[PATCH] backlight: l4f00242t03: Fix reset sequence

[PATCH] backlight: l4f00242t03: Fix reset sequence

[Alberto Panizzo]

The reset command is part of the init sequence and it take effect
only if the lcd is powered.

Signed-off-by: Alberto Panizzo <maramaopercheseimorto@gmail.com>
---

This patch is built on top of:
http://userweb.kernel.org/~akpm/mmotm/broken-out/drivers-video-backlight-l4f00242t03c-full-implement-fb-power-states-for-this-lcd.patch

 drivers/video/backlight/l4f00242t03.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/drivers/video/backlight/l4f00242t03.c b/drivers/video/backlight/l4f00242t03.c
index 5d25743..c04aa90 100644
--- a/drivers/video/backlight/l4f00242t03.c
+++ b/drivers/video/backlight/l4f00242t03.c
@@ -62,6 +62,8 @@ static void l4f00242t03_lcd_init(struct spi_device *spi)
 		regulator_enable(priv->core_reg);
 	}
 
+	l4f00242t03_reset(pdata->reset_gpio);
+
 	gpio_set_value(pdata->data_enable_gpio, 1);
 	msleep(60);
 	spi_write(spi, (const u8 *)cmd, ARRAY_SIZE(cmd) * sizeof(u16));
@@ -225,7 +227,6 @@ static int __devinit l4f00242t03_probe(struct spi_device *spi)
 	}
 
 	/* Init the LCD */
-	l4f00242t03_reset(pdata->reset_gpio);
 	l4f00242t03_lcd_init(spi);
 	priv->lcd_state = FB_BLANK_VSYNC_SUSPEND;
 	l4f00242t03_lcd_power_set(priv->ld, FB_BLANK_UNBLANK);
-- 
1.6.3.3

[PATCH] backlight: l4f00242t03: Prevent unbalanced calls to regulator enable/disable

[Alberto Panizzo]

Otherwise a double call to:
$ echo 4 > /sys/class/lcd/l4f00242t03/lcd_power
Will, the first power down the lcd and regulators correctly and the
second produce an unbalanced call to regulator disable.

Signed-off-by: Alberto Panizzo <maramaopercheseimorto@gmail.com>
---

This patch follow the one this mail is answering.
Thanks for reviewing.

Best regards,
Alberto!

 drivers/video/backlight/l4f00242t03.c |   10 ++++++----
 1 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/video/backlight/l4f00242t03.c b/drivers/video/backlight/l4f00242t03.c
index c04aa90..98ad3e5 100644
--- a/drivers/video/backlight/l4f00242t03.c
+++ b/drivers/video/backlight/l4f00242t03.c
@@ -136,10 +136,12 @@ static int l4f00242t03_lcd_power_set(struct lcd_device *ld, int power)
 		}
 	} else {
 		/* power == FB_BLANK_POWERDOWN */
-		/* Clear the screen before shutting down */
-		spi_write(spi, (const u8 *)&disoff, sizeof(u16));
-		msleep(60);
-		l4f00242t03_lcd_powerdown(spi);
+		if (priv->lcd_state != FB_BLANK_POWERDOWN) {
+			/* Clear the screen before shutting down */
+			spi_write(spi, (const u8 *)&disoff, sizeof(u16));
+			msleep(60);
+			l4f00242t03_lcd_powerdown(spi);
+		}
 	}
 
 	priv->lcd_state = power;
-- 
1.6.3.3

Re: [PATCH] backlight: l4f00242t03: Fix reset sequence

[Andrew Morton]

On Sun, 28 Nov 2010 18:13:53 +0100
Alberto Panizzo <maramaopercheseimorto@gmail.com> wrote:

> The reset command is part of the init sequence and it take effect
> only if the lcd is powered.
> 
> Signed-off-by: Alberto Panizzo <maramaopercheseimorto@gmail.com>
> ---
> 
> This patch is built on top of:
> http://userweb.kernel.org/~akpm/mmotm/broken-out/drivers-video-backlight-l4f00242t03c-full-implement-fb-power-states-for-this-lcd.patch
> 
>  drivers/video/backlight/l4f00242t03.c |    3 ++-
>  1 files changed, 2 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/video/backlight/l4f00242t03.c b/drivers/video/backlight/l4f00242t03.c
> index 5d25743..c04aa90 100644
> --- a/drivers/video/backlight/l4f00242t03.c
> +++ b/drivers/video/backlight/l4f00242t03.c
> @@ -62,6 +62,8 @@ static void l4f00242t03_lcd_init(struct spi_device *spi)
>  		regulator_enable(priv->core_reg);
>  	}
>  
> +	l4f00242t03_reset(pdata->reset_gpio);
> +
>  	gpio_set_value(pdata->data_enable_gpio, 1);
>  	msleep(60);
>  	spi_write(spi, (const u8 *)cmd, ARRAY_SIZE(cmd) * sizeof(u16));
> @@ -225,7 +227,6 @@ static int __devinit l4f00242t03_probe(struct spi_device *spi)
>  	}
>  
>  	/* Init the LCD */
> -	l4f00242t03_reset(pdata->reset_gpio);
>  	l4f00242t03_lcd_init(spi);
>  	priv->lcd_state = FB_BLANK_VSYNC_SUSPEND;
>  	l4f00242t03_lcd_power_set(priv->ld, FB_BLANK_UNBLANK);

What were the user-visible effects of the bug which you just fixed?

Please always include this info when sending a bugfix!  Because without
that info it's much harder to work out whether we should backport the
fix into earlier kernels and it's harder for end-users and distro
maintainers to work out if the patch will fix particular behaviour
which they or their customers are seeing.

Re: [PATCH] backlight: l4f00242t03: Fix reset sequence

[Alberto Panizzo]

On Tue, 2011-01-04 at 15:59 -0800, Andrew Morton wrote:
> On Sun, 28 Nov 2010 18:13:53 +0100
> Alberto Panizzo <maramaopercheseimorto@gmail.com> wrote:
> 
> > The reset command is part of the init sequence and it take effect
> > only if the lcd is powered.
> > 
> > Signed-off-by: Alberto Panizzo <maramaopercheseimorto@gmail.com>
> > ---
> > 
> > This patch is built on top of:
> > http://userweb.kernel.org/~akpm/mmotm/broken-out/drivers-video-backlight-l4f00242t03c-full-implement-fb-power-states-for-this-lcd.patch
> > 
> >  drivers/video/backlight/l4f00242t03.c |    3 ++-
> >  1 files changed, 2 insertions(+), 1 deletions(-)
> > 
> > diff --git a/drivers/video/backlight/l4f00242t03.c b/drivers/video/backlight/l4f00242t03.c
> > index 5d25743..c04aa90 100644
> > --- a/drivers/video/backlight/l4f00242t03.c
> > +++ b/drivers/video/backlight/l4f00242t03.c
> > @@ -62,6 +62,8 @@ static void l4f00242t03_lcd_init(struct spi_device *spi)
> >  		regulator_enable(priv->core_reg);
> >  	}
> >  
> > +	l4f00242t03_reset(pdata->reset_gpio);
> > +
> >  	gpio_set_value(pdata->data_enable_gpio, 1);
> >  	msleep(60);
> >  	spi_write(spi, (const u8 *)cmd, ARRAY_SIZE(cmd) * sizeof(u16));
> > @@ -225,7 +227,6 @@ static int __devinit l4f00242t03_probe(struct spi_device *spi)
> >  	}
> >  
> >  	/* Init the LCD */
> > -	l4f00242t03_reset(pdata->reset_gpio);
> >  	l4f00242t03_lcd_init(spi);
> >  	priv->lcd_state = FB_BLANK_VSYNC_SUSPEND;
> >  	l4f00242t03_lcd_power_set(priv->ld, FB_BLANK_UNBLANK);
> 
> What were the user-visible effects of the bug which you just fixed?
> 
> Please always include this info when sending a bugfix!  Because without
> that info it's much harder to work out whether we should backport the
> fix into earlier kernels and it's harder for end-users and distro
> maintainers to work out if the patch will fix particular behaviour
> which they or their customers are seeing.
> 

The effect of the bug was that the sequence:
set lcd power_state to FB_BLANK_POWERDOWN
set lcd power_state to FB_BLANK_UNBLANK
Did not produced a complete reboot of the LCD which was showing fuzzy 
colours.

This was not experienced before implementing correctly all the LCD power
states with the patch [1]. Since before the patch [1] the regulators
were not touched and the LCD shutdown was reached with a register 
write. After the patch [1] a complete boot sequence with an initial 
reset is needed for the display every time the LCD is powered up.

drivers-video-backlight-l4f00242t03c-full-implement-fb-power-states-for-this-lcd.patch

Best Regards,
-- 
Alberto!

        Be Persistent!
                - Greg Kroah-Hartman (FOSDEM 2010)