linux-kernel - Re: [PATCH v2] iio: core: register chardev only if needed

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: <5d3d6bad-934d-41c5-d167-dd69734e8015@metafoo.de>
Date:   Wed, 15 Apr 2020 16:55:40 +0200
From:   Lars-Peter Clausen <lars@...afoo.de>
To:     "Ardelean, Alexandru" <alexandru.Ardelean@...log.com>,
        "jic23@...nel.org" <jic23@...nel.org>
Cc:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "linux-iio@...r.kernel.org" <linux-iio@...r.kernel.org>
Subject: Re: [PATCH v2] iio: core: register chardev only if needed

On 4/15/20 3:56 PM, Ardelean, Alexandru wrote:
> On Tue, 2020-04-14 at 19:06 +0100, Jonathan Cameron wrote:
>> [External]
>>
>> On Tue, 14 Apr 2020 11:36:56 +0300
>> Alexandru Ardelean <alexandru.ardelean@...log.com> wrote:
>>
>>> The final intent is to localize all buffer ops into the
>>> industrialio-buffer.c file, to be able to add support for multiple buffers
>>> per IIO device.
>>>
>>> We only need a chardev if we need to support buffers and/or events.
>>>
>>> With this change, a chardev will be created:
>>> 1. if there is an IIO buffer attached OR
>>> 2. if there is an event_interface configured
>>>
>>> Otherwise, no chardev will be created.
>>> Quite a lot of IIO devices don't really need a chardev, so this is a minor
>>> improvement to the IIO core, as the IIO device will take up fewer
>>> resources.
>>>
>>> Signed-off-by: Alexandru Ardelean <alexandru.ardelean@...log.com>
>>> ---
>>>
>>> Changelog v1 -> v2:
>>> * split away from series 'iio: core,buffer: re-organize chardev creation';
>>>    i'm getting the feeling that this has some value on it's own;
>>>    no idea if it needs 'Fixes' tag; it is a bit fuzzy to point to a patch
>>>    which this would be fixed by this; i'm guessing it would be fine
>>>    without one
>> I'd argue it's an 'optimization' rather than a fix :)
>>
>> Still looks good to me but I'd like it to sit for a little while to
>> see if anyone points out something we are both missing!
>>
> This is not good.
> It seems that I did not properly test all cases.
> I had to break a device to not have an event_interface to notice that the sysfs
> doesn't get instantiated either because device_add is missing.
>
> Will do another try.

I think you also have to make the `indio_dev->dev.devt = ...` 
conditional. Or conditionally use device_add() instead of device_add_cdev().

If you go for the former you need to call cdev_device_del() 
unconditionally, for the latter call device_del() or cdev_device_del() 
depending on whether the cdev was registered.

- Lars