Re: [PATCH v2] pwm: tiehrpwm: Enable EHRPWM controller before setting configuration

["Rafael V. Volkmer" <rafael.v.volkmer@gmail.com>]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2008 bytes --]

Hello Uwe, Gokul,

Sorry for the late response, I was on vacation and away from my setup.

On Fri, 9 Jan 2026 23:53:22 +0100, Uwe Kleine-König wrote:
> That makes me think the problem isn't understood well yet and needs more
> research. @Rafael, does the problem reproduce for you with Gokul's
> recipe? (Or did you try that already? I understood your reply as "I
> didn't encounter the issue but also didn't test specifically for that.")

Right, my previous reply meant I hadn't explicitly targeted this issue yet.
I have now re-tested using Gokul's sysfs configuration sequence, but I still
cannot reproduce it on my setup.

> As I cannot reproduce the issue, can you please check if adding
>
>       pm_runtime_get_sync(pwmchip_parent(chip));
>
> to the probe function makes the problem disappear? Also please boot with
>
>       trace_event=pwm
>
> on the command line and provide the content of
> /sys/kernel/debug/tracing/trace after reproducing the problem.

Since I cannot reproduce the issue here, I can't validate whether adding
pm_runtime_get_sync() changes the behavior, and I don't have a failing
trace to share.

For reference, I ran the tests on an AM62P EVM using TI's default SDK
userspace, with a custom kernel on top, and U-Boot from the SDK. The
board was booted from SD card.

I used pwm1 instead of pwm0, since the PWM pin routed to the EVM 40-pin
header is ball B21 (SPI0_CLK / EHRPWM1_A). The signal was verified with a
logic analyzer at 24 MHz sampling rate.

This makes me suspect the behavior Gokul observed might depend on another
configuration interacting in parallel.

If possible, could Gokul try the same recipe on an AM62 EVM using TI's
default images and confirm whether the issue reproduces there? That is the
platform I am currently working with. This should either match the AM62P
results or help identify a relevant configuration difference.

Best regards,
Rafael V. Volkmer