Re: [PATCH] kcov, usb, vhost: specify contexts for remote coverage sections

From: Andrey Konovalov
Date: Wed Oct 07 2020 - 15:55:10 EST


On Wed, Oct 7, 2020 at 9:29 PM Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote:
>
> On Wed, Oct 07, 2020 at 07:30:51PM +0200, Andrey Konovalov wrote:
> > Currently there's a KCOV remote coverage collection section in
> > __usb_hcd_giveback_urb(). Initially that section was added based on the
> > assumption that usb_hcd_giveback_urb() can only be called in interrupt
> > context as indicated by a comment before it.
> >
> > As it turns out, it's actually valid to call usb_hcd_giveback_urb() in task
> > context, provided that the caller turned off the interrupts; USB/IP actually
> > does that. This can lead to a nested KCOV remote coverage collection
> > sections both trying to collect coverage in task context. This isn't
> > supported by KCOV, and leads to a WARNING.
> >
> > The approach this patch takes is to annotate every call of kcov_remote_*()
> > callbacks with the context those callbacks are supposed to be executed in.
> > If the current context doesn't match the mask provided to a callback,
> > that callback is ignored. KCOV currently only supports collecting remote
> > coverage in two contexts: task and softirq.
> >
> > As the result, the coverage from USB/IP related usb_hcd_giveback_urb() calls
> > won't be collected, but the WARNING is fixed.
> >
> > A potential future improvement would be to support nested remote coverage
> > collection sections, but this patch doesn't address that.
> >
> > Signed-off-by: Andrey Konovalov <andreyknvl@xxxxxxxxxx>
> > ---
>
> > --- a/drivers/usb/core/hcd.c
> > +++ b/drivers/usb/core/hcd.c
> > @@ -1646,9 +1646,9 @@ static void __usb_hcd_giveback_urb(struct urb *urb)
> >
> > /* pass ownership to the completion handler */
> > urb->status = status;
> > - kcov_remote_start_usb((u64)urb->dev->bus->busnum);
> > + kcov_remote_start_usb((u64)urb->dev->bus->busnum, KCOV_CONTEXT_SOFTIRQ);
> > urb->complete(urb);
> > - kcov_remote_stop();
> > + kcov_remote_stop(KCOV_CONTEXT_SOFTIRQ);
>
> This isn't right. __usb_hcd_giveback_urb() can execute in pretty much
> any context; its constraint is that interrupts must be disabled.

You're right, but here we constraint kcov to only collect coverage in
case __usb_hcd_giveback_urb() is executed in softirq context. This is
what happens when we're fuzzing USB with the dummy driver, which is
the case we currently take care of. Whenever someone has a desire to
collect coverage in other contexts, it will need to be implemented
separately.