Merge commit 'origin/master' into radeon-rewrite

Conflicts:
	src/mesa/drivers/dri/r200/r200_state.c
	src/mesa/drivers/dri/r300/r300_context.h
	src/mesa/drivers/dri/r300/r300_fragprog.c
	src/mesa/drivers/dri/r300/r300_state.c
	src/mesa/drivers/dri/r300/r300_texmem.c
	src/mesa/drivers/dri/r300/r300_texstate.c
	src/mesa/drivers/dri/r300/r500_fragprog.c
	src/mesa/drivers/dri/radeon/radeon_screen.c
	src/mesa/drivers/dri/radeon/radeon_state.c

8 files changed, 239 insertions, 42 deletions

diff --git a/src/gallium/auxiliary/util/u_debug.c b/src/gallium/auxiliary/util/u_debug.c
index ae47a274a6..18597ef839 100644
--- a/src/gallium/auxiliary/util/u_debug.c
+++ b/src/gallium/auxiliary/util/u_debug.c
@@ -721,6 +721,7 @@ void
 debug_dump_surface_bmp(const char *filename,
                        struct pipe_surface *surface)
 {
+#ifndef PIPE_SUBSYSTEM_WINDOWS_MINIPORT
    struct pipe_transfer *transfer;
    struct pipe_texture *texture = surface->texture;
    struct pipe_screen *screen = texture->screen;
@@ -733,6 +734,7 @@ debug_dump_surface_bmp(const char *filename,
    debug_dump_transfer_bmp(filename, transfer);
 
    screen->tex_transfer_destroy(transfer);
+#endif
 }
 
 void
@@ -740,11 +742,7 @@ debug_dump_transfer_bmp(const char *filename,
                         struct pipe_transfer *transfer)
 {
 #ifndef PIPE_SUBSYSTEM_WINDOWS_MINIPORT
-   struct util_stream *stream;
-   struct bmp_file_header bmfh;
-   struct bmp_info_header bmih;
    float *rgba;
-   unsigned x, y;
 
    if (!transfer)
       goto error1;
@@ -753,19 +751,47 @@ debug_dump_transfer_bmp(const char *filename,
    if(!rgba)
       goto error1;
 
+   pipe_get_tile_rgba(transfer, 0, 0,
+                      transfer->width, transfer->height,
+                      rgba);
+
+   debug_dump_float_rgba_bmp(filename,
+                             transfer->width, transfer->height,
+                             rgba, transfer->width);
+
+   FREE(rgba);
+error1:
+   ;
+#endif
+}
+
+void
+debug_dump_float_rgba_bmp(const char *filename,
+                          unsigned width, unsigned height,
+                          float *rgba, unsigned stride)
+{
+#ifndef PIPE_SUBSYSTEM_WINDOWS_MINIPORT
+   struct util_stream *stream;
+   struct bmp_file_header bmfh;
+   struct bmp_info_header bmih;
+   unsigned x, y;
+
+   if(!rgba)
+      goto error1;
+
    bmfh.bfType = 0x4d42;
-   bmfh.bfSize = 14 + 40 + transfer->height*transfer->width*4;
+   bmfh.bfSize = 14 + 40 + height*width*4;
    bmfh.bfReserved1 = 0;
    bmfh.bfReserved2 = 0;
    bmfh.bfOffBits = 14 + 40;
 
    bmih.biSize = 40;
-   bmih.biWidth = transfer->width;
-   bmih.biHeight = transfer->height;
+   bmih.biWidth = width;
+   bmih.biHeight = height;
    bmih.biPlanes = 1;
    bmih.biBitCount = 32;
    bmih.biCompression = 0;
-   bmih.biSizeImage = transfer->height*transfer->width*4;
+   bmih.biSizeImage = height*width*4;
    bmih.biXPelsPerMeter = 0;
    bmih.biYPelsPerMeter = 0;
    bmih.biClrUsed = 0;
@@ -773,19 +799,15 @@ debug_dump_transfer_bmp(const char *filename,
 
    stream = util_stream_create(filename, bmfh.bfSize);
    if(!stream)
-      goto error2;
+      goto error1;
 
    util_stream_write(stream, &bmfh, 14);
    util_stream_write(stream, &bmih, 40);
 
-   pipe_get_tile_rgba(transfer, 0, 0,
-                      transfer->width, transfer->height,
-                      rgba);
-
-   y = transfer->height;
+   y = height;
    while(y--) {
-      float *ptr = rgba + (transfer->width * y * 4);
-      for(x = 0; x < transfer->width; ++x)
+      float *ptr = rgba + (stride * y * 4);
+      for(x = 0; x < width; ++x)
       {
          struct bmp_rgb_quad pixel;
          pixel.rgbRed   = float_to_ubyte(ptr[x*4 + 0]);
@@ -797,8 +819,6 @@ debug_dump_transfer_bmp(const char *filename,
    }
 
    util_stream_close(stream);
-error2:
-   FREE(rgba);
 error1:
    ;
 #endif
diff --git a/src/gallium/auxiliary/util/u_debug.h b/src/gallium/auxiliary/util/u_debug.h
index 8d703e47fc..d42b65ce28 100644
--- a/src/gallium/auxiliary/util/u_debug.h
+++ b/src/gallium/auxiliary/util/u_debug.h
@@ -102,6 +102,22 @@ debug_printf(const char *format, ...)
 }
 
 
+/*
+ * ... isn't portable so we need to pass arguments in parentheses.
+ *
+ * usage:
+ *    debug_printf_once(("awnser: %i\n", 42));
+ */
+#define debug_printf_once(args) \
+   do { \
+      static boolean once = TRUE; \
+      if (once) { \
+         once = FALSE; \
+         debug_printf args; \
+      } \
+   } while (0)
+
+
 #ifdef DEBUG
 #define debug_vprintf(_format, _ap) _debug_vprintf(_format, _ap)
 #else
@@ -347,10 +363,15 @@ void debug_dump_surface_bmp(const char *filename,
                             struct pipe_surface *surface);
 void debug_dump_transfer_bmp(const char *filename,
                              struct pipe_transfer *transfer);
+void debug_dump_float_rgba_bmp(const char *filename,
+                               unsigned width, unsigned height,
+                               float *rgba, unsigned stride);
 #else
 #define debug_dump_image(prefix, format, cpp, width, height, stride, data) ((void)0)
 #define debug_dump_surface(prefix, surface) ((void)0)
 #define debug_dump_surface_bmp(filename, surface) ((void)0)
+#define debug_dump_transfer_bmp(filename, transfer) ((void)0)
+#define debug_dump_rgba_float_bmp(filename, width, height, rgba, stride) ((void)0)
 #endif
 
 
diff --git a/src/gallium/auxiliary/util/u_debug_stack.c b/src/gallium/auxiliary/util/u_debug_stack.c
index e9891fde8a..528a1c394b 100644
--- a/src/gallium/auxiliary/util/u_debug_stack.c
+++ b/src/gallium/auxiliary/util/u_debug_stack.c
@@ -62,6 +62,8 @@ debug_backtrace_capture(struct debug_stack_frame *backtrace,
    
 #ifdef PIPE_ARCH_X86
    while(nr_frames) {
+      const void **next_frame_pointer;
+
       if(!frame_pointer)
          break;
       
@@ -72,7 +74,14 @@ debug_backtrace_capture(struct debug_stack_frame *backtrace,
          --nr_frames;
       }
       
-      frame_pointer = (const void **)frame_pointer[0];
+      next_frame_pointer = (const void **)frame_pointer[0];
+      
+      /* Limit the stack walk to avoid referencing undefined memory */
+      if((uintptr_t)next_frame_pointer <= (uintptr_t)frame_pointer ||
+         (uintptr_t)next_frame_pointer > (uintptr_t)frame_pointer + 64*1024)
+         break;
+      
+      frame_pointer = next_frame_pointer;
    }
 #endif
    
diff --git a/src/gallium/auxiliary/util/u_gen_mipmap.c b/src/gallium/auxiliary/util/u_gen_mipmap.c
index 690412ae7d..6fa13a8ce1 100644
--- a/src/gallium/auxiliary/util/u_gen_mipmap.c
+++ b/src/gallium/auxiliary/util/u_gen_mipmap.c
@@ -1316,7 +1316,6 @@ util_create_gen_mipmap(struct pipe_context *pipe,
    for (i = 0; i < 4; i++) {
       ctx->vertices[i][0][2] = 0.0f; /* z */
       ctx->vertices[i][0][3] = 1.0f; /* w */
-      ctx->vertices[i][1][2] = 0.0f; /* r */
       ctx->vertices[i][1][3] = 1.0f; /* q */
    }
 
@@ -1350,29 +1349,104 @@ get_next_slot(struct gen_mipmap_state *ctx)
 
 
 static unsigned
-set_vertex_data(struct gen_mipmap_state *ctx, float width, float height)
+set_vertex_data(struct gen_mipmap_state *ctx,
+                enum pipe_texture_target tex_target,
+                uint face, float width, float height)
 {
    unsigned offset;
 
+   /* vert[0].position */
    ctx->vertices[0][0][0] = 0.0f; /*x*/
    ctx->vertices[0][0][1] = 0.0f; /*y*/
-   ctx->vertices[0][1][0] = 0.0f; /*s*/
-   ctx->vertices[0][1][1] = 0.0f; /*t*/
 
+   /* vert[1].position */
    ctx->vertices[1][0][0] = width;
    ctx->vertices[1][0][1] = 0.0f;
-   ctx->vertices[1][1][0] = 1.0f;
-   ctx->vertices[1][1][1] = 0.0f;
 
+   /* vert[2].position */
    ctx->vertices[2][0][0] = width;
    ctx->vertices[2][0][1] = height;
-   ctx->vertices[2][1][0] = 1.0f;
-   ctx->vertices[2][1][1] = 1.0f;
 
+   /* vert[3].position */
    ctx->vertices[3][0][0] = 0.0f;
    ctx->vertices[3][0][1] = height;
-   ctx->vertices[3][1][0] = 0.0f;
-   ctx->vertices[3][1][1] = 1.0f;
+
+   /* Setup vertex texcoords.  This is a little tricky for cube maps. */
+   if (tex_target == PIPE_TEXTURE_CUBE) {
+      static const float st[4][2] = {
+         {0.0f, 0.0f}, {1.0f, 0.0f}, {1.0f, 1.0f}, {0.0f, 1.0f}
+      };
+      float rx, ry, rz;
+      uint i;
+
+      /* loop over quad verts */
+      for (i = 0; i < 4; i++) {
+         /* Compute sc = +/-scale and tc = +/-scale.
+          * Not +/-1 to avoid cube face selection ambiguity near the edges,
+          * though that can still sometimes happen with this scale factor...
+          */
+         const float scale = 0.9999;
+         const float sc = (2.0f * st[i][0] - 1.0f) * scale;
+         const float tc = (2.0f * st[i][1] - 1.0f) * scale;
+
+         switch (face) {
+         case PIPE_TEX_FACE_POS_X:
+            rx = 1.0f;
+            ry = -tc;
+            rz = -sc;
+            break;
+         case PIPE_TEX_FACE_NEG_X:
+            rx = -1.0f;
+            ry = -tc;
+            rz = sc;
+            break;
+         case PIPE_TEX_FACE_POS_Y:
+            rx = sc;
+            ry = 1.0f;
+            rz = tc;
+            break;
+         case PIPE_TEX_FACE_NEG_Y:
+            rx = sc;
+            ry = -1.0f;
+            rz = -tc;
+            break;
+         case PIPE_TEX_FACE_POS_Z:
+            rx = sc;
+            ry = -tc;
+            rz = 1.0f;
+            break;
+         case PIPE_TEX_FACE_NEG_Z:
+            rx = -sc;
+            ry = -tc;
+            rz = -1.0f;
+            break;
+         default:
+            assert(0);
+         }
+
+         ctx->vertices[i][1][0] = rx; /*s*/
+         ctx->vertices[i][1][1] = ry; /*t*/
+         ctx->vertices[i][1][2] = rz; /*r*/
+      }
+   }
+   else {
+      /* 1D/2D */
+      ctx->vertices[0][1][0] = 0.0f; /*s*/
+      ctx->vertices[0][1][1] = 0.0f; /*t*/
+      ctx->vertices[0][1][2] = 0.0f; /*r*/
+
+      ctx->vertices[1][1][0] = 1.0f;
+      ctx->vertices[1][1][1] = 0.0f;
+      ctx->vertices[1][1][2] = 0.0f;
+
+      ctx->vertices[2][1][0] = 1.0f;
+      ctx->vertices[2][1][1] = 1.0f;
+      ctx->vertices[2][1][2] = 0.0f;
+
+      ctx->vertices[3][1][0] = 0.0f;
+      ctx->vertices[3][1][1] = 1.0f;
+      ctx->vertices[3][1][2] = 0.0f;
+   }
 
    offset = get_next_slot( ctx );
 
@@ -1503,6 +1577,8 @@ util_gen_mipmap(struct gen_mipmap_state *ctx,
 
       /* quad coords in window coords (bypassing vs, clip and viewport) */
       offset = set_vertex_data(ctx,
+                               pt->target,
+                               face,
                                (float) pt->width[dstLevel],
                                (float) pt->height[dstLevel]);
 
diff --git a/src/gallium/auxiliary/util/u_linear.c b/src/gallium/auxiliary/util/u_linear.c
index 6be365e53b..a1dce3f5cf 100644
--- a/src/gallium/auxiliary/util/u_linear.c
+++ b/src/gallium/auxiliary/util/u_linear.c
@@ -1,3 +1,34 @@
+/**************************************************************************
+ * 
+ * Copyright 2009 VMware, Inc.
+ * All Rights Reserved.
+ * 
+ * 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, sub license, 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 NON-INFRINGEMENT.
+ * IN NO EVENT SHALL VMWARE AND/OR ITS SUPPLIERS 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.
+ * 
+ **************************************************************************/
+
+/**
+ * Functions for converting tiled data to linear and vice versa.
+ */
+
 
 #include "util/u_debug.h"
 #include "u_linear.h"
diff --git a/src/gallium/auxiliary/util/u_linear.h b/src/gallium/auxiliary/util/u_linear.h
index 1589f029bc..b74308ffa3 100644
--- a/src/gallium/auxiliary/util/u_linear.h
+++ b/src/gallium/auxiliary/util/u_linear.h
@@ -1,3 +1,34 @@
+/**************************************************************************
+ * 
+ * Copyright 2009 VMware, Inc.
+ * All Rights Reserved.
+ * 
+ * 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, sub license, 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 NON-INFRINGEMENT.
+ * IN NO EVENT SHALL VMWARE AND/OR ITS SUPPLIERS 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.
+ * 
+ **************************************************************************/
+
+/**
+ * Functions for converting tiled data to linear and vice versa.
+ */
+
 
 #ifndef U_LINEAR_H
 #define U_LINEAR_H
diff --git a/src/gallium/auxiliary/util/u_time.c b/src/gallium/auxiliary/util/u_time.c
index 8afe4fccf7..5268cbf79c 100644
--- a/src/gallium/auxiliary/util/u_time.c
+++ b/src/gallium/auxiliary/util/u_time.c
@@ -217,4 +217,9 @@ void util_time_sleep(unsigned usecs)
    } while(start <= curr && curr < end || 
 	   end < start && (curr < end || start <= curr));
 }
+#elif defined(PIPE_SUBSYSTEM_WINDOWS_USER)
+void util_time_sleep(unsigned usecs)
+{
+   Sleep((usecs + 999)/ 1000);
+}
 #endif
diff --git a/src/gallium/auxiliary/util/u_upload_mgr.c b/src/gallium/auxiliary/util/u_upload_mgr.c
index d9c0d7afa8..2eb98068c8 100644
--- a/src/gallium/auxiliary/util/u_upload_mgr.c
+++ b/src/gallium/auxiliary/util/u_upload_mgr.c
@@ -70,7 +70,7 @@ struct u_upload_mgr *u_upload_create( struct pipe_screen *screen,
 }
 
 
-static INLINE void
+static INLINE enum pipe_error
 my_buffer_write(struct pipe_screen *screen,
                 struct pipe_buffer *buf,
                 unsigned offset, unsigned size, unsigned dirty_size,
@@ -84,12 +84,14 @@ my_buffer_write(struct pipe_screen *screen,
    assert(size);
 
    map = pipe_buffer_map_range(screen, buf, offset, size, PIPE_BUFFER_USAGE_CPU_WRITE);
-   assert(map);
-   if(map) {
-      memcpy(map + offset, data, size);
-      pipe_buffer_flush_mapped_range(screen, buf, offset, dirty_size);
-      pipe_buffer_unmap(screen, buf);
-   }
+   if (map == NULL) 
+      return PIPE_ERROR_OUT_OF_MEMORY;
+
+   memcpy(map + offset, data, size);
+   pipe_buffer_flush_mapped_range(screen, buf, offset, dirty_size);
+   pipe_buffer_unmap(screen, buf);
+
+   return PIPE_OK;
 }
 
 /* Release old buffer.
@@ -162,12 +164,14 @@ enum pipe_error u_upload_data( struct u_upload_mgr *upload,
 
    /* Copy the data, using map_range if available:
     */
-   my_buffer_write( upload->screen, 
-                    upload->buffer,
-                    upload->offset,
-                    size, 
-                    alloc_size,
-                    data );
+   ret = my_buffer_write( upload->screen, 
+                          upload->buffer,
+                          upload->offset,
+                          size, 
+                          alloc_size,
+                          data );
+   if (ret)
+      return ret;
 
    /* Emit the return values:
     */