Hello Greg,
On Thu, Jan 08, 2026 at 01:45:23PM +0100, Sean Nyekjaer wrote:
Commit 7346e7a058a2 ("pwm: stm32: Always do lazy disabling") triggered a regression where PWM polarity changes could be ignored.
stm32_pwm_set_polarity() was skipped due to a mismatch between the cached pwm->state.polarity and the actual hardware state, leaving the hardware polarity unchanged.
Fixes: 7edf7369205b ("pwm: Add driver for STM32 plaftorm") Cc: stable@vger.kernel.org # <= 6.12 Signed-off-by: Sean Nyekjaer sean@geanix.com Co-developed-by: Uwe Kleine-König ukleinek@kernel.org
Signed-off-by: Uwe Kleine-König ukleinek@kernel.org
This patch is only applicable for stable tree's <= 6.12
Can you please apply this patch to v6.12.x and older? The problem fixed here doesn't happen in mainline since commit deaba9cff809 ("pwm: stm32: Implementation of the waveform callbacks"). That mainline commit however is too intrusive to backport it to stable, not even considering its dependencies (e.g. 17e40c25158f ("pwm: New abstraction for PWM waveforms")).
I assume such an exception to the sable process is ok and the patch is simple enough to be easily reviewable? If not, tell me, then I can apply the patch and merge it (with `-s ours`) into my next PWM pull request to Linus :-)
Best regards Uwe