Skip to contentSkip to conversation

pinctrl: intel: platform: Add INTC10B5 ACPI ID for Lunar Lake

PATCH In review Vasu 2 participants 15 messages Apr 4 – Apr 5

Subsystems (5)
lore
Share & tools
lore

Vasu

author
#0 lore raw

PATCH 1/2 pinctrl: intel: platform: Add INTC10B5 ACPI ID for Lunar Lake

From: Vasu Hansalia <[email protected]>

Intel Lunar Lake platforms (Core Ultra 200V series) use INTC10B5 as the ACPI hardware ID for the GPIO controller that manages power to the integrated MIPI camera sensor.

Without this entry, the pinctrl-intel-platform driver does not bind to the GPIO controller, causing int3472-discrete to report:

int3472-discrete INT3472:00: cannot find GPIO chip INTC10B5:00, deferring

This prevents the camera sensor from being powered on entirely.

Tested on: Dell Pro 14 Premium PA14250, Intel Core Ultra 7 266V

           (Lunar Lake), Ubuntu 24.04, kernel 6.11.0-1016-oem

---

drivers/pinctrl/intel/pinctrl-intel-platform.c +1 -0
@@ -204,6 +204,7 @@
 
 static const struct acpi_device_id intel_platform_pinctrl_acpi_match[] = {
 	{ "INTC105F" },
+	{ "INTC10B5" },
 	{ }
 };
 MODULE_DEVICE_TABLE(acpi, intel_platform_pinctrl_acpi_match);
author
#1 lore raw

PATCH 2/2 platform/x86: int3472: Handle GPIO types 0x02 and 0x12 for Lunar Lake

From: Vasu Hansalia <[email protected]>

Intel Lunar Lake platforms (Core Ultra 200V series) use two GPIO types in their INT3472 ACPI tables that are not handled by the discrete driver:

- Type 0x02: Strobe GPIO, used for IR LED/flash control. Map to sensor.
- Type 0x12: AVDD/sensor-supply GPIO, used on Lunar Lake to control sensor power. Treat like RESET/POWERDOWN and map to sensor.

Without these handlers the driver logs:

int3472-discrete INT3472:00: GPIO type 0x02 unknown; the sensor may not work int3472-discrete INT3472:00: GPIO type 0x12 unknown; the sensor may not work

Type 0x12 handling was confirmed by inspecting the OEM kernel module for Ubuntu 24.04 (6.11.0-1016-oem) which handles this type by mapping it directly to the sensor via skl_int3472_fill_gpiod_lookup().

Tested on: Dell Pro 14 Premium PA14250, Intel Core Ultra 7 266V

           (Lunar Lake), Ubuntu 24.04, kernel 6.11.0-1016-oem
           Camera confirmed working at 1280x720 30fps after fix.

---

drivers/platform/x86/intel/int3472/discrete.c +7 -0
@@ -239,6 +239,7 @@
 	switch (type) {
 	case INT3472_GPIO_TYPE_RESET:
 	case INT3472_GPIO_TYPE_POWERDOWN:
+	case 0x12: /* Lunar Lake AVDD/sensor-supply GPIO */
 		ret = skl_int3472_map_gpio_to_sensor(int3472, agpio, func, polarity);
 		if (ret)
 			err_msg = "Failed to map GPIO pin to sensor\n";
@@ -278,6 +279,12 @@
 			break;
 		}
 		break;
+	case 0x02:
+		/* Strobe GPIO - used for IR LED/flash, map to sensor */
+		ret = skl_int3472_map_gpio_to_sensor(int3472, agpio, func, polarity);
+		if (ret)
+			err_msg = "Failed to map strobe GPIO to sensor\n";
+		break;
 	default:
 		dev_warn(int3472->dev,
 			 "GPIO type 0x%02x unknown; the sensor may not work\n",
#2 lore raw

Re: [PATCH 2/2] platform/x86: int3472: Handle GPIO types 0x02 and 0x12 for Lunar Lake

Quoted 12 lines · Vasu

Intel Lunar Lake platforms (Core Ultra 200V series) use two GPIO types in their INT3472 ACPI tables that are not handled by the discrete driver:

- Type 0x02: Strobe GPIO, used for IR LED/flash control. Map to sensor.
- Type 0x12: AVDD/sensor-supply GPIO, used on Lunar Lake to control sensor power. Treat like RESET/POWERDOWN and map to sensor.

Without these handlers the driver logs:

int3472-discrete INT3472:00: GPIO type 0x02 unknown; the sensor may not work int3472-discrete INT3472:00: GPIO type 0x12 unknown; the sensor may not work

Quoted 7 lines

Type 0x12 handling was confirmed by inspecting the OEM kernel module for Ubuntu 24.04 (6.11.0-1016-oem) which handles this type by mapping it directly to the sensor via skl_int3472_fill_gpiod_lookup().

Tested on: Dell Pro 14 Premium PA14250, Intel Core Ultra 7 266V

           (Lunar Lake), Ubuntu 24.04, kernel 6.11.0-1016-oem
           Camera confirmed working at 1280x720 30fps after fix.

STROBE is about to be added by another series. The 0x12 needs a proper define.

Signature · 2 lines
With Best Regards,
Andy Shevchenko
author
#3 lore raw

Re: [PATCH 2/2] platform/x86: int3472: Handle GPIO types 0x02 and 0x12 for Lunar Lake

From: Vasu Hansalia <[email protected]>

Thank you for the feedback Andy.

Could you point me to the series that is adding STROBE support so I can avoid duplicating that work?

For 0x12, I will add a proper define. Based on the OEM kernel module behaviour, would INT3472_GPIO_TYPE_AVDD be an appropriate name, or do you have a preferred name for this type?

Regards,
Vasu Hansalia

#4 lore raw

Re: [PATCH 2/2] platform/x86: int3472: Handle GPIO types 0x02 and 0x12 for Lunar Lake

Quoted 4 lines · Vasu

Thank you for the feedback Andy.

Could you point me to the series that is adding STROBE support so I can avoid duplicating that work?

Sure, you can find it on lore archive:
https://lore.kernel.org/platform-driver-x86/[email protected]/
(It took less than a minute to find :-)

Quoted 3 lines

For 0x12, I will add a proper define. Based on the OEM kernel module behaviour, would INT3472_GPIO_TYPE_AVDD be an appropriate name, or do you have a preferred name for this type?

Better to follow the name in accordance with the specification(s).

Signature · 2 lines
With Best Regards,
Andy Shevchenko
author
#5 lore raw

Re: [PATCH 2/2] platform/x86: int3472: Handle GPIO types 0x02 and 0x12 for Lunar Lake

From: Vasu Hansalia <[email protected]>

Thank you for the pointer to the STROBE series.

Based on the specification, the correct name for type 0x12 appears to be "handshake". I will use INT3472_GPIO_TYPE_HANDSHAKE in v2.

I will prepare v2 dropping type 0x02 (handled by the STROBE series) and adding the proper #define for type 0x12.

Regards,
Vasu Hansalia

author
#6 lore raw

Re: [PATCH 2/2] platform/x86: int3472: Handle GPIO types 0x02 and 0x12 for Lunar Lake

In reply to Andy Shevchenko

From: Vasu Hansalia <[email protected]>

After checking the mainline kernel source, I found that INT3472_GPIO_TYPE_HANDSHAKE (0x12) is already defined and fully handled in mainline. So no v2 patch is needed for mainline.

The fix just needs to be backported to the Ubuntu OEM kernels (6.11-oem, 6.17-oem) where affected users are running Ubuntu 24.04.

Should I send a backport request to the Ubuntu kernel team, or is there another process you would recommend?

Regards,
Vasu Hansalia

#7 lore raw

Re: [PATCH 2/2] platform/x86: int3472: Handle GPIO types 0x02 and 0x12 for Lunar Lake

Quoted 9 lines · Vasu

After checking the mainline kernel source, I found that INT3472_GPIO_TYPE_HANDSHAKE (0x12) is already defined and fully handled in mainline. So no v2 patch is needed for mainline.

The fix just needs to be backported to the Ubuntu OEM kernels (6.11-oem, 6.17-oem) where affected users are running Ubuntu 24.04.

Should I send a backport request to the Ubuntu kernel team, or is there another process you would recommend?

It's documented:
Documentation/process/stable-kernel-rules.rst

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#8 lore raw
In reply to Vasu
Quoted 7 lines · Vasu

Intel Lunar Lake platforms (Core Ultra 200V series) use INTC10B5 as the ACPI hardware ID for the GPIO controller that manages power to the integrated MIPI camera sensor.

Without this entry, the pinctrl-intel-platform driver does not bind to the GPIO controller, causing int3472-discrete to report:

Quoted 6 lines

int3472-discrete INT3472:00: cannot find GPIO chip INTC10B5:00, deferring

This prevents the camera sensor from being powered on entirely.

Tested on: Dell Pro 14 Premium PA14250, Intel Core Ultra 7 266V

           (Lunar Lake), Ubuntu 24.04, kernel 6.11.0-1016-oem

Huh?!

The currently mentioned ID is CID in the driver, the BIOS has to set _CID properly. Can you file the bug against BIOS and make OEM aware of this?

Signature · 2 lines
With Best Regards,
Andy Shevchenko
author
#9 lore raw

From: Vasu Hansalia <[email protected]>

Thank you for the feedback Andy.

Understood - this is a BIOS bug where the firmware is not setting _CID properly for INTC10B5. I will file a bug with Dell to get their BIOS fixed.

Should I drop this patch entirely, or would a kernel workaround still be acceptable with a comment noting the BIOS bug?

Regards,
Vasu Hansalia

#10 lore raw
Quoted 8 lines · Vasu

Thank you for the feedback Andy.

Understood - this is a BIOS bug where the firmware is not setting _CID properly for INTC10B5. I will file a bug with Dell to get their BIOS fixed.

Should I drop this patch entirely, or would a kernel workaround still be acceptable with a comment noting the BIOS bug?

Is this machine already on the market?

Signature · 2 lines
With Best Regards,
Andy Shevchenko
author
#11 lore raw

From: Vasu Hansalia <[email protected]>

Yes, multiple PA14250 variants are already on the market and shipping. The Core Ultra 5 variants ship with Ubuntu 24.04 and are Ubuntu-certified. The Dell Pro 13 Plus (PB13250) with Core Ultra 7 266V is also Ubuntu-certified. My device, the PA14250 with Core Ultra 7 266V, ships with Windows 11.

All of these Lunar Lake Dell laptops share the same ACPI BIOS issue with INTC10B5 not having _CID set properly.

Regards,
Vasu Hansalia

#12 lore raw
Quoted 9 lines · Vasu

Yes, multiple PA14250 variants are already on the market and shipping. The Core Ultra 5 variants ship with Ubuntu 24.04 and are Ubuntu-certified. The Dell Pro 13 Plus (PB13250) with Core Ultra 7 266V is also Ubuntu-certified. My device, the PA14250 with Core Ultra 7 266V, ships with Windows 11.

All of these Lunar Lake Dell laptops share the same ACPI BIOS issue with INTC10B5 not having _CID set properly.

Okay, you misinterpreted the SoC's GPIO with external component. The problem is you forgot (or distro, in that case file a bug in their respective bugzilla) to have the proper LJCA driver(s) be enabled.

Another possibility that they are using old kernel.

You can spent 5 minutes by checking yourself:

$ git grep -n -w INTC10B5
drivers/gpio/gpio-usbio.c:31:   { "INTC10B5" }, /* LNL */
drivers/usb/misc/usb-ljca.c:182:        { "INTC10B5" }, /* LNL LJCA GPIO */

Overall, your series is not needed at all.

Signature · 2 lines
With Best Regards,
Andy Shevchenko
author
#13 lore raw

From: Vasu Hansalia <[email protected]>

Thank you for the clarification Andy.

Confirmed — the LJCA drivers (gpio-usbio and usb-ljca) are present and loaded on the system:

  $ lsmod | grep -E "usbio|ljca"
  i2c_usbio    12288  0
  gpio_usbio   12288  2
  usbio        20480  2 gpio_usbio,i2c_usbio

You are correct, Patch 1 is not needed. I will drop it.

For Patch 2, I will follow stable-kernel-rules.rst to request the backport of INT3472_GPIO_TYPE_HANDSHAKE to the Ubuntu OEM kernels.

Thank you for your patience and guidance throughout this process.

Regards,
Vasu Hansalia

author
#14 lore raw
In reply to Andy Shevchenko

From: Vasu Hansalia <[email protected]>

Hi Andy,

Thank you for confirming this is a BIOS bug.

I have filed the issue in the following places:

1. Ubuntu Launchpad bug tracker:
https://bugs.launchpad.net/ubuntu/+source/linux-firmware/+bug/2131016

2. Intel ipu7-drivers issue tracker:
https://github.com/intel/ipu7-drivers/issues/26

Unfortunately I could not find an official Dell bug tracker or public repository where BIOS/firmware bugs can be reported. Dell does not appear to have a public GitHub repository for BIOS bugs unlike other vendors.

Since Intel has a direct relationship with Dell OEM, would it be possible for you or someone at Intel to escalate this BIOS _CID issue to Dell's firmware team? This affects all Lunar Lake Dell laptops (PA14250, XPS 13 9350) running Linux.

The affected users are currently unable to use the camera in any native Linux application. A BIOS fix from Dell would resolve this permanently without requiring a kernel workaround.

Thank you for your help.

Regards,
Vasu Hansalia