linux-kernel - Re: [PATCH] leds: rgb: leds-qcom-lpg: Only enable TRILED when LPG is used by LED

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] [day] [month] [year] [list]

Message-ID: <t453czpauswdttsl5cqxwk5ryc7aau3bz4jfwfe4istkffgp43@ffoiicvnbxzh>
Date: Fri, 14 Nov 2025 08:56:34 -0600
From: Bjorn Andersson <andersson@...nel.org>
To: Fenglin Wu <fenglin.wu@....qualcomm.com>
Cc: kernel@....qualcomm.com, Lee Jones <lee@...nel.org>, 
	Pavel Machek <pavel@...nel.org>, Marijn Suijten <marijn.suijten@...ainline.org>, 
	Subbaraman Narayanamurthy <subbaraman.narayanamurthy@....qualcomm.com>, Pavel Machek <pavel@....cz>, linux-leds@...r.kernel.org, 
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] leds: rgb: leds-qcom-lpg: Only enable TRILED when LPG is
 used by LED

On Fri, Nov 14, 2025 at 03:13:18PM +0800, Fenglin Wu wrote:
> 
> On 11/14/2025 12:58 PM, Bjorn Andersson wrote:
> > If chan->in_use, then the channel is exposed as a LED and
> > lpg_pwm_request() should have returned -EBUSY, so we should never reach
> > lpg_pwm_apply()?
> 
> Yes, I agree.
> 
> Change is trying to ignore enabling TRILED channel when the LPG channel
> mapping to TRILED is not used for controlling the LED (not defining the LED
> child nodes).
> 
> So the fix should be just removing this line instead of adding the if check.
> 

Sorry, it's been a while since I looked at this code, but isn't it
possible to configure a channel going through the triled to be exposed
as a PWM channel and if so, don't we need to enable the TRILED driver
for this channel in those cases?

> I will update it in patch v2.
> 
> > 
> > Why do you check chan->triled_mask? I guess we will still read/write the
> > triled regiter, but don't make any changes if this is 0?
> > 
> > Or is this the actual issue that you're fixing, that we read/write the
> > registers when we shouldn't? If so this should be clarified in the
> > commit message.
> 
> Yes, there was a case that a LPG channel mapping to TRILED is repurposed to
> control a fan, and it was seen that the BOB1 (supplies to TRILED) voltage
> bumped to higher voltage when the PWM channel was enabled.
> 

Is the signal still routed through the TRILED, or is it muxed to another
driver?

Regards,
Bjorn

> > 
> > Regards,
> > Bjorn