is OK.

Mark: the hunk hitting the include/linux/spi/corgi_lcd.h
file looks like this:

@@ -11,9 +11,6 @@ struct corgi_lcd_platform_data {
int default_intensity;
int limit_mask;

-   int gpio_backlight_on;  /* -1 if n/a */
-   int gpio_backlight_cont;/* -1 if n/a */
-
void (*notify)(int intensity);
void (*kick_battery)(void);
 };

OK with you? If you want I can resend the whole patch
as well so you can have a look.

Yours,
Linus Walleij
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel


Re: [PATCH] backlight: corgi: Convert to use GPIO descriptors

2019-12-16 Thread Lee Jones
On Tue, 03 Dec 2019, Linus Walleij wrote:

> The code in the Corgi backlight driver can be considerably
> simplified by moving to GPIO descriptors and lookup tables
> from the board files instead of passing GPIO numbers using
> the old API.
> 
> Make sure to encode inversion semantics for the Akita and
> Spitz platforms inside the GPIO lookup table and drop the
> custom inversion semantics from the driver.
> 
> All in-tree users are converted in this patch.
> 
> Cc: Andrea Adami 
> Cc: Robert Jarzmik 
> Signed-off-by: Linus Walleij 
> ---
>  arch/arm/mach-pxa/corgi.c   | 12 -
>  arch/arm/mach-pxa/spitz.c   | 34 +++

Does this have an Arm Ack yet?

>  drivers/video/backlight/corgi_lcd.c | 68 -
>  include/linux/spi/corgi_lcd.h   |  3 --

What about SPI?

>  4 files changed, 54 insertions(+), 63 deletions(-)

-- 
Lee Jones [李琼斯]
Linaro Services Technical Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel


Re: [PATCH] backlight: corgi: Convert to use GPIO descriptors

2019-12-16 Thread Lee Jones
On Sat, 14 Dec 2019, Linus Walleij wrote:

> On Fri, Dec 13, 2019 at 6:24 PM Robert Jarzmik  wrote:
> > Linus Walleij  writes:
> > > On Sun, Dec 8, 2019 at 9:06 PM Robert Jarzmik  
> > > wrote:
> > >> Linus Walleij  writes:
> 
> > > So it will theoretically "spi0.1"
> > >
> > > Beware about bugs in the above interpreter because it is
> > > just my brain.
> >
> > Well, if nobody complains because of testing, why not, your explanation 
> > seems as
> > good as it could be.
> 
> :D
> 
> > If you would be so kind as to carry these changes through your tree instead 
> > of
> > the PXA one, please have my :
> > Acked-by: Robert Jarzmik 
> 
> I was hoping the backlight people would pick it up to the backlight tree.

Yep, that's the plan.

Big update today, bear with.

-- 
Lee Jones [李琼斯]
Linaro Services Technical Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel


Re: [PATCH] backlight: corgi: Convert to use GPIO descriptors

2019-12-14 Thread Robert Jarzmik
Linus Walleij  writes:

> On Sun, Dec 8, 2019 at 9:06 PM Robert Jarzmik  wrote:
>
>> Linus Walleij  writes:
>> > @@ -525,13 +525,33 @@ static void spitz_bl_kick_battery(void)
>> >   }
>> >  }
>> >
>> > +static struct gpiod_lookup_table spitz_lcdcon_gpio_table = {
>> > + .dev_id = "spi0.1",
>> How do you know the correct device name is "spi0.1" ?
>
> With SPI devices it is always hard to know without access to the
> actual hardware, so every patch is a request for testing...
>
> I looked at arch/arm/mach-pxa/spitz.c and
> it registers just one spi bus (AFAICT) with 3 chip
> selects so that will be "spi0", and then
> spi_register_board_info() is called with an array of 3
> devices (spitz_spi_devices[]). Those are in order of
> chip select so chip select 0, 1, 2. This is the second
> device so chip select 1.
>
> The code in drivers/spi/spi.c names the devices
> using spi_dev_set_name() like this:
> dev_set_name(&spi->dev, "%s.%u", dev_name(&spi->controller->dev),
>  spi->chip_select);
>
> So it will theoretically "spi0.1"
>
> Beware about bugs in the above interpreter because it is
> just my brain.
Well, if nobody complains because of testing, why not, your explanation seems as
good as it could be.

Moreover my boards are actually packed for a house move, so I hope for the best
with all the patches for gpiod conversion, as I won't be able to test much for a
month.

If you would be so kind as to carry these changes through your tree instead of
the PXA one, please have my :
Acked-by: Robert Jarzmik 

Cheers.

-- 
Robert
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel


Re: [PATCH] backlight: corgi: Convert to use GPIO descriptors

2019-12-13 Thread Linus Walleij
On Fri, Dec 13, 2019 at 6:24 PM Robert Jarzmik  wrote:
> Linus Walleij  writes:
> > On Sun, Dec 8, 2019 at 9:06 PM Robert Jarzmik  
> > wrote:
> >> Linus Walleij  writes:

> > So it will theoretically "spi0.1"
> >
> > Beware about bugs in the above interpreter because it is
> > just my brain.
>
> Well, if nobody complains because of testing, why not, your explanation seems 
> as
> good as it could be.

:D

> If you would be so kind as to carry these changes through your tree instead of
> the PXA one, please have my :
> Acked-by: Robert Jarzmik 

I was hoping the backlight people would pick it up to the backlight tree.

Yours,
Linus Walleij
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel


Re: [PATCH] backlight: corgi: Convert to use GPIO descriptors

2019-12-10 Thread Linus Walleij
On Sun, Dec 8, 2019 at 9:06 PM Robert Jarzmik  wrote:

> Linus Walleij  writes:
> > @@ -525,13 +525,33 @@ static void spitz_bl_kick_battery(void)
> >   }
> >  }
> >
> > +static struct gpiod_lookup_table spitz_lcdcon_gpio_table = {
> > + .dev_id = "spi0.1",
> How do you know the correct device name is "spi0.1" ?

With SPI devices it is always hard to know without access to the
actual hardware, so every patch is a request for testing...

I looked at arch/arm/mach-pxa/spitz.c and
it registers just one spi bus (AFAICT) with 3 chip
selects so that will be "spi0", and then
spi_register_board_info() is called with an array of 3
devices (spitz_spi_devices[]). Those are in order of
chip select so chip select 0, 1, 2. This is the second
device so chip select 1.

The code in drivers/spi/spi.c names the devices
using spi_dev_set_name() like this:
dev_set_name(&spi->dev, "%s.%u", dev_name(&spi->controller->dev),
 spi->chip_select);

So it will theoretically "spi0.1"

Beware about bugs in the above interpreter because it is
just my brain.

Yours,
Linus Walleij
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel


Re: [PATCH] backlight: corgi: Convert to use GPIO descriptors

2019-12-09 Thread Robert Jarzmik
Linus Walleij  writes:

Hi Linus,

> @@ -525,13 +525,33 @@ static void spitz_bl_kick_battery(void)
>   }
>  }
>  
> +static struct gpiod_lookup_table spitz_lcdcon_gpio_table = {
> + .dev_id = "spi0.1",
How do you know the correct device name is "spi0.1" ?

Cheers.

--
Robert
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH] backlight: corgi: Convert to use GPIO descriptors

2019-12-05 Thread Daniel Thompson
On Tue, Dec 03, 2019 at 01:31:43PM +0100, Linus Walleij wrote:
> The code in the Corgi backlight driver can be considerably
> simplified by moving to GPIO descriptors and lookup tables
> from the board files instead of passing GPIO numbers using
> the old API.
> 
> Make sure to encode inversion semantics for the Akita and
> Spitz platforms inside the GPIO lookup table and drop the
> custom inversion semantics from the driver.
> 
> All in-tree users are converted in this patch.
> 
> Cc: Andrea Adami 
> Cc: Robert Jarzmik 
> Signed-off-by: Linus Walleij 

Reviewed-by: Daniel Thompson 

> ---
>  arch/arm/mach-pxa/corgi.c   | 12 -
>  arch/arm/mach-pxa/spitz.c   | 34 +++
>  drivers/video/backlight/corgi_lcd.c | 68 -
>  include/linux/spi/corgi_lcd.h   |  3 --
>  4 files changed, 54 insertions(+), 63 deletions(-)
> 
> diff --git a/arch/arm/mach-pxa/corgi.c b/arch/arm/mach-pxa/corgi.c
> index f2d73289230f..c9625dcae27c 100644
> --- a/arch/arm/mach-pxa/corgi.c
> +++ b/arch/arm/mach-pxa/corgi.c
> @@ -563,13 +563,20 @@ static void corgi_bl_kick_battery(void)
>   }
>  }
>  
> +static struct gpiod_lookup_table corgi_lcdcon_gpio_table = {
> + .dev_id = "spi0.1",
> + .table = {
> + GPIO_LOOKUP("gpio-pxa", CORGI_GPIO_BACKLIGHT_CONT,
> + "BL_CONT", GPIO_ACTIVE_HIGH),
> + { },
> + },
> +};
> +
>  static struct corgi_lcd_platform_data corgi_lcdcon_info = {
>   .init_mode  = CORGI_LCD_MODE_VGA,
>   .max_intensity  = 0x2f,
>   .default_intensity  = 0x1f,
>   .limit_mask = 0x0b,
> - .gpio_backlight_cont= CORGI_GPIO_BACKLIGHT_CONT,
> - .gpio_backlight_on  = -1,
>   .kick_battery   = corgi_bl_kick_battery,
>  };
>  
> @@ -609,6 +616,7 @@ static struct spi_board_info corgi_spi_devices[] = {
>  static void __init corgi_init_spi(void)
>  {
>   pxa2xx_set_spi_info(1, &corgi_spi_info);
> + gpiod_add_lookup_table(&corgi_lcdcon_gpio_table);
>   spi_register_board_info(ARRAY_AND_SIZE(corgi_spi_devices));
>  }
>  #else
> diff --git a/arch/arm/mach-pxa/spitz.c b/arch/arm/mach-pxa/spitz.c
> index a4fdc399d152..82e80a257c0f 100644
> --- a/arch/arm/mach-pxa/spitz.c
> +++ b/arch/arm/mach-pxa/spitz.c
> @@ -525,13 +525,33 @@ static void spitz_bl_kick_battery(void)
>   }
>  }
>  
> +static struct gpiod_lookup_table spitz_lcdcon_gpio_table = {
> + .dev_id = "spi0.1",
> + .table = {
> + GPIO_LOOKUP("gpio-pxa", SPITZ_GPIO_BACKLIGHT_CONT,
> + "BL_CONT", GPIO_ACTIVE_LOW),
> + GPIO_LOOKUP("gpio-pxa", SPITZ_GPIO_BACKLIGHT_ON,
> + "BL_ON", GPIO_ACTIVE_HIGH),
> + { },
> + },
> +};
> +
> +static struct gpiod_lookup_table akita_lcdcon_gpio_table = {
> + .dev_id = "spi0.1",
> + .table = {
> + GPIO_LOOKUP("gpio-pxa", AKITA_GPIO_BACKLIGHT_CONT,
> + "BL_CONT", GPIO_ACTIVE_LOW),
> + GPIO_LOOKUP("gpio-pxa", AKITA_GPIO_BACKLIGHT_ON,
> + "BL_ON", GPIO_ACTIVE_HIGH),
> + { },
> + },
> +};
> +
>  static struct corgi_lcd_platform_data spitz_lcdcon_info = {
>   .init_mode  = CORGI_LCD_MODE_VGA,
>   .max_intensity  = 0x2f,
>   .default_intensity  = 0x1f,
>   .limit_mask = 0x0b,
> - .gpio_backlight_cont= SPITZ_GPIO_BACKLIGHT_CONT,
> - .gpio_backlight_on  = SPITZ_GPIO_BACKLIGHT_ON,
>   .kick_battery   = spitz_bl_kick_battery,
>  };
>  
> @@ -574,12 +594,10 @@ static struct pxa2xx_spi_controller spitz_spi_info = {
>  
>  static void __init spitz_spi_init(void)
>  {
> - struct corgi_lcd_platform_data *lcd_data = &spitz_lcdcon_info;
> -
> - if (machine_is_akita()) {
> - lcd_data->gpio_backlight_cont = AKITA_GPIO_BACKLIGHT_CONT;
> - lcd_data->gpio_backlight_on = AKITA_GPIO_BACKLIGHT_ON;
> - }
> + if (machine_is_akita())
> + gpiod_add_lookup_table(&akita_lcdcon_gpio_table);
> + else
> + gpiod_add_lookup_table(&spitz_lcdcon_gpio_table);
>  
>   pxa2xx_set_spi_info(2, &spitz_spi_info);
>   spi_register_board_info(ARRAY_AND_SIZE(spitz_spi_devices));
> diff --git a/drivers/video/backlight/corgi_lcd.c 
> b/drivers/video/backlight/corgi_lcd.c
> index 68f7592c5060..25ef0cbd7583 100644
> --- a/drivers/video/backlight/corgi_lcd.c
> +++ b/drivers/video/backlight/corgi_lcd.c
> @@ -15,7 +15,7 @@
>  #include 
>  #include 
>  #include 
> -#include 
> +#include 
>  #include 
>  #include 
>  #include 
> @@ -90,9 +90,8 @@ struct corgi_lcd {
>   int mode;
>   charbuf[2];
>  
> - int gpio_backlight_on;
> - int gpio_backlight_cont;
> - int gpio_backlight_cont_inverted;
> + struct gpio_desc *backlight_on;
> + struct gpio_desc *backlight_cont;
>  
>   void (*kick_battery)(void);
>  };
> 

[PATCH] backlight: corgi: Convert to use GPIO descriptors

2019-12-03 Thread Linus Walleij
The code in the Corgi backlight driver can be considerably
simplified by moving to GPIO descriptors and lookup tables
from the board files instead of passing GPIO numbers using
the old API.

Make sure to encode inversion semantics for the Akita and
Spitz platforms inside the GPIO lookup table and drop the
custom inversion semantics from the driver.

All in-tree users are converted in this patch.

Cc: Andrea Adami 
Cc: Robert Jarzmik 
Signed-off-by: Linus Walleij 
---
 arch/arm/mach-pxa/corgi.c   | 12 -
 arch/arm/mach-pxa/spitz.c   | 34 +++
 drivers/video/backlight/corgi_lcd.c | 68 -
 include/linux/spi/corgi_lcd.h   |  3 --
 4 files changed, 54 insertions(+), 63 deletions(-)

diff --git a/arch/arm/mach-pxa/corgi.c b/arch/arm/mach-pxa/corgi.c
index f2d73289230f..c9625dcae27c 100644
--- a/arch/arm/mach-pxa/corgi.c
+++ b/arch/arm/mach-pxa/corgi.c
@@ -563,13 +563,20 @@ static void corgi_bl_kick_battery(void)
}
 }
 
+static struct gpiod_lookup_table corgi_lcdcon_gpio_table = {
+   .dev_id = "spi0.1",
+   .table = {
+   GPIO_LOOKUP("gpio-pxa", CORGI_GPIO_BACKLIGHT_CONT,
+   "BL_CONT", GPIO_ACTIVE_HIGH),
+   { },
+   },
+};
+
 static struct corgi_lcd_platform_data corgi_lcdcon_info = {
.init_mode  = CORGI_LCD_MODE_VGA,
.max_intensity  = 0x2f,
.default_intensity  = 0x1f,
.limit_mask = 0x0b,
-   .gpio_backlight_cont= CORGI_GPIO_BACKLIGHT_CONT,
-   .gpio_backlight_on  = -1,
.kick_battery   = corgi_bl_kick_battery,
 };
 
@@ -609,6 +616,7 @@ static struct spi_board_info corgi_spi_devices[] = {
 static void __init corgi_init_spi(void)
 {
pxa2xx_set_spi_info(1, &corgi_spi_info);
+   gpiod_add_lookup_table(&corgi_lcdcon_gpio_table);
spi_register_board_info(ARRAY_AND_SIZE(corgi_spi_devices));
 }
 #else
diff --git a/arch/arm/mach-pxa/spitz.c b/arch/arm/mach-pxa/spitz.c
index a4fdc399d152..82e80a257c0f 100644
--- a/arch/arm/mach-pxa/spitz.c
+++ b/arch/arm/mach-pxa/spitz.c
@@ -525,13 +525,33 @@ static void spitz_bl_kick_battery(void)
}
 }
 
+static struct gpiod_lookup_table spitz_lcdcon_gpio_table = {
+   .dev_id = "spi0.1",
+   .table = {
+   GPIO_LOOKUP("gpio-pxa", SPITZ_GPIO_BACKLIGHT_CONT,
+   "BL_CONT", GPIO_ACTIVE_LOW),
+   GPIO_LOOKUP("gpio-pxa", SPITZ_GPIO_BACKLIGHT_ON,
+   "BL_ON", GPIO_ACTIVE_HIGH),
+   { },
+   },
+};
+
+static struct gpiod_lookup_table akita_lcdcon_gpio_table = {
+   .dev_id = "spi0.1",
+   .table = {
+   GPIO_LOOKUP("gpio-pxa", AKITA_GPIO_BACKLIGHT_CONT,
+   "BL_CONT", GPIO_ACTIVE_LOW),
+   GPIO_LOOKUP("gpio-pxa", AKITA_GPIO_BACKLIGHT_ON,
+   "BL_ON", GPIO_ACTIVE_HIGH),
+   { },
+   },
+};
+
 static struct corgi_lcd_platform_data spitz_lcdcon_info = {
.init_mode  = CORGI_LCD_MODE_VGA,
.max_intensity  = 0x2f,
.default_intensity  = 0x1f,
.limit_mask = 0x0b,
-   .gpio_backlight_cont= SPITZ_GPIO_BACKLIGHT_CONT,
-   .gpio_backlight_on  = SPITZ_GPIO_BACKLIGHT_ON,
.kick_battery   = spitz_bl_kick_battery,
 };
 
@@ -574,12 +594,10 @@ static struct pxa2xx_spi_controller spitz_spi_info = {
 
 static void __init spitz_spi_init(void)
 {
-   struct corgi_lcd_platform_data *lcd_data = &spitz_lcdcon_info;
-
-   if (machine_is_akita()) {
-   lcd_data->gpio_backlight_cont = AKITA_GPIO_BACKLIGHT_CONT;
-   lcd_data->gpio_backlight_on = AKITA_GPIO_BACKLIGHT_ON;
-   }
+   if (machine_is_akita())
+   gpiod_add_lookup_table(&akita_lcdcon_gpio_table);
+   else
+   gpiod_add_lookup_table(&spitz_lcdcon_gpio_table);
 
pxa2xx_set_spi_info(2, &spitz_spi_info);
spi_register_board_info(ARRAY_AND_SIZE(spitz_spi_devices));
diff --git a/drivers/video/backlight/corgi_lcd.c 
b/drivers/video/backlight/corgi_lcd.c
index 68f7592c5060..25ef0cbd7583 100644
--- a/drivers/video/backlight/corgi_lcd.c
+++ b/drivers/video/backlight/corgi_lcd.c
@@ -15,7 +15,7 @@
 #include 
 #include 
 #include 
-#include 
+#include 
 #include 
 #include 
 #include 
@@ -90,9 +90,8 @@ struct corgi_lcd {
int mode;
charbuf[2];
 
-   int gpio_backlight_on;
-   int gpio_backlight_cont;
-   int gpio_backlight_cont_inverted;
+   struct gpio_desc *backlight_on;
+   struct gpio_desc *backlight_cont;
 
void (*kick_battery)(void);
 };
@@ -403,13 +402,13 @@ static int corgi_bl_set_intensity(struct corgi_lcd *lcd, 
int intensity)
corgi_ssp_lcdtg_send(lcd, DUTYCTRL_ADRS, intensity);
 
/* Bit 5 via GPIO_BACKLIGHT_CONT */
-   cont = !!(intensity & 0x20) ^ lcd->gpio_backl