Re: [PATCH 2/2] drm/gma500: Add page flip support on psb/cdv

[Daniel Vetter <daniel@ffwll.ch>]

On Wed, Nov 06, 2019 at 04:24:59PM +0100, Patrik Jakobsson wrote:
> On Wed, Nov 6, 2019 at 10:44 AM Paul Kocialkowski
> <paul.kocialkowski@bootlin.com> wrote:
> >
> > Legacy (non-atomic) page flip support is added to the driver by using the
> > mode_set_base CRTC function, that allows configuring a new framebuffer for
> > display. Since the function requires the primary plane's fb to be set
> > already, this is done prior to calling the function in the page flip helper
> > and reverted if the flip fails.
> >
> > The vblank interrupt handler is also refactored to support passing an event.
> > The PIPE_TE_STATUS bit is also considered to indicate vblank on medfield
> > only, as explained in psb_enable_vblank.
> >
> > It was tested by running weston on both poulsbo and cedartrail.
> >
> > Signed-off-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com>
> > ---
> >  drivers/gpu/drm/gma500/cdv_intel_display.c |  1 +
> >  drivers/gpu/drm/gma500/gma_display.c       | 46 ++++++++++++++++++++++
> >  drivers/gpu/drm/gma500/gma_display.h       |  6 +++
> >  drivers/gpu/drm/gma500/psb_intel_display.c |  1 +
> >  drivers/gpu/drm/gma500/psb_intel_drv.h     |  3 ++
> >  drivers/gpu/drm/gma500/psb_irq.c           | 18 +++++++--
> >  6 files changed, 72 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/gma500/cdv_intel_display.c b/drivers/gpu/drm/gma500/cdv_intel_display.c
> > index 8b784947ed3b..7109d3d19be0 100644
> > --- a/drivers/gpu/drm/gma500/cdv_intel_display.c
> > +++ b/drivers/gpu/drm/gma500/cdv_intel_display.c
> > @@ -979,6 +979,7 @@ const struct drm_crtc_funcs cdv_intel_crtc_funcs = {
> >         .gamma_set = gma_crtc_gamma_set,
> >         .set_config = gma_crtc_set_config,
> >         .destroy = gma_crtc_destroy,
> > +       .page_flip = gma_crtc_page_flip,
> >  };
> >
> >  const struct gma_clock_funcs cdv_clock_funcs = {
> > diff --git a/drivers/gpu/drm/gma500/gma_display.c b/drivers/gpu/drm/gma500/gma_display.c
> > index bc07ae2a9a1d..17f136985d21 100644
> > --- a/drivers/gpu/drm/gma500/gma_display.c
> > +++ b/drivers/gpu/drm/gma500/gma_display.c
> > @@ -503,6 +503,52 @@ void gma_crtc_destroy(struct drm_crtc *crtc)
> >         kfree(gma_crtc);
> >  }
> >
> > +int gma_crtc_page_flip(struct drm_crtc *crtc,
> > +                      struct drm_framebuffer *fb,
> > +                      struct drm_pending_vblank_event *event,
> > +                      uint32_t page_flip_flags,
> > +                      struct drm_modeset_acquire_ctx *ctx)
> > +{
> > +       struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
> > +       struct drm_framebuffer *current_fb = crtc->primary->fb;
> > +       struct drm_framebuffer *old_fb = crtc->primary->old_fb;
> > +       const struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private;
> > +       struct drm_device *dev = crtc->dev;
> > +       unsigned long flags;
> > +       int ret;
> > +
> > +       if (!crtc_funcs->mode_set_base)
> > +               return -EINVAL;
> > +
> > +       /* Using mode_set_base requires the new fb to be set already. */
> > +       crtc->primary->fb = fb;
> > +
> > +       if (event) {
> > +               spin_lock_irqsave(&dev->event_lock, flags);
> > +
> > +               WARN_ON(drm_crtc_vblank_get(crtc) != 0);
> > +
> > +               gma_crtc->page_flip_event = event;
> > +
> > +               /* Call this locked if we want an event at vblank interrupt. */
> > +               ret = crtc_funcs->mode_set_base(crtc, crtc->x, crtc->y, old_fb);
> > +               if (ret) {
> > +                       gma_crtc->page_flip_event = NULL;
> > +                       drm_crtc_vblank_put(crtc);
> > +               }
> > +
> > +               spin_unlock_irqrestore(&dev->event_lock, flags);
> > +       } else {
> > +               ret = crtc_funcs->mode_set_base(crtc, crtc->x, crtc->y, old_fb);
> > +       }
> > +
> > +       /* Restore previous fb in case of failure. */
> > +       if (ret)
> > +               crtc->primary->fb = current_fb;
> > +
> > +       return ret;
> > +}
> > +
> >  int gma_crtc_set_config(struct drm_mode_set *set,
> >                         struct drm_modeset_acquire_ctx *ctx)
> >  {
> > diff --git a/drivers/gpu/drm/gma500/gma_display.h b/drivers/gpu/drm/gma500/gma_display.h
> > index fdbd7ecaa59c..7bd6c1ee8b21 100644
> > --- a/drivers/gpu/drm/gma500/gma_display.h
> > +++ b/drivers/gpu/drm/gma500/gma_display.h
> > @@ -11,6 +11,7 @@
> >  #define _GMA_DISPLAY_H_
> >
> >  #include <linux/pm_runtime.h>
> > +#include <drm/drm_vblank.h>
> >
> >  struct drm_encoder;
> >  struct drm_mode_set;
> > @@ -71,6 +72,11 @@ extern void gma_crtc_prepare(struct drm_crtc *crtc);
> >  extern void gma_crtc_commit(struct drm_crtc *crtc);
> >  extern void gma_crtc_disable(struct drm_crtc *crtc);
> >  extern void gma_crtc_destroy(struct drm_crtc *crtc);
> > +extern int gma_crtc_page_flip(struct drm_crtc *crtc,
> > +                             struct drm_framebuffer *fb,
> > +                             struct drm_pending_vblank_event *event,
> > +                             uint32_t page_flip_flags,
> > +                             struct drm_modeset_acquire_ctx *ctx);
> >  extern int gma_crtc_set_config(struct drm_mode_set *set,
> >                                struct drm_modeset_acquire_ctx *ctx);
> >
> > diff --git a/drivers/gpu/drm/gma500/psb_intel_display.c b/drivers/gpu/drm/gma500/psb_intel_display.c
> > index 4256410535f0..fed3b563e62e 100644
> > --- a/drivers/gpu/drm/gma500/psb_intel_display.c
> > +++ b/drivers/gpu/drm/gma500/psb_intel_display.c
> > @@ -432,6 +432,7 @@ const struct drm_crtc_funcs psb_intel_crtc_funcs = {
> >         .gamma_set = gma_crtc_gamma_set,
> >         .set_config = gma_crtc_set_config,
> >         .destroy = gma_crtc_destroy,
> > +       .page_flip = gma_crtc_page_flip,
> >  };
> >
> >  const struct gma_clock_funcs psb_clock_funcs = {
> > diff --git a/drivers/gpu/drm/gma500/psb_intel_drv.h b/drivers/gpu/drm/gma500/psb_intel_drv.h
> > index cdf10333d1c2..16c6136f778b 100644
> > --- a/drivers/gpu/drm/gma500/psb_intel_drv.h
> > +++ b/drivers/gpu/drm/gma500/psb_intel_drv.h
> > @@ -12,6 +12,7 @@
> >  #include <drm/drm_crtc_helper.h>
> >  #include <drm/drm_encoder.h>
> >  #include <drm/drm_probe_helper.h>
> > +#include <drm/drm_vblank.h>
> >  #include <linux/gpio.h>
> >  #include "gma_display.h"
> >
> > @@ -182,6 +183,8 @@ struct gma_crtc {
> >         struct psb_intel_crtc_state *crtc_state;
> >
> >         const struct gma_clock_funcs *clock_funcs;
> > +
> > +       struct drm_pending_vblank_event *page_flip_event;
> >  };
> >
> >  #define to_gma_crtc(x) \
> > diff --git a/drivers/gpu/drm/gma500/psb_irq.c b/drivers/gpu/drm/gma500/psb_irq.c
> > index e6265fb85626..f787a51f6335 100644
> > --- a/drivers/gpu/drm/gma500/psb_irq.c
> > +++ b/drivers/gpu/drm/gma500/psb_irq.c
> > @@ -165,11 +165,23 @@ static void mid_pipe_event_handler(struct drm_device *dev, int pipe)
> >                 "%s, can't clear status bits for pipe %d, its value = 0x%x.\n",
> >                 __func__, pipe, PSB_RVDC32(pipe_stat_reg));
> >
> > -       if (pipe_stat_val & PIPE_VBLANK_STATUS)
> > -               drm_handle_vblank(dev, pipe);
> > +       if (pipe_stat_val & PIPE_VBLANK_STATUS ||
> > +           (IS_MFLD(dev) && pipe_stat_val & PIPE_TE_STATUS)) {
> > +               struct drm_crtc *crtc = drm_crtc_from_index(dev, pipe);
> > +               struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
> > +               unsigned long flags;
> >
> > -       if (pipe_stat_val & PIPE_TE_STATUS)
> >                 drm_handle_vblank(dev, pipe);
> > +
> > +               spin_lock_irqsave(&dev->event_lock, flags);
> > +               if (gma_crtc->page_flip_event) {
> > +                       drm_crtc_send_vblank_event(crtc,
> > +                                                  gma_crtc->page_flip_event);
> > +                       gma_crtc->page_flip_event = NULL;
> > +                       drm_crtc_vblank_put(crtc);
> > +               }
> > +               spin_unlock_irqrestore(&dev->event_lock, flags);
> > +       }
> >  }
> >
> >  /*
> > --
> > 2.23.0
> >
> 
> Looks good!
> 
> Reviewed-by: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>

I'm assuming you'll also push these?

Always confusing when maintainer/committers r-b but don't say anything
about pushing the patch. Good chances it'll fall through cracks if that
happens.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch