linux-kernel - Re: [PATCH] Input: psmouse - add resync_on

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]

Message-ID: <b9f08bfb-4c1c-4d1b-9061-8a4b1013497d@redhat.com>
Date: Mon, 5 Aug 2024 16:18:57 +0200
From: Hans de Goede <hdegoede@...hat.com>
To: Dmitry Torokhov <dmitry.torokhov@...il.com>,
 Jonathan Denose <jdenose@...omium.org>
Cc: LKML <linux-kernel@...r.kernel.org>, jefferymiller@...gle.com,
 Jonathan Denose <jdenose@...gle.com>, Raul Rangel <rrangel@...omium.org>,
 linux-input@...r.kernel.org, Ike Panhc <ike.pan@...onical.com>,
 platform-driver-x86@...r.kernel.org
Subject: Re: [PATCH] Input: psmouse - add resync_on_resume dmi check

Hi Dmitry,

On 3/7/24 7:29 PM, Dmitry Torokhov wrote:
> On Mon, Mar 04, 2024 at 11:17:31AM -0600, Jonathan Denose wrote:
>> I disabled the ideapad driver by rebuilding the kernel without the
>> ideapad_laptop module. That does fix the suspend/resume issue!
>>
>> Attached are the logs. Is there a way to make this permanent?
>>
>> On Thu, Feb 29, 2024 at 12:23 PM Dmitry Torokhov
>> <dmitry.torokhov@...il.com> wrote:
>>>
>>> On Mon, Feb 12, 2024 at 02:57:08PM -0600, Jonathan Denose wrote:
>>> ...
>>>> [   50.241235] ideapad_acpi VPC2004:00: PM: calling acpi_subsys_resume+0x0/0x5d @ 4492, parent: PNP0C09:00
>>>> [   50.242055] snd_hda_intel 0000:00:0e.0: PM: pci_pm_resume+0x0/0xed returned 0 after 13511 usecs
>>>> [   50.242120] snd_hda_codec_realtek hdaudioC0D0: PM: calling hda_codec_pm_resume+0x0/0x19 [snd_hda_codec] @ 4518, parent: 0000:00:0e.0
>>>> [   50.247406] i8042: [49434] a8 -> i8042 (command)
>>>> [   50.247468] ideapad_acpi VPC2004:00: PM: acpi_subsys_resume+0x0/0x5d returned 0 after 6220 usecs
>>> ...
>>>> [   50.247883] i8042 kbd 00:01: PM: calling pnp_bus_resume+0x0/0x9d @ 4492, parent: pnp0
>>>> [   50.247894] i8042 kbd 00:01: PM: pnp_bus_resume+0x0/0x9d returned 0 after 0 usecs
>>>> [   50.247906] i8042 aux 00:02: PM: calling pnp_bus_resume+0x0/0x9d @ 4492, parent: pnp0
>>>> [   50.247916] i8042 aux 00:02: PM: pnp_bus_resume+0x0/0x9d returned 0 after 0 usecs
>>> ...
>>>> [   50.248301] i8042 i8042: PM: calling platform_pm_resume+0x0/0x41 @ 4492, parent: platform
>>>> [   50.248377] i8042: [49434] 55 <- i8042 (flush, kbd)
>>>> [   50.248407] i8042: [49435] aa -> i8042 (command)
>>>> [   50.248601] i8042: [49435] 00 <- i8042 (return)
>>>> [   50.248604] i8042: [49435] i8042 controller selftest: 0x0 != 0x55
>>>
>>> So here I see the ideapad-laptop driver trying to access i8042 before it
>>> even starts resuming. I wonder, does it help if you disable
>>> (temporarily) the ideapad driver?
> 
> OK, so I tried to cook up a patch that would allow ideapad-laptop driver
> to establish device link with i8042 so that the resume will be processed
> after i8042 resumes, but the longer I think about it, the more I think
> that ideapad driver should not be messing with the touchpad state
> directly. The disable event may come up in a middle of the touchpad
> resume transition, or when we decide to change touchpad mode for one
> reason or another. It also does not respect inhibit/uninhibit controls
> for input devices. I think that the proper way for ideapad driver to
> handle this is to only send KEY_TOUCHPAD_OFF/KEY_TOUCHPAD_ON to
> userspace and let userspace deal with toggling touchpad input (via
> inhibit or by other means).
> 
> CC-ing ideapad maintainers for their thoughts.

Sorry for the very slow reply.

The interesting thing is that sometime ago I already removed the i8042_command()
command being done on most models now the ideapad driver already only
sends KEY_TOUCHPAD_OFF/KEY_TOUCHPAD_ON except on the ideapad Z570 for
which the i8042_command() call was initially added.

I agree that this should probably just be removed.

Jonathan, I presume that you are seeing this on an Ideapad Z570 ?
(since that is the only model where this is still done by default).

Since the i8042_command() call has already been disabled on all other
ideapad models I agree that it would be best to just remove it entirely
relying on userspace filtering out touchpad events after receiving
a KEY_TOUCHPAD_OFF.

I have submitted a patch to do just that:

https://lore.kernel.org/platform-driver-x86/20240805141608.170844-1-hdegoede@redhat.com/

Jonathan can you give this patch a try (with a kernel with
the ideapad-laptop module re-enabled) and then confirm that this
fixes things ?

Regards,

Hans