X86 platform drivers
 help / color / mirror / Atom feed
From: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
To: Armin Wolf <W_Armin@gmx.de>,
	hdegoede@redhat.com, ilpo.jarvinen@linux.intel.com
Cc: corbet@lwn.net, linux-doc@vger.kernel.org,
	platform-driver-x86@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/4] platform/x86: xiaomi-wmi: Fix race condition when reporting key events
Date: Thu, 28 Mar 2024 18:37:33 -0700	[thread overview]
Message-ID: <9e7a081f-b727-4ead-84b9-3c80530a1039@linux.intel.com> (raw)
In-Reply-To: <3e41e132-98b9-496d-987f-6838289926b6@gmx.de>


On 3/28/24 5:26 PM, Armin Wolf wrote:
> Am 28.03.24 um 03:58 schrieb Kuppuswamy Sathyanarayanan:
>
>> On 3/27/24 6:23 PM, Armin Wolf wrote:
>>> Multiple WMI events can be received concurrently, so multiple instances
>>> of xiaomi_wmi_notify() can be active at the same time. Since the input
>>> device is shared between those handlers, the key input sequence can be
>>> disturbed.
>> Since locking is needed for all notify related calls in all WMI drivers,
>> is there a generic way to add this support in WMI core driver? Like
>> defining some common function which will hold the lock and call
>> driver specific notify handler? I am just thinking aloud.. If it is not
>> feasible, then it is fine.
>
> Hi,
>
> actually, not all notify-related calls need locking. It just so happens that
> input drivers must protect their key sequence, other WMI drivers might not need
> to use any locking at all.

Got it.

>
> I would prefer if the WMI driver does the locking, so it can be optimized for
> each WMI driver.

Why not implement this support?

>
> Thanks,
> Armin Wolf
>
>>> Fix this by protecting the key input sequence with a mutex.
>>>
>>> Compile-tested only.
>>>
>>> Fixes: edb73f4f0247 ("platform/x86: wmi: add Xiaomi WMI key driver")
>>> Signed-off-by: Armin Wolf <W_Armin@gmx.de>
>>> ---
>>>   drivers/platform/x86/xiaomi-wmi.c | 18 ++++++++++++++++++
>>>   1 file changed, 18 insertions(+)
>>>
>>> diff --git a/drivers/platform/x86/xiaomi-wmi.c b/drivers/platform/x86/xiaomi-wmi.c
>>> index 1f5f108d87c0..7efbdc111803 100644
>>> --- a/drivers/platform/x86/xiaomi-wmi.c
>>> +++ b/drivers/platform/x86/xiaomi-wmi.c
>>> @@ -2,8 +2,10 @@
>>>   /* WMI driver for Xiaomi Laptops */
>>>
>>>   #include <linux/acpi.h>
>>> +#include <linux/device.h>
>>>   #include <linux/input.h>
>>>   #include <linux/module.h>
>>> +#include <linux/mutex.h>
>>>   #include <linux/wmi.h>
>>>
>>>   #include <uapi/linux/input-event-codes.h>
>>> @@ -20,12 +22,21 @@
>>>
>>>   struct xiaomi_wmi {
>>>       struct input_dev *input_dev;
>>> +    struct mutex key_lock;    /* Protects the key event sequence */
>>>       unsigned int key_code;
>>>   };
>>>
>>> +static void xiaomi_mutex_destroy(void *data)
>>> +{
>>> +    struct mutex *lock = data;
>>> +
>>> +    mutex_destroy(lock);
>>> +}
>>> +
>>>   static int xiaomi_wmi_probe(struct wmi_device *wdev, const void *context)
>>>   {
>>>       struct xiaomi_wmi *data;
>>> +    int ret;
>>>
>>>       if (wdev == NULL || context == NULL)
>>>           return -EINVAL;
>>> @@ -35,6 +46,11 @@ static int xiaomi_wmi_probe(struct wmi_device *wdev, const void *context)
>>>           return -ENOMEM;
>>>       dev_set_drvdata(&wdev->dev, data);
>>>
>>> +    mutex_init(&data->key_lock);
>>> +    ret = devm_add_action_or_reset(&wdev->dev, xiaomi_mutex_destroy, &data->key_lock);
>>> +    if (ret < 0)
>>> +        return ret;
>>> +
>>>       data->input_dev = devm_input_allocate_device(&wdev->dev);
>>>       if (data->input_dev == NULL)
>>>           return -ENOMEM;
>>> @@ -59,10 +75,12 @@ static void xiaomi_wmi_notify(struct wmi_device *wdev, union acpi_object *dummy)
>>>       if (data == NULL)
>>>           return;
>>>
>>> +    mutex_lock(&data->key_lock);
>>>       input_report_key(data->input_dev, data->key_code, 1);
>>>       input_sync(data->input_dev);
>>>       input_report_key(data->input_dev, data->key_code, 0);
>>>       input_sync(data->input_dev);
>>> +    mutex_unlock(&data->key_lock);
>>>   }
>>>
>>>   static const struct wmi_device_id xiaomi_wmi_id_table[] = {
>>> -- 
>>> 2.39.2
>>>
>>>
-- 
Sathyanarayanan Kuppuswamy
Linux Kernel Developer


  reply	other threads:[~2024-03-29  1:37 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-28  1:23 [PATCH 1/4] platform/x86: wmi: Mark simple WMI drivers as legacy-free Armin Wolf
2024-03-28  1:23 ` [PATCH 2/4] platform/x86: xiaomi-wmi: Fix race condition when reporting key events Armin Wolf
2024-03-28  2:58   ` Kuppuswamy Sathyanarayanan
2024-03-29  0:26     ` Armin Wolf
2024-03-29  1:37       ` Kuppuswamy Sathyanarayanan [this message]
2024-04-02 14:13         ` Armin Wolf
2024-03-28  1:23 ` [PATCH 3/4] platform/x86: xiaomi-wmi: Drop unnecessary NULL checks Armin Wolf
2024-03-28  1:23 ` [PATCH 4/4] platform/x86: wmi: Add driver development guide Armin Wolf
2024-03-28  2:54   ` Kuppuswamy Sathyanarayanan
2024-03-28  2:54 ` [PATCH 1/4] platform/x86: wmi: Mark simple WMI drivers as legacy-free Kuppuswamy Sathyanarayanan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=9e7a081f-b727-4ead-84b9-3c80530a1039@linux.intel.com \
    --to=sathyanarayanan.kuppuswamy@linux.intel.com \
    --cc=W_Armin@gmx.de \
    --cc=corbet@lwn.net \
    --cc=hdegoede@redhat.com \
    --cc=ilpo.jarvinen@linux.intel.com \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=platform-driver-x86@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox