Re: BUG: scheduling while atomic in f_fs when gadget remove driver

[John Stultz]

On Wed, Sep 28, 2016 at 2:38 PM, Michal Nazarewicz <mina86@xxxxxxxxxx> wrote:
> On Wed, Sep 28 2016, Michal Nazarewicz wrote:
>> With that done, the only thing which needs a mutex is
>> epfile->read_buffer.
>
> Perhaps this would do:
>
> ---- >8 -------------------------------------------------- -------------
> From 6416a1065203a39328311f6c58083089efe169aa Mon Sep 17 00:00:00 2001
> From: Michal Nazarewicz <mina86@xxxxxxxxxx>
> Date: Wed, 28 Sep 2016 23:36:56 +0200
> Subject: [RFC] usb: gadget: f_fs: stop sleeping in ffs_func_eps_disable
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
>
> ffs_func_eps_disable is called from atomic context so it cannot sleep
> thus cannot grab a mutex.  Change the handling of epfile->read_buffer
> to use non-sleeping synchronisation method.
>
> Reported-by: Chen Yu <chenyu56@xxxxxxxxxx>
> Signed-off-by: MichaÅ Nazarewicz <mina86@xxxxxxxxxx>
> Fixes: 9353afbbfa7b ("buffer data from âoversizedâ OUT requests")

So the patch here seems to be in some odd encoding? Can you resend it
using git-send-email or in some way other then embedding it inline
here? Maybe just point me to a git tree that has it?

thanks
-john