[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20170228184558.GG20776@dtor-ws>
Date: Tue, 28 Feb 2017 10:45:58 -0800
From: Dmitry Torokhov <dmitry.torokhov@...il.com>
To: Michał Kępień <kernel@...pniu.pl>
Cc: linux-input@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] Input: sparse-keymap - add managed version of
sparse_keymap_setup()
On Tue, Feb 28, 2017 at 10:45:25AM +0100, Michał Kępień wrote:
> Some platform drivers use devm_input_allocate_device() together with
> sparse_keymap_setup() in their .probe callbacks. While using the former
> simplifies error handling, using the latter necessitates calling
> sparse_keymap_free() in the error path and upon module unloading to
> avoid leaking the copy of the keymap allocated by sparse_keymap_setup().
>
> To help prevent such leaks and enable simpler error handling in these
> drivers, add a new function which allows automatic freeing of the keymap
> copy upon probe failure and on driver detach.
>
> As devm_input_allocate_device() adds its devres to the device owning the
> input device, we do the same for managed input devices to ensure freeing
> the keymap copy is properly slotted in the devres stack.
>
> The new function can also be used by non-managed input devices, though
> in this case the devres is attached to the struct device embedded inside
> the input device itself.
This is wrong and does not work as input devices are never probed and
never unbound, so the cleanup will never happen.
Either pass device explicitly, or always take input's parent.
Thanks.
--
Dmitry
Powered by blists - more mailing lists