Re: [PATCH v3 4/8] platform/x86: wmi: create character devices when requested by drivers

[Andy Lutomirski]

I think David Hermann had some review comments.

--Andy

On Oct 3, 2017, at 11:38 AM, <Mario.Limonciello@xxxxxxxx> <Mario.Limonciello@xxxxxxxx> wrote:

>> -----Original Message-----
>> From: Andy Lutomirski [mailto:luto@xxxxxxxxxx]
>> Sent: Tuesday, October 3, 2017 11:48 AM
>> To: Darren Hart <dvhart@xxxxxxxxxxxxx>
>> Cc: Greg KH <greg@xxxxxxxxx>; Limonciello, Mario
>> <Mario_Limonciello@xxxxxxxx>; Andy Shevchenko
>> <andy.shevchenko@xxxxxxxxx>; LKML <linux-kernel@xxxxxxxxxxxxxxx>;
>> Platform Driver <platform-driver-x86@xxxxxxxxxxxxxxx>; Andy Lutomirski
>> <luto@xxxxxxxxxx>; quasisec@xxxxxxxxxx; Pali RohÃr <pali.rohar@xxxxxxxxx>
>> Subject: Re: [PATCH v3 4/8] platform/x86: wmi: create character devices when
>> requested by drivers
>> 
>>> On Tue, Oct 3, 2017 at 8:10 AM, Darren Hart <dvhart@xxxxxxxxxxxxx> wrote:
>>>> On Tue, Oct 03, 2017 at 11:23:23AM +0200, Greg KH wrote:
>>>>> On Wed, Sep 27, 2017 at 11:02:16PM -0500, Mario Limonciello wrote:
>>>>> For WMI operations that are only Set or Query read or write sysfs
>>>>> attributes created by WMI vendor drivers make sense.
>>>>> 
>>>>> For other WMI operations that are run on Method, there needs to be a
>>>>> way to guarantee to userspace that the results from the method call
>>>>> belong to the data request to the method call.  Sysfs attributes don't
>>>>> work well in this scenario because two userspace processes may be
>>>>> competing at reading/writing an attribute and step on each other's
>>>>> data.
>>>>> 
>>>>> When a WMI vendor driver declares a set of functions in a
>>>>> file_operations object the WMI bus driver will create a character
>>>>> device that maps to those file operations.
>>>>> 
>>>>> That character device will correspond to this path:
>>>>> /dev/wmi/$driver
>>>>> 
>>>>> This policy is selected as one driver may map and use multiple
>>>>> GUIDs and it would be better to only expose a single character
>>>>> device.
>>>>> 
>>>>> The WMI vendor drivers will be responsible for managing access to
>>>>> this character device and proper locking on it.
>>>>> 
>>>>> When a WMI vendor driver is unloaded the WMI bus driver will clean
>>>>> up the character device.
>>>> 
>>>> Ok, thanks to Darren, I've gone and dug these up while my boxes were
>>>> building stable kernels...
>>>> 
>>>> Why are you not just using the misc device interface here?  Why do you
>>>> need a whole new major and minor range?  Why not just register misc
>>>> devices dynamically as-needed?  Should be much simpler and easier to
>>>> maintain and reduce your code size a lot.
>>> 
>>> Thank you Greg, this simplifies things quite a bit.
>>> 
>>> Mario, the misc device interface will remove a lot of the boiler plate
>>> setup and eliminate the need to allocate a new major number.
>>> 
>> 
>> In my mind, the problem with misc is that you may end up forever stuck
>> with a misc device, and they're distinct (visibly to userspace) from
>> all other character devices.
>> 
>> If you really want to be fancy, you could try to dust off a non-awful
>> character device API, a la:
>> 
>> https://git.kernel.org/pub/scm/linux/kernel/git/luto/linux.git/commit/?h=u2f&i
>> d=d3ab93173d51cebf00dd2263fd0ce9f8cd6258f7
> 
> That's two years old.  What's the history of it?  Did you try to get it merged?