linux-kernel - Re: [regression] stm32mp1xx based targets stopped entering suspend if pwm-leds exist

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Tue, 16 Apr 2024 10:28:36 +0200
From: "Linux regression tracking (Thorsten Leemhuis)"
 <regressions@...mhuis.info>
To: Uwe Kleine-König <u.kleine-koenig@...gutronix.de>,
 Linux regressions mailing list <regressions@...ts.linux.dev>
Cc: Lee Jones <lee@...nel.org>, linux-leds@...r.kernel.org,
 linux-pwm@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>
Subject: Re: [regression] stm32mp1xx based targets stopped entering suspend if
 pwm-leds exist

On 06.03.24 09:18, Uwe Kleine-König wrote:
> On Wed, Mar 06, 2024 at 08:05:15AM +0100, Linux regression tracking (Thorsten Leemhuis) wrote:
>>
>> Uwe, I noticed a report about a regression in bugzilla.kernel.org that
>> apparently is caused by a change of yours. As many (most?) kernel
>> developers don't keep an eye on it, I decided to forward it by mail.
>>
>> Note, you have to use bugzilla to reach the reporter, as I sadly[1] can
>> not CCed them in mails like this.
>>
>> Quoting from https://bugzilla.kernel.org/show_bug.cgi?id=218559 :
>>
>>> Commit 76fe464c8e64e71b2e4af11edeef0e5d85eeb6aa ("leds: pwm: Don't
>>> disable the PWM when the LED should be off") prevents stm32mp1xx based
>>> targets from entering suspend if pwm-leds exist, as the stm32 PWM driver
>>> refuses to enter suspend if any PWM channels are still active ("PWM 0
>>> still in use by consumer" see stm32_pwm_suspend in drivers/pwm/stm32-pwm.c).
>>>
>>> Reverting the mentioned commit fixes this behaviour but I'm not
>>> certain if this is a problem with stm32-pwm or pwm-leds (what is the
>>> usual behaviour for suspend with active PWM channels?).
> 
> I'd assume the following patch fixes this report. I didn't test it
> though.

Jakob confirmed it helped in the bugzilla ticket. But the patch since
then didn't make any progress afaics -- or did it and I just missed it
in my search?

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)
--
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
If I did something stupid, please tell me, as explained on that page.

#regzbot poke

> ---->8----
> From: Uwe Kleine-König <u.kleine-koenig@...gutronix.de>
> Subject: [PATCH] leds: pwm: Disable PWM when going to suspend
> 
> On stm32mp1xx based machines (and others) a PWM consumer has to disable
> the PWM because an enabled PWM refuses to suspend. So check the
> LED_SUSPENDED flag and depending on that set the .enabled property.
> 
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=218559
> Fixes: 76fe464c8e64 ("leds: pwm: Don't disable the PWM when the LED should be off")
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@...gutronix.de>
> ---
>  drivers/leds/leds-pwm.c | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/leds/leds-pwm.c b/drivers/leds/leds-pwm.c
> index 4e3936a39d0e..e1b414b40353 100644
> --- a/drivers/leds/leds-pwm.c
> +++ b/drivers/leds/leds-pwm.c
> @@ -53,7 +53,13 @@ static int led_pwm_set(struct led_classdev *led_cdev,
>  		duty = led_dat->pwmstate.period - duty;
>  
>  	led_dat->pwmstate.duty_cycle = duty;
> -	led_dat->pwmstate.enabled = true;
> +	/*
> +	 * Disabling a PWM doesn't guarantee that it emits the inactive level.
> +	 * So keep it on. Only for suspending the PWM should be disabled because
> +	 * otherwise it refuses to suspend. The possible downside is that the
> +	 * LED might stay (or even go) on.
> +	 */
> +	led_dat->pwmstate.enabled = !(led_cdev->flags & LED_SUSPENDED);
>  	return pwm_apply_might_sleep(led_dat->pwm, &led_dat->pwmstate);
>  }
> 
> base-commit: 15facbd7bd3dbfa04721cb71e69954eb4686cb9e
> ---->8----
>