drivers: gpio: intel: Enable support for new ACPI GINF method #95920

akanisetti · 2025-09-12T17:26:51Z

This PR adds GPIO support for Intel Panther Lake H SoC, implementing a new ACPI GINF method for GPIO resource enumeration and adding corresponding device tree definitions.

Adds GPIO device tree definitions for Panther Lake H with new ACPI GINF method support
Implements enhanced ACPI GPIO resource enumeration with support for both old and new GINF methods
Adds GPIO configuration overlay for test board intel_ptl_h_crb

edersondisouza · 2025-09-22T16:17:52Z

drivers/gpio/Kconfig.intel

+	  as it is observed that the offset varies from
+	  platform to platform.
+
+config GPIO_INTEL_INT_STAT_OFFSET


I wonder if these two actually belong to devicetree, as they are per platform (so they would be added to dts/bindings/gpio/intel,gpio.yaml).

yes, adding it as a dts parameter is more suitable but the value is common for all the GPIO instances of a platform. That's why, it is added as a CONFIG to be updated once for a platform instead of adding same value for all the instances in dts file.

Hmmm... not sure how this is usually (supposed to be) handled. @henrikbrixandersen thoughts?

Typically, we would describe that in the devicetree instance, but there's no good way of setting the same value across a number of devicetree nodes. Only thing I can think of is having a parent node with this property, but that would be kind of superficial.

Is this GPIO controller supported in Linux? if yes, how does the Linux kernel handle/provide this offset?

This GPIO controller is supported in Linux, a different ACPI method is used to fetch this offset directly instead of deriving from another offset like what we are doing hereLinux Ref. That ACPI method is not enabled in zephyr.

edersondisouza · 2025-09-22T16:41:11Z

dts/bindings/gpio/intel,gpio.yaml

    type: int
    description: Pin offset of this GPIO entry

+  acpi-ginf:


Not familiar with ACPI at all, but it seems weird that a boolean describes a version. Maybe rename it to acpi-ginf-3-param or so?

Will update it to acpi-ginf-3-param

edersondisouza · 2025-09-22T16:42:56Z

soc/intel/common/soc_gpio.c

 	acpi_child = acpi_device_get(hid, uid);
 	if (!acpi_child) {
-		printk("acpi_device_get failed\n");
+		LOG_ERR("acpi_device_get failed");


Nitpick: maybe have these s/printk/LOG_ERR on a different patch, so it doesn't distract from the functional changes?

Shall I add this change in a different comment in this PR?

That would be ideal, yes (assuming you meant "commit").

Yes, commit

edersondisouza · 2025-10-06T18:24:24Z

@akanisetti will you update this PR?

akanisetti · 2025-10-06T18:36:33Z

@akanisetti will you update this PR?

Hi @edersondisouza, I am waiting for @henrikbrixandersen's response regarding #95920 (comment).
If the approach is finalized, I could add all changes at a time, instead of adding changes one by one.

henrikbrixandersen · 2025-10-06T18:43:15Z

@akanisetti will you update this PR?

Hi @edersondisouza, I am waiting for @henrikbrixandersen's response regarding #95920 (comment). If the approach is finalized, I could add all changes at a time, instead of adding changes one by one.

I wasn't aware you were waiting for a response from me, my apologies. I don't think I have a good solution on how to obtain this offset in Zephyr.

akanisetti · 2025-10-06T18:55:14Z

@akanisetti will you update this PR?

Hi @edersondisouza, I am waiting for @henrikbrixandersen's response regarding #95920 (comment). If the approach is finalized, I could add all changes at a time, instead of adding changes one by one.

I wasn't aware you were waiting for a response from me, my apologies. I don't think I have a good solution on how to obtain this offset in Zephyr.

No worries @henrikbrixandersen.
If the approach of using Kconfigs to update the offset is not appropriate, I will take the other approach of adding it as a dts parameter and add it in all the instances. Just that same value has to be added for all GPIO instances of a board.
Please confirm if this (dts-parameter) approach is okay if not perfect for this scenario @henrikbrixandersen @edersondisouza.

henrikbrixandersen · 2025-10-06T18:57:27Z

If the approach of using Kconfigs to update the offset is not appropriate, I will take the other approach of adding it as a dts parameter and add it in all the instances. Just that same value has to be added for all GPIO instances of a board.
Please confirm if this approach is okay if not perfect for this scenario @henrikbrixandersen @edersondisouza.

I would prefer a devicetree-based solution for specifying the offset.

akanisetti · 2025-10-06T19:00:10Z

If the approach of using Kconfigs to update the offset is not appropriate, I will take the other approach of adding it as a dts parameter and add it in all the instances. Just that same value has to be added for all GPIO instances of a board.
Please confirm if this approach is okay if not perfect for this scenario @henrikbrixandersen @edersondisouza.

I would prefer a devicetree-based solution for specifying the offset.

Thank you @henrikbrixandersen,
@edersondisouza I will update the PR with all the changes within 24hours.

Enable support for ACPI_RESOURCE_TYPE_ADDRESS64 in acpi_device_mmio_get to fetch 64bit address from resource of an ACPI device. Signed-off-by: Anisetti Avinash Krishna <[email protected]>

Enable support for latest GINF method which requires 3 paramters for each GPIO group and enables gpio support for intel_ptl_h platform. Signed-off-by: Anisetti Avinash Krishna <[email protected]>

Replace printk with LOG_ERR by adding a log module. Signed-off-by: Anisetti Avinash Krishna <[email protected]>

sonarqubecloud · 2025-10-07T17:10:52Z

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

akanisetti force-pushed the gpio_ptlh_main branch 4 times, most recently from 684d7fc to 8812842 Compare September 15, 2025 06:10

akanisetti marked this pull request as ready for review September 15, 2025 06:57

zephyrbot added platform: Intel Intel Corporation area: ACPI ACPI support platform: X86 x86 and x86-64 area: GPIO labels Sep 15, 2025

zephyrbot requested review from ceolin, dcpleung, edersondisouza, henrikbrixandersen, mnkp and teburd September 15, 2025 06:58

edersondisouza reviewed Sep 22, 2025

View reviewed changes

edersondisouza self-assigned this Oct 6, 2025

lib: acpi: Enable support for ACPI_RESOURCE_TYPE_ADDRESS64

08a6c2f

Enable support for ACPI_RESOURCE_TYPE_ADDRESS64 in acpi_device_mmio_get to fetch 64bit address from resource of an ACPI device. Signed-off-by: Anisetti Avinash Krishna <[email protected]>

akanisetti force-pushed the gpio_ptlh_main branch from 8812842 to 9dd44e2 Compare October 7, 2025 04:57

zephyrbot added area: Tests Issues related to a particular existing or missing test area: Devicetree Bindings area: Boards/SoCs labels Oct 7, 2025

zephyrbot requested review from najumon1980 and nashif October 7, 2025 04:58

akanisetti added 2 commits October 7, 2025 21:08

drivers: gpio: Enable support for latest GINF method

587b40f

Enable support for latest GINF method which requires 3 paramters for each GPIO group and enables gpio support for intel_ptl_h platform. Signed-off-by: Anisetti Avinash Krishna <[email protected]>

soc: intel: common: Replace printk with LOG_ERR

2e4aaa6

Replace printk with LOG_ERR by adding a log module. Signed-off-by: Anisetti Avinash Krishna <[email protected]>

akanisetti force-pushed the gpio_ptlh_main branch from 9dd44e2 to 2e4aaa6 Compare October 7, 2025 17:03

akanisetti requested a review from edersondisouza October 7, 2025 17:44

edersondisouza approved these changes Oct 7, 2025

View reviewed changes

nashif approved these changes Oct 8, 2025

View reviewed changes

cfriedt merged commit ea1a839 into zephyrproject-rtos:main Oct 8, 2025
26 checks passed

drivers: gpio: intel: Enable support for new ACPI GINF method #95920

drivers: gpio: intel: Enable support for new ACPI GINF method #95920

Uh oh!

Conversation

akanisetti commented Sep 12, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

akanisetti Sep 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

edersondisouza commented Oct 6, 2025

Uh oh!

akanisetti commented Oct 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

henrikbrixandersen commented Oct 6, 2025

Uh oh!

akanisetti commented Oct 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

henrikbrixandersen commented Oct 6, 2025

Uh oh!

akanisetti commented Oct 6, 2025

Uh oh!

sonarqubecloud bot commented Oct 7, 2025

Quality Gate passed

Uh oh!

Uh oh!

Uh oh!

akanisetti Sep 29, 2025 •

edited

Loading

akanisetti commented Oct 6, 2025 •

edited

Loading

akanisetti commented Oct 6, 2025 •

edited

Loading