Wakeup for IRQ1 should be disabled only in cases where i8042 had actually
enabled it, otherwise "wake_depth" for this IRQ will try to drop below zero
and there will be an unpleasant WARN() logged:
kernel: atkbd serio0: Disabling IRQ1 wakeup source to avoid platform firmware bug
kernel: ------------[ cut here ]------------
kernel: Unbalanced IRQ 1 wake disable
kernel: WARNING: CPU: 10 PID: 6431 at kernel/irq/manage.c:920 irq_set_irq_wake+0x147/0x1a0

To fix this call the PMC suspend handler only from the same set of
dev_pm_ops handlers as i8042_pm_suspend() is called, which currently means
just the ".suspend" handler.
Before this patch the driver used DEFINE_SIMPLE_DEV_PM_OPS() to define its
dev_pm_ops, which also called this handler on ".freeze" and ".poweroff".

To reproduce this issue try hibernating (S4) the machine after a fresh boot
without putting it into s2idle first.

Fixes: 8e60615e8932 ("platform/x86/amd: pmc: Disable IRQ1 wakeup for RN/CZN")
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
---

Changes from v1:
* Reword the commit message slightly, as suggested by Ilpo.

* Add Mario's "Reviewed-by:" tag.

 drivers/platform/x86/amd/pmc/pmc.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/platform/x86/amd/pmc/pmc.c b/drivers/platform/x86/amd/pmc/pmc.c
index 26b878ee5191..a254debb9256 100644
--- a/drivers/platform/x86/amd/pmc/pmc.c
+++ b/drivers/platform/x86/amd/pmc/pmc.c
@@ -947,6 +947,10 @@ static int amd_pmc_suspend_handler(struct device *dev)
 {
 	struct amd_pmc_dev *pdev = dev_get_drvdata(dev);
 
+	/*
+	 * Must be called only from the same set of dev_pm_ops handlers
+	 * as i8042_pm_suspend() is called: currently just from .suspend.
+	 */
 	if (pdev->disable_8042_wakeup && !disable_workarounds) {
 		int rc = amd_pmc_wa_irq1(pdev);
 
@@ -959,7 +963,9 @@ static int amd_pmc_suspend_handler(struct device *dev)
 	return 0;
 }
 
-static DEFINE_SIMPLE_DEV_PM_OPS(amd_pmc_pm, amd_pmc_suspend_handler, NULL);
+static const struct dev_pm_ops amd_pmc_pm = {
+	.suspend = amd_pmc_suspend_handler,
+};
 
 static const struct pci_device_id pmc_pci_ids[] = {
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, AMD_CPU_ID_PS) },

On Mon, 06 Jan 2025 18:40:34 +0100, Maciej S. Szmigiero wrote:

> Wakeup for IRQ1 should be disabled only in cases where i8042 had actually
> enabled it, otherwise "wake_depth" for this IRQ will try to drop below zero
> and there will be an unpleasant WARN() logged:
> kernel: atkbd serio0: Disabling IRQ1 wakeup source to avoid platform firmware bug
> kernel: ------------[ cut here ]------------
> kernel: Unbalanced IRQ 1 wake disable
> kernel: WARNING: CPU: 10 PID: 6431 at kernel/irq/manage.c:920 irq_set_irq_wake+0x147/0x1a0
> 
> [...]


Thank you for your contribution, it has been applied to my local
review-ilpo-fixes branch. Note it will show up in the public
platform-drivers-x86/review-ilpo-fixes branch only once I've pushed my
local branch there, which might take a while.

The list of commits applied:
[1/1] platform/x86/amd/pmc: Only disable IRQ1 wakeup where i8042 actually enabled it
      commit: dd410d784402c5775f66faf8b624e85e41c38aaf

--
 i.