[Mesa-dev] [PATCH 13/13] i965/blorp: Convert state setup to C

Jason Ekstrand jason at jlekstrand.net
Fri Apr 22 23:19:20 UTC 2016


---
 src/mesa/drivers/dri/i965/Makefile.sources |    6 +-
 src/mesa/drivers/dri/i965/gen6_blorp.c     | 1097 ++++++++++++++++++++++++++++
 src/mesa/drivers/dri/i965/gen6_blorp.cpp   | 1097 ----------------------------
 src/mesa/drivers/dri/i965/gen7_blorp.c     |  884 ++++++++++++++++++++++
 src/mesa/drivers/dri/i965/gen7_blorp.cpp   |  884 ----------------------
 src/mesa/drivers/dri/i965/gen8_blorp.c     |  713 ++++++++++++++++++
 src/mesa/drivers/dri/i965/gen8_blorp.cpp   |  713 ------------------
 7 files changed, 2697 insertions(+), 2697 deletions(-)
 create mode 100644 src/mesa/drivers/dri/i965/gen6_blorp.c
 delete mode 100644 src/mesa/drivers/dri/i965/gen6_blorp.cpp
 create mode 100644 src/mesa/drivers/dri/i965/gen7_blorp.c
 delete mode 100644 src/mesa/drivers/dri/i965/gen7_blorp.cpp
 create mode 100644 src/mesa/drivers/dri/i965/gen8_blorp.c
 delete mode 100644 src/mesa/drivers/dri/i965/gen8_blorp.cpp

diff --git a/src/mesa/drivers/dri/i965/Makefile.sources b/src/mesa/drivers/dri/i965/Makefile.sources
index fe70d3a..441d727 100644
--- a/src/mesa/drivers/dri/i965/Makefile.sources
+++ b/src/mesa/drivers/dri/i965/Makefile.sources
@@ -177,7 +177,7 @@ i965_FILES = \
 	brw_wm.h \
 	brw_wm_state.c \
 	brw_wm_surface_state.c \
-	gen6_blorp.cpp \
+	gen6_blorp.c \
 	gen6_cc.c \
 	gen6_clip_state.c \
 	gen6_constant_state.c \
@@ -195,7 +195,7 @@ i965_FILES = \
 	gen6_viewport_state.c \
 	gen6_vs_state.c \
 	gen6_wm_state.c \
-	gen7_blorp.cpp \
+	gen7_blorp.c \
 	gen7_cs_state.c \
 	gen7_ds_state.c \
 	gen7_gs_state.c \
@@ -211,7 +211,7 @@ i965_FILES = \
 	gen7_wm_state.c \
 	gen7_wm_surface_state.c \
 	gen8_blend_state.c \
-	gen8_blorp.cpp \
+	gen8_blorp.c \
 	gen8_depth_state.c \
 	gen8_disable.c \
 	gen8_draw_upload.c \
diff --git a/src/mesa/drivers/dri/i965/gen6_blorp.c b/src/mesa/drivers/dri/i965/gen6_blorp.c
new file mode 100644
index 0000000..dabe8b5
--- /dev/null
+++ b/src/mesa/drivers/dri/i965/gen6_blorp.c
@@ -0,0 +1,1097 @@
+/*
+ * Copyright © 2011 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include <assert.h>
+
+#include "intel_batchbuffer.h"
+#include "intel_mipmap_tree.h"
+
+#include "brw_context.h"
+#include "brw_state.h"
+
+#include "brw_blorp.h"
+
+/**
+ * CMD_STATE_BASE_ADDRESS
+ *
+ * From the Sandy Bridge PRM, Volume 1, Part 1, Table STATE_BASE_ADDRESS:
+ *     The following commands must be reissued following any change to the
+ *     base addresses:
+ *         3DSTATE_CC_POINTERS
+ *         3DSTATE_BINDING_TABLE_POINTERS
+ *         3DSTATE_SAMPLER_STATE_POINTERS
+ *         3DSTATE_VIEWPORT_STATE_POINTERS
+ *         MEDIA_STATE_POINTERS
+ */
+void
+gen6_blorp_emit_state_base_address(struct brw_context *brw,
+                                   const struct brw_blorp_params *params)
+{
+   uint8_t mocs = brw->gen == 7 ? GEN7_MOCS_L3 : 0;
+
+   BEGIN_BATCH(10);
+   OUT_BATCH(CMD_STATE_BASE_ADDRESS << 16 | (10 - 2));
+   OUT_BATCH(mocs << 8 | /* GeneralStateMemoryObjectControlState */
+             mocs << 4 | /* StatelessDataPortAccessMemoryObjectControlState */
+             1); /* GeneralStateBaseAddressModifyEnable */
+
+   /* SurfaceStateBaseAddress */
+   OUT_RELOC(brw->batch.bo, I915_GEM_DOMAIN_SAMPLER, 0, 1);
+   /* DynamicStateBaseAddress */
+   OUT_RELOC(brw->batch.bo, (I915_GEM_DOMAIN_RENDER |
+                               I915_GEM_DOMAIN_INSTRUCTION), 0, 1);
+   OUT_BATCH(1); /* IndirectObjectBaseAddress */
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
+      OUT_RELOC(brw->cache.bo, I915_GEM_DOMAIN_INSTRUCTION, 0,
+                1); /* Instruction base address: shader kernels */
+   } else {
+      OUT_BATCH(1); /* InstructionBaseAddress */
+   }
+   OUT_BATCH(1); /* GeneralStateUpperBound */
+   /* Dynamic state upper bound.  Although the documentation says that
+    * programming it to zero will cause it to be ignored, that is a lie.
+    * If this isn't programmed to a real bound, the sampler border color
+    * pointer is rejected, causing border color to mysteriously fail.
+    */
+   OUT_BATCH(0xfffff001);
+   OUT_BATCH(1); /* IndirectObjectUpperBound*/
+   OUT_BATCH(1); /* InstructionAccessUpperBound */
+   ADVANCE_BATCH();
+}
+
+static void
+gen6_blorp_emit_vertex_buffer_state(struct brw_context *brw,
+                                    unsigned num_elems,
+                                    unsigned vbo_size,
+                                    uint32_t vertex_offset)
+{
+   /* 3DSTATE_VERTEX_BUFFERS */
+   const int num_buffers = 1;
+   const int batch_length = 1 + 4 * num_buffers;
+
+   uint32_t dw0 = GEN6_VB0_ACCESS_VERTEXDATA |
+                  (num_elems * sizeof(float)) << BRW_VB0_PITCH_SHIFT;
+
+   if (brw->gen >= 7)
+      dw0 |= GEN7_VB0_ADDRESS_MODIFYENABLE;
+
+   switch (brw->gen) {
+   case 7:
+      dw0 |= GEN7_MOCS_L3 << 16;
+      break;
+   case 8:
+      dw0 |= BDW_MOCS_WB << 16;
+      break;
+   case 9:
+      dw0 |= SKL_MOCS_WB << 16;
+      break;
+   }
+
+   BEGIN_BATCH(batch_length);
+   OUT_BATCH((_3DSTATE_VERTEX_BUFFERS << 16) | (batch_length - 2));
+   OUT_BATCH(dw0);
+   if (brw->gen >= 8) {
+      OUT_RELOC64(brw->batch.bo, I915_GEM_DOMAIN_VERTEX, 0, vertex_offset);
+      OUT_BATCH(vbo_size);
+   } else {
+      /* start address */
+      OUT_RELOC(brw->batch.bo, I915_GEM_DOMAIN_VERTEX, 0,
+                vertex_offset);
+      /* end address */
+      OUT_RELOC(brw->batch.bo, I915_GEM_DOMAIN_VERTEX, 0,
+                vertex_offset + vbo_size - 1);
+      OUT_BATCH(0);
+   }
+   ADVANCE_BATCH();
+}
+
+void
+gen6_blorp_emit_vertices(struct brw_context *brw,
+                         const struct brw_blorp_params *params)
+{
+   uint32_t vertex_offset;
+
+   /* Setup VBO for the rectangle primitive..
+    *
+    * A rectangle primitive (3DPRIM_RECTLIST) consists of only three
+    * vertices. The vertices reside in screen space with DirectX coordinates
+    * (that is, (0, 0) is the upper left corner).
+    *
+    *   v2 ------ implied
+    *    |        |
+    *    |        |
+    *   v0 ----- v1
+    *
+    * Since the VS is disabled, the clipper loads each VUE directly from
+    * the URB. This is controlled by the 3DSTATE_VERTEX_BUFFERS and
+    * 3DSTATE_VERTEX_ELEMENTS packets below. The VUE contents are as follows:
+    *   dw0: Reserved, MBZ.
+    *   dw1: Render Target Array Index. The HiZ op does not use indexed
+    *        vertices, so set the dword to 0.
+    *   dw2: Viewport Index. The HiZ op disables viewport mapping and
+    *        scissoring, so set the dword to 0.
+    *   dw3: Point Width: The HiZ op does not emit the POINTLIST primitive, so
+    *        set the dword to 0.
+    *   dw4: Vertex Position X.
+    *   dw5: Vertex Position Y.
+    *   dw6: Vertex Position Z.
+    *   dw7: Vertex Position W.
+    *
+    * For details, see the Sandybridge PRM, Volume 2, Part 1, Section 1.5.1
+    * "Vertex URB Entry (VUE) Formats".
+    *
+    * Only vertex position X and Y are going to be variable, Z is fixed to
+    * zero and W to one. Header words dw0-3 are all zero. There is no need to
+    * include the fixed values in the vertex buffer. Vertex fetcher can be
+    * instructed to fill vertex elements with constant values of one and zero
+    * instead of reading them from the buffer. See the vertex element setup
+    * below.
+    */
+   {
+      float *vertex_data;
+
+      const float vertices[] = {
+         /* v0 */ (float)params->x0, (float)params->y1,
+         /* v1 */ (float)params->x1, (float)params->y1,
+         /* v2 */ (float)params->x0, (float)params->y0,
+      };
+
+      vertex_data = (float *) brw_state_batch(brw, AUB_TRACE_VERTEX_BUFFER,
+                                              sizeof(vertices), 32,
+                                              &vertex_offset);
+      memcpy(vertex_data, vertices, sizeof(vertices));
+
+      const unsigned blorp_num_vue_elems = 2;
+      gen6_blorp_emit_vertex_buffer_state(brw, blorp_num_vue_elems,
+                                          sizeof(vertices), vertex_offset);
+   }
+
+   /* 3DSTATE_VERTEX_ELEMENTS
+    *
+    * Fetch dwords 0 - 7 from each VUE. See the comments above where
+    * the vertex_bo is filled with data.
+    */
+   {
+      const int num_elements = 2;
+      const int batch_length = 1 + 2 * num_elements;
+
+      BEGIN_BATCH(batch_length);
+      OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | (batch_length - 2));
+      /* Element 0 */
+      OUT_BATCH(GEN6_VE0_VALID |
+                BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT |
+                0 << BRW_VE0_SRC_OFFSET_SHIFT);
+      OUT_BATCH(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_0_SHIFT |
+                BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT |
+                BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT |
+                BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_3_SHIFT);
+      /* Element 1 */
+      OUT_BATCH(GEN6_VE0_VALID |
+                BRW_SURFACEFORMAT_R32G32_FLOAT << BRW_VE0_FORMAT_SHIFT |
+                0 << BRW_VE0_SRC_OFFSET_SHIFT);
+      OUT_BATCH(BRW_VE1_COMPONENT_STORE_SRC << BRW_VE1_COMPONENT_0_SHIFT |
+                BRW_VE1_COMPONENT_STORE_SRC << BRW_VE1_COMPONENT_1_SHIFT |
+                BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT |
+                BRW_VE1_COMPONENT_STORE_1_FLT << BRW_VE1_COMPONENT_3_SHIFT);
+      ADVANCE_BATCH();
+   }
+}
+
+
+/* 3DSTATE_URB
+ *
+ * Assign the entire URB to the VS. Even though the VS disabled, URB space
+ * is still needed because the clipper loads the VUE's from the URB. From
+ * the Sandybridge PRM, Volume 2, Part 1, Section 3DSTATE,
+ * Dword 1.15:0 "VS Number of URB Entries":
+ *     This field is always used (even if VS Function Enable is DISABLED).
+ *
+ * The warning below appears in the PRM (Section 3DSTATE_URB), but we can
+ * safely ignore it because this batch contains only one draw call.
+ *     Because of URB corruption caused by allocating a previous GS unit
+ *     URB entry to the VS unit, software is required to send a “GS NULL
+ *     Fence” (Send URB fence with VS URB size == 1 and GS URB size == 0)
+ *     plus a dummy DRAW call before any case where VS will be taking over
+ *     GS URB space.
+ */
+static void
+gen6_blorp_emit_urb_config(struct brw_context *brw,
+                           const struct brw_blorp_params *params)
+{
+   BEGIN_BATCH(3);
+   OUT_BATCH(_3DSTATE_URB << 16 | (3 - 2));
+   OUT_BATCH(brw->urb.max_vs_entries << GEN6_URB_VS_ENTRIES_SHIFT);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/* BLEND_STATE */
+uint32_t
+gen6_blorp_emit_blend_state(struct brw_context *brw,
+                            const struct brw_blorp_params *params)
+{
+   uint32_t cc_blend_state_offset;
+
+   assume(params->num_draw_buffers);
+
+   const unsigned size = params->num_draw_buffers *
+                         sizeof(struct gen6_blend_state);
+   struct gen6_blend_state *blend = (struct gen6_blend_state *)
+      brw_state_batch(brw, AUB_TRACE_BLEND_STATE, size, 64,
+                      &cc_blend_state_offset);
+
+   memset(blend, 0, size);
+
+   for (unsigned i = 0; i < params->num_draw_buffers; ++i) {
+      blend[i].blend1.pre_blend_clamp_enable = 1;
+      blend[i].blend1.post_blend_clamp_enable = 1;
+      blend[i].blend1.clamp_range = BRW_RENDERTARGET_CLAMPRANGE_FORMAT;
+
+      blend[i].blend1.write_disable_r = params->color_write_disable[0];
+      blend[i].blend1.write_disable_g = params->color_write_disable[1];
+      blend[i].blend1.write_disable_b = params->color_write_disable[2];
+      blend[i].blend1.write_disable_a = params->color_write_disable[3];
+   }
+
+   return cc_blend_state_offset;
+}
+
+
+/* CC_STATE */
+uint32_t
+gen6_blorp_emit_cc_state(struct brw_context *brw)
+{
+   uint32_t cc_state_offset;
+
+   struct gen6_color_calc_state *cc = (struct gen6_color_calc_state *)
+      brw_state_batch(brw, AUB_TRACE_CC_STATE,
+                      sizeof(gen6_color_calc_state), 64,
+                      &cc_state_offset);
+   memset(cc, 0, sizeof(*cc));
+
+   return cc_state_offset;
+}
+
+
+/**
+ * \param out_offset is relative to
+ *        CMD_STATE_BASE_ADDRESS.DynamicStateBaseAddress.
+ */
+uint32_t
+gen6_blorp_emit_depth_stencil_state(struct brw_context *brw,
+                                    const struct brw_blorp_params *params)
+{
+   uint32_t depthstencil_offset;
+
+   struct gen6_depth_stencil_state *state;
+   state = (struct gen6_depth_stencil_state *)
+      brw_state_batch(brw, AUB_TRACE_DEPTH_STENCIL_STATE,
+                      sizeof(*state), 64,
+                      &depthstencil_offset);
+   memset(state, 0, sizeof(*state));
+
+   /* See the following sections of the Sandy Bridge PRM, Volume 1, Part2:
+    *   - 7.5.3.1 Depth Buffer Clear
+    *   - 7.5.3.2 Depth Buffer Resolve
+    *   - 7.5.3.3 Hierarchical Depth Buffer Resolve
+    */
+   state->ds2.depth_write_enable = 1;
+   if (params->hiz_op == GEN6_HIZ_OP_DEPTH_RESOLVE) {
+      state->ds2.depth_test_enable = 1;
+      state->ds2.depth_test_func = BRW_COMPAREFUNCTION_NEVER;
+   }
+
+   return depthstencil_offset;
+}
+
+
+/* 3DSTATE_CC_STATE_POINTERS
+ *
+ * The pointer offsets are relative to
+ * CMD_STATE_BASE_ADDRESS.DynamicStateBaseAddress.
+ *
+ * The HiZ op doesn't use BLEND_STATE or COLOR_CALC_STATE.
+ */
+static void
+gen6_blorp_emit_cc_state_pointers(struct brw_context *brw,
+                                  const struct brw_blorp_params *params,
+                                  uint32_t cc_blend_state_offset,
+                                  uint32_t depthstencil_offset,
+                                  uint32_t cc_state_offset)
+{
+   BEGIN_BATCH(4);
+   OUT_BATCH(_3DSTATE_CC_STATE_POINTERS << 16 | (4 - 2));
+   OUT_BATCH(cc_blend_state_offset | 1); /* BLEND_STATE offset */
+   OUT_BATCH(depthstencil_offset | 1); /* DEPTH_STENCIL_STATE offset */
+   OUT_BATCH(cc_state_offset | 1); /* COLOR_CALC_STATE offset */
+   ADVANCE_BATCH();
+}
+
+
+/* WM push constants */
+uint32_t
+gen6_blorp_emit_wm_constants(struct brw_context *brw,
+                             const struct brw_blorp_params *params)
+{
+   uint32_t wm_push_const_offset;
+
+   void *constants = brw_state_batch(brw, AUB_TRACE_WM_CONSTANTS,
+                                     sizeof(params->wm_push_consts),
+                                     32, &wm_push_const_offset);
+   memcpy(constants, &params->wm_push_consts,
+          sizeof(params->wm_push_consts));
+
+   return wm_push_const_offset;
+}
+
+
+/* SURFACE_STATE for renderbuffer or texture surface (see
+ * brw_update_renderbuffer_surface and brw_update_texture_surface)
+ */
+static uint32_t
+gen6_blorp_emit_surface_state(struct brw_context *brw,
+                              const struct brw_blorp_params *params,
+                              const struct brw_blorp_surface_info *surface,
+                              uint32_t read_domains, uint32_t write_domain)
+{
+   uint32_t wm_surf_offset;
+   uint32_t width = surface->width;
+   uint32_t height = surface->height;
+   if (surface->num_samples > 1) {
+      /* Since gen6 uses INTEL_MSAA_LAYOUT_IMS, width and height are measured
+       * in samples.  But SURFACE_STATE wants them in pixels, so we need to
+       * divide them each by 2.
+       */
+      width /= 2;
+      height /= 2;
+   }
+   struct intel_mipmap_tree *mt = surface->mt;
+   uint32_t tile_x, tile_y;
+
+   uint32_t *surf = (uint32_t *)
+      brw_state_batch(brw, AUB_TRACE_SURFACE_STATE, 6 * 4, 32,
+                      &wm_surf_offset);
+
+   surf[0] = (BRW_SURFACE_2D << BRW_SURFACE_TYPE_SHIFT |
+              BRW_SURFACE_MIPMAPLAYOUT_BELOW << BRW_SURFACE_MIPLAYOUT_SHIFT |
+              BRW_SURFACE_CUBEFACE_ENABLES |
+              surface->brw_surfaceformat << BRW_SURFACE_FORMAT_SHIFT);
+
+   /* reloc */
+   surf[1] = (brw_blorp_compute_tile_offsets(surface, &tile_x, &tile_y) +
+              mt->bo->offset64);
+
+   surf[2] = (0 << BRW_SURFACE_LOD_SHIFT |
+              (width - 1) << BRW_SURFACE_WIDTH_SHIFT |
+              (height - 1) << BRW_SURFACE_HEIGHT_SHIFT);
+
+   uint32_t tiling = surface->map_stencil_as_y_tiled
+      ? BRW_SURFACE_TILED | BRW_SURFACE_TILED_Y
+      : brw_get_surface_tiling_bits(mt->tiling);
+   uint32_t pitch_bytes = mt->pitch;
+   if (surface->map_stencil_as_y_tiled)
+      pitch_bytes *= 2;
+   surf[3] = (tiling |
+              0 << BRW_SURFACE_DEPTH_SHIFT |
+              (pitch_bytes - 1) << BRW_SURFACE_PITCH_SHIFT);
+
+   surf[4] = brw_get_surface_num_multisamples(surface->num_samples);
+
+   /* Note that the low bits of these fields are missing, so
+    * there's the possibility of getting in trouble.
+    */
+   assert(tile_x % 4 == 0);
+   assert(tile_y % 2 == 0);
+   surf[5] = ((tile_x / 4) << BRW_SURFACE_X_OFFSET_SHIFT |
+              (tile_y / 2) << BRW_SURFACE_Y_OFFSET_SHIFT |
+              (surface->mt->valign == 4 ?
+               BRW_SURFACE_VERTICAL_ALIGN_ENABLE : 0));
+
+   /* Emit relocation to surface contents */
+   drm_intel_bo_emit_reloc(brw->batch.bo,
+                           wm_surf_offset + 4,
+                           mt->bo,
+                           surf[1] - mt->bo->offset64,
+                           read_domains, write_domain);
+
+   return wm_surf_offset;
+}
+
+
+/* BINDING_TABLE.  See brw_wm_binding_table(). */
+uint32_t
+gen6_blorp_emit_binding_table(struct brw_context *brw,
+                              uint32_t wm_surf_offset_renderbuffer,
+                              uint32_t wm_surf_offset_texture)
+{
+   uint32_t wm_bind_bo_offset;
+   uint32_t *bind = (uint32_t *)
+      brw_state_batch(brw, AUB_TRACE_BINDING_TABLE,
+                      sizeof(uint32_t) *
+                      BRW_BLORP_NUM_BINDING_TABLE_ENTRIES,
+                      32, /* alignment */
+                      &wm_bind_bo_offset);
+   bind[BRW_BLORP_RENDERBUFFER_BINDING_TABLE_INDEX] =
+      wm_surf_offset_renderbuffer;
+   bind[BRW_BLORP_TEXTURE_BINDING_TABLE_INDEX] = wm_surf_offset_texture;
+
+   return wm_bind_bo_offset;
+}
+
+
+/**
+ * SAMPLER_STATE.  See brw_update_sampler_state().
+ */
+uint32_t
+gen6_blorp_emit_sampler_state(struct brw_context *brw,
+                              unsigned tex_filter, unsigned max_lod,
+                              bool non_normalized_coords)
+{
+   uint32_t sampler_offset;
+   uint32_t *sampler_state = (uint32_t *)
+      brw_state_batch(brw, AUB_TRACE_SAMPLER_STATE, 16, 32, &sampler_offset);
+
+   unsigned address_rounding = BRW_ADDRESS_ROUNDING_ENABLE_U_MIN |
+                               BRW_ADDRESS_ROUNDING_ENABLE_V_MIN |
+                               BRW_ADDRESS_ROUNDING_ENABLE_R_MIN |
+                               BRW_ADDRESS_ROUNDING_ENABLE_U_MAG |
+                               BRW_ADDRESS_ROUNDING_ENABLE_V_MAG |
+                               BRW_ADDRESS_ROUNDING_ENABLE_R_MAG;
+
+   /* XXX: I don't think that using firstLevel, lastLevel works,
+    * because we always setup the surface state as if firstLevel ==
+    * level zero.  Probably have to subtract firstLevel from each of
+    * these:
+    */
+   brw_emit_sampler_state(brw,
+                          sampler_state,
+                          sampler_offset,
+                          tex_filter, /* min filter */
+                          tex_filter, /* mag filter */
+                          BRW_MIPFILTER_NONE,
+                          BRW_ANISORATIO_2,
+                          address_rounding,
+                          BRW_TEXCOORDMODE_CLAMP,
+                          BRW_TEXCOORDMODE_CLAMP,
+                          BRW_TEXCOORDMODE_CLAMP,
+                          0, /* min LOD */
+                          max_lod,
+                          0, /* LOD bias */
+                          0, /* shadow function */
+                          non_normalized_coords,
+                          0); /* border color offset - unused */
+
+   return sampler_offset;
+}
+
+
+/**
+ * 3DSTATE_SAMPLER_STATE_POINTERS.  See upload_sampler_state_pointers().
+ */
+static void
+gen6_blorp_emit_sampler_state_pointers(struct brw_context *brw,
+                                       uint32_t sampler_offset)
+{
+   BEGIN_BATCH(4);
+   OUT_BATCH(_3DSTATE_SAMPLER_STATE_POINTERS << 16 |
+             VS_SAMPLER_STATE_CHANGE |
+             GS_SAMPLER_STATE_CHANGE |
+             PS_SAMPLER_STATE_CHANGE |
+             (4 - 2));
+   OUT_BATCH(0); /* VS */
+   OUT_BATCH(0); /* GS */
+   OUT_BATCH(sampler_offset);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_VS
+ *
+ * Disable vertex shader.
+ */
+void
+gen6_blorp_emit_vs_disable(struct brw_context *brw,
+                           const struct brw_blorp_params *params)
+{
+   /* From the BSpec, 3D Pipeline > Geometry > Vertex Shader > State,
+    * 3DSTATE_VS, Dword 5.0 "VS Function Enable":
+    *
+    *   [DevSNB] A pipeline flush must be programmed prior to a
+    *   3DSTATE_VS command that causes the VS Function Enable to
+    *   toggle. Pipeline flush can be executed by sending a PIPE_CONTROL
+    *   command with CS stall bit set and a post sync operation.
+    *
+    * We've already done one at the start of the BLORP operation.
+    */
+
+   /* Disable the push constant buffers. */
+   BEGIN_BATCH(5);
+   OUT_BATCH(_3DSTATE_CONSTANT_VS << 16 | (5 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(6);
+   OUT_BATCH(_3DSTATE_VS << 16 | (6 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_GS
+ *
+ * Disable the geometry shader.
+ */
+void
+gen6_blorp_emit_gs_disable(struct brw_context *brw,
+                           const struct brw_blorp_params *params)
+{
+   /* Disable all the constant buffers. */
+   BEGIN_BATCH(5);
+   OUT_BATCH(_3DSTATE_CONSTANT_GS << 16 | (5 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_GS << 16 | (7 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+   brw->gs.enabled = false;
+}
+
+
+/* 3DSTATE_CLIP
+ *
+ * Disable the clipper.
+ *
+ * The BLORP op emits a rectangle primitive, which requires clipping to
+ * be disabled. From page 10 of the Sandy Bridge PRM Volume 2 Part 1
+ * Section 1.3 "3D Primitives Overview":
+ *    RECTLIST:
+ *    Either the CLIP unit should be DISABLED, or the CLIP unit's Clip
+ *    Mode should be set to a value other than CLIPMODE_NORMAL.
+ *
+ * Also disable perspective divide. This doesn't change the clipper's
+ * output, but does spare a few electrons.
+ */
+void
+gen6_blorp_emit_clip_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(4);
+   OUT_BATCH(_3DSTATE_CLIP << 16 | (4 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(GEN6_CLIP_PERSPECTIVE_DIVIDE_DISABLE);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_SF
+ *
+ * Disable ViewportTransformEnable (dw2.1)
+ *
+ * From the SandyBridge PRM, Volume 2, Part 1, Section 1.3, "3D
+ * Primitives Overview":
+ *     RECTLIST: Viewport Mapping must be DISABLED (as is typical with the
+ *     use of screen- space coordinates).
+ *
+ * A solid rectangle must be rendered, so set FrontFaceFillMode (dw2.4:3)
+ * and BackFaceFillMode (dw2.5:6) to SOLID(0).
+ *
+ * From the Sandy Bridge PRM, Volume 2, Part 1, Section
+ * 6.4.1.1 3DSTATE_SF, Field FrontFaceFillMode:
+ *     SOLID: Any triangle or rectangle object found to be front-facing
+ *     is rendered as a solid object. This setting is required when
+ *     (rendering rectangle (RECTLIST) objects.
+ */
+static void
+gen6_blorp_emit_sf_config(struct brw_context *brw,
+                          const struct brw_blorp_params *params)
+{
+   BEGIN_BATCH(20);
+   OUT_BATCH(_3DSTATE_SF << 16 | (20 - 2));
+   OUT_BATCH(params->num_varyings << GEN6_SF_NUM_OUTPUTS_SHIFT |
+             1 << GEN6_SF_URB_ENTRY_READ_LENGTH_SHIFT |
+             BRW_SF_URB_ENTRY_READ_OFFSET <<
+                GEN6_SF_URB_ENTRY_READ_OFFSET_SHIFT);
+   OUT_BATCH(0); /* dw2 */
+   OUT_BATCH(params->dst.num_samples > 1 ? GEN6_SF_MSRAST_ON_PATTERN : 0);
+   for (int i = 0; i < 16; ++i)
+      OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/**
+ * Enable or disable thread dispatch and set the HiZ op appropriately.
+ */
+static void
+gen6_blorp_emit_wm_config(struct brw_context *brw,
+                          const struct brw_blorp_params *params)
+{
+   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
+   uint32_t dw2, dw4, dw5, dw6;
+
+   /* Even when thread dispatch is disabled, max threads (dw5.25:31) must be
+    * nonzero to prevent the GPU from hanging.  While the documentation doesn't
+    * mention this explicitly, it notes that the valid range for the field is
+    * [1,39] = [2,40] threads, which excludes zero.
+    *
+    * To be safe (and to minimize extraneous code) we go ahead and fully
+    * configure the WM state whether or not there is a WM program.
+    */
+
+   dw2 = dw4 = dw5 = dw6 = 0;
+   switch (params->hiz_op) {
+   case GEN6_HIZ_OP_DEPTH_CLEAR:
+      dw4 |= GEN6_WM_DEPTH_CLEAR;
+      break;
+   case GEN6_HIZ_OP_DEPTH_RESOLVE:
+      dw4 |= GEN6_WM_DEPTH_RESOLVE;
+      break;
+   case GEN6_HIZ_OP_HIZ_RESOLVE:
+      dw4 |= GEN6_WM_HIERARCHICAL_DEPTH_RESOLVE;
+      break;
+   case GEN6_HIZ_OP_NONE:
+      break;
+   default:
+      unreachable("not reached");
+   }
+   dw5 |= GEN6_WM_LINE_AA_WIDTH_1_0;
+   dw5 |= GEN6_WM_LINE_END_CAP_AA_WIDTH_0_5;
+   dw5 |= (brw->max_wm_threads - 1) << GEN6_WM_MAX_THREADS_SHIFT;
+   dw6 |= 0 << GEN6_WM_BARYCENTRIC_INTERPOLATION_MODE_SHIFT; /* No interp */
+   dw6 |= 0 << GEN6_WM_NUM_SF_OUTPUTS_SHIFT; /* No inputs from SF */
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
+      dw4 |= prog_data->first_curbe_grf << GEN6_WM_DISPATCH_START_GRF_SHIFT_0;
+      dw5 |= GEN6_WM_16_DISPATCH_ENABLE;
+      dw5 |= GEN6_WM_DISPATCH_ENABLE; /* We are rendering */
+   }
+
+   if (params->src.mt) {
+      dw5 |= GEN6_WM_KILL_ENABLE; /* TODO: temporarily smash on */
+      dw2 |= 1 << GEN6_WM_SAMPLER_COUNT_SHIFT; /* Up to 4 samplers */
+   }
+
+   if (params->dst.num_samples > 1) {
+      dw6 |= GEN6_WM_MSRAST_ON_PATTERN;
+      if (prog_data && prog_data->persample_msaa_dispatch)
+         dw6 |= GEN6_WM_MSDISPMODE_PERSAMPLE;
+      else
+         dw6 |= GEN6_WM_MSDISPMODE_PERPIXEL;
+   } else {
+      dw6 |= GEN6_WM_MSRAST_OFF_PIXEL;
+      dw6 |= GEN6_WM_MSDISPMODE_PERSAMPLE;
+   }
+
+   BEGIN_BATCH(9);
+   OUT_BATCH(_3DSTATE_WM << 16 | (9 - 2));
+   OUT_BATCH(params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG ?
+             params->wm_prog_kernel : 0);
+   OUT_BATCH(dw2);
+   OUT_BATCH(0); /* No scratch needed */
+   OUT_BATCH(dw4);
+   OUT_BATCH(dw5);
+   OUT_BATCH(dw6);
+   OUT_BATCH(0); /* No other programs */
+   OUT_BATCH(0); /* No other programs */
+   ADVANCE_BATCH();
+}
+
+
+static void
+gen6_blorp_emit_constant_ps(struct brw_context *brw,
+                            const struct brw_blorp_params *params,
+                            uint32_t wm_push_const_offset)
+{
+   /* Make sure the push constants fill an exact integer number of
+    * registers.
+    */
+   assert(sizeof(struct brw_blorp_wm_push_constants) % 32 == 0);
+
+   /* There must be at least one register worth of push constant data. */
+   assert(BRW_BLORP_NUM_PUSH_CONST_REGS > 0);
+
+   /* Enable push constant buffer 0. */
+   BEGIN_BATCH(5);
+   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 |
+             GEN6_CONSTANT_BUFFER_0_ENABLE |
+             (5 - 2));
+   OUT_BATCH(wm_push_const_offset + (BRW_BLORP_NUM_PUSH_CONST_REGS - 1));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+static void
+gen6_blorp_emit_constant_ps_disable(struct brw_context *brw,
+                                    const struct brw_blorp_params *params)
+{
+   /* Disable the push constant buffers. */
+   BEGIN_BATCH(5);
+   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 | (5 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+/**
+ * 3DSTATE_BINDING_TABLE_POINTERS
+ */
+static void
+gen6_blorp_emit_binding_table_pointers(struct brw_context *brw,
+                                       uint32_t wm_bind_bo_offset)
+{
+   BEGIN_BATCH(4);
+   OUT_BATCH(_3DSTATE_BINDING_TABLE_POINTERS << 16 |
+             GEN6_BINDING_TABLE_MODIFY_PS |
+             (4 - 2));
+   OUT_BATCH(0); /* vs -- ignored */
+   OUT_BATCH(0); /* gs -- ignored */
+   OUT_BATCH(wm_bind_bo_offset); /* wm/ps */
+   ADVANCE_BATCH();
+}
+
+
+static void
+gen6_blorp_emit_depth_stencil_config(struct brw_context *brw,
+                                     const struct brw_blorp_params *params)
+{
+   uint32_t surfwidth, surfheight;
+   uint32_t surftype;
+   unsigned int depth = MAX2(params->depth.mt->logical_depth0, 1);
+   GLenum gl_target = params->depth.mt->target;
+   unsigned int lod;
+
+   switch (gl_target) {
+   case GL_TEXTURE_CUBE_MAP_ARRAY:
+   case GL_TEXTURE_CUBE_MAP:
+      /* The PRM claims that we should use BRW_SURFACE_CUBE for this
+       * situation, but experiments show that gl_Layer doesn't work when we do
+       * this.  So we use BRW_SURFACE_2D, since for rendering purposes this is
+       * equivalent.
+       */
+      surftype = BRW_SURFACE_2D;
+      depth *= 6;
+      break;
+   default:
+      surftype = translate_tex_target(gl_target);
+      break;
+   }
+
+   const unsigned min_array_element = params->depth.layer;
+
+   lod = params->depth.level - params->depth.mt->first_level;
+
+   if (params->hiz_op != GEN6_HIZ_OP_NONE && lod == 0) {
+      /* HIZ ops for lod 0 may set the width & height a little
+       * larger to allow the fast depth clear to fit the hardware
+       * alignment requirements. (8x4)
+       */
+      surfwidth = params->depth.width;
+      surfheight = params->depth.height;
+   } else {
+      surfwidth = params->depth.mt->logical_width0;
+      surfheight = params->depth.mt->logical_height0;
+   }
+
+   /* 3DSTATE_DEPTH_BUFFER */
+   {
+      brw_emit_depth_stall_flushes(brw);
+
+      BEGIN_BATCH(7);
+      /* 3DSTATE_DEPTH_BUFFER dw0 */
+      OUT_BATCH(_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
+
+      /* 3DSTATE_DEPTH_BUFFER dw1 */
+      OUT_BATCH((params->depth.mt->pitch - 1) |
+                params->depth_format << 18 |
+                1 << 21 | /* separate stencil enable */
+                1 << 22 | /* hiz enable */
+                BRW_TILEWALK_YMAJOR << 26 |
+                1 << 27 | /* y-tiled */
+                surftype << 29);
+
+      /* 3DSTATE_DEPTH_BUFFER dw2 */
+      OUT_RELOC(params->depth.mt->bo,
+                I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
+                0);
+
+      /* 3DSTATE_DEPTH_BUFFER dw3 */
+      OUT_BATCH(BRW_SURFACE_MIPMAPLAYOUT_BELOW << 1 |
+                (surfwidth - 1) << 6 |
+                (surfheight - 1) << 19 |
+                lod << 2);
+
+      /* 3DSTATE_DEPTH_BUFFER dw4 */
+      OUT_BATCH((depth - 1) << 21 |
+                min_array_element << 10 |
+                (depth - 1) << 1);
+
+      /* 3DSTATE_DEPTH_BUFFER dw5 */
+      OUT_BATCH(0);
+
+      /* 3DSTATE_DEPTH_BUFFER dw6 */
+      OUT_BATCH(0);
+      ADVANCE_BATCH();
+   }
+
+   /* 3DSTATE_HIER_DEPTH_BUFFER */
+   {
+      struct intel_mipmap_tree *hiz_mt = params->depth.mt->hiz_buf->mt;
+      uint32_t offset = 0;
+
+      if (hiz_mt->array_layout == ALL_SLICES_AT_EACH_LOD) {
+         offset = intel_miptree_get_aligned_offset(hiz_mt,
+                                                   hiz_mt->level[lod].level_x,
+                                                   hiz_mt->level[lod].level_y,
+                                                   false);
+      }
+
+      BEGIN_BATCH(3);
+      OUT_BATCH((_3DSTATE_HIER_DEPTH_BUFFER << 16) | (3 - 2));
+      OUT_BATCH(hiz_mt->pitch - 1);
+      OUT_RELOC(hiz_mt->bo,
+                I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
+                offset);
+      ADVANCE_BATCH();
+   }
+
+   /* 3DSTATE_STENCIL_BUFFER */
+   {
+      BEGIN_BATCH(3);
+      OUT_BATCH((_3DSTATE_STENCIL_BUFFER << 16) | (3 - 2));
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      ADVANCE_BATCH();
+   }
+}
+
+
+static void
+gen6_blorp_emit_depth_disable(struct brw_context *brw,
+                              const struct brw_blorp_params *params)
+{
+   brw_emit_depth_stall_flushes(brw);
+
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
+   OUT_BATCH((BRW_DEPTHFORMAT_D32_FLOAT << 18) |
+             (BRW_SURFACE_NULL << 29));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(3);
+   OUT_BATCH(_3DSTATE_HIER_DEPTH_BUFFER << 16 | (3 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(3);
+   OUT_BATCH(_3DSTATE_STENCIL_BUFFER << 16 | (3 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_CLEAR_PARAMS
+ *
+ * From the Sandybridge PRM, Volume 2, Part 1, Section 3DSTATE_CLEAR_PARAMS:
+ *   [DevSNB] 3DSTATE_CLEAR_PARAMS packet must follow the DEPTH_BUFFER_STATE
+ *   packet when HiZ is enabled and the DEPTH_BUFFER_STATE changes.
+ */
+static void
+gen6_blorp_emit_clear_params(struct brw_context *brw,
+                             const struct brw_blorp_params *params)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_CLEAR_PARAMS << 16 |
+	     GEN5_DEPTH_CLEAR_VALID |
+	     (2 - 2));
+   OUT_BATCH(params->depth.mt ? params->depth.mt->depth_clear_value : 0);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_DRAWING_RECTANGLE */
+void
+gen6_blorp_emit_drawing_rectangle(struct brw_context *brw,
+                                  const struct brw_blorp_params *params)
+{
+   BEGIN_BATCH(4);
+   OUT_BATCH(_3DSTATE_DRAWING_RECTANGLE << 16 | (4 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(((MAX2(params->x1, params->x0) - 1) & 0xffff) |
+             ((MAX2(params->y1, params->y0) - 1) << 16));
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+/* 3DSTATE_VIEWPORT_STATE_POINTERS */
+static void
+gen6_blorp_emit_viewport_state(struct brw_context *brw,
+			       const struct brw_blorp_params *params)
+{
+   struct brw_cc_viewport *ccv;
+   uint32_t cc_vp_offset;
+
+   ccv = (struct brw_cc_viewport *)brw_state_batch(brw, AUB_TRACE_CC_VP_STATE,
+						   sizeof(*ccv), 32,
+						   &cc_vp_offset);
+
+   ccv->min_depth = 0.0;
+   ccv->max_depth = 1.0;
+
+   BEGIN_BATCH(4);
+   OUT_BATCH(_3DSTATE_VIEWPORT_STATE_POINTERS << 16 | (4 - 2) |
+	     GEN6_CC_VIEWPORT_MODIFY);
+   OUT_BATCH(0); /* clip VP */
+   OUT_BATCH(0); /* SF VP */
+   OUT_BATCH(cc_vp_offset);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DPRIMITIVE */
+static void
+gen6_blorp_emit_primitive(struct brw_context *brw,
+                          const struct brw_blorp_params *params)
+{
+   BEGIN_BATCH(6);
+   OUT_BATCH(CMD_3D_PRIM << 16 | (6 - 2) |
+             _3DPRIM_RECTLIST << GEN4_3DPRIM_TOPOLOGY_TYPE_SHIFT |
+             GEN4_3DPRIM_VERTEXBUFFER_ACCESS_SEQUENTIAL);
+   OUT_BATCH(3); /* vertex count per instance */
+   OUT_BATCH(0);
+   OUT_BATCH(params->num_layers); /* instance count */
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+/**
+ * \brief Execute a blit or render pass operation.
+ *
+ * To execute the operation, this function manually constructs and emits a
+ * batch to draw a rectangle primitive. The batchbuffer is flushed before
+ * constructing and after emitting the batch.
+ *
+ * This function alters no GL state.
+ */
+void
+gen6_blorp_exec(struct brw_context *brw,
+                const struct brw_blorp_params *params)
+{
+   uint32_t cc_blend_state_offset = 0;
+   uint32_t cc_state_offset = 0;
+   uint32_t depthstencil_offset;
+   uint32_t wm_push_const_offset = 0;
+   uint32_t wm_bind_bo_offset = 0;
+
+   /* Emit workaround flushes when we switch from drawing to blorping. */
+   brw_emit_post_sync_nonzero_flush(brw);
+
+   gen6_emit_3dstate_multisample(brw, params->dst.num_samples);
+   gen6_emit_3dstate_sample_mask(brw,
+                                 params->dst.num_samples > 1 ?
+                                 (1 << params->dst.num_samples) - 1 : 1);
+   gen6_blorp_emit_state_base_address(brw, params);
+   gen6_blorp_emit_vertices(brw, params);
+   gen6_blorp_emit_urb_config(brw, params);
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
+      cc_blend_state_offset = gen6_blorp_emit_blend_state(brw, params);
+      cc_state_offset = gen6_blorp_emit_cc_state(brw);
+   }
+   depthstencil_offset = gen6_blorp_emit_depth_stencil_state(brw, params);
+   gen6_blorp_emit_cc_state_pointers(brw, params, cc_blend_state_offset,
+                                     depthstencil_offset, cc_state_offset);
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
+      uint32_t wm_surf_offset_renderbuffer;
+      uint32_t wm_surf_offset_texture = 0;
+      wm_push_const_offset = gen6_blorp_emit_wm_constants(brw, params);
+      intel_miptree_used_for_rendering(params->dst.mt);
+      wm_surf_offset_renderbuffer =
+         gen6_blorp_emit_surface_state(brw, params, &params->dst,
+                                       I915_GEM_DOMAIN_RENDER,
+                                       I915_GEM_DOMAIN_RENDER);
+      if (params->src.mt) {
+         wm_surf_offset_texture =
+            gen6_blorp_emit_surface_state(brw, params, &params->src,
+                                          I915_GEM_DOMAIN_SAMPLER, 0);
+      }
+      wm_bind_bo_offset =
+         gen6_blorp_emit_binding_table(brw,
+                                       wm_surf_offset_renderbuffer,
+                                       wm_surf_offset_texture);
+   }
+
+   if (params->src.mt) {
+      const uint32_t sampler_offset =
+         gen6_blorp_emit_sampler_state(brw, BRW_MAPFILTER_LINEAR, 0, true);
+      gen6_blorp_emit_sampler_state_pointers(brw, sampler_offset);
+   }
+   gen6_blorp_emit_vs_disable(brw, params);
+   gen6_blorp_emit_gs_disable(brw, params);
+   gen6_blorp_emit_clip_disable(brw);
+   gen6_blorp_emit_sf_config(brw, params);
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG)
+      gen6_blorp_emit_constant_ps(brw, params, wm_push_const_offset);
+   else
+      gen6_blorp_emit_constant_ps_disable(brw, params);
+   gen6_blorp_emit_wm_config(brw, params);
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG)
+      gen6_blorp_emit_binding_table_pointers(brw, wm_bind_bo_offset);
+   gen6_blorp_emit_viewport_state(brw, params);
+
+   if (params->depth.mt)
+      gen6_blorp_emit_depth_stencil_config(brw, params);
+   else
+      gen6_blorp_emit_depth_disable(brw, params);
+   gen6_blorp_emit_clear_params(brw, params);
+   gen6_blorp_emit_drawing_rectangle(brw, params);
+   gen6_blorp_emit_primitive(brw, params);
+}
+
diff --git a/src/mesa/drivers/dri/i965/gen6_blorp.cpp b/src/mesa/drivers/dri/i965/gen6_blorp.cpp
deleted file mode 100644
index dabe8b5..0000000
--- a/src/mesa/drivers/dri/i965/gen6_blorp.cpp
+++ /dev/null
@@ -1,1097 +0,0 @@
-/*
- * Copyright © 2011 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- */
-
-#include <assert.h>
-
-#include "intel_batchbuffer.h"
-#include "intel_mipmap_tree.h"
-
-#include "brw_context.h"
-#include "brw_state.h"
-
-#include "brw_blorp.h"
-
-/**
- * CMD_STATE_BASE_ADDRESS
- *
- * From the Sandy Bridge PRM, Volume 1, Part 1, Table STATE_BASE_ADDRESS:
- *     The following commands must be reissued following any change to the
- *     base addresses:
- *         3DSTATE_CC_POINTERS
- *         3DSTATE_BINDING_TABLE_POINTERS
- *         3DSTATE_SAMPLER_STATE_POINTERS
- *         3DSTATE_VIEWPORT_STATE_POINTERS
- *         MEDIA_STATE_POINTERS
- */
-void
-gen6_blorp_emit_state_base_address(struct brw_context *brw,
-                                   const struct brw_blorp_params *params)
-{
-   uint8_t mocs = brw->gen == 7 ? GEN7_MOCS_L3 : 0;
-
-   BEGIN_BATCH(10);
-   OUT_BATCH(CMD_STATE_BASE_ADDRESS << 16 | (10 - 2));
-   OUT_BATCH(mocs << 8 | /* GeneralStateMemoryObjectControlState */
-             mocs << 4 | /* StatelessDataPortAccessMemoryObjectControlState */
-             1); /* GeneralStateBaseAddressModifyEnable */
-
-   /* SurfaceStateBaseAddress */
-   OUT_RELOC(brw->batch.bo, I915_GEM_DOMAIN_SAMPLER, 0, 1);
-   /* DynamicStateBaseAddress */
-   OUT_RELOC(brw->batch.bo, (I915_GEM_DOMAIN_RENDER |
-                               I915_GEM_DOMAIN_INSTRUCTION), 0, 1);
-   OUT_BATCH(1); /* IndirectObjectBaseAddress */
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
-      OUT_RELOC(brw->cache.bo, I915_GEM_DOMAIN_INSTRUCTION, 0,
-                1); /* Instruction base address: shader kernels */
-   } else {
-      OUT_BATCH(1); /* InstructionBaseAddress */
-   }
-   OUT_BATCH(1); /* GeneralStateUpperBound */
-   /* Dynamic state upper bound.  Although the documentation says that
-    * programming it to zero will cause it to be ignored, that is a lie.
-    * If this isn't programmed to a real bound, the sampler border color
-    * pointer is rejected, causing border color to mysteriously fail.
-    */
-   OUT_BATCH(0xfffff001);
-   OUT_BATCH(1); /* IndirectObjectUpperBound*/
-   OUT_BATCH(1); /* InstructionAccessUpperBound */
-   ADVANCE_BATCH();
-}
-
-static void
-gen6_blorp_emit_vertex_buffer_state(struct brw_context *brw,
-                                    unsigned num_elems,
-                                    unsigned vbo_size,
-                                    uint32_t vertex_offset)
-{
-   /* 3DSTATE_VERTEX_BUFFERS */
-   const int num_buffers = 1;
-   const int batch_length = 1 + 4 * num_buffers;
-
-   uint32_t dw0 = GEN6_VB0_ACCESS_VERTEXDATA |
-                  (num_elems * sizeof(float)) << BRW_VB0_PITCH_SHIFT;
-
-   if (brw->gen >= 7)
-      dw0 |= GEN7_VB0_ADDRESS_MODIFYENABLE;
-
-   switch (brw->gen) {
-   case 7:
-      dw0 |= GEN7_MOCS_L3 << 16;
-      break;
-   case 8:
-      dw0 |= BDW_MOCS_WB << 16;
-      break;
-   case 9:
-      dw0 |= SKL_MOCS_WB << 16;
-      break;
-   }
-
-   BEGIN_BATCH(batch_length);
-   OUT_BATCH((_3DSTATE_VERTEX_BUFFERS << 16) | (batch_length - 2));
-   OUT_BATCH(dw0);
-   if (brw->gen >= 8) {
-      OUT_RELOC64(brw->batch.bo, I915_GEM_DOMAIN_VERTEX, 0, vertex_offset);
-      OUT_BATCH(vbo_size);
-   } else {
-      /* start address */
-      OUT_RELOC(brw->batch.bo, I915_GEM_DOMAIN_VERTEX, 0,
-                vertex_offset);
-      /* end address */
-      OUT_RELOC(brw->batch.bo, I915_GEM_DOMAIN_VERTEX, 0,
-                vertex_offset + vbo_size - 1);
-      OUT_BATCH(0);
-   }
-   ADVANCE_BATCH();
-}
-
-void
-gen6_blorp_emit_vertices(struct brw_context *brw,
-                         const struct brw_blorp_params *params)
-{
-   uint32_t vertex_offset;
-
-   /* Setup VBO for the rectangle primitive..
-    *
-    * A rectangle primitive (3DPRIM_RECTLIST) consists of only three
-    * vertices. The vertices reside in screen space with DirectX coordinates
-    * (that is, (0, 0) is the upper left corner).
-    *
-    *   v2 ------ implied
-    *    |        |
-    *    |        |
-    *   v0 ----- v1
-    *
-    * Since the VS is disabled, the clipper loads each VUE directly from
-    * the URB. This is controlled by the 3DSTATE_VERTEX_BUFFERS and
-    * 3DSTATE_VERTEX_ELEMENTS packets below. The VUE contents are as follows:
-    *   dw0: Reserved, MBZ.
-    *   dw1: Render Target Array Index. The HiZ op does not use indexed
-    *        vertices, so set the dword to 0.
-    *   dw2: Viewport Index. The HiZ op disables viewport mapping and
-    *        scissoring, so set the dword to 0.
-    *   dw3: Point Width: The HiZ op does not emit the POINTLIST primitive, so
-    *        set the dword to 0.
-    *   dw4: Vertex Position X.
-    *   dw5: Vertex Position Y.
-    *   dw6: Vertex Position Z.
-    *   dw7: Vertex Position W.
-    *
-    * For details, see the Sandybridge PRM, Volume 2, Part 1, Section 1.5.1
-    * "Vertex URB Entry (VUE) Formats".
-    *
-    * Only vertex position X and Y are going to be variable, Z is fixed to
-    * zero and W to one. Header words dw0-3 are all zero. There is no need to
-    * include the fixed values in the vertex buffer. Vertex fetcher can be
-    * instructed to fill vertex elements with constant values of one and zero
-    * instead of reading them from the buffer. See the vertex element setup
-    * below.
-    */
-   {
-      float *vertex_data;
-
-      const float vertices[] = {
-         /* v0 */ (float)params->x0, (float)params->y1,
-         /* v1 */ (float)params->x1, (float)params->y1,
-         /* v2 */ (float)params->x0, (float)params->y0,
-      };
-
-      vertex_data = (float *) brw_state_batch(brw, AUB_TRACE_VERTEX_BUFFER,
-                                              sizeof(vertices), 32,
-                                              &vertex_offset);
-      memcpy(vertex_data, vertices, sizeof(vertices));
-
-      const unsigned blorp_num_vue_elems = 2;
-      gen6_blorp_emit_vertex_buffer_state(brw, blorp_num_vue_elems,
-                                          sizeof(vertices), vertex_offset);
-   }
-
-   /* 3DSTATE_VERTEX_ELEMENTS
-    *
-    * Fetch dwords 0 - 7 from each VUE. See the comments above where
-    * the vertex_bo is filled with data.
-    */
-   {
-      const int num_elements = 2;
-      const int batch_length = 1 + 2 * num_elements;
-
-      BEGIN_BATCH(batch_length);
-      OUT_BATCH((_3DSTATE_VERTEX_ELEMENTS << 16) | (batch_length - 2));
-      /* Element 0 */
-      OUT_BATCH(GEN6_VE0_VALID |
-                BRW_SURFACEFORMAT_R32G32B32A32_FLOAT << BRW_VE0_FORMAT_SHIFT |
-                0 << BRW_VE0_SRC_OFFSET_SHIFT);
-      OUT_BATCH(BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_0_SHIFT |
-                BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_1_SHIFT |
-                BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT |
-                BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_3_SHIFT);
-      /* Element 1 */
-      OUT_BATCH(GEN6_VE0_VALID |
-                BRW_SURFACEFORMAT_R32G32_FLOAT << BRW_VE0_FORMAT_SHIFT |
-                0 << BRW_VE0_SRC_OFFSET_SHIFT);
-      OUT_BATCH(BRW_VE1_COMPONENT_STORE_SRC << BRW_VE1_COMPONENT_0_SHIFT |
-                BRW_VE1_COMPONENT_STORE_SRC << BRW_VE1_COMPONENT_1_SHIFT |
-                BRW_VE1_COMPONENT_STORE_0 << BRW_VE1_COMPONENT_2_SHIFT |
-                BRW_VE1_COMPONENT_STORE_1_FLT << BRW_VE1_COMPONENT_3_SHIFT);
-      ADVANCE_BATCH();
-   }
-}
-
-
-/* 3DSTATE_URB
- *
- * Assign the entire URB to the VS. Even though the VS disabled, URB space
- * is still needed because the clipper loads the VUE's from the URB. From
- * the Sandybridge PRM, Volume 2, Part 1, Section 3DSTATE,
- * Dword 1.15:0 "VS Number of URB Entries":
- *     This field is always used (even if VS Function Enable is DISABLED).
- *
- * The warning below appears in the PRM (Section 3DSTATE_URB), but we can
- * safely ignore it because this batch contains only one draw call.
- *     Because of URB corruption caused by allocating a previous GS unit
- *     URB entry to the VS unit, software is required to send a “GS NULL
- *     Fence” (Send URB fence with VS URB size == 1 and GS URB size == 0)
- *     plus a dummy DRAW call before any case where VS will be taking over
- *     GS URB space.
- */
-static void
-gen6_blorp_emit_urb_config(struct brw_context *brw,
-                           const struct brw_blorp_params *params)
-{
-   BEGIN_BATCH(3);
-   OUT_BATCH(_3DSTATE_URB << 16 | (3 - 2));
-   OUT_BATCH(brw->urb.max_vs_entries << GEN6_URB_VS_ENTRIES_SHIFT);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/* BLEND_STATE */
-uint32_t
-gen6_blorp_emit_blend_state(struct brw_context *brw,
-                            const struct brw_blorp_params *params)
-{
-   uint32_t cc_blend_state_offset;
-
-   assume(params->num_draw_buffers);
-
-   const unsigned size = params->num_draw_buffers *
-                         sizeof(struct gen6_blend_state);
-   struct gen6_blend_state *blend = (struct gen6_blend_state *)
-      brw_state_batch(brw, AUB_TRACE_BLEND_STATE, size, 64,
-                      &cc_blend_state_offset);
-
-   memset(blend, 0, size);
-
-   for (unsigned i = 0; i < params->num_draw_buffers; ++i) {
-      blend[i].blend1.pre_blend_clamp_enable = 1;
-      blend[i].blend1.post_blend_clamp_enable = 1;
-      blend[i].blend1.clamp_range = BRW_RENDERTARGET_CLAMPRANGE_FORMAT;
-
-      blend[i].blend1.write_disable_r = params->color_write_disable[0];
-      blend[i].blend1.write_disable_g = params->color_write_disable[1];
-      blend[i].blend1.write_disable_b = params->color_write_disable[2];
-      blend[i].blend1.write_disable_a = params->color_write_disable[3];
-   }
-
-   return cc_blend_state_offset;
-}
-
-
-/* CC_STATE */
-uint32_t
-gen6_blorp_emit_cc_state(struct brw_context *brw)
-{
-   uint32_t cc_state_offset;
-
-   struct gen6_color_calc_state *cc = (struct gen6_color_calc_state *)
-      brw_state_batch(brw, AUB_TRACE_CC_STATE,
-                      sizeof(gen6_color_calc_state), 64,
-                      &cc_state_offset);
-   memset(cc, 0, sizeof(*cc));
-
-   return cc_state_offset;
-}
-
-
-/**
- * \param out_offset is relative to
- *        CMD_STATE_BASE_ADDRESS.DynamicStateBaseAddress.
- */
-uint32_t
-gen6_blorp_emit_depth_stencil_state(struct brw_context *brw,
-                                    const struct brw_blorp_params *params)
-{
-   uint32_t depthstencil_offset;
-
-   struct gen6_depth_stencil_state *state;
-   state = (struct gen6_depth_stencil_state *)
-      brw_state_batch(brw, AUB_TRACE_DEPTH_STENCIL_STATE,
-                      sizeof(*state), 64,
-                      &depthstencil_offset);
-   memset(state, 0, sizeof(*state));
-
-   /* See the following sections of the Sandy Bridge PRM, Volume 1, Part2:
-    *   - 7.5.3.1 Depth Buffer Clear
-    *   - 7.5.3.2 Depth Buffer Resolve
-    *   - 7.5.3.3 Hierarchical Depth Buffer Resolve
-    */
-   state->ds2.depth_write_enable = 1;
-   if (params->hiz_op == GEN6_HIZ_OP_DEPTH_RESOLVE) {
-      state->ds2.depth_test_enable = 1;
-      state->ds2.depth_test_func = BRW_COMPAREFUNCTION_NEVER;
-   }
-
-   return depthstencil_offset;
-}
-
-
-/* 3DSTATE_CC_STATE_POINTERS
- *
- * The pointer offsets are relative to
- * CMD_STATE_BASE_ADDRESS.DynamicStateBaseAddress.
- *
- * The HiZ op doesn't use BLEND_STATE or COLOR_CALC_STATE.
- */
-static void
-gen6_blorp_emit_cc_state_pointers(struct brw_context *brw,
-                                  const struct brw_blorp_params *params,
-                                  uint32_t cc_blend_state_offset,
-                                  uint32_t depthstencil_offset,
-                                  uint32_t cc_state_offset)
-{
-   BEGIN_BATCH(4);
-   OUT_BATCH(_3DSTATE_CC_STATE_POINTERS << 16 | (4 - 2));
-   OUT_BATCH(cc_blend_state_offset | 1); /* BLEND_STATE offset */
-   OUT_BATCH(depthstencil_offset | 1); /* DEPTH_STENCIL_STATE offset */
-   OUT_BATCH(cc_state_offset | 1); /* COLOR_CALC_STATE offset */
-   ADVANCE_BATCH();
-}
-
-
-/* WM push constants */
-uint32_t
-gen6_blorp_emit_wm_constants(struct brw_context *brw,
-                             const struct brw_blorp_params *params)
-{
-   uint32_t wm_push_const_offset;
-
-   void *constants = brw_state_batch(brw, AUB_TRACE_WM_CONSTANTS,
-                                     sizeof(params->wm_push_consts),
-                                     32, &wm_push_const_offset);
-   memcpy(constants, &params->wm_push_consts,
-          sizeof(params->wm_push_consts));
-
-   return wm_push_const_offset;
-}
-
-
-/* SURFACE_STATE for renderbuffer or texture surface (see
- * brw_update_renderbuffer_surface and brw_update_texture_surface)
- */
-static uint32_t
-gen6_blorp_emit_surface_state(struct brw_context *brw,
-                              const struct brw_blorp_params *params,
-                              const struct brw_blorp_surface_info *surface,
-                              uint32_t read_domains, uint32_t write_domain)
-{
-   uint32_t wm_surf_offset;
-   uint32_t width = surface->width;
-   uint32_t height = surface->height;
-   if (surface->num_samples > 1) {
-      /* Since gen6 uses INTEL_MSAA_LAYOUT_IMS, width and height are measured
-       * in samples.  But SURFACE_STATE wants them in pixels, so we need to
-       * divide them each by 2.
-       */
-      width /= 2;
-      height /= 2;
-   }
-   struct intel_mipmap_tree *mt = surface->mt;
-   uint32_t tile_x, tile_y;
-
-   uint32_t *surf = (uint32_t *)
-      brw_state_batch(brw, AUB_TRACE_SURFACE_STATE, 6 * 4, 32,
-                      &wm_surf_offset);
-
-   surf[0] = (BRW_SURFACE_2D << BRW_SURFACE_TYPE_SHIFT |
-              BRW_SURFACE_MIPMAPLAYOUT_BELOW << BRW_SURFACE_MIPLAYOUT_SHIFT |
-              BRW_SURFACE_CUBEFACE_ENABLES |
-              surface->brw_surfaceformat << BRW_SURFACE_FORMAT_SHIFT);
-
-   /* reloc */
-   surf[1] = (brw_blorp_compute_tile_offsets(surface, &tile_x, &tile_y) +
-              mt->bo->offset64);
-
-   surf[2] = (0 << BRW_SURFACE_LOD_SHIFT |
-              (width - 1) << BRW_SURFACE_WIDTH_SHIFT |
-              (height - 1) << BRW_SURFACE_HEIGHT_SHIFT);
-
-   uint32_t tiling = surface->map_stencil_as_y_tiled
-      ? BRW_SURFACE_TILED | BRW_SURFACE_TILED_Y
-      : brw_get_surface_tiling_bits(mt->tiling);
-   uint32_t pitch_bytes = mt->pitch;
-   if (surface->map_stencil_as_y_tiled)
-      pitch_bytes *= 2;
-   surf[3] = (tiling |
-              0 << BRW_SURFACE_DEPTH_SHIFT |
-              (pitch_bytes - 1) << BRW_SURFACE_PITCH_SHIFT);
-
-   surf[4] = brw_get_surface_num_multisamples(surface->num_samples);
-
-   /* Note that the low bits of these fields are missing, so
-    * there's the possibility of getting in trouble.
-    */
-   assert(tile_x % 4 == 0);
-   assert(tile_y % 2 == 0);
-   surf[5] = ((tile_x / 4) << BRW_SURFACE_X_OFFSET_SHIFT |
-              (tile_y / 2) << BRW_SURFACE_Y_OFFSET_SHIFT |
-              (surface->mt->valign == 4 ?
-               BRW_SURFACE_VERTICAL_ALIGN_ENABLE : 0));
-
-   /* Emit relocation to surface contents */
-   drm_intel_bo_emit_reloc(brw->batch.bo,
-                           wm_surf_offset + 4,
-                           mt->bo,
-                           surf[1] - mt->bo->offset64,
-                           read_domains, write_domain);
-
-   return wm_surf_offset;
-}
-
-
-/* BINDING_TABLE.  See brw_wm_binding_table(). */
-uint32_t
-gen6_blorp_emit_binding_table(struct brw_context *brw,
-                              uint32_t wm_surf_offset_renderbuffer,
-                              uint32_t wm_surf_offset_texture)
-{
-   uint32_t wm_bind_bo_offset;
-   uint32_t *bind = (uint32_t *)
-      brw_state_batch(brw, AUB_TRACE_BINDING_TABLE,
-                      sizeof(uint32_t) *
-                      BRW_BLORP_NUM_BINDING_TABLE_ENTRIES,
-                      32, /* alignment */
-                      &wm_bind_bo_offset);
-   bind[BRW_BLORP_RENDERBUFFER_BINDING_TABLE_INDEX] =
-      wm_surf_offset_renderbuffer;
-   bind[BRW_BLORP_TEXTURE_BINDING_TABLE_INDEX] = wm_surf_offset_texture;
-
-   return wm_bind_bo_offset;
-}
-
-
-/**
- * SAMPLER_STATE.  See brw_update_sampler_state().
- */
-uint32_t
-gen6_blorp_emit_sampler_state(struct brw_context *brw,
-                              unsigned tex_filter, unsigned max_lod,
-                              bool non_normalized_coords)
-{
-   uint32_t sampler_offset;
-   uint32_t *sampler_state = (uint32_t *)
-      brw_state_batch(brw, AUB_TRACE_SAMPLER_STATE, 16, 32, &sampler_offset);
-
-   unsigned address_rounding = BRW_ADDRESS_ROUNDING_ENABLE_U_MIN |
-                               BRW_ADDRESS_ROUNDING_ENABLE_V_MIN |
-                               BRW_ADDRESS_ROUNDING_ENABLE_R_MIN |
-                               BRW_ADDRESS_ROUNDING_ENABLE_U_MAG |
-                               BRW_ADDRESS_ROUNDING_ENABLE_V_MAG |
-                               BRW_ADDRESS_ROUNDING_ENABLE_R_MAG;
-
-   /* XXX: I don't think that using firstLevel, lastLevel works,
-    * because we always setup the surface state as if firstLevel ==
-    * level zero.  Probably have to subtract firstLevel from each of
-    * these:
-    */
-   brw_emit_sampler_state(brw,
-                          sampler_state,
-                          sampler_offset,
-                          tex_filter, /* min filter */
-                          tex_filter, /* mag filter */
-                          BRW_MIPFILTER_NONE,
-                          BRW_ANISORATIO_2,
-                          address_rounding,
-                          BRW_TEXCOORDMODE_CLAMP,
-                          BRW_TEXCOORDMODE_CLAMP,
-                          BRW_TEXCOORDMODE_CLAMP,
-                          0, /* min LOD */
-                          max_lod,
-                          0, /* LOD bias */
-                          0, /* shadow function */
-                          non_normalized_coords,
-                          0); /* border color offset - unused */
-
-   return sampler_offset;
-}
-
-
-/**
- * 3DSTATE_SAMPLER_STATE_POINTERS.  See upload_sampler_state_pointers().
- */
-static void
-gen6_blorp_emit_sampler_state_pointers(struct brw_context *brw,
-                                       uint32_t sampler_offset)
-{
-   BEGIN_BATCH(4);
-   OUT_BATCH(_3DSTATE_SAMPLER_STATE_POINTERS << 16 |
-             VS_SAMPLER_STATE_CHANGE |
-             GS_SAMPLER_STATE_CHANGE |
-             PS_SAMPLER_STATE_CHANGE |
-             (4 - 2));
-   OUT_BATCH(0); /* VS */
-   OUT_BATCH(0); /* GS */
-   OUT_BATCH(sampler_offset);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_VS
- *
- * Disable vertex shader.
- */
-void
-gen6_blorp_emit_vs_disable(struct brw_context *brw,
-                           const struct brw_blorp_params *params)
-{
-   /* From the BSpec, 3D Pipeline > Geometry > Vertex Shader > State,
-    * 3DSTATE_VS, Dword 5.0 "VS Function Enable":
-    *
-    *   [DevSNB] A pipeline flush must be programmed prior to a
-    *   3DSTATE_VS command that causes the VS Function Enable to
-    *   toggle. Pipeline flush can be executed by sending a PIPE_CONTROL
-    *   command with CS stall bit set and a post sync operation.
-    *
-    * We've already done one at the start of the BLORP operation.
-    */
-
-   /* Disable the push constant buffers. */
-   BEGIN_BATCH(5);
-   OUT_BATCH(_3DSTATE_CONSTANT_VS << 16 | (5 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(6);
-   OUT_BATCH(_3DSTATE_VS << 16 | (6 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_GS
- *
- * Disable the geometry shader.
- */
-void
-gen6_blorp_emit_gs_disable(struct brw_context *brw,
-                           const struct brw_blorp_params *params)
-{
-   /* Disable all the constant buffers. */
-   BEGIN_BATCH(5);
-   OUT_BATCH(_3DSTATE_CONSTANT_GS << 16 | (5 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_GS << 16 | (7 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-   brw->gs.enabled = false;
-}
-
-
-/* 3DSTATE_CLIP
- *
- * Disable the clipper.
- *
- * The BLORP op emits a rectangle primitive, which requires clipping to
- * be disabled. From page 10 of the Sandy Bridge PRM Volume 2 Part 1
- * Section 1.3 "3D Primitives Overview":
- *    RECTLIST:
- *    Either the CLIP unit should be DISABLED, or the CLIP unit's Clip
- *    Mode should be set to a value other than CLIPMODE_NORMAL.
- *
- * Also disable perspective divide. This doesn't change the clipper's
- * output, but does spare a few electrons.
- */
-void
-gen6_blorp_emit_clip_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(4);
-   OUT_BATCH(_3DSTATE_CLIP << 16 | (4 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(GEN6_CLIP_PERSPECTIVE_DIVIDE_DISABLE);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_SF
- *
- * Disable ViewportTransformEnable (dw2.1)
- *
- * From the SandyBridge PRM, Volume 2, Part 1, Section 1.3, "3D
- * Primitives Overview":
- *     RECTLIST: Viewport Mapping must be DISABLED (as is typical with the
- *     use of screen- space coordinates).
- *
- * A solid rectangle must be rendered, so set FrontFaceFillMode (dw2.4:3)
- * and BackFaceFillMode (dw2.5:6) to SOLID(0).
- *
- * From the Sandy Bridge PRM, Volume 2, Part 1, Section
- * 6.4.1.1 3DSTATE_SF, Field FrontFaceFillMode:
- *     SOLID: Any triangle or rectangle object found to be front-facing
- *     is rendered as a solid object. This setting is required when
- *     (rendering rectangle (RECTLIST) objects.
- */
-static void
-gen6_blorp_emit_sf_config(struct brw_context *brw,
-                          const struct brw_blorp_params *params)
-{
-   BEGIN_BATCH(20);
-   OUT_BATCH(_3DSTATE_SF << 16 | (20 - 2));
-   OUT_BATCH(params->num_varyings << GEN6_SF_NUM_OUTPUTS_SHIFT |
-             1 << GEN6_SF_URB_ENTRY_READ_LENGTH_SHIFT |
-             BRW_SF_URB_ENTRY_READ_OFFSET <<
-                GEN6_SF_URB_ENTRY_READ_OFFSET_SHIFT);
-   OUT_BATCH(0); /* dw2 */
-   OUT_BATCH(params->dst.num_samples > 1 ? GEN6_SF_MSRAST_ON_PATTERN : 0);
-   for (int i = 0; i < 16; ++i)
-      OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/**
- * Enable or disable thread dispatch and set the HiZ op appropriately.
- */
-static void
-gen6_blorp_emit_wm_config(struct brw_context *brw,
-                          const struct brw_blorp_params *params)
-{
-   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
-   uint32_t dw2, dw4, dw5, dw6;
-
-   /* Even when thread dispatch is disabled, max threads (dw5.25:31) must be
-    * nonzero to prevent the GPU from hanging.  While the documentation doesn't
-    * mention this explicitly, it notes that the valid range for the field is
-    * [1,39] = [2,40] threads, which excludes zero.
-    *
-    * To be safe (and to minimize extraneous code) we go ahead and fully
-    * configure the WM state whether or not there is a WM program.
-    */
-
-   dw2 = dw4 = dw5 = dw6 = 0;
-   switch (params->hiz_op) {
-   case GEN6_HIZ_OP_DEPTH_CLEAR:
-      dw4 |= GEN6_WM_DEPTH_CLEAR;
-      break;
-   case GEN6_HIZ_OP_DEPTH_RESOLVE:
-      dw4 |= GEN6_WM_DEPTH_RESOLVE;
-      break;
-   case GEN6_HIZ_OP_HIZ_RESOLVE:
-      dw4 |= GEN6_WM_HIERARCHICAL_DEPTH_RESOLVE;
-      break;
-   case GEN6_HIZ_OP_NONE:
-      break;
-   default:
-      unreachable("not reached");
-   }
-   dw5 |= GEN6_WM_LINE_AA_WIDTH_1_0;
-   dw5 |= GEN6_WM_LINE_END_CAP_AA_WIDTH_0_5;
-   dw5 |= (brw->max_wm_threads - 1) << GEN6_WM_MAX_THREADS_SHIFT;
-   dw6 |= 0 << GEN6_WM_BARYCENTRIC_INTERPOLATION_MODE_SHIFT; /* No interp */
-   dw6 |= 0 << GEN6_WM_NUM_SF_OUTPUTS_SHIFT; /* No inputs from SF */
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
-      dw4 |= prog_data->first_curbe_grf << GEN6_WM_DISPATCH_START_GRF_SHIFT_0;
-      dw5 |= GEN6_WM_16_DISPATCH_ENABLE;
-      dw5 |= GEN6_WM_DISPATCH_ENABLE; /* We are rendering */
-   }
-
-   if (params->src.mt) {
-      dw5 |= GEN6_WM_KILL_ENABLE; /* TODO: temporarily smash on */
-      dw2 |= 1 << GEN6_WM_SAMPLER_COUNT_SHIFT; /* Up to 4 samplers */
-   }
-
-   if (params->dst.num_samples > 1) {
-      dw6 |= GEN6_WM_MSRAST_ON_PATTERN;
-      if (prog_data && prog_data->persample_msaa_dispatch)
-         dw6 |= GEN6_WM_MSDISPMODE_PERSAMPLE;
-      else
-         dw6 |= GEN6_WM_MSDISPMODE_PERPIXEL;
-   } else {
-      dw6 |= GEN6_WM_MSRAST_OFF_PIXEL;
-      dw6 |= GEN6_WM_MSDISPMODE_PERSAMPLE;
-   }
-
-   BEGIN_BATCH(9);
-   OUT_BATCH(_3DSTATE_WM << 16 | (9 - 2));
-   OUT_BATCH(params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG ?
-             params->wm_prog_kernel : 0);
-   OUT_BATCH(dw2);
-   OUT_BATCH(0); /* No scratch needed */
-   OUT_BATCH(dw4);
-   OUT_BATCH(dw5);
-   OUT_BATCH(dw6);
-   OUT_BATCH(0); /* No other programs */
-   OUT_BATCH(0); /* No other programs */
-   ADVANCE_BATCH();
-}
-
-
-static void
-gen6_blorp_emit_constant_ps(struct brw_context *brw,
-                            const struct brw_blorp_params *params,
-                            uint32_t wm_push_const_offset)
-{
-   /* Make sure the push constants fill an exact integer number of
-    * registers.
-    */
-   assert(sizeof(struct brw_blorp_wm_push_constants) % 32 == 0);
-
-   /* There must be at least one register worth of push constant data. */
-   assert(BRW_BLORP_NUM_PUSH_CONST_REGS > 0);
-
-   /* Enable push constant buffer 0. */
-   BEGIN_BATCH(5);
-   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 |
-             GEN6_CONSTANT_BUFFER_0_ENABLE |
-             (5 - 2));
-   OUT_BATCH(wm_push_const_offset + (BRW_BLORP_NUM_PUSH_CONST_REGS - 1));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-static void
-gen6_blorp_emit_constant_ps_disable(struct brw_context *brw,
-                                    const struct brw_blorp_params *params)
-{
-   /* Disable the push constant buffers. */
-   BEGIN_BATCH(5);
-   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 | (5 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-/**
- * 3DSTATE_BINDING_TABLE_POINTERS
- */
-static void
-gen6_blorp_emit_binding_table_pointers(struct brw_context *brw,
-                                       uint32_t wm_bind_bo_offset)
-{
-   BEGIN_BATCH(4);
-   OUT_BATCH(_3DSTATE_BINDING_TABLE_POINTERS << 16 |
-             GEN6_BINDING_TABLE_MODIFY_PS |
-             (4 - 2));
-   OUT_BATCH(0); /* vs -- ignored */
-   OUT_BATCH(0); /* gs -- ignored */
-   OUT_BATCH(wm_bind_bo_offset); /* wm/ps */
-   ADVANCE_BATCH();
-}
-
-
-static void
-gen6_blorp_emit_depth_stencil_config(struct brw_context *brw,
-                                     const struct brw_blorp_params *params)
-{
-   uint32_t surfwidth, surfheight;
-   uint32_t surftype;
-   unsigned int depth = MAX2(params->depth.mt->logical_depth0, 1);
-   GLenum gl_target = params->depth.mt->target;
-   unsigned int lod;
-
-   switch (gl_target) {
-   case GL_TEXTURE_CUBE_MAP_ARRAY:
-   case GL_TEXTURE_CUBE_MAP:
-      /* The PRM claims that we should use BRW_SURFACE_CUBE for this
-       * situation, but experiments show that gl_Layer doesn't work when we do
-       * this.  So we use BRW_SURFACE_2D, since for rendering purposes this is
-       * equivalent.
-       */
-      surftype = BRW_SURFACE_2D;
-      depth *= 6;
-      break;
-   default:
-      surftype = translate_tex_target(gl_target);
-      break;
-   }
-
-   const unsigned min_array_element = params->depth.layer;
-
-   lod = params->depth.level - params->depth.mt->first_level;
-
-   if (params->hiz_op != GEN6_HIZ_OP_NONE && lod == 0) {
-      /* HIZ ops for lod 0 may set the width & height a little
-       * larger to allow the fast depth clear to fit the hardware
-       * alignment requirements. (8x4)
-       */
-      surfwidth = params->depth.width;
-      surfheight = params->depth.height;
-   } else {
-      surfwidth = params->depth.mt->logical_width0;
-      surfheight = params->depth.mt->logical_height0;
-   }
-
-   /* 3DSTATE_DEPTH_BUFFER */
-   {
-      brw_emit_depth_stall_flushes(brw);
-
-      BEGIN_BATCH(7);
-      /* 3DSTATE_DEPTH_BUFFER dw0 */
-      OUT_BATCH(_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
-
-      /* 3DSTATE_DEPTH_BUFFER dw1 */
-      OUT_BATCH((params->depth.mt->pitch - 1) |
-                params->depth_format << 18 |
-                1 << 21 | /* separate stencil enable */
-                1 << 22 | /* hiz enable */
-                BRW_TILEWALK_YMAJOR << 26 |
-                1 << 27 | /* y-tiled */
-                surftype << 29);
-
-      /* 3DSTATE_DEPTH_BUFFER dw2 */
-      OUT_RELOC(params->depth.mt->bo,
-                I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
-                0);
-
-      /* 3DSTATE_DEPTH_BUFFER dw3 */
-      OUT_BATCH(BRW_SURFACE_MIPMAPLAYOUT_BELOW << 1 |
-                (surfwidth - 1) << 6 |
-                (surfheight - 1) << 19 |
-                lod << 2);
-
-      /* 3DSTATE_DEPTH_BUFFER dw4 */
-      OUT_BATCH((depth - 1) << 21 |
-                min_array_element << 10 |
-                (depth - 1) << 1);
-
-      /* 3DSTATE_DEPTH_BUFFER dw5 */
-      OUT_BATCH(0);
-
-      /* 3DSTATE_DEPTH_BUFFER dw6 */
-      OUT_BATCH(0);
-      ADVANCE_BATCH();
-   }
-
-   /* 3DSTATE_HIER_DEPTH_BUFFER */
-   {
-      struct intel_mipmap_tree *hiz_mt = params->depth.mt->hiz_buf->mt;
-      uint32_t offset = 0;
-
-      if (hiz_mt->array_layout == ALL_SLICES_AT_EACH_LOD) {
-         offset = intel_miptree_get_aligned_offset(hiz_mt,
-                                                   hiz_mt->level[lod].level_x,
-                                                   hiz_mt->level[lod].level_y,
-                                                   false);
-      }
-
-      BEGIN_BATCH(3);
-      OUT_BATCH((_3DSTATE_HIER_DEPTH_BUFFER << 16) | (3 - 2));
-      OUT_BATCH(hiz_mt->pitch - 1);
-      OUT_RELOC(hiz_mt->bo,
-                I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
-                offset);
-      ADVANCE_BATCH();
-   }
-
-   /* 3DSTATE_STENCIL_BUFFER */
-   {
-      BEGIN_BATCH(3);
-      OUT_BATCH((_3DSTATE_STENCIL_BUFFER << 16) | (3 - 2));
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      ADVANCE_BATCH();
-   }
-}
-
-
-static void
-gen6_blorp_emit_depth_disable(struct brw_context *brw,
-                              const struct brw_blorp_params *params)
-{
-   brw_emit_depth_stall_flushes(brw);
-
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
-   OUT_BATCH((BRW_DEPTHFORMAT_D32_FLOAT << 18) |
-             (BRW_SURFACE_NULL << 29));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(3);
-   OUT_BATCH(_3DSTATE_HIER_DEPTH_BUFFER << 16 | (3 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(3);
-   OUT_BATCH(_3DSTATE_STENCIL_BUFFER << 16 | (3 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_CLEAR_PARAMS
- *
- * From the Sandybridge PRM, Volume 2, Part 1, Section 3DSTATE_CLEAR_PARAMS:
- *   [DevSNB] 3DSTATE_CLEAR_PARAMS packet must follow the DEPTH_BUFFER_STATE
- *   packet when HiZ is enabled and the DEPTH_BUFFER_STATE changes.
- */
-static void
-gen6_blorp_emit_clear_params(struct brw_context *brw,
-                             const struct brw_blorp_params *params)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_CLEAR_PARAMS << 16 |
-	     GEN5_DEPTH_CLEAR_VALID |
-	     (2 - 2));
-   OUT_BATCH(params->depth.mt ? params->depth.mt->depth_clear_value : 0);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_DRAWING_RECTANGLE */
-void
-gen6_blorp_emit_drawing_rectangle(struct brw_context *brw,
-                                  const struct brw_blorp_params *params)
-{
-   BEGIN_BATCH(4);
-   OUT_BATCH(_3DSTATE_DRAWING_RECTANGLE << 16 | (4 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(((MAX2(params->x1, params->x0) - 1) & 0xffff) |
-             ((MAX2(params->y1, params->y0) - 1) << 16));
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-/* 3DSTATE_VIEWPORT_STATE_POINTERS */
-static void
-gen6_blorp_emit_viewport_state(struct brw_context *brw,
-			       const struct brw_blorp_params *params)
-{
-   struct brw_cc_viewport *ccv;
-   uint32_t cc_vp_offset;
-
-   ccv = (struct brw_cc_viewport *)brw_state_batch(brw, AUB_TRACE_CC_VP_STATE,
-						   sizeof(*ccv), 32,
-						   &cc_vp_offset);
-
-   ccv->min_depth = 0.0;
-   ccv->max_depth = 1.0;
-
-   BEGIN_BATCH(4);
-   OUT_BATCH(_3DSTATE_VIEWPORT_STATE_POINTERS << 16 | (4 - 2) |
-	     GEN6_CC_VIEWPORT_MODIFY);
-   OUT_BATCH(0); /* clip VP */
-   OUT_BATCH(0); /* SF VP */
-   OUT_BATCH(cc_vp_offset);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DPRIMITIVE */
-static void
-gen6_blorp_emit_primitive(struct brw_context *brw,
-                          const struct brw_blorp_params *params)
-{
-   BEGIN_BATCH(6);
-   OUT_BATCH(CMD_3D_PRIM << 16 | (6 - 2) |
-             _3DPRIM_RECTLIST << GEN4_3DPRIM_TOPOLOGY_TYPE_SHIFT |
-             GEN4_3DPRIM_VERTEXBUFFER_ACCESS_SEQUENTIAL);
-   OUT_BATCH(3); /* vertex count per instance */
-   OUT_BATCH(0);
-   OUT_BATCH(params->num_layers); /* instance count */
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-/**
- * \brief Execute a blit or render pass operation.
- *
- * To execute the operation, this function manually constructs and emits a
- * batch to draw a rectangle primitive. The batchbuffer is flushed before
- * constructing and after emitting the batch.
- *
- * This function alters no GL state.
- */
-void
-gen6_blorp_exec(struct brw_context *brw,
-                const struct brw_blorp_params *params)
-{
-   uint32_t cc_blend_state_offset = 0;
-   uint32_t cc_state_offset = 0;
-   uint32_t depthstencil_offset;
-   uint32_t wm_push_const_offset = 0;
-   uint32_t wm_bind_bo_offset = 0;
-
-   /* Emit workaround flushes when we switch from drawing to blorping. */
-   brw_emit_post_sync_nonzero_flush(brw);
-
-   gen6_emit_3dstate_multisample(brw, params->dst.num_samples);
-   gen6_emit_3dstate_sample_mask(brw,
-                                 params->dst.num_samples > 1 ?
-                                 (1 << params->dst.num_samples) - 1 : 1);
-   gen6_blorp_emit_state_base_address(brw, params);
-   gen6_blorp_emit_vertices(brw, params);
-   gen6_blorp_emit_urb_config(brw, params);
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
-      cc_blend_state_offset = gen6_blorp_emit_blend_state(brw, params);
-      cc_state_offset = gen6_blorp_emit_cc_state(brw);
-   }
-   depthstencil_offset = gen6_blorp_emit_depth_stencil_state(brw, params);
-   gen6_blorp_emit_cc_state_pointers(brw, params, cc_blend_state_offset,
-                                     depthstencil_offset, cc_state_offset);
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
-      uint32_t wm_surf_offset_renderbuffer;
-      uint32_t wm_surf_offset_texture = 0;
-      wm_push_const_offset = gen6_blorp_emit_wm_constants(brw, params);
-      intel_miptree_used_for_rendering(params->dst.mt);
-      wm_surf_offset_renderbuffer =
-         gen6_blorp_emit_surface_state(brw, params, &params->dst,
-                                       I915_GEM_DOMAIN_RENDER,
-                                       I915_GEM_DOMAIN_RENDER);
-      if (params->src.mt) {
-         wm_surf_offset_texture =
-            gen6_blorp_emit_surface_state(brw, params, &params->src,
-                                          I915_GEM_DOMAIN_SAMPLER, 0);
-      }
-      wm_bind_bo_offset =
-         gen6_blorp_emit_binding_table(brw,
-                                       wm_surf_offset_renderbuffer,
-                                       wm_surf_offset_texture);
-   }
-
-   if (params->src.mt) {
-      const uint32_t sampler_offset =
-         gen6_blorp_emit_sampler_state(brw, BRW_MAPFILTER_LINEAR, 0, true);
-      gen6_blorp_emit_sampler_state_pointers(brw, sampler_offset);
-   }
-   gen6_blorp_emit_vs_disable(brw, params);
-   gen6_blorp_emit_gs_disable(brw, params);
-   gen6_blorp_emit_clip_disable(brw);
-   gen6_blorp_emit_sf_config(brw, params);
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG)
-      gen6_blorp_emit_constant_ps(brw, params, wm_push_const_offset);
-   else
-      gen6_blorp_emit_constant_ps_disable(brw, params);
-   gen6_blorp_emit_wm_config(brw, params);
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG)
-      gen6_blorp_emit_binding_table_pointers(brw, wm_bind_bo_offset);
-   gen6_blorp_emit_viewport_state(brw, params);
-
-   if (params->depth.mt)
-      gen6_blorp_emit_depth_stencil_config(brw, params);
-   else
-      gen6_blorp_emit_depth_disable(brw, params);
-   gen6_blorp_emit_clear_params(brw, params);
-   gen6_blorp_emit_drawing_rectangle(brw, params);
-   gen6_blorp_emit_primitive(brw, params);
-}
-
diff --git a/src/mesa/drivers/dri/i965/gen7_blorp.c b/src/mesa/drivers/dri/i965/gen7_blorp.c
new file mode 100644
index 0000000..0dcdcf6
--- /dev/null
+++ b/src/mesa/drivers/dri/i965/gen7_blorp.c
@@ -0,0 +1,884 @@
+/*
+ * Copyright © 2011 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include <assert.h>
+
+#include "intel_batchbuffer.h"
+#include "intel_mipmap_tree.h"
+
+#include "brw_context.h"
+#include "brw_state.h"
+
+#include "brw_blorp.h"
+
+static bool
+gen7_blorp_skip_urb_config(const struct brw_context *brw)
+{
+   if (brw->ctx.NewDriverState & (BRW_NEW_CONTEXT | BRW_NEW_URB_SIZE))
+      return false;
+
+   /* Vertex buffer takes 24 bytes. As the size is expressed in 64 bytes,
+    * one will suffice, otherwise the setup can be any valid configuration.
+    */
+   return brw->urb.vsize > 0;
+}
+
+/* 3DSTATE_URB_VS
+ * 3DSTATE_URB_HS
+ * 3DSTATE_URB_DS
+ * 3DSTATE_URB_GS
+ *
+ * If the 3DSTATE_URB_VS is emitted, than the others must be also.
+ * From the Ivybridge PRM, Volume 2 Part 1, section 1.7.1 3DSTATE_URB_VS:
+ *
+ *     3DSTATE_URB_HS, 3DSTATE_URB_DS, and 3DSTATE_URB_GS must also be
+ *     programmed in order for the programming of this state to be
+ *     valid.
+ */
+void
+gen7_blorp_emit_urb_config(struct brw_context *brw)
+{
+   /* URB allocations must be done in 8k chunks. */
+   const unsigned chunk_size_bytes = 8192;
+   const unsigned urb_size =
+      (brw->gen >= 8 || (brw->is_haswell && brw->gt == 3)) ? 32 : 16;
+   const unsigned push_constant_bytes = 1024 * urb_size;
+   const unsigned push_constant_chunks =
+      push_constant_bytes / chunk_size_bytes;
+   const unsigned vs_size = 1;
+   const unsigned vs_start = push_constant_chunks;
+   const unsigned vs_chunks =
+      DIV_ROUND_UP(brw->urb.min_vs_entries * vs_size * 64, chunk_size_bytes);
+
+   if (gen7_blorp_skip_urb_config(brw))
+      return;
+
+   brw->ctx.NewDriverState |= BRW_NEW_URB_SIZE;
+
+   gen7_emit_push_constant_state(brw,
+                                 urb_size / 2 /* vs_size */,
+                                 0 /* hs_size */,
+                                 0 /* ds_size */,
+                                 0 /* gs_size */,
+                                 urb_size / 2 /* fs_size */);
+
+   gen7_emit_urb_state(brw,
+                       brw->urb.min_vs_entries /* num_vs_entries */,
+                       vs_size,
+                       vs_start,
+                       0 /* num_hs_entries */,
+                       1 /* hs_size */,
+                       vs_start + vs_chunks /* hs_start */,
+                       0 /* num_ds_entries */,
+                       1 /* ds_size */,
+                       vs_start + vs_chunks /* ds_start */,
+                       0 /* num_gs_entries */,
+                       1 /* gs_size */,
+                       vs_start + vs_chunks /* gs_start */);
+}
+
+
+/* 3DSTATE_BLEND_STATE_POINTERS */
+void
+gen7_blorp_emit_blend_state_pointer(struct brw_context *brw,
+                                    uint32_t cc_blend_state_offset)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_BLEND_STATE_POINTERS << 16 | (2 - 2));
+   OUT_BATCH(cc_blend_state_offset | 1);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_CC_STATE_POINTERS */
+void
+gen7_blorp_emit_cc_state_pointer(struct brw_context *brw,
+                                 uint32_t cc_state_offset)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_CC_STATE_POINTERS << 16 | (2 - 2));
+   OUT_BATCH(cc_state_offset | 1);
+   ADVANCE_BATCH();
+}
+
+void
+gen7_blorp_emit_cc_viewport(struct brw_context *brw)
+{
+   struct brw_cc_viewport *ccv;
+   uint32_t cc_vp_offset;
+
+   ccv = (struct brw_cc_viewport *)brw_state_batch(brw, AUB_TRACE_CC_VP_STATE,
+						   sizeof(*ccv), 32,
+						   &cc_vp_offset);
+   ccv->min_depth = 0.0;
+   ccv->max_depth = 1.0;
+
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_VIEWPORT_STATE_POINTERS_CC << 16 | (2 - 2));
+   OUT_BATCH(cc_vp_offset);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_DEPTH_STENCIL_STATE_POINTERS
+ *
+ * The offset is relative to CMD_STATE_BASE_ADDRESS.DynamicStateBaseAddress.
+ */
+static void
+gen7_blorp_emit_depth_stencil_state_pointers(struct brw_context *brw,
+                                             uint32_t depthstencil_offset)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_DEPTH_STENCIL_STATE_POINTERS << 16 | (2 - 2));
+   OUT_BATCH(depthstencil_offset | 1);
+   ADVANCE_BATCH();
+}
+
+
+/* SURFACE_STATE for renderbuffer or texture surface (see
+ * brw_update_renderbuffer_surface and brw_update_texture_surface)
+ */
+static uint32_t
+gen7_blorp_emit_surface_state(struct brw_context *brw,
+                              const struct brw_blorp_surface_info *surface,
+                              uint32_t read_domains, uint32_t write_domain,
+                              bool is_render_target)
+{
+   uint32_t wm_surf_offset;
+   uint32_t width = surface->width;
+   uint32_t height = surface->height;
+   /* Note: since gen7 uses INTEL_MSAA_LAYOUT_CMS or INTEL_MSAA_LAYOUT_UMS for
+    * color surfaces, width and height are measured in pixels; we don't need
+    * to divide them by 2 as we do for Gen6 (see
+    * gen6_blorp_emit_surface_state).
+    */
+   struct intel_mipmap_tree *mt = surface->mt;
+   uint32_t tile_x, tile_y;
+   const uint8_t mocs = GEN7_MOCS_L3;
+
+   uint32_t tiling = surface->map_stencil_as_y_tiled
+      ? I915_TILING_Y : mt->tiling;
+
+   uint32_t *surf = (uint32_t *)
+      brw_state_batch(brw, AUB_TRACE_SURFACE_STATE, 8 * 4, 32, &wm_surf_offset);
+   memset(surf, 0, 8 * 4);
+
+   surf[0] = BRW_SURFACE_2D << BRW_SURFACE_TYPE_SHIFT |
+             surface->brw_surfaceformat << BRW_SURFACE_FORMAT_SHIFT |
+             gen7_surface_tiling_mode(tiling);
+
+   if (surface->mt->valign == 4)
+      surf[0] |= GEN7_SURFACE_VALIGN_4;
+   if (surface->mt->halign == 8)
+      surf[0] |= GEN7_SURFACE_HALIGN_8;
+
+   if (surface->array_layout == ALL_SLICES_AT_EACH_LOD)
+      surf[0] |= GEN7_SURFACE_ARYSPC_LOD0;
+   else
+      surf[0] |= GEN7_SURFACE_ARYSPC_FULL;
+
+   /* reloc */
+   surf[1] = brw_blorp_compute_tile_offsets(surface, &tile_x, &tile_y) +
+             mt->bo->offset64;
+
+   /* Note that the low bits of these fields are missing, so
+    * there's the possibility of getting in trouble.
+    */
+   assert(tile_x % 4 == 0);
+   assert(tile_y % 2 == 0);
+   surf[5] = SET_FIELD(tile_x / 4, BRW_SURFACE_X_OFFSET) |
+             SET_FIELD(tile_y / 2, BRW_SURFACE_Y_OFFSET) |
+             SET_FIELD(mocs, GEN7_SURFACE_MOCS);
+
+   surf[2] = SET_FIELD(width - 1, GEN7_SURFACE_WIDTH) |
+             SET_FIELD(height - 1, GEN7_SURFACE_HEIGHT);
+
+   uint32_t pitch_bytes = mt->pitch;
+   if (surface->map_stencil_as_y_tiled)
+      pitch_bytes *= 2;
+   surf[3] = pitch_bytes - 1;
+
+   surf[4] = gen7_surface_msaa_bits(surface->num_samples, surface->msaa_layout);
+   if (surface->mt->mcs_mt) {
+      gen7_set_surface_mcs_info(brw, surf, wm_surf_offset, surface->mt->mcs_mt,
+                                is_render_target);
+   }
+
+   surf[7] = surface->mt->fast_clear_color_value;
+
+   if (brw->is_haswell) {
+      surf[7] |= (SET_FIELD(HSW_SCS_RED,   GEN7_SURFACE_SCS_R) |
+                  SET_FIELD(HSW_SCS_GREEN, GEN7_SURFACE_SCS_G) |
+                  SET_FIELD(HSW_SCS_BLUE,  GEN7_SURFACE_SCS_B) |
+                  SET_FIELD(HSW_SCS_ALPHA, GEN7_SURFACE_SCS_A));
+   }
+
+   /* Emit relocation to surface contents */
+   drm_intel_bo_emit_reloc(brw->batch.bo,
+                           wm_surf_offset + 4,
+                           mt->bo,
+                           surf[1] - mt->bo->offset64,
+                           read_domains, write_domain);
+
+   gen7_check_surface_setup(surf, is_render_target);
+
+   return wm_surf_offset;
+}
+
+
+/* 3DSTATE_VS
+ *
+ * Disable vertex shader.
+ */
+static void
+gen7_blorp_emit_vs_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_CONSTANT_VS << 16 | (7 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(6);
+   OUT_BATCH(_3DSTATE_VS << 16 | (6 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_HS
+ *
+ * Disable the hull shader.
+ */
+static void
+gen7_blorp_emit_hs_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_CONSTANT_HS << 16 | (7 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_HS << 16 | (7 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_TE
+ *
+ * Disable the tesselation engine.
+ */
+void
+gen7_blorp_emit_te_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(4);
+   OUT_BATCH(_3DSTATE_TE << 16 | (4 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_DS
+ *
+ * Disable the domain shader.
+ */
+static void
+gen7_blorp_emit_ds_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_CONSTANT_DS << 16 | (7 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(6);
+   OUT_BATCH(_3DSTATE_DS << 16 | (6 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+/* 3DSTATE_GS
+ *
+ * Disable the geometry shader.
+ */
+static void
+gen7_blorp_emit_gs_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_CONSTANT_GS << 16 | (7 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   /**
+    * From Graphics BSpec: 3D-Media-GPGPU Engine > 3D Pipeline Stages >
+    * Geometry > Geometry Shader > State:
+    *
+    *     "Note: Because of corruption in IVB:GT2, software needs to flush the
+    *     whole fixed function pipeline when the GS enable changes value in
+    *     the 3DSTATE_GS."
+    *
+    * The hardware architects have clarified that in this context "flush the
+    * whole fixed function pipeline" means to emit a PIPE_CONTROL with the "CS
+    * Stall" bit set.
+    */
+   if (brw->gen < 8 && !brw->is_haswell && brw->gt == 2 && brw->gs.enabled)
+      gen7_emit_cs_stall_flush(brw);
+
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_GS << 16 | (7 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+   brw->gs.enabled = false;
+}
+
+/* 3DSTATE_STREAMOUT
+ *
+ * Disable streamout.
+ */
+static void
+gen7_blorp_emit_streamout_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(3);
+   OUT_BATCH(_3DSTATE_STREAMOUT << 16 | (3 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+static void
+gen7_blorp_emit_sf_config(struct brw_context *brw,
+                          const struct brw_blorp_params *params)
+{
+   /* 3DSTATE_SF
+    *
+    * Disable ViewportTransformEnable (dw1.1)
+    *
+    * From the SandyBridge PRM, Volume 2, Part 1, Section 1.3, "3D
+    * Primitives Overview":
+    *     RECTLIST: Viewport Mapping must be DISABLED (as is typical with the
+    *     use of screen- space coordinates).
+    *
+    * A solid rectangle must be rendered, so set FrontFaceFillMode (dw1.6:5)
+    * and BackFaceFillMode (dw1.4:3) to SOLID(0).
+    *
+    * From the Sandy Bridge PRM, Volume 2, Part 1, Section
+    * 6.4.1.1 3DSTATE_SF, Field FrontFaceFillMode:
+    *     SOLID: Any triangle or rectangle object found to be front-facing
+    *     is rendered as a solid object. This setting is required when
+    *     (rendering rectangle (RECTLIST) objects.
+    */
+   {
+      BEGIN_BATCH(7);
+      OUT_BATCH(_3DSTATE_SF << 16 | (7 - 2));
+      OUT_BATCH(params->depth_format <<
+                GEN7_SF_DEPTH_BUFFER_SURFACE_FORMAT_SHIFT);
+      OUT_BATCH(params->dst.num_samples > 1 ? GEN6_SF_MSRAST_ON_PATTERN : 0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      ADVANCE_BATCH();
+   }
+
+   /* 3DSTATE_SBE */
+   {
+      BEGIN_BATCH(14);
+      OUT_BATCH(_3DSTATE_SBE << 16 | (14 - 2));
+      OUT_BATCH(GEN7_SBE_SWIZZLE_ENABLE |
+                params->num_varyings << GEN7_SBE_NUM_OUTPUTS_SHIFT |
+                1 << GEN7_SBE_URB_ENTRY_READ_LENGTH_SHIFT |
+                BRW_SF_URB_ENTRY_READ_OFFSET <<
+                   GEN7_SBE_URB_ENTRY_READ_OFFSET_SHIFT);
+      for (int i = 0; i < 12; ++i)
+         OUT_BATCH(0);
+      ADVANCE_BATCH();
+   }
+}
+
+
+/**
+ * Disable thread dispatch (dw5.19) and enable the HiZ op.
+ */
+static void
+gen7_blorp_emit_wm_config(struct brw_context *brw,
+                          const struct brw_blorp_params *params)
+{
+   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
+   uint32_t dw1 = 0, dw2 = 0;
+
+   switch (params->hiz_op) {
+   case GEN6_HIZ_OP_DEPTH_CLEAR:
+      dw1 |= GEN7_WM_DEPTH_CLEAR;
+      break;
+   case GEN6_HIZ_OP_DEPTH_RESOLVE:
+      dw1 |= GEN7_WM_DEPTH_RESOLVE;
+      break;
+   case GEN6_HIZ_OP_HIZ_RESOLVE:
+      dw1 |= GEN7_WM_HIERARCHICAL_DEPTH_RESOLVE;
+      break;
+   case GEN6_HIZ_OP_NONE:
+      break;
+   default:
+      unreachable("not reached");
+   }
+   dw1 |= GEN7_WM_LINE_AA_WIDTH_1_0;
+   dw1 |= GEN7_WM_LINE_END_CAP_AA_WIDTH_0_5;
+   dw1 |= 0 << GEN7_WM_BARYCENTRIC_INTERPOLATION_MODE_SHIFT; /* No interp */
+
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG)
+      dw1 |= GEN7_WM_DISPATCH_ENABLE; /* We are rendering */
+
+   if (params->src.mt)
+      dw1 |= GEN7_WM_KILL_ENABLE; /* TODO: temporarily smash on */
+
+      if (params->dst.num_samples > 1) {
+         dw1 |= GEN7_WM_MSRAST_ON_PATTERN;
+         if (prog_data && prog_data->persample_msaa_dispatch)
+            dw2 |= GEN7_WM_MSDISPMODE_PERSAMPLE;
+         else
+            dw2 |= GEN7_WM_MSDISPMODE_PERPIXEL;
+      } else {
+         dw1 |= GEN7_WM_MSRAST_OFF_PIXEL;
+         dw2 |= GEN7_WM_MSDISPMODE_PERSAMPLE;
+      }
+
+   BEGIN_BATCH(3);
+   OUT_BATCH(_3DSTATE_WM << 16 | (3 - 2));
+   OUT_BATCH(dw1);
+   OUT_BATCH(dw2);
+   ADVANCE_BATCH();
+}
+
+
+/**
+ * 3DSTATE_PS
+ *
+ * Pixel shader dispatch is disabled above in 3DSTATE_WM, dw1.29. Despite
+ * that, thread dispatch info must still be specified.
+ *     - Maximum Number of Threads (dw4.24:31) must be nonzero, as the
+ *       valid range for this field is [0x3, 0x2f].
+ *     - A dispatch mode must be given; that is, at least one of the
+ *       "N Pixel Dispatch Enable" (N=8,16,32) fields must be set. This was
+ *       discovered through simulator error messages.
+ */
+static void
+gen7_blorp_emit_ps_config(struct brw_context *brw,
+                          const struct brw_blorp_params *params)
+{
+   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
+   uint32_t dw2, dw4, dw5;
+   const int max_threads_shift = brw->is_haswell ?
+      HSW_PS_MAX_THREADS_SHIFT : IVB_PS_MAX_THREADS_SHIFT;
+
+   dw2 = dw4 = dw5 = 0;
+   dw4 |= (brw->max_wm_threads - 1) << max_threads_shift;
+
+   /* If there's a WM program, we need to do 16-pixel dispatch since that's
+    * what the program is compiled for.  If there isn't, then it shouldn't
+    * matter because no program is actually being run.  However, the hardware
+    * gets angry if we don't enable at least one dispatch mode, so just enable
+    * 16-pixel dispatch unconditionally.
+    */
+   dw4 |= GEN7_PS_16_DISPATCH_ENABLE;
+
+   if (brw->is_haswell)
+      dw4 |= SET_FIELD(1, HSW_PS_SAMPLE_MASK); /* 1 sample for now */
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
+      dw4 |= GEN7_PS_PUSH_CONSTANT_ENABLE;
+      dw5 |= prog_data->first_curbe_grf << GEN7_PS_DISPATCH_START_GRF_SHIFT_0;
+   }
+
+   if (params->src.mt)
+      dw2 |= 1 << GEN7_PS_SAMPLER_COUNT_SHIFT; /* Up to 4 samplers */
+
+   dw4 |= params->fast_clear_op;
+
+   BEGIN_BATCH(8);
+   OUT_BATCH(_3DSTATE_PS << 16 | (8 - 2));
+   OUT_BATCH(params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG ?
+             params->wm_prog_kernel : 0);
+   OUT_BATCH(dw2);
+   OUT_BATCH(0);
+   OUT_BATCH(dw4);
+   OUT_BATCH(dw5);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+void
+gen7_blorp_emit_binding_table_pointers_ps(struct brw_context *brw,
+                                          uint32_t wm_bind_bo_offset)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_BINDING_TABLE_POINTERS_PS << 16 | (2 - 2));
+   OUT_BATCH(wm_bind_bo_offset);
+   ADVANCE_BATCH();
+}
+
+
+void
+gen7_blorp_emit_sampler_state_pointers_ps(struct brw_context *brw,
+                                          uint32_t sampler_offset)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_SAMPLER_STATE_POINTERS_PS << 16 | (2 - 2));
+   OUT_BATCH(sampler_offset);
+   ADVANCE_BATCH();
+}
+
+
+void
+gen7_blorp_emit_constant_ps(struct brw_context *brw,
+                            uint32_t wm_push_const_offset)
+{
+   const uint8_t mocs = GEN7_MOCS_L3;
+
+   /* Make sure the push constants fill an exact integer number of
+    * registers.
+    */
+   assert(sizeof(struct brw_blorp_wm_push_constants) % 32 == 0);
+
+   /* There must be at least one register worth of push constant data. */
+   assert(BRW_BLORP_NUM_PUSH_CONST_REGS > 0);
+
+   /* Enable push constant buffer 0. */
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 |
+             (7 - 2));
+   OUT_BATCH(BRW_BLORP_NUM_PUSH_CONST_REGS);
+   OUT_BATCH(0);
+   OUT_BATCH(wm_push_const_offset | mocs);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+void
+gen7_blorp_emit_constant_ps_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(7);
+   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 | (7 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+static void
+gen7_blorp_emit_depth_stencil_config(struct brw_context *brw,
+                                     const struct brw_blorp_params *params)
+{
+   const uint8_t mocs = GEN7_MOCS_L3;
+   uint32_t surfwidth, surfheight;
+   uint32_t surftype;
+   unsigned int depth = MAX2(params->depth.mt->logical_depth0, 1);
+   unsigned int min_array_element;
+   GLenum gl_target = params->depth.mt->target;
+   unsigned int lod;
+
+   switch (gl_target) {
+   case GL_TEXTURE_CUBE_MAP_ARRAY:
+   case GL_TEXTURE_CUBE_MAP:
+      /* The PRM claims that we should use BRW_SURFACE_CUBE for this
+       * situation, but experiments show that gl_Layer doesn't work when we do
+       * this.  So we use BRW_SURFACE_2D, since for rendering purposes this is
+       * equivalent.
+       */
+      surftype = BRW_SURFACE_2D;
+      depth *= 6;
+      break;
+   default:
+      surftype = translate_tex_target(gl_target);
+      break;
+   }
+
+   min_array_element = params->depth.layer;
+   if (params->depth.mt->num_samples > 1) {
+      /* Convert physical layer to logical layer. */
+      min_array_element /= params->depth.mt->num_samples;
+   }
+
+   lod = params->depth.level - params->depth.mt->first_level;
+
+   if (params->hiz_op != GEN6_HIZ_OP_NONE && lod == 0) {
+      /* HIZ ops for lod 0 may set the width & height a little
+       * larger to allow the fast depth clear to fit the hardware
+       * alignment requirements. (8x4)
+       */
+      surfwidth = params->depth.width;
+      surfheight = params->depth.height;
+   } else {
+      surfwidth = params->depth.mt->logical_width0;
+      surfheight = params->depth.mt->logical_height0;
+   }
+
+   /* 3DSTATE_DEPTH_BUFFER */
+   {
+      brw_emit_depth_stall_flushes(brw);
+
+      BEGIN_BATCH(7);
+      OUT_BATCH(GEN7_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
+      OUT_BATCH((params->depth.mt->pitch - 1) |
+                params->depth_format << 18 |
+                1 << 22 | /* hiz enable */
+                1 << 28 | /* depth write */
+                surftype << 29);
+      OUT_RELOC(params->depth.mt->bo,
+                I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
+                0);
+      OUT_BATCH((surfwidth - 1) << 4 |
+                (surfheight - 1) << 18 |
+                lod);
+      OUT_BATCH(((depth - 1) << 21) |
+                (min_array_element << 10) |
+                mocs);
+      OUT_BATCH(0);
+      OUT_BATCH((depth - 1) << 21);
+      ADVANCE_BATCH();
+   }
+
+   /* 3DSTATE_HIER_DEPTH_BUFFER */
+   {
+      struct intel_miptree_aux_buffer *hiz_buf = params->depth.mt->hiz_buf;
+
+      BEGIN_BATCH(3);
+      OUT_BATCH((GEN7_3DSTATE_HIER_DEPTH_BUFFER << 16) | (3 - 2));
+      OUT_BATCH((mocs << 25) |
+                (hiz_buf->pitch - 1));
+      OUT_RELOC(hiz_buf->bo,
+                I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
+                0);
+      ADVANCE_BATCH();
+   }
+
+   /* 3DSTATE_STENCIL_BUFFER */
+   {
+      BEGIN_BATCH(3);
+      OUT_BATCH((GEN7_3DSTATE_STENCIL_BUFFER << 16) | (3 - 2));
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      ADVANCE_BATCH();
+   }
+}
+
+
+static void
+gen7_blorp_emit_depth_disable(struct brw_context *brw)
+{
+   brw_emit_depth_stall_flushes(brw);
+
+   BEGIN_BATCH(7);
+   OUT_BATCH(GEN7_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
+   OUT_BATCH(BRW_DEPTHFORMAT_D32_FLOAT << 18 | (BRW_SURFACE_NULL << 29));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(3);
+   OUT_BATCH(GEN7_3DSTATE_HIER_DEPTH_BUFFER << 16 | (3 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(3);
+   OUT_BATCH(GEN7_3DSTATE_STENCIL_BUFFER << 16 | (3 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DSTATE_CLEAR_PARAMS
+ *
+ * From the Ivybridge PRM, Volume 2 Part 1, Section 11.5.5.4
+ * 3DSTATE_CLEAR_PARAMS:
+ *    3DSTATE_CLEAR_PARAMS must always be programmed in the along
+ *    with the other Depth/Stencil state commands(i.e.  3DSTATE_DEPTH_BUFFER,
+ *    3DSTATE_STENCIL_BUFFER, or 3DSTATE_HIER_DEPTH_BUFFER).
+ */
+void
+gen7_blorp_emit_clear_params(struct brw_context *brw,
+                             const struct brw_blorp_params *params)
+{
+   BEGIN_BATCH(3);
+   OUT_BATCH(GEN7_3DSTATE_CLEAR_PARAMS << 16 | (3 - 2));
+   OUT_BATCH(params->depth.mt ? params->depth.mt->depth_clear_value : 0);
+   OUT_BATCH(GEN7_DEPTH_CLEAR_VALID);
+   ADVANCE_BATCH();
+}
+
+
+/* 3DPRIMITIVE */
+void
+gen7_blorp_emit_primitive(struct brw_context *brw,
+                          const struct brw_blorp_params *params)
+{
+   BEGIN_BATCH(7);
+   OUT_BATCH(CMD_3D_PRIM << 16 | (7 - 2));
+   OUT_BATCH(GEN7_3DPRIM_VERTEXBUFFER_ACCESS_SEQUENTIAL |
+             _3DPRIM_RECTLIST);
+   OUT_BATCH(3); /* vertex count per instance */
+   OUT_BATCH(0);
+   OUT_BATCH(params->num_layers); /* instance count */
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+
+/**
+ * \copydoc gen6_blorp_exec()
+ */
+void
+gen7_blorp_exec(struct brw_context *brw,
+                const struct brw_blorp_params *params)
+{
+   if (brw->gen >= 8)
+      return;
+
+   uint32_t cc_blend_state_offset = 0;
+   uint32_t cc_state_offset = 0;
+   uint32_t depthstencil_offset;
+   uint32_t wm_push_const_offset = 0;
+   uint32_t wm_bind_bo_offset = 0;
+
+   gen6_emit_3dstate_multisample(brw, params->dst.num_samples);
+   gen6_emit_3dstate_sample_mask(brw,
+                                 params->dst.num_samples > 1 ?
+                                 (1 << params->dst.num_samples) - 1 : 1);
+   gen6_blorp_emit_state_base_address(brw, params);
+   gen6_blorp_emit_vertices(brw, params);
+   gen7_blorp_emit_urb_config(brw);
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
+      cc_blend_state_offset = gen6_blorp_emit_blend_state(brw, params);
+      cc_state_offset = gen6_blorp_emit_cc_state(brw);
+      gen7_blorp_emit_blend_state_pointer(brw, cc_blend_state_offset);
+      gen7_blorp_emit_cc_state_pointer(brw, cc_state_offset);
+   }
+   depthstencil_offset = gen6_blorp_emit_depth_stencil_state(brw, params);
+   gen7_blorp_emit_depth_stencil_state_pointers(brw, depthstencil_offset);
+   if (brw->use_resource_streamer)
+      gen7_disable_hw_binding_tables(brw);
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
+      uint32_t wm_surf_offset_renderbuffer;
+      uint32_t wm_surf_offset_texture = 0;
+      wm_push_const_offset = gen6_blorp_emit_wm_constants(brw, params);
+      intel_miptree_used_for_rendering(params->dst.mt);
+      wm_surf_offset_renderbuffer =
+         gen7_blorp_emit_surface_state(brw, &params->dst,
+                                       I915_GEM_DOMAIN_RENDER,
+                                       I915_GEM_DOMAIN_RENDER,
+                                       true /* is_render_target */);
+      if (params->src.mt) {
+         wm_surf_offset_texture =
+            gen7_blorp_emit_surface_state(brw, &params->src,
+                                          I915_GEM_DOMAIN_SAMPLER, 0,
+                                          false /* is_render_target */);
+      }
+      wm_bind_bo_offset =
+         gen6_blorp_emit_binding_table(brw,
+                                       wm_surf_offset_renderbuffer,
+                                       wm_surf_offset_texture);
+   }
+   gen7_blorp_emit_vs_disable(brw);
+   gen7_blorp_emit_hs_disable(brw);
+   gen7_blorp_emit_te_disable(brw);
+   gen7_blorp_emit_ds_disable(brw);
+   gen7_blorp_emit_gs_disable(brw);
+   gen7_blorp_emit_streamout_disable(brw);
+   gen6_blorp_emit_clip_disable(brw);
+   gen7_blorp_emit_sf_config(brw, params);
+   gen7_blorp_emit_wm_config(brw, params);
+   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
+      gen7_blorp_emit_binding_table_pointers_ps(brw, wm_bind_bo_offset);
+      gen7_blorp_emit_constant_ps(brw, wm_push_const_offset);
+   } else {
+      gen7_blorp_emit_constant_ps_disable(brw);
+   }
+
+   if (params->src.mt) {
+      const uint32_t sampler_offset =
+         gen6_blorp_emit_sampler_state(brw, BRW_MAPFILTER_LINEAR, 0, true);
+      gen7_blorp_emit_sampler_state_pointers_ps(brw, sampler_offset);
+   }
+
+   gen7_blorp_emit_ps_config(brw, params);
+   gen7_blorp_emit_cc_viewport(brw);
+
+   if (params->depth.mt)
+      gen7_blorp_emit_depth_stencil_config(brw, params);
+   else
+      gen7_blorp_emit_depth_disable(brw);
+   gen7_blorp_emit_clear_params(brw, params);
+   gen6_blorp_emit_drawing_rectangle(brw, params);
+   gen7_blorp_emit_primitive(brw, params);
+}
diff --git a/src/mesa/drivers/dri/i965/gen7_blorp.cpp b/src/mesa/drivers/dri/i965/gen7_blorp.cpp
deleted file mode 100644
index 0dcdcf6..0000000
--- a/src/mesa/drivers/dri/i965/gen7_blorp.cpp
+++ /dev/null
@@ -1,884 +0,0 @@
-/*
- * Copyright © 2011 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- */
-
-#include <assert.h>
-
-#include "intel_batchbuffer.h"
-#include "intel_mipmap_tree.h"
-
-#include "brw_context.h"
-#include "brw_state.h"
-
-#include "brw_blorp.h"
-
-static bool
-gen7_blorp_skip_urb_config(const struct brw_context *brw)
-{
-   if (brw->ctx.NewDriverState & (BRW_NEW_CONTEXT | BRW_NEW_URB_SIZE))
-      return false;
-
-   /* Vertex buffer takes 24 bytes. As the size is expressed in 64 bytes,
-    * one will suffice, otherwise the setup can be any valid configuration.
-    */
-   return brw->urb.vsize > 0;
-}
-
-/* 3DSTATE_URB_VS
- * 3DSTATE_URB_HS
- * 3DSTATE_URB_DS
- * 3DSTATE_URB_GS
- *
- * If the 3DSTATE_URB_VS is emitted, than the others must be also.
- * From the Ivybridge PRM, Volume 2 Part 1, section 1.7.1 3DSTATE_URB_VS:
- *
- *     3DSTATE_URB_HS, 3DSTATE_URB_DS, and 3DSTATE_URB_GS must also be
- *     programmed in order for the programming of this state to be
- *     valid.
- */
-void
-gen7_blorp_emit_urb_config(struct brw_context *brw)
-{
-   /* URB allocations must be done in 8k chunks. */
-   const unsigned chunk_size_bytes = 8192;
-   const unsigned urb_size =
-      (brw->gen >= 8 || (brw->is_haswell && brw->gt == 3)) ? 32 : 16;
-   const unsigned push_constant_bytes = 1024 * urb_size;
-   const unsigned push_constant_chunks =
-      push_constant_bytes / chunk_size_bytes;
-   const unsigned vs_size = 1;
-   const unsigned vs_start = push_constant_chunks;
-   const unsigned vs_chunks =
-      DIV_ROUND_UP(brw->urb.min_vs_entries * vs_size * 64, chunk_size_bytes);
-
-   if (gen7_blorp_skip_urb_config(brw))
-      return;
-
-   brw->ctx.NewDriverState |= BRW_NEW_URB_SIZE;
-
-   gen7_emit_push_constant_state(brw,
-                                 urb_size / 2 /* vs_size */,
-                                 0 /* hs_size */,
-                                 0 /* ds_size */,
-                                 0 /* gs_size */,
-                                 urb_size / 2 /* fs_size */);
-
-   gen7_emit_urb_state(brw,
-                       brw->urb.min_vs_entries /* num_vs_entries */,
-                       vs_size,
-                       vs_start,
-                       0 /* num_hs_entries */,
-                       1 /* hs_size */,
-                       vs_start + vs_chunks /* hs_start */,
-                       0 /* num_ds_entries */,
-                       1 /* ds_size */,
-                       vs_start + vs_chunks /* ds_start */,
-                       0 /* num_gs_entries */,
-                       1 /* gs_size */,
-                       vs_start + vs_chunks /* gs_start */);
-}
-
-
-/* 3DSTATE_BLEND_STATE_POINTERS */
-void
-gen7_blorp_emit_blend_state_pointer(struct brw_context *brw,
-                                    uint32_t cc_blend_state_offset)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_BLEND_STATE_POINTERS << 16 | (2 - 2));
-   OUT_BATCH(cc_blend_state_offset | 1);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_CC_STATE_POINTERS */
-void
-gen7_blorp_emit_cc_state_pointer(struct brw_context *brw,
-                                 uint32_t cc_state_offset)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_CC_STATE_POINTERS << 16 | (2 - 2));
-   OUT_BATCH(cc_state_offset | 1);
-   ADVANCE_BATCH();
-}
-
-void
-gen7_blorp_emit_cc_viewport(struct brw_context *brw)
-{
-   struct brw_cc_viewport *ccv;
-   uint32_t cc_vp_offset;
-
-   ccv = (struct brw_cc_viewport *)brw_state_batch(brw, AUB_TRACE_CC_VP_STATE,
-						   sizeof(*ccv), 32,
-						   &cc_vp_offset);
-   ccv->min_depth = 0.0;
-   ccv->max_depth = 1.0;
-
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_VIEWPORT_STATE_POINTERS_CC << 16 | (2 - 2));
-   OUT_BATCH(cc_vp_offset);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_DEPTH_STENCIL_STATE_POINTERS
- *
- * The offset is relative to CMD_STATE_BASE_ADDRESS.DynamicStateBaseAddress.
- */
-static void
-gen7_blorp_emit_depth_stencil_state_pointers(struct brw_context *brw,
-                                             uint32_t depthstencil_offset)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_DEPTH_STENCIL_STATE_POINTERS << 16 | (2 - 2));
-   OUT_BATCH(depthstencil_offset | 1);
-   ADVANCE_BATCH();
-}
-
-
-/* SURFACE_STATE for renderbuffer or texture surface (see
- * brw_update_renderbuffer_surface and brw_update_texture_surface)
- */
-static uint32_t
-gen7_blorp_emit_surface_state(struct brw_context *brw,
-                              const struct brw_blorp_surface_info *surface,
-                              uint32_t read_domains, uint32_t write_domain,
-                              bool is_render_target)
-{
-   uint32_t wm_surf_offset;
-   uint32_t width = surface->width;
-   uint32_t height = surface->height;
-   /* Note: since gen7 uses INTEL_MSAA_LAYOUT_CMS or INTEL_MSAA_LAYOUT_UMS for
-    * color surfaces, width and height are measured in pixels; we don't need
-    * to divide them by 2 as we do for Gen6 (see
-    * gen6_blorp_emit_surface_state).
-    */
-   struct intel_mipmap_tree *mt = surface->mt;
-   uint32_t tile_x, tile_y;
-   const uint8_t mocs = GEN7_MOCS_L3;
-
-   uint32_t tiling = surface->map_stencil_as_y_tiled
-      ? I915_TILING_Y : mt->tiling;
-
-   uint32_t *surf = (uint32_t *)
-      brw_state_batch(brw, AUB_TRACE_SURFACE_STATE, 8 * 4, 32, &wm_surf_offset);
-   memset(surf, 0, 8 * 4);
-
-   surf[0] = BRW_SURFACE_2D << BRW_SURFACE_TYPE_SHIFT |
-             surface->brw_surfaceformat << BRW_SURFACE_FORMAT_SHIFT |
-             gen7_surface_tiling_mode(tiling);
-
-   if (surface->mt->valign == 4)
-      surf[0] |= GEN7_SURFACE_VALIGN_4;
-   if (surface->mt->halign == 8)
-      surf[0] |= GEN7_SURFACE_HALIGN_8;
-
-   if (surface->array_layout == ALL_SLICES_AT_EACH_LOD)
-      surf[0] |= GEN7_SURFACE_ARYSPC_LOD0;
-   else
-      surf[0] |= GEN7_SURFACE_ARYSPC_FULL;
-
-   /* reloc */
-   surf[1] = brw_blorp_compute_tile_offsets(surface, &tile_x, &tile_y) +
-             mt->bo->offset64;
-
-   /* Note that the low bits of these fields are missing, so
-    * there's the possibility of getting in trouble.
-    */
-   assert(tile_x % 4 == 0);
-   assert(tile_y % 2 == 0);
-   surf[5] = SET_FIELD(tile_x / 4, BRW_SURFACE_X_OFFSET) |
-             SET_FIELD(tile_y / 2, BRW_SURFACE_Y_OFFSET) |
-             SET_FIELD(mocs, GEN7_SURFACE_MOCS);
-
-   surf[2] = SET_FIELD(width - 1, GEN7_SURFACE_WIDTH) |
-             SET_FIELD(height - 1, GEN7_SURFACE_HEIGHT);
-
-   uint32_t pitch_bytes = mt->pitch;
-   if (surface->map_stencil_as_y_tiled)
-      pitch_bytes *= 2;
-   surf[3] = pitch_bytes - 1;
-
-   surf[4] = gen7_surface_msaa_bits(surface->num_samples, surface->msaa_layout);
-   if (surface->mt->mcs_mt) {
-      gen7_set_surface_mcs_info(brw, surf, wm_surf_offset, surface->mt->mcs_mt,
-                                is_render_target);
-   }
-
-   surf[7] = surface->mt->fast_clear_color_value;
-
-   if (brw->is_haswell) {
-      surf[7] |= (SET_FIELD(HSW_SCS_RED,   GEN7_SURFACE_SCS_R) |
-                  SET_FIELD(HSW_SCS_GREEN, GEN7_SURFACE_SCS_G) |
-                  SET_FIELD(HSW_SCS_BLUE,  GEN7_SURFACE_SCS_B) |
-                  SET_FIELD(HSW_SCS_ALPHA, GEN7_SURFACE_SCS_A));
-   }
-
-   /* Emit relocation to surface contents */
-   drm_intel_bo_emit_reloc(brw->batch.bo,
-                           wm_surf_offset + 4,
-                           mt->bo,
-                           surf[1] - mt->bo->offset64,
-                           read_domains, write_domain);
-
-   gen7_check_surface_setup(surf, is_render_target);
-
-   return wm_surf_offset;
-}
-
-
-/* 3DSTATE_VS
- *
- * Disable vertex shader.
- */
-static void
-gen7_blorp_emit_vs_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_CONSTANT_VS << 16 | (7 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(6);
-   OUT_BATCH(_3DSTATE_VS << 16 | (6 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_HS
- *
- * Disable the hull shader.
- */
-static void
-gen7_blorp_emit_hs_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_CONSTANT_HS << 16 | (7 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_HS << 16 | (7 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_TE
- *
- * Disable the tesselation engine.
- */
-void
-gen7_blorp_emit_te_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(4);
-   OUT_BATCH(_3DSTATE_TE << 16 | (4 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_DS
- *
- * Disable the domain shader.
- */
-static void
-gen7_blorp_emit_ds_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_CONSTANT_DS << 16 | (7 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(6);
-   OUT_BATCH(_3DSTATE_DS << 16 | (6 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-/* 3DSTATE_GS
- *
- * Disable the geometry shader.
- */
-static void
-gen7_blorp_emit_gs_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_CONSTANT_GS << 16 | (7 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   /**
-    * From Graphics BSpec: 3D-Media-GPGPU Engine > 3D Pipeline Stages >
-    * Geometry > Geometry Shader > State:
-    *
-    *     "Note: Because of corruption in IVB:GT2, software needs to flush the
-    *     whole fixed function pipeline when the GS enable changes value in
-    *     the 3DSTATE_GS."
-    *
-    * The hardware architects have clarified that in this context "flush the
-    * whole fixed function pipeline" means to emit a PIPE_CONTROL with the "CS
-    * Stall" bit set.
-    */
-   if (brw->gen < 8 && !brw->is_haswell && brw->gt == 2 && brw->gs.enabled)
-      gen7_emit_cs_stall_flush(brw);
-
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_GS << 16 | (7 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-   brw->gs.enabled = false;
-}
-
-/* 3DSTATE_STREAMOUT
- *
- * Disable streamout.
- */
-static void
-gen7_blorp_emit_streamout_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(3);
-   OUT_BATCH(_3DSTATE_STREAMOUT << 16 | (3 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-static void
-gen7_blorp_emit_sf_config(struct brw_context *brw,
-                          const struct brw_blorp_params *params)
-{
-   /* 3DSTATE_SF
-    *
-    * Disable ViewportTransformEnable (dw1.1)
-    *
-    * From the SandyBridge PRM, Volume 2, Part 1, Section 1.3, "3D
-    * Primitives Overview":
-    *     RECTLIST: Viewport Mapping must be DISABLED (as is typical with the
-    *     use of screen- space coordinates).
-    *
-    * A solid rectangle must be rendered, so set FrontFaceFillMode (dw1.6:5)
-    * and BackFaceFillMode (dw1.4:3) to SOLID(0).
-    *
-    * From the Sandy Bridge PRM, Volume 2, Part 1, Section
-    * 6.4.1.1 3DSTATE_SF, Field FrontFaceFillMode:
-    *     SOLID: Any triangle or rectangle object found to be front-facing
-    *     is rendered as a solid object. This setting is required when
-    *     (rendering rectangle (RECTLIST) objects.
-    */
-   {
-      BEGIN_BATCH(7);
-      OUT_BATCH(_3DSTATE_SF << 16 | (7 - 2));
-      OUT_BATCH(params->depth_format <<
-                GEN7_SF_DEPTH_BUFFER_SURFACE_FORMAT_SHIFT);
-      OUT_BATCH(params->dst.num_samples > 1 ? GEN6_SF_MSRAST_ON_PATTERN : 0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      ADVANCE_BATCH();
-   }
-
-   /* 3DSTATE_SBE */
-   {
-      BEGIN_BATCH(14);
-      OUT_BATCH(_3DSTATE_SBE << 16 | (14 - 2));
-      OUT_BATCH(GEN7_SBE_SWIZZLE_ENABLE |
-                params->num_varyings << GEN7_SBE_NUM_OUTPUTS_SHIFT |
-                1 << GEN7_SBE_URB_ENTRY_READ_LENGTH_SHIFT |
-                BRW_SF_URB_ENTRY_READ_OFFSET <<
-                   GEN7_SBE_URB_ENTRY_READ_OFFSET_SHIFT);
-      for (int i = 0; i < 12; ++i)
-         OUT_BATCH(0);
-      ADVANCE_BATCH();
-   }
-}
-
-
-/**
- * Disable thread dispatch (dw5.19) and enable the HiZ op.
- */
-static void
-gen7_blorp_emit_wm_config(struct brw_context *brw,
-                          const struct brw_blorp_params *params)
-{
-   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
-   uint32_t dw1 = 0, dw2 = 0;
-
-   switch (params->hiz_op) {
-   case GEN6_HIZ_OP_DEPTH_CLEAR:
-      dw1 |= GEN7_WM_DEPTH_CLEAR;
-      break;
-   case GEN6_HIZ_OP_DEPTH_RESOLVE:
-      dw1 |= GEN7_WM_DEPTH_RESOLVE;
-      break;
-   case GEN6_HIZ_OP_HIZ_RESOLVE:
-      dw1 |= GEN7_WM_HIERARCHICAL_DEPTH_RESOLVE;
-      break;
-   case GEN6_HIZ_OP_NONE:
-      break;
-   default:
-      unreachable("not reached");
-   }
-   dw1 |= GEN7_WM_LINE_AA_WIDTH_1_0;
-   dw1 |= GEN7_WM_LINE_END_CAP_AA_WIDTH_0_5;
-   dw1 |= 0 << GEN7_WM_BARYCENTRIC_INTERPOLATION_MODE_SHIFT; /* No interp */
-
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG)
-      dw1 |= GEN7_WM_DISPATCH_ENABLE; /* We are rendering */
-
-   if (params->src.mt)
-      dw1 |= GEN7_WM_KILL_ENABLE; /* TODO: temporarily smash on */
-
-      if (params->dst.num_samples > 1) {
-         dw1 |= GEN7_WM_MSRAST_ON_PATTERN;
-         if (prog_data && prog_data->persample_msaa_dispatch)
-            dw2 |= GEN7_WM_MSDISPMODE_PERSAMPLE;
-         else
-            dw2 |= GEN7_WM_MSDISPMODE_PERPIXEL;
-      } else {
-         dw1 |= GEN7_WM_MSRAST_OFF_PIXEL;
-         dw2 |= GEN7_WM_MSDISPMODE_PERSAMPLE;
-      }
-
-   BEGIN_BATCH(3);
-   OUT_BATCH(_3DSTATE_WM << 16 | (3 - 2));
-   OUT_BATCH(dw1);
-   OUT_BATCH(dw2);
-   ADVANCE_BATCH();
-}
-
-
-/**
- * 3DSTATE_PS
- *
- * Pixel shader dispatch is disabled above in 3DSTATE_WM, dw1.29. Despite
- * that, thread dispatch info must still be specified.
- *     - Maximum Number of Threads (dw4.24:31) must be nonzero, as the
- *       valid range for this field is [0x3, 0x2f].
- *     - A dispatch mode must be given; that is, at least one of the
- *       "N Pixel Dispatch Enable" (N=8,16,32) fields must be set. This was
- *       discovered through simulator error messages.
- */
-static void
-gen7_blorp_emit_ps_config(struct brw_context *brw,
-                          const struct brw_blorp_params *params)
-{
-   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
-   uint32_t dw2, dw4, dw5;
-   const int max_threads_shift = brw->is_haswell ?
-      HSW_PS_MAX_THREADS_SHIFT : IVB_PS_MAX_THREADS_SHIFT;
-
-   dw2 = dw4 = dw5 = 0;
-   dw4 |= (brw->max_wm_threads - 1) << max_threads_shift;
-
-   /* If there's a WM program, we need to do 16-pixel dispatch since that's
-    * what the program is compiled for.  If there isn't, then it shouldn't
-    * matter because no program is actually being run.  However, the hardware
-    * gets angry if we don't enable at least one dispatch mode, so just enable
-    * 16-pixel dispatch unconditionally.
-    */
-   dw4 |= GEN7_PS_16_DISPATCH_ENABLE;
-
-   if (brw->is_haswell)
-      dw4 |= SET_FIELD(1, HSW_PS_SAMPLE_MASK); /* 1 sample for now */
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
-      dw4 |= GEN7_PS_PUSH_CONSTANT_ENABLE;
-      dw5 |= prog_data->first_curbe_grf << GEN7_PS_DISPATCH_START_GRF_SHIFT_0;
-   }
-
-   if (params->src.mt)
-      dw2 |= 1 << GEN7_PS_SAMPLER_COUNT_SHIFT; /* Up to 4 samplers */
-
-   dw4 |= params->fast_clear_op;
-
-   BEGIN_BATCH(8);
-   OUT_BATCH(_3DSTATE_PS << 16 | (8 - 2));
-   OUT_BATCH(params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG ?
-             params->wm_prog_kernel : 0);
-   OUT_BATCH(dw2);
-   OUT_BATCH(0);
-   OUT_BATCH(dw4);
-   OUT_BATCH(dw5);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-void
-gen7_blorp_emit_binding_table_pointers_ps(struct brw_context *brw,
-                                          uint32_t wm_bind_bo_offset)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_BINDING_TABLE_POINTERS_PS << 16 | (2 - 2));
-   OUT_BATCH(wm_bind_bo_offset);
-   ADVANCE_BATCH();
-}
-
-
-void
-gen7_blorp_emit_sampler_state_pointers_ps(struct brw_context *brw,
-                                          uint32_t sampler_offset)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_SAMPLER_STATE_POINTERS_PS << 16 | (2 - 2));
-   OUT_BATCH(sampler_offset);
-   ADVANCE_BATCH();
-}
-
-
-void
-gen7_blorp_emit_constant_ps(struct brw_context *brw,
-                            uint32_t wm_push_const_offset)
-{
-   const uint8_t mocs = GEN7_MOCS_L3;
-
-   /* Make sure the push constants fill an exact integer number of
-    * registers.
-    */
-   assert(sizeof(struct brw_blorp_wm_push_constants) % 32 == 0);
-
-   /* There must be at least one register worth of push constant data. */
-   assert(BRW_BLORP_NUM_PUSH_CONST_REGS > 0);
-
-   /* Enable push constant buffer 0. */
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 |
-             (7 - 2));
-   OUT_BATCH(BRW_BLORP_NUM_PUSH_CONST_REGS);
-   OUT_BATCH(0);
-   OUT_BATCH(wm_push_const_offset | mocs);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-void
-gen7_blorp_emit_constant_ps_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(7);
-   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 | (7 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-static void
-gen7_blorp_emit_depth_stencil_config(struct brw_context *brw,
-                                     const struct brw_blorp_params *params)
-{
-   const uint8_t mocs = GEN7_MOCS_L3;
-   uint32_t surfwidth, surfheight;
-   uint32_t surftype;
-   unsigned int depth = MAX2(params->depth.mt->logical_depth0, 1);
-   unsigned int min_array_element;
-   GLenum gl_target = params->depth.mt->target;
-   unsigned int lod;
-
-   switch (gl_target) {
-   case GL_TEXTURE_CUBE_MAP_ARRAY:
-   case GL_TEXTURE_CUBE_MAP:
-      /* The PRM claims that we should use BRW_SURFACE_CUBE for this
-       * situation, but experiments show that gl_Layer doesn't work when we do
-       * this.  So we use BRW_SURFACE_2D, since for rendering purposes this is
-       * equivalent.
-       */
-      surftype = BRW_SURFACE_2D;
-      depth *= 6;
-      break;
-   default:
-      surftype = translate_tex_target(gl_target);
-      break;
-   }
-
-   min_array_element = params->depth.layer;
-   if (params->depth.mt->num_samples > 1) {
-      /* Convert physical layer to logical layer. */
-      min_array_element /= params->depth.mt->num_samples;
-   }
-
-   lod = params->depth.level - params->depth.mt->first_level;
-
-   if (params->hiz_op != GEN6_HIZ_OP_NONE && lod == 0) {
-      /* HIZ ops for lod 0 may set the width & height a little
-       * larger to allow the fast depth clear to fit the hardware
-       * alignment requirements. (8x4)
-       */
-      surfwidth = params->depth.width;
-      surfheight = params->depth.height;
-   } else {
-      surfwidth = params->depth.mt->logical_width0;
-      surfheight = params->depth.mt->logical_height0;
-   }
-
-   /* 3DSTATE_DEPTH_BUFFER */
-   {
-      brw_emit_depth_stall_flushes(brw);
-
-      BEGIN_BATCH(7);
-      OUT_BATCH(GEN7_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
-      OUT_BATCH((params->depth.mt->pitch - 1) |
-                params->depth_format << 18 |
-                1 << 22 | /* hiz enable */
-                1 << 28 | /* depth write */
-                surftype << 29);
-      OUT_RELOC(params->depth.mt->bo,
-                I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
-                0);
-      OUT_BATCH((surfwidth - 1) << 4 |
-                (surfheight - 1) << 18 |
-                lod);
-      OUT_BATCH(((depth - 1) << 21) |
-                (min_array_element << 10) |
-                mocs);
-      OUT_BATCH(0);
-      OUT_BATCH((depth - 1) << 21);
-      ADVANCE_BATCH();
-   }
-
-   /* 3DSTATE_HIER_DEPTH_BUFFER */
-   {
-      struct intel_miptree_aux_buffer *hiz_buf = params->depth.mt->hiz_buf;
-
-      BEGIN_BATCH(3);
-      OUT_BATCH((GEN7_3DSTATE_HIER_DEPTH_BUFFER << 16) | (3 - 2));
-      OUT_BATCH((mocs << 25) |
-                (hiz_buf->pitch - 1));
-      OUT_RELOC(hiz_buf->bo,
-                I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
-                0);
-      ADVANCE_BATCH();
-   }
-
-   /* 3DSTATE_STENCIL_BUFFER */
-   {
-      BEGIN_BATCH(3);
-      OUT_BATCH((GEN7_3DSTATE_STENCIL_BUFFER << 16) | (3 - 2));
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      ADVANCE_BATCH();
-   }
-}
-
-
-static void
-gen7_blorp_emit_depth_disable(struct brw_context *brw)
-{
-   brw_emit_depth_stall_flushes(brw);
-
-   BEGIN_BATCH(7);
-   OUT_BATCH(GEN7_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
-   OUT_BATCH(BRW_DEPTHFORMAT_D32_FLOAT << 18 | (BRW_SURFACE_NULL << 29));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(3);
-   OUT_BATCH(GEN7_3DSTATE_HIER_DEPTH_BUFFER << 16 | (3 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(3);
-   OUT_BATCH(GEN7_3DSTATE_STENCIL_BUFFER << 16 | (3 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DSTATE_CLEAR_PARAMS
- *
- * From the Ivybridge PRM, Volume 2 Part 1, Section 11.5.5.4
- * 3DSTATE_CLEAR_PARAMS:
- *    3DSTATE_CLEAR_PARAMS must always be programmed in the along
- *    with the other Depth/Stencil state commands(i.e.  3DSTATE_DEPTH_BUFFER,
- *    3DSTATE_STENCIL_BUFFER, or 3DSTATE_HIER_DEPTH_BUFFER).
- */
-void
-gen7_blorp_emit_clear_params(struct brw_context *brw,
-                             const struct brw_blorp_params *params)
-{
-   BEGIN_BATCH(3);
-   OUT_BATCH(GEN7_3DSTATE_CLEAR_PARAMS << 16 | (3 - 2));
-   OUT_BATCH(params->depth.mt ? params->depth.mt->depth_clear_value : 0);
-   OUT_BATCH(GEN7_DEPTH_CLEAR_VALID);
-   ADVANCE_BATCH();
-}
-
-
-/* 3DPRIMITIVE */
-void
-gen7_blorp_emit_primitive(struct brw_context *brw,
-                          const struct brw_blorp_params *params)
-{
-   BEGIN_BATCH(7);
-   OUT_BATCH(CMD_3D_PRIM << 16 | (7 - 2));
-   OUT_BATCH(GEN7_3DPRIM_VERTEXBUFFER_ACCESS_SEQUENTIAL |
-             _3DPRIM_RECTLIST);
-   OUT_BATCH(3); /* vertex count per instance */
-   OUT_BATCH(0);
-   OUT_BATCH(params->num_layers); /* instance count */
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-
-/**
- * \copydoc gen6_blorp_exec()
- */
-void
-gen7_blorp_exec(struct brw_context *brw,
-                const struct brw_blorp_params *params)
-{
-   if (brw->gen >= 8)
-      return;
-
-   uint32_t cc_blend_state_offset = 0;
-   uint32_t cc_state_offset = 0;
-   uint32_t depthstencil_offset;
-   uint32_t wm_push_const_offset = 0;
-   uint32_t wm_bind_bo_offset = 0;
-
-   gen6_emit_3dstate_multisample(brw, params->dst.num_samples);
-   gen6_emit_3dstate_sample_mask(brw,
-                                 params->dst.num_samples > 1 ?
-                                 (1 << params->dst.num_samples) - 1 : 1);
-   gen6_blorp_emit_state_base_address(brw, params);
-   gen6_blorp_emit_vertices(brw, params);
-   gen7_blorp_emit_urb_config(brw);
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
-      cc_blend_state_offset = gen6_blorp_emit_blend_state(brw, params);
-      cc_state_offset = gen6_blorp_emit_cc_state(brw);
-      gen7_blorp_emit_blend_state_pointer(brw, cc_blend_state_offset);
-      gen7_blorp_emit_cc_state_pointer(brw, cc_state_offset);
-   }
-   depthstencil_offset = gen6_blorp_emit_depth_stencil_state(brw, params);
-   gen7_blorp_emit_depth_stencil_state_pointers(brw, depthstencil_offset);
-   if (brw->use_resource_streamer)
-      gen7_disable_hw_binding_tables(brw);
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
-      uint32_t wm_surf_offset_renderbuffer;
-      uint32_t wm_surf_offset_texture = 0;
-      wm_push_const_offset = gen6_blorp_emit_wm_constants(brw, params);
-      intel_miptree_used_for_rendering(params->dst.mt);
-      wm_surf_offset_renderbuffer =
-         gen7_blorp_emit_surface_state(brw, &params->dst,
-                                       I915_GEM_DOMAIN_RENDER,
-                                       I915_GEM_DOMAIN_RENDER,
-                                       true /* is_render_target */);
-      if (params->src.mt) {
-         wm_surf_offset_texture =
-            gen7_blorp_emit_surface_state(brw, &params->src,
-                                          I915_GEM_DOMAIN_SAMPLER, 0,
-                                          false /* is_render_target */);
-      }
-      wm_bind_bo_offset =
-         gen6_blorp_emit_binding_table(brw,
-                                       wm_surf_offset_renderbuffer,
-                                       wm_surf_offset_texture);
-   }
-   gen7_blorp_emit_vs_disable(brw);
-   gen7_blorp_emit_hs_disable(brw);
-   gen7_blorp_emit_te_disable(brw);
-   gen7_blorp_emit_ds_disable(brw);
-   gen7_blorp_emit_gs_disable(brw);
-   gen7_blorp_emit_streamout_disable(brw);
-   gen6_blorp_emit_clip_disable(brw);
-   gen7_blorp_emit_sf_config(brw, params);
-   gen7_blorp_emit_wm_config(brw, params);
-   if (params->wm_prog_kernel != BRW_BLORP_NO_WM_PROG) {
-      gen7_blorp_emit_binding_table_pointers_ps(brw, wm_bind_bo_offset);
-      gen7_blorp_emit_constant_ps(brw, wm_push_const_offset);
-   } else {
-      gen7_blorp_emit_constant_ps_disable(brw);
-   }
-
-   if (params->src.mt) {
-      const uint32_t sampler_offset =
-         gen6_blorp_emit_sampler_state(brw, BRW_MAPFILTER_LINEAR, 0, true);
-      gen7_blorp_emit_sampler_state_pointers_ps(brw, sampler_offset);
-   }
-
-   gen7_blorp_emit_ps_config(brw, params);
-   gen7_blorp_emit_cc_viewport(brw);
-
-   if (params->depth.mt)
-      gen7_blorp_emit_depth_stencil_config(brw, params);
-   else
-      gen7_blorp_emit_depth_disable(brw);
-   gen7_blorp_emit_clear_params(brw, params);
-   gen6_blorp_emit_drawing_rectangle(brw, params);
-   gen7_blorp_emit_primitive(brw, params);
-}
diff --git a/src/mesa/drivers/dri/i965/gen8_blorp.c b/src/mesa/drivers/dri/i965/gen8_blorp.c
new file mode 100644
index 0000000..8da51e1
--- /dev/null
+++ b/src/mesa/drivers/dri/i965/gen8_blorp.c
@@ -0,0 +1,713 @@
+/*
+ * Copyright © 2016 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include <assert.h>
+
+#include "intel_batchbuffer.h"
+#include "intel_fbo.h"
+#include "intel_mipmap_tree.h"
+
+#include "brw_context.h"
+#include "brw_defines.h"
+#include "brw_state.h"
+
+#include "brw_blorp.h"
+
+
+/* SURFACE_STATE for renderbuffer or texture surface (see
+ * brw_update_renderbuffer_surface and brw_update_texture_surface)
+ */
+static uint32_t
+gen8_blorp_emit_surface_state(struct brw_context *brw,
+                              const struct brw_blorp_surface_info *surface,
+                              uint32_t read_domains, uint32_t write_domain,
+                              bool is_render_target)
+{
+   uint32_t wm_surf_offset;
+   const struct intel_mipmap_tree *mt = surface->mt;
+   const uint32_t mocs_wb = is_render_target ?
+                               (brw->gen >= 9 ? SKL_MOCS_PTE : BDW_MOCS_PTE) :
+                               (brw->gen >= 9 ? SKL_MOCS_WB : BDW_MOCS_WB);
+   const uint32_t tiling = surface->map_stencil_as_y_tiled
+      ? I915_TILING_Y : mt->tiling;
+   uint32_t tile_x, tile_y;
+
+   uint32_t *surf = gen8_allocate_surface_state(brw, &wm_surf_offset, -1);
+
+   surf[0] = BRW_SURFACE_2D << BRW_SURFACE_TYPE_SHIFT |
+             surface->brw_surfaceformat << BRW_SURFACE_FORMAT_SHIFT |
+             gen8_vertical_alignment(brw, mt, BRW_SURFACE_2D) |
+             gen8_horizontal_alignment(brw, mt, BRW_SURFACE_2D) |
+             gen8_surface_tiling_mode(tiling);
+
+   surf[1] = SET_FIELD(mocs_wb, GEN8_SURFACE_MOCS) | mt->qpitch >> 2;
+
+   surf[2] = SET_FIELD(surface->width - 1, GEN7_SURFACE_WIDTH) |
+             SET_FIELD(surface->height - 1, GEN7_SURFACE_HEIGHT);
+
+   uint32_t pitch_bytes = mt->pitch;
+   if (surface->map_stencil_as_y_tiled)
+      pitch_bytes *= 2;
+   surf[3] = pitch_bytes - 1;
+
+   surf[4] = gen7_surface_msaa_bits(surface->num_samples,
+                                    surface->msaa_layout);
+
+   if (surface->mt->mcs_mt) {
+      surf[6] = SET_FIELD(surface->mt->qpitch / 4, GEN8_SURFACE_AUX_QPITCH) |
+                SET_FIELD((surface->mt->mcs_mt->pitch / 128) - 1,
+                          GEN8_SURFACE_AUX_PITCH) |
+                GEN8_SURFACE_AUX_MODE_MCS;
+   } else {
+      surf[6] = 0;
+   }
+
+   gen8_emit_fast_clear_color(brw, mt, surf);
+   surf[7] |= SET_FIELD(HSW_SCS_RED,   GEN7_SURFACE_SCS_R) |
+              SET_FIELD(HSW_SCS_GREEN, GEN7_SURFACE_SCS_G) |
+              SET_FIELD(HSW_SCS_BLUE,  GEN7_SURFACE_SCS_B) |
+              SET_FIELD(HSW_SCS_ALPHA, GEN7_SURFACE_SCS_A);
+
+    /* reloc */
+   *((uint64_t *)&surf[8]) =
+      brw_blorp_compute_tile_offsets(surface, &tile_x, &tile_y) +
+      mt->bo->offset64;
+
+   /* Note that the low bits of these fields are missing, so there's the
+    * possibility of getting in trouble.
+    */
+   assert(tile_x % 4 == 0);
+   assert(tile_y % 4 == 0);
+   surf[5] = SET_FIELD(tile_x / 4, BRW_SURFACE_X_OFFSET) |
+             SET_FIELD(tile_y / 4, GEN8_SURFACE_Y_OFFSET);
+
+   if (brw->gen >= 9) {
+      /* Disable Mip Tail by setting a large value. */
+      surf[5] |= SET_FIELD(15, GEN9_SURFACE_MIP_TAIL_START_LOD);
+   }
+
+   if (surface->mt->mcs_mt) {
+      *((uint64_t *) &surf[10]) = surface->mt->mcs_mt->bo->offset64;
+      drm_intel_bo_emit_reloc(brw->batch.bo,
+                              wm_surf_offset + 10 * 4,
+                              surface->mt->mcs_mt->bo, 0,
+                              read_domains, write_domain);
+   }
+
+   /* Emit relocation to surface contents */
+   drm_intel_bo_emit_reloc(brw->batch.bo,
+                           wm_surf_offset + 8 * 4,
+                           mt->bo,
+                           surf[8] - mt->bo->offset64,
+                           read_domains, write_domain);
+
+   return wm_surf_offset;
+}
+
+static uint32_t
+gen8_blorp_emit_blend_state(struct brw_context *brw,
+                            const struct brw_blorp_params *params)
+{
+   uint32_t blend_state_offset;
+
+   assume(params->num_draw_buffers);
+
+   const unsigned size = 4 + 8 * params->num_draw_buffers;
+   uint32_t *blend = (uint32_t *)brw_state_batch(brw, AUB_TRACE_BLEND_STATE,
+                                                 size, 64,
+                                                 &blend_state_offset);
+   memset(blend, 0, size);
+
+   for (unsigned i = 0; i < params->num_draw_buffers; ++i) {
+      if (params->color_write_disable[0])
+         blend[1 + 2 * i] |= GEN8_BLEND_WRITE_DISABLE_RED;
+      if (params->color_write_disable[1])
+         blend[1 + 2 * i] |= GEN8_BLEND_WRITE_DISABLE_GREEN;
+      if (params->color_write_disable[2])
+         blend[1 + 2 * i] |= GEN8_BLEND_WRITE_DISABLE_BLUE;
+      if (params->color_write_disable[3])
+         blend[1 + 2 * i] |= GEN8_BLEND_WRITE_DISABLE_ALPHA;
+
+      blend[1 + 2 * i + 1] = GEN8_BLEND_PRE_BLEND_COLOR_CLAMP_ENABLE |
+                             GEN8_BLEND_POST_BLEND_COLOR_CLAMP_ENABLE |
+                             GEN8_BLEND_COLOR_CLAMP_RANGE_RTFORMAT;
+   }
+
+   return blend_state_offset;
+}
+
+static void
+gen8_blorp_emit_disable_constant_state(struct brw_context *brw,
+                                       unsigned opcode)
+{
+   BEGIN_BATCH(11);
+   OUT_BATCH(opcode << 16 | (11 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_disable_binding_table(struct brw_context *brw,
+                                      unsigned opcode)
+{
+
+   BEGIN_BATCH(2);
+   OUT_BATCH(opcode << 16 | (2 - 2));
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+/* 3DSTATE_VS
+ *
+ * Disable vertex shader.
+ */
+static void
+gen8_blorp_emit_vs_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(9);
+   OUT_BATCH(_3DSTATE_VS << 16 | (9 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+/* 3DSTATE_HS
+ *
+ * Disable the hull shader.
+ */
+static void
+gen8_blorp_emit_hs_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(9);
+   OUT_BATCH(_3DSTATE_HS << 16 | (9 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+/* 3DSTATE_DS
+ *
+ * Disable the domain shader.
+ */
+static void
+gen8_blorp_emit_ds_disable(struct brw_context *brw)
+{
+   const int ds_pkt_len = brw->gen >= 9 ? 11 : 9;
+   BEGIN_BATCH(ds_pkt_len);
+   OUT_BATCH(_3DSTATE_DS << 16 | (ds_pkt_len - 2));
+   for (int i = 0; i < ds_pkt_len - 1; i++)
+      OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+/* 3DSTATE_GS
+ *
+ * Disable the geometry shader.
+ */
+static void
+gen8_blorp_emit_gs_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(10);
+   OUT_BATCH(_3DSTATE_GS << 16 | (10 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+/* 3DSTATE_STREAMOUT
+ *
+ * Disable streamout.
+ */
+static void
+gen8_blorp_emit_streamout_disable(struct brw_context *brw)
+{
+   BEGIN_BATCH(5);
+   OUT_BATCH(_3DSTATE_STREAMOUT << 16 | (5 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_raster_state(struct brw_context *brw)
+{
+   BEGIN_BATCH(5);
+   OUT_BATCH(_3DSTATE_RASTER << 16 | (5 - 2));
+   OUT_BATCH(GEN8_RASTER_CULL_NONE);
+   OUT_BATCH_F(0);
+   OUT_BATCH_F(0);
+   OUT_BATCH_F(0);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_sbe_state(struct brw_context *brw,
+                          const struct brw_blorp_params *params)
+{
+   /* 3DSTATE_SBE */
+   {
+      const unsigned sbe_cmd_length = brw->gen == 8 ? 4 : 6;
+      BEGIN_BATCH(sbe_cmd_length);
+      OUT_BATCH(_3DSTATE_SBE << 16 | (sbe_cmd_length - 2));
+      OUT_BATCH(GEN7_SBE_SWIZZLE_ENABLE |
+                params->num_varyings << GEN7_SBE_NUM_OUTPUTS_SHIFT |
+                1 << GEN7_SBE_URB_ENTRY_READ_LENGTH_SHIFT |
+                BRW_SF_URB_ENTRY_READ_OFFSET <<
+                   GEN8_SBE_URB_ENTRY_READ_OFFSET_SHIFT |
+                GEN8_SBE_FORCE_URB_ENTRY_READ_LENGTH |
+                GEN8_SBE_FORCE_URB_ENTRY_READ_OFFSET);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      if (sbe_cmd_length >= 6) {
+         OUT_BATCH(GEN9_SBE_ACTIVE_COMPONENT_XYZW << (0 << 1));
+         OUT_BATCH(0);
+      }
+      ADVANCE_BATCH();
+   }
+
+   {
+      BEGIN_BATCH(11);
+      OUT_BATCH(_3DSTATE_SBE_SWIZ << 16 | (11 - 2));
+
+      /* Output DWords 1 through 8: */
+      for (int i = 0; i < 8; i++) {
+         OUT_BATCH(0);
+      }
+
+      OUT_BATCH(0); /* wrapshortest enables 0-7 */
+      OUT_BATCH(0); /* wrapshortest enables 8-15 */
+      ADVANCE_BATCH();
+   }
+}
+
+static void
+gen8_blorp_emit_sf_config(struct brw_context *brw)
+{
+   /* See gen6_blorp_emit_sf_config() */
+   BEGIN_BATCH(4);
+   OUT_BATCH(_3DSTATE_SF << 16 | (4 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(GEN6_SF_LINE_AA_MODE_TRUE);
+   ADVANCE_BATCH();
+}
+
+/**
+ * Disable thread dispatch (dw5.19) and enable the HiZ op.
+ */
+static void
+gen8_blorp_emit_wm_state(struct brw_context *brw)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_WM << 16 | (2 - 2));
+   OUT_BATCH(GEN7_WM_LINE_AA_WIDTH_1_0 |
+             GEN7_WM_LINE_END_CAP_AA_WIDTH_0_5 |
+             GEN7_WM_POINT_RASTRULE_UPPER_RIGHT);
+   ADVANCE_BATCH();
+}
+
+/**
+ * 3DSTATE_PS
+ *
+ * Pixel shader dispatch is disabled above in 3DSTATE_WM, dw1.29. Despite
+ * that, thread dispatch info must still be specified.
+ *     - Maximum Number of Threads (dw4.24:31) must be nonzero, as the
+ *       valid range for this field is [0x3, 0x2f].
+ *     - A dispatch mode must be given; that is, at least one of the
+ *       "N Pixel Dispatch Enable" (N=8,16,32) fields must be set. This was
+ *       discovered through simulator error messages.
+ */
+static void
+gen8_blorp_emit_ps_config(struct brw_context *brw,
+                          const struct brw_blorp_params *params)
+{
+   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
+   uint32_t dw3, dw5, dw6, dw7;
+
+   dw3 = dw5 = dw6 = dw7 = 0;
+   dw3 |= GEN7_PS_VECTOR_MASK_ENABLE;
+
+   dw6 |= GEN7_PS_16_DISPATCH_ENABLE;
+
+   if (params->src.mt) {
+      dw3 |= 1 << GEN7_PS_SAMPLER_COUNT_SHIFT; /* Up to 4 samplers */
+      dw3 |= 2 << GEN7_PS_BINDING_TABLE_ENTRY_COUNT_SHIFT; /* Two surfaces */
+   } else {
+      dw3 |= 1 << GEN7_PS_BINDING_TABLE_ENTRY_COUNT_SHIFT; /* One surface */
+   }
+
+   dw6 |= GEN7_PS_PUSH_CONSTANT_ENABLE;
+   dw7 |= prog_data->first_curbe_grf << GEN7_PS_DISPATCH_START_GRF_SHIFT_0;
+
+   /* 3DSTATE_PS expects the number of threads per PSD, which is always 64;
+    * it implicitly scales for different GT levels (which have some # of PSDs).
+    *
+    * In Gen8 the format is U8-2 whereas in Gen9 it is U8-1.
+    */
+   if (brw->gen >= 9)
+      dw6 |= (64 - 1) << HSW_PS_MAX_THREADS_SHIFT;
+   else
+      dw6 |= (64 - 2) << HSW_PS_MAX_THREADS_SHIFT;
+
+   dw6 |= GEN7_PS_POSOFFSET_NONE;
+   dw6 |= params->fast_clear_op;
+
+   BEGIN_BATCH(12);
+   OUT_BATCH(_3DSTATE_PS << 16 | (12 - 2));
+   OUT_BATCH(params->wm_prog_kernel);
+   OUT_BATCH(0);
+   OUT_BATCH(dw3);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(dw6);
+   OUT_BATCH(dw7);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_ps_blend(struct brw_context *brw)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_PS_BLEND << 16 | (2 - 2));
+   OUT_BATCH(GEN8_PS_BLEND_HAS_WRITEABLE_RT);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_ps_extra(struct brw_context *brw,
+                         const struct brw_blorp_params *params)
+{
+   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
+   uint32_t dw1 = 0;
+
+   dw1 |= GEN8_PSX_PIXEL_SHADER_VALID;
+
+   if (params->src.mt) {
+      dw1 |= GEN8_PSX_KILL_ENABLE;
+      dw1 |= GEN8_PSX_ATTRIBUTE_ENABLE;
+   }
+
+   if (params->dst.num_samples > 1 && prog_data &&
+       prog_data->persample_msaa_dispatch)
+      dw1 |= GEN8_PSX_SHADER_IS_PER_SAMPLE;
+
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_PS_EXTRA << 16 | (2 - 2));
+   OUT_BATCH(dw1);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_depth_disable(struct brw_context *brw)
+{
+   /* Skip repeated NULL depth/stencil emits (think 2D rendering). */
+   if (brw->no_depth_or_stencil)
+      return;
+
+   brw_emit_depth_stall_flushes(brw);
+
+   BEGIN_BATCH(8);
+   OUT_BATCH(GEN7_3DSTATE_DEPTH_BUFFER << 16 | (8 - 2));
+   OUT_BATCH((BRW_DEPTHFORMAT_D32_FLOAT << 18) | (BRW_SURFACE_NULL << 29));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(5);
+   OUT_BATCH(GEN7_3DSTATE_HIER_DEPTH_BUFFER << 16 | (5 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+
+   BEGIN_BATCH(5);
+   OUT_BATCH(GEN7_3DSTATE_STENCIL_BUFFER << 16 | (5 - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_vf_topology(struct brw_context *brw)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_VF_TOPOLOGY << 16 | (2 - 2));
+   OUT_BATCH(_3DPRIM_RECTLIST);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_vf_sys_gen_vals_state(struct brw_context *brw)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_VF_SGVS << 16 | (2 - 2));
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_vf_instancing_state(struct brw_context *brw,
+                                    unsigned num_elems)
+{
+   for (unsigned i = 0; i < num_elems; ++i) {
+      BEGIN_BATCH(3);
+      OUT_BATCH(_3DSTATE_VF_INSTANCING << 16 | (3 - 2));
+      OUT_BATCH(i);
+      OUT_BATCH(0);
+      ADVANCE_BATCH();
+   }
+}
+
+static void
+gen8_blorp_emit_vf_state(struct brw_context *brw)
+{
+   BEGIN_BATCH(2);
+   OUT_BATCH(_3DSTATE_VF << 16 | (2 - 2));
+   OUT_BATCH(0);
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_depth_stencil_state(struct brw_context *brw,
+                                    const struct brw_blorp_params *params)
+{
+   const unsigned pkt_len = brw->gen >= 9 ? 4 : 3;
+
+   BEGIN_BATCH(pkt_len);
+   OUT_BATCH(_3DSTATE_WM_DEPTH_STENCIL << 16 | (pkt_len - 2));
+   OUT_BATCH(0);
+   OUT_BATCH(0);
+   if (pkt_len > 3) {
+      OUT_BATCH(0);
+   }
+   ADVANCE_BATCH();
+}
+
+static void
+gen8_blorp_emit_constant_ps(struct brw_context *brw,
+                            uint32_t wm_push_const_offset)
+{
+   const int dwords = brw->gen >= 8 ? 11 : 7;
+   BEGIN_BATCH(dwords);
+   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 | (dwords - 2));
+
+   if (brw->gen >= 9) {
+      OUT_BATCH(0);
+      OUT_BATCH(BRW_BLORP_NUM_PUSH_CONST_REGS);
+   } else {
+      OUT_BATCH(BRW_BLORP_NUM_PUSH_CONST_REGS);
+      OUT_BATCH(0);
+   }
+
+   if (brw->gen >= 9) {
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_RELOC64(brw->batch.bo, I915_GEM_DOMAIN_RENDER, 0,
+                  wm_push_const_offset);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+   } else {
+      OUT_BATCH(wm_push_const_offset);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+      OUT_BATCH(0);
+   }
+
+   ADVANCE_BATCH();
+}
+
+static uint32_t
+gen8_blorp_emit_surface_states(struct brw_context *brw,
+                               const struct brw_blorp_params *params)
+{
+   uint32_t wm_surf_offset_renderbuffer;
+   uint32_t wm_surf_offset_texture = 0;
+
+   intel_miptree_used_for_rendering(params->dst.mt);
+
+   wm_surf_offset_renderbuffer =
+      gen8_blorp_emit_surface_state(brw, &params->dst,
+                                    I915_GEM_DOMAIN_RENDER,
+                                    I915_GEM_DOMAIN_RENDER,
+                                    true /* is_render_target */);
+   if (params->src.mt) {
+      const struct brw_blorp_surface_info *surface = &params->src;
+      struct intel_mipmap_tree *mt = surface->mt;
+
+      /* Textures are always sampled as 2D. */
+      const bool is_cube = mt->target == GL_TEXTURE_CUBE_MAP_ARRAY ||
+                           mt->target == GL_TEXTURE_CUBE_MAP;
+      const unsigned depth = (is_cube ? 6 : 1) * mt->logical_depth0;
+      const GLenum target = is_cube ? GL_TEXTURE_2D_ARRAY : mt->target;
+      const unsigned max_level = surface->level + mt->last_level + 1;
+      const unsigned layer = mt->target != GL_TEXTURE_3D ?
+                                surface->layer / MAX2(mt->num_samples, 1) : 0;
+
+      brw->vtbl.emit_texture_surface_state(brw, mt, target,
+                                           layer, layer + depth,
+                                           surface->level, max_level,
+                                           surface->brw_surfaceformat,
+                                           surface->swizzle,
+                                           &wm_surf_offset_texture,
+                                           -1, false, false);
+   }
+
+   return gen6_blorp_emit_binding_table(brw,
+                                        wm_surf_offset_renderbuffer,
+                                        wm_surf_offset_texture);
+}
+
+/**
+ * \copydoc gen6_blorp_exec()
+ */
+void
+gen8_blorp_exec(struct brw_context *brw, const struct brw_blorp_params *params)
+{
+   uint32_t wm_bind_bo_offset = 0;
+
+   gen8_upload_state_base_address(brw);
+   gen7_blorp_emit_cc_viewport(brw);
+   gen7_l3_state.emit(brw);
+
+   gen7_blorp_emit_urb_config(brw);
+
+   const uint32_t cc_blend_state_offset =
+      gen8_blorp_emit_blend_state(brw, params);
+   gen7_blorp_emit_blend_state_pointer(brw, cc_blend_state_offset);
+
+   const uint32_t cc_state_offset = gen6_blorp_emit_cc_state(brw);
+   gen7_blorp_emit_cc_state_pointer(brw, cc_state_offset);
+
+   gen8_blorp_emit_disable_constant_state(brw, _3DSTATE_CONSTANT_VS);
+   gen8_blorp_emit_disable_constant_state(brw, _3DSTATE_CONSTANT_HS);
+   gen8_blorp_emit_disable_constant_state(brw, _3DSTATE_CONSTANT_DS);
+   gen8_blorp_emit_disable_constant_state(brw, _3DSTATE_CONSTANT_GS);
+
+   const uint32_t wm_push_const_offset =
+      gen6_blorp_emit_wm_constants(brw, params);
+   gen8_blorp_emit_constant_ps(brw, wm_push_const_offset);
+   wm_bind_bo_offset = gen8_blorp_emit_surface_states(brw, params);
+
+   gen8_blorp_emit_disable_binding_table(brw,
+                                         _3DSTATE_BINDING_TABLE_POINTERS_VS);
+   gen8_blorp_emit_disable_binding_table(brw,
+                                         _3DSTATE_BINDING_TABLE_POINTERS_HS);
+   gen8_blorp_emit_disable_binding_table(brw,
+                                         _3DSTATE_BINDING_TABLE_POINTERS_DS);
+   gen8_blorp_emit_disable_binding_table(brw,
+                                         _3DSTATE_BINDING_TABLE_POINTERS_GS);
+
+   gen7_blorp_emit_binding_table_pointers_ps(brw, wm_bind_bo_offset);
+
+   if (params->src.mt) {
+      const uint32_t sampler_offset =
+         gen6_blorp_emit_sampler_state(brw, BRW_MAPFILTER_LINEAR, 0, true);
+      gen7_blorp_emit_sampler_state_pointers_ps(brw, sampler_offset);
+   }
+
+   gen8_emit_3dstate_multisample(brw, params->dst.num_samples);
+   gen6_emit_3dstate_sample_mask(brw,
+                                 params->dst.num_samples > 1 ?
+                                    (1 << params->dst.num_samples) - 1 : 1);
+
+   gen8_disable_stages.emit(brw);
+   gen8_blorp_emit_vs_disable(brw);
+   gen8_blorp_emit_hs_disable(brw);
+   gen7_blorp_emit_te_disable(brw);
+   gen8_blorp_emit_ds_disable(brw);
+   gen8_blorp_emit_gs_disable(brw);
+
+   gen8_blorp_emit_streamout_disable(brw);
+   gen6_blorp_emit_clip_disable(brw);
+   gen8_blorp_emit_raster_state(brw);
+   gen8_blorp_emit_sbe_state(brw, params);
+   gen8_blorp_emit_sf_config(brw);
+
+   gen8_blorp_emit_ps_blend(brw);
+   gen8_blorp_emit_ps_extra(brw, params);
+
+   gen8_blorp_emit_ps_config(brw, params);
+
+   gen8_blorp_emit_depth_stencil_state(brw, params);
+   gen8_blorp_emit_wm_state(brw);
+
+   gen8_blorp_emit_depth_disable(brw);
+   gen7_blorp_emit_clear_params(brw, params);
+   gen6_blorp_emit_drawing_rectangle(brw, params);
+   gen8_blorp_emit_vf_topology(brw);
+   gen8_blorp_emit_vf_sys_gen_vals_state(brw);
+   gen6_blorp_emit_vertices(brw, params);
+   gen8_blorp_emit_vf_instancing_state(brw, 2);
+   gen8_blorp_emit_vf_state(brw);
+   gen7_blorp_emit_primitive(brw, params);
+
+   if (brw->gen < 9)
+      gen8_write_pma_stall_bits(brw, 0);
+}
diff --git a/src/mesa/drivers/dri/i965/gen8_blorp.cpp b/src/mesa/drivers/dri/i965/gen8_blorp.cpp
deleted file mode 100644
index 8da51e1..0000000
--- a/src/mesa/drivers/dri/i965/gen8_blorp.cpp
+++ /dev/null
@@ -1,713 +0,0 @@
-/*
- * Copyright © 2016 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- */
-
-#include <assert.h>
-
-#include "intel_batchbuffer.h"
-#include "intel_fbo.h"
-#include "intel_mipmap_tree.h"
-
-#include "brw_context.h"
-#include "brw_defines.h"
-#include "brw_state.h"
-
-#include "brw_blorp.h"
-
-
-/* SURFACE_STATE for renderbuffer or texture surface (see
- * brw_update_renderbuffer_surface and brw_update_texture_surface)
- */
-static uint32_t
-gen8_blorp_emit_surface_state(struct brw_context *brw,
-                              const struct brw_blorp_surface_info *surface,
-                              uint32_t read_domains, uint32_t write_domain,
-                              bool is_render_target)
-{
-   uint32_t wm_surf_offset;
-   const struct intel_mipmap_tree *mt = surface->mt;
-   const uint32_t mocs_wb = is_render_target ?
-                               (brw->gen >= 9 ? SKL_MOCS_PTE : BDW_MOCS_PTE) :
-                               (brw->gen >= 9 ? SKL_MOCS_WB : BDW_MOCS_WB);
-   const uint32_t tiling = surface->map_stencil_as_y_tiled
-      ? I915_TILING_Y : mt->tiling;
-   uint32_t tile_x, tile_y;
-
-   uint32_t *surf = gen8_allocate_surface_state(brw, &wm_surf_offset, -1);
-
-   surf[0] = BRW_SURFACE_2D << BRW_SURFACE_TYPE_SHIFT |
-             surface->brw_surfaceformat << BRW_SURFACE_FORMAT_SHIFT |
-             gen8_vertical_alignment(brw, mt, BRW_SURFACE_2D) |
-             gen8_horizontal_alignment(brw, mt, BRW_SURFACE_2D) |
-             gen8_surface_tiling_mode(tiling);
-
-   surf[1] = SET_FIELD(mocs_wb, GEN8_SURFACE_MOCS) | mt->qpitch >> 2;
-
-   surf[2] = SET_FIELD(surface->width - 1, GEN7_SURFACE_WIDTH) |
-             SET_FIELD(surface->height - 1, GEN7_SURFACE_HEIGHT);
-
-   uint32_t pitch_bytes = mt->pitch;
-   if (surface->map_stencil_as_y_tiled)
-      pitch_bytes *= 2;
-   surf[3] = pitch_bytes - 1;
-
-   surf[4] = gen7_surface_msaa_bits(surface->num_samples,
-                                    surface->msaa_layout);
-
-   if (surface->mt->mcs_mt) {
-      surf[6] = SET_FIELD(surface->mt->qpitch / 4, GEN8_SURFACE_AUX_QPITCH) |
-                SET_FIELD((surface->mt->mcs_mt->pitch / 128) - 1,
-                          GEN8_SURFACE_AUX_PITCH) |
-                GEN8_SURFACE_AUX_MODE_MCS;
-   } else {
-      surf[6] = 0;
-   }
-
-   gen8_emit_fast_clear_color(brw, mt, surf);
-   surf[7] |= SET_FIELD(HSW_SCS_RED,   GEN7_SURFACE_SCS_R) |
-              SET_FIELD(HSW_SCS_GREEN, GEN7_SURFACE_SCS_G) |
-              SET_FIELD(HSW_SCS_BLUE,  GEN7_SURFACE_SCS_B) |
-              SET_FIELD(HSW_SCS_ALPHA, GEN7_SURFACE_SCS_A);
-
-    /* reloc */
-   *((uint64_t *)&surf[8]) =
-      brw_blorp_compute_tile_offsets(surface, &tile_x, &tile_y) +
-      mt->bo->offset64;
-
-   /* Note that the low bits of these fields are missing, so there's the
-    * possibility of getting in trouble.
-    */
-   assert(tile_x % 4 == 0);
-   assert(tile_y % 4 == 0);
-   surf[5] = SET_FIELD(tile_x / 4, BRW_SURFACE_X_OFFSET) |
-             SET_FIELD(tile_y / 4, GEN8_SURFACE_Y_OFFSET);
-
-   if (brw->gen >= 9) {
-      /* Disable Mip Tail by setting a large value. */
-      surf[5] |= SET_FIELD(15, GEN9_SURFACE_MIP_TAIL_START_LOD);
-   }
-
-   if (surface->mt->mcs_mt) {
-      *((uint64_t *) &surf[10]) = surface->mt->mcs_mt->bo->offset64;
-      drm_intel_bo_emit_reloc(brw->batch.bo,
-                              wm_surf_offset + 10 * 4,
-                              surface->mt->mcs_mt->bo, 0,
-                              read_domains, write_domain);
-   }
-
-   /* Emit relocation to surface contents */
-   drm_intel_bo_emit_reloc(brw->batch.bo,
-                           wm_surf_offset + 8 * 4,
-                           mt->bo,
-                           surf[8] - mt->bo->offset64,
-                           read_domains, write_domain);
-
-   return wm_surf_offset;
-}
-
-static uint32_t
-gen8_blorp_emit_blend_state(struct brw_context *brw,
-                            const struct brw_blorp_params *params)
-{
-   uint32_t blend_state_offset;
-
-   assume(params->num_draw_buffers);
-
-   const unsigned size = 4 + 8 * params->num_draw_buffers;
-   uint32_t *blend = (uint32_t *)brw_state_batch(brw, AUB_TRACE_BLEND_STATE,
-                                                 size, 64,
-                                                 &blend_state_offset);
-   memset(blend, 0, size);
-
-   for (unsigned i = 0; i < params->num_draw_buffers; ++i) {
-      if (params->color_write_disable[0])
-         blend[1 + 2 * i] |= GEN8_BLEND_WRITE_DISABLE_RED;
-      if (params->color_write_disable[1])
-         blend[1 + 2 * i] |= GEN8_BLEND_WRITE_DISABLE_GREEN;
-      if (params->color_write_disable[2])
-         blend[1 + 2 * i] |= GEN8_BLEND_WRITE_DISABLE_BLUE;
-      if (params->color_write_disable[3])
-         blend[1 + 2 * i] |= GEN8_BLEND_WRITE_DISABLE_ALPHA;
-
-      blend[1 + 2 * i + 1] = GEN8_BLEND_PRE_BLEND_COLOR_CLAMP_ENABLE |
-                             GEN8_BLEND_POST_BLEND_COLOR_CLAMP_ENABLE |
-                             GEN8_BLEND_COLOR_CLAMP_RANGE_RTFORMAT;
-   }
-
-   return blend_state_offset;
-}
-
-static void
-gen8_blorp_emit_disable_constant_state(struct brw_context *brw,
-                                       unsigned opcode)
-{
-   BEGIN_BATCH(11);
-   OUT_BATCH(opcode << 16 | (11 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_disable_binding_table(struct brw_context *brw,
-                                      unsigned opcode)
-{
-
-   BEGIN_BATCH(2);
-   OUT_BATCH(opcode << 16 | (2 - 2));
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-/* 3DSTATE_VS
- *
- * Disable vertex shader.
- */
-static void
-gen8_blorp_emit_vs_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(9);
-   OUT_BATCH(_3DSTATE_VS << 16 | (9 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-/* 3DSTATE_HS
- *
- * Disable the hull shader.
- */
-static void
-gen8_blorp_emit_hs_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(9);
-   OUT_BATCH(_3DSTATE_HS << 16 | (9 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-/* 3DSTATE_DS
- *
- * Disable the domain shader.
- */
-static void
-gen8_blorp_emit_ds_disable(struct brw_context *brw)
-{
-   const int ds_pkt_len = brw->gen >= 9 ? 11 : 9;
-   BEGIN_BATCH(ds_pkt_len);
-   OUT_BATCH(_3DSTATE_DS << 16 | (ds_pkt_len - 2));
-   for (int i = 0; i < ds_pkt_len - 1; i++)
-      OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-/* 3DSTATE_GS
- *
- * Disable the geometry shader.
- */
-static void
-gen8_blorp_emit_gs_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(10);
-   OUT_BATCH(_3DSTATE_GS << 16 | (10 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-/* 3DSTATE_STREAMOUT
- *
- * Disable streamout.
- */
-static void
-gen8_blorp_emit_streamout_disable(struct brw_context *brw)
-{
-   BEGIN_BATCH(5);
-   OUT_BATCH(_3DSTATE_STREAMOUT << 16 | (5 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_raster_state(struct brw_context *brw)
-{
-   BEGIN_BATCH(5);
-   OUT_BATCH(_3DSTATE_RASTER << 16 | (5 - 2));
-   OUT_BATCH(GEN8_RASTER_CULL_NONE);
-   OUT_BATCH_F(0);
-   OUT_BATCH_F(0);
-   OUT_BATCH_F(0);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_sbe_state(struct brw_context *brw,
-                          const struct brw_blorp_params *params)
-{
-   /* 3DSTATE_SBE */
-   {
-      const unsigned sbe_cmd_length = brw->gen == 8 ? 4 : 6;
-      BEGIN_BATCH(sbe_cmd_length);
-      OUT_BATCH(_3DSTATE_SBE << 16 | (sbe_cmd_length - 2));
-      OUT_BATCH(GEN7_SBE_SWIZZLE_ENABLE |
-                params->num_varyings << GEN7_SBE_NUM_OUTPUTS_SHIFT |
-                1 << GEN7_SBE_URB_ENTRY_READ_LENGTH_SHIFT |
-                BRW_SF_URB_ENTRY_READ_OFFSET <<
-                   GEN8_SBE_URB_ENTRY_READ_OFFSET_SHIFT |
-                GEN8_SBE_FORCE_URB_ENTRY_READ_LENGTH |
-                GEN8_SBE_FORCE_URB_ENTRY_READ_OFFSET);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      if (sbe_cmd_length >= 6) {
-         OUT_BATCH(GEN9_SBE_ACTIVE_COMPONENT_XYZW << (0 << 1));
-         OUT_BATCH(0);
-      }
-      ADVANCE_BATCH();
-   }
-
-   {
-      BEGIN_BATCH(11);
-      OUT_BATCH(_3DSTATE_SBE_SWIZ << 16 | (11 - 2));
-
-      /* Output DWords 1 through 8: */
-      for (int i = 0; i < 8; i++) {
-         OUT_BATCH(0);
-      }
-
-      OUT_BATCH(0); /* wrapshortest enables 0-7 */
-      OUT_BATCH(0); /* wrapshortest enables 8-15 */
-      ADVANCE_BATCH();
-   }
-}
-
-static void
-gen8_blorp_emit_sf_config(struct brw_context *brw)
-{
-   /* See gen6_blorp_emit_sf_config() */
-   BEGIN_BATCH(4);
-   OUT_BATCH(_3DSTATE_SF << 16 | (4 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(GEN6_SF_LINE_AA_MODE_TRUE);
-   ADVANCE_BATCH();
-}
-
-/**
- * Disable thread dispatch (dw5.19) and enable the HiZ op.
- */
-static void
-gen8_blorp_emit_wm_state(struct brw_context *brw)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_WM << 16 | (2 - 2));
-   OUT_BATCH(GEN7_WM_LINE_AA_WIDTH_1_0 |
-             GEN7_WM_LINE_END_CAP_AA_WIDTH_0_5 |
-             GEN7_WM_POINT_RASTRULE_UPPER_RIGHT);
-   ADVANCE_BATCH();
-}
-
-/**
- * 3DSTATE_PS
- *
- * Pixel shader dispatch is disabled above in 3DSTATE_WM, dw1.29. Despite
- * that, thread dispatch info must still be specified.
- *     - Maximum Number of Threads (dw4.24:31) must be nonzero, as the
- *       valid range for this field is [0x3, 0x2f].
- *     - A dispatch mode must be given; that is, at least one of the
- *       "N Pixel Dispatch Enable" (N=8,16,32) fields must be set. This was
- *       discovered through simulator error messages.
- */
-static void
-gen8_blorp_emit_ps_config(struct brw_context *brw,
-                          const struct brw_blorp_params *params)
-{
-   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
-   uint32_t dw3, dw5, dw6, dw7;
-
-   dw3 = dw5 = dw6 = dw7 = 0;
-   dw3 |= GEN7_PS_VECTOR_MASK_ENABLE;
-
-   dw6 |= GEN7_PS_16_DISPATCH_ENABLE;
-
-   if (params->src.mt) {
-      dw3 |= 1 << GEN7_PS_SAMPLER_COUNT_SHIFT; /* Up to 4 samplers */
-      dw3 |= 2 << GEN7_PS_BINDING_TABLE_ENTRY_COUNT_SHIFT; /* Two surfaces */
-   } else {
-      dw3 |= 1 << GEN7_PS_BINDING_TABLE_ENTRY_COUNT_SHIFT; /* One surface */
-   }
-
-   dw6 |= GEN7_PS_PUSH_CONSTANT_ENABLE;
-   dw7 |= prog_data->first_curbe_grf << GEN7_PS_DISPATCH_START_GRF_SHIFT_0;
-
-   /* 3DSTATE_PS expects the number of threads per PSD, which is always 64;
-    * it implicitly scales for different GT levels (which have some # of PSDs).
-    *
-    * In Gen8 the format is U8-2 whereas in Gen9 it is U8-1.
-    */
-   if (brw->gen >= 9)
-      dw6 |= (64 - 1) << HSW_PS_MAX_THREADS_SHIFT;
-   else
-      dw6 |= (64 - 2) << HSW_PS_MAX_THREADS_SHIFT;
-
-   dw6 |= GEN7_PS_POSOFFSET_NONE;
-   dw6 |= params->fast_clear_op;
-
-   BEGIN_BATCH(12);
-   OUT_BATCH(_3DSTATE_PS << 16 | (12 - 2));
-   OUT_BATCH(params->wm_prog_kernel);
-   OUT_BATCH(0);
-   OUT_BATCH(dw3);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(dw6);
-   OUT_BATCH(dw7);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_ps_blend(struct brw_context *brw)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_PS_BLEND << 16 | (2 - 2));
-   OUT_BATCH(GEN8_PS_BLEND_HAS_WRITEABLE_RT);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_ps_extra(struct brw_context *brw,
-                         const struct brw_blorp_params *params)
-{
-   const struct brw_blorp_prog_data *prog_data = params->wm_prog_data;
-   uint32_t dw1 = 0;
-
-   dw1 |= GEN8_PSX_PIXEL_SHADER_VALID;
-
-   if (params->src.mt) {
-      dw1 |= GEN8_PSX_KILL_ENABLE;
-      dw1 |= GEN8_PSX_ATTRIBUTE_ENABLE;
-   }
-
-   if (params->dst.num_samples > 1 && prog_data &&
-       prog_data->persample_msaa_dispatch)
-      dw1 |= GEN8_PSX_SHADER_IS_PER_SAMPLE;
-
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_PS_EXTRA << 16 | (2 - 2));
-   OUT_BATCH(dw1);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_depth_disable(struct brw_context *brw)
-{
-   /* Skip repeated NULL depth/stencil emits (think 2D rendering). */
-   if (brw->no_depth_or_stencil)
-      return;
-
-   brw_emit_depth_stall_flushes(brw);
-
-   BEGIN_BATCH(8);
-   OUT_BATCH(GEN7_3DSTATE_DEPTH_BUFFER << 16 | (8 - 2));
-   OUT_BATCH((BRW_DEPTHFORMAT_D32_FLOAT << 18) | (BRW_SURFACE_NULL << 29));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(5);
-   OUT_BATCH(GEN7_3DSTATE_HIER_DEPTH_BUFFER << 16 | (5 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-
-   BEGIN_BATCH(5);
-   OUT_BATCH(GEN7_3DSTATE_STENCIL_BUFFER << 16 | (5 - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_vf_topology(struct brw_context *brw)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_VF_TOPOLOGY << 16 | (2 - 2));
-   OUT_BATCH(_3DPRIM_RECTLIST);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_vf_sys_gen_vals_state(struct brw_context *brw)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_VF_SGVS << 16 | (2 - 2));
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_vf_instancing_state(struct brw_context *brw,
-                                    unsigned num_elems)
-{
-   for (unsigned i = 0; i < num_elems; ++i) {
-      BEGIN_BATCH(3);
-      OUT_BATCH(_3DSTATE_VF_INSTANCING << 16 | (3 - 2));
-      OUT_BATCH(i);
-      OUT_BATCH(0);
-      ADVANCE_BATCH();
-   }
-}
-
-static void
-gen8_blorp_emit_vf_state(struct brw_context *brw)
-{
-   BEGIN_BATCH(2);
-   OUT_BATCH(_3DSTATE_VF << 16 | (2 - 2));
-   OUT_BATCH(0);
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_depth_stencil_state(struct brw_context *brw,
-                                    const struct brw_blorp_params *params)
-{
-   const unsigned pkt_len = brw->gen >= 9 ? 4 : 3;
-
-   BEGIN_BATCH(pkt_len);
-   OUT_BATCH(_3DSTATE_WM_DEPTH_STENCIL << 16 | (pkt_len - 2));
-   OUT_BATCH(0);
-   OUT_BATCH(0);
-   if (pkt_len > 3) {
-      OUT_BATCH(0);
-   }
-   ADVANCE_BATCH();
-}
-
-static void
-gen8_blorp_emit_constant_ps(struct brw_context *brw,
-                            uint32_t wm_push_const_offset)
-{
-   const int dwords = brw->gen >= 8 ? 11 : 7;
-   BEGIN_BATCH(dwords);
-   OUT_BATCH(_3DSTATE_CONSTANT_PS << 16 | (dwords - 2));
-
-   if (brw->gen >= 9) {
-      OUT_BATCH(0);
-      OUT_BATCH(BRW_BLORP_NUM_PUSH_CONST_REGS);
-   } else {
-      OUT_BATCH(BRW_BLORP_NUM_PUSH_CONST_REGS);
-      OUT_BATCH(0);
-   }
-
-   if (brw->gen >= 9) {
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_RELOC64(brw->batch.bo, I915_GEM_DOMAIN_RENDER, 0,
-                  wm_push_const_offset);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-   } else {
-      OUT_BATCH(wm_push_const_offset);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-      OUT_BATCH(0);
-   }
-
-   ADVANCE_BATCH();
-}
-
-static uint32_t
-gen8_blorp_emit_surface_states(struct brw_context *brw,
-                               const struct brw_blorp_params *params)
-{
-   uint32_t wm_surf_offset_renderbuffer;
-   uint32_t wm_surf_offset_texture = 0;
-
-   intel_miptree_used_for_rendering(params->dst.mt);
-
-   wm_surf_offset_renderbuffer =
-      gen8_blorp_emit_surface_state(brw, &params->dst,
-                                    I915_GEM_DOMAIN_RENDER,
-                                    I915_GEM_DOMAIN_RENDER,
-                                    true /* is_render_target */);
-   if (params->src.mt) {
-      const struct brw_blorp_surface_info *surface = &params->src;
-      struct intel_mipmap_tree *mt = surface->mt;
-
-      /* Textures are always sampled as 2D. */
-      const bool is_cube = mt->target == GL_TEXTURE_CUBE_MAP_ARRAY ||
-                           mt->target == GL_TEXTURE_CUBE_MAP;
-      const unsigned depth = (is_cube ? 6 : 1) * mt->logical_depth0;
-      const GLenum target = is_cube ? GL_TEXTURE_2D_ARRAY : mt->target;
-      const unsigned max_level = surface->level + mt->last_level + 1;
-      const unsigned layer = mt->target != GL_TEXTURE_3D ?
-                                surface->layer / MAX2(mt->num_samples, 1) : 0;
-
-      brw->vtbl.emit_texture_surface_state(brw, mt, target,
-                                           layer, layer + depth,
-                                           surface->level, max_level,
-                                           surface->brw_surfaceformat,
-                                           surface->swizzle,
-                                           &wm_surf_offset_texture,
-                                           -1, false, false);
-   }
-
-   return gen6_blorp_emit_binding_table(brw,
-                                        wm_surf_offset_renderbuffer,
-                                        wm_surf_offset_texture);
-}
-
-/**
- * \copydoc gen6_blorp_exec()
- */
-void
-gen8_blorp_exec(struct brw_context *brw, const struct brw_blorp_params *params)
-{
-   uint32_t wm_bind_bo_offset = 0;
-
-   gen8_upload_state_base_address(brw);
-   gen7_blorp_emit_cc_viewport(brw);
-   gen7_l3_state.emit(brw);
-
-   gen7_blorp_emit_urb_config(brw);
-
-   const uint32_t cc_blend_state_offset =
-      gen8_blorp_emit_blend_state(brw, params);
-   gen7_blorp_emit_blend_state_pointer(brw, cc_blend_state_offset);
-
-   const uint32_t cc_state_offset = gen6_blorp_emit_cc_state(brw);
-   gen7_blorp_emit_cc_state_pointer(brw, cc_state_offset);
-
-   gen8_blorp_emit_disable_constant_state(brw, _3DSTATE_CONSTANT_VS);
-   gen8_blorp_emit_disable_constant_state(brw, _3DSTATE_CONSTANT_HS);
-   gen8_blorp_emit_disable_constant_state(brw, _3DSTATE_CONSTANT_DS);
-   gen8_blorp_emit_disable_constant_state(brw, _3DSTATE_CONSTANT_GS);
-
-   const uint32_t wm_push_const_offset =
-      gen6_blorp_emit_wm_constants(brw, params);
-   gen8_blorp_emit_constant_ps(brw, wm_push_const_offset);
-   wm_bind_bo_offset = gen8_blorp_emit_surface_states(brw, params);
-
-   gen8_blorp_emit_disable_binding_table(brw,
-                                         _3DSTATE_BINDING_TABLE_POINTERS_VS);
-   gen8_blorp_emit_disable_binding_table(brw,
-                                         _3DSTATE_BINDING_TABLE_POINTERS_HS);
-   gen8_blorp_emit_disable_binding_table(brw,
-                                         _3DSTATE_BINDING_TABLE_POINTERS_DS);
-   gen8_blorp_emit_disable_binding_table(brw,
-                                         _3DSTATE_BINDING_TABLE_POINTERS_GS);
-
-   gen7_blorp_emit_binding_table_pointers_ps(brw, wm_bind_bo_offset);
-
-   if (params->src.mt) {
-      const uint32_t sampler_offset =
-         gen6_blorp_emit_sampler_state(brw, BRW_MAPFILTER_LINEAR, 0, true);
-      gen7_blorp_emit_sampler_state_pointers_ps(brw, sampler_offset);
-   }
-
-   gen8_emit_3dstate_multisample(brw, params->dst.num_samples);
-   gen6_emit_3dstate_sample_mask(brw,
-                                 params->dst.num_samples > 1 ?
-                                    (1 << params->dst.num_samples) - 1 : 1);
-
-   gen8_disable_stages.emit(brw);
-   gen8_blorp_emit_vs_disable(brw);
-   gen8_blorp_emit_hs_disable(brw);
-   gen7_blorp_emit_te_disable(brw);
-   gen8_blorp_emit_ds_disable(brw);
-   gen8_blorp_emit_gs_disable(brw);
-
-   gen8_blorp_emit_streamout_disable(brw);
-   gen6_blorp_emit_clip_disable(brw);
-   gen8_blorp_emit_raster_state(brw);
-   gen8_blorp_emit_sbe_state(brw, params);
-   gen8_blorp_emit_sf_config(brw);
-
-   gen8_blorp_emit_ps_blend(brw);
-   gen8_blorp_emit_ps_extra(brw, params);
-
-   gen8_blorp_emit_ps_config(brw, params);
-
-   gen8_blorp_emit_depth_stencil_state(brw, params);
-   gen8_blorp_emit_wm_state(brw);
-
-   gen8_blorp_emit_depth_disable(brw);
-   gen7_blorp_emit_clear_params(brw, params);
-   gen6_blorp_emit_drawing_rectangle(brw, params);
-   gen8_blorp_emit_vf_topology(brw);
-   gen8_blorp_emit_vf_sys_gen_vals_state(brw);
-   gen6_blorp_emit_vertices(brw, params);
-   gen8_blorp_emit_vf_instancing_state(brw, 2);
-   gen8_blorp_emit_vf_state(brw);
-   gen7_blorp_emit_primitive(brw, params);
-
-   if (brw->gen < 9)
-      gen8_write_pma_stall_bits(brw, 0);
-}
-- 
2.5.0.400.gff86faf



More information about the mesa-dev mailing list