linux-kernel - Re: [PATCH] arm64: dts: qcom: sdm845-shift-axolotl: fix touchscreen properties

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: <2a014801-e1ef-47fa-b067-9b566b9d523e@joelselvaraj.com>
Date: Sat, 18 Oct 2025 03:43:39 +0000
From: Joel Selvaraj <foss@...lselvaraj.com>
To: Konrad Dybcio <konrad.dybcio@....qualcomm.com>, Bjorn Andersson <andersson@...nel.org>, Konrad Dybcio <konradybcio@...nel.org>, Rob Herring <robh@...nel.org>, Krzysztof Kozlowski <krzk+dt@...nel.org>, Conor Dooley <conor+dt@...nel.org>
Cc: linux-arm-msm@...r.kernel.org, devicetree@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] arm64: dts: qcom: sdm845-shift-axolotl: fix touchscreen properties

Hi Konrad Dybcio,

On 10/17/25 4:57 AM, Konrad Dybcio wrote:
>>> All the changes look good given your commit message, but you dropped
>>> this wakeup-source property without explanation. It's fine to do so
>>> if it's intended, but please mention it if so
>>
>> In reference to the touchscreen/edt-ft5x06.c driver which is used here,
>> I am bit confused how wakeup-source works. Does specifying wakeup-source
>> in dts automatically makes "device_may_wakeup(dev)" return true, even if
>> device_init_wakeup is NOT configured in the driver? I noticed some
>> drivers do:
>>
>> device_init_wakeup(dev,device_property_read_bool(dev, "wakeup-source"));
>>
>> but the edt-ft5x06 driver doesnt do the init, but directly checks for
>> may_wakeup in suspend/resume.
>>
>> Few scenarios based on the driver code and my understanding:
>> 1. if device_may_wakeup will return true when wakeup-source is
>> specified, I probably want to just remove it, because irq and regulator
>> is not disabled during suspend and this will likely cause power drain.
> 
> I think this may be the case
> 
>> 2. The driver has an option to specify wake-gpio. In which case, the
>> touchscreen is put in some low power hibernate mode with irq and
>> regulators still enabled. But the touchscreen controller used in this
>> device doesn't seem to have support for a wake-gpio (atleast based on
>> downstream code). So that is not an option.
> 
> IIRC Shift was pretty open about development collaboration.. maybe you
> could reach out to them to confirm on schematics that the GPIO is
> absent?
> 
>> 3. if device_may_wakeup will always return false since
>> device_init_wakeup is not configured and since no wake-gpio is
>> available, the irq and regulators will be disabled during suspend.
>> Therefore, the device will not wake up from sleep even if wakeup-source
>> is specified as the irq is not going to be triggered.
>>
>> So probably no point in specifying wakeup-source either way I think? But
>> I am not sure which of these explanation is correct and thus not sure
>> what to mention in the v2 patch commit message. Also, there is a
>> possibility I am not understanding something. A little help from someone
>> will be very nice and sorry if I am obviously missing something.
> 
> I think this is intended for things like double-tap-to-wake, which
> obviously need some hw backing if you don't want to just keep the
> touchscreen online at "full power" 24/7
> 

Yeah. I think I will remove the wakeup-source for now to avoid potential 
battery drain. Once I get more info on wake-gpio/low-power mode and test 
with users that tap-to-wake work as expected, it will be added. For now, 
this is not something expected to be working. Will send a v2 with 
updated commit message.

Thank you,
Joel Selvaraj

> Konrad