r300g: implement hyper-z support. (v4)

This implements fast Z clear, Z compression, and HiZ support for r300->r500
GPUs.

It also allows cbzb clears when fast Z clears are being used for the ZB.

It requires a kernel with hyper-z support.

Thanks to Marek Olšák <maraeo@gmail.com>, who started this off, and Alex Deucher at AMD for providing lots of hints.

v2:
squashed zmask ram size fix]
squashed r300g/blitter: fix Z readback when compressed]

v3:
rebase around texture changes in master - .1 fix more bits

v4:
migrated to using u_mm in r300_texture to manage hiz/zmask rams consistently
disabled HiZ when using OQ
flush z-cache before turning hyper-z off
update hyper-z state on dsa state change
store depthclearvalue across cbzb clears and replace it afterwards.

Signed-off-by: Dave Airlie <airlied@redhat.com>

2 files changed, 46 insertions, 0 deletions

diff --git a/src/gallium/auxiliary/util/u_blitter.c b/src/gallium/auxiliary/util/u_blitter.c
index 0d94aaae95..b5b86b7214 100644
--- a/src/gallium/auxiliary/util/u_blitter.c
+++ b/src/gallium/auxiliary/util/u_blitter.c
@@ -87,6 +87,7 @@ struct blitter_context_priv
    void *dsa_write_depth_keep_stencil;
    void *dsa_keep_depth_stencil;
    void *dsa_keep_depth_write_stencil;
+   void *dsa_flush_depth_stencil;
 
    void *velem_state;
 
@@ -156,6 +157,10 @@ struct blitter_context *util_blitter_create(struct pipe_context *pipe)
    ctx->dsa_keep_depth_stencil =
       pipe->create_depth_stencil_alpha_state(pipe, &dsa);
 
+   dsa.depth.writemask = 1;
+   ctx->dsa_flush_depth_stencil =
+      pipe->create_depth_stencil_alpha_state(pipe, &dsa);
+
    dsa.depth.enabled = 1;
    dsa.depth.writemask = 1;
    dsa.depth.func = PIPE_FUNC_ALWAYS;
@@ -940,3 +945,42 @@ void util_blitter_clear_depth_stencil(struct blitter_context *blitter,
                            UTIL_BLITTER_ATTRIB_NONE, NULL);
    blitter_restore_CSOs(ctx);
 }
+
+/* Clear a region of a depth stencil surface. */
+void util_blitter_flush_depth_stencil(struct blitter_context *blitter,
+                                      struct pipe_surface *dstsurf)
+{
+   struct blitter_context_priv *ctx = (struct blitter_context_priv*)blitter;
+   struct pipe_context *pipe = ctx->base.pipe;
+   struct pipe_framebuffer_state fb_state;
+
+   assert(dstsurf->texture);
+   if (!dstsurf->texture)
+      return;
+
+   /* check the saved state */
+   blitter_check_saved_CSOs(ctx);
+   assert(blitter->saved_fb_state.nr_cbufs != ~0);
+
+   /* bind CSOs */
+   pipe->bind_blend_state(pipe, ctx->blend_keep_color);
+   pipe->bind_depth_stencil_alpha_state(pipe, ctx->dsa_flush_depth_stencil);
+
+   pipe->bind_rasterizer_state(pipe, ctx->rs_state);
+   pipe->bind_fs_state(pipe, blitter_get_fs_col(ctx, 0));
+   pipe->bind_vs_state(pipe, ctx->vs_col);
+   pipe->bind_vertex_elements_state(pipe, ctx->velem_state);
+
+   /* set a framebuffer state */
+   fb_state.width = dstsurf->width;
+   fb_state.height = dstsurf->height;
+   fb_state.nr_cbufs = 0;
+   fb_state.cbufs[0] = 0;
+   fb_state.zsbuf = dstsurf;
+   pipe->set_framebuffer_state(pipe, &fb_state);
+
+   blitter_set_dst_dimensions(ctx, dstsurf->width, dstsurf->height);
+   blitter->draw_rectangle(blitter, 0, 0, dstsurf->width, dstsurf->height, 0,
+                           UTIL_BLITTER_ATTRIB_NONE, NULL);
+   blitter_restore_CSOs(ctx);
+}
diff --git a/src/gallium/auxiliary/util/u_blitter.h b/src/gallium/auxiliary/util/u_blitter.h
index ba3f92eca8..f316587dea 100644
--- a/src/gallium/auxiliary/util/u_blitter.h
+++ b/src/gallium/auxiliary/util/u_blitter.h
@@ -200,6 +200,8 @@ void util_blitter_clear_depth_stencil(struct blitter_context *blitter,
                                       unsigned dstx, unsigned dsty,
                                       unsigned width, unsigned height);
 
+void util_blitter_flush_depth_stencil(struct blitter_context *blitter,
+                                      struct pipe_surface *dstsurf);
 /* The functions below should be used to save currently bound constant state
  * objects inside a driver. The objects are automatically restored at the end
  * of the util_blitter_{clear, copy_region, fill_region} functions and then