[PATCH 0/2] Wakeup the i2c-omap controller during suspend stage

[PATCH 0/2] Wakeup the i2c-omap controller during suspend stage

[Thomas Richard]

The patches of this series were originally in the series "Add suspend to
ram support for PCIe on J7200" [1].
There is no changes compared to the patches in the series [1].

The goal is to wakeup the controller during the suspend stage to be able
to use it during suspend_noirq stage.
As autosuspend is enabled, the controller can be suspended at suspend_noirq
stage. But runtime pm is disabled in suspend_late stage. So it's not
possible to wakeup the controller during suspend_noirq.
The trick is to force the wakeup of the controller during suspend stage.

[1] https://lore.kernel.org/all/20240102-j7200-pcie-s2r-v6-0-4656ef6e6d66@bootlin.com/

Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
---
Thomas Richard (2):
      i2c: omap: switch to NOIRQ_SYSTEM_SLEEP_PM_OPS() and RUNTIME_PM_OPS()
      i2c: omap: wakeup the controller during suspend() callback

 drivers/i2c/busses/i2c-omap.c | 36 +++++++++++++++++++++++++++++-------
 1 file changed, 29 insertions(+), 7 deletions(-)
---
base-commit: 0719fea199617c798ad3a9158916066ac17219b2
change-id: 20240613-i2c-omap-wakeup-controller-during-suspend-895adf95a005

Best regards,
-- 
Thomas Richard <thomas.richard@bootlin.com>

[PATCH 1/2] i2c: omap: switch to NOIRQ_SYSTEM_SLEEP_PM_OPS() and RUNTIME_PM_OPS()

[Thomas Richard]

Replace SET_NOIRQ_SYSTEM_SLEEP_PM_OPS() and SET_RUNTIME_PM_OPS() by
NOIRQ_SYSTEM_SLEEP_PM_OPS() and RUNTIME_PM_OPS().
The usage of pm_ptr() and these more recent macros allows the compiler to
see and drop the dead code.
The unnecessary '__maybe_unused' annotations on PM functions can be
removed.

Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
---
 drivers/i2c/busses/i2c-omap.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index 30a5ea282a8b..410c8b37f768 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -1534,7 +1534,7 @@ static void omap_i2c_remove(struct platform_device *pdev)
 	pm_runtime_disable(&pdev->dev);
 }
 
-static int __maybe_unused omap_i2c_runtime_suspend(struct device *dev)
+static int omap_i2c_runtime_suspend(struct device *dev)
 {
 	struct omap_i2c_dev *omap = dev_get_drvdata(dev);
 
@@ -1560,7 +1560,7 @@ static int __maybe_unused omap_i2c_runtime_suspend(struct device *dev)
 	return 0;
 }
 
-static int __maybe_unused omap_i2c_runtime_resume(struct device *dev)
+static int omap_i2c_runtime_resume(struct device *dev)
 {
 	struct omap_i2c_dev *omap = dev_get_drvdata(dev);
 
@@ -1575,10 +1575,10 @@ static int __maybe_unused omap_i2c_runtime_resume(struct device *dev)
 }
 
 static const struct dev_pm_ops omap_i2c_pm_ops = {
-	SET_NOIRQ_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
-				      pm_runtime_force_resume)
-	SET_RUNTIME_PM_OPS(omap_i2c_runtime_suspend,
-			   omap_i2c_runtime_resume, NULL)
+	NOIRQ_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
+				  pm_runtime_force_resume)
+	RUNTIME_PM_OPS(omap_i2c_runtime_suspend,
+		       omap_i2c_runtime_resume, NULL)
 };
 
 static struct platform_driver omap_i2c_driver = {
@@ -1586,7 +1586,7 @@ static struct platform_driver omap_i2c_driver = {
 	.remove_new	= omap_i2c_remove,
 	.driver		= {
 		.name	= "omap_i2c",
-		.pm	= &omap_i2c_pm_ops,
+		.pm	= pm_ptr(&omap_i2c_pm_ops),
 		.of_match_table = of_match_ptr(omap_i2c_of_match),
 	},
 };

-- 
2.39.2

[PATCH 2/2] i2c: omap: wakeup the controller during suspend() callback

[Thomas Richard]

A device may need the controller up during suspend_noirq() or
resume_noirq().
But if the controller is autosuspended, there is no way to wakeup it during
suspend_noirq() or resume_noirq() because runtime pm is disabled at this
time.

The suspend() callback wakes up the controller, so it is available until
its suspend_noirq() callback (pm_runtime_force_suspend()).
During the resume, it's restored by resume_noirq() callback
(pm_runtime_force_resume()). Then resume() callback enables autosuspend.

So the controller is up during a little time slot in suspend and resume
sequences even if it's not used.

Reviewed-by: Andi Shyti <andi.shyti@kernel.org>
Acked-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Thomas Richard <thomas.richard@bootlin.com>
---
 drivers/i2c/busses/i2c-omap.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index 410c8b37f768..35a3f0a64986 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -1574,9 +1574,31 @@ static int omap_i2c_runtime_resume(struct device *dev)
 	return 0;
 }
 
+static int omap_i2c_suspend(struct device *dev)
+{
+	/*
+	 * If the controller is autosuspended, there is no way to wakeup it once
+	 * runtime pm is disabled (in suspend_late()).
+	 * But a device may need the controller up during suspend_noirq() or
+	 * resume_noirq().
+	 * Wakeup the controller while runtime pm is enabled, so it is available
+	 * until its suspend_noirq(), and from resume_noirq().
+	 */
+	return pm_runtime_resume_and_get(dev);
+}
+
+static int omap_i2c_resume(struct device *dev)
+{
+	pm_runtime_mark_last_busy(dev);
+	pm_runtime_put_autosuspend(dev);
+
+	return 0;
+}
+
 static const struct dev_pm_ops omap_i2c_pm_ops = {
 	NOIRQ_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
 				  pm_runtime_force_resume)
+	SYSTEM_SLEEP_PM_OPS(omap_i2c_suspend, omap_i2c_resume)
 	RUNTIME_PM_OPS(omap_i2c_runtime_suspend,
 		       omap_i2c_runtime_resume, NULL)
 };

-- 
2.39.2

Re: [PATCH 0/2] Wakeup the i2c-omap controller during suspend stage

[Andi Shyti]

Hi

On Thu, 13 Jun 2024 15:13:26 +0200, Thomas Richard wrote:
> The patches of this series were originally in the series "Add suspend to
> ram support for PCIe on J7200" [1].
> There is no changes compared to the patches in the series [1].
> 
> The goal is to wakeup the controller during the suspend stage to be able
> to use it during suspend_noirq stage.
> As autosuspend is enabled, the controller can be suspended at suspend_noirq
> stage. But runtime pm is disabled in suspend_late stage. So it's not
> possible to wakeup the controller during suspend_noirq.
> The trick is to force the wakeup of the controller during suspend stage.
> 
> [...]

Applied to i2c/i2c-host on

git://git.kernel.org/pub/scm/linux/kernel/git/andi.shyti/linux.git

Thank you,
Andi

Patches applied
===============
[1/2] i2c: omap: switch to NOIRQ_SYSTEM_SLEEP_PM_OPS() and RUNTIME_PM_OPS()
      commit: 659eb16247d04f442a014fe2dd8b7fa936dfccb5
[2/2] i2c: omap: wakeup the controller during suspend() callback
      commit: 8a2ebb775a497a95a79a8ac83027a4f159c36f32