[Intel-gfx] [PATCH v3] drm/i915/pxp: don't start pxp without mei_pxp bind

[Juston Li]

On Fri, Aug 19, 2022 at 4:53 AM Andrzej Hajda <andrzej.hajda at intel.com> wrote:
>
> On 18.08.2022 19:42, Juston Li wrote:
> > pxp will not start correctly until after mei_pxp bind completes and
> > intel_pxp_init_hw() is called.
> > Wait for the bind to complete before proceeding with startup.
> >
> > This fixes a race condition during bootup where we observed a small
> > window for pxp commands to be sent, starting pxp before mei_pxp bind
> > completed.
> >
> > Changes since v2:
> > - wait for pxp_component to bind instead of returning -EAGAIN (Daniele)
> >
> > Changes since v1:
> > - check pxp_component instead of pxp_component_added (Daniele)
> > - pxp_component needs tee_mutex (Daniele)
> > - return -EAGAIN so caller knows to retry (Daniele)
> >
> > Signed-off-by: Juston Li <justonli at chromium.org>
>
> In typical usage of component framework driver postpones initialization
> till component is bound. In such case checking/waiting for component as
> in this patch is not necessary and the code is more straightforward.
> I wonder how it behaves on component unbind.
>
> Anyway:
> Reviewed-by: Andrzej Hajda <andrzej.hajda at intel.com>

Thanks Andrzej!

Any other comments Daniele?
Otherwise, need some help from someone to merge this :)

Thanks
Juston

> Regards
> Andrzej
>
>
> > ---
> >   drivers/gpu/drm/i915/pxp/intel_pxp.c | 15 +++++++++++++++
> >   1 file changed, 15 insertions(+)
> >
> > diff --git a/drivers/gpu/drm/i915/pxp/intel_pxp.c b/drivers/gpu/drm/i915/pxp/intel_pxp.c
> > index 15311eaed848..17109c513259 100644
> > --- a/drivers/gpu/drm/i915/pxp/intel_pxp.c
> > +++ b/drivers/gpu/drm/i915/pxp/intel_pxp.c
> > @@ -176,6 +176,18 @@ static void pxp_queue_termination(struct intel_pxp *pxp)
> >       spin_unlock_irq(&gt->irq_lock);
> >   }
> >
> > +static bool pxp_component_bound(struct intel_pxp *pxp)
> > +{
> > +     bool bound = false;
> > +
> > +     mutex_lock(&pxp->tee_mutex);
> > +     if (pxp->pxp_component)
> > +             bound = true;
> > +     mutex_unlock(&pxp->tee_mutex);
> > +
> > +     return bound;
> > +}
> > +
> >   /*
> >    * the arb session is restarted from the irq work when we receive the
> >    * termination completion interrupt
> > @@ -187,6 +199,9 @@ int intel_pxp_start(struct intel_pxp *pxp)
> >       if (!intel_pxp_is_enabled(pxp))
> >               return -ENODEV;
> >
> > +     if (wait_for(pxp_component_bound(pxp), 250))
> > +             return -ENXIO;
> > +
> >       mutex_lock(&pxp->arb_mutex);
> >
> >       if (pxp->arb_is_valid)
>