This series was triggered by "Runtime PM usage count underflow!" when
unloading the module(s).
By testing the driver in various use cases and reading code it was
obvious that it could need some tiding up.

I'm still not 100% satisfied with suspend/resume is calling directly to
vddio_enable/disable. In my mind it should be managed by pm_runtime.

Signed-off-by: Sean Nyekjaer <sean@geanix.com>
---
Sean Nyekjaer (6):
      iio: imu: inv_icm42600: Use inv_icm42600_disable_vddio_reg()
      iio: imu: inv_icm42600: Use devm_regulator_get_enable() for vdd regulator
      iio: imu: inv_icm42600: Remove redundant error msg on regulator_disable()
      iio: imu: inv_icm42600: Simplify pm_runtime setup
      iio: imu: inv_icm42600: Drop redundant pm_runtime reinitialization in resume
      iio: imu: inv_icm42600: Avoid configuring if already pm_runtime suspended

 drivers/iio/imu/inv_icm42600/inv_icm42600.h      |  1 -
 drivers/iio/imu/inv_icm42600/inv_icm42600_core.c | 54 ++++++------------------
 2 files changed, 14 insertions(+), 41 deletions(-)
---
base-commit: 3e28fa06444e7031aba0b3552cce332b776fe267
change-id: 20250708-icm42pmreg-24d824d978c4

Best regards,
-- 
Sean Nyekjaer <sean@geanix.com>

On Wed, Jul 09, 2025 at 02:35:08PM +0200, Sean Nyekjaer wrote:
> This series was triggered by "Runtime PM usage count underflow!" when
> unloading the module(s).
> By testing the driver in various use cases and reading code it was
> obvious that it could need some tiding up.

> I'm still not 100% satisfied with suspend/resume is calling directly to
> vddio_enable/disable. In my mind it should be managed by pm_runtime.

This patch series touches a quite sensitive area and needs a good,
comprehensive tests (better done independently from the author).
What have been done so far?

-- 
With Best Regards,
Andy Shevchenko