drm/i915: Allow mmio updates on all platforms, v2.
authorMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Tue, 17 May 2016 13:07:51 +0000 (15:07 +0200)
committerMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Thu, 19 May 2016 12:37:37 +0000 (14:37 +0200)
With intel_pipe_update begin/end we ensure that the mmio updates
don't run during vblank interrupt, using the hw counter we can
be sure that when current vblank count != vblank count at the time
of pipe_update_end the mmio update is complete.

This allows us to use mmio updates on all platforms, using the
update_plane call.

With Chris Wilson's patch to skip waiting for vblanks for
legacy_cursor_update this potentially leaves a small race
condition, in which update_plane can be called with a freed
crtc_state. Because of this commit acf4e84d61673
("drm/i915: Avoid stalling on pending flips for legacy cursor updates")
is temporarily reverted.

Changes since v1:
- Split out the flip_work rename.

Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1463490484-19540-9-git-send-email-maarten.lankhorst@linux.intel.com
Reviewed-by: Patrik Jakobsson <patrik.jakobsson@linux.intel.com>
drivers/gpu/drm/i915/intel_display.c
drivers/gpu/drm/i915/intel_drv.h

index a432fdc6fe683be079d0d4c854ec641ca671d550..9c57736fc521d6738b27332a77b145a6e8083f49 100644 (file)
@@ -11269,9 +11269,6 @@ static bool use_mmio_flip(struct intel_engine_cs *engine,
        if (engine == NULL)
                return true;
 
-       if (INTEL_GEN(engine->i915) < 5)
-               return false;
-
        if (i915.use_mmio_flip < 0)
                return false;
        else if (i915.use_mmio_flip > 0)
@@ -11286,92 +11283,15 @@ static bool use_mmio_flip(struct intel_engine_cs *engine,
                return engine != i915_gem_request_get_engine(obj->last_write_req);
 }
 
-static void skl_do_mmio_flip(struct intel_crtc *intel_crtc,
-                            unsigned int rotation,
-                            struct intel_flip_work *work)
-{
-       struct drm_device *dev = intel_crtc->base.dev;
-       struct drm_i915_private *dev_priv = dev->dev_private;
-       struct drm_framebuffer *fb = intel_crtc->base.primary->fb;
-       const enum pipe pipe = intel_crtc->pipe;
-       u32 ctl, stride, tile_height;
-
-       ctl = I915_READ(PLANE_CTL(pipe, 0));
-       ctl &= ~PLANE_CTL_TILED_MASK;
-       switch (fb->modifier[0]) {
-       case DRM_FORMAT_MOD_NONE:
-               break;
-       case I915_FORMAT_MOD_X_TILED:
-               ctl |= PLANE_CTL_TILED_X;
-               break;
-       case I915_FORMAT_MOD_Y_TILED:
-               ctl |= PLANE_CTL_TILED_Y;
-               break;
-       case I915_FORMAT_MOD_Yf_TILED:
-               ctl |= PLANE_CTL_TILED_YF;
-               break;
-       default:
-               MISSING_CASE(fb->modifier[0]);
-       }
-
-       /*
-        * The stride is either expressed as a multiple of 64 bytes chunks for
-        * linear buffers or in number of tiles for tiled buffers.
-        */
-       if (intel_rotation_90_or_270(rotation)) {
-               /* stride = Surface height in tiles */
-               tile_height = intel_tile_height(dev_priv, fb->modifier[0], 0);
-               stride = DIV_ROUND_UP(fb->height, tile_height);
-       } else {
-               stride = fb->pitches[0] /
-                       intel_fb_stride_alignment(dev_priv, fb->modifier[0],
-                                                 fb->pixel_format);
-       }
-
-       /*
-        * Both PLANE_CTL and PLANE_STRIDE are not updated on vblank but on
-        * PLANE_SURF updates, the update is then guaranteed to be atomic.
-        */
-       I915_WRITE(PLANE_CTL(pipe, 0), ctl);
-       I915_WRITE(PLANE_STRIDE(pipe, 0), stride);
-
-       I915_WRITE(PLANE_SURF(pipe, 0), work->gtt_offset);
-       POSTING_READ(PLANE_SURF(pipe, 0));
-}
-
-static void ilk_do_mmio_flip(struct intel_crtc *intel_crtc,
-                            struct intel_flip_work *work)
-{
-       struct drm_device *dev = intel_crtc->base.dev;
-       struct drm_i915_private *dev_priv = dev->dev_private;
-       struct intel_framebuffer *intel_fb =
-               to_intel_framebuffer(intel_crtc->base.primary->fb);
-       struct drm_i915_gem_object *obj = intel_fb->obj;
-       i915_reg_t reg = DSPCNTR(intel_crtc->plane);
-       u32 dspcntr;
-
-       dspcntr = I915_READ(reg);
-
-       if (obj->tiling_mode != I915_TILING_NONE)
-               dspcntr |= DISPPLANE_TILED;
-       else
-               dspcntr &= ~DISPPLANE_TILED;
-
-       I915_WRITE(reg, dspcntr);
-
-       I915_WRITE(DSPSURF(intel_crtc->plane), work->gtt_offset);
-       POSTING_READ(DSPSURF(intel_crtc->plane));
-}
-
 static void intel_mmio_flip_work_func(struct work_struct *w)
 {
        struct intel_flip_work *work =
                container_of(w, struct intel_flip_work, mmio_work);
        struct intel_crtc *crtc = to_intel_crtc(work->crtc);
-       struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-       struct intel_framebuffer *intel_fb =
-               to_intel_framebuffer(crtc->base.primary->fb);
-       struct drm_i915_gem_object *obj = intel_fb->obj;
+       struct drm_device *dev = crtc->base.dev;
+       struct drm_i915_private *dev_priv = dev->dev_private;
+       struct intel_plane *primary = to_intel_plane(crtc->base.primary);
+       struct drm_i915_gem_object *obj = intel_fb_obj(primary->base.state->fb);
 
        if (work->flip_queued_req)
                WARN_ON(__i915_wait_request(work->flip_queued_req,
@@ -11385,13 +11305,9 @@ static void intel_mmio_flip_work_func(struct work_struct *w)
                                                            MAX_SCHEDULE_TIMEOUT) < 0);
 
        intel_pipe_update_start(crtc);
-
-       if (INTEL_GEN(dev_priv) >= 9)
-               skl_do_mmio_flip(crtc, work->rotation, work);
-       else
-               /* use_mmio_flip() retricts MMIO flips to ilk+ */
-               ilk_do_mmio_flip(crtc, work);
-
+       primary->update_plane(&primary->base,
+                             crtc->config,
+                             to_intel_plane_state(primary->base.state));
        intel_pipe_update_end(crtc, work);
 }
 
@@ -11616,7 +11532,6 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
        work->gtt_offset = intel_plane_obj_offset(to_intel_plane(primary),
                                                  obj, 0);
        work->gtt_offset += intel_crtc->dspaddr_offset;
-       work->rotation = crtc->primary->state->rotation;
 
        if (mmio_flip) {
                INIT_WORK(&work->mmio_work, intel_mmio_flip_work_func);
index 97de5e05890a0805bcd9451f3be1a672ad5543be..9502e79b2f5b62c06ac44c0be1c5a1ce442d7e7c 100644 (file)
@@ -983,7 +983,6 @@ struct intel_flip_work {
        struct drm_i915_gem_request *flip_queued_req;
        u32 flip_queued_vblank;
        u32 flip_ready_vblank;
-       unsigned int rotation;
 };
 
 struct intel_load_detect_pipe {