[Intel-gfx] [PATCH v2] drm/i915: Support more QGV points
Saarinen, Jani
jani.saarinen at intel.com
Mon Nov 25 16:03:47 UTC 2019
Hi,
> -----Original Message-----
> From: Lisovskiy, Stanislav <stanislav.lisovskiy at intel.com>
> Sent: maanantai 25. marraskuuta 2019 17.55
> To: ville.syrjala at linux.intel.com
> Cc: Saarinen, Jani <jani.saarinen at intel.com>; intel-gfx at lists.freedesktop.org;
> Roper, Matthew D <matthew.d.roper at intel.com>; jani.nikula at linux.intel.com;
> Sarvela, Tomi P <tomi.p.sarvela at intel.com>
> Subject: Re: [PATCH v2] drm/i915: Support more QGV points
>
> On Mon, 2019-11-25 at 17:31 +0200, Ville Syrjälä wrote:
> > On Fri, Nov 22, 2019 at 04:10:49PM +0200, Stanislav Lisovskiy wrote:
> > > According to BSpec 53998, there is a mask of max 8 SAGV/QGV points
> > > we need to support.
> > >
> > > Bumping this up to keep the CI happy(currently preventing tests to
> > > run), until all SAGV changes land.
> > >
> > > Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=112189
> > > Signed-off-by: Stanislav Lisovskiy <stanislav.lisovskiy at intel.com>
> > > ---
> > > drivers/gpu/drm/i915/display/intel_bw.c | 22 +++++++++++++++++--
> > > ---
> > > drivers/gpu/drm/i915/i915_drv.h | 6 +++++-
> > > 2 files changed, 22 insertions(+), 6 deletions(-)
> > >
> > > diff --git a/drivers/gpu/drm/i915/display/intel_bw.c
> > > b/drivers/gpu/drm/i915/display/intel_bw.c
> > > index 86e75e858008..d1b805b30177 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_bw.c
> > > +++ b/drivers/gpu/drm/i915/display/intel_bw.c
> > > @@ -15,7 +15,7 @@ struct intel_qgv_point { };
> > >
> > > struct intel_qgv_info {
> > > - struct intel_qgv_point points[3];
> > > + struct intel_qgv_point points[I915_NUM_SAGV_POINTS];
> > > u8 num_points;
> > > u8 num_channels;
> > > u8 t_bl;
> > > @@ -276,15 +276,27 @@ void intel_bw_init_hw(struct drm_i915_private
> > > *dev_priv)
> > > static unsigned int intel_max_data_rate(struct drm_i915_private
> > > *dev_priv,
> > > int
> num_planes)
> > > {
> > > - if (INTEL_GEN(dev_priv) >= 11)
> > > + if (INTEL_GEN(dev_priv) >= 11) {
> > > + /*
> > > + * Any bw group has same amount of QGV points
> > > + */
> > > + const struct intel_bw_info *bi =
> > > + &dev_priv->max_bw[0];
> > > + unsigned int min_bw = UINT_MAX;
> > > + int i;
> > > +
> > > /*
> > > * FIXME with SAGV disabled maybe we can assume
> > > * point 1 will always be used? Seems to match
> > > * the behaviour observed in the wild.
> > > */
> > > - return min3(icl_max_bw(dev_priv, num_planes, 0),
> > > - icl_max_bw(dev_priv, num_planes,
> 1),
> > > - icl_max_bw(dev_priv, num_planes,
> 2));
> > > + for (i = 0; i < bi->num_qgv_points; i++) {
> > > + unsigned int bw = icl_max_bw(dev_priv,
> > > num_planes, i);
> > > +
> > > + min_bw = min(bw, min_bw);
> > > + }
> > > + return min_bw;
> > > + }
> > > else
> > > return UINT_MAX;
> > > }
> > > diff --git a/drivers/gpu/drm/i915/i915_drv.h
> > > b/drivers/gpu/drm/i915/i915_drv.h index fdae5a919bc8..d45a9ffaed4f
> > > 100644
> > > --- a/drivers/gpu/drm/i915/i915_drv.h
> > > +++ b/drivers/gpu/drm/i915/i915_drv.h
> > > @@ -621,6 +621,9 @@ struct i915_gem_mm {
> > >
> > > #define I915_ENGINE_WEDGED_TIMEOUT (60 * HZ) /* Reset but no
> > > recovery? */
> > >
> > > +/* BSpec precisely defines this */
> > > +#define I915_NUM_SAGV_POINTS 8
> >
> > I think everything else talks about "QGV points" rather than "SAGV
> > points" now? Would be nice to be consistent.
>
> Agree, _SAGV_ is not that accurate here.
> Will change the naming.
Can we do that while merging?
>
> Stan
>
> >
> > Apart from that
> > Reviewed-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
> >
> > > +
> > > struct ddi_vbt_port_info {
> > > /* Non-NULL if port present. */
> > > const struct child_device_config *child; @@ -1232,7 +1235,8 @@
> > > struct drm_i915_private {
> > > } dram_info;
> > >
> > > struct intel_bw_info {
> > > - unsigned int deratedbw[3]; /* for each QGV point */
> > > + /* for each QGV point */
> > > + unsigned int deratedbw[I915_NUM_SAGV_POINTS];
> > > u8 num_qgv_points;
> > > u8 num_planes;
> > > } max_bw[6];
> > > --
> > > 2.17.1
> >
> >
More information about the Intel-gfx
mailing list