mesa: Rename _mesa_lookup_enum_by_nr() to _mesa_enum_to_string().
[mesa.git] / src / mesa / main / dlist.c
index 83c56a7e7286577b35f8c0eb8b82d9cd81e6739c..5554738d1a3154a26eee2a1540426fe19ab7eb0b 100644 (file)
@@ -29,6 +29,7 @@
  * Display lists management functions.
  */
 
+#include "c99_math.h"
 #include "glheader.h"
 #include "imports.h"
 #include "api_arrayelt.h"
@@ -81,7 +82,7 @@ struct gl_list_instruction
    GLuint Size;
    void (*Execute)( struct gl_context *ctx, void *data );
    void (*Destroy)( struct gl_context *ctx, void *data );
-   void (*Print)( struct gl_context *ctx, void *data );
+   void (*Print)( struct gl_context *ctx, void *data, FILE *f );
 };
 
 
@@ -317,8 +318,6 @@ typedef enum
    /* GL_ARB_draw_buffers */
    OPCODE_DRAW_BUFFERS_ARB,
    /* GL_ATI_fragment_shader */
-   OPCODE_TEX_BUMP_PARAMETER_ATI,
-   /* GL_ATI_fragment_shader */
    OPCODE_BIND_FRAGMENT_SHADER_ATI,
    OPCODE_SET_FRAGMENT_SHADER_CONSTANTS_ATI,
    /* OpenGL 2.0 */
@@ -364,6 +363,45 @@ typedef enum
    OPCODE_UNIFORM_3UIV,
    OPCODE_UNIFORM_4UIV,
 
+   /* OpenGL 4.2 / GL_ARB_separate_shader_objects */
+   OPCODE_USE_PROGRAM_STAGES,
+   OPCODE_PROGRAM_UNIFORM_1F,
+   OPCODE_PROGRAM_UNIFORM_2F,
+   OPCODE_PROGRAM_UNIFORM_3F,
+   OPCODE_PROGRAM_UNIFORM_4F,
+   OPCODE_PROGRAM_UNIFORM_1FV,
+   OPCODE_PROGRAM_UNIFORM_2FV,
+   OPCODE_PROGRAM_UNIFORM_3FV,
+   OPCODE_PROGRAM_UNIFORM_4FV,
+   OPCODE_PROGRAM_UNIFORM_1I,
+   OPCODE_PROGRAM_UNIFORM_2I,
+   OPCODE_PROGRAM_UNIFORM_3I,
+   OPCODE_PROGRAM_UNIFORM_4I,
+   OPCODE_PROGRAM_UNIFORM_1IV,
+   OPCODE_PROGRAM_UNIFORM_2IV,
+   OPCODE_PROGRAM_UNIFORM_3IV,
+   OPCODE_PROGRAM_UNIFORM_4IV,
+   OPCODE_PROGRAM_UNIFORM_1UI,
+   OPCODE_PROGRAM_UNIFORM_2UI,
+   OPCODE_PROGRAM_UNIFORM_3UI,
+   OPCODE_PROGRAM_UNIFORM_4UI,
+   OPCODE_PROGRAM_UNIFORM_1UIV,
+   OPCODE_PROGRAM_UNIFORM_2UIV,
+   OPCODE_PROGRAM_UNIFORM_3UIV,
+   OPCODE_PROGRAM_UNIFORM_4UIV,
+   OPCODE_PROGRAM_UNIFORM_MATRIX22F,
+   OPCODE_PROGRAM_UNIFORM_MATRIX33F,
+   OPCODE_PROGRAM_UNIFORM_MATRIX44F,
+   OPCODE_PROGRAM_UNIFORM_MATRIX23F,
+   OPCODE_PROGRAM_UNIFORM_MATRIX32F,
+   OPCODE_PROGRAM_UNIFORM_MATRIX24F,
+   OPCODE_PROGRAM_UNIFORM_MATRIX42F,
+   OPCODE_PROGRAM_UNIFORM_MATRIX34F,
+   OPCODE_PROGRAM_UNIFORM_MATRIX43F,
+
+   /* GL_ARB_clip_control */
+   OPCODE_CLIP_CONTROL,
+
    /* GL_ARB_color_buffer_float */
    OPCODE_CLAMP_COLOR,
 
@@ -407,10 +445,6 @@ typedef enum
    OPCODE_TEXPARAMETER_I,
    OPCODE_TEXPARAMETER_UI,
 
-   /* GL_EXT_separate_shader_objects */
-   OPCODE_ACTIVE_PROGRAM_EXT,
-   OPCODE_USE_SHADER_PROGRAM_EXT,
-
    /* GL_ARB_instanced_arrays */
    OPCODE_VERTEX_ATTRIB_DIVISOR,
 
@@ -451,9 +485,13 @@ typedef enum
    /* ARB_uniform_buffer_object */
    OPCODE_UNIFORM_BLOCK_BINDING,
 
+   /* EXT_polygon_offset_clamp */
+   OPCODE_POLYGON_OFFSET_CLAMP,
+
    /* The following three are meta instructions */
    OPCODE_ERROR,                /* raise compiled-in error */
    OPCODE_CONTINUE,
+   OPCODE_NOP,                  /* No-op (used for 8-byte alignment */
    OPCODE_END_OF_LIST,
    OPCODE_EXT_0
 } OpCode;
@@ -470,6 +508,10 @@ typedef enum
  * Each instruction in the display list is stored as a sequence of
  * contiguous nodes in memory.
  * Each node is the union of a variety of data types.
+ *
+ * Note, all of these members should be 4 bytes in size or less for the
+ * sake of compact display lists.  We store 8-byte pointers in a pair of
+ * these nodes using the save/get_pointer() functions below.
  */
 union gl_dlist_node
 {
@@ -484,17 +526,64 @@ union gl_dlist_node
    GLenum e;
    GLfloat f;
    GLsizei si;
-   GLvoid *data;
-   void *next;                  /* If prev node's opcode==OPCODE_CONTINUE */
 };
 
 
 typedef union gl_dlist_node Node;
 
 
+/** How many 4-byte dwords to store a pointer */
+#define POINTER_DWORDS (sizeof(void *) / 4)
+
+/* We want to keep sizeof(union gl_dlist_node) == 4 to minimize
+ * space for display lists.  The following types and functions are
+ * used to help store 4- and 8-byte pointers in 1 or 2 dlist_nodes.
+ */
+union pointer
+{
+   void *ptr;
+   GLuint dwords[POINTER_DWORDS];
+};
+
+
+/**
+ * Save a 4 or 8-byte pointer at dest (and dest+1).
+ */
+static inline void
+save_pointer(Node *dest, void *src)
+{
+   union pointer p;
+   unsigned i;
+
+   STATIC_ASSERT(POINTER_DWORDS == 1 || POINTER_DWORDS == 2);
+   STATIC_ASSERT(sizeof(Node) == 4);
+
+   p.ptr = src;
+
+   for (i = 0; i < POINTER_DWORDS; i++)
+      dest[i].ui = p.dwords[i];
+}
+
+
+/**
+ * Retrieve a 4 or 8-byte pointer from node (node+1).
+ */
+static inline void *
+get_pointer(const Node *node)
+{
+   union pointer p;
+   unsigned i;
+
+   for (i = 0; i < POINTER_DWORDS; i++)
+      p.dwords[i] = node[i].ui;
+
+   return p.ptr;
+}
+
+
 /**
  * Used to store a 64-bit uint in a pair of "Nodes" for the sake of 32-bit
- * environment.  In 64-bit env, sizeof(Node)==8 anyway.
+ * environment.
  */
 union uint64_pair
 {
@@ -504,9 +593,9 @@ union uint64_pair
 
 
 /**
- * How many nodes to allocate at a time.
- *
- * \note Reduced now that we hold vertices etc. elsewhere.
+ * How many nodes to allocate at a time.  Note that bulk vertex data
+ * from glBegin/glVertex/glEnd primitives will typically wind up in
+ * a VBO, and not directly in the display list itself.
  */
 #define BLOCK_SIZE 256
 
@@ -522,14 +611,9 @@ static GLuint InstSize[OPCODE_END_OF_LIST + 1];
 void mesa_print_display_list(GLuint list);
 
 
-/**********************************************************************/
-/*****                           Private                          *****/
-/**********************************************************************/
-
-
 /**
- * Make an empty display list.  This is used by glGenLists() to
- * reserve display list IDs.
+ * Allocate a gl_display_list object with an initial block of storage.
+ * \param count  how many display list nodes/tokes to allocate
  */
 static struct gl_display_list *
 make_list(GLuint name, GLuint count)
@@ -587,11 +671,11 @@ ext_opcode_execute(struct gl_context *ctx, Node *node)
 
 /** Print an extended opcode instruction */
 static GLint
-ext_opcode_print(struct gl_context *ctx, Node *node)
+ext_opcode_print(struct gl_context *ctx, Node *node, FILE *f)
 {
    const GLint i = node[0].opcode - OPCODE_EXT_0;
    GLint step;
-   ctx->ListExt->Opcode[i].Print(ctx, &node[1]);
+   ctx->ListExt->Opcode[i].Print(ctx, &node[1], f);
    step = ctx->ListExt->Opcode[i].Size;
    return step;
 }
@@ -621,75 +705,75 @@ _mesa_delete_list(struct gl_context *ctx, struct gl_display_list *dlist)
          switch (opcode) {
             /* for some commands, we need to free malloc'd memory */
          case OPCODE_MAP1:
-            free(n[6].data);
+            free(get_pointer(&n[6]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_MAP2:
-            free(n[10].data);
+            free(get_pointer(&n[10]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_DRAW_PIXELS:
-            free(n[5].data);
+            free(get_pointer(&n[5]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_BITMAP:
-            free(n[7].data);
+            free(get_pointer(&n[7]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_POLYGON_STIPPLE:
-            free(n[1].data);
+            free(get_pointer(&n[1]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_TEX_IMAGE1D:
-            free(n[8].data);
+            free(get_pointer(&n[8]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_TEX_IMAGE2D:
-            free(n[9].data);
+            free(get_pointer(&n[9]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_TEX_IMAGE3D:
-            free(n[10].data);
+            free(get_pointer(&n[10]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_TEX_SUB_IMAGE1D:
-            free(n[7].data);
+            free(get_pointer(&n[7]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_TEX_SUB_IMAGE2D:
-            free(n[9].data);
+            free(get_pointer(&n[9]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_TEX_SUB_IMAGE3D:
-            free(n[11].data);
+            free(get_pointer(&n[11]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_COMPRESSED_TEX_IMAGE_1D:
-            free(n[7].data);
+            free(get_pointer(&n[7]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_COMPRESSED_TEX_IMAGE_2D:
-            free(n[8].data);
+            free(get_pointer(&n[8]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_COMPRESSED_TEX_IMAGE_3D:
-            free(n[9].data);
+            free(get_pointer(&n[9]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_COMPRESSED_TEX_SUB_IMAGE_1D:
-            free(n[7].data);
+            free(get_pointer(&n[7]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_COMPRESSED_TEX_SUB_IMAGE_2D:
-            free(n[9].data);
+            free(get_pointer(&n[9]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_COMPRESSED_TEX_SUB_IMAGE_3D:
-            free(n[11].data);
+            free(get_pointer(&n[11]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_PROGRAM_STRING_ARB:
-            free(n[4].data);      /* program string */
+            free(get_pointer(&n[4]));      /* program string */
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_UNIFORM_1FV:
@@ -704,7 +788,7 @@ _mesa_delete_list(struct gl_context *ctx, struct gl_display_list *dlist)
          case OPCODE_UNIFORM_2UIV:
          case OPCODE_UNIFORM_3UIV:
          case OPCODE_UNIFORM_4UIV:
-            free(n[3].data);
+            free(get_pointer(&n[3]));
             n += InstSize[n[0].opcode];
             break;
          case OPCODE_UNIFORM_MATRIX22:
@@ -716,12 +800,43 @@ _mesa_delete_list(struct gl_context *ctx, struct gl_display_list *dlist)
          case OPCODE_UNIFORM_MATRIX32:
          case OPCODE_UNIFORM_MATRIX34:
          case OPCODE_UNIFORM_MATRIX43:
-            free(n[4].data);
+            free(get_pointer(&n[4]));
+            n += InstSize[n[0].opcode];
+            break;
+         case OPCODE_PROGRAM_UNIFORM_1FV:
+         case OPCODE_PROGRAM_UNIFORM_2FV:
+         case OPCODE_PROGRAM_UNIFORM_3FV:
+         case OPCODE_PROGRAM_UNIFORM_4FV:
+         case OPCODE_PROGRAM_UNIFORM_1IV:
+         case OPCODE_PROGRAM_UNIFORM_2IV:
+         case OPCODE_PROGRAM_UNIFORM_3IV:
+         case OPCODE_PROGRAM_UNIFORM_4IV:
+         case OPCODE_PROGRAM_UNIFORM_1UIV:
+         case OPCODE_PROGRAM_UNIFORM_2UIV:
+         case OPCODE_PROGRAM_UNIFORM_3UIV:
+         case OPCODE_PROGRAM_UNIFORM_4UIV:
+            free(get_pointer(&n[4]));
+            n += InstSize[n[0].opcode];
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX22F:
+         case OPCODE_PROGRAM_UNIFORM_MATRIX33F:
+         case OPCODE_PROGRAM_UNIFORM_MATRIX44F:
+         case OPCODE_PROGRAM_UNIFORM_MATRIX24F:
+         case OPCODE_PROGRAM_UNIFORM_MATRIX42F:
+         case OPCODE_PROGRAM_UNIFORM_MATRIX23F:
+         case OPCODE_PROGRAM_UNIFORM_MATRIX32F:
+         case OPCODE_PROGRAM_UNIFORM_MATRIX34F:
+         case OPCODE_PROGRAM_UNIFORM_MATRIX43F:
+            free(get_pointer(&n[5]));
+            n += InstSize[n[0].opcode];
+            break;
+         case OPCODE_PIXEL_MAP:
+            free(get_pointer(&n[3]));
             n += InstSize[n[0].opcode];
             break;
 
          case OPCODE_CONTINUE:
-            n = (Node *) n[1].next;
+            n = (Node *) get_pointer(&n[1]);
             free(block);
             block = n;
             break;
@@ -798,7 +913,7 @@ translate_id(GLsizei n, GLenum type, const GLvoid * list)
       return (GLint) uiptr[n];
    case GL_FLOAT:
       fptr = (GLfloat *) list;
-      return (GLint) FLOORF(fptr[n]);
+      return (GLint) floorf(fptr[n]);
    case GL_2_BYTES:
       ubptr = ((GLubyte *) list) + 2 * n;
       return (GLint) ubptr[0] * 256
@@ -820,12 +935,6 @@ translate_id(GLsizei n, GLenum type, const GLvoid * list)
 }
 
 
-
-
-/**********************************************************************/
-/*****                        Public                              *****/
-/**********************************************************************/
-
 /**
  * Wrapper for _mesa_unpack_image/bitmap() that handles pixel buffer objects.
  * If width < 0 or height < 0 or format or type are invalid we'll just
@@ -853,11 +962,8 @@ unpack_image(struct gl_context *ctx, GLuint dimensions,
       /* no PBO */
       GLvoid *image;
 
-      if (type == GL_BITMAP)
-         image = _mesa_unpack_bitmap(width, height, pixels, unpack);
-      else
-         image = _mesa_unpack_image(dimensions, width, height, depth,
-                                    format, type, pixels, unpack);
+      image = _mesa_unpack_image(dimensions, width, height, depth,
+                                 format, type, pixels, unpack);
       if (pixels && !image) {
          _mesa_error(ctx, GL_OUT_OF_MEMORY, "display list construction");
       }
@@ -870,7 +976,8 @@ unpack_image(struct gl_context *ctx, GLuint dimensions,
 
       map = (GLubyte *)
          ctx->Driver.MapBufferRange(ctx, 0, unpack->BufferObj->Size,
-                                   GL_MAP_READ_BIT, unpack->BufferObj);
+                                   GL_MAP_READ_BIT, unpack->BufferObj,
+                                    MAP_INTERNAL);
       if (!map) {
          /* unable to map src buffer! */
          _mesa_error(ctx, GL_INVALID_OPERATION, "unable to map PBO");
@@ -878,13 +985,10 @@ unpack_image(struct gl_context *ctx, GLuint dimensions,
       }
 
       src = ADD_POINTERS(map, pixels);
-      if (type == GL_BITMAP)
-         image = _mesa_unpack_bitmap(width, height, src, unpack);
-      else
-         image = _mesa_unpack_image(dimensions, width, height, depth,
-                                    format, type, src, unpack);
+      image = _mesa_unpack_image(dimensions, width, height, depth,
+                                 format, type, src, unpack);
 
-      ctx->Driver.UnmapBuffer(ctx, unpack->BufferObj);
+      ctx->Driver.UnmapBuffer(ctx, unpack->BufferObj, MAP_INTERNAL);
 
       if (!image) {
          _mesa_error(ctx, GL_OUT_OF_MEMORY, "display list construction");
@@ -897,30 +1001,59 @@ unpack_image(struct gl_context *ctx, GLuint dimensions,
    return NULL;
 }
 
+
+/** Return copy of memory */
+static void *
+memdup(const void *src, GLsizei bytes)
+{
+   void *b = bytes >= 0 ? malloc(bytes) : NULL;
+   if (b)
+      memcpy(b, src, bytes);
+   return b;
+}
+
+
 /**
  * Allocate space for a display list instruction (opcode + payload space).
  * \param opcode  the instruction opcode (OPCODE_* value)
  * \param bytes   instruction payload size (not counting opcode)
- * \return pointer to allocated memory (the opcode space)
+ * \param align8  does the payload need to be 8-byte aligned?
+ *                This is only relevant in 64-bit environments.
+ * \return pointer to allocated memory (the payload will be at pointer+1)
  */
 static Node *
-dlist_alloc(struct gl_context *ctx, OpCode opcode, GLuint bytes)
+dlist_alloc(struct gl_context *ctx, OpCode opcode, GLuint bytes, bool align8)
 {
    const GLuint numNodes = 1 + (bytes + sizeof(Node) - 1) / sizeof(Node);
+   const GLuint contNodes = 1 + POINTER_DWORDS;  /* size of continue info */
+   GLuint nopNode;
    Node *n;
 
-   if (opcode < (GLuint) OPCODE_EXT_0) {
+   if (opcode < OPCODE_EXT_0) {
       if (InstSize[opcode] == 0) {
          /* save instruction size now */
          InstSize[opcode] = numNodes;
       }
       else {
          /* make sure instruction size agrees */
-         ASSERT(numNodes == InstSize[opcode]);
+         assert(numNodes == InstSize[opcode]);
       }
    }
 
-   if (ctx->ListState.CurrentPos + numNodes + 2 > BLOCK_SIZE) {
+   if (sizeof(void *) > sizeof(Node) && align8
+       && ctx->ListState.CurrentPos % 2 == 0) {
+      /* The opcode would get placed at node[0] and the payload would start
+       * at node[1].  But the payload needs to be at an even offset (8-byte
+       * multiple).
+       */
+      nopNode = 1;
+   }
+   else {
+      nopNode = 0;
+   }
+
+   if (ctx->ListState.CurrentPos + nopNode + numNodes + contNodes
+       > BLOCK_SIZE) {
       /* This block is full.  Allocate a new block and chain to it */
       Node *newblock;
       n = ctx->ListState.CurrentBlock + ctx->ListState.CurrentPos;
@@ -930,13 +1063,34 @@ dlist_alloc(struct gl_context *ctx, OpCode opcode, GLuint bytes)
          _mesa_error(ctx, GL_OUT_OF_MEMORY, "Building display list");
          return NULL;
       }
-      n[1].next = (Node *) newblock;
+
+      /* a fresh block should be 8-byte aligned on 64-bit systems */
+      assert(((GLintptr) newblock) % sizeof(void *) == 0);
+
+      save_pointer(&n[1], newblock);
       ctx->ListState.CurrentBlock = newblock;
       ctx->ListState.CurrentPos = 0;
+
+      /* Display list nodes are always 4 bytes.  If we need 8-byte alignment
+       * we have to insert a NOP so that the payload of the real opcode lands
+       * on an even location:
+       *   node[0] = OPCODE_NOP
+       *   node[1] = OPCODE_x;
+       *   node[2] = start of payload
+       */
+      nopNode = sizeof(void *) > sizeof(Node) && align8;
    }
 
    n = ctx->ListState.CurrentBlock + ctx->ListState.CurrentPos;
-   ctx->ListState.CurrentPos += numNodes;
+   if (nopNode) {
+      assert(ctx->ListState.CurrentPos % 2 == 0); /* even value */
+      n[0].opcode = OPCODE_NOP;
+      n++;
+      /* The "real" opcode will now be at an odd location and the payload
+       * will be at an even location.
+       */
+   }
+   ctx->ListState.CurrentPos += nopNode + numNodes;
 
    n[0].opcode = opcode;
 
@@ -957,7 +1111,22 @@ dlist_alloc(struct gl_context *ctx, OpCode opcode, GLuint bytes)
 void *
 _mesa_dlist_alloc(struct gl_context *ctx, GLuint opcode, GLuint bytes)
 {
-   Node *n = dlist_alloc(ctx, (OpCode) opcode, bytes);
+   Node *n = dlist_alloc(ctx, (OpCode) opcode, bytes, false);
+   if (n)
+      return n + 1;  /* return pointer to payload area, after opcode */
+   else
+      return NULL;
+}
+
+
+/**
+ * Same as _mesa_dlist_alloc(), but return a pointer which is 8-byte
+ * aligned in 64-bit environments, 4-byte aligned otherwise.
+ */
+void *
+_mesa_dlist_alloc_aligned(struct gl_context *ctx, GLuint opcode, GLuint bytes)
+{
+   Node *n = dlist_alloc(ctx, (OpCode) opcode, bytes, true);
    if (n)
       return n + 1;  /* return pointer to payload area, after opcode */
    else
@@ -980,7 +1149,7 @@ _mesa_dlist_alloc_opcode(struct gl_context *ctx,
                          GLuint size,
                          void (*execute) (struct gl_context *, void *),
                          void (*destroy) (struct gl_context *, void *),
-                         void (*print) (struct gl_context *, void *))
+                         void (*print) (struct gl_context *, void *, FILE *))
 {
    if (ctx->ListExt->NumOpcodes < MAX_DLIST_EXT_OPCODES) {
       const GLuint i = ctx->ListExt->NumOpcodes++;
@@ -1007,7 +1176,38 @@ _mesa_dlist_alloc_opcode(struct gl_context *ctx,
 static inline Node *
 alloc_instruction(struct gl_context *ctx, OpCode opcode, GLuint nparams)
 {
-   return dlist_alloc(ctx, opcode, nparams * sizeof(Node));
+   return dlist_alloc(ctx, opcode, nparams * sizeof(Node), false);
+}
+
+
+/**
+ * Called by EndList to try to reduce memory used for the list.
+ */
+static void
+trim_list(struct gl_context *ctx)
+{
+   /* If the list we're ending only has one allocated block of nodes/tokens
+    * and its size isn't a full block size, realloc the block to use less
+    * memory.  This is important for apps that create many small display
+    * lists and apps that use glXUseXFont (many lists each containing one
+    * glBitmap call).
+    * Note: we currently only trim display lists that allocated one block
+    * of tokens.  That hits the short list case which is what we're mainly
+    * concerned with.  Trimming longer lists would involve traversing the
+    * linked list of blocks.
+    */
+   struct gl_dlist_state *list = &ctx->ListState;
+
+   if ((list->CurrentList->Head == list->CurrentBlock) &&
+       (list->CurrentPos < BLOCK_SIZE)) {
+      /* There's only one block and it's not full, so realloc */
+      GLuint newSize = list->CurrentPos * sizeof(Node);
+      list->CurrentList->Head =
+      list->CurrentBlock = realloc(list->CurrentBlock, newSize);
+      if (!list->CurrentBlock) {
+         _mesa_error(ctx, GL_OUT_OF_MEMORY, "glEndList");
+      }
+   }
 }
 
 
@@ -1074,7 +1274,7 @@ save_Bitmap(GLsizei width, GLsizei height,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_BITMAP, 7);
+   n = alloc_instruction(ctx, OPCODE_BITMAP, 6 + POINTER_DWORDS);
    if (n) {
       n[1].i = (GLint) width;
       n[2].i = (GLint) height;
@@ -1082,8 +1282,9 @@ save_Bitmap(GLsizei width, GLsizei height,
       n[4].f = yorig;
       n[5].f = xmove;
       n[6].f = ymove;
-      n[7].data = unpack_image(ctx, 2, width, height, 1, GL_COLOR_INDEX,
-                               GL_BITMAP, pixels, &ctx->Unpack);
+      save_pointer(&n[7],
+                   unpack_image(ctx, 2, width, height, 1, GL_COLOR_INDEX,
+                                GL_BITMAP, pixels, &ctx->Unpack));
    }
    if (ctx->ExecuteFlag) {
       CALL_Bitmap(ctx->Exec, (width, height,
@@ -1975,14 +2176,15 @@ save_DrawPixels(GLsizei width, GLsizei height,
 
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-   n = alloc_instruction(ctx, OPCODE_DRAW_PIXELS, 5);
+   n = alloc_instruction(ctx, OPCODE_DRAW_PIXELS, 4 + POINTER_DWORDS);
    if (n) {
       n[1].i = width;
       n[2].i = height;
       n[3].e = format;
       n[4].e = type;
-      n[5].data = unpack_image(ctx, 2, width, height, 1, format, type,
-                               pixels, &ctx->Unpack);
+      save_pointer(&n[5],
+                   unpack_image(ctx, 2, width, height, 1, format, type,
+                                pixels, &ctx->Unpack));
    }
    if (ctx->ExecuteFlag) {
       CALL_DrawPixels(ctx->Exec, (width, height, format, type, pixels));
@@ -2530,7 +2732,7 @@ save_Map1d(GLenum target, GLdouble u1, GLdouble u2, GLint stride,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_MAP1, 6);
+   n = alloc_instruction(ctx, OPCODE_MAP1, 5 + POINTER_DWORDS);
    if (n) {
       GLfloat *pnts = _mesa_copy_map_points1d(target, stride, order, points);
       n[1].e = target;
@@ -2538,7 +2740,7 @@ save_Map1d(GLenum target, GLdouble u1, GLdouble u2, GLint stride,
       n[3].f = (GLfloat) u2;
       n[4].i = _mesa_evaluator_components(target);      /* stride */
       n[5].i = order;
-      n[6].data = (void *) pnts;
+      save_pointer(&n[6], pnts);
    }
    if (ctx->ExecuteFlag) {
       CALL_Map1d(ctx->Exec, (target, u1, u2, stride, order, points));
@@ -2552,7 +2754,7 @@ save_Map1f(GLenum target, GLfloat u1, GLfloat u2, GLint stride,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_MAP1, 6);
+   n = alloc_instruction(ctx, OPCODE_MAP1, 5 + POINTER_DWORDS);
    if (n) {
       GLfloat *pnts = _mesa_copy_map_points1f(target, stride, order, points);
       n[1].e = target;
@@ -2560,7 +2762,7 @@ save_Map1f(GLenum target, GLfloat u1, GLfloat u2, GLint stride,
       n[3].f = u2;
       n[4].i = _mesa_evaluator_components(target);      /* stride */
       n[5].i = order;
-      n[6].data = (void *) pnts;
+      save_pointer(&n[6], pnts);
    }
    if (ctx->ExecuteFlag) {
       CALL_Map1f(ctx->Exec, (target, u1, u2, stride, order, points));
@@ -2577,7 +2779,7 @@ save_Map2d(GLenum target,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_MAP2, 10);
+   n = alloc_instruction(ctx, OPCODE_MAP2, 9 + POINTER_DWORDS);
    if (n) {
       GLfloat *pnts = _mesa_copy_map_points2d(target, ustride, uorder,
                                               vstride, vorder, points);
@@ -2591,7 +2793,7 @@ save_Map2d(GLenum target,
       n[7].i = _mesa_evaluator_components(target);      /*vstride */
       n[8].i = uorder;
       n[9].i = vorder;
-      n[10].data = (void *) pnts;
+      save_pointer(&n[10], pnts);
    }
    if (ctx->ExecuteFlag) {
       CALL_Map2d(ctx->Exec, (target,
@@ -2610,7 +2812,7 @@ save_Map2f(GLenum target,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_MAP2, 10);
+   n = alloc_instruction(ctx, OPCODE_MAP2, 9 + POINTER_DWORDS);
    if (n) {
       GLfloat *pnts = _mesa_copy_map_points2f(target, ustride, uorder,
                                               vstride, vorder, points);
@@ -2624,7 +2826,7 @@ save_Map2f(GLenum target,
       n[7].i = _mesa_evaluator_components(target);      /*vstride */
       n[8].i = uorder;
       n[9].i = vorder;
-      n[10].data = (void *) pnts;
+      save_pointer(&n[10], pnts);
    }
    if (ctx->ExecuteFlag) {
       CALL_Map2f(ctx->Exec, (target, u1, u2, ustride, uorder,
@@ -2777,12 +2979,11 @@ save_PixelMapfv(GLenum map, GLint mapsize, const GLfloat *values)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_PIXEL_MAP, 3);
+   n = alloc_instruction(ctx, OPCODE_PIXEL_MAP, 2 + POINTER_DWORDS);
    if (n) {
       n[1].e = map;
       n[2].i = mapsize;
-      n[3].data = malloc(mapsize * sizeof(GLfloat));
-      memcpy(n[3].data, (void *) values, mapsize * sizeof(GLfloat));
+      save_pointer(&n[3], memdup(values, mapsize * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_PixelMapfv(ctx->Exec, (map, mapsize, values));
@@ -2957,10 +3158,11 @@ save_PolygonStipple(const GLubyte * pattern)
 
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-   n = alloc_instruction(ctx, OPCODE_POLYGON_STIPPLE, 1);
+   n = alloc_instruction(ctx, OPCODE_POLYGON_STIPPLE, POINTER_DWORDS);
    if (n) {
-      n[1].data = unpack_image(ctx, 2, 32, 32, 1, GL_COLOR_INDEX, GL_BITMAP,
-                               pattern, &ctx->Unpack);
+      save_pointer(&n[1],
+                   unpack_image(ctx, 2, 32, 32, 1, GL_COLOR_INDEX, GL_BITMAP,
+                                pattern, &ctx->Unpack));
    }
    if (ctx->ExecuteFlag) {
       CALL_PolygonStipple(ctx->Exec, ((GLubyte *) pattern));
@@ -2993,6 +3195,22 @@ save_PolygonOffsetEXT(GLfloat factor, GLfloat bias)
    save_PolygonOffset(factor, ctx->DrawBuffer->_DepthMaxF * bias);
 }
 
+static void GLAPIENTRY
+save_PolygonOffsetClampEXT(GLfloat factor, GLfloat units, GLfloat clamp)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_POLYGON_OFFSET_CLAMP, 3);
+   if (n) {
+      n[1].f = factor;
+      n[2].f = units;
+      n[3].f = clamp;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_PolygonOffsetClampEXT(ctx->Exec, (factor, units, clamp));
+   }
+}
 
 static void GLAPIENTRY
 save_PopAttrib(void)
@@ -3729,7 +3947,7 @@ save_TexImage1D(GLenum target,
    else {
       Node *n;
       ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-      n = alloc_instruction(ctx, OPCODE_TEX_IMAGE1D, 8);
+      n = alloc_instruction(ctx, OPCODE_TEX_IMAGE1D, 7 + POINTER_DWORDS);
       if (n) {
          n[1].e = target;
          n[2].i = level;
@@ -3738,8 +3956,9 @@ save_TexImage1D(GLenum target,
          n[5].i = border;
          n[6].e = format;
          n[7].e = type;
-         n[8].data = unpack_image(ctx, 1, width, 1, 1, format, type,
-                                  pixels, &ctx->Unpack);
+         save_pointer(&n[8],
+                      unpack_image(ctx, 1, width, 1, 1, format, type,
+                                   pixels, &ctx->Unpack));
       }
       if (ctx->ExecuteFlag) {
          CALL_TexImage1D(ctx->Exec, (target, level, components, width,
@@ -3764,7 +3983,7 @@ save_TexImage2D(GLenum target,
    else {
       Node *n;
       ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-      n = alloc_instruction(ctx, OPCODE_TEX_IMAGE2D, 9);
+      n = alloc_instruction(ctx, OPCODE_TEX_IMAGE2D, 8 + POINTER_DWORDS);
       if (n) {
          n[1].e = target;
          n[2].i = level;
@@ -3774,8 +3993,9 @@ save_TexImage2D(GLenum target,
          n[6].i = border;
          n[7].e = format;
          n[8].e = type;
-         n[9].data = unpack_image(ctx, 2, width, height, 1, format, type,
-                                  pixels, &ctx->Unpack);
+         save_pointer(&n[9],
+                      unpack_image(ctx, 2, width, height, 1, format, type,
+                                   pixels, &ctx->Unpack));
       }
       if (ctx->ExecuteFlag) {
          CALL_TexImage2D(ctx->Exec, (target, level, components, width,
@@ -3802,7 +4022,7 @@ save_TexImage3D(GLenum target,
    else {
       Node *n;
       ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-      n = alloc_instruction(ctx, OPCODE_TEX_IMAGE3D, 10);
+      n = alloc_instruction(ctx, OPCODE_TEX_IMAGE3D, 9 + POINTER_DWORDS);
       if (n) {
          n[1].e = target;
          n[2].i = level;
@@ -3813,8 +4033,9 @@ save_TexImage3D(GLenum target,
          n[7].i = border;
          n[8].e = format;
          n[9].e = type;
-         n[10].data = unpack_image(ctx, 3, width, height, depth, format, type,
-                                   pixels, &ctx->Unpack);
+         save_pointer(&n[10],
+                      unpack_image(ctx, 3, width, height, depth, format, type,
+                                   pixels, &ctx->Unpack));
       }
       if (ctx->ExecuteFlag) {
          CALL_TexImage3D(ctx->Exec, (target, level, internalFormat, width,
@@ -3835,7 +4056,7 @@ save_TexSubImage1D(GLenum target, GLint level, GLint xoffset,
 
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-   n = alloc_instruction(ctx, OPCODE_TEX_SUB_IMAGE1D, 7);
+   n = alloc_instruction(ctx, OPCODE_TEX_SUB_IMAGE1D, 6 + POINTER_DWORDS);
    if (n) {
       n[1].e = target;
       n[2].i = level;
@@ -3843,8 +4064,9 @@ save_TexSubImage1D(GLenum target, GLint level, GLint xoffset,
       n[4].i = (GLint) width;
       n[5].e = format;
       n[6].e = type;
-      n[7].data = unpack_image(ctx, 1, width, 1, 1, format, type,
-                               pixels, &ctx->Unpack);
+      save_pointer(&n[7],
+                   unpack_image(ctx, 1, width, 1, 1, format, type,
+                                pixels, &ctx->Unpack));
    }
    if (ctx->ExecuteFlag) {
       CALL_TexSubImage1D(ctx->Exec, (target, level, xoffset, width,
@@ -3864,7 +4086,7 @@ save_TexSubImage2D(GLenum target, GLint level,
 
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-   n = alloc_instruction(ctx, OPCODE_TEX_SUB_IMAGE2D, 9);
+   n = alloc_instruction(ctx, OPCODE_TEX_SUB_IMAGE2D, 8 + POINTER_DWORDS);
    if (n) {
       n[1].e = target;
       n[2].i = level;
@@ -3874,8 +4096,9 @@ save_TexSubImage2D(GLenum target, GLint level,
       n[6].i = (GLint) height;
       n[7].e = format;
       n[8].e = type;
-      n[9].data = unpack_image(ctx, 2, width, height, 1, format, type,
-                               pixels, &ctx->Unpack);
+      save_pointer(&n[9],
+                   unpack_image(ctx, 2, width, height, 1, format, type,
+                                pixels, &ctx->Unpack));
    }
    if (ctx->ExecuteFlag) {
       CALL_TexSubImage2D(ctx->Exec, (target, level, xoffset, yoffset,
@@ -3895,7 +4118,7 @@ save_TexSubImage3D(GLenum target, GLint level,
 
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-   n = alloc_instruction(ctx, OPCODE_TEX_SUB_IMAGE3D, 11);
+   n = alloc_instruction(ctx, OPCODE_TEX_SUB_IMAGE3D, 10 + POINTER_DWORDS);
    if (n) {
       n[1].e = target;
       n[2].i = level;
@@ -3907,8 +4130,9 @@ save_TexSubImage3D(GLenum target, GLint level,
       n[8].i = (GLint) depth;
       n[9].e = format;
       n[10].e = type;
-      n[11].data = unpack_image(ctx, 3, width, height, depth, format, type,
-                                pixels, &ctx->Unpack);
+      save_pointer(&n[11],
+                   unpack_image(ctx, 3, width, height, depth, format, type,
+                                pixels, &ctx->Unpack));
    }
    if (ctx->ExecuteFlag) {
       CALL_TexSubImage3D(ctx->Exec, (target, level,
@@ -4215,7 +4439,8 @@ save_CompressedTexImage1DARB(GLenum target, GLint level,
       Node *n;
       ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-      n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_IMAGE_1D, 7);
+      n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_IMAGE_1D,
+                            6 + POINTER_DWORDS);
       if (n) {
          n[1].e = target;
          n[2].i = level;
@@ -4223,7 +4448,8 @@ save_CompressedTexImage1DARB(GLenum target, GLint level,
          n[4].i = (GLint) width;
          n[5].i = border;
          n[6].i = imageSize;
-         n[7].data = copy_data(data, imageSize, "glCompressedTexImage1DARB");
+         save_pointer(&n[7],
+                      copy_data(data, imageSize, "glCompressedTexImage1DARB"));
       }
       if (ctx->ExecuteFlag) {
          CALL_CompressedTexImage1D(ctx->Exec,
@@ -4251,7 +4477,8 @@ save_CompressedTexImage2DARB(GLenum target, GLint level,
       Node *n;
       ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-      n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_IMAGE_2D, 8);
+      n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_IMAGE_2D,
+                            7 + POINTER_DWORDS);
       if (n) {
          n[1].e = target;
          n[2].i = level;
@@ -4260,7 +4487,8 @@ save_CompressedTexImage2DARB(GLenum target, GLint level,
          n[5].i = (GLint) height;
          n[6].i = border;
          n[7].i = imageSize;
-         n[8].data = copy_data(data, imageSize, "glCompressedTexImage2DARB");
+         save_pointer(&n[8],
+                      copy_data(data, imageSize, "glCompressedTexImage2DARB"));
       }
       if (ctx->ExecuteFlag) {
          CALL_CompressedTexImage2D(ctx->Exec,
@@ -4288,7 +4516,8 @@ save_CompressedTexImage3DARB(GLenum target, GLint level,
       Node *n;
       ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-      n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_IMAGE_3D, 9);
+      n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_IMAGE_3D,
+                            8 + POINTER_DWORDS);
       if (n) {
          n[1].e = target;
          n[2].i = level;
@@ -4298,7 +4527,8 @@ save_CompressedTexImage3DARB(GLenum target, GLint level,
          n[6].i = (GLint) depth;
          n[7].i = border;
          n[8].i = imageSize;
-         n[9].data = copy_data(data, imageSize, "glCompressedTexImage3DARB");
+         save_pointer(&n[9],
+                      copy_data(data, imageSize, "glCompressedTexImage3DARB"));
       }
       if (ctx->ExecuteFlag) {
          CALL_CompressedTexImage3D(ctx->Exec,
@@ -4319,7 +4549,8 @@ save_CompressedTexSubImage1DARB(GLenum target, GLint level, GLint xoffset,
    GET_CURRENT_CONTEXT(ctx);
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-   n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_SUB_IMAGE_1D, 7);
+   n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_SUB_IMAGE_1D,
+                         6 + POINTER_DWORDS);
    if (n) {
       n[1].e = target;
       n[2].i = level;
@@ -4327,7 +4558,8 @@ save_CompressedTexSubImage1DARB(GLenum target, GLint level, GLint xoffset,
       n[4].i = (GLint) width;
       n[5].e = format;
       n[6].i = imageSize;
-      n[7].data = copy_data(data, imageSize, "glCompressedTexSubImage1DARB");
+      save_pointer(&n[7],
+                   copy_data(data, imageSize, "glCompressedTexSubImage1DARB"));
    }
    if (ctx->ExecuteFlag) {
       CALL_CompressedTexSubImage1D(ctx->Exec, (target, level, xoffset,
@@ -4347,7 +4579,8 @@ save_CompressedTexSubImage2DARB(GLenum target, GLint level, GLint xoffset,
    GET_CURRENT_CONTEXT(ctx);
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-   n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_SUB_IMAGE_2D, 9);
+   n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_SUB_IMAGE_2D,
+                         8 + POINTER_DWORDS);
    if (n) {
       n[1].e = target;
       n[2].i = level;
@@ -4357,7 +4590,8 @@ save_CompressedTexSubImage2DARB(GLenum target, GLint level, GLint xoffset,
       n[6].i = (GLint) height;
       n[7].e = format;
       n[8].i = imageSize;
-      n[9].data = copy_data(data, imageSize, "glCompressedTexSubImage2DARB");
+      save_pointer(&n[9],
+                   copy_data(data, imageSize, "glCompressedTexSubImage2DARB"));
    }
    if (ctx->ExecuteFlag) {
       CALL_CompressedTexSubImage2D(ctx->Exec,
@@ -4377,7 +4611,8 @@ save_CompressedTexSubImage3DARB(GLenum target, GLint level, GLint xoffset,
    GET_CURRENT_CONTEXT(ctx);
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-   n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_SUB_IMAGE_3D, 11);
+   n = alloc_instruction(ctx, OPCODE_COMPRESSED_TEX_SUB_IMAGE_3D,
+                         10 + POINTER_DWORDS);
    if (n) {
       n[1].e = target;
       n[2].i = level;
@@ -4389,7 +4624,8 @@ save_CompressedTexSubImage3DARB(GLenum target, GLint level, GLint xoffset,
       n[8].i = (GLint) depth;
       n[9].e = format;
       n[10].i = imageSize;
-      n[11].data = copy_data(data, imageSize, "glCompressedTexSubImage3DARB");
+      save_pointer(&n[11],
+                   copy_data(data, imageSize, "glCompressedTexSubImage3DARB"));
    }
    if (ctx->ExecuteFlag) {
       CALL_CompressedTexSubImage3D(ctx->Exec,
@@ -4687,7 +4923,7 @@ save_ProgramStringARB(GLenum target, GLenum format, GLsizei len,
 
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
 
-   n = alloc_instruction(ctx, OPCODE_PROGRAM_STRING_ARB, 4);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_STRING_ARB, 3 + POINTER_DWORDS);
    if (n) {
       GLubyte *programCopy = malloc(len);
       if (!programCopy) {
@@ -4698,7 +4934,7 @@ save_ProgramStringARB(GLenum target, GLenum format, GLsizei len,
       n[1].e = target;
       n[2].e = format;
       n[3].i = len;
-      n[4].data = programCopy;
+      save_pointer(&n[4], programCopy);
    }
    if (ctx->ExecuteFlag) {
       CALL_ProgramStringARB(ctx->Exec, (target, format, len, string));
@@ -4808,36 +5044,6 @@ save_DrawBuffersARB(GLsizei count, const GLenum * buffers)
    }
 }
 
-static void GLAPIENTRY
-save_TexBumpParameterfvATI(GLenum pname, const GLfloat *param)
-{
-   GET_CURRENT_CONTEXT(ctx);
-   Node *n;
-
-   n = alloc_instruction(ctx, OPCODE_TEX_BUMP_PARAMETER_ATI, 5);
-   if (n) {
-      n[1].ui = pname;
-      n[2].f = param[0];
-      n[3].f = param[1];
-      n[4].f = param[2];
-      n[5].f = param[3];
-   }
-   if (ctx->ExecuteFlag) {
-      CALL_TexBumpParameterfvATI(ctx->Exec, (pname, param));
-   }
-}
-
-static void GLAPIENTRY
-save_TexBumpParameterivATI(GLenum pname, const GLint *param)
-{
-   GLfloat p[4];
-   p[0] = INT_TO_FLOAT(param[0]);
-   p[1] = INT_TO_FLOAT(param[1]);
-   p[2] = INT_TO_FLOAT(param[2]);
-   p[3] = INT_TO_FLOAT(param[3]);
-   save_TexBumpParameterfvATI(pname, p);
-}
-
 static void GLAPIENTRY
 save_BindFragmentShaderATI(GLuint id)
 {
@@ -4884,7 +5090,7 @@ save_Attr1fNV(GLenum attr, GLfloat x)
       n[2].f = x;
    }
 
-   ASSERT(attr < MAX_VERTEX_GENERIC_ATTRIBS);
+   assert(attr < MAX_VERTEX_GENERIC_ATTRIBS);
    ctx->ListState.ActiveAttribSize[attr] = 1;
    ASSIGN_4V(ctx->ListState.CurrentAttrib[attr], x, 0, 0, 1);
 
@@ -4906,7 +5112,7 @@ save_Attr2fNV(GLenum attr, GLfloat x, GLfloat y)
       n[3].f = y;
    }
 
-   ASSERT(attr < MAX_VERTEX_GENERIC_ATTRIBS);
+   assert(attr < MAX_VERTEX_GENERIC_ATTRIBS);
    ctx->ListState.ActiveAttribSize[attr] = 2;
    ASSIGN_4V(ctx->ListState.CurrentAttrib[attr], x, y, 0, 1);
 
@@ -4929,7 +5135,7 @@ save_Attr3fNV(GLenum attr, GLfloat x, GLfloat y, GLfloat z)
       n[4].f = z;
    }
 
-   ASSERT(attr < MAX_VERTEX_GENERIC_ATTRIBS);
+   assert(attr < MAX_VERTEX_GENERIC_ATTRIBS);
    ctx->ListState.ActiveAttribSize[attr] = 3;
    ASSIGN_4V(ctx->ListState.CurrentAttrib[attr], x, y, z, 1);
 
@@ -4953,7 +5159,7 @@ save_Attr4fNV(GLenum attr, GLfloat x, GLfloat y, GLfloat z, GLfloat w)
       n[5].f = w;
    }
 
-   ASSERT(attr < MAX_VERTEX_GENERIC_ATTRIBS);
+   assert(attr < MAX_VERTEX_GENERIC_ATTRIBS);
    ctx->ListState.ActiveAttribSize[attr] = 4;
    ASSIGN_4V(ctx->ListState.CurrentAttrib[attr], x, y, z, w);
 
@@ -4975,7 +5181,7 @@ save_Attr1fARB(GLenum attr, GLfloat x)
       n[2].f = x;
    }
 
-   ASSERT(attr < MAX_VERTEX_GENERIC_ATTRIBS);
+   assert(attr < MAX_VERTEX_GENERIC_ATTRIBS);
    ctx->ListState.ActiveAttribSize[attr] = 1;
    ASSIGN_4V(ctx->ListState.CurrentAttrib[attr], x, 0, 0, 1);
 
@@ -4997,7 +5203,7 @@ save_Attr2fARB(GLenum attr, GLfloat x, GLfloat y)
       n[3].f = y;
    }
 
-   ASSERT(attr < MAX_VERTEX_GENERIC_ATTRIBS);
+   assert(attr < MAX_VERTEX_GENERIC_ATTRIBS);
    ctx->ListState.ActiveAttribSize[attr] = 2;
    ASSIGN_4V(ctx->ListState.CurrentAttrib[attr], x, y, 0, 1);
 
@@ -5020,7 +5226,7 @@ save_Attr3fARB(GLenum attr, GLfloat x, GLfloat y, GLfloat z)
       n[4].f = z;
    }
 
-   ASSERT(attr < MAX_VERTEX_GENERIC_ATTRIBS);
+   assert(attr < MAX_VERTEX_GENERIC_ATTRIBS);
    ctx->ListState.ActiveAttribSize[attr] = 3;
    ASSIGN_4V(ctx->ListState.CurrentAttrib[attr], x, y, z, 1);
 
@@ -5044,7 +5250,7 @@ save_Attr4fARB(GLenum attr, GLfloat x, GLfloat y, GLfloat z, GLfloat w)
       n[5].f = w;
    }
 
-   ASSERT(attr < MAX_VERTEX_GENERIC_ATTRIBS);
+   assert(attr < MAX_VERTEX_GENERIC_ATTRIBS);
    ctx->ListState.ActiveAttribSize[attr] = 4;
    ASSIGN_4V(ctx->ListState.CurrentAttrib[attr], x, y, z, w);
 
@@ -5873,28 +6079,17 @@ save_Uniform4fARB(GLint location, GLfloat x, GLfloat y, GLfloat z, GLfloat w)
 }
 
 
-/** Return copy of memory */
-static void *
-memdup(const void *src, GLsizei bytes)
-{
-   void *b = bytes >= 0 ? malloc(bytes) : NULL;
-   if (b)
-      memcpy(b, src, bytes);
-   return b;
-}
-
-
 static void GLAPIENTRY
 save_Uniform1fvARB(GLint location, GLsizei count, const GLfloat *v)
 {
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_1FV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_1FV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 1 * sizeof(GLfloat));
+      save_pointer(&n[3], memdup(v, count * 1 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_Uniform1fv(ctx->Exec, (location, count, v));
@@ -5907,11 +6102,11 @@ save_Uniform2fvARB(GLint location, GLsizei count, const GLfloat *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_2FV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_2FV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 2 * sizeof(GLfloat));
+      save_pointer(&n[3], memdup(v, count * 2 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_Uniform2fv(ctx->Exec, (location, count, v));
@@ -5924,11 +6119,11 @@ save_Uniform3fvARB(GLint location, GLsizei count, const GLfloat *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_3FV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_3FV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 3 * sizeof(GLfloat));
+      save_pointer(&n[3], memdup(v, count * 3 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_Uniform3fv(ctx->Exec, (location, count, v));
@@ -5941,11 +6136,11 @@ save_Uniform4fvARB(GLint location, GLsizei count, const GLfloat *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_4FV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_4FV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 4 * sizeof(GLfloat));
+      save_pointer(&n[3], memdup(v, count * 4 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_Uniform4fv(ctx->Exec, (location, count, v));
@@ -6031,11 +6226,11 @@ save_Uniform1ivARB(GLint location, GLsizei count, const GLint *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_1IV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_1IV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 1 * sizeof(GLint));
+      save_pointer(&n[3], memdup(v, count * 1 * sizeof(GLint)));
    }
    if (ctx->ExecuteFlag) {
       CALL_Uniform1iv(ctx->Exec, (location, count, v));
@@ -6048,11 +6243,11 @@ save_Uniform2ivARB(GLint location, GLsizei count, const GLint *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_2IV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_2IV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 2 * sizeof(GLint));
+      save_pointer(&n[3], memdup(v, count * 2 * sizeof(GLint)));
    }
    if (ctx->ExecuteFlag) {
       CALL_Uniform2iv(ctx->Exec, (location, count, v));
@@ -6065,11 +6260,11 @@ save_Uniform3ivARB(GLint location, GLsizei count, const GLint *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_3IV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_3IV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 3 * sizeof(GLint));
+      save_pointer(&n[3], memdup(v, count * 3 * sizeof(GLint)));
    }
    if (ctx->ExecuteFlag) {
       CALL_Uniform3iv(ctx->Exec, (location, count, v));
@@ -6082,11 +6277,11 @@ save_Uniform4ivARB(GLint location, GLsizei count, const GLint *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_4IV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_4IV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 4 * sizeof(GLfloat));
+      save_pointer(&n[3], memdup(v, count * 4 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_Uniform4iv(ctx->Exec, (location, count, v));
@@ -6173,11 +6368,11 @@ save_Uniform1uiv(GLint location, GLsizei count, const GLuint *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_1UIV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_1UIV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 1 * sizeof(*v));
+      save_pointer(&n[3], memdup(v, count * 1 * sizeof(*v)));
    }
    if (ctx->ExecuteFlag) {
       /*CALL_Uniform1uiv(ctx->Exec, (location, count, v));*/
@@ -6190,11 +6385,11 @@ save_Uniform2uiv(GLint location, GLsizei count, const GLuint *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_2UIV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_2UIV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 2 * sizeof(*v));
+      save_pointer(&n[3], memdup(v, count * 2 * sizeof(*v)));
    }
    if (ctx->ExecuteFlag) {
       /*CALL_Uniform2uiv(ctx->Exec, (location, count, v));*/
@@ -6207,11 +6402,11 @@ save_Uniform3uiv(GLint location, GLsizei count, const GLuint *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_3UIV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_3UIV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 3 * sizeof(*v));
+      save_pointer(&n[3], memdup(v, count * 3 * sizeof(*v)));
    }
    if (ctx->ExecuteFlag) {
       /*CALL_Uniform3uiv(ctx->Exec, (location, count, v));*/
@@ -6224,11 +6419,11 @@ save_Uniform4uiv(GLint location, GLsizei count, const GLuint *v)
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_4UIV, 3);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_4UIV, 2 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
-      n[3].data = memdup(v, count * 4 * sizeof(*v));
+      save_pointer(&n[3], memdup(v, count * 4 * sizeof(*v)));
    }
    if (ctx->ExecuteFlag) {
       /*CALL_Uniform4uiv(ctx->Exec, (location, count, v));*/
@@ -6244,12 +6439,12 @@ save_UniformMatrix2fvARB(GLint location, GLsizei count, GLboolean transpose,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX22, 4);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX22, 3 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
       n[3].b = transpose;
-      n[4].data = memdup(m, count * 2 * 2 * sizeof(GLfloat));
+      save_pointer(&n[4], memdup(m, count * 2 * 2 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_UniformMatrix2fv(ctx->Exec, (location, count, transpose, m));
@@ -6263,12 +6458,12 @@ save_UniformMatrix3fvARB(GLint location, GLsizei count, GLboolean transpose,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX33, 4);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX33, 3 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
       n[3].b = transpose;
-      n[4].data = memdup(m, count * 3 * 3 * sizeof(GLfloat));
+      save_pointer(&n[4], memdup(m, count * 3 * 3 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_UniformMatrix3fv(ctx->Exec, (location, count, transpose, m));
@@ -6282,12 +6477,12 @@ save_UniformMatrix4fvARB(GLint location, GLsizei count, GLboolean transpose,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX44, 4);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX44, 3 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
       n[3].b = transpose;
-      n[4].data = memdup(m, count * 4 * 4 * sizeof(GLfloat));
+      save_pointer(&n[4], memdup(m, count * 4 * 4 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_UniformMatrix4fv(ctx->Exec, (location, count, transpose, m));
@@ -6302,12 +6497,12 @@ save_UniformMatrix2x3fv(GLint location, GLsizei count, GLboolean transpose,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX23, 4);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX23, 3 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
       n[3].b = transpose;
-      n[4].data = memdup(m, count * 2 * 3 * sizeof(GLfloat));
+      save_pointer(&n[4], memdup(m, count * 2 * 3 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_UniformMatrix2x3fv(ctx->Exec, (location, count, transpose, m));
@@ -6321,12 +6516,12 @@ save_UniformMatrix3x2fv(GLint location, GLsizei count, GLboolean transpose,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX32, 4);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX32, 3 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
       n[3].b = transpose;
-      n[4].data = memdup(m, count * 3 * 2 * sizeof(GLfloat));
+      save_pointer(&n[4], memdup(m, count * 3 * 2 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_UniformMatrix3x2fv(ctx->Exec, (location, count, transpose, m));
@@ -6341,12 +6536,12 @@ save_UniformMatrix2x4fv(GLint location, GLsizei count, GLboolean transpose,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX24, 4);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX24, 3 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
       n[3].b = transpose;
-      n[4].data = memdup(m, count * 2 * 4 * sizeof(GLfloat));
+      save_pointer(&n[4], memdup(m, count * 2 * 4 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_UniformMatrix2x4fv(ctx->Exec, (location, count, transpose, m));
@@ -6360,12 +6555,12 @@ save_UniformMatrix4x2fv(GLint location, GLsizei count, GLboolean transpose,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX42, 4);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX42, 3 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
       n[3].b = transpose;
-      n[4].data = memdup(m, count * 4 * 2 * sizeof(GLfloat));
+      save_pointer(&n[4], memdup(m, count * 4 * 2 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_UniformMatrix4x2fv(ctx->Exec, (location, count, transpose, m));
@@ -6380,12 +6575,12 @@ save_UniformMatrix3x4fv(GLint location, GLsizei count, GLboolean transpose,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX34, 4);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX34, 3 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
       n[3].b = transpose;
-      n[4].data = memdup(m, count * 3 * 4 * sizeof(GLfloat));
+      save_pointer(&n[4], memdup(m, count * 3 * 4 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_UniformMatrix3x4fv(ctx->Exec, (location, count, transpose, m));
@@ -6399,12 +6594,12 @@ save_UniformMatrix4x3fv(GLint location, GLsizei count, GLboolean transpose,
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX43, 4);
+   n = alloc_instruction(ctx, OPCODE_UNIFORM_MATRIX43, 3 + POINTER_DWORDS);
    if (n) {
       n[1].i = location;
       n[2].i = count;
       n[3].b = transpose;
-      n[4].data = memdup(m, count * 4 * 3 * sizeof(GLfloat));
+      save_pointer(&n[4], memdup(m, count * 4 * 3 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
       CALL_UniformMatrix4x3fv(ctx->Exec, (location, count, transpose, m));
@@ -6412,236 +6607,892 @@ save_UniformMatrix4x3fv(GLint location, GLsizei count, GLboolean transpose,
 }
 
 static void GLAPIENTRY
-save_ClampColorARB(GLenum target, GLenum clamp)
+save_UseProgramStages(GLuint pipeline, GLbitfield stages, GLuint program)
 {
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_CLAMP_COLOR, 2);
+   n = alloc_instruction(ctx, OPCODE_USE_PROGRAM_STAGES, 3);
    if (n) {
-      n[1].e = target;
-      n[2].e = clamp;
+      n[1].ui = pipeline;
+      n[2].ui = stages;
+      n[3].ui = program;
    }
    if (ctx->ExecuteFlag) {
-      CALL_ClampColor(ctx->Exec, (target, clamp));
+      CALL_UseProgramStages(ctx->Exec, (pipeline, stages, program));
    }
 }
 
 static void GLAPIENTRY
-save_UseShaderProgramEXT(GLenum type, GLuint program)
+save_ProgramUniform1f(GLuint program, GLint location, GLfloat x)
 {
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_USE_SHADER_PROGRAM_EXT, 2);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_1F, 3);
    if (n) {
-      n[1].ui = type;
-      n[2].ui = program;
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].f = x;
    }
    if (ctx->ExecuteFlag) {
-      CALL_UseShaderProgramEXT(ctx->Exec, (type, program));
+      CALL_ProgramUniform1f(ctx->Exec, (program, location, x));
    }
 }
 
 static void GLAPIENTRY
-save_ActiveProgramEXT(GLuint program)
+save_ProgramUniform2f(GLuint program, GLint location, GLfloat x, GLfloat y)
 {
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_ACTIVE_PROGRAM_EXT, 1);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_2F, 4);
    if (n) {
       n[1].ui = program;
+      n[2].i = location;
+      n[3].f = x;
+      n[4].f = y;
    }
    if (ctx->ExecuteFlag) {
-      CALL_ActiveProgramEXT(ctx->Exec, (program));
+      CALL_ProgramUniform2f(ctx->Exec, (program, location, x, y));
    }
 }
 
-/** GL_EXT_texture_integer */
 static void GLAPIENTRY
-save_ClearColorIi(GLint red, GLint green, GLint blue, GLint alpha)
+save_ProgramUniform3f(GLuint program, GLint location,
+                      GLfloat x, GLfloat y, GLfloat z)
 {
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_CLEARCOLOR_I, 4);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_3F, 5);
    if (n) {
-      n[1].i = red;
-      n[2].i = green;
-      n[3].i = blue;
-      n[4].i = alpha;
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].f = x;
+      n[4].f = y;
+      n[5].f = z;
    }
    if (ctx->ExecuteFlag) {
-      CALL_ClearColorIiEXT(ctx->Exec, (red, green, blue, alpha));
+      CALL_ProgramUniform3f(ctx->Exec, (program, location, x, y, z));
    }
 }
 
-/** GL_EXT_texture_integer */
 static void GLAPIENTRY
-save_ClearColorIui(GLuint red, GLuint green, GLuint blue, GLuint alpha)
+save_ProgramUniform4f(GLuint program, GLint location,
+                      GLfloat x, GLfloat y, GLfloat z, GLfloat w)
 {
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_CLEARCOLOR_UI, 4);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_4F, 6);
    if (n) {
-      n[1].ui = red;
-      n[2].ui = green;
-      n[3].ui = blue;
-      n[4].ui = alpha;
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].f = x;
+      n[4].f = y;
+      n[5].f = z;
+      n[6].f = w;
    }
    if (ctx->ExecuteFlag) {
-      CALL_ClearColorIuiEXT(ctx->Exec, (red, green, blue, alpha));
+      CALL_ProgramUniform4f(ctx->Exec, (program, location, x, y, z, w));
    }
 }
 
-/** GL_EXT_texture_integer */
 static void GLAPIENTRY
-save_TexParameterIiv(GLenum target, GLenum pname, const GLint *params)
+save_ProgramUniform1fv(GLuint program, GLint location, GLsizei count,
+                       const GLfloat *v)
 {
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_TEXPARAMETER_I, 6);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_1FV, 3 + POINTER_DWORDS);
    if (n) {
-      n[1].e = target;
-      n[2].e = pname;
-      n[3].i = params[0];
-      n[4].i = params[1];
-      n[5].i = params[2];
-      n[6].i = params[3];
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 1 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
-      CALL_TexParameterIiv(ctx->Exec, (target, pname, params));
+      CALL_ProgramUniform1fv(ctx->Exec, (program, location, count, v));
    }
 }
 
-/** GL_EXT_texture_integer */
 static void GLAPIENTRY
-save_TexParameterIuiv(GLenum target, GLenum pname, const GLuint *params)
+save_ProgramUniform2fv(GLuint program, GLint location, GLsizei count,
+                       const GLfloat *v)
 {
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_TEXPARAMETER_UI, 6);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_2FV, 3 + POINTER_DWORDS);
    if (n) {
-      n[1].e = target;
-      n[2].e = pname;
-      n[3].ui = params[0];
-      n[4].ui = params[1];
-      n[5].ui = params[2];
-      n[6].ui = params[3];
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 2 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
-      CALL_TexParameterIuiv(ctx->Exec, (target, pname, params));
+      CALL_ProgramUniform2fv(ctx->Exec, (program, location, count, v));
    }
 }
 
-/* GL_ARB_instanced_arrays */
 static void GLAPIENTRY
-save_VertexAttribDivisor(GLuint index, GLuint divisor)
+save_ProgramUniform3fv(GLuint program, GLint location, GLsizei count,
+                       const GLfloat *v)
 {
    GET_CURRENT_CONTEXT(ctx);
    Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_VERTEX_ATTRIB_DIVISOR, 2);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_3FV, 3 + POINTER_DWORDS);
    if (n) {
-      n[1].ui = index;
-      n[2].ui = divisor;
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 3 * sizeof(GLfloat)));
    }
    if (ctx->ExecuteFlag) {
-      CALL_VertexAttribDivisor(ctx->Exec, (index, divisor));
+      CALL_ProgramUniform3fv(ctx->Exec, (program, location, count, v));
    }
 }
 
-
-/* GL_NV_texture_barrier */
 static void GLAPIENTRY
-save_TextureBarrierNV(void)
+save_ProgramUniform4fv(GLuint program, GLint location, GLsizei count,
+                       const GLfloat *v)
 {
    GET_CURRENT_CONTEXT(ctx);
+   Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   alloc_instruction(ctx, OPCODE_TEXTURE_BARRIER_NV, 0);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_4FV, 3 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 4 * sizeof(GLfloat)));
+   }
    if (ctx->ExecuteFlag) {
-      CALL_TextureBarrierNV(ctx->Exec, ());
+      CALL_ProgramUniform4fv(ctx->Exec, (program, location, count, v));
    }
 }
 
-
-/* GL_ARB_sampler_objects */
 static void GLAPIENTRY
-save_BindSampler(GLuint unit, GLuint sampler)
+save_ProgramUniform1i(GLuint program, GLint location, GLint x)
 {
-   Node *n;
    GET_CURRENT_CONTEXT(ctx);
+   Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_BIND_SAMPLER, 2);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_1I, 3);
    if (n) {
-      n[1].ui = unit;
-      n[2].ui = sampler;
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = x;
    }
    if (ctx->ExecuteFlag) {
-      CALL_BindSampler(ctx->Exec, (unit, sampler));
+      CALL_ProgramUniform1i(ctx->Exec, (program, location, x));
    }
 }
 
 static void GLAPIENTRY
-save_SamplerParameteriv(GLuint sampler, GLenum pname, const GLint *params)
+save_ProgramUniform2i(GLuint program, GLint location, GLint x, GLint y)
 {
-   Node *n;
    GET_CURRENT_CONTEXT(ctx);
+   Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_SAMPLER_PARAMETERIV, 6);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_2I, 4);
    if (n) {
-      n[1].ui = sampler;
-      n[2].e = pname;
-      n[3].i = params[0];
-      if (pname == GL_TEXTURE_BORDER_COLOR) {
-         n[4].i = params[1];
-         n[5].i = params[2];
-         n[6].i = params[3];
-      }
-      else {
-         n[4].i = n[5].i = n[6].i = 0;
-      }
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = x;
+      n[4].i = y;
    }
    if (ctx->ExecuteFlag) {
-      CALL_SamplerParameteriv(ctx->Exec, (sampler, pname, params));
+      CALL_ProgramUniform2i(ctx->Exec, (program, location, x, y));
    }
 }
 
 static void GLAPIENTRY
-save_SamplerParameteri(GLuint sampler, GLenum pname, GLint param)
+save_ProgramUniform3i(GLuint program, GLint location,
+                      GLint x, GLint y, GLint z)
 {
-   GLint parray[4];
-   parray[0] = param;
-   parray[1] = parray[2] = parray[3] = 0;
-   save_SamplerParameteriv(sampler, pname, parray);
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_3I, 5);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = x;
+      n[4].i = y;
+      n[5].i = z;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform3i(ctx->Exec, (program, location, x, y, z));
+   }
 }
 
 static void GLAPIENTRY
-save_SamplerParameterfv(GLuint sampler, GLenum pname, const GLfloat *params)
+save_ProgramUniform4i(GLuint program, GLint location,
+                      GLint x, GLint y, GLint z, GLint w)
 {
-   Node *n;
    GET_CURRENT_CONTEXT(ctx);
+   Node *n;
    ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_SAMPLER_PARAMETERFV, 6);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_4I, 6);
    if (n) {
-      n[1].ui = sampler;
-      n[2].e = pname;
-      n[3].f = params[0];
-      if (pname == GL_TEXTURE_BORDER_COLOR) {
-         n[4].f = params[1];
-         n[5].f = params[2];
-         n[6].f = params[3];
-      }
-      else {
-         n[4].f = n[5].f = n[6].f = 0.0F;
-      }
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = x;
+      n[4].i = y;
+      n[5].i = z;
+      n[6].i = w;
    }
    if (ctx->ExecuteFlag) {
-      CALL_SamplerParameterfv(ctx->Exec, (sampler, pname, params));
+      CALL_ProgramUniform4i(ctx->Exec, (program, location, x, y, z, w));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform1iv(GLuint program, GLint location, GLsizei count,
+                       const GLint *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_1IV, 3 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 1 * sizeof(GLint)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform1iv(ctx->Exec, (program, location, count, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform2iv(GLuint program, GLint location, GLsizei count,
+                       const GLint *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_2IV, 3 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 2 * sizeof(GLint)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform2iv(ctx->Exec, (program, location, count, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform3iv(GLuint program, GLint location, GLsizei count,
+                       const GLint *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_3IV, 3 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 3 * sizeof(GLint)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform3iv(ctx->Exec, (program, location, count, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform4iv(GLuint program, GLint location, GLsizei count,
+                       const GLint *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_4IV, 3 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 4 * sizeof(GLint)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform4iv(ctx->Exec, (program, location, count, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform1ui(GLuint program, GLint location, GLuint x)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_1UI, 3);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].ui = x;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform1ui(ctx->Exec, (program, location, x));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform2ui(GLuint program, GLint location, GLuint x, GLuint y)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_2UI, 4);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].ui = x;
+      n[4].ui = y;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform2ui(ctx->Exec, (program, location, x, y));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform3ui(GLuint program, GLint location,
+                       GLuint x, GLuint y, GLuint z)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_3UI, 5);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].ui = x;
+      n[4].ui = y;
+      n[5].ui = z;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform3ui(ctx->Exec, (program, location, x, y, z));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform4ui(GLuint program, GLint location,
+                       GLuint x, GLuint y, GLuint z, GLuint w)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_4UI, 6);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].ui = x;
+      n[4].ui = y;
+      n[5].ui = z;
+      n[6].ui = w;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform4ui(ctx->Exec, (program, location, x, y, z, w));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform1uiv(GLuint program, GLint location, GLsizei count,
+                        const GLuint *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_1UIV, 3 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 1 * sizeof(GLuint)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform1uiv(ctx->Exec, (program, location, count, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform2uiv(GLuint program, GLint location, GLsizei count,
+                        const GLuint *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_2UIV, 3 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 2 * sizeof(GLuint)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform2uiv(ctx->Exec, (program, location, count, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform3uiv(GLuint program, GLint location, GLsizei count,
+                        const GLuint *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_3UIV, 3 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 3 * sizeof(GLuint)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform3uiv(ctx->Exec, (program, location, count, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniform4uiv(GLuint program, GLint location, GLsizei count,
+                        const GLuint *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_4UIV, 3 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      save_pointer(&n[4], memdup(v, count * 4 * sizeof(GLuint)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniform4uiv(ctx->Exec, (program, location, count, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniformMatrix2fv(GLuint program, GLint location, GLsizei count,
+                             GLboolean transpose, const GLfloat *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_MATRIX22F,
+                         4 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      n[4].b = transpose;
+      save_pointer(&n[5], memdup(v, count * 2 * 2 * sizeof(GLfloat)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniformMatrix2fv(ctx->Exec,
+                                   (program, location, count, transpose, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniformMatrix2x3fv(GLuint program, GLint location, GLsizei count,
+                               GLboolean transpose, const GLfloat *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_MATRIX23F,
+                         4 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      n[4].b = transpose;
+      save_pointer(&n[5], memdup(v, count * 2 * 3 * sizeof(GLfloat)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniformMatrix2x3fv(ctx->Exec,
+                                     (program, location, count, transpose, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniformMatrix2x4fv(GLuint program, GLint location, GLsizei count,
+                               GLboolean transpose, const GLfloat *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_MATRIX24F,
+                         4 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      n[4].b = transpose;
+      save_pointer(&n[5], memdup(v, count * 2 * 4 * sizeof(GLfloat)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniformMatrix2x4fv(ctx->Exec,
+                                     (program, location, count, transpose, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniformMatrix3x2fv(GLuint program, GLint location, GLsizei count,
+                               GLboolean transpose, const GLfloat *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_MATRIX32F,
+                         4 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      n[4].b = transpose;
+      save_pointer(&n[5], memdup(v, count * 3 * 2 * sizeof(GLfloat)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniformMatrix3x2fv(ctx->Exec,
+                                     (program, location, count, transpose, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniformMatrix3fv(GLuint program, GLint location, GLsizei count,
+                             GLboolean transpose, const GLfloat *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_MATRIX33F,
+                         4 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      n[4].b = transpose;
+      save_pointer(&n[5], memdup(v, count * 3 * 3 * sizeof(GLfloat)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniformMatrix3fv(ctx->Exec,
+                                   (program, location, count, transpose, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniformMatrix3x4fv(GLuint program, GLint location, GLsizei count,
+                               GLboolean transpose, const GLfloat *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_MATRIX34F,
+                         4 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      n[4].b = transpose;
+      save_pointer(&n[5], memdup(v, count * 3 * 4 * sizeof(GLfloat)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniformMatrix3x4fv(ctx->Exec,
+                                     (program, location, count, transpose, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniformMatrix4x2fv(GLuint program, GLint location, GLsizei count,
+                               GLboolean transpose, const GLfloat *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_MATRIX42F,
+                         4 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      n[4].b = transpose;
+      save_pointer(&n[5], memdup(v, count * 4 * 2 * sizeof(GLfloat)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniformMatrix4x2fv(ctx->Exec,
+                                     (program, location, count, transpose, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniformMatrix4x3fv(GLuint program, GLint location, GLsizei count,
+                               GLboolean transpose, const GLfloat *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_MATRIX43F,
+                         4 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      n[4].b = transpose;
+      save_pointer(&n[5], memdup(v, count * 4 * 3 * sizeof(GLfloat)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniformMatrix4x3fv(ctx->Exec,
+                                     (program, location, count, transpose, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ProgramUniformMatrix4fv(GLuint program, GLint location, GLsizei count,
+                             GLboolean transpose, const GLfloat *v)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_PROGRAM_UNIFORM_MATRIX44F,
+                         4 + POINTER_DWORDS);
+   if (n) {
+      n[1].ui = program;
+      n[2].i = location;
+      n[3].i = count;
+      n[4].b = transpose;
+      save_pointer(&n[5], memdup(v, count * 4 * 4 * sizeof(GLfloat)));
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ProgramUniformMatrix4fv(ctx->Exec,
+                                   (program, location, count, transpose, v));
+   }
+}
+
+static void GLAPIENTRY
+save_ClipControl(GLenum origin, GLenum depth)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_CLIP_CONTROL, 2);
+   if (n) {
+      n[1].e = origin;
+      n[2].e = depth;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ClipControl(ctx->Exec, (origin, depth));
+   }
+}
+
+static void GLAPIENTRY
+save_ClampColorARB(GLenum target, GLenum clamp)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_CLAMP_COLOR, 2);
+   if (n) {
+      n[1].e = target;
+      n[2].e = clamp;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ClampColor(ctx->Exec, (target, clamp));
+   }
+}
+
+/** GL_EXT_texture_integer */
+static void GLAPIENTRY
+save_ClearColorIi(GLint red, GLint green, GLint blue, GLint alpha)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_CLEARCOLOR_I, 4);
+   if (n) {
+      n[1].i = red;
+      n[2].i = green;
+      n[3].i = blue;
+      n[4].i = alpha;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ClearColorIiEXT(ctx->Exec, (red, green, blue, alpha));
+   }
+}
+
+/** GL_EXT_texture_integer */
+static void GLAPIENTRY
+save_ClearColorIui(GLuint red, GLuint green, GLuint blue, GLuint alpha)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_CLEARCOLOR_UI, 4);
+   if (n) {
+      n[1].ui = red;
+      n[2].ui = green;
+      n[3].ui = blue;
+      n[4].ui = alpha;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_ClearColorIuiEXT(ctx->Exec, (red, green, blue, alpha));
+   }
+}
+
+/** GL_EXT_texture_integer */
+static void GLAPIENTRY
+save_TexParameterIiv(GLenum target, GLenum pname, const GLint *params)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_TEXPARAMETER_I, 6);
+   if (n) {
+      n[1].e = target;
+      n[2].e = pname;
+      n[3].i = params[0];
+      n[4].i = params[1];
+      n[5].i = params[2];
+      n[6].i = params[3];
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_TexParameterIiv(ctx->Exec, (target, pname, params));
+   }
+}
+
+/** GL_EXT_texture_integer */
+static void GLAPIENTRY
+save_TexParameterIuiv(GLenum target, GLenum pname, const GLuint *params)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_TEXPARAMETER_UI, 6);
+   if (n) {
+      n[1].e = target;
+      n[2].e = pname;
+      n[3].ui = params[0];
+      n[4].ui = params[1];
+      n[5].ui = params[2];
+      n[6].ui = params[3];
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_TexParameterIuiv(ctx->Exec, (target, pname, params));
+   }
+}
+
+/* GL_ARB_instanced_arrays */
+static void GLAPIENTRY
+save_VertexAttribDivisor(GLuint index, GLuint divisor)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   Node *n;
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_VERTEX_ATTRIB_DIVISOR, 2);
+   if (n) {
+      n[1].ui = index;
+      n[2].ui = divisor;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_VertexAttribDivisor(ctx->Exec, (index, divisor));
+   }
+}
+
+
+/* GL_NV_texture_barrier */
+static void GLAPIENTRY
+save_TextureBarrierNV(void)
+{
+   GET_CURRENT_CONTEXT(ctx);
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   alloc_instruction(ctx, OPCODE_TEXTURE_BARRIER_NV, 0);
+   if (ctx->ExecuteFlag) {
+      CALL_TextureBarrierNV(ctx->Exec, ());
+   }
+}
+
+
+/* GL_ARB_sampler_objects */
+static void GLAPIENTRY
+save_BindSampler(GLuint unit, GLuint sampler)
+{
+   Node *n;
+   GET_CURRENT_CONTEXT(ctx);
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_BIND_SAMPLER, 2);
+   if (n) {
+      n[1].ui = unit;
+      n[2].ui = sampler;
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_BindSampler(ctx->Exec, (unit, sampler));
+   }
+}
+
+static void GLAPIENTRY
+save_SamplerParameteriv(GLuint sampler, GLenum pname, const GLint *params)
+{
+   Node *n;
+   GET_CURRENT_CONTEXT(ctx);
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_SAMPLER_PARAMETERIV, 6);
+   if (n) {
+      n[1].ui = sampler;
+      n[2].e = pname;
+      n[3].i = params[0];
+      if (pname == GL_TEXTURE_BORDER_COLOR) {
+         n[4].i = params[1];
+         n[5].i = params[2];
+         n[6].i = params[3];
+      }
+      else {
+         n[4].i = n[5].i = n[6].i = 0;
+      }
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_SamplerParameteriv(ctx->Exec, (sampler, pname, params));
+   }
+}
+
+static void GLAPIENTRY
+save_SamplerParameteri(GLuint sampler, GLenum pname, GLint param)
+{
+   GLint parray[4];
+   parray[0] = param;
+   parray[1] = parray[2] = parray[3] = 0;
+   save_SamplerParameteriv(sampler, pname, parray);
+}
+
+static void GLAPIENTRY
+save_SamplerParameterfv(GLuint sampler, GLenum pname, const GLfloat *params)
+{
+   Node *n;
+   GET_CURRENT_CONTEXT(ctx);
+   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
+   n = alloc_instruction(ctx, OPCODE_SAMPLER_PARAMETERFV, 6);
+   if (n) {
+      n[1].ui = sampler;
+      n[2].e = pname;
+      n[3].f = params[0];
+      if (pname == GL_TEXTURE_BORDER_COLOR) {
+         n[4].f = params[1];
+         n[5].f = params[2];
+         n[6].f = params[3];
+      }
+      else {
+         n[4].f = n[5].f = n[6].f = 0.0F;
+      }
+   }
+   if (ctx->ExecuteFlag) {
+      CALL_SamplerParameterfv(ctx->Exec, (sampler, pname, params));
    }
 }
 
@@ -6741,28 +7592,6 @@ save_FramebufferTexture(GLenum target, GLenum attachment,
    }
 }
 
-static void GLAPIENTRY
-save_FramebufferTextureFace(GLenum target, GLenum attachment,
-                            GLuint texture, GLint level, GLenum face)
-{
-   Node *n;
-   GET_CURRENT_CONTEXT(ctx);
-   ASSERT_OUTSIDE_SAVE_BEGIN_END_AND_FLUSH(ctx);
-   n = alloc_instruction(ctx, OPCODE_FRAMEBUFFER_TEXTURE_FACE, 5);
-   if (n) {
-      n[1].e = target;
-      n[2].e = attachment;
-      n[3].ui = texture;
-      n[4].i = level;
-      n[5].e = face;
-   }
-   if (ctx->ExecuteFlag) {
-      CALL_FramebufferTextureFaceARB(ctx->Exec, (target, attachment, texture,
-                                                 level, face));
-   }
-}
-
-
 
 static void GLAPIENTRY
 save_WaitSync(GLsync sync, GLbitfield flags, GLuint64 timeout)
@@ -6774,10 +7603,10 @@ save_WaitSync(GLsync sync, GLbitfield flags, GLuint64 timeout)
    if (n) {
       union uint64_pair p;
       p.uint64 = timeout;
-      n[1].data = sync;
-      n[2].bf = flags;
-      n[3].ui = p.uint32[0];
-      n[4].ui = p.uint32[1];
+      n[1].bf = flags;
+      n[2].ui = p.uint32[0];
+      n[3].ui = p.uint32[1];
+      save_pointer(&n[4], sync);
    }
    if (ctx->ExecuteFlag) {
       CALL_WaitSync(ctx->Exec, (sync, flags, timeout));
@@ -6841,10 +7670,10 @@ static void
 save_error(struct gl_context *ctx, GLenum error, const char *s)
 {
    Node *n;
-   n = alloc_instruction(ctx, OPCODE_ERROR, 2);
+   n = alloc_instruction(ctx, OPCODE_ERROR, 1 + POINTER_DWORDS);
    if (n) {
       n[1].e = error;
-      n[2].data = (void *) s;
+      save_pointer(&n[2], (void *) s);
       /* note: the data/string here doesn't have to be freed in
        * _mesa_delete_list() since the string is never dynamically
        * allocated.
@@ -6929,7 +7758,7 @@ execute_list(struct gl_context *ctx, GLuint list)
       else {
          switch (opcode) {
          case OPCODE_ERROR:
-            _mesa_error(ctx, n[1].e, "%s", (const char *) n[2].data);
+            _mesa_error(ctx, n[1].e, "%s", (const char *) get_pointer(&n[2]));
             break;
          case OPCODE_ACCUM:
             CALL_Accum(ctx->Exec, (n[1].e, n[2].f));
@@ -6946,7 +7775,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                ctx->Unpack = ctx->DefaultPacking;
                CALL_Bitmap(ctx->Exec, ((GLsizei) n[1].i, (GLsizei) n[2].i,
                                        n[3].f, n[4].f, n[5].f, n[6].f,
-                                       (const GLubyte *) n[7].data));
+                                       get_pointer(&n[7])));
                ctx->Unpack = save;      /* restore */
             }
             break;
@@ -7125,7 +7954,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                const struct gl_pixelstore_attrib save = ctx->Unpack;
                ctx->Unpack = ctx->DefaultPacking;
                CALL_DrawPixels(ctx->Exec, (n[1].i, n[2].i, n[3].e, n[4].e,
-                                           n[5].data));
+                                           get_pointer(&n[5])));
                ctx->Unpack = save;      /* restore */
             }
             break;
@@ -7201,17 +8030,8 @@ execute_list(struct gl_context *ctx, GLuint list)
             CALL_LoadIdentity(ctx->Exec, ());
             break;
          case OPCODE_LOAD_MATRIX:
-            if (sizeof(Node) == sizeof(GLfloat)) {
-               CALL_LoadMatrixf(ctx->Exec, (&n[1].f));
-            }
-            else {
-               GLfloat m[16];
-               GLuint i;
-               for (i = 0; i < 16; i++) {
-                  m[i] = n[1 + i].f;
-               }
-               CALL_LoadMatrixf(ctx->Exec, (m));
-            }
+            STATIC_ASSERT(sizeof(Node) == sizeof(GLfloat));
+            CALL_LoadMatrixf(ctx->Exec, (&n[1].f));
             break;
          case OPCODE_LOAD_NAME:
             CALL_LoadName(ctx->Exec, (n[1].ui));
@@ -7227,7 +8047,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                GLfloat u1 = n[2].f;
                GLfloat u2 = n[3].f;
                CALL_Map1f(ctx->Exec, (target, u1, u2, ustride, uorder,
-                                      (GLfloat *) n[6].data));
+                                      (GLfloat *) get_pointer(&n[6])));
             }
             break;
          case OPCODE_MAP2:
@@ -7243,7 +8063,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                GLint vorder = n[9].i;
                CALL_Map2f(ctx->Exec, (target, u1, u2, ustride, uorder,
                                       v1, v2, vstride, vorder,
-                                      (GLfloat *) n[10].data));
+                                      (GLfloat *) get_pointer(&n[10])));
             }
             break;
          case OPCODE_MAPGRID1:
@@ -7257,17 +8077,7 @@ execute_list(struct gl_context *ctx, GLuint list)
             CALL_MatrixMode(ctx->Exec, (n[1].e));
             break;
          case OPCODE_MULT_MATRIX:
-            if (sizeof(Node) == sizeof(GLfloat)) {
-               CALL_MultMatrixf(ctx->Exec, (&n[1].f));
-            }
-            else {
-               GLfloat m[16];
-               GLuint i;
-               for (i = 0; i < 16; i++) {
-                  m[i] = n[1 + i].f;
-               }
-               CALL_MultMatrixf(ctx->Exec, (m));
-            }
+            CALL_MultMatrixf(ctx->Exec, (&n[1].f));
             break;
          case OPCODE_ORTHO:
             CALL_Ortho(ctx->Exec,
@@ -7278,7 +8088,7 @@ execute_list(struct gl_context *ctx, GLuint list)
             break;
          case OPCODE_PIXEL_MAP:
             CALL_PixelMapfv(ctx->Exec,
-                            (n[1].e, n[2].i, (GLfloat *) n[3].data));
+                            (n[1].e, n[2].i, get_pointer(&n[3])));
             break;
          case OPCODE_PIXEL_TRANSFER:
             CALL_PixelTransferf(ctx->Exec, (n[1].e, n[2].f));
@@ -7305,13 +8115,16 @@ execute_list(struct gl_context *ctx, GLuint list)
             {
                const struct gl_pixelstore_attrib save = ctx->Unpack;
                ctx->Unpack = ctx->DefaultPacking;
-               CALL_PolygonStipple(ctx->Exec, ((GLubyte *) n[1].data));
+               CALL_PolygonStipple(ctx->Exec, (get_pointer(&n[1])));
                ctx->Unpack = save;      /* restore */
             }
             break;
          case OPCODE_POLYGON_OFFSET:
             CALL_PolygonOffset(ctx->Exec, (n[1].f, n[2].f));
             break;
+         case OPCODE_POLYGON_OFFSET_CLAMP:
+            CALL_PolygonOffsetClampEXT(ctx->Exec, (n[1].f, n[2].f, n[3].f));
+            break;
          case OPCODE_POP_ATTRIB:
             CALL_PopAttrib(ctx->Exec, ());
             break;
@@ -7415,7 +8228,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                                            n[5].e,      /* border */
                                            n[6].e,      /* format */
                                            n[7].e,      /* type */
-                                           n[8].data));
+                                           get_pointer(&n[8])));
                ctx->Unpack = save;      /* restore */
             }
             break;
@@ -7431,7 +8244,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                                            n[6].e,      /* border */
                                            n[7].e,      /* format */
                                            n[8].e,      /* type */
-                                           n[9].data));
+                                           get_pointer(&n[9])));
                ctx->Unpack = save;      /* restore */
             }
             break;
@@ -7448,7 +8261,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                                            n[7].e,      /* border */
                                            n[8].e,      /* format */
                                            n[9].e,      /* type */
-                                           n[10].data));
+                                           get_pointer(&n[10])));
                ctx->Unpack = save;      /* restore */
             }
             break;
@@ -7458,7 +8271,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                ctx->Unpack = ctx->DefaultPacking;
                CALL_TexSubImage1D(ctx->Exec, (n[1].e, n[2].i, n[3].i,
                                               n[4].i, n[5].e,
-                                              n[6].e, n[7].data));
+                                              n[6].e, get_pointer(&n[7])));
                ctx->Unpack = save;      /* restore */
             }
             break;
@@ -7469,7 +8282,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                CALL_TexSubImage2D(ctx->Exec, (n[1].e, n[2].i, n[3].i,
                                               n[4].i, n[5].e,
                                               n[6].i, n[7].e, n[8].e,
-                                              n[9].data));
+                                              get_pointer(&n[9])));
                ctx->Unpack = save;      /* restore */
             }
             break;
@@ -7480,7 +8293,7 @@ execute_list(struct gl_context *ctx, GLuint list)
                CALL_TexSubImage3D(ctx->Exec, (n[1].e, n[2].i, n[3].i,
                                               n[4].i, n[5].i, n[6].i, n[7].i,
                                               n[8].i, n[9].e, n[10].e,
-                                              n[11].data));
+                                              get_pointer(&n[11])));
                ctx->Unpack = save;      /* restore */
             }
             break;
@@ -7499,36 +8312,38 @@ execute_list(struct gl_context *ctx, GLuint list)
             break;
          case OPCODE_COMPRESSED_TEX_IMAGE_1D:  /* GL_ARB_texture_compression */
             CALL_CompressedTexImage1D(ctx->Exec, (n[1].e, n[2].i, n[3].e,
-                                                     n[4].i, n[5].i, n[6].i,
-                                                     n[7].data));
+                                                  n[4].i, n[5].i, n[6].i,
+                                                  get_pointer(&n[7])));
             break;
          case OPCODE_COMPRESSED_TEX_IMAGE_2D:  /* GL_ARB_texture_compression */
             CALL_CompressedTexImage2D(ctx->Exec, (n[1].e, n[2].i, n[3].e,
-                                                     n[4].i, n[5].i, n[6].i,
-                                                     n[7].i, n[8].data));
+                                                  n[4].i, n[5].i, n[6].i,
+                                                  n[7].i, get_pointer(&n[8])));
             break;
          case OPCODE_COMPRESSED_TEX_IMAGE_3D:  /* GL_ARB_texture_compression */
             CALL_CompressedTexImage3D(ctx->Exec, (n[1].e, n[2].i, n[3].e,
-                                                     n[4].i, n[5].i, n[6].i,
-                                                     n[7].i, n[8].i,
-                                                     n[9].data));
+                                                  n[4].i, n[5].i, n[6].i,
+                                                  n[7].i, n[8].i,
+                                                  get_pointer(&n[9])));
             break;
          case OPCODE_COMPRESSED_TEX_SUB_IMAGE_1D:      /* GL_ARB_texture_compress */
             CALL_CompressedTexSubImage1D(ctx->Exec,
                                             (n[1].e, n[2].i, n[3].i, n[4].i,
-                                             n[5].e, n[6].i, n[7].data));
+                                             n[5].e, n[6].i,
+                                             get_pointer(&n[7])));
             break;
          case OPCODE_COMPRESSED_TEX_SUB_IMAGE_2D:      /* GL_ARB_texture_compress */
             CALL_CompressedTexSubImage2D(ctx->Exec,
                                             (n[1].e, n[2].i, n[3].i, n[4].i,
                                              n[5].i, n[6].i, n[7].e, n[8].i,
-                                             n[9].data));
+                                             get_pointer(&n[9])));
             break;
          case OPCODE_COMPRESSED_TEX_SUB_IMAGE_3D:      /* GL_ARB_texture_compress */
             CALL_CompressedTexSubImage3D(ctx->Exec,
                                             (n[1].e, n[2].i, n[3].i, n[4].i,
                                              n[5].i, n[6].i, n[7].i, n[8].i,
-                                             n[9].e, n[10].i, n[11].data));
+                                             n[9].e, n[10].i,
+                                             get_pointer(&n[11])));
             break;
          case OPCODE_SAMPLE_COVERAGE:  /* GL_ARB_multisample */
             CALL_SampleCoverage(ctx->Exec, (n[1].f, n[2].b));
@@ -7552,7 +8367,8 @@ execute_list(struct gl_context *ctx, GLuint list)
             break;
          case OPCODE_PROGRAM_STRING_ARB:
             CALL_ProgramStringARB(ctx->Exec,
-                                  (n[1].e, n[2].e, n[3].i, n[4].data));
+                                  (n[1].e, n[2].e, n[3].i,
+                                   get_pointer(&n[4])));
             break;
          case OPCODE_PROGRAM_ENV_PARAMETER_ARB:
             CALL_ProgramEnvParameter4fARB(ctx->Exec, (n[1].e, n[2].ui, n[3].f,
@@ -7591,12 +8407,6 @@ execute_list(struct gl_context *ctx, GLuint list)
         case OPCODE_USE_PROGRAM:
            CALL_UseProgram(ctx->Exec, (n[1].ui));
            break;
-        case OPCODE_USE_SHADER_PROGRAM_EXT:
-           CALL_UseShaderProgramEXT(ctx->Exec, (n[1].ui, n[2].ui));
-           break;
-        case OPCODE_ACTIVE_PROGRAM_EXT:
-           CALL_ActiveProgramEXT(ctx->Exec, (n[1].ui));
-           break;
         case OPCODE_UNIFORM_1F:
            CALL_Uniform1f(ctx->Exec, (n[1].i, n[2].f));
            break;
@@ -7611,16 +8421,16 @@ execute_list(struct gl_context *ctx, GLuint list)
                               (n[1].i, n[2].f, n[3].f, n[4].f, n[5].f));
            break;
         case OPCODE_UNIFORM_1FV:
-           CALL_Uniform1fv(ctx->Exec, (n[1].i, n[2].i, n[3].data));
+           CALL_Uniform1fv(ctx->Exec, (n[1].i, n[2].i, get_pointer(&n[3])));
            break;
         case OPCODE_UNIFORM_2FV:
-           CALL_Uniform2fv(ctx->Exec, (n[1].i, n[2].i, n[3].data));
+           CALL_Uniform2fv(ctx->Exec, (n[1].i, n[2].i, get_pointer(&n[3])));
            break;
         case OPCODE_UNIFORM_3FV:
-           CALL_Uniform3fv(ctx->Exec, (n[1].i, n[2].i, n[3].data));
+           CALL_Uniform3fv(ctx->Exec, (n[1].i, n[2].i, get_pointer(&n[3])));
            break;
         case OPCODE_UNIFORM_4FV:
-           CALL_Uniform4fv(ctx->Exec, (n[1].i, n[2].i, n[3].data));
+           CALL_Uniform4fv(ctx->Exec, (n[1].i, n[2].i, get_pointer(&n[3])));
            break;
         case OPCODE_UNIFORM_1I:
            CALL_Uniform1i(ctx->Exec, (n[1].i, n[2].i));
@@ -7636,16 +8446,16 @@ execute_list(struct gl_context *ctx, GLuint list)
                               (n[1].i, n[2].i, n[3].i, n[4].i, n[5].i));
            break;
         case OPCODE_UNIFORM_1IV:
-           CALL_Uniform1iv(ctx->Exec, (n[1].i, n[2].i, n[3].data));
+           CALL_Uniform1iv(ctx->Exec, (n[1].i, n[2].i, get_pointer(&n[3])));
            break;
         case OPCODE_UNIFORM_2IV:
-           CALL_Uniform2iv(ctx->Exec, (n[1].i, n[2].i, n[3].data));
+           CALL_Uniform2iv(ctx->Exec, (n[1].i, n[2].i, get_pointer(&n[3])));
            break;
         case OPCODE_UNIFORM_3IV:
-           CALL_Uniform3iv(ctx->Exec, (n[1].i, n[2].i, n[3].data));
+           CALL_Uniform3iv(ctx->Exec, (n[1].i, n[2].i, get_pointer(&n[3])));
            break;
         case OPCODE_UNIFORM_4IV:
-           CALL_Uniform4iv(ctx->Exec, (n[1].i, n[2].i, n[3].data));
+           CALL_Uniform4iv(ctx->Exec, (n[1].i, n[2].i, get_pointer(&n[3])));
            break;
         case OPCODE_UNIFORM_1UI:
            /*CALL_Uniform1uiARB(ctx->Exec, (n[1].i, n[2].i));*/
@@ -7662,150 +8472,239 @@ execute_list(struct gl_context *ctx, GLuint list)
             */
            break;
         case OPCODE_UNIFORM_1UIV:
-           /*CALL_Uniform1uivARB(ctx->Exec, (n[1].i, n[2].i, n[3].data));*/
+           /*CALL_Uniform1uivARB(ctx->Exec, (n[1].i, n[2].i,
+                                              get_pointer(&n[3])));*/
            break;
         case OPCODE_UNIFORM_2UIV:
-           /*CALL_Uniform2uivARB(ctx->Exec, (n[1].i, n[2].i, n[3].data));*/
+           /*CALL_Uniform2uivARB(ctx->Exec, (n[1].i, n[2].i,
+                                              get_pointer(&n[3])));*/
            break;
         case OPCODE_UNIFORM_3UIV:
-           /*CALL_Uniform3uivARB(ctx->Exec, (n[1].i, n[2].i, n[3].data));*/
+           /*CALL_Uniform3uivARB(ctx->Exec, (n[1].i, n[2].i,
+                                              get_pointer(&n[3])));*/
            break;
         case OPCODE_UNIFORM_4UIV:
-           /*CALL_Uniform4uivARB(ctx->Exec, (n[1].i, n[2].i, n[3].data));*/
+           /*CALL_Uniform4uivARB(ctx->Exec, (n[1].i, n[2].i,
+                                              get_pointer(&n[3])));*/
            break;
         case OPCODE_UNIFORM_MATRIX22:
            CALL_UniformMatrix2fv(ctx->Exec,
-                                     (n[1].i, n[2].i, n[3].b, n[4].data));
+                                  (n[1].i, n[2].i, n[3].b, get_pointer(&n[4])));
            break;
         case OPCODE_UNIFORM_MATRIX33:
            CALL_UniformMatrix3fv(ctx->Exec,
-                                     (n[1].i, n[2].i, n[3].b, n[4].data));
+                                  (n[1].i, n[2].i, n[3].b, get_pointer(&n[4])));
            break;
         case OPCODE_UNIFORM_MATRIX44:
            CALL_UniformMatrix4fv(ctx->Exec,
-                                     (n[1].i, n[2].i, n[3].b, n[4].data));
+                                  (n[1].i, n[2].i, n[3].b, get_pointer(&n[4])));
            break;
         case OPCODE_UNIFORM_MATRIX23:
            CALL_UniformMatrix2x3fv(ctx->Exec,
-                                    (n[1].i, n[2].i, n[3].b, n[4].data));
+                                    (n[1].i, n[2].i, n[3].b, get_pointer(&n[4])));
            break;
         case OPCODE_UNIFORM_MATRIX32:
            CALL_UniformMatrix3x2fv(ctx->Exec,
-                                    (n[1].i, n[2].i, n[3].b, n[4].data));
+                                    (n[1].i, n[2].i, n[3].b, get_pointer(&n[4])));
            break;
         case OPCODE_UNIFORM_MATRIX24:
            CALL_UniformMatrix2x4fv(ctx->Exec,
-                                    (n[1].i, n[2].i, n[3].b, n[4].data));
+                                    (n[1].i, n[2].i, n[3].b, get_pointer(&n[4])));
            break;
         case OPCODE_UNIFORM_MATRIX42:
            CALL_UniformMatrix4x2fv(ctx->Exec,
-                                    (n[1].i, n[2].i, n[3].b, n[4].data));
+                                    (n[1].i, n[2].i, n[3].b, get_pointer(&n[4])));
            break;
         case OPCODE_UNIFORM_MATRIX34:
            CALL_UniformMatrix3x4fv(ctx->Exec,
-                                    (n[1].i, n[2].i, n[3].b, n[4].data));
+                                    (n[1].i, n[2].i, n[3].b, get_pointer(&n[4])));
            break;
         case OPCODE_UNIFORM_MATRIX43:
            CALL_UniformMatrix4x3fv(ctx->Exec,
-                                    (n[1].i, n[2].i, n[3].b, n[4].data));
+                                    (n[1].i, n[2].i, n[3].b, get_pointer(&n[4])));
            break;
 
-         case OPCODE_CLAMP_COLOR:
-            CALL_ClampColor(ctx->Exec, (n[1].e, n[2].e));
+        case OPCODE_USE_PROGRAM_STAGES:
+           CALL_UseProgramStages(ctx->Exec, (n[1].ui, n[2].ui, n[3].ui));
+           break;
+         case OPCODE_PROGRAM_UNIFORM_1F:
+            CALL_ProgramUniform1f(ctx->Exec, (n[1].ui, n[2].i, n[3].f));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_2F:
+            CALL_ProgramUniform2f(ctx->Exec, (n[1].ui, n[2].i, n[3].f, n[4].f));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_3F:
+            CALL_ProgramUniform3f(ctx->Exec, (n[1].ui, n[2].i,
+                                              n[3].f, n[4].f, n[5].f));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_4F:
+            CALL_ProgramUniform4f(ctx->Exec, (n[1].ui, n[2].i,
+                                              n[3].f, n[4].f, n[5].f, n[6].f));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_1FV:
+            CALL_ProgramUniform1fv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                               get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_2FV:
+            CALL_ProgramUniform2fv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                               get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_3FV:
+            CALL_ProgramUniform3fv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                               get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_4FV:
+            CALL_ProgramUniform4fv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                               get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_1I:
+            CALL_ProgramUniform1i(ctx->Exec, (n[1].ui, n[2].i, n[3].i));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_2I:
+            CALL_ProgramUniform2i(ctx->Exec, (n[1].ui, n[2].i, n[3].i, n[4].i));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_3I:
+            CALL_ProgramUniform3i(ctx->Exec, (n[1].ui, n[2].i,
+                                              n[3].i, n[4].i, n[5].i));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_4I:
+            CALL_ProgramUniform4i(ctx->Exec, (n[1].ui, n[2].i,
+                                              n[3].i, n[4].i, n[5].i, n[6].i));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_1IV:
+            CALL_ProgramUniform1iv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                               get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_2IV:
+            CALL_ProgramUniform2iv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                               get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_3IV:
+            CALL_ProgramUniform3iv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                               get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_4IV:
+            CALL_ProgramUniform4iv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                               get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_1UI:
+            CALL_ProgramUniform1ui(ctx->Exec, (n[1].ui, n[2].i, n[3].ui));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_2UI:
+            CALL_ProgramUniform2ui(ctx->Exec, (n[1].ui, n[2].i,
+                                               n[3].ui, n[4].ui));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_3UI:
+            CALL_ProgramUniform3ui(ctx->Exec, (n[1].ui, n[2].i,
+                                               n[3].ui, n[4].ui, n[5].ui));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_4UI:
+            CALL_ProgramUniform4ui(ctx->Exec, (n[1].ui, n[2].i,
+                                               n[3].ui,
+                                               n[4].ui, n[5].ui, n[6].ui));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_1UIV:
+            CALL_ProgramUniform1uiv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                                get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_2UIV:
+            CALL_ProgramUniform2uiv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                                get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_3UIV:
+            CALL_ProgramUniform3uiv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                                get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_4UIV:
+            CALL_ProgramUniform4uiv(ctx->Exec, (n[1].ui, n[2].i, n[3].i,
+                                                get_pointer(&n[4])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX22F:
+            CALL_ProgramUniformMatrix2fv(ctx->Exec,
+                                         (n[1].ui, n[2].i, n[3].i, n[4].b,
+                                          get_pointer(&n[5])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX23F:
+            CALL_ProgramUniformMatrix2x3fv(ctx->Exec,
+                                           (n[1].ui, n[2].i, n[3].i, n[4].b,
+                                            get_pointer(&n[5])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX24F:
+            CALL_ProgramUniformMatrix2x4fv(ctx->Exec,
+                                           (n[1].ui, n[2].i, n[3].i, n[4].b,
+                                            get_pointer(&n[5])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX32F:
+            CALL_ProgramUniformMatrix3x2fv(ctx->Exec,
+                                           (n[1].ui, n[2].i, n[3].i, n[4].b,
+                                            get_pointer(&n[5])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX33F:
+            CALL_ProgramUniformMatrix3fv(ctx->Exec,
+                                         (n[1].ui, n[2].i, n[3].i, n[4].b,
+                                          get_pointer(&n[5])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX34F:
+            CALL_ProgramUniformMatrix3x4fv(ctx->Exec,
+                                           (n[1].ui, n[2].i, n[3].i, n[4].b,
+                                            get_pointer(&n[5])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX42F:
+            CALL_ProgramUniformMatrix4x2fv(ctx->Exec,
+                                           (n[1].ui, n[2].i, n[3].i, n[4].b,
+                                            get_pointer(&n[5])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX43F:
+            CALL_ProgramUniformMatrix4x3fv(ctx->Exec,
+                                           (n[1].ui, n[2].i, n[3].i, n[4].b,
+                                            get_pointer(&n[5])));
+            break;
+         case OPCODE_PROGRAM_UNIFORM_MATRIX44F:
+            CALL_ProgramUniformMatrix4fv(ctx->Exec,
+                                         (n[1].ui, n[2].i, n[3].i, n[4].b,
+                                          get_pointer(&n[5])));
             break;
 
-         case OPCODE_TEX_BUMP_PARAMETER_ATI:
-            {
-               GLfloat values[4];
-               GLuint i, pname = n[1].ui;
+         case OPCODE_CLIP_CONTROL:
+            CALL_ClipControl(ctx->Exec, (n[1].e, n[2].e));
+            break;
 
-               for (i = 0; i < 4; i++)
-                  values[i] = n[1 + i].f;
-               CALL_TexBumpParameterfvATI(ctx->Exec, (pname, values));
-            }
+         case OPCODE_CLAMP_COLOR:
+            CALL_ClampColor(ctx->Exec, (n[1].e, n[2].e));
             break;
+
          case OPCODE_BIND_FRAGMENT_SHADER_ATI:
             CALL_BindFragmentShaderATI(ctx->Exec, (n[1].i));
             break;
          case OPCODE_SET_FRAGMENT_SHADER_CONSTANTS_ATI:
-            {
-               GLfloat values[4];
-               GLuint i, dst = n[1].ui;
-
-               for (i = 0; i < 4; i++)
-                  values[i] = n[1 + i].f;
-               CALL_SetFragmentShaderConstantATI(ctx->Exec, (dst, values));
-            }
+            CALL_SetFragmentShaderConstantATI(ctx->Exec, (n[1].ui, &n[2].f));
             break;
          case OPCODE_ATTR_1F_NV:
             CALL_VertexAttrib1fNV(ctx->Exec, (n[1].e, n[2].f));
             break;
          case OPCODE_ATTR_2F_NV:
-            /* Really shouldn't have to do this - the Node structure
-             * is convenient, but it would be better to store the data
-             * packed appropriately so that it can be sent directly
-             * on.  With x86_64 becoming common, this will start to
-             * matter more.
-             */
-            if (sizeof(Node) == sizeof(GLfloat))
-               CALL_VertexAttrib2fvNV(ctx->Exec, (n[1].e, &n[2].f));
-            else
-               CALL_VertexAttrib2fNV(ctx->Exec, (n[1].e, n[2].f, n[3].f));
+            CALL_VertexAttrib2fvNV(ctx->Exec, (n[1].e, &n[2].f));
             break;
          case OPCODE_ATTR_3F_NV:
-            if (sizeof(Node) == sizeof(GLfloat))
-               CALL_VertexAttrib3fvNV(ctx->Exec, (n[1].e, &n[2].f));
-            else
-               CALL_VertexAttrib3fNV(ctx->Exec, (n[1].e, n[2].f, n[3].f,
-                                                 n[4].f));
+            CALL_VertexAttrib3fvNV(ctx->Exec, (n[1].e, &n[2].f));
             break;
          case OPCODE_ATTR_4F_NV:
-            if (sizeof(Node) == sizeof(GLfloat))
-               CALL_VertexAttrib4fvNV(ctx->Exec, (n[1].e, &n[2].f));
-            else
-               CALL_VertexAttrib4fNV(ctx->Exec, (n[1].e, n[2].f, n[3].f,
-                                                 n[4].f, n[5].f));
+            CALL_VertexAttrib4fvNV(ctx->Exec, (n[1].e, &n[2].f));
             break;
          case OPCODE_ATTR_1F_ARB:
             CALL_VertexAttrib1fARB(ctx->Exec, (n[1].e, n[2].f));
             break;
          case OPCODE_ATTR_2F_ARB:
-            /* Really shouldn't have to do this - the Node structure
-             * is convenient, but it would be better to store the data
-             * packed appropriately so that it can be sent directly
-             * on.  With x86_64 becoming common, this will start to
-             * matter more.
-             */
-            if (sizeof(Node) == sizeof(GLfloat))
-               CALL_VertexAttrib2fvARB(ctx->Exec, (n[1].e, &n[2].f));
-            else
-               CALL_VertexAttrib2fARB(ctx->Exec, (n[1].e, n[2].f, n[3].f));
+            CALL_VertexAttrib2fvARB(ctx->Exec, (n[1].e, &n[2].f));
             break;
          case OPCODE_ATTR_3F_ARB:
-            if (sizeof(Node) == sizeof(GLfloat))
-               CALL_VertexAttrib3fvARB(ctx->Exec, (n[1].e, &n[2].f));
-            else
-               CALL_VertexAttrib3fARB(ctx->Exec, (n[1].e, n[2].f, n[3].f,
-                                                  n[4].f));
+            CALL_VertexAttrib3fvARB(ctx->Exec, (n[1].e, &n[2].f));
             break;
          case OPCODE_ATTR_4F_ARB:
-            if (sizeof(Node) == sizeof(GLfloat))
-               CALL_VertexAttrib4fvARB(ctx->Exec, (n[1].e, &n[2].f));
-            else
-               CALL_VertexAttrib4fARB(ctx->Exec, (n[1].e, n[2].f, n[3].f,
-                                                  n[4].f, n[5].f));
+            CALL_VertexAttrib4fvARB(ctx->Exec, (n[1].e, &n[2].f));
             break;
          case OPCODE_MATERIAL:
-            if (sizeof(Node) == sizeof(GLfloat))
-               CALL_Materialfv(ctx->Exec, (n[1].e, n[2].e, &n[3].f));
-            else {
-               GLfloat f[4];
-               f[0] = n[3].f;
-               f[1] = n[4].f;
-               f[2] = n[5].f;
-               f[3] = n[6].f;
-               CALL_Materialfv(ctx->Exec, (n[1].e, n[2].e, f));
-            }
+            CALL_Materialfv(ctx->Exec, (n[1].e, n[2].e, &n[3].f));
             break;
          case OPCODE_BEGIN:
             CALL_Begin(ctx->Exec, (n[1].e));
@@ -7952,18 +8851,14 @@ execute_list(struct gl_context *ctx, GLuint list)
             CALL_FramebufferTexture(ctx->Exec, (n[1].e, n[2].e,
                                                    n[3].ui, n[4].i));
             break;
-         case OPCODE_FRAMEBUFFER_TEXTURE_FACE:
-            CALL_FramebufferTextureFaceARB(ctx->Exec, (n[1].e, n[2].e,
-                                                       n[3].ui, n[4].i, n[5].e));
-            break;
-
          /* GL_ARB_sync */
          case OPCODE_WAIT_SYNC:
             {
                union uint64_pair p;
-               p.uint32[0] = n[3].ui;
-               p.uint32[1] = n[4].ui;
-               CALL_WaitSync(ctx->Exec, (n[1].data, n[2].bf, p.uint64));
+               p.uint32[0] = n[2].ui;
+               p.uint32[1] = n[3].ui;
+               CALL_WaitSync(ctx->Exec,
+                             (get_pointer(&n[4]), n[1].bf, p.uint64));
             }
             break;
 
@@ -7980,7 +8875,10 @@ execute_list(struct gl_context *ctx, GLuint list)
             break;
 
          case OPCODE_CONTINUE:
-            n = (Node *) n[1].next;
+            n = (Node *) get_pointer(&n[1]);
+            break;
+         case OPCODE_NOP:
+            /* no-op */
             break;
          case OPCODE_END_OF_LIST:
             done = GL_TRUE;
@@ -8071,7 +8969,7 @@ _mesa_GenLists(GLsizei range)
    /*
     * Make this an atomic operation
     */
-   _glthread_LOCK_MUTEX(ctx->Shared->Mutex);
+   mtx_lock(&ctx->Shared->Mutex);
 
    base = _mesa_HashFindFreeKeyBlock(ctx->Shared->DisplayList, range);
    if (base) {
@@ -8083,7 +8981,7 @@ _mesa_GenLists(GLsizei range)
       }
    }
 
-   _glthread_UNLOCK_MUTEX(ctx->Shared->Mutex);
+   mtx_unlock(&ctx->Shared->Mutex);
 
    return base;
 }
@@ -8102,7 +9000,7 @@ _mesa_NewList(GLuint name, GLenum mode)
 
    if (MESA_VERBOSE & VERBOSE_API)
       _mesa_debug(ctx, "glNewList %u %s\n", name,
-                  _mesa_lookup_enum_by_nr(mode));
+                  _mesa_enum_to_string(mode));
 
    if (name == 0) {
       _mesa_error(ctx, GL_INVALID_VALUE, "glNewList");
@@ -8169,6 +9067,8 @@ _mesa_EndList(void)
 
    (void) alloc_instruction(ctx, OPCODE_END_OF_LIST, 0);
 
+   trim_list(ctx);
+
    /* Destroy old list, if any */
    destroy_list(ctx, ctx->ListState.CurrentList->Name);
 
@@ -8182,6 +9082,8 @@ _mesa_EndList(void)
       mesa_print_display_list(ctx->ListState.CurrentList->Name);
 
    ctx->ListState.CurrentList = NULL;
+   ctx->ListState.CurrentBlock = NULL;
+   ctx->ListState.CurrentPos = 0;
    ctx->ExecuteFlag = GL_TRUE;
    ctx->CompileFlag = GL_FALSE;
 
@@ -8544,10 +9446,6 @@ _mesa_initialize_save_table(const struct gl_context *ctx)
     */
    SET_BindProgramARB(table, save_BindProgramNV);
 
-   /* 244. GL_ATI_envmap_bumpmap */
-   SET_TexBumpParameterivATI(table, save_TexBumpParameterivATI);
-   SET_TexBumpParameterfvATI(table, save_TexBumpParameterfvATI);
-
    /* 245. GL_ATI_fragment_shader */
    SET_BindFragmentShaderATI(table, save_BindFragmentShaderATI);
    SET_SetFragmentShaderConstantATI(table, save_SetFragmentShaderConstantATI);
@@ -8653,9 +9551,8 @@ _mesa_initialize_save_table(const struct gl_context *ctx)
    SET_TexParameterIiv(table, save_TexParameterIiv);
    SET_TexParameterIuiv(table, save_TexParameterIuiv);
 
-   /* 377. GL_EXT_separate_shader_objects */
-   SET_UseShaderProgramEXT(table, save_UseShaderProgramEXT);
-   SET_ActiveProgramEXT(table, save_ActiveProgramEXT);
+   /* GL_ARB_clip_control */
+   SET_ClipControl(table, save_ClipControl);
 
    /* GL_ARB_color_buffer_float */
    SET_ClampColor(table, save_ClampColorARB);
@@ -8720,10 +9617,9 @@ _mesa_initialize_save_table(const struct gl_context *ctx)
    SET_BlendEquationiARB(table, save_BlendEquationi);
    SET_BlendEquationSeparateiARB(table, save_BlendEquationSeparatei);
 
-   /* GL_ARB_geometry_shader4 */
+   /* OpenGL 3.2 */
    SET_ProgramParameteri(table, save_ProgramParameteri);
    SET_FramebufferTexture(table, save_FramebufferTexture);
-   SET_FramebufferTextureFaceARB(table, save_FramebufferTextureFace);
 
    /* GL_NV_conditional_render */
    SET_BeginConditionalRender(table, save_BeginConditionalRender);
@@ -8746,6 +9642,45 @@ _mesa_initialize_save_table(const struct gl_context *ctx)
    SET_DrawArraysInstancedBaseInstance(table, save_DrawArraysInstancedBaseInstance);
    SET_DrawElementsInstancedBaseInstance(table, save_DrawElementsInstancedBaseInstance);
    SET_DrawElementsInstancedBaseVertexBaseInstance(table, save_DrawElementsInstancedBaseVertexBaseInstance);
+
+   /* OpenGL 4.2 / GL_ARB_separate_shader_objects */
+   SET_UseProgramStages(table, save_UseProgramStages);
+   SET_ProgramUniform1f(table, save_ProgramUniform1f);
+   SET_ProgramUniform2f(table, save_ProgramUniform2f);
+   SET_ProgramUniform3f(table, save_ProgramUniform3f);
+   SET_ProgramUniform4f(table, save_ProgramUniform4f);
+   SET_ProgramUniform1fv(table, save_ProgramUniform1fv);
+   SET_ProgramUniform2fv(table, save_ProgramUniform2fv);
+   SET_ProgramUniform3fv(table, save_ProgramUniform3fv);
+   SET_ProgramUniform4fv(table, save_ProgramUniform4fv);
+   SET_ProgramUniform1i(table, save_ProgramUniform1i);
+   SET_ProgramUniform2i(table, save_ProgramUniform2i);
+   SET_ProgramUniform3i(table, save_ProgramUniform3i);
+   SET_ProgramUniform4i(table, save_ProgramUniform4i);
+   SET_ProgramUniform1iv(table, save_ProgramUniform1iv);
+   SET_ProgramUniform2iv(table, save_ProgramUniform2iv);
+   SET_ProgramUniform3iv(table, save_ProgramUniform3iv);
+   SET_ProgramUniform4iv(table, save_ProgramUniform4iv);
+   SET_ProgramUniform1ui(table, save_ProgramUniform1ui);
+   SET_ProgramUniform2ui(table, save_ProgramUniform2ui);
+   SET_ProgramUniform3ui(table, save_ProgramUniform3ui);
+   SET_ProgramUniform4ui(table, save_ProgramUniform4ui);
+   SET_ProgramUniform1uiv(table, save_ProgramUniform1uiv);
+   SET_ProgramUniform2uiv(table, save_ProgramUniform2uiv);
+   SET_ProgramUniform3uiv(table, save_ProgramUniform3uiv);
+   SET_ProgramUniform4uiv(table, save_ProgramUniform4uiv);
+   SET_ProgramUniformMatrix2fv(table, save_ProgramUniformMatrix2fv);
+   SET_ProgramUniformMatrix3fv(table, save_ProgramUniformMatrix3fv);
+   SET_ProgramUniformMatrix4fv(table, save_ProgramUniformMatrix4fv);
+   SET_ProgramUniformMatrix2x3fv(table, save_ProgramUniformMatrix2x3fv);
+   SET_ProgramUniformMatrix3x2fv(table, save_ProgramUniformMatrix3x2fv);
+   SET_ProgramUniformMatrix2x4fv(table, save_ProgramUniformMatrix2x4fv);
+   SET_ProgramUniformMatrix4x2fv(table, save_ProgramUniformMatrix4x2fv);
+   SET_ProgramUniformMatrix3x4fv(table, save_ProgramUniformMatrix3x4fv);
+   SET_ProgramUniformMatrix4x3fv(table, save_ProgramUniformMatrix4x3fv);
+
+   /* GL_EXT_polygon_offset_clamp */
+   SET_PolygonOffsetClampEXT(table, save_PolygonOffsetClampEXT);
 }
 
 
@@ -8753,23 +9688,31 @@ _mesa_initialize_save_table(const struct gl_context *ctx)
 static const char *
 enum_string(GLenum k)
 {
-   return _mesa_lookup_enum_by_nr(k);
+   return _mesa_enum_to_string(k);
 }
 
 
 /**
  * Print the commands in a display list.  For debugging only.
  * TODO: many commands aren't handled yet.
+ * \param fname  filename to write display list to.  If null, use stdout.
  */
 static void GLAPIENTRY
-print_list(struct gl_context *ctx, GLuint list)
+print_list(struct gl_context *ctx, GLuint list, const char *fname)
 {
    struct gl_display_list *dlist;
    Node *n;
    GLboolean done;
+   FILE *f = stdout;
+
+   if (fname) {
+      f = fopen(fname, "w");
+      if (!f)
+         return;
+   }
 
    if (!islist(ctx, list)) {
-      printf("%u is not a display list ID\n", list);
+      fprintf(f, "%u is not a display list ID\n", list);
       return;
    }
 
@@ -8779,214 +9722,221 @@ print_list(struct gl_context *ctx, GLuint list)
 
    n = dlist->Head;
 
-   printf("START-LIST %u, address %p\n", list, (void *) n);
+   fprintf(f, "START-LIST %u, address %p\n", list, (void *) n);
 
    done = n ? GL_FALSE : GL_TRUE;
    while (!done) {
       const OpCode opcode = n[0].opcode;
 
       if (is_ext_opcode(opcode)) {
-         n += ext_opcode_print(ctx, n);
+         n += ext_opcode_print(ctx, n, f);
       }
       else {
          switch (opcode) {
          case OPCODE_ACCUM:
-            printf("Accum %s %g\n", enum_string(n[1].e), n[2].f);
+            fprintf(f, "Accum %s %g\n", enum_string(n[1].e), n[2].f);
+            break;
+         case OPCODE_ACTIVE_TEXTURE:
+            fprintf(f, "ActiveTexture(%s)\n", enum_string(n[1].e));
             break;
          case OPCODE_BITMAP:
-            printf("Bitmap %d %d %g %g %g %g %p\n", n[1].i, n[2].i,
-                         n[3].f, n[4].f, n[5].f, n[6].f, (void *) n[7].data);
+            fprintf(f, "Bitmap %d %d %g %g %g %g %p\n", n[1].i, n[2].i,
+                   n[3].f, n[4].f, n[5].f, n[6].f,
+                   get_pointer(&n[7]));
             break;
          case OPCODE_CALL_LIST:
-            printf("CallList %d\n", (int) n[1].ui);
+            fprintf(f, "CallList %d\n", (int) n[1].ui);
             break;
          case OPCODE_CALL_LIST_OFFSET:
-            printf("CallList %d + offset %u = %u\n", (int) n[1].ui,
+            fprintf(f, "CallList %d + offset %u = %u\n", (int) n[1].ui,
                          ctx->List.ListBase, ctx->List.ListBase + n[1].ui);
             break;
          case OPCODE_DISABLE:
-            printf("Disable %s\n", enum_string(n[1].e));
+            fprintf(f, "Disable %s\n", enum_string(n[1].e));
             break;
          case OPCODE_ENABLE:
-            printf("Enable %s\n", enum_string(n[1].e));
+            fprintf(f, "Enable %s\n", enum_string(n[1].e));
             break;
          case OPCODE_FRUSTUM:
-            printf("Frustum %g %g %g %g %g %g\n",
+            fprintf(f, "Frustum %g %g %g %g %g %g\n",
                          n[1].f, n[2].f, n[3].f, n[4].f, n[5].f, n[6].f);
             break;
          case OPCODE_LINE_STIPPLE:
-            printf("LineStipple %d %x\n", n[1].i, (int) n[2].us);
+            fprintf(f, "LineStipple %d %x\n", n[1].i, (int) n[2].us);
             break;
          case OPCODE_LOAD_IDENTITY:
-            printf("LoadIdentity\n");
+            fprintf(f, "LoadIdentity\n");
             break;
          case OPCODE_LOAD_MATRIX:
-            printf("LoadMatrix\n");
-            printf("  %8f %8f %8f %8f\n",
+            fprintf(f, "LoadMatrix\n");
+            fprintf(f, "  %8f %8f %8f %8f\n",
                          n[1].f, n[5].f, n[9].f, n[13].f);
-            printf("  %8f %8f %8f %8f\n",
+            fprintf(f, "  %8f %8f %8f %8f\n",
                          n[2].f, n[6].f, n[10].f, n[14].f);
-            printf("  %8f %8f %8f %8f\n",
+            fprintf(f, "  %8f %8f %8f %8f\n",
                          n[3].f, n[7].f, n[11].f, n[15].f);
-            printf("  %8f %8f %8f %8f\n",
+            fprintf(f, "  %8f %8f %8f %8f\n",
                          n[4].f, n[8].f, n[12].f, n[16].f);
             break;
          case OPCODE_MULT_MATRIX:
-            printf("MultMatrix (or Rotate)\n");
-            printf("  %8f %8f %8f %8f\n",
+            fprintf(f, "MultMatrix (or Rotate)\n");
+            fprintf(f, "  %8f %8f %8f %8f\n",
                          n[1].f, n[5].f, n[9].f, n[13].f);
-            printf("  %8f %8f %8f %8f\n",
+            fprintf(f, "  %8f %8f %8f %8f\n",
                          n[2].f, n[6].f, n[10].f, n[14].f);
-            printf("  %8f %8f %8f %8f\n",
+            fprintf(f, "  %8f %8f %8f %8f\n",
                          n[3].f, n[7].f, n[11].f, n[15].f);
-            printf("  %8f %8f %8f %8f\n",
+            fprintf(f, "  %8f %8f %8f %8f\n",
                          n[4].f, n[8].f, n[12].f, n[16].f);
             break;
          case OPCODE_ORTHO:
-            printf("Ortho %g %g %g %g %g %g\n",
+            fprintf(f, "Ortho %g %g %g %g %g %g\n",
                          n[1].f, n[2].f, n[3].f, n[4].f, n[5].f, n[6].f);
             break;
          case OPCODE_POP_ATTRIB:
-            printf("PopAttrib\n");
+            fprintf(f, "PopAttrib\n");
             break;
          case OPCODE_POP_MATRIX:
-            printf("PopMatrix\n");
+            fprintf(f, "PopMatrix\n");
             break;
          case OPCODE_POP_NAME:
-            printf("PopName\n");
+            fprintf(f, "PopName\n");
             break;
          case OPCODE_PUSH_ATTRIB:
-            printf("PushAttrib %x\n", n[1].bf);
+            fprintf(f, "PushAttrib %x\n", n[1].bf);
             break;
          case OPCODE_PUSH_MATRIX:
-            printf("PushMatrix\n");
+            fprintf(f, "PushMatrix\n");
             break;
          case OPCODE_PUSH_NAME:
-            printf("PushName %d\n", (int) n[1].ui);
+            fprintf(f, "PushName %d\n", (int) n[1].ui);
             break;
          case OPCODE_RASTER_POS:
-            printf("RasterPos %g %g %g %g\n",
+            fprintf(f, "RasterPos %g %g %g %g\n",
                          n[1].f, n[2].f, n[3].f, n[4].f);
             break;
          case OPCODE_ROTATE:
-            printf("Rotate %g %g %g %g\n",
+            fprintf(f, "Rotate %g %g %g %g\n",
                          n[1].f, n[2].f, n[3].f, n[4].f);
             break;
          case OPCODE_SCALE:
-            printf("Scale %g %g %g\n", n[1].f, n[2].f, n[3].f);
+            fprintf(f, "Scale %g %g %g\n", n[1].f, n[2].f, n[3].f);
             break;
          case OPCODE_TRANSLATE:
-            printf("Translate %g %g %g\n", n[1].f, n[2].f, n[3].f);
+            fprintf(f, "Translate %g %g %g\n", n[1].f, n[2].f, n[3].f);
             break;
          case OPCODE_BIND_TEXTURE:
-            printf("BindTexture %s %d\n",
-                         _mesa_lookup_enum_by_nr(n[1].ui), n[2].ui);
+            fprintf(f, "BindTexture %s %d\n",
+                         _mesa_enum_to_string(n[1].ui), n[2].ui);
             break;
          case OPCODE_SHADE_MODEL:
-            printf("ShadeModel %s\n", _mesa_lookup_enum_by_nr(n[1].ui));
+            fprintf(f, "ShadeModel %s\n", _mesa_enum_to_string(n[1].ui));
             break;
          case OPCODE_MAP1:
-            printf("Map1 %s %.3f %.3f %d %d\n",
-                         _mesa_lookup_enum_by_nr(n[1].ui),
+            fprintf(f, "Map1 %s %.3f %.3f %d %d\n",
+                         _mesa_enum_to_string(n[1].ui),
                          n[2].f, n[3].f, n[4].i, n[5].i);
             break;
          case OPCODE_MAP2:
-            printf("Map2 %s %.3f %.3f %.3f %.3f %d %d %d %d\n",
-                         _mesa_lookup_enum_by_nr(n[1].ui),
+            fprintf(f, "Map2 %s %.3f %.3f %.3f %.3f %d %d %d %d\n",
+                         _mesa_enum_to_string(n[1].ui),
                          n[2].f, n[3].f, n[4].f, n[5].f,
                          n[6].i, n[7].i, n[8].i, n[9].i);
             break;
          case OPCODE_MAPGRID1:
-            printf("MapGrid1 %d %.3f %.3f\n", n[1].i, n[2].f, n[3].f);
+            fprintf(f, "MapGrid1 %d %.3f %.3f\n", n[1].i, n[2].f, n[3].f);
             break;
          case OPCODE_MAPGRID2:
-            printf("MapGrid2 %d %.3f %.3f, %d %.3f %.3f\n",
+            fprintf(f, "MapGrid2 %d %.3f %.3f, %d %.3f %.3f\n",
                          n[1].i, n[2].f, n[3].f, n[4].i, n[5].f, n[6].f);
             break;
          case OPCODE_EVALMESH1:
-            printf("EvalMesh1 %d %d\n", n[1].i, n[2].i);
+            fprintf(f, "EvalMesh1 %d %d\n", n[1].i, n[2].i);
             break;
          case OPCODE_EVALMESH2:
-            printf("EvalMesh2 %d %d %d %d\n",
+            fprintf(f, "EvalMesh2 %d %d %d %d\n",
                          n[1].i, n[2].i, n[3].i, n[4].i);
             break;
 
          case OPCODE_ATTR_1F_NV:
-            printf("ATTR_1F_NV attr %d: %f\n", n[1].i, n[2].f);
+            fprintf(f, "ATTR_1F_NV attr %d: %f\n", n[1].i, n[2].f);
             break;
          case OPCODE_ATTR_2F_NV:
-            printf("ATTR_2F_NV attr %d: %f %f\n",
+            fprintf(f, "ATTR_2F_NV attr %d: %f %f\n",
                          n[1].i, n[2].f, n[3].f);
             break;
          case OPCODE_ATTR_3F_NV:
-            printf("ATTR_3F_NV attr %d: %f %f %f\n",
+            fprintf(f, "ATTR_3F_NV attr %d: %f %f %f\n",
                          n[1].i, n[2].f, n[3].f, n[4].f);
             break;
          case OPCODE_ATTR_4F_NV:
-            printf("ATTR_4F_NV attr %d: %f %f %f %f\n",
+            fprintf(f, "ATTR_4F_NV attr %d: %f %f %f %f\n",
                          n[1].i, n[2].f, n[3].f, n[4].f, n[5].f);
             break;
          case OPCODE_ATTR_1F_ARB:
-            printf("ATTR_1F_ARB attr %d: %f\n", n[1].i, n[2].f);
+            fprintf(f, "ATTR_1F_ARB attr %d: %f\n", n[1].i, n[2].f);
             break;
          case OPCODE_ATTR_2F_ARB:
-            printf("ATTR_2F_ARB attr %d: %f %f\n",
+            fprintf(f, "ATTR_2F_ARB attr %d: %f %f\n",
                          n[1].i, n[2].f, n[3].f);
             break;
          case OPCODE_ATTR_3F_ARB:
-            printf("ATTR_3F_ARB attr %d: %f %f %f\n",
+            fprintf(f, "ATTR_3F_ARB attr %d: %f %f %f\n",
                          n[1].i, n[2].f, n[3].f, n[4].f);
             break;
          case OPCODE_ATTR_4F_ARB:
-            printf("ATTR_4F_ARB attr %d: %f %f %f %f\n",
+            fprintf(f, "ATTR_4F_ARB attr %d: %f %f %f %f\n",
                          n[1].i, n[2].f, n[3].f, n[4].f, n[5].f);
             break;
 
          case OPCODE_MATERIAL:
-            printf("MATERIAL %x %x: %f %f %f %f\n",
+            fprintf(f, "MATERIAL %x %x: %f %f %f %f\n",
                          n[1].i, n[2].i, n[3].f, n[4].f, n[5].f, n[6].f);
             break;
          case OPCODE_BEGIN:
-            printf("BEGIN %x\n", n[1].i);
+            fprintf(f, "BEGIN %x\n", n[1].i);
             break;
          case OPCODE_END:
-            printf("END\n");
+            fprintf(f, "END\n");
             break;
          case OPCODE_RECTF:
-            printf("RECTF %f %f %f %f\n", n[1].f, n[2].f, n[3].f,
+            fprintf(f, "RECTF %f %f %f %f\n", n[1].f, n[2].f, n[3].f,
                          n[4].f);
             break;
          case OPCODE_EVAL_C1:
-            printf("EVAL_C1 %f\n", n[1].f);
+            fprintf(f, "EVAL_C1 %f\n", n[1].f);
             break;
          case OPCODE_EVAL_C2:
-            printf("EVAL_C2 %f %f\n", n[1].f, n[2].f);
+            fprintf(f, "EVAL_C2 %f %f\n", n[1].f, n[2].f);
             break;
          case OPCODE_EVAL_P1:
-            printf("EVAL_P1 %d\n", n[1].i);
+            fprintf(f, "EVAL_P1 %d\n", n[1].i);
             break;
          case OPCODE_EVAL_P2:
-            printf("EVAL_P2 %d %d\n", n[1].i, n[2].i);
+            fprintf(f, "EVAL_P2 %d %d\n", n[1].i, n[2].i);
             break;
 
          case OPCODE_PROVOKING_VERTEX:
-            printf("ProvokingVertex %s\n",
-                         _mesa_lookup_enum_by_nr(n[1].ui));
+            fprintf(f, "ProvokingVertex %s\n",
+                         _mesa_enum_to_string(n[1].ui));
             break;
 
             /*
              * meta opcodes/commands
              */
          case OPCODE_ERROR:
-            printf("Error: %s %s\n",
-                         enum_string(n[1].e), (const char *) n[2].data);
+            fprintf(f, "Error: %s %s\n", enum_string(n[1].e),
+                   (const char *) get_pointer(&n[2]));
             break;
          case OPCODE_CONTINUE:
-            printf("DISPLAY-LIST-CONTINUE\n");
-            n = (Node *) n[1].next;
+            fprintf(f, "DISPLAY-LIST-CONTINUE\n");
+            n = (Node *) get_pointer(&n[1]);
+            break;
+         case OPCODE_NOP:
+            fprintf(f, "NOP\n");
             break;
          case OPCODE_END_OF_LIST:
-            printf("END-LIST %u\n", list);
+            fprintf(f, "END-LIST %u\n", list);
             done = GL_TRUE;
             break;
          default:
@@ -8997,7 +9947,7 @@ print_list(struct gl_context *ctx, GLuint list)
                return;
             }
             else {
-               printf("command %d, %u operands\n", opcode,
+               fprintf(f, "command %d, %u operands\n", opcode,
                             InstSize[opcode]);
             }
          }
@@ -9007,6 +9957,10 @@ print_list(struct gl_context *ctx, GLuint list)
          }
       }
    }
+
+   fflush(f);
+   if (fname)
+      fclose(f);
 }
 
 
@@ -9020,7 +9974,7 @@ void
 mesa_print_display_list(GLuint list)
 {
    GET_CURRENT_CONTEXT(ctx);
-   print_list(ctx, list);
+   print_list(ctx, list, NULL);
 }
 
 
@@ -9131,6 +10085,8 @@ _mesa_init_display_list(struct gl_context *ctx)
    ctx->List.ListBase = 0;
 
    save_vtxfmt_init(&ctx->ListState.ListVtxfmt);
+
+   InstSize[OPCODE_NOP] = 1;
 }