diff mbox series

platform/chrome: cros_kbd_led_backlight: support Google Link

Message ID 20241113023800.6263-2-drafnel@gmail.com (mailing list archive)
State New
Headers show
Series platform/chrome: cros_kbd_led_backlight: support Google Link | expand

Commit Message

Brandon Casey Nov. 13, 2024, 2:29 a.m. UTC
On the Google Pixel chromebook (aka Google Link), the keyboard led
backlight has the ACPI device id "GGL0002".  Let's add it to the
configuration array.

Signed-off-by: Brandon Casey <drafnel@gmail.com>
---
 drivers/platform/chrome/cros_kbd_led_backlight.c | 1 +
 1 file changed, 1 insertion(+)

Comments

Tzung-Bi Shih Nov. 13, 2024, 6:46 a.m. UTC | #1
On Tue, Nov 12, 2024 at 08:29:05PM -0600, Brandon Casey wrote:
> On the Google Pixel chromebook (aka Google Link), the keyboard led
> backlight has the ACPI device id "GGL0002".  Let's add it to the
> configuration array.

Why it can't be matched by "GOOG0002"? [1]  Have you tried to update your AP
firmware?

[1]: https://review.coreboot.org/c/coreboot/+/11478
Brandon Casey Nov. 13, 2024, 8:44 a.m. UTC | #2
On November 13, 2024 12:46:18 AM CST, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
>On Tue, Nov 12, 2024 at 08:29:05PM -0600, Brandon Casey wrote:
>> On the Google Pixel chromebook (aka Google Link), the keyboard led
>> backlight has the ACPI device id "GGL0002".  Let's add it to the
>> configuration array.
>
>Why it can't be matched by "GOOG0002"? [1] 

I'm not sure I understand this question. On my device the backlight has ACPI id "GGL0002", not "GOOG0002"?

>Have you tried to update your AP firmware?

I have not explicitly updated any firmware. It has whatever firmware Chromeos has installed.

-Brandon
Tzung-Bi Shih Nov. 14, 2024, 2:07 a.m. UTC | #3
On Wed, Nov 13, 2024 at 02:44:55AM -0600, Brandon Casey wrote:
> 
> 
> On November 13, 2024 12:46:18 AM CST, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
> >On Tue, Nov 12, 2024 at 08:29:05PM -0600, Brandon Casey wrote:
> >> On the Google Pixel chromebook (aka Google Link), the keyboard led
> >> backlight has the ACPI device id "GGL0002".  Let's add it to the
> >> configuration array.
> >
> >Why it can't be matched by "GOOG0002"? [1] 
> 
> I'm not sure I understand this question. On my device the backlight has ACPI id "GGL0002", not "GOOG0002"?
> 
> >Have you tried to update your AP firmware?
> 
> I have not explicitly updated any firmware. It has whatever firmware Chromeos has installed.

IIUC, if your AP firmware includes patch [1] (submitted in 2015), it should
report "GOOG0002" instead of "GGL0002".

[1]: https://review.coreboot.org/c/coreboot/+/11478
Brandon Casey Nov. 14, 2024, 4:38 a.m. UTC | #4
On November 13, 2024 8:07:42 PM CST, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
>On Wed, Nov 13, 2024 at 02:44:55AM -0600, Brandon Casey wrote:
>> 
>> 
>> On November 13, 2024 12:46:18 AM CST, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
>> >On Tue, Nov 12, 2024 at 08:29:05PM -0600, Brandon Casey wrote:
>> >> On the Google Pixel chromebook (aka Google Link), the keyboard led
>> >> backlight has the ACPI device id "GGL0002".  Let's add it to the
>> >> configuration array.
>> >
>> >Why it can't be matched by "GOOG0002"? [1] 
>> 
>> I'm not sure I understand this question. On my device the backlight has ACPI id "GGL0002", not "GOOG0002"?
>> 
>> >Have you tried to update your AP firmware?
>> 
>> I have not explicitly updated any firmware. It has whatever firmware Chromeos has installed.
>
>IIUC, if your AP firmware includes patch [1] (submitted in 2015), it should
>report "GOOG0002" instead of "GGL0002".
>
>[1]: https://review.coreboot.org/c/coreboot/+/11478

I can only assume that Google did not ever ship a firmware update for Link that included that patch. I believe this device has all ChromeOS updates applied, so it represents a stock system, though no longer supported, and aside from being in developer mode and having a Linux partition.

My firmware also reports the acpi device as GGL0001.

    $ cat /sys/bus/acpi/devices/GGL000*/hid
    GGL0001
    GGL0002

    $ cat /sys/bus/acpi/devices/GGL000*/path
    \CRHW
    \_SB_.KBLT

Is there any reason why we shouldn't do the same thing here for the keyboard backlight that was done in drivers/platform/chrome/chromeos_acpi.c and support both the legacy ID and the official ACPI ID? I see your commit 703e77134 which added GOOG0016 in addition to GGL0001.

-Brandon
Tzung-Bi Shih Nov. 14, 2024, 8:53 a.m. UTC | #5
On Wed, Nov 13, 2024 at 10:38:36PM -0600, Brandon Casey wrote:
> 
> 
> On November 13, 2024 8:07:42 PM CST, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
> >On Wed, Nov 13, 2024 at 02:44:55AM -0600, Brandon Casey wrote:
> >> 
> >> 
> >> On November 13, 2024 12:46:18 AM CST, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
> >> >On Tue, Nov 12, 2024 at 08:29:05PM -0600, Brandon Casey wrote:
> >> >> On the Google Pixel chromebook (aka Google Link), the keyboard led
> >> >> backlight has the ACPI device id "GGL0002".  Let's add it to the
> >> >> configuration array.
> >> >
> >> >Why it can't be matched by "GOOG0002"? [1] 
> >> 
> >> I'm not sure I understand this question. On my device the backlight has ACPI id "GGL0002", not "GOOG0002"?
> >> 
> >> >Have you tried to update your AP firmware?
> >> 
> >> I have not explicitly updated any firmware. It has whatever firmware Chromeos has installed.
> >
> >IIUC, if your AP firmware includes patch [1] (submitted in 2015), it should
> >report "GOOG0002" instead of "GGL0002".
> >
> >[1]: https://review.coreboot.org/c/coreboot/+/11478
> 
> I can only assume that Google did not ever ship a firmware update for Link that included that patch. I believe this device has all ChromeOS updates applied, so it represents a stock system, though no longer supported, and aside from being in developer mode and having a Linux partition.

Please try to wrap your message to fit to 80-column.  See [2].

[2]: https://people.kernel.org/tglx/notes-about-netiquette

> My firmware also reports the acpi device as GGL0001.
> 
>     $ cat /sys/bus/acpi/devices/GGL000*/hid
>     GGL0001
>     GGL0002
> 
>     $ cat /sys/bus/acpi/devices/GGL000*/path
>     \CRHW
>     \_SB_.KBLT
> 
> Is there any reason why we shouldn't do the same thing here for the keyboard backlight that was done in drivers/platform/chrome/chromeos_acpi.c and support both the legacy ID and the official ACPI ID? I see your commit 703e77134 which added GOOG0016 in addition to GGL0001.

- The most important reason: "GGL0002" is assigned to other purpose.
- For some reason, your machine is using a quite old firmware.  It makes
  less sense to me for supporting a should-be-deprecated PNP ID by [1].
- The case is different from 703e77134.  We prefer HID to PNP ID.
Brandon Casey Nov. 18, 2024, 6:58 p.m. UTC | #6
On November 14, 2024 2:53:47 AM CST, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
>On Wed, Nov 13, 2024 at 10:38:36PM -0600, Brandon Casey wrote:

>> I can only assume that Google did not ever ship a firmware update
>> for Link that included that patch. I believe this device has all
>> ChromeOS updates applied, so it represents a stock
>> system, though no longer supported, and aside from being in
>> developer mode and having a Linux partition.

>Please try to wrap your message to fit to 80-column.  See [2].

Sorry about that. I didn't realize my email client was doing that.

>> Is there any reason why we shouldn't do the same thing here for the 
>> keyboard backlight that was done in drivers/platform/chrome/chromeos_acpi.c
>> and support both the legacy ID and the official ACPI ID? I see your
>> commit 703e77134 which added GOOG0016 in addition to GGL0001.
>
>- The most important reason: "GGL0002" is assigned to other purpose.

Ah, so this seems to be the crux of it :-(

>- For some reason, your machine is using a quite old firmware.  It makes
>  less sense to me for supporting a should-be-deprecated PNP ID by [1].

You seem to be suggesting that my machine is in some odd/abnormal state.
Do you know whether Google has actually ever pushed an updated firmware
for this device (LINK) that included the patch referenced in [1]? If not, then
shouldn't my machine be considered to be in the "normal" state, not the
oddity?

-Brandon
Tzung-Bi Shih Nov. 21, 2024, 12:54 a.m. UTC | #7
On Mon, Nov 18, 2024 at 12:58:52PM -0600, Brandon Casey wrote:
> You seem to be suggesting that my machine is in some odd/abnormal state.
> Do you know whether Google has actually ever pushed an updated firmware
> for this device (LINK) that included the patch referenced in [1]? If not, then
> shouldn't my machine be considered to be in the "normal" state, not the
> oddity?

No, I don't know.
diff mbox series

Patch

diff --git a/drivers/platform/chrome/cros_kbd_led_backlight.c b/drivers/platform/chrome/cros_kbd_led_backlight.c
index 78097c8a4966..5b7766feb10d 100644
--- a/drivers/platform/chrome/cros_kbd_led_backlight.c
+++ b/drivers/platform/chrome/cros_kbd_led_backlight.c
@@ -268,6 +268,7 @@  static int keyboard_led_probe(struct platform_device *pdev)
 #ifdef CONFIG_ACPI
 static const struct acpi_device_id keyboard_led_acpi_match[] = {
 	{ "GOOG0002", (kernel_ulong_t)&keyboard_led_drvdata_acpi },
+	{ "GGL0002", (kernel_ulong_t)&keyboard_led_drvdata_acpi },
 	{ }
 };
 MODULE_DEVICE_TABLE(acpi, keyboard_led_acpi_match);