Merge remote branch 'origin/master' into lp-binning

Conflicts:
	Makefile
	src/gallium/auxiliary/util/u_surface.c
	src/gallium/drivers/llvmpipe/lp_flush.c
	src/gallium/drivers/llvmpipe/lp_setup.c
	src/gallium/drivers/llvmpipe/lp_state_derived.c
	src/gallium/drivers/llvmpipe/lp_state_fs.c
	src/gallium/drivers/llvmpipe/lp_state_surface.c
	src/gallium/drivers/llvmpipe/lp_tex_cache.c
	src/gallium/drivers/llvmpipe/lp_texture.c
	src/gallium/drivers/llvmpipe/lp_tile_cache.c
	src/mesa/state_tracker/st_cb_condrender.c

22 files changed, 285 insertions, 699 deletions

diff --git a/src/mesa/drivers/dri/intel/intel_batchbuffer.c b/src/mesa/drivers/dri/intel/intel_batchbuffer.c
index 3a4b21a844..ae0f8a16f9 100644
--- a/src/mesa/drivers/dri/intel/intel_batchbuffer.c
+++ b/src/mesa/drivers/dri/intel/intel_batchbuffer.c
@@ -32,44 +32,6 @@
 #include "intel_bufmgr.h"
 #include "intel_buffers.h"
 
-/* Relocations in kernel space:
- *    - pass dma buffer seperately
- *    - memory manager knows how to patch
- *    - pass list of dependent buffers
- *    - pass relocation list
- *
- * Either:
- *    - get back an offset for buffer to fire
- *    - memory manager knows how to fire buffer
- *
- * Really want the buffer to be AGP and pinned.
- *
- */
-
-/* Cliprect fence: The highest fence protecting a dma buffer
- * containing explicit cliprect information.  Like the old drawable
- * lock but irq-driven.  X server must wait for this fence to expire
- * before changing cliprects [and then doing sw rendering?].  For
- * other dma buffers, the scheduler will grab current cliprect info
- * and mix into buffer.  X server must hold the lock while changing
- * cliprects???  Make per-drawable.  Need cliprects in shared memory
- * -- beats storing them with every cmd buffer in the queue.
- *
- * ==> X server must wait for this fence to expire before touching the
- * framebuffer with new cliprects.
- *
- * ==> Cliprect-dependent buffers associated with a
- * cliprect-timestamp.  All of the buffers associated with a timestamp
- * must go to hardware before any buffer with a newer timestamp.
- *
- * ==> Dma should be queued per-drawable for correct X/GL
- * synchronization.  Or can fences be used for this?
- *
- * Applies to: Blit operations, metaops, X server operations -- X
- * server automatically waits on its own dma to complete before
- * modifying cliprects ???
- */
-
 void
 intel_batchbuffer_reset(struct intel_batchbuffer *batch)
 {
@@ -167,7 +129,8 @@ _intel_batchbuffer_flush(struct intel_batchbuffer *batch, const char *file,
    struct intel_context *intel = batch->intel;
    GLuint used = batch->ptr - batch->map;
 
-   if (intel->first_post_swapbuffers_batch == NULL) {
+   if (!intel->using_dri2_swapbuffers &&
+       intel->first_post_swapbuffers_batch == NULL) {
       intel->first_post_swapbuffers_batch = intel->batch->buf;
       drm_intel_bo_reference(intel->first_post_swapbuffers_batch);
    }
diff --git a/src/mesa/drivers/dri/intel/intel_blit.c b/src/mesa/drivers/dri/intel/intel_blit.c
index 55bee0084c..31d4f3d421 100644
--- a/src/mesa/drivers/dri/intel/intel_blit.c
+++ b/src/mesa/drivers/dri/intel/intel_blit.c
@@ -38,7 +38,6 @@
 #include "intel_reg.h"
 #include "intel_regions.h"
 #include "intel_batchbuffer.h"
-#include "intel_chipset.h"
 
 #define FILE_DEBUG_FLAG DEBUG_BLIT
 
@@ -103,7 +102,7 @@ intelEmitCopyBlit(struct intel_context *intel,
 	 return GL_FALSE;
    }
 
-   /* do space/cliprects check before going any further */
+   /* do space check before going any further */
    do {
        aper_array[0] = intel->batch->buf;
        aper_array[1] = dst_buffer;
@@ -213,10 +212,8 @@ intelClearWithBlit(GLcontext *ctx, GLbitfield mask)
    struct intel_context *intel = intel_context(ctx);
    struct gl_framebuffer *fb = ctx->DrawBuffer;
    GLuint clear_depth;
-   GLbitfield skipBuffers = 0;
-   unsigned int num_cliprects;
-   struct drm_clip_rect *cliprects;
-   int x_off, y_off;
+   GLboolean all;
+   GLint cx, cy, cw, ch;
    BATCH_LOCALS;
 
    /*
@@ -230,183 +227,134 @@ intelClearWithBlit(GLcontext *ctx, GLbitfield mask)
       clear_depth |= (ctx->Stencil.Clear & 0xff) << 24;
    }
 
-   /* If clearing both depth and stencil, skip BUFFER_BIT_STENCIL in
-    * the loop below.
-    */
-   if ((mask & BUFFER_BIT_DEPTH) && (mask & BUFFER_BIT_STENCIL)) {
-      skipBuffers = BUFFER_BIT_STENCIL;
-   }
-
-   intel_get_cliprects(intel, &cliprects, &num_cliprects, &x_off, &y_off);
-   if (num_cliprects) {
-      GLint cx, cy, cw, ch;
-      drm_clip_rect_t clear;
-      int i;
-
-      /* Get clear bounds after locking */
-      cx = fb->_Xmin;
+   cx = fb->_Xmin;
+   if (fb->Name == 0)
+      cy = ctx->DrawBuffer->Height - fb->_Ymax;
+   else
       cy = fb->_Ymin;
-      cw = fb->_Xmax - cx;
-      ch = fb->_Ymax - cy;
-
-      if (fb->Name == 0) {
-         /* clearing a window */
-
-         /* flip top to bottom */
-         clear.x1 = cx + x_off;
-         clear.y1 = intel->driDrawable->y + intel->driDrawable->h - cy - ch;
-         clear.x2 = clear.x1 + cw;
-         clear.y2 = clear.y1 + ch;
-      }
-      else {
-         /* clearing FBO */
-         assert(num_cliprects == 1);
-         assert(cliprects == &intel->fboRect);
-         clear.x1 = cx;
-         clear.y1 = cy;
-         clear.x2 = clear.x1 + cw;
-         clear.y2 = clear.y1 + ch;
-         /* no change to mask */
+   cw = fb->_Xmax - fb->_Xmin;
+   ch = fb->_Ymax - fb->_Ymin;
+
+   if (cw == 0 || ch == 0)
+      return;
+
+   GLuint buf;
+   all = (cw == fb->Width && ch == fb->Height);
+
+   /* Loop over all renderbuffers */
+   for (buf = 0; buf < BUFFER_COUNT && mask; buf++) {
+      const GLbitfield bufBit = 1 << buf;
+      struct intel_renderbuffer *irb;
+      drm_intel_bo *write_buffer;
+      int x1, y1, x2, y2;
+      uint32_t clear_val;
+      uint32_t BR13, CMD;
+      int pitch, cpp;
+
+      if (!(mask & bufBit))
+	 continue;
+
+      /* OK, clear this renderbuffer */
+      irb = intel_get_renderbuffer(fb, buf);
+      write_buffer = intel_region_buffer(intel, irb->region,
+					 all ? INTEL_WRITE_FULL :
+					 INTEL_WRITE_PART);
+      x1 = cx + irb->region->draw_x;
+      y1 = cy + irb->region->draw_y;
+      x2 = cx + cw + irb->region->draw_x;
+      y2 = cy + ch + irb->region->draw_y;
+
+      pitch = irb->region->pitch;
+      cpp = irb->region->cpp;
+
+      DBG("%s dst:buf(%p)/%d %d,%d sz:%dx%d\n",
+	  __FUNCTION__,
+	  irb->region->buffer, (pitch * cpp),
+	  x1, y1, x2 - x1, y2 - y1);
+
+      BR13 = 0xf0 << 16;
+      CMD = XY_COLOR_BLT_CMD;
+
+      /* Setup the blit command */
+      if (cpp == 4) {
+	 BR13 |= BR13_8888;
+	 if (buf == BUFFER_DEPTH || buf == BUFFER_STENCIL) {
+	    if (mask & BUFFER_BIT_DEPTH)
+	       CMD |= XY_BLT_WRITE_RGB;
+	    if (mask & BUFFER_BIT_STENCIL)
+	       CMD |= XY_BLT_WRITE_ALPHA;
+	 } else {
+	    /* clearing RGBA */
+	    CMD |= XY_BLT_WRITE_ALPHA | XY_BLT_WRITE_RGB;
+	 }
+      } else {
+	 ASSERT(cpp == 2);
+	 BR13 |= BR13_565;
       }
 
-      for (i = 0; i < num_cliprects; i++) {
-         const drm_clip_rect_t *box = &cliprects[i];
-         drm_clip_rect_t b;
-         GLuint buf;
-         GLuint clearMask = mask;      /* use copy, since we modify it below */
-         GLboolean all = (cw == fb->Width && ch == fb->Height);
-
-         if (!all) {
-            intel_intersect_cliprects(&b, &clear, box);
-         }
-         else {
-            b = *box;
-         }
-
-         if (b.x1 >= b.x2 || b.y1 >= b.y2)
-            continue;
-
-         if (0)
-            _mesa_printf("clear %d,%d..%d,%d, mask %x\n",
-                         b.x1, b.y1, b.x2, b.y2, mask);
-
-         /* Loop over all renderbuffers */
-         for (buf = 0; buf < BUFFER_COUNT && clearMask; buf++) {
-            const GLbitfield bufBit = 1 << buf;
-            if ((clearMask & bufBit) && !(bufBit & skipBuffers)) {
-               /* OK, clear this renderbuffer */
-	       struct intel_renderbuffer *irb = intel_get_renderbuffer(fb, buf);
-               dri_bo *write_buffer =
-                  intel_region_buffer(intel, irb->region,
-                                      all ? INTEL_WRITE_FULL :
-                                      INTEL_WRITE_PART);
-	       int x1 = b.x1 + irb->region->draw_x;
-	       int y1 = b.y1 + irb->region->draw_y;
-	       int x2 = b.x2 + irb->region->draw_x;
-	       int y2 = b.y2 + irb->region->draw_y;
-
-               GLuint clearVal;
-               GLint pitch, cpp;
-               GLuint BR13, CMD;
-
-               pitch = irb->region->pitch;
-               cpp = irb->region->cpp;
-
-               DBG("%s dst:buf(%p)/%d %d,%d sz:%dx%d\n",
-                   __FUNCTION__,
-                   irb->region->buffer, (pitch * cpp),
-                   x1, y1, x2 - x1, y2 - y1);
-
-	       BR13 = 0xf0 << 16;
-	       CMD = XY_COLOR_BLT_CMD;
-
-               /* Setup the blit command */
-               if (cpp == 4) {
-                  BR13 |= BR13_8888;
-                  if (buf == BUFFER_DEPTH || buf == BUFFER_STENCIL) {
-                     if (clearMask & BUFFER_BIT_DEPTH)
-                        CMD |= XY_BLT_WRITE_RGB;
-                     if (clearMask & BUFFER_BIT_STENCIL)
-                        CMD |= XY_BLT_WRITE_ALPHA;
-                  }
-                  else {
-                     /* clearing RGBA */
-                     CMD |= XY_BLT_WRITE_ALPHA | XY_BLT_WRITE_RGB;
-                  }
-               }
-               else {
-                  ASSERT(cpp == 2);
-                  BR13 |= BR13_565;
-               }
-
-	       assert(irb->region->tiling != I915_TILING_Y);
+      assert(irb->region->tiling != I915_TILING_Y);
 
 #ifndef I915
-	       if (irb->region->tiling != I915_TILING_NONE) {
-		  CMD |= XY_DST_TILED;
-		  pitch /= 4;
-	       }
+      if (irb->region->tiling != I915_TILING_NONE) {
+	 CMD |= XY_DST_TILED;
+	 pitch /= 4;
+      }
 #endif
-	       BR13 |= (pitch * cpp);
-
-               if (buf == BUFFER_DEPTH || buf == BUFFER_STENCIL) {
-                  clearVal = clear_depth;
-               }
-               else {
-		  uint8_t clear[4];
-		  GLclampf *color = ctx->Color.ClearColor;
-
-		  CLAMPED_FLOAT_TO_UBYTE(clear[0], color[0]);
-		  CLAMPED_FLOAT_TO_UBYTE(clear[1], color[1]);
-		  CLAMPED_FLOAT_TO_UBYTE(clear[2], color[2]);
-		  CLAMPED_FLOAT_TO_UBYTE(clear[3], color[3]);
-
-		  switch (irb->Base.Format) {
-		  case MESA_FORMAT_ARGB8888:
-		  case MESA_FORMAT_XRGB8888:
-		     clearVal = PACK_COLOR_8888(clear[3], clear[0],
-						clear[1], clear[2]);
-		     break;
-		  case MESA_FORMAT_RGB565:
-		     clearVal = PACK_COLOR_565(clear[0], clear[1], clear[2]);
-		     break;
-		  case MESA_FORMAT_ARGB4444:
-		     clearVal = PACK_COLOR_4444(clear[3], clear[0],
-						clear[1], clear[2]);
-		     break;
-		  case MESA_FORMAT_ARGB1555:
-		     clearVal = PACK_COLOR_1555(clear[3], clear[0],
-						clear[1], clear[2]);
-		     break;
-		  default:
-		     _mesa_problem(ctx, "Unexpected renderbuffer format: %d\n",
-				   irb->Base.Format);
-		     clearVal = 0;
-		  }
-	       }
-
-               /*
-                  _mesa_debug(ctx, "hardware blit clear buf %d rb id %d\n",
-                  buf, irb->Base.Name);
-                */
-
-               assert(x1 < x2);
-               assert(y1 < y2);
-
-               BEGIN_BATCH(6);
-               OUT_BATCH(CMD);
-               OUT_BATCH(BR13);
-               OUT_BATCH((y1 << 16) | x1);
-               OUT_BATCH((y2 << 16) | x2);
-               OUT_RELOC(write_buffer,
-			 I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
-                         0);
-               OUT_BATCH(clearVal);
-               ADVANCE_BATCH();
-               clearMask &= ~bufBit;    /* turn off bit, for faster loop exit */
-            }
-         }
+      BR13 |= (pitch * cpp);
+
+      if (buf == BUFFER_DEPTH || buf == BUFFER_STENCIL) {
+	 clear_val = clear_depth;
+      } else {
+	 uint8_t clear[4];
+	 GLclampf *color = ctx->Color.ClearColor;
+
+	 CLAMPED_FLOAT_TO_UBYTE(clear[0], color[0]);
+	 CLAMPED_FLOAT_TO_UBYTE(clear[1], color[1]);
+	 CLAMPED_FLOAT_TO_UBYTE(clear[2], color[2]);
+	 CLAMPED_FLOAT_TO_UBYTE(clear[3], color[3]);
+
+	 switch (irb->Base.Format) {
+	 case MESA_FORMAT_ARGB8888:
+	 case MESA_FORMAT_XRGB8888:
+	    clear_val = PACK_COLOR_8888(clear[3], clear[0],
+					clear[1], clear[2]);
+	    break;
+	 case MESA_FORMAT_RGB565:
+	    clear_val = PACK_COLOR_565(clear[0], clear[1], clear[2]);
+	    break;
+	 case MESA_FORMAT_ARGB4444:
+	    clear_val = PACK_COLOR_4444(clear[3], clear[0],
+					clear[1], clear[2]);
+	    break;
+	 case MESA_FORMAT_ARGB1555:
+	    clear_val = PACK_COLOR_1555(clear[3], clear[0],
+					clear[1], clear[2]);
+	    break;
+	 default:
+	    _mesa_problem(ctx, "Unexpected renderbuffer format: %d\n",
+			  irb->Base.Format);
+	    clear_val = 0;
+	 }
       }
+
+      assert(x1 < x2);
+      assert(y1 < y2);
+
+      BEGIN_BATCH(6);
+      OUT_BATCH(CMD);
+      OUT_BATCH(BR13);
+      OUT_BATCH((y1 << 16) | x1);
+      OUT_BATCH((y2 << 16) | x2);
+      OUT_RELOC(write_buffer,
+		I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
+		0);
+      OUT_BATCH(clear_val);
+      ADVANCE_BATCH();
+
+      if (buf == BUFFER_DEPTH || buf == BUFFER_STENCIL)
+	 mask &= ~(BUFFER_BIT_DEPTH | BUFFER_BIT_STENCIL);
+      else
+	 mask &= ~bufBit;    /* turn off bit, for faster loop exit */
    }
 }
 
diff --git a/src/mesa/drivers/dri/intel/intel_buffers.c b/src/mesa/drivers/dri/intel/intel_buffers.c
index 7c4b79f743..2f03e45005 100644
--- a/src/mesa/drivers/dri/intel/intel_buffers.c
+++ b/src/mesa/drivers/dri/intel/intel_buffers.c
@@ -28,45 +28,7 @@
 #include "intel_context.h"
 #include "intel_buffers.h"
 #include "intel_fbo.h"
-#include "intel_regions.h"
-#include "intel_batchbuffer.h"
 #include "main/framebuffer.h"
-#include "drirenderbuffer.h"
-
-
-/**
- * XXX move this into a new dri/common/cliprects.c file.
- */
-GLboolean
-intel_intersect_cliprects(drm_clip_rect_t * dst,
-                          const drm_clip_rect_t * a,
-                          const drm_clip_rect_t * b)
-{
-   GLint bx = b->x1;
-   GLint by = b->y1;
-   GLint bw = b->x2 - bx;
-   GLint bh = b->y2 - by;
-
-   if (bx < a->x1)
-      bw -= a->x1 - bx, bx = a->x1;
-   if (by < a->y1)
-      bh -= a->y1 - by, by = a->y1;
-   if (bx + bw > a->x2)
-      bw = a->x2 - bx;
-   if (by + bh > a->y2)
-      bh = a->y2 - by;
-   if (bw <= 0)
-      return GL_FALSE;
-   if (bh <= 0)
-      return GL_FALSE;
-
-   dst->x1 = bx;
-   dst->y1 = by;
-   dst->x2 = bx + bw;
-   dst->y2 = by + bh;
-
-   return GL_TRUE;
-}
 
 /**
  * Return pointer to current color drawing region, or NULL.
@@ -96,24 +58,6 @@ intel_readbuf_region(struct intel_context *intel)
       return NULL;
 }
 
-void
-intel_get_cliprects(struct intel_context *intel,
-		    struct drm_clip_rect **cliprects,
-		    unsigned int *num_cliprects,
-		    int *x_off, int *y_off)
-{
-   intel->fboRect.x1 = 0;
-   intel->fboRect.y1 = 0;
-   intel->fboRect.x2 = intel->ctx.DrawBuffer->Width;
-   intel->fboRect.y2 = intel->ctx.DrawBuffer->Height;
-
-   *cliprects = &intel->fboRect;
-   *num_cliprects = 1;
-   *x_off = 0;
-   *y_off = 0;
-}
-
-
 /**
  * Check if we're about to draw into the front color buffer.
  * If so, set the intel->front_buffer_dirty field to true.
diff --git a/src/mesa/drivers/dri/intel/intel_buffers.h b/src/mesa/drivers/dri/intel/intel_buffers.h
index d7800f2ca2..abb86aade6 100644
--- a/src/mesa/drivers/dri/intel/intel_buffers.h
+++ b/src/mesa/drivers/dri/intel/intel_buffers.h
@@ -35,12 +35,6 @@
 struct intel_context;
 struct intel_framebuffer;
 
-
-extern GLboolean
-intel_intersect_cliprects(drm_clip_rect_t * dest,
-                          const drm_clip_rect_t * a,
-                          const drm_clip_rect_t * b);
-
 extern struct intel_region *intel_readbuf_region(struct intel_context *intel);
 
 extern struct intel_region *intel_drawbuf_region(struct intel_context *intel);
diff --git a/src/mesa/drivers/dri/intel/intel_clear.c b/src/mesa/drivers/dri/intel/intel_clear.c
index 956f2339ff..ca78681538 100644
--- a/src/mesa/drivers/dri/intel/intel_clear.c
+++ b/src/mesa/drivers/dri/intel/intel_clear.c
@@ -33,12 +33,9 @@
 
 #include "intel_context.h"
 #include "intel_blit.h"
-#include "intel_chipset.h"
 #include "intel_clear.h"
 #include "intel_fbo.h"
-#include "intel_pixel.h"
 #include "intel_regions.h"
-#include "intel_batchbuffer.h"
 
 #define FILE_DEBUG_FLAG DEBUG_BLIT
 
diff --git a/src/mesa/drivers/dri/intel/intel_context.c b/src/mesa/drivers/dri/intel/intel_context.c
index 3f6634c65a..0adee6131e 100644
--- a/src/mesa/drivers/dri/intel/intel_context.c
+++ b/src/mesa/drivers/dri/intel/intel_context.c
@@ -28,7 +28,6 @@
 
 #include "main/glheader.h"
 #include "main/context.h"
-#include "main/arrayobj.h"
 #include "main/extensions.h"
 #include "main/framebuffer.h"
 #include "main/imports.h"
@@ -52,13 +51,11 @@
 #include "intel_regions.h"
 #include "intel_buffer_objects.h"
 #include "intel_fbo.h"
-#include "intel_decode.h"
 #include "intel_bufmgr.h"
 #include "intel_screen.h"
 
 #include "drirenderbuffer.h"
 #include "utils.h"
-#include "xmlpool.h"            /* for symbolic values of enum-type options */
 
 
 #ifndef INTEL_DEBUG
@@ -522,7 +519,8 @@ intel_glFlush(GLcontext *ctx)
     * and getting our hands on that doesn't seem worth it, so we just us the
     * first batch we emitted after the last swap.
     */
-   if (intel->first_post_swapbuffers_batch != NULL) {
+   if (!intel->using_dri2_swapbuffers &&
+       intel->first_post_swapbuffers_batch != NULL) {
       drm_intel_bo_wait_rendering(intel->first_post_swapbuffers_batch);
       drm_intel_bo_unreference(intel->first_post_swapbuffers_batch);
       intel->first_post_swapbuffers_batch = NULL;
diff --git a/src/mesa/drivers/dri/intel/intel_context.h b/src/mesa/drivers/dri/intel/intel_context.h
index 07207bfbec..36abef470a 100644
--- a/src/mesa/drivers/dri/intel/intel_context.h
+++ b/src/mesa/drivers/dri/intel/intel_context.h
@@ -107,7 +107,6 @@ struct intel_context
       void (*finish_batch) (struct intel_context * intel);
       void (*new_batch) (struct intel_context * intel);
       void (*emit_invarient_state) (struct intel_context * intel);
-      void (*note_fence) (struct intel_context *intel, GLuint fence);
       void (*update_texture_state) (struct intel_context * intel);
 
       void (*render_start) (struct intel_context * intel);
@@ -125,40 +124,6 @@ struct intel_context
       void (*invalidate_state) (struct intel_context *intel,
 				GLuint new_state);
 
-
-      /* Metaops: 
-       */
-      void (*install_meta_state) (struct intel_context * intel);
-      void (*leave_meta_state) (struct intel_context * intel);
-
-      void (*meta_draw_region) (struct intel_context * intel,
-                                struct intel_region * draw_region,
-                                struct intel_region * depth_region);
-
-      void (*meta_color_mask) (struct intel_context * intel, GLboolean);
-
-      void (*meta_stencil_replace) (struct intel_context * intel,
-                                    GLuint mask, GLuint clear);
-
-      void (*meta_depth_replace) (struct intel_context * intel);
-
-      void (*meta_texture_blend_replace) (struct intel_context * intel);
-
-      void (*meta_no_stencil_write) (struct intel_context * intel);
-      void (*meta_no_depth_write) (struct intel_context * intel);
-      void (*meta_no_texture) (struct intel_context * intel);
-
-      void (*meta_import_pixel_state) (struct intel_context * intel);
-      void (*meta_frame_buffer_texture) (struct intel_context *intel,
-					 GLint xoff, GLint yoff);
-
-      GLboolean(*meta_tex_rect_source) (struct intel_context * intel,
-					dri_bo * buffer,
-					GLuint offset,
-					GLuint pitch,
-					GLuint height,
-					GLenum format, GLenum type);
-
       void (*assert_not_dirty) (struct intel_context *intel);
 
       void (*debug_batch)(struct intel_context *intel);
@@ -187,6 +152,7 @@ struct intel_context
    struct intel_batchbuffer *batch;
    drm_intel_bo *first_post_swapbuffers_batch;
    GLboolean no_batch_wrap;
+   GLboolean using_dri2_swapbuffers;
 
    struct
    {
@@ -273,10 +239,6 @@ struct intel_context
 
    GLboolean use_texture_tiling;
    GLboolean use_early_z;
-   drm_clip_rect_t fboRect;     /**< cliprect for FBO rendering */
-
-   drm_clip_rect_t draw_rect;
-   drm_clip_rect_t scissor_rect;
 
    int driFd;
 
diff --git a/src/mesa/drivers/dri/intel/intel_extensions.c b/src/mesa/drivers/dri/intel/intel_extensions.c
index 5ac5ce10af..84c8d013e3 100644
--- a/src/mesa/drivers/dri/intel/intel_extensions.c
+++ b/src/mesa/drivers/dri/intel/intel_extensions.c
@@ -48,6 +48,7 @@
 #define need_GL_EXT_blend_func_separate
 #define need_GL_EXT_blend_minmax
 #define need_GL_EXT_cull_vertex
+#define need_GL_EXT_draw_buffers2
 #define need_GL_EXT_fog_coord
 #define need_GL_EXT_framebuffer_object
 #define need_GL_EXT_framebuffer_blit
@@ -154,12 +155,14 @@ static const struct dri_extension brw_extensions[] = {
    { "GL_ARB_fragment_program_shadow",    NULL },
    { "GL_ARB_fragment_shader",            NULL },
    { "GL_ARB_framebuffer_object",         GL_ARB_framebuffer_object_functions},
+   { "GL_ARB_half_float_vertex",          NULL },
    { "GL_ARB_occlusion_query",            GL_ARB_occlusion_query_functions },
    { "GL_ARB_point_sprite", 		  NULL },
    { "GL_ARB_seamless_cube_map",          NULL },
    { "GL_ARB_shadow",                     NULL },
    { "GL_MESA_texture_signed_rgba",       NULL },
    { "GL_ARB_texture_non_power_of_two",   NULL },
+   { "GL_EXT_draw_buffers2",              GL_EXT_draw_buffers2_functions },
    { "GL_EXT_shadow_funcs",               NULL },
    { "GL_EXT_stencil_two_side",           GL_EXT_stencil_two_side_functions },
    { "GL_EXT_texture_sRGB",		  NULL },
diff --git a/src/mesa/drivers/dri/intel/intel_mipmap_tree.c b/src/mesa/drivers/dri/intel/intel_mipmap_tree.c
index 82e4150c6a..cb5a341050 100644
--- a/src/mesa/drivers/dri/intel/intel_mipmap_tree.c
+++ b/src/mesa/drivers/dri/intel/intel_mipmap_tree.c
@@ -29,7 +29,6 @@
 #include "intel_mipmap_tree.h"
 #include "intel_regions.h"
 #include "intel_tex_layout.h"
-#include "intel_chipset.h"
 #ifndef I915
 #include "brw_state.h"
 #endif
diff --git a/src/mesa/drivers/dri/intel/intel_pixel.c b/src/mesa/drivers/dri/intel/intel_pixel.c
index 5142f3dcd9..cb088e4032 100644
--- a/src/mesa/drivers/dri/intel/intel_pixel.c
+++ b/src/mesa/drivers/dri/intel/intel_pixel.c
@@ -29,14 +29,7 @@
 #include "main/state.h"
 #include "main/bufferobj.h"
 #include "main/context.h"
-#include "main/enable.h"
-#include "main/matrix.h"
-#include "main/texstate.h"
-#include "main/varray.h"
-#include "main/viewport.h"
 #include "swrast/swrast.h"
-#include "shader/arbprogram.h"
-#include "shader/program.h"
 
 #include "intel_context.h"
 #include "intel_pixel.h"
diff --git a/src/mesa/drivers/dri/intel/intel_pixel_bitmap.c b/src/mesa/drivers/dri/intel/intel_pixel_bitmap.c
index 85e5ad2cdd..1e517650b7 100644
--- a/src/mesa/drivers/dri/intel/intel_pixel_bitmap.c
+++ b/src/mesa/drivers/dri/intel/intel_pixel_bitmap.c
@@ -37,7 +37,6 @@
 #include "main/polygon.h"
 #include "main/state.h"
 #include "main/teximage.h"
-#include "main/texenv.h"
 #include "main/texobj.h"
 #include "main/texstate.h"
 #include "main/texparam.h"
@@ -46,7 +45,6 @@
 #include "main/enable.h"
 #include "main/viewport.h"
 #include "shader/arbprogram.h"
-#include "glapi/dispatch.h"
 #include "swrast/swrast.h"
 
 #include "intel_screen.h"
@@ -54,7 +52,6 @@
 #include "intel_batchbuffer.h"
 #include "intel_blit.h"
 #include "intel_regions.h"
-#include "intel_buffer_objects.h"
 #include "intel_buffers.h"
 #include "intel_pixel.h"
 #include "intel_reg.h"
@@ -106,7 +103,7 @@ static void set_bit( GLubyte *dest, GLuint bit )
 }
 
 /* Extract a rectangle's worth of data from the bitmap.  Called
- * per-cliprect.
+ * per chunk of HW-sized bitmap.
  */
 static GLuint get_bitmap_rect(GLsizei width, GLsizei height,
 			      const struct gl_pixelstore_attrib *unpack,
@@ -191,11 +188,12 @@ do_blit_bitmap( GLcontext *ctx,
    GLfloat tmpColor[4];
    GLubyte ubcolor[4];
    GLuint color;
-   unsigned int num_cliprects;
-   drm_clip_rect_t *cliprects;
-   int x_off, y_off;
    GLsizei bitmap_width = width;
    GLsizei bitmap_height = height;
+   GLint px, py;
+   GLuint stipple[32];
+   GLint orig_dstx = dstx;
+   GLint orig_dsty = dsty;
 
    /* Update draw buffer bounds */
    _mesa_update_state(ctx);
@@ -236,90 +234,60 @@ do_blit_bitmap( GLcontext *ctx,
    if (!intel_check_blit_fragment_ops(ctx, tmpColor[3] == 1.0F))
       return GL_FALSE;
 
-   intel_get_cliprects(intel, &cliprects, &num_cliprects, &x_off, &y_off);
-   if (num_cliprects != 0) {
-      GLuint i;
-      GLint orig_dstx = dstx;
-      GLint orig_dsty = dsty;
-
-      /* Clip to buffer bounds and scissor. */
-      if (!_mesa_clip_to_region(fb->_Xmin, fb->_Ymin,
-				fb->_Xmax, fb->_Ymax,
-				&dstx, &dsty, &width, &height))
-            goto out;
-
-      dstx = x_off + dstx;
-      dsty = y_off + y_flip(fb, dsty, height);
-
-      for (i = 0; i < num_cliprects; i++) {
-	 int box_x, box_y, box_w, box_h;
-	 GLint px, py;
-	 GLuint stipple[32];  
-
-	 box_x = dstx;
-	 box_y = dsty;
-	 box_w = width;
-	 box_h = height;
-
-	 /* Clip to drawable cliprect */
-         if (!_mesa_clip_to_region(cliprects[i].x1,
-				   cliprects[i].y1,
-				   cliprects[i].x2,
-				   cliprects[i].y2,
-				   &box_x, &box_y, &box_w, &box_h))
-	    continue;
+   /* Clip to buffer bounds and scissor. */
+   if (!_mesa_clip_to_region(fb->_Xmin, fb->_Ymin,
+			     fb->_Xmax, fb->_Ymax,
+			     &dstx, &dsty, &width, &height))
+      goto out;
+
+   dsty = y_flip(fb, dsty, height);
 
 #define DY 32
 #define DX 32
 
-	 /* Then, finally, chop it all into chunks that can be
-	  * digested by hardware:
+   /* Chop it all into chunks that can be digested by hardware: */
+   for (py = 0; py < height; py += DY) {
+      for (px = 0; px < width; px += DX) {
+	 int h = MIN2(DY, height - py);
+	 int w = MIN2(DX, width - px);
+	 GLuint sz = ALIGN(ALIGN(w,8) * h, 64)/8;
+	 GLenum logic_op = ctx->Color.ColorLogicOpEnabled ?
+	    ctx->Color.LogicOp : GL_COPY;
+
+	 assert(sz <= sizeof(stipple));
+	 memset(stipple, 0, sz);
+
+	 /* May need to adjust this when padding has been introduced in
+	  * sz above:
+	  *
+	  * Have to translate destination coordinates back into source
+	  * coordinates.
 	  */
-	 for (py = 0; py < box_h; py += DY) { 
-	    for (px = 0; px < box_w; px += DX) { 
-	       int h = MIN2(DY, box_h - py);
-	       int w = MIN2(DX, box_w - px); 
-	       GLuint sz = ALIGN(ALIGN(w,8) * h, 64)/8;
-	       GLenum logic_op = ctx->Color.ColorLogicOpEnabled ?
-		  ctx->Color.LogicOp : GL_COPY;
-
-	       assert(sz <= sizeof(stipple));
-	       memset(stipple, 0, sz);
-
-	       /* May need to adjust this when padding has been introduced in
-		* sz above:
-		*
-		* Have to translate destination coordinates back into source
-		* coordinates.
-		*/
-	       if (get_bitmap_rect(bitmap_width, bitmap_height, unpack,
-				   bitmap,
-				   -orig_dstx + (box_x + px - x_off),
-				   -orig_dsty + y_flip(fb,
-						       box_y + py - y_off, h),
-				   w, h,
-				   (GLubyte *)stipple,
-				   8,
-				   fb->Name == 0 ? GL_TRUE : GL_FALSE) == 0)
-		  continue;
-
-	       if (!intelEmitImmediateColorExpandBlit(intel,
-						      dst->cpp,
-						      (GLubyte *)stipple,
-						      sz,
-						      color,
-						      dst->pitch,
-						      dst->buffer,
-						      0,
-						      dst->tiling,
-						      box_x + px,
-						      box_y + py,
-						      w, h,
-						      logic_op)) {
-		  return GL_FALSE;
-	       }
-	    } 
-	 } 
+	 if (get_bitmap_rect(bitmap_width, bitmap_height, unpack,
+			     bitmap,
+			     -orig_dstx + (dstx + px),
+			     -orig_dsty + y_flip(fb, dsty + py, h),
+			     w, h,
+			     (GLubyte *)stipple,
+			     8,
+			     fb->Name == 0 ? GL_TRUE : GL_FALSE) == 0)
+	    continue;
+
+	 if (!intelEmitImmediateColorExpandBlit(intel,
+						dst->cpp,
+						(GLubyte *)stipple,
+						sz,
+						color,
+						dst->pitch,
+						dst->buffer,
+						0,
+						dst->tiling,
+						dstx + px,
+						dsty + py,
+						w, h,
+						logic_op)) {
+	    return GL_FALSE;
+	 }
       }
    }
 out:
diff --git a/src/mesa/drivers/dri/intel/intel_pixel_copy.c b/src/mesa/drivers/dri/intel/intel_pixel_copy.c
index e002516cdd..87c293d8b0 100644
--- a/src/mesa/drivers/dri/intel/intel_pixel_copy.c
+++ b/src/mesa/drivers/dri/intel/intel_pixel_copy.c
@@ -112,9 +112,10 @@ do_blit_copypixels(GLcontext * ctx,
    struct intel_region *src = copypix_src_region(intel, type);
    struct gl_framebuffer *fb = ctx->DrawBuffer;
    struct gl_framebuffer *read_fb = ctx->ReadBuffer;
-   unsigned int num_cliprects;
-   drm_clip_rect_t *cliprects;
-   int x_off, y_off;
+   GLint orig_dstx;
+   GLint orig_dsty;
+   GLint orig_srcx;
+   GLint orig_srcy;
 
    if (type == GL_DEPTH || type == GL_STENCIL) {
       if (INTEL_DEBUG & DEBUG_FALLBACKS)
@@ -135,94 +136,56 @@ do_blit_copypixels(GLcontext * ctx,
    if (!src || !dst)
       return GL_FALSE;
 
-
-
    intelFlush(&intel->ctx);
 
-   intel_get_cliprects(intel, &cliprects, &num_cliprects, &x_off, &y_off);
-   if (num_cliprects != 0) {
-      GLint delta_x;
-      GLint delta_y;
-      GLint orig_dstx;
-      GLint orig_dsty;
-      GLint orig_srcx;
-      GLint orig_srcy;
-      GLuint i;
-
-      /* XXX: We fail to handle different inversion between read and draw framebuffer. */
-
-      /* Clip to destination buffer. */
-      orig_dstx = dstx;
-      orig_dsty = dsty;
-      if (!_mesa_clip_to_region(fb->_Xmin, fb->_Ymin,
-				fb->_Xmax, fb->_Ymax,
-				&dstx, &dsty, &width, &height))
-	 goto out;
-      /* Adjust src coords for our post-clipped destination origin */
-      srcx += dstx - orig_dstx;
-      srcy += dsty - orig_dsty;
-
-      /* Clip to source buffer. */
-      orig_srcx = srcx;
-      orig_srcy = srcy;
-      if (!_mesa_clip_to_region(0, 0,
-				read_fb->Width, read_fb->Height,
-				&srcx, &srcy, &width, &height))
-	 goto out;
-      /* Adjust dst coords for our post-clipped source origin */
-      dstx += srcx - orig_srcx;
-      dsty += srcy - orig_srcy;
-
-      /* Convert from GL to hardware coordinates:
-       */
-      if (fb->Name == 0) {
-	 /* copypixels to a system framebuffer */
-	 dstx = x_off + dstx;
-	 dsty = y_off + (fb->Height - dsty - height);
-      } else {
-	 /* copypixels to a user framebuffer object */
-	 dstx = x_off + dstx;
-	 dsty = y_off + dsty;
-      }
-
-      /* Flip source Y if it's a system framebuffer. */
-      if (read_fb->Name == 0) {
-	 srcx = intel->driReadDrawable->x + srcx;
-	 srcy = intel->driReadDrawable->y + (fb->Height - srcy - height);
-      }
-
-      delta_x = srcx - dstx;
-      delta_y = srcy - dsty;
-      /* Could do slightly more clipping: Eg, take the intersection of
-       * the destination cliprects and the read drawable cliprects
-       *
-       * This code will not overwrite other windows, but will
-       * introduce garbage when copying from obscured window regions.
-       */
-      for (i = 0; i < num_cliprects; i++) {
-	 GLint clip_x = dstx;
-	 GLint clip_y = dsty;
-	 GLint clip_w = width;
-	 GLint clip_h = height;
-
-         if (!_mesa_clip_to_region(cliprects[i].x1, cliprects[i].y1,
-				   cliprects[i].x2, cliprects[i].y2,
-				   &clip_x, &clip_y, &clip_w, &clip_h))
-            continue;
-
-	 if (!intel_region_copy(intel,
-				dst, 0, clip_x, clip_y,
-				src, 0, clip_x + delta_x, clip_y + delta_y,
-				clip_w, clip_h,
-				ctx->Color.ColorLogicOpEnabled ?
-				ctx->Color.LogicOp : GL_COPY)) {
-	    DBG("%s: blit failure\n", __FUNCTION__);
-	    return GL_FALSE;
-	 }
-      }
+   /* XXX: We fail to handle different inversion between read and draw framebuffer. */
+
+   /* Clip to destination buffer. */
+   orig_dstx = dstx;
+   orig_dsty = dsty;
+   if (!_mesa_clip_to_region(fb->_Xmin, fb->_Ymin,
+			     fb->_Xmax, fb->_Ymax,
+			     &dstx, &dsty, &width, &height))
+      goto out;
+   /* Adjust src coords for our post-clipped destination origin */
+   srcx += dstx - orig_dstx;
+   srcy += dsty - orig_dsty;
+
+   /* Clip to source buffer. */
+   orig_srcx = srcx;
+   orig_srcy = srcy;
+   if (!_mesa_clip_to_region(0, 0,
+			     read_fb->Width, read_fb->Height,
+			     &srcx, &srcy, &width, &height))
+      goto out;
+   /* Adjust dst coords for our post-clipped source origin */
+   dstx += srcx - orig_srcx;
+   dsty += srcy - orig_srcy;
+
+   /* Convert from GL to hardware coordinates: */
+   if (fb->Name == 0) {
+      /* copypixels to a system framebuffer */
+      dsty = fb->Height - dsty - height;
+   } else {
+      /* copypixels to a user framebuffer object */
+      dsty = dsty;
    }
-out:
 
+   /* Flip source Y if it's a system framebuffer. */
+   if (read_fb->Name == 0)
+      srcy = fb->Height - srcy - height;
+
+   if (!intel_region_copy(intel,
+			  dst, 0, dstx, dsty,
+			  src, 0, srcx, srcy,
+			  width, height,
+			  ctx->Color.ColorLogicOpEnabled ?
+			  ctx->Color.LogicOp : GL_COPY)) {
+      DBG("%s: blit failure\n", __FUNCTION__);
+      return GL_FALSE;
+   }
+
+out:
    intel_check_front_buffer_rendering(intel);
 
    DBG("%s: success\n", __FUNCTION__);
diff --git a/src/mesa/drivers/dri/intel/intel_pixel_draw.c b/src/mesa/drivers/dri/intel/intel_pixel_draw.c
index b870e9315e..10177228bd 100644
--- a/src/mesa/drivers/dri/intel/intel_pixel_draw.c
+++ b/src/mesa/drivers/dri/intel/intel_pixel_draw.c
@@ -46,10 +46,6 @@
 #include "drivers/common/meta.h"
 
 #include "intel_context.h"
-#include "intel_batchbuffer.h"
-#include "intel_blit.h"
-#include "intel_buffers.h"
-#include "intel_regions.h"
 #include "intel_pixel.h"
 #include "intel_fbo.h"
 
diff --git a/src/mesa/drivers/dri/intel/intel_pixel_read.c b/src/mesa/drivers/dri/intel/intel_pixel_read.c
index 9c0fdc6067..a98e8e16c2 100644
--- a/src/mesa/drivers/dri/intel/intel_pixel_read.c
+++ b/src/mesa/drivers/dri/intel/intel_pixel_read.c
@@ -36,7 +36,6 @@
 
 #include "intel_screen.h"
 #include "intel_context.h"
-#include "intel_batchbuffer.h"
 #include "intel_blit.h"
 #include "intel_buffers.h"
 #include "intel_regions.h"
@@ -65,99 +64,6 @@
  * any case.
  */
 
-
-static GLboolean
-do_texture_readpixels(GLcontext * ctx,
-                      GLint x, GLint y, GLsizei width, GLsizei height,
-                      GLenum format, GLenum type,
-                      const struct gl_pixelstore_attrib *pack,
-                      struct intel_region *dest_region)
-{
-#if 0
-   struct intel_context *intel = intel_context(ctx);
-   intelScreenPrivate *screen = intel->intelScreen;
-   GLint pitch = pack->RowLength ? pack->RowLength : width;
-   __DRIdrawable *dPriv = intel->driDrawable;
-   int textureFormat;
-   GLenum glTextureFormat;
-   int destFormat, depthFormat, destPitch;
-   drm_clip_rect_t tmp;
-
-   if (INTEL_DEBUG & DEBUG_PIXEL)
-      fprintf(stderr, "%s\n", __FUNCTION__);
-
-
-   if (ctx->_ImageTransferState ||
-       pack->SwapBytes || pack->LsbFirst || !pack->Invert) {
-      if (INTEL_DEBUG & DEBUG_PIXEL)
-         fprintf(stderr, "%s: check_color failed\n", __FUNCTION__);
-      return GL_FALSE;
-   }
-
-   intel->vtbl.meta_texrect_source(intel, intel_readbuf_region(intel));
-
-   if (!intel->vtbl.meta_render_dest(intel, dest_region, type, format)) {
-      if (INTEL_DEBUG & DEBUG_PIXEL)
-         fprintf(stderr, "%s: couldn't set dest %s/%s\n",
-                 __FUNCTION__,
-                 _mesa_lookup_enum_by_nr(type),
-                 _mesa_lookup_enum_by_nr(format));
-      return GL_FALSE;
-   }
-
-   if (intel->driDrawable->numClipRects) {
-      intel->vtbl.install_meta_state(intel);
-      intel->vtbl.meta_no_depth_write(intel);
-      intel->vtbl.meta_no_stencil_write(intel);
-
-      if (!driClipRectToFramebuffer(ctx->ReadBuffer, &x, &y, &width, &height)) {
-         SET_STATE(i830, state);
-         if (INTEL_DEBUG & DEBUG_PIXEL)
-            fprintf(stderr, "%s: cliprect failed\n", __FUNCTION__);
-         return GL_TRUE;
-      }
-
-      y = dPriv->h - y - height;
-      x += dPriv->x;
-      y += dPriv->y;
-
-
-      /* Set the frontbuffer up as a large rectangular texture.
-       */
-      intel->vtbl.meta_tex_rect_source(intel, src_region, textureFormat);
-
-
-      intel->vtbl.meta_texture_blend_replace(i830, glTextureFormat);
-
-
-      /* Set the 3d engine to draw into the destination region:
-       */
-
-      intel->vtbl.meta_draw_region(intel, dest_region);
-      intel->vtbl.meta_draw_format(intel, destFormat, depthFormat);     /* ?? */
-
-
-      /* Draw a single quad, no cliprects:
-       */
-      intel->vtbl.meta_disable_cliprects(intel);
-
-      intel->vtbl.draw_quad(intel,
-                            0, width, 0, height,
-                            0x00ff00ff, x, x + width, y, y + height);
-
-      intel->vtbl.leave_meta_state(intel);
-   }
-
-   intel_region_wait_fence(ctx, dest_region);   /* required by GL */
-   return GL_TRUE;
-#endif
-
-   return GL_FALSE;
-}
-
-
-
-
 static GLboolean
 do_blit_readpixels(GLcontext * ctx,
                    GLint x, GLint y, GLsizei width, GLsizei height,
@@ -169,6 +75,9 @@ do_blit_readpixels(GLcontext * ctx,
    struct intel_buffer_object *dst = intel_buffer_object(pack->BufferObj);
    GLuint dst_offset;
    GLuint rowLength;
+   drm_intel_bo *dst_buffer;
+   GLboolean all;
+   GLint dst_x, dst_y;
 
    if (INTEL_DEBUG & DEBUG_PIXEL)
       _mesa_printf("%s\n", __FUNCTION__);
@@ -209,56 +118,42 @@ do_blit_readpixels(GLcontext * ctx,
       return GL_FALSE;
    }
    else {
-      rowLength = -rowLength;
+      if (ctx->ReadBuffer->Name == 0)
+	 rowLength = -rowLength;
    }
 
    dst_offset = (GLintptr) _mesa_image_address(2, pack, pixels, width, height,
 					       format, type, 0, 0, 0);
 
+   if (!_mesa_clip_copytexsubimage(ctx,
+				   &dst_x, &dst_y,
+				   &x, &y,
+				   &width, &height)) {
+      return GL_TRUE;
+   }
 
-   /* Although the blits go on the command buffer, need to do this and
-    * fire with lock held to guarentee cliprects are correct.
-    */
-   intelFlush(&intel->ctx);
-
-   if (intel->driReadDrawable->numClipRects) {
-      GLboolean all = (width * height * src->cpp == dst->Base.Size &&
-                       x == 0 && dst_offset == 0);
-
-      dri_bo *dst_buffer = intel_bufferobj_buffer(intel, dst,
-						  all ? INTEL_WRITE_FULL :
-						  INTEL_WRITE_PART);
-      __DRIdrawable *dPriv = intel->driReadDrawable;
-      int nbox = dPriv->numClipRects;
-      drm_clip_rect_t *box = dPriv->pClipRects;
-      drm_clip_rect_t rect;
-      drm_clip_rect_t src_rect;
-      int i;
-
-      src_rect.x1 = dPriv->x + x;
-      src_rect.y1 = dPriv->y + dPriv->h - (y + height);
-      src_rect.x2 = src_rect.x1 + width;
-      src_rect.y2 = src_rect.y1 + height;
+   all = (width * height * src->cpp == dst->Base.Size &&
+	  x == 0 && dst_offset == 0);
 
+   dst_x = 0;
+   dst_y = 0;
 
+   dst_buffer = intel_bufferobj_buffer(intel, dst,
+					       all ? INTEL_WRITE_FULL :
+					       INTEL_WRITE_PART);
 
-      for (i = 0; i < nbox; i++) {
-         if (!intel_intersect_cliprects(&rect, &src_rect, &box[i]))
-            continue;
+   if (ctx->ReadBuffer->Name == 0)
+      y = ctx->ReadBuffer->Height - (y + height);
 
-         if (!intelEmitCopyBlit(intel,
-				src->cpp,
-				src->pitch, src->buffer, 0, src->tiling,
-				rowLength, dst_buffer, dst_offset, GL_FALSE,
-				rect.x1,
-				rect.y1,
-				rect.x1 - src_rect.x1,
-				rect.y2 - src_rect.y2,
-				rect.x2 - rect.x1, rect.y2 - rect.y1,
-				GL_COPY)) {
-	    return GL_FALSE;
-	 }
-      }
+   if (!intelEmitCopyBlit(intel,
+			  src->cpp,
+			  src->pitch, src->buffer, 0, src->tiling,
+			  rowLength, dst_buffer, dst_offset, GL_FALSE,
+			  x, y,
+			  dst_x, dst_y,
+			  width, height,
+			  GL_COPY)) {
+      return GL_FALSE;
    }
 
    if (INTEL_DEBUG & DEBUG_PIXEL)
@@ -282,15 +177,6 @@ intelReadPixels(GLcontext * ctx,
        (ctx, x, y, width, height, format, type, pack, pixels))
       return;
 
-#ifdef I915
-   if (do_texture_readpixels
-       (ctx, x, y, width, height, format, type, pack, pixels))
-      return;
-#else
-   (void)do_blit_readpixels;
-   (void)do_texture_readpixels;
-#endif
-
    if (INTEL_DEBUG & DEBUG_PIXEL)
       _mesa_printf("%s: fallback to swrast\n", __FUNCTION__);
 
diff --git a/src/mesa/drivers/dri/intel/intel_regions.c b/src/mesa/drivers/dri/intel/intel_regions.c
index 61aefa01b8..301c3df17c 100644
--- a/src/mesa/drivers/dri/intel/intel_regions.c
+++ b/src/mesa/drivers/dri/intel/intel_regions.c
@@ -48,7 +48,6 @@
 #include "intel_buffer_objects.h"
 #include "intel_bufmgr.h"
 #include "intel_batchbuffer.h"
-#include "intel_chipset.h"
 
 #define FILE_DEBUG_FLAG DEBUG_REGION
 
@@ -443,6 +442,7 @@ intel_region_attach_pbo(struct intel_context *intel,
    region->pbo->region = region;
    dri_bo_reference(buffer);
    region->buffer = buffer;
+   region->tiling = I915_TILING_NONE;
 }
 
 
diff --git a/src/mesa/drivers/dri/intel/intel_screen.c b/src/mesa/drivers/dri/intel/intel_screen.c
index e240957197..e8cc202f88 100644
--- a/src/mesa/drivers/dri/intel/intel_screen.c
+++ b/src/mesa/drivers/dri/intel/intel_screen.c
@@ -37,15 +37,11 @@
 #include "intel_buffers.h"
 #include "intel_bufmgr.h"
 #include "intel_chipset.h"
-#include "intel_extensions.h"
 #include "intel_fbo.h"
-#include "intel_regions.h"
 #include "intel_screen.h"
-#include "intel_span.h"
 #include "intel_tex.h"
 
 #include "i915_drm.h"
-#include "i830_dri.h"
 
 #define DRI_CONF_TEXTURE_TILING(def) \
 	DRI_CONF_OPT_BEGIN(texture_tiling, bool, def)		\
@@ -128,8 +124,14 @@ intelDRI2Flush(__DRIdrawable *drawable)
 static void
 intelDRI2FlushInvalidate(__DRIdrawable *drawable)
 {
+   struct intel_context *intel = drawable->driContextPriv->driverPrivate;
+
+   intel->using_dri2_swapbuffers = GL_TRUE;
+
    intelDRI2Flush(drawable);
    drawable->validBuffers = GL_FALSE;
+
+   intel_update_renderbuffers(intel->driContext, drawable);
 }
 
 static const struct __DRI2flushExtensionRec intelFlushExtension = {
diff --git a/src/mesa/drivers/dri/intel/intel_span.c b/src/mesa/drivers/dri/intel/intel_span.c
index 605734d8e5..d925cb9997 100644
--- a/src/mesa/drivers/dri/intel/intel_span.c
+++ b/src/mesa/drivers/dri/intel/intel_span.c
@@ -259,36 +259,20 @@ static uint32_t y_tile_swizzle(struct intel_renderbuffer *irb,
 #define DBG 0
 
 #define LOCAL_VARS							\
-   struct intel_context *intel = intel_context(ctx);			\
    struct intel_renderbuffer *irb = intel_renderbuffer(rb);		\
    const GLint yScale = ctx->DrawBuffer->Name ? 1 : -1;			\
    const GLint yBias = ctx->DrawBuffer->Name ? 0 : irb->Base.Height - 1;\
-   unsigned int num_cliprects;						\
-   struct drm_clip_rect *cliprects;					\
-   int x_off, y_off;							\
+   int minx = 0, miny = 0;						\
+   int maxx = ctx->DrawBuffer->Width;					\
+   int maxy = ctx->DrawBuffer->Height;					\
    int pitch = irb->region->pitch * irb->region->cpp;			\
    void *buf = irb->region->buffer->virtual;				\
    GLuint p;								\
    (void) p;								\
    (void)buf; (void)pitch; /* unused for non-gttmap. */			\
-   intel_get_cliprects(intel, &cliprects, &num_cliprects, &x_off, &y_off);
 
-/* XXX FBO: this is identical to the macro in spantmp2.h except we get
- * the cliprect info from the context, not the driDrawable.
- * Move this into spantmp2.h someday.
- */
-#define HW_CLIPLOOP()							\
-   do {									\
-      int _nc = num_cliprects;						\
-      while ( _nc-- ) {							\
-	 int minx = cliprects[_nc].x1 - x_off;				\
-	 int miny = cliprects[_nc].y1 - y_off;				\
-	 int maxx = cliprects[_nc].x2 - x_off;				\
-	 int maxy = cliprects[_nc].y2 - y_off;
-	
-#if 0
-      }}
-#endif
+#define HW_CLIPLOOP()
+#define HW_ENDCLIPLOOP()
 
 #define Y_FLIP(_y) ((_y) * yScale + yBias)
 
@@ -297,9 +281,9 @@ static uint32_t y_tile_swizzle(struct intel_renderbuffer *irb,
 #define HW_UNLOCK()
 
 /* Convenience macros to avoid typing the swizzle argument over and over */
-#define NO_TILE(_X, _Y) no_tile_swizzle(irb, (_X) + x_off, (_Y) + y_off)
-#define X_TILE(_X, _Y) x_tile_swizzle(irb, (_X) + x_off, (_Y) + y_off)
-#define Y_TILE(_X, _Y) y_tile_swizzle(irb, (_X) + x_off, (_Y) + y_off)
+#define NO_TILE(_X, _Y) no_tile_swizzle(irb, (_X), (_Y))
+#define X_TILE(_X, _Y) x_tile_swizzle(irb, (_X), (_Y))
+#define Y_TILE(_X, _Y) y_tile_swizzle(irb, (_X), (_Y))
 
 /* r5g6b5 color span and pixel functions */
 #define INTEL_PIXEL_FMT GL_RGB
@@ -342,18 +326,15 @@ static uint32_t y_tile_swizzle(struct intel_renderbuffer *irb,
 #include "intel_spantmp.h"
 
 #define LOCAL_DEPTH_VARS						\
-   struct intel_context *intel = intel_context(ctx);			\
    struct intel_renderbuffer *irb = intel_renderbuffer(rb);		\
    const GLint yScale = ctx->DrawBuffer->Name ? 1 : -1;			\
    const GLint yBias = ctx->DrawBuffer->Name ? 0 : irb->Base.Height - 1;\
-   unsigned int num_cliprects;						\
-   struct drm_clip_rect *cliprects;					\
-   int x_off, y_off;							\
+   int minx = 0, miny = 0;						\
+   int maxx = ctx->DrawBuffer->Width;					\
+   int maxy = ctx->DrawBuffer->Height;					\
    int pitch = irb->region->pitch * irb->region->cpp;			\
    void *buf = irb->region->buffer->virtual;				\
    (void)buf; (void)pitch; /* unused for non-gttmap. */			\
-   intel_get_cliprects(intel, &cliprects, &num_cliprects, &x_off, &y_off);
-
 
 #define LOCAL_STENCIL_VARS LOCAL_DEPTH_VARS
 
diff --git a/src/mesa/drivers/dri/intel/intel_state.c b/src/mesa/drivers/dri/intel/intel_state.c
index aefae53eb2..c5ef909dbf 100644
--- a/src/mesa/drivers/dri/intel/intel_state.c
+++ b/src/mesa/drivers/dri/intel/intel_state.c
@@ -35,8 +35,6 @@
 
 #include "intel_screen.h"
 #include "intel_context.h"
-#include "intel_regions.h"
-#include "swrast/swrast.h"
 
 int
 intel_translate_shadow_compare_func(GLenum func)
diff --git a/src/mesa/drivers/dri/intel/intel_tex_copy.c b/src/mesa/drivers/dri/intel/intel_tex_copy.c
index d8e71093c4..d67451cf8e 100644
--- a/src/mesa/drivers/dri/intel/intel_tex_copy.c
+++ b/src/mesa/drivers/dri/intel/intel_tex_copy.c
@@ -36,7 +36,6 @@
 
 #include "intel_screen.h"
 #include "intel_context.h"
-#include "intel_batchbuffer.h"
 #include "intel_buffers.h"
 #include "intel_mipmap_tree.h"
 #include "intel_regions.h"
@@ -114,8 +113,6 @@ do_copy_texsubimage(struct intel_context *intel,
       drm_intel_bo *dst_bo = intel_region_buffer(intel,
 						 intelImage->mt->region,
 						 INTEL_WRITE_PART);
-      const GLint orig_x = x;
-      const GLint orig_y = y;
       GLuint image_x, image_y;
       GLshort src_pitch;
 
@@ -125,9 +122,6 @@ do_copy_texsubimage(struct intel_context *intel,
 				     intelImage->face,
 				     0,
 				     &image_x, &image_y);
-      /* Update dst for clipped src.  Need to also clip the source rect. */
-      dstx += x - orig_x;
-      dsty += y - orig_y;
 
       /* Can't blit to tiled buffers with non-tile-aligned offset. */
       if (intelImage->mt->region->tiling == I915_TILING_Y) {
diff --git a/src/mesa/drivers/dri/intel/intel_tex_format.c b/src/mesa/drivers/dri/intel/intel_tex_format.c
index a7c6c45ffe..7be5231eae 100644
--- a/src/mesa/drivers/dri/intel/intel_tex_format.c
+++ b/src/mesa/drivers/dri/intel/intel_tex_format.c
@@ -1,6 +1,5 @@
 #include "intel_context.h"
 #include "intel_tex.h"
-#include "intel_chipset.h"
 #include "main/enums.h"
 
 
diff --git a/src/mesa/drivers/dri/intel/intel_tex_image.c b/src/mesa/drivers/dri/intel/intel_tex_image.c
index 6f41eafd0e..a56e085b90 100644
--- a/src/mesa/drivers/dri/intel/intel_tex_image.c
+++ b/src/mesa/drivers/dri/intel/intel_tex_image.c
@@ -7,7 +7,6 @@
 #include "main/convolve.h"
 #include "main/context.h"
 #include "main/formats.h"
-#include "main/image.h"
 #include "main/texcompress.h"
 #include "main/texstore.h"
 #include "main/texgetimage.h"
@@ -241,7 +240,8 @@ try_pbo_upload(struct intel_context *intel,
       if (!intelEmitCopyBlit(intel,
 			     intelImage->mt->cpp,
 			     src_stride, src_buffer, src_offset, GL_FALSE,
-			     dst_stride, dst_buffer, 0, GL_FALSE,
+			     dst_stride, dst_buffer, 0,
+			     intelImage->mt->region->tiling,
 			     0, 0, dst_x, dst_y, width, height,
 			     GL_COPY)) {
 	 return GL_FALSE;
diff --git a/src/mesa/drivers/dri/intel/intel_tex_validate.c b/src/mesa/drivers/dri/intel/intel_tex_validate.c
index c9a24ac398..ed5c5d896b 100644
--- a/src/mesa/drivers/dri/intel/intel_tex_validate.c
+++ b/src/mesa/drivers/dri/intel/intel_tex_validate.c
@@ -2,10 +2,8 @@
 #include "main/macros.h"
 
 #include "intel_context.h"
-#include "intel_batchbuffer.h"
 #include "intel_mipmap_tree.h"
 #include "intel_tex.h"
-#include "intel_chipset.h"
 
 #define FILE_DEBUG_FLAG DEBUG_TEXTURE