PCIe WAKE# interrupt is needed for bringing back PCIe device state from
D3cold to D0.

This is pending from long time, there was two attempts done previously to
add WAKE# support[1], [2]. Those series tried to add support for legacy
interrupts along with WAKE#. Legacy interrupts are already available in
the latest kernel and we can ignore them. For the wake IRQ the series is
trying to use interrupts property define in the device tree.

This series is using gpio property instead of interrupts, from
gpio desc driver will allocate the dedicate IRQ.

Bjorn,
Can you take this series through PCI branch, once other subsystem
maintainers give us ACK. 

WAKE# is added in dts schema and merged based on this patch.
https://lore.kernel.org/all/20250515090517.3506772-1-krishna.chundru@oss.qualcomm.com/

[1]: https://lore.kernel.org/all/b2b91240-95fe-145d-502c-d52225497a34@nvidia.com/T/
[2]: https://lore.kernel.org/all/20171226023646.17722-1-jeffy.chen@rock-chips.com/

Signed-off-by: Krishna Chaitanya Chundru <krishna.chundru@oss.qualcomm.com>
---
Changes in v10:
- As sashiko pointed, shared irq has plenty of race conditions.
- So we are moving away from the shared IRQ patch and registering with
  dedicated wake irq only, as part of wake irq the link will come to D0
  as the parent controller driver will be runtime resume first and then
  pme service will kick in wake up correct endpoint driver.
- Removed device_init_wakeup() since it enabling wakeup explicitly,
  which is not intended as this should be set by endpoint driver only.
- Link to v9: https://lore.kernel.org/r/20260403-wakeirq_support-v9-0-1cbecf3b58d7@oss.qualcomm.com

Changes in v9:
- Call device_init_wakeup() only if
  dev_pm_set_dedicated_shared_wake_irq() succeeds (Mani).
- Change the IRQ_TYPE from IRQ_TYPE_EDGE_FALLING to IRQ_TYPE_LEVEL_LOW (Mani).
- Link to v8: https://lore.kernel.org/r/20260313-wakeirq_support-v8-0-48a0a702518a@oss.qualcomm.com

Changes in v8:
- Moved the stub functions under CONFIG_OF_IRQ(mani).
- Added the description of how dev_pm_set_dedicated_shared_wake_irq()
  works.
- Link to v7: https://lore.kernel.org/r/20260218-wakeirq_support-v7-0-0d4689830207@oss.qualcomm.com

Changes in v7:
- Updated the commit text (Mani).
- Couple of nits like using pci_err instead of dev_err,
  use platform_pci_configure_wake(), platform_pci_remove_wake() instead
  of calling directly calling pci_configure_of_wake_gpio() & pci_remove_of_wake_gpio() etc (Mani).
- Add a new fwnode_gpiod_get() API that wraps fwnode_gpiod_get_index(..0..), similar to
  devm_fwnode_gpiod_get() (Mani).
- Link to v6: https://lore.kernel.org/r/20251127-wakeirq_support-v6-0-60f581f94205@oss.qualcomm.com

Changes in v6:
- Change the name to dev_pm_set_dedicated_shared_wake_irq() and make the
  changes pointed by (Rafael). 
- Link to v5: https://lore.kernel.org/r/20251107-wakeirq_support-v5-0-464e17f2c20c@oss.qualcomm.com

Changes in v5:
- Enable WAKE# irq only when there is wake -gpios defined in its device
  tree node (Bjorn).
- For legacy bindings for direct atach check in root port if we haven't
  find the wake in the endpoint node.
- Instead of hooking wake in driver bound case, do it in the framework
  irrespective of the driver state (Bjorn).
- Link to v4: https://lore.kernel.org/r/20250801-wake_irq_support-v4-0-6b6639013a1a@oss.qualcomm.com

Changes in v4:
- Move wake from portdrv to core framework to endpoint (Bjorn).
- Added support for multiple WAKE# case (Bjorn). But traverse from
  endpoint upstream port to root port till you get WAKE#. And use
  IRQF_SHARED flag for requesting interrupts.
- Link to v3: https://lore.kernel.org/r/20250605-wake_irq_support-v3-0-7ba56dc909a5@oss.qualcomm.com

Changes in v3:
- Update the commit messages, function names etc as suggested by Mani.
- return wake_irq if returns error (Neil).
- Link to v2: https://lore.kernel.org/r/20250419-wake_irq_support-v2-0-06baed9a87a1@oss.qualcomm.com

Changes in v2:
- Move the wake irq teardown after pcie_port_device_remove
  and move of_pci_setup_wake_irq before pcie_link_rcec (Lukas)
- teardown wake irq in shutdown also.
- Link to v1: https://lore.kernel.org/r/20250401-wake_irq_support-v1-0-d2e22f4a0efd@oss.qualcomm.com

---
Krishna Chaitanya Chundru (2):
      gpio: Add fwnode_gpiod_get() helper
      PCI: Add support for PCIe WAKE# interrupt

 drivers/pci/of.c              | 72 +++++++++++++++++++++++++++++++++++++++++++
 drivers/pci/pci.c             | 11 +++++++
 drivers/pci/pci.h             |  2 ++
 drivers/pci/probe.c           |  2 ++
 drivers/pci/remove.c          |  1 +
 include/linux/gpio/consumer.h |  9 ++++++
 include/linux/of_pci.h        |  6 ++++
 include/linux/pci.h           |  2 ++
 8 files changed, 105 insertions(+)
---
base-commit: 2e68039281932e6dc37718a1ea7cbb8e2cda42e6
change-id: 20251104-wakeirq_support-f54c4baa18c5

Best regards,
--  
Krishna Chaitanya Chundru <krishna.chundru@oss.qualcomm.com>

On Mon, May 11, 2026 at 9:26 AM Krishna Chaitanya Chundru
<krishna.chundru@oss.qualcomm.com> wrote:
>
> PCIe WAKE# interrupt is needed for bringing back PCIe device state from
> D3cold to D0.
>
> This is pending from long time, there was two attempts done previously to
> add WAKE# support[1], [2]. Those series tried to add support for legacy
> interrupts along with WAKE#. Legacy interrupts are already available in
> the latest kernel and we can ignore them. For the wake IRQ the series is
> trying to use interrupts property define in the device tree.
>
> This series is using gpio property instead of interrupts, from
> gpio desc driver will allocate the dedicate IRQ.
>
> Bjorn,
> Can you take this series through PCI branch, once other subsystem
> maintainers give us ACK.
>

Actually, it's pretty early into the cycle, I would prefer to queue
patch 1/2 myself, I'll provide an immutable branch to Bjorn.

Bart

On Mon, 11 May 2026 12:55:36 +0530, Krishna Chaitanya Chundru wrote:
> PCIe WAKE# interrupt is needed for bringing back PCIe device state from
> D3cold to D0.
> 
> This is pending from long time, there was two attempts done previously to
> add WAKE# support[1], [2]. Those series tried to add support for legacy
> interrupts along with WAKE#. Legacy interrupts are already available in
> the latest kernel and we can ignore them. For the wake IRQ the series is
> trying to use interrupts property define in the device tree.
> 
> [...]

Applied, thanks!

[1/2] gpio: Add fwnode_gpiod_get() helper
      https://git.kernel.org/brgl/c/8a46bd2638f1ad6d1ed73dc3ab10919e67274738

Best regards,
-- 
Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>