On 28-05-21, 13:29, Dmitry Baryshkov wrote:
> On 21/05/2021 15:49, Vinod Koul wrote:
> > DSC needs some configuration from device tree, add support to read and
> > store these params and add DSC structures in msm_drv
> >
> > Signed-off-by: Vinod Koul <vkoul@xxxxxxxxxx>
> > ---
> > drivers/gpu/drm/msm/dsi/dsi_host.c | 170 +++++++++++++++++++++++++++++
> > drivers/gpu/drm/msm/msm_drv.h | 32 ++++++
> > 2 files changed, 202 insertions(+)
> >
> > diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
> > index 8a10e4343281..864d3c655e73 100644
> > --- a/drivers/gpu/drm/msm/dsi/dsi_host.c
> > +++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
> > @@ -156,6 +156,7 @@ struct msm_dsi_host {
> > struct regmap *sfpb;
> > struct drm_display_mode *mode;
> > + struct msm_display_dsc_config *dsc;
> > /* connected device info */
> > struct device_node *device_node;
> > @@ -1744,6 +1745,168 @@ static int dsi_host_parse_lane_data(struct msm_dsi_host *msm_host,
> > return -EINVAL;
> > }
> > +static u32 dsi_dsc_rc_buf_thresh[DSC_NUM_BUF_RANGES - 1] = {
> > + 0x0e, 0x1c, 0x2a, 0x38, 0x46, 0x54, 0x62,
> > + 0x69, 0x70, 0x77, 0x79, 0x7b, 0x7d, 0x7e
> > +};
>
> I think we should move this table to a generic place. AMD and Intel DSC code
> uses the same table, shifted by 6 (and both of those drivers shift it before
> writing to the HW). Intel modifies this table for 6bpp case. AMD seems to
> use it as is.
>
> > +
> > +/* only 8bpc, 8bpp added */
> > +static char min_qp[DSC_NUM_BUF_RANGES] = {
> > + 0, 0, 1, 1, 3, 3, 3, 3, 3, 3, 5, 5, 5, 7, 13
> > +};
> > +
> > +static char max_qp[DSC_NUM_BUF_RANGES] = {
> > + 4, 4, 5, 6, 7, 7, 7, 8, 9, 10, 11, 12, 13, 13, 15
> > +};
> > +
> > +static char bpg_offset[DSC_NUM_BUF_RANGES] = {
> > + 2, 0, 0, -2, -4, -6, -8, -8, -8, -10, -10, -12, -12, -12, -12
> > +};
>
> And these parameters seem to be generic too. Intel DSC code contains them in
> a bit different form. Should we probably move them to the drm_dsc.c and use
> the tables the generic location?
>
> AMD drivers uses a bit different values at the first glance, so let's stick
> with Intel version.
Yeah I think this is a good suggestion. I did look into and had this in
my todo. Yes drm_dsc.c would be apt for the move..
--
~Vinod
