On Fri, 2017-01-13 at 09:55 +0200, Jani Nikula wrote:
> On Thu, 12 Jan 2017, Mika Kahola <mika.kahola@xxxxxxxxx> wrote:
> >
> > This is definitely needed to pass igt test on bxt
> >
> > 'gem_exec_suspend --run-subtest basic-S3'
> >
> > Tested-by: Mika Kahola <mika.kahola@xxxxxxxxx>
> >
> > On Mon, 2017-01-09 at 14:46 +0530, Vidya Srinivas wrote:
> > >
> > > From: Uma Shankar <uma.shankar@xxxxxxxxx>
> > >
> > > Enable MIPI IO WA for BXT DSI as per bspec.
> > >
> > > Signed-off-by: Uma Shankar <uma.shankar@xxxxxxxxx>
> > > ---
> > > drivers/gpu/drm/i915/i915_reg.h | 3 +++
> > > drivers/gpu/drm/i915/intel_dsi.c | 9 +++++++++
> > > 2 files changed, 12 insertions(+)
> > >
> > > diff --git a/drivers/gpu/drm/i915/i915_reg.h
> > > b/drivers/gpu/drm/i915/i915_reg.h
> > > index 71b978a..b9d7e98 100644
> > > --- a/drivers/gpu/drm/i915/i915_reg.h
> > > +++ b/drivers/gpu/drm/i915/i915_reg.h
> > > @@ -8301,6 +8301,9 @@ enum {
> > > #define _BXT_MIPIC_PORT_CTRL 0x6B8C0
> > > #define BXT_MIPI_PORT_CTRL(tc) _MMIO_MIPI(tc,
> > > _BXT_MIPIA_PORT_CTRL, _BXT_MIPIC_PORT_CTRL)
> > >
> > > +#define BXT_P_CR_GT_DISP_PWRON_0_2_0_GTTMMADR _MMIO(0
> > > x138090)
> Observe that this register is already defined as BXT_P_CR_GT_DISP_PWRON,
> and already used in intel_dpio_phy.c. It seems to me changing the bits
> in this register should be hooked at the dpio level.
AFAIK this is an uncore register and not exactly part of DPIO. The DPIO phy bits
are power requests that go to the P-Unit, so somewhat similar to power well
enabling. The DSI usage seems orthogonal to the DPIO phys, so I don't think it
makes a lot of sense to do it in intel_dpio_phy.c.
Ander
>
> Imre?
>
> >
> > >
> > > +#define MIPIO_RST_CTRL (1 <<
> > > 2)
> > > +
> > > #define DPI_ENABLE (1 << 31)
> > > /* A + C */
> > > #define MIPIA_MIPI4DPHY_DELAY_COUNT_SHIFT 27
> > > #define MIPIA_MIPI4DPHY_DELAY_COUNT_MASK (0xf << 27)
> > > diff --git a/drivers/gpu/drm/i915/intel_dsi.c
> > > b/drivers/gpu/drm/i915/intel_dsi.c
> > > index a4bda92..9252490 100644
> > > --- a/drivers/gpu/drm/i915/intel_dsi.c
> > > +++ b/drivers/gpu/drm/i915/intel_dsi.c
> > > @@ -366,6 +366,11 @@ static void bxt_dsi_device_ready(struct
> > > intel_encoder *encoder)
> > >
> > > DRM_DEBUG_KMS("\n");
> > >
> > > + /* Add MIPI IO reset programming for modeset */
> > > + val = I915_READ(BXT_P_CR_GT_DISP_PWRON_0_2_0_GTTMMADR);
> > > + I915_WRITE(BXT_P_CR_GT_DISP_PWRON_0_2_0_GTTMMADR,
> > > + val | MIPIO_RST_CTRL);
> > > +
> > Should we move this WA to intel_dsi_pre_enable() as the counterpart of
> > this WA is defined intel_dsi_post_disable()?
> As I said, this should probably be managed in intel_dpio_phy.c.
>
> And if not, this is BXT specific, and this hunk runs it on everything
> else too.
>
> BR,
> Jani.
>
>
> >
> >
> > >
> > > /* Enable MIPI PHY transparent latch */
> > > for_each_dsi_port(port, intel_dsi->ports) {
> > > val = I915_READ(BXT_MIPI_PORT_CTRL(port));
> > > @@ -757,6 +762,10 @@ static void intel_dsi_post_disable(struct
> > > intel_encoder *encoder,
> > > drm_panel_power_off(intel_dsi->panel);
> > > msleep(intel_dsi->panel_off_delay);
> > >
> > > + val = I915_READ(BXT_P_CR_GT_DISP_PWRON_0_2_0_GTTMMADR);
> > > + I915_WRITE(BXT_P_CR_GT_DISP_PWRON_0_2_0_GTTMMADR,
> > > + val & ~MIPIO_RST_CTRL);
> > > +
> > > intel_disable_dsi_pll(encoder);
> > >
> > > /* Panel Disable over CRC PMIC */
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
