[PATCH v2 1/2] mfd: wm8994: Fix driver operation if loaded as modules

[PATCH v2 1/2] mfd: wm8994: Fix driver operation if loaded as modules

[Marek Szyprowski]

WM8994 chip has built-in regulators, which might be used for chip
operation. They are controlled by a separate wm8994-regulator driver,
which should be loaded before this driver calls regulator_get(), because
that driver also provides consumer-supply mapping for the them. If that
driver is not yet loaded, regulator core substitute them with dummy
regulator, what breaks chip operation, because the built-in regulators are
never enabled. Fix this by annotating this driver with MODULE_SOFTDEP()
"pre" dependency to "wm8994_regulator" module.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
v2:
- replaced request_module() call with MODULE_SOFTDEP() annotation
---
 drivers/mfd/wm8994-core.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/mfd/wm8994-core.c b/drivers/mfd/wm8994-core.c
index 1e9fe7d92597..737dede4a95c 100644
--- a/drivers/mfd/wm8994-core.c
+++ b/drivers/mfd/wm8994-core.c
@@ -690,3 +690,4 @@ module_i2c_driver(wm8994_i2c_driver);
 MODULE_DESCRIPTION("Core support for the WM8994 audio CODEC");
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Mark Brown <broonie@opensource.wolfsonmicro.com>");
+MODULE_SOFTDEP("pre: wm8994_regulator");
-- 
2.17.1

[PATCH v2 2/2] mfd: wm8994: Fix unbalanced calls to regulator_bulk_disable()

[Marek Szyprowski]

When runtime PM is enabled, regulators are being controlled by the
driver's suspend and resume callbacks. They are also unconditionally
enabled at driver's probe(), and disabled in remove() functions. Add
more calls to runtime PM framework to ensure that the device's runtime
PM state matches the regulators state:
1. at the end of probe() function: set runtime PM state to active, so
there will be no spurious call to resume();
2. in remove(), ensure that resume() is called before disabling runtime PM
management and unconditionally disabling the regulators.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
 drivers/mfd/wm8994-core.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/mfd/wm8994-core.c b/drivers/mfd/wm8994-core.c
index 737dede4a95c..69d973ec42bf 100644
--- a/drivers/mfd/wm8994-core.c
+++ b/drivers/mfd/wm8994-core.c
@@ -584,6 +584,7 @@ static int wm8994_device_init(struct wm8994 *wm8994, int irq)
 		goto err_irq;
 	}
 
+	pm_runtime_set_active(wm8994->dev);
 	pm_runtime_enable(wm8994->dev);
 	pm_runtime_idle(wm8994->dev);
 
@@ -603,7 +604,9 @@ static int wm8994_device_init(struct wm8994 *wm8994, int irq)
 
 static void wm8994_device_exit(struct wm8994 *wm8994)
 {
+	pm_runtime_get_sync(wm8994->dev);
 	pm_runtime_disable(wm8994->dev);
+	pm_runtime_put_noidle(wm8994->dev);
 	wm8994_irq_exit(wm8994);
 	regulator_bulk_disable(wm8994->num_supplies, wm8994->supplies);
 	regulator_bulk_free(wm8994->num_supplies, wm8994->supplies);
-- 
2.17.1

Re: [PATCH v2 1/2] mfd: wm8994: Fix driver operation if loaded as modules

[Charles Keepax]

On Wed, Feb 26, 2020 at 11:08:01AM +0100, Marek Szyprowski wrote:
> WM8994 chip has built-in regulators, which might be used for chip
> operation. They are controlled by a separate wm8994-regulator driver,
> which should be loaded before this driver calls regulator_get(), because
> that driver also provides consumer-supply mapping for the them. If that
> driver is not yet loaded, regulator core substitute them with dummy
> regulator, what breaks chip operation, because the built-in regulators are
> never enabled. Fix this by annotating this driver with MODULE_SOFTDEP()
> "pre" dependency to "wm8994_regulator" module.
> 
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---

Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>

Be good to ensure patches@opensource.cirrus.com is CCed on
patches for the old Wolfson CODECs.

Thanks,
Charles

Re: [PATCH v2 2/2] mfd: wm8994: Fix unbalanced calls to regulator_bulk_disable()

[Charles Keepax]

On Wed, Feb 26, 2020 at 11:08:02AM +0100, Marek Szyprowski wrote:
> When runtime PM is enabled, regulators are being controlled by the
> driver's suspend and resume callbacks. They are also unconditionally
> enabled at driver's probe(), and disabled in remove() functions. Add
> more calls to runtime PM framework to ensure that the device's runtime
> PM state matches the regulators state:
> 1. at the end of probe() function: set runtime PM state to active, so
> there will be no spurious call to resume();
> 2. in remove(), ensure that resume() is called before disabling runtime PM
> management and unconditionally disabling the regulators.
> 
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---

Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>

Thanks,
Charles