Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <CAAeHK+wb4k-LGTjK9F5YbJNviF_+yU+wE_=Vpo9Rn7KFN8vG6Q@mail.gmail.com>
 <20201005151857.GA2309511@kroah.com> <CAAeHK+zes2Y00+EJ6SVtOHj8YCADw5WSXUEFHWCRgxi=H42+4w@mail.gmail.com>
 <20201005152540.GG376584@rowland.harvard.edu> <65b4ff62-f9c8-b9cf-50bb-c9b08cce7230@linuxfoundation.org>
 <20201006012333.GA399825@rowland.harvard.edu>
In-Reply-To: <20201006012333.GA399825@rowland.harvard.edu>
From:   Andrey Konovalov <andreyknvl@google.com>
Date:   Tue, 6 Oct 2020 17:44:33 +0200
Message-ID: <CAAeHK+w3hUbNO5PBcoZX2cJsmRPZ8bq+8j8-Xs7nZLFyKbdd+A@mail.gmail.com>
Subject: Re: Is usb_hcd_giveback_urb() allowed in task context?
To:     Alan Stern <stern@rowland.harvard.edu>
Cc:     Shuah Khan <skhan@linuxfoundation.org>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Valentina Manea <valentina.manea.m@gmail.com>,
        Shuah Khan <shuah@kernel.org>,
        USB list <linux-usb@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Dmitry Vyukov <dvyukov@google.com>,
        Nazime Hande Harputluoglu <handeharputlu@google.com>,
        syzkaller <syzkaller@googlegroups.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Tue, Oct 6, 2020 at 3:23 AM Alan Stern <stern@rowland.harvard.edu> wrote:
>
> On Mon, Oct 05, 2020 at 05:38:22PM -0600, Shuah Khan wrote:
> > On 10/5/20 9:25 AM, Alan Stern wrote:
> > > On Mon, Oct 05, 2020 at 05:21:30PM +0200, Andrey Konovalov wrote:
> > > No, no -- it won't work right if it's called in process context.  Not
> > > only do the spinlock calls leave the interrupt flag unchanged, also the
> > > driver callback routines may expect to be invoked with interrupts
> > > disabled.  (We have tried to fix this, but I'm not at all certain that
> > > all the cases have been updated.)
> > >
> >
> > In the case of vhci case, usb_hcd_giveback_urb() is called from vhci's
> > urb_enqueue, when it determines it doesn't need to xmit the urb and can give
> > it back. This path runs in task context.
> >
> > Do you have any recommendation on how this case can be handled?
>
> Just call local_irq_disable() before usb_hcd_giveback_urb(), and
> local_irq_enable() afterward.

OK, so overall it's possible to call usb_hcd_giveback_urb() in task
context, but only with irqs disabled.

This means we do need a fix for kcov as well, thank you!