linux-kernel - Re: Regression: Dell XPS 13 9360 keyboard no longer works

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: <01000161bf0c3bf3-3d6f8448-0bfd-40f6-bf2c-f7034ea47002-000000@email.amazonses.com>
Date:   Thu, 22 Feb 2018 19:45:05 +0000
From:   Jeremy Cline <jeremy@...ine.org>
To:     Mario.Limonciello@...l.com, notmart@...il.com
Cc:     pali.rohar@...il.com, andriy.shevchenko@...ux.intel.com,
        mjg59@...f.ucam.org, dvhart@...radead.org,
        platform-driver-x86@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: Regression: Dell XPS 13 9360 keyboard no longer works

On 02/22/2018 12:05 PM, Mario.Limonciello@...l.com wrote:
>> -----Original Message-----
>> From: platform-driver-x86-owner@...r.kernel.org [mailto:platform-driver-x86-
>> owner@...r.kernel.org] On Behalf Of Jeremy Cline
>> Sent: Thursday, February 22, 2018 10:42 AM
>> To: Limonciello, Mario <Mario_Limonciello@...l.com>; notmart@...il.com
>> Cc: pali.rohar@...il.com; andriy.shevchenko@...ux.intel.com;
>> mjg59@...f.ucam.org; dvhart@...radead.org; platform-driver-
>> x86@...r.kernel.org; linux-kernel@...r.kernel.org
>> Subject: Re: Regression: Dell XPS 13 9360 keyboard no longer works
>>
>> On 02/22/2018 11:17 AM, Mario.Limonciello@...l.com wrote:
>>>> -----Original Message-----
>>>> From: platform-driver-x86-owner@...r.kernel.org [mailto:platform-driver-x86-
>>>> owner@...r.kernel.org] On Behalf Of Jeremy Cline
>>>> Sent: Thursday, February 22, 2018 10:17 AM
>>>> To: Limonciello, Mario <Mario_Limonciello@...l.com>; notmart@...il.com
>>>> Cc: pali.rohar@...il.com; andriy.shevchenko@...ux.intel.com;
>>>> mjg59@...f.ucam.org; dvhart@...radead.org; platform-driver-
>>>> x86@...r.kernel.org; linux-kernel@...r.kernel.org
>>>> Subject: Re: Regression: Dell XPS 13 9360 keyboard no longer works
>>>>
>>>> On 02/22/2018 11:14 AM, Mario.Limonciello@...l.com wrote:
>>>>>> -----Original Message-----
>>>>>> From: Jeremy Cline [mailto:jeremy@...ine.org]
>>>>>> Sent: Thursday, February 22, 2018 9:59 AM
>>>>>> To: Limonciello, Mario <Mario_Limonciello@...l.com>; notmart@...il.com
>>>>>> Cc: pali.rohar@...il.com; andriy.shevchenko@...ux.intel.com;
>>>>>> mjg59@...f.ucam.org; dvhart@...radead.org; platform-driver-
>>>>>> x86@...r.kernel.org; linux-kernel@...r.kernel.org
>>>>>> Subject: Re: Regression: Dell XPS 13 9360 keyboard no longer works
>>>>>>
>>>>>> On 02/22/2018 10:21 AM, Mario.Limonciello@...l.com wrote:> I guess that
>>>>>> means we got this wrong and the patch should be reverted
>>>>>>> until we figure this out.
>>>>>>>
>>>>>>> Jeremy,
>>>>>>>
>>>>>>> Can you please confirm what BIOS version you are on?
>>>>>>> Also Is this a 9360 with 7th or 8th gen Intel CPU?
>>>>>>
>>>>>> Hi Mario,
>>>>>>
>>>>>> I've got BIOS version 2.5.0 with the 7th gen Intel CPU.
>>>>>>
>>>>>>
>>>>>> Regards,
>>>>>> Jeremy
>>>>>
>>>>> Jeremy,
>>>>>
>>>>> Thanks.  Do you have any of the Dell docks (TB16/WD15)?  If so are you
>> connected
>>>> to any dock
>>>>> when reproducing this problem?
>>>>
>>>> Mario,
>>>>
>>>> I do have a TB16. I can reproduce this whether or not I'm connected to
>>>> the dock, though.
>>>>
>>>>
>>>> Regards,
>>>> Jeremy
>>>
>>> Jeremy,
>>>
>>> Can you try booting up from a cold boot with it connected to see if it still
>> happens?
>>>
>>
>> Mario,
>>
>> Yup, it still happens from a cold boot when connected to the dock.
> 
> OK thanks for confirming.  Here's what I've concluded:
> 
> * So looking through the ACPI tables on the 9360 it initializes that status
> (slate vs laptop mode) bit to "slate" mode.  The 9360 isn't a 2-in1- so that
> seems wrong to me, but that's what it does.
> It only gets updated based on dock status.
> 
> The 9365 (which is a 2 in 1) however seems to initialize the status properly.
> 
> So that's an impasse of what to do.
> It's not clear to me what is really happening:
> a) We're missing something else in this driver (eg something else that 
> indicates whether to trust VGBS output)
> 
> b) Mis-interpreting the results from it (we shouldn't report the switch for 
> tablet mode based on what we do)
> 
> c) 9360 has a BIOS bug (seems unlikely since Windows doesn't freak out and
> show virtual keyboard at wrong time)
> 
> I'm leaning on it's probably <a>.
> We should check for tablet mode should only be run if chassis type 
> matches 2-in-1 (chassis type 0x1F).  
> 
> I believe that should fix the problem on the 9360, let it continue to work on
> the 9365 (and other 2-in-1's).
> 
> Can you guys please test this?  If that works I'll split up the routine and submit
> it.
> 

Hi Mario,

There's one small error in the patch, after I fixed it, it works.
Thanks!

> diff --git a/drivers/platform/x86/intel-vbtn.c b/drivers/platform/x86/intel-vbtn.c
> index b703d6f..07bc489 100644
> --- a/drivers/platform/x86/intel-vbtn.c
> +++ b/drivers/platform/x86/intel-vbtn.c
> @@ -7,6 +7,7 @@
>   */
>  
>  #include <linux/acpi.h>
> +#include <linux/dmi.h>
>  #include <linux/input.h>
>  #include <linux/input/sparse-keymap.h>
>  #include <linux/kernel.h>
> @@ -102,6 +103,7 @@ static int intel_vbtn_probe(struct platform_device *device)
>         struct acpi_buffer vgbs_output = { ACPI_ALLOCATE_BUFFER, NULL };
>         acpi_handle handle = ACPI_HANDLE(&device->dev);
>         struct intel_vbtn_priv *priv;
> +       const char *chassis_type;
>         acpi_status status;
>         int err;
>  
> @@ -123,22 +125,24 @@ static int intel_vbtn_probe(struct platform_device *device)
>         }
>  
>         /*
> -        * VGBS being present and returning something means we have
> -        * a tablet mode switch.
> +        * Running on 2-in-1 chassis, VGBS being present and
> +        * returning something means we have a tablet mode switch.
>          */
> -       status = acpi_evaluate_object(handle, "VGBS", NULL, &vgbs_output);
> -       if (ACPI_SUCCESS(status)) {
> -               union acpi_object *obj = vgbs_output.pointer;
> +       chassis_type = dmi_get_system_info(DMI_CHASSIS_TYPE);
> +        if (chassis_type && strcmp(chassis_type, "31")) {This should be "chassis_type && strcmp(chassis_type, "31") == 0" since
strcmp returns 0 for equality.

> +               status = acpi_evaluate_object(handle, "VGBS", NULL, &vgbs_output);
> +               if (ACPI_SUCCESS(status)) {
> +                       union acpi_object *obj = vgbs_output.pointer;
>  
> -               if (obj && obj->type == ACPI_TYPE_INTEGER) {
> -                       int m = !(obj->integer.value & TABLET_MODE_FLAG);
> +                       if (obj && obj->type == ACPI_TYPE_INTEGER) {
> +                               int m = !(obj->integer.value & TABLET_MODE_FLAG);
>  
> -                       input_report_switch(priv->input_dev, SW_TABLET_MODE, m);
> +                               input_report_switch(priv->input_dev, SW_TABLET_MODE, m);
> +                       }
>                 }
> +               kfree(vgbs_output.pointer);
>         }
> -       kfree(vgbs_output.pointer);
> -
>         status = acpi_install_notify_handler(handle,
>                                              ACPI_DEVICE_NOTIFY,
>                                              notify_handler,
> 

Regards,
Jeremy