device tree bindings for lp5523 leds on N950?

device tree bindings for lp5523 leds on N950?

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 428 bytes --]

Hi!

I'm (still) trying to debug "N950 not booting" situation. I'm stuck
with black screen, so I thought it would be cool to use RGB led for
debugging...

Unfortunately, I don't see required dts bindings in the mainline.

Does anyone have that working?

Thanks,
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

Re: device tree bindings for lp5523 leds on N950?

[Sebastian Reichel]

[-- Attachment #1: Type: text/plain, Size: 568 bytes --]

Hi Pavel,

On Mon, Aug 14, 2017 at 03:38:16PM +0200, Pavel Machek wrote:
> I'm (still) trying to debug "N950 not booting" situation. I'm stuck
> with black screen, so I thought it would be cool to use RGB led for
> debugging...
> 
> Unfortunately, I don't see required dts bindings in the mainline.
> 
> Does anyone have that working?

I have some N950 DTS patches lying around, I think RGB led was one
of them. I'm currently in Canda for a company event without my N950,
but I should be able to send something tested before the weekend.

-- Sebastian

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: device tree bindings for lp5523 leds on N950?

[Sebastian Reichel]

[-- Attachment #1: Type: text/plain, Size: 3618 bytes --]

Hi,

On Tue, Aug 15, 2017 at 09:15:28AM -0400, Sebastian Reichel wrote:
> Hi Pavel,
> 
> On Mon, Aug 14, 2017 at 03:38:16PM +0200, Pavel Machek wrote:
> > I'm (still) trying to debug "N950 not booting" situation. I'm stuck
> > with black screen, so I thought it would be cool to use RGB led for
> > debugging...
> > 
> > Unfortunately, I don't see required dts bindings in the mainline.
> > 
> > Does anyone have that working?
> 
> I have some N950 DTS patches lying around, I think RGB led was one
> of them. I'm currently in Canda for a company event without my N950,
> but I should be able to send something tested before the weekend.

Below is a DTS snippet for N950, that has been tested by me
on top of 4.13-rc6. The left and right keyboard LEDs are
also debug LEDs for cpu idle modes, so they will be lit
without being enabled in lp5523. The status LED of N950 is
not RGB, but just white. Also you need this to get the driver
loading. Otherwise detection after reset fails even after
quite increases waiting times. Looks like the lp5523 reset
(which is not properly Ack'd by the device) breaks the omap-i2c
state horribly.

-- Sebastian

--- a/drivers/leds/leds-lp55xx-common.c
+++ b/drivers/leds/leds-lp55xx-common.c
@@ -413,7 +413,7 @@ int lp55xx_init_device(struct lp55xx_chip *chip)
                usleep_range(1000, 2000); /* 500us abs min. */
        }
 
-       lp55xx_reset_device(chip);
+       //lp55xx_reset_device(chip); // FIXME: breaks omap-i2c
 
        /*
         * Exact value is not available. 10 - 20ms

----------

&i2c2 {
       rgb-led@32 {
               compatible = "national,lp5523";
               reg = <0x32>;

               enable-gpio = <&gpio2 9 GPIO_ACTIVE_HIGH>; /* 41 */
               clock-mode = /bits/ 8 <2>; /* external */

               chan0 {
                       chan-name = "lp5523:kb5"; /* Keyboard ooooox */
                       led-cur = /bits/ 8 <50>;
                       max-cur = /bits/ 8 <130>;
               };

               chan1 {
                       chan-name = "lp5523:kb4"; /* Keyboard ooooxo */
                       led-cur = /bits/ 8 <50>;
                       max-cur = /bits/ 8 <130>;
               };

               chan2 {
                       chan-name = "lp5523:kb3"; /* Keyboard oooxoo */
                       led-cur = /bits/ 8 <50>;
                       max-cur = /bits/ 8 <130>;
               };

               chan3 {
                       chan-name = "lp5523:kb2"; /* Keyboard ooxooo */
                       led-cur = /bits/ 8 <50>;
                       max-cur = /bits/ 8 <130>;
               };

               chan4 {
                       chan-name = "lp5523:kb1"; /* Keyboard oxoooo */
                       led-cur = /bits/ 8 <50>;
                       max-cur = /bits/ 8 <130>;
               };

               chan5 {
                       chan-name = "lp5523:kb0"; /* Keyboard xooooo */
                       led-cur = /bits/ 8 <50>;
                       max-cur = /bits/ 8 <130>;
               };

               chan6 {
                       led-cur = /bits/ 8 <0>;
                       max-cur = /bits/ 8 <0>;
               };

               chan7 {
                       led-cur = /bits/ 8 <0>;
                       max-cur = /bits/ 8 <0>;
               };

               chan8 {
                       chan-name = "lp5523:status-led";
                       led-cur = /bits/ 8 <50>;
                       max-cur = /bits/ 8 <130>;
               };
       };
};

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: device tree bindings for lp5523 leds on N950?

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 1602 bytes --]

Hi!

> On Tue, Aug 15, 2017 at 09:15:28AM -0400, Sebastian Reichel wrote:
> > Hi Pavel,
> > 
> > On Mon, Aug 14, 2017 at 03:38:16PM +0200, Pavel Machek wrote:
> > > I'm (still) trying to debug "N950 not booting" situation. I'm stuck
> > > with black screen, so I thought it would be cool to use RGB led for
> > > debugging...
> > > 
> > > Unfortunately, I don't see required dts bindings in the mainline.
> > > 
> > > Does anyone have that working?
> > 
> > I have some N950 DTS patches lying around, I think RGB led was one
> > of them. I'm currently in Canda for a company event without my N950,
> > but I should be able to send something tested before the weekend.
> 
> Below is a DTS snippet for N950, that has been tested by me
> on top of 4.13-rc6. The left and right keyboard LEDs are
> also debug LEDs for cpu idle modes, so they will be lit
> without being enabled in lp5523. The status LED of N950 is
> not RGB, but just white. Also you need this to get the driver
> loading. Otherwise detection after reset fails even after
> quite increases waiting times. Looks like the lp5523 reset
> (which is not properly Ack'd by the device) breaks the omap-i2c
> state horribly.

Thanks a lot, LED now works. It took me a while, but I now have n950
usable with nfsroot. (For a while, anyway.)

N9 is usable too. Hardware is similar enough than it could be possible
to debug without serial cable.

Best regards,
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]