Import vtx-0-2-branch

41 files changed, 409 insertions, 470 deletions

diff --git a/src/mesa/drivers/dri/fb/Makefile.X11 b/src/mesa/drivers/dri/fb/Makefile.X11
index 671de313c08..c1eb713bda9 100644
--- a/src/mesa/drivers/dri/fb/Makefile.X11
+++ b/src/mesa/drivers/dri/fb/Makefile.X11
@@ -1,4 +1,3 @@
-# $Id: Makefile.X11,v 1.3 2003/10/20 02:17:32 jonsmirl Exp $
 
 # Mesa 3-D graphics library
 # Version:  5.0
diff --git a/src/mesa/drivers/dri/fb/fb_dri.c b/src/mesa/drivers/dri/fb/fb_dri.c
index 794e3d1fc8c..7068b94f688 100644
--- a/src/mesa/drivers/dri/fb/fb_dri.c
+++ b/src/mesa/drivers/dri/fb/fb_dri.c
@@ -1,4 +1,3 @@
-/* $Id: fb_dri.c,v 1.2 2003/10/08 01:53:30 kendallb Exp $ */
 
 /*
  * Mesa 3-D graphics library
diff --git a/src/mesa/drivers/dri/gamma/Makefile.X11 b/src/mesa/drivers/dri/gamma/Makefile.X11
index f779bf336cc..8364ebaecf0 100644
--- a/src/mesa/drivers/dri/gamma/Makefile.X11
+++ b/src/mesa/drivers/dri/gamma/Makefile.X11
@@ -1,4 +1,4 @@
-# $Id: Makefile.X11,v 1.2 2003/10/20 02:17:32 jonsmirl Exp $
+# $Id: Makefile.X11,v 1.3 2003/11/24 15:21:14 keithw Exp $
 
 # Mesa 3-D graphics library
 # Version:  5.0
diff --git a/src/mesa/drivers/dri/gamma/gamma_context.c b/src/mesa/drivers/dri/gamma/gamma_context.c
index f7db2adea37..b27851be8f8 100644
--- a/src/mesa/drivers/dri/gamma/gamma_context.c
+++ b/src/mesa/drivers/dri/gamma/gamma_context.c
@@ -48,9 +48,9 @@
 #include "gamma_vb.h"
 #include "gamma_tris.h"
 
-extern const struct gl_pipeline_stage _gamma_render_stage;
+extern const struct tnl_pipeline_stage _gamma_render_stage;
 
-static const struct gl_pipeline_stage *gamma_pipeline[] = {
+static const struct tnl_pipeline_stage *gamma_pipeline[] = {
    &_tnl_vertex_transform_stage,
    &_tnl_normal_transform_stage,
    &_tnl_lighting_stage,
diff --git a/src/mesa/drivers/dri/gamma/gamma_render.c b/src/mesa/drivers/dri/gamma/gamma_render.c
index b8bf6171564..67063c2e525 100644
--- a/src/mesa/drivers/dri/gamma/gamma_render.c
+++ b/src/mesa/drivers/dri/gamma/gamma_render.c
@@ -46,18 +46,15 @@ static void gamma_emit( GLcontext *ctx, GLuint start, GLuint end)
    struct vertex_buffer *VB = &TNL_CONTEXT(ctx)->vb;
    GLfloat (*coord)[4];
    GLuint coord_stride;
-   GLubyte (*col)[4];
+   GLfloat (*col)[4];
    GLuint col_stride;
    int i;
    GLuint tc0_stride = 0;
    GLfloat (*tc0)[4] = 0;
    GLuint tc0_size = 0;
 
-   if (VB->ColorPtr[0]->Type != GL_UNSIGNED_BYTE) 
-      gamma_import_float_colors( ctx );
-   
-   col = VB->ColorPtr[0]->Ptr;
-   col_stride = VB->ColorPtr[0]->StrideB;
+   col = VB->ColorPtr[0]->data;
+   col_stride = VB->ColorPtr[0]->stride;
 
    if (ctx->Texture.Unit[0]._ReallyEnabled) {
       tc0_stride = VB->TexCoordPtr[0]->stride;
@@ -70,58 +67,8 @@ static void gamma_emit( GLcontext *ctx, GLuint start, GLuint end)
       coord_stride = VB->NdcPtr->stride;
    }
 
-   if (VB->importable_data) {
-      if (start) {
-         coord =  (GLfloat (*)[4])((GLubyte *)coord + start * coord_stride);
-         STRIDE_4UB(col, start * col_stride);
-         if (ctx->Texture.Unit[0]._ReallyEnabled)
-	    tc0 =  (GLfloat (*)[4])((GLubyte *)tc0 + start * tc0_stride);
-      }
-
-      if (ctx->Texture.Unit[0]._ReallyEnabled && tc0_size == 4) {
-         for (i=start; i < end; i++) {
-	    CHECK_DMA_BUFFER(gmesa, 9);
-	    WRITEF(gmesa->buf, Tq4, tc0[0][3]);
-	    WRITEF(gmesa->buf, Tr4, tc0[0][2]);
-	    WRITEF(gmesa->buf, Tt4, tc0[0][0]);
-	    WRITEF(gmesa->buf, Ts4, tc0[0][1]);
-	    WRITE(gmesa->buf, PackedColor4, *(CARD32*)col[0]);
-	    WRITEF(gmesa->buf, Vw, coord[0][3]);
-	    WRITEF(gmesa->buf, Vz, coord[0][2]);
-	    WRITEF(gmesa->buf, Vy, coord[0][1]);
-	    WRITEF(gmesa->buf, Vx4, coord[0][0]);
-	    STRIDE_4UB(col, col_stride);	    
-	    tc0 =  (GLfloat (*)[4])((GLubyte *)tc0 +  tc0_stride);
-	    coord =  (GLfloat (*)[4])((GLubyte *)coord +  coord_stride);
-         }
-      } else if (ctx->Texture.Unit[0]._ReallyEnabled && tc0_size == 2) {
-         for (i=start; i < end; i++) {
-	    CHECK_DMA_BUFFER(gmesa, 7);
-	    WRITEF(gmesa->buf, Tt2, tc0[0][0]);
-	    WRITEF(gmesa->buf, Ts2, tc0[0][1]);
-	    WRITE(gmesa->buf, PackedColor4, *(CARD32*)col[0]);
-	    WRITEF(gmesa->buf, Vw, coord[0][3]);
-	    WRITEF(gmesa->buf, Vz, coord[0][2]);
-	    WRITEF(gmesa->buf, Vy, coord[0][1]);
-	    WRITEF(gmesa->buf, Vx4, coord[0][0]);
-	    STRIDE_4UB(col, col_stride);	    
-	    tc0 =  (GLfloat (*)[4])((GLubyte *)tc0 +  tc0_stride);
-	    coord =  (GLfloat (*)[4])((GLubyte *)coord +  coord_stride);
-         }
-      } else {
-         for (i=start; i < end; i++) {
-	    CHECK_DMA_BUFFER(gmesa, 4);
-	    WRITE(gmesa->buf, PackedColor4, *(CARD32*)col[0]);
-	    WRITEF(gmesa->buf, Vz, coord[0][2]);
-	    WRITEF(gmesa->buf, Vy, coord[0][1]);
-	    WRITEF(gmesa->buf, Vx3, coord[0][0]);
-	    STRIDE_4UB(col, col_stride);	    
-	    coord =  (GLfloat (*)[4])((GLubyte *)coord +  coord_stride);
-         }
-      }
-   } else {
-      if (ctx->Texture.Unit[0]._ReallyEnabled && tc0_size == 4) {
-         for (i=start; i < end; i++) {
+   if (ctx->Texture.Unit[0]._ReallyEnabled && tc0_size == 4) {
+      for (i=start; i < end; i++) {
 	    CHECK_DMA_BUFFER(gmesa, 9);
 	    WRITEF(gmesa->buf, Tq4, tc0[i][3]);
 	    WRITEF(gmesa->buf, Tr4, tc0[i][2]);
@@ -132,9 +79,9 @@ static void gamma_emit( GLcontext *ctx, GLuint start, GLuint end)
 	    WRITEF(gmesa->buf, Vz, coord[i][2]);
 	    WRITEF(gmesa->buf, Vy, coord[i][1]);
 	    WRITEF(gmesa->buf, Vx4, coord[i][0]);
-         }
-      } else if (ctx->Texture.Unit[0]._ReallyEnabled && tc0_size == 2) {
-         for (i=start; i < end; i++) {
+      }
+   } else if (ctx->Texture.Unit[0]._ReallyEnabled && tc0_size == 2) {
+      for (i=start; i < end; i++) {
 	    CHECK_DMA_BUFFER(gmesa, 7);
 	    WRITEF(gmesa->buf, Tt2, tc0[i][0]);
 	    WRITEF(gmesa->buf, Ts2, tc0[i][1]);
@@ -143,15 +90,14 @@ static void gamma_emit( GLcontext *ctx, GLuint start, GLuint end)
 	    WRITEF(gmesa->buf, Vz, coord[i][2]);
 	    WRITEF(gmesa->buf, Vy, coord[i][1]);
 	    WRITEF(gmesa->buf, Vx4, coord[i][0]);
-         }
-      } else {
-         for (i=start; i < end; i++) {
+      }
+   } else {
+      for (i=start; i < end; i++) {
 	    CHECK_DMA_BUFFER(gmesa, 4);
 	    WRITE(gmesa->buf, PackedColor4, *(CARD32*)col[i]);
 	    WRITEF(gmesa->buf, Vz, coord[i][2]);
 	    WRITEF(gmesa->buf, Vy, coord[i][1]);
 	    WRITEF(gmesa->buf, Vx3, coord[i][0]);
-         }
       }
    }
 }
@@ -178,7 +124,7 @@ static void VERT_FALLBACK( GLcontext *ctx,
    tnl->Driver.Render.PrimitiveNotify( ctx, flags & PRIM_MODE_MASK );
    tnl->Driver.Render.BuildVertices( ctx, start, count, ~0 );
    tnl->Driver.Render.PrimTabVerts[flags&PRIM_MODE_MASK]( ctx, start, count, flags );
-   GAMMA_CONTEXT(ctx)->SetupNewInputs = VERT_BIT_CLIP;
+   GAMMA_CONTEXT(ctx)->SetupNewInputs = VERT_BIT_POS;
 }
 
 static const GLuint hw_prim[GL_POLYGON+1] = {
@@ -236,7 +182,7 @@ static __inline void gammaEndPrimitive( gammaContextPtr gmesa )
 
 
 static GLboolean gamma_run_render( GLcontext *ctx,
-				  struct gl_pipeline_stage *stage )
+				  struct tnl_pipeline_stage *stage )
 {
    gammaContextPtr gmesa = GAMMA_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
@@ -256,12 +202,16 @@ static GLboolean gamma_run_render( GLcontext *ctx,
 
    tnl->Driver.Render.Start( ctx );
 
-   for (i = VB->FirstPrimitive ; !(flags & PRIM_LAST) ; i += length)
+   for (i = 0 ; i < VB->PrimitiveCount ; i++)
    {
-      flags = VB->Primitive[i];
-      length = VB->PrimitiveLength[i];
-      if (length)
-	 tab[flags & PRIM_MODE_MASK]( ctx, i, i + length, flags );
+      GLuint prim = VB->Primitive[i].mode;
+      GLuint start = VB->Primitive[i].start;
+      GLuint length = VB->Primitive[i].count;
+
+      if (!length)
+	 continue;
+
+      tab[prim & PRIM_MODE_MASK]( ctx, start, start + length, prim);
    }
 
    tnl->Driver.Render.Finish( ctx );
@@ -271,9 +221,9 @@ static GLboolean gamma_run_render( GLcontext *ctx,
 
 
 static void gamma_check_render( GLcontext *ctx,
-				 struct gl_pipeline_stage *stage )
+				 struct tnl_pipeline_stage *stage )
 {
-   GLuint inputs = VERT_BIT_CLIP | VERT_BIT_COLOR0;
+   GLuint inputs = VERT_BIT_POS | VERT_BIT_COLOR0;
 
    if (ctx->RenderMode == GL_RENDER) {
       if (ctx->_TriangleCaps & DD_SEPARATE_SPECULAR)
@@ -293,13 +243,13 @@ static void gamma_check_render( GLcontext *ctx,
 }
 
 
-static void dtr( struct gl_pipeline_stage *stage )
+static void dtr( struct tnl_pipeline_stage *stage )
 {
    (void)stage;
 }
 
 
-const struct gl_pipeline_stage _gamma_render_stage =
+const struct tnl_pipeline_stage _gamma_render_stage =
 {
    "gamma render",
    (_DD_NEW_SEPARATE_SPECULAR |
diff --git a/src/mesa/drivers/dri/gamma/gamma_vb.c b/src/mesa/drivers/dri/gamma/gamma_vb.c
index 2e2f9ede457..f4ed08b140f 100644
--- a/src/mesa/drivers/dri/gamma/gamma_vb.c
+++ b/src/mesa/drivers/dri/gamma/gamma_vb.c
@@ -278,7 +278,7 @@ void gammaBuildVertices( GLcontext *ctx,
    if (!newinputs)
       return;
 
-   if (newinputs & VERT_BIT_CLIP) {
+   if (newinputs & VERT_BIT_POS) {
       setup_tab[gmesa->SetupIndex].emit( ctx, start, count, v, stride );
    } else {
       GLuint ind = 0;
diff --git a/src/mesa/drivers/dri/i810/Makefile.X11 b/src/mesa/drivers/dri/i810/Makefile.X11
index 9fd5a5a3e59..8f9c7958ba2 100644
--- a/src/mesa/drivers/dri/i810/Makefile.X11
+++ b/src/mesa/drivers/dri/i810/Makefile.X11
@@ -1,4 +1,3 @@
-# $Id: Makefile.X11,v 1.2 2003/10/20 02:17:32 jonsmirl Exp $
 
 # Mesa 3-D graphics library
 # Version:  5.0
diff --git a/src/mesa/drivers/dri/i810/i810context.c b/src/mesa/drivers/dri/i810/i810context.c
index da1764cefad..d678aabc824 100644
--- a/src/mesa/drivers/dri/i810/i810context.c
+++ b/src/mesa/drivers/dri/i810/i810context.c
@@ -107,9 +107,9 @@ static const char * const card_extensions[] =
    NULL
 };
 
-extern const struct gl_pipeline_stage _i810_render_stage;
+extern const struct tnl_pipeline_stage _i810_render_stage;
 
-static const struct gl_pipeline_stage *i810_pipeline[] = {
+static const struct tnl_pipeline_stage *i810_pipeline[] = {
    &_tnl_vertex_transform_stage,
    &_tnl_normal_transform_stage,
    &_tnl_lighting_stage,
diff --git a/src/mesa/drivers/dri/i810/i810render.c b/src/mesa/drivers/dri/i810/i810render.c
index 8d388d88b3f..d1c1adc3b7c 100644
--- a/src/mesa/drivers/dri/i810/i810render.c
+++ b/src/mesa/drivers/dri/i810/i810render.c
@@ -105,7 +105,7 @@ static void VERT_FALLBACK( GLcontext *ctx,
    tnl->Driver.Render.BuildVertices( ctx, start, count, ~0 );
    tnl->Driver.Render.PrimTabVerts[flags&PRIM_MODE_MASK]( ctx, start, 
 							  count, flags );
-   I810_CONTEXT(ctx)->SetupNewInputs = VERT_BIT_CLIP;
+   I810_CONTEXT(ctx)->SetupNewInputs = VERT_BIT_POS;
 }
 
 
@@ -137,12 +137,12 @@ static void VERT_FALLBACK( GLcontext *ctx,
 
 
 static GLboolean i810_run_render( GLcontext *ctx,
-				  struct gl_pipeline_stage *stage )
+				  struct tnl_pipeline_stage *stage )
 {
    i810ContextPtr imesa = I810_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
    struct vertex_buffer *VB = &tnl->vb;
-   GLuint i, length, flags = 0;
+   GLuint i;
 
    /* Don't handle clipping or indexed vertices.
     */
@@ -150,17 +150,21 @@ static GLboolean i810_run_render( GLcontext *ctx,
       return GL_TRUE;
    }
 
-   imesa->SetupNewInputs = VERT_BIT_CLIP;
+   imesa->SetupNewInputs = VERT_BIT_POS;
 
    tnl->Driver.Render.Start( ctx );
 
-   for (i = VB->FirstPrimitive ; !(flags & PRIM_LAST) ; i += length)
+   for (i = 0 ; i < VB->PrimitiveCount ; i++)
    {
-      flags = VB->Primitive[i];
-      length= VB->PrimitiveLength[i];
-      if (length)
-	 i810_render_tab_verts[flags & PRIM_MODE_MASK]( ctx, i, i + length,
-							flags );
+      GLuint prim = VB->Primitive[i].mode;
+      GLuint start = VB->Primitive[i].start;
+      GLuint length = VB->Primitive[i].count;
+
+      if (!length)
+	 continue;
+
+      i810_render_tab_verts[prim & PRIM_MODE_MASK]( ctx, start, start + length,
+						    prim );
    }
 
    tnl->Driver.Render.Finish( ctx );
@@ -169,9 +173,9 @@ static GLboolean i810_run_render( GLcontext *ctx,
 }
 
 
-static void i810_check_render( GLcontext *ctx, struct gl_pipeline_stage *stage )
+static void i810_check_render( GLcontext *ctx, struct tnl_pipeline_stage *stage )
 {
-   GLuint inputs = VERT_BIT_CLIP | VERT_BIT_COLOR0;
+   GLuint inputs = VERT_BIT_POS | VERT_BIT_COLOR0;
 
    if (ctx->RenderMode == GL_RENDER) {
       if (ctx->_TriangleCaps & DD_SEPARATE_SPECULAR)
@@ -191,13 +195,13 @@ static void i810_check_render( GLcontext *ctx, struct gl_pipeline_stage *stage )
 }
 
 
-static void dtr( struct gl_pipeline_stage *stage )
+static void dtr( struct tnl_pipeline_stage *stage )
 {
    (void)stage;
 }
 
 
-const struct gl_pipeline_stage _i810_render_stage =
+const struct tnl_pipeline_stage _i810_render_stage =
 {
    "i810 render",
    (_DD_NEW_SEPARATE_SPECULAR |
diff --git a/src/mesa/drivers/dri/i810/i810screen.c b/src/mesa/drivers/dri/i810/i810screen.c
index 604d56e2c27..53f1251fa9d 100644
--- a/src/mesa/drivers/dri/i810/i810screen.c
+++ b/src/mesa/drivers/dri/i810/i810screen.c
@@ -74,7 +74,7 @@ static drmBufMapPtr i810_create_empty_buffers(void)
    memset(retval, 0, sizeof(drmBufMap));
    retval->list = (drmBufPtr)ALIGN_MALLOC(sizeof(drmBuf) * I810_DMA_BUF_NR, 32);
    if(retval->list == NULL) {
-      FREE(retval);
+      ALIGN_FREE(retval);
       return NULL;
    }
    memset(retval->list, 0, sizeof(drmBuf) * I810_DMA_BUF_NR);
diff --git a/src/mesa/drivers/dri/i810/i810vb.c b/src/mesa/drivers/dri/i810/i810vb.c
index ada7f6db853..58979ceac23 100644
--- a/src/mesa/drivers/dri/i810/i810vb.c
+++ b/src/mesa/drivers/dri/i810/i810vb.c
@@ -383,7 +383,7 @@ void i810BuildVertices( GLcontext *ctx,
    if (!newinputs)
       return;
 
-   if (newinputs & VERT_BIT_CLIP) {
+   if (newinputs & VERT_BIT_POS) {
       setup_tab[imesa->SetupIndex].emit( ctx, start, count, v, stride );
    } else {
       GLuint ind = 0;
diff --git a/src/mesa/drivers/dri/i830/Makefile.X11 b/src/mesa/drivers/dri/i830/Makefile.X11
index 3c6a4036822..cfc513690db 100644
--- a/src/mesa/drivers/dri/i830/Makefile.X11
+++ b/src/mesa/drivers/dri/i830/Makefile.X11
@@ -1,4 +1,3 @@
-# $Id: Makefile.X11,v 1.3 2003/10/20 02:17:33 jonsmirl Exp $
 
 # Mesa 3-D graphics library
 # Version:  5.0
diff --git a/src/mesa/drivers/dri/i830/i830_context.c b/src/mesa/drivers/dri/i830/i830_context.c
index 254fa284626..36abd2fdbc4 100644
--- a/src/mesa/drivers/dri/i830/i830_context.c
+++ b/src/mesa/drivers/dri/i830/i830_context.c
@@ -167,9 +167,9 @@ static const char * const card_extensions[] =
 };
 
 
-extern const struct gl_pipeline_stage _i830_render_stage;
+extern const struct tnl_pipeline_stage _i830_render_stage;
 
-static const struct gl_pipeline_stage *i830_pipeline[] = {
+static const struct tnl_pipeline_stage *i830_pipeline[] = {
    &_tnl_vertex_transform_stage,
    &_tnl_normal_transform_stage,
    &_tnl_lighting_stage,
diff --git a/src/mesa/drivers/dri/i830/i830_render.c b/src/mesa/drivers/dri/i830/i830_render.c
index 5df05dfde87..f66acd3b564 100644
--- a/src/mesa/drivers/dri/i830/i830_render.c
+++ b/src/mesa/drivers/dri/i830/i830_render.c
@@ -122,7 +122,7 @@ static void VERT_FALLBACK( GLcontext *ctx,
    tnl->Driver.Render.BuildVertices( ctx, start, count, ~0 );
    tnl->Driver.Render.PrimTabVerts[flags&PRIM_MODE_MASK]( ctx, start, 
 							  count, flags );
-   I830_CONTEXT(ctx)->SetupNewInputs = VERT_BIT_CLIP;
+   I830_CONTEXT(ctx)->SetupNewInputs = VERT_BIT_POS;
 }
 
 
@@ -159,24 +159,27 @@ static GLboolean choose_render( struct vertex_buffer *VB, int bufsz )
    int nr_rprims = 0;
    int nr_rverts = 0;
    int rprim = 0;
-   int i = 0, length, flags = 0;
+   int i;
 
    
-   for (i = VB->FirstPrimitive ; !(flags & PRIM_LAST) ; i += length) {
-      flags = VB->Primitive[i];
-      length = VB->PrimitiveLength[i];
+   for (i = 0 ; i < VB->PrimitiveCount ; i++)
+   {
+      GLuint prim = VB->Primitive[i].mode;
+      GLuint start = VB->Primitive[i].start;
+      GLuint length = VB->Primitive[i].count;
+
       if (!length)
 	 continue;
 
-      if (!hw_prim[flags & PRIM_MODE_MASK])
+      if (!hw_prim[prim & PRIM_MODE_MASK])
 	 return GL_FALSE;
 
       nr_prims++;
-      nr_rverts += length * scale_prim[flags & PRIM_MODE_MASK];
+      nr_rverts += length * scale_prim[prim & PRIM_MODE_MASK];
 
-      if (reduced_prim[flags&PRIM_MODE_MASK] != rprim) {
+      if (reduced_prim[prim&PRIM_MODE_MASK] != rprim) {
 	 nr_rprims++;
-	 rprim = reduced_prim[flags&PRIM_MODE_MASK];
+	 rprim = reduced_prim[prim&PRIM_MODE_MASK];
       }
    }
 
@@ -192,7 +195,7 @@ static GLboolean choose_render( struct vertex_buffer *VB, int bufsz )
 
 
 static GLboolean i830_run_render( GLcontext *ctx, 
-				 struct gl_pipeline_stage *stage )
+				 struct tnl_pipeline_stage *stage )
 {
    i830ContextPtr imesa = I830_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
@@ -205,16 +208,21 @@ static GLboolean i830_run_render( GLcontext *ctx,
       return GL_TRUE;
    }
 
-   imesa->SetupNewInputs = VERT_BIT_CLIP;
+   imesa->SetupNewInputs = VERT_BIT_POS;
 
    tnl->Driver.Render.Start( ctx );
    
-   for (i = VB->FirstPrimitive ; !(flags & PRIM_LAST) ; i += length) {
-      flags = VB->Primitive[i];
-      length= VB->PrimitiveLength[i];
-      if (length)
-	 i830_render_tab_verts[flags & PRIM_MODE_MASK]( ctx, i, i + length,
-						        flags );
+   for (i = 0 ; i < VB->PrimitiveCount ; i++)
+   {
+      GLuint prim = VB->Primitive[i].mode;
+      GLuint start = VB->Primitive[i].start;
+      GLuint length = VB->Primitive[i].count;
+
+      if (!length)
+	 continue;
+
+      i830_render_tab_verts[prim & PRIM_MODE_MASK]( ctx, start, start + length,
+						    prim );
    }
       
    tnl->Driver.Render.Finish( ctx );
@@ -224,9 +232,9 @@ static GLboolean i830_run_render( GLcontext *ctx,
 
 
 static void i830_check_render( GLcontext *ctx, 
-			       struct gl_pipeline_stage *stage )
+			       struct tnl_pipeline_stage *stage )
 {
-   GLuint inputs = VERT_BIT_CLIP | VERT_BIT_COLOR0;
+   GLuint inputs = VERT_BIT_POS | VERT_BIT_COLOR0;
    if (ctx->RenderMode == GL_RENDER) {
       if (ctx->_TriangleCaps & DD_SEPARATE_SPECULAR)
 	 inputs |= VERT_BIT_COLOR1;
@@ -244,13 +252,13 @@ static void i830_check_render( GLcontext *ctx,
    stage->inputs = inputs;
 }
 
-static void dtr( struct gl_pipeline_stage *stage )
+static void dtr( struct tnl_pipeline_stage *stage )
 {
    (void)stage;
 }
 
 
-const struct gl_pipeline_stage _i830_render_stage =
+const struct tnl_pipeline_stage _i830_render_stage =
 {
    "i830 render",
    (_DD_NEW_SEPARATE_SPECULAR |
diff --git a/src/mesa/drivers/dri/i830/i830_vb.c b/src/mesa/drivers/dri/i830/i830_vb.c
index a7ac054c230..c9c5408dc99 100644
--- a/src/mesa/drivers/dri/i830/i830_vb.c
+++ b/src/mesa/drivers/dri/i830/i830_vb.c
@@ -445,7 +445,7 @@ void i830BuildVertices( GLcontext *ctx,
    if (!newinputs)
       return;
 
-   if (newinputs & VERT_BIT_CLIP) {
+   if (newinputs & VERT_BIT_POS) {
       setup_tab[imesa->SetupIndex].emit( ctx, start, count, v, stride );
    } else {
       GLuint ind = 0;
diff --git a/src/mesa/drivers/dri/mga/Makefile.X11 b/src/mesa/drivers/dri/mga/Makefile.X11
index cb769838227..246a7451840 100644
--- a/src/mesa/drivers/dri/mga/Makefile.X11
+++ b/src/mesa/drivers/dri/mga/Makefile.X11
@@ -1,4 +1,3 @@
-# $Id: Makefile.X11,v 1.5 2003/10/21 06:05:41 jonsmirl Exp $
 
 # Mesa 3-D graphics library
 # Version:  5.0
diff --git a/src/mesa/drivers/dri/mga/mga_xmesa.c b/src/mesa/drivers/dri/mga/mga_xmesa.c
index d95c0197ebb..11daf9aaf4f 100644
--- a/src/mesa/drivers/dri/mga/mga_xmesa.c
+++ b/src/mesa/drivers/dri/mga/mga_xmesa.c
@@ -261,9 +261,9 @@ mgaDestroyScreen(__DRIscreenPrivate *sPriv)
 }
 
 
-extern const struct gl_pipeline_stage _mga_render_stage;
+extern const struct tnl_pipeline_stage _mga_render_stage;
 
-static const struct gl_pipeline_stage *mga_pipeline[] = {
+static const struct tnl_pipeline_stage *mga_pipeline[] = {
    &_tnl_vertex_transform_stage, 
    &_tnl_normal_transform_stage, 
    &_tnl_lighting_stage,	
@@ -683,7 +683,7 @@ void mgaGetLock( mgaContextPtr mmesa, GLuint flags )
 
    if (*(dPriv->pStamp) != mmesa->lastStamp) {
       mmesa->lastStamp = *(dPriv->pStamp);
-      mmesa->SetupNewInputs |= VERT_BIT_CLIP;
+      mmesa->SetupNewInputs |= VERT_BIT_POS;
       mmesa->dirty_cliprects = (MGA_FRONT|MGA_BACK);
       mgaUpdateRects( mmesa, (MGA_FRONT|MGA_BACK) );
    }
diff --git a/src/mesa/drivers/dri/mga/mgarender.c b/src/mesa/drivers/dri/mga/mgarender.c
index 516a99099fd..5fb7af908c4 100644
--- a/src/mesa/drivers/dri/mga/mgarender.c
+++ b/src/mesa/drivers/dri/mga/mgarender.c
@@ -101,7 +101,7 @@ static void VERT_FALLBACK( GLcontext *ctx, GLuint start, GLuint count,
    tnl->Driver.Render.PrimitiveNotify( ctx, flags & PRIM_MODE_MASK );
    tnl->Driver.Render.BuildVertices( ctx, start, count, ~0 );
    tnl->Driver.Render.PrimTabVerts[flags&PRIM_MODE_MASK]( ctx, start, count, flags );
-   MGA_CONTEXT(ctx)->SetupNewInputs |= VERT_BIT_CLIP;
+   MGA_CONTEXT(ctx)->SetupNewInputs |= VERT_BIT_POS;
 }
 
 #define LOCAL_VARS mgaContextPtr mmesa = MGA_CONTEXT(ctx) 
@@ -131,12 +131,12 @@ static void VERT_FALLBACK( GLcontext *ctx, GLuint start, GLuint count,
 
 
 static GLboolean mga_run_render( GLcontext *ctx,
-				  struct gl_pipeline_stage *stage )
+				  struct tnl_pipeline_stage *stage )
 {
    mgaContextPtr mmesa = MGA_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
    struct vertex_buffer *VB = &tnl->vb; 
-   GLuint i, length, flags = 0;
+   GLuint i;
 
    /* Don't handle clipping or indexed vertices or vertex manipulations.
     */
@@ -147,13 +147,17 @@ static GLboolean mga_run_render( GLcontext *ctx,
    tnl->Driver.Render.Start( ctx );
    mmesa->SetupNewInputs = ~0;      
 
-   for (i = VB->FirstPrimitive ; !(flags & PRIM_LAST) ; i += length)
+   for (i = 0 ; i < VB->PrimitiveCount ; i++)
    {
-      flags = VB->Primitive[i];
-      length= VB->PrimitiveLength[i];	
-      if (length)
-	 mga_render_tab_verts[flags & PRIM_MODE_MASK]( ctx, i, i + length,
-						       flags );
+      GLuint prim = VB->Primitive[i].mode;
+      GLuint start = VB->Primitive[i].start;
+      GLuint length = VB->Primitive[i].count;
+
+      if (!length)
+	 continue;
+
+      mga_render_tab_verts[prim & PRIM_MODE_MASK]( ctx, start, start + length, 
+						   prim);
    } 
 
    tnl->Driver.Render.Finish( ctx );
@@ -162,9 +166,9 @@ static GLboolean mga_run_render( GLcontext *ctx,
 }
 
 
-static void mga_check_render( GLcontext *ctx, struct gl_pipeline_stage *stage )
+static void mga_check_render( GLcontext *ctx, struct tnl_pipeline_stage *stage )
 {
-   GLuint inputs = VERT_BIT_CLIP | VERT_BIT_COLOR0;
+   GLuint inputs = VERT_BIT_POS | VERT_BIT_COLOR0;
 
    if (ctx->RenderMode == GL_RENDER) {
       if (ctx->_TriangleCaps & DD_SEPARATE_SPECULAR) 
@@ -184,13 +188,13 @@ static void mga_check_render( GLcontext *ctx, struct gl_pipeline_stage *stage )
 }
 
 
-static void dtr( struct gl_pipeline_stage *stage )
+static void dtr( struct tnl_pipeline_stage *stage )
 {
    (void)stage;
 }
 
 
-const struct gl_pipeline_stage _mga_render_stage = 
+const struct tnl_pipeline_stage _mga_render_stage = 
 { 
    "mga render",
    (_DD_NEW_SEPARATE_SPECULAR |
diff --git a/src/mesa/drivers/dri/mga/mgavb.c b/src/mesa/drivers/dri/mga/mgavb.c
index 17dde2bba60..63de3563c45 100644
--- a/src/mesa/drivers/dri/mga/mgavb.c
+++ b/src/mesa/drivers/dri/mga/mgavb.c
@@ -362,7 +362,7 @@ void mgaBuildVertices( GLcontext *ctx,
    if (!newinputs)
       return;
 
-   if (newinputs & VERT_BIT_CLIP) {
+   if (newinputs & VERT_BIT_POS) {
       setup_tab[mmesa->SetupIndex].emit( ctx, start, count, v, stride );   
    } else {
       GLuint ind = 0;
diff --git a/src/mesa/drivers/dri/r128/Makefile.X11 b/src/mesa/drivers/dri/r128/Makefile.X11
index e8f07a5e3b1..c77b0e44459 100644
--- a/src/mesa/drivers/dri/r128/Makefile.X11
+++ b/src/mesa/drivers/dri/r128/Makefile.X11
@@ -1,4 +1,3 @@
-# $Id: Makefile.X11,v 1.3 2003/10/21 06:05:43 jonsmirl Exp $
 
 # Mesa 3-D graphics library
 # Version:  5.0
diff --git a/src/mesa/drivers/dri/r128/r128_vb.c b/src/mesa/drivers/dri/r128/r128_vb.c
index 0387e5dc977..6c52975286f 100644
--- a/src/mesa/drivers/dri/r128/r128_vb.c
+++ b/src/mesa/drivers/dri/r128/r128_vb.c
@@ -382,7 +382,7 @@ void r128BuildVertices( GLcontext *ctx,
    if (!newinputs)
       return;
 
-   if (newinputs & VERT_BIT_CLIP) {
+   if (newinputs & VERT_BIT_POS) {
       setup_tab[rmesa->SetupIndex].emit( ctx, start, count, v, stride );
    } else {
       GLuint ind = 0;
diff --git a/src/mesa/drivers/dri/r200/Makefile.X11 b/src/mesa/drivers/dri/r200/Makefile.X11
index aad26a70acc..bfa332af25e 100644
--- a/src/mesa/drivers/dri/r200/Makefile.X11
+++ b/src/mesa/drivers/dri/r200/Makefile.X11
@@ -1,4 +1,3 @@
-# $Id: Makefile.X11,v 1.6 2003/10/21 06:05:46 jonsmirl Exp $
 
 # Mesa 3-D graphics library
 # Version:  5.0
diff --git a/src/mesa/drivers/dri/r200/r200_context.c b/src/mesa/drivers/dri/r200/r200_context.c
index b64ee2b1f29..53aa711cc1e 100644
--- a/src/mesa/drivers/dri/r200/r200_context.c
+++ b/src/mesa/drivers/dri/r200/r200_context.c
@@ -169,10 +169,10 @@ static const char * const card_extensions[] =
     NULL
 };
 
-extern const struct gl_pipeline_stage _r200_render_stage;
-extern const struct gl_pipeline_stage _r200_tcl_stage;
+extern const struct tnl_pipeline_stage _r200_render_stage;
+extern const struct tnl_pipeline_stage _r200_tcl_stage;
 
-static const struct gl_pipeline_stage *r200_pipeline[] = {
+static const struct tnl_pipeline_stage *r200_pipeline[] = {
 
    /* Try and go straight to t&l
     */
diff --git a/src/mesa/drivers/dri/r200/r200_maos_arrays.c b/src/mesa/drivers/dri/r200/r200_maos_arrays.c
index 7842c0c594f..b442dce207a 100644
--- a/src/mesa/drivers/dri/r200/r200_maos_arrays.c
+++ b/src/mesa/drivers/dri/r200/r200_maos_arrays.c
@@ -43,7 +43,6 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 #include "math/m_translate.h"
 #include "tnl/tnl.h"
 #include "tnl/t_context.h"
-#include "tnl/t_imm_debug.h"
 
 #include "r200_context.h"
 #include "r200_ioctl.h"
@@ -341,9 +340,6 @@ void r200EmitArrays( GLcontext *ctx, GLuint inputs )
    GLuint vfmt0 = 0, vfmt1 = 0;
    GLuint count = VB->Count;
    
-   if (R200_DEBUG & DEBUG_VERTS) 
-      _tnl_print_vert_flags( __FUNCTION__, inputs );
-
    if (1) {
       if (!rmesa->tcl.obj.buf) 
 	 emit_vector( ctx, 
@@ -358,7 +354,7 @@ void r200EmitArrays( GLcontext *ctx, GLuint inputs )
       case 3: vfmt0 |= R200_VTX_Z0;
       case 2: 
       default:
-	break;
+	 break;
       }
       component[nr++] = &rmesa->tcl.obj;
    }
@@ -378,60 +374,44 @@ void r200EmitArrays( GLcontext *ctx, GLuint inputs )
    }
 
    if (inputs & VERT_BIT_COLOR0) {
-      if (VB->ColorPtr[0]->Type == GL_UNSIGNED_BYTE) {
-	 if (!rmesa->tcl.rgba.buf)
-	    emit_ubyte_rgba( ctx, 
-			     &rmesa->tcl.rgba, 
-			     (char *)VB->ColorPtr[0]->Ptr,
-			     VB->ColorPtr[0]->Size,
-			     VB->ColorPtr[0]->StrideB,
-			     count);
-
-	 vfmt0 |= R200_VTX_PK_RGBA << R200_VTX_COLOR_0_SHIFT; 
+      int emitsize;
+
+      if (VB->ColorPtr[0]->size == 4 &&
+	  (VB->ColorPtr[0]->stride != 0 ||
+	   VB->ColorPtr[0]->data[0][3] != 1.0)) { 
+	 vfmt0 |= R200_VTX_FP_RGBA << R200_VTX_COLOR_0_SHIFT; 
+	 emitsize = 4;
       }
-      else {
-	 int emitsize;
-
-	 if (VB->ColorPtr[0]->Size == 4 &&
-	     (VB->ColorPtr[0]->StrideB != 0 ||
-	      ((GLfloat *)VB->ColorPtr[0]->Ptr)[3] != 1.0)) { 
-	    vfmt0 |= R200_VTX_FP_RGBA << R200_VTX_COLOR_0_SHIFT; 
-	    emitsize = 4;
-	 }
-	 else { 
-	    vfmt0 |= R200_VTX_FP_RGB << R200_VTX_COLOR_0_SHIFT; 
-	    emitsize = 3;
-	 }
-
-	 if (!rmesa->tcl.rgba.buf)
-	    emit_vector( ctx, 
-			 &(rmesa->tcl.rgba), 
-			 (char *)VB->ColorPtr[0]->Ptr,
-			 emitsize,
-			 VB->ColorPtr[0]->StrideB,
-			 count);
+      else { 
+	 vfmt0 |= R200_VTX_FP_RGB << R200_VTX_COLOR_0_SHIFT; 
+	 emitsize = 3;
       }
 
+      if (!rmesa->tcl.rgba.buf)
+	 emit_vector( ctx, 
+		      &(rmesa->tcl.rgba), 
+		      (char *)VB->ColorPtr[0]->data,
+		      emitsize,
+		      VB->ColorPtr[0]->stride,
+		      count);
+
       component[nr++] = &rmesa->tcl.rgba;
    }
 
 
    if (inputs & VERT_BIT_COLOR1) {
       if (!rmesa->tcl.spec.buf) {
-	 if (VB->SecondaryColorPtr[0]->Type != GL_UNSIGNED_BYTE)
-	    r200_import_float_spec_colors( ctx );
-
-	 emit_ubyte_rgba( ctx, 
-			  &rmesa->tcl.spec, 
-			  (char *)VB->SecondaryColorPtr[0]->Ptr,
-			  3,
-			  VB->SecondaryColorPtr[0]->StrideB,
-			  count);
+	 emit_vector( ctx, 
+		      &rmesa->tcl.spec, 
+		      (char *)VB->SecondaryColorPtr[0]->data,
+		      3,
+		      VB->SecondaryColorPtr[0]->stride,
+		      count);
       }
 
       /* How does this work?
        */
-      vfmt0 |= R200_VTX_PK_RGBA << R200_VTX_COLOR_1_SHIFT; 
+      vfmt0 |= R200_VTX_FP_RGB << R200_VTX_COLOR_1_SHIFT; 
       component[nr++] = &rmesa->tcl.spec;
    }
 
@@ -480,8 +460,8 @@ void r200ReleaseArrays( GLcontext *ctx, GLuint newinputs )
 {
    r200ContextPtr rmesa = R200_CONTEXT( ctx );
 
-   if (R200_DEBUG & DEBUG_VERTS) 
-      _tnl_print_vert_flags( __FUNCTION__, newinputs );
+/*    if (R200_DEBUG & DEBUG_VERTS)  */
+/*       _tnl_print_vert_flags( __FUNCTION__, newinputs ); */
 
    if (newinputs & VERT_BIT_POS) 
      r200ReleaseDmaRegion( rmesa, &rmesa->tcl.obj, __FUNCTION__ );
diff --git a/src/mesa/drivers/dri/r200/r200_state.c b/src/mesa/drivers/dri/r200/r200_state.c
index 55a00cd4de7..27dc6fd1f2b 100644
--- a/src/mesa/drivers/dri/r200/r200_state.c
+++ b/src/mesa/drivers/dri/r200/r200_state.c
@@ -2089,10 +2089,30 @@ static void r200InvalidateState( GLcontext *ctx, GLuint new_state )
    r200VtxfmtInvalidate( ctx );
 }
 
+/* A hack.  The r200 can actually cope just fine with materials
+ * between begin/ends, so fix this.
+ */
+static GLboolean check_material( GLcontext *ctx )
+{
+   TNLcontext *tnl = TNL_CONTEXT(ctx);
+   GLint i;
+
+   for (i = _TNL_ATTRIB_MAT_FRONT_AMBIENT; 
+	i < _TNL_ATTRIB_MAT_BACK_INDEXES; 
+	i++)
+      if (tnl->vb.AttribPtr[i] &&
+	  tnl->vb.AttribPtr[i]->stride)
+	 return GL_TRUE;
+
+   return GL_FALSE;
+}
+      
+
 static void r200WrapRunPipeline( GLcontext *ctx )
 {
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
+   GLboolean has_material;
 
    if (0)
       fprintf(stderr, "%s, newstate: %x\n", __FUNCTION__, rmesa->NewGLState);
@@ -2102,7 +2122,9 @@ static void r200WrapRunPipeline( GLcontext *ctx )
    if (rmesa->NewGLState)
       r200ValidateState( ctx );
 
-   if (tnl->vb.Material) {
+   has_material = (ctx->Light.Enabled && check_material( ctx ));
+
+   if (has_material) {
       TCL_FALLBACK( ctx, R200_TCL_FALLBACK_MATERIAL, GL_TRUE );
    }
 
@@ -2110,7 +2132,7 @@ static void r200WrapRunPipeline( GLcontext *ctx )
     */ 
    _tnl_run_pipeline( ctx );
 
-   if (tnl->vb.Material) {
+   if (has_material) {
       TCL_FALLBACK( ctx, R200_TCL_FALLBACK_MATERIAL, GL_FALSE );
       r200UpdateMaterial( ctx ); /* not needed any more? */
    }
diff --git a/src/mesa/drivers/dri/r200/r200_swtcl.c b/src/mesa/drivers/dri/r200/r200_swtcl.c
index 9d98358d45e..ca5c456889d 100644
--- a/src/mesa/drivers/dri/r200/r200_swtcl.c
+++ b/src/mesa/drivers/dri/r200/r200_swtcl.c
@@ -47,7 +47,6 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 #include "math/m_translate.h"
 #include "tnl/tnl.h"
 #include "tnl/t_context.h"
-#include "tnl/t_imm_exec.h"
 #include "tnl/t_pipeline.h"
 
 #include "r200_context.h"
@@ -537,7 +536,7 @@ static void VERT_FALLBACK( GLcontext *ctx,
    tnl->Driver.Render.PrimitiveNotify( ctx, flags & PRIM_MODE_MASK );
    tnl->Driver.Render.BuildVertices( ctx, start, count, ~0 );
    tnl->Driver.Render.PrimTabVerts[flags&PRIM_MODE_MASK]( ctx, start, count, flags );
-   R200_CONTEXT(ctx)->swtcl.SetupNewInputs = VERT_BIT_CLIP;
+   R200_CONTEXT(ctx)->swtcl.SetupNewInputs = _TNL_BIT_POS;
 }
 
 static void ELT_FALLBACK( GLcontext *ctx,
@@ -549,7 +548,7 @@ static void ELT_FALLBACK( GLcontext *ctx,
    tnl->Driver.Render.PrimitiveNotify( ctx, flags & PRIM_MODE_MASK );
    tnl->Driver.Render.BuildVertices( ctx, start, count, ~0 );
    tnl->Driver.Render.PrimTabElts[flags&PRIM_MODE_MASK]( ctx, start, count, flags );
-   R200_CONTEXT(ctx)->swtcl.SetupNewInputs = VERT_BIT_CLIP;
+   R200_CONTEXT(ctx)->swtcl.SetupNewInputs = _TNL_BIT_POS;
 }
 
 
@@ -631,7 +630,7 @@ do {									\
 
 
 static GLboolean r200_run_render( GLcontext *ctx,
-				    struct gl_pipeline_stage *stage )
+				    struct tnl_pipeline_stage *stage )
 {
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
@@ -659,18 +658,21 @@ static GLboolean r200_run_render( GLcontext *ctx,
 
    tnl->Driver.Render.Start( ctx );
 
-   for (i = 0 ; !(flags & PRIM_LAST) ; i += length)
+   for (i = 0 ; i < VB->PrimitiveCount ; i++)
    {
-      flags = VB->Primitive[i];
-      length = VB->PrimitiveLength[i];
+      GLuint prim = VB->Primitive[i].mode;
+      GLuint start = VB->Primitive[i].start;
+      GLuint length = VB->Primitive[i].count;
+
+      if (!length)
+	 continue;
 
       if (R200_DEBUG & DEBUG_PRIMS)
 	 fprintf(stderr, "r200_render.c: prim %s %d..%d\n", 
-		 _mesa_lookup_enum_by_nr(flags & PRIM_MODE_MASK), 
-		 i, i+length);
+		 _mesa_lookup_enum_by_nr(prim & PRIM_MODE_MASK), 
+		 start, start+length);
 
-      if (length)
-	 tab[flags & PRIM_MODE_MASK]( ctx, i, i + length, flags );
+      tab[prim & PRIM_MODE_MASK]( ctx, start, start + length, flags );
    }
 
    tnl->Driver.Render.Finish( ctx );
@@ -681,35 +683,35 @@ static GLboolean r200_run_render( GLcontext *ctx,
 
 
 static void r200_check_render( GLcontext *ctx,
-				 struct gl_pipeline_stage *stage )
+				 struct tnl_pipeline_stage *stage )
 {
-   GLuint inputs = VERT_BIT_POS | VERT_BIT_CLIP | VERT_BIT_COLOR0;
+   GLuint inputs = _TNL_BIT_POS | _TNL_BIT_COLOR0;
 
    if (ctx->RenderMode == GL_RENDER) {
       if (ctx->_TriangleCaps & DD_SEPARATE_SPECULAR)
-	 inputs |= VERT_BIT_COLOR1;
+	 inputs |= _TNL_BIT_COLOR1;
 
       if (ctx->Texture.Unit[0]._ReallyEnabled)
-	 inputs |= VERT_BIT_TEX0;
+	 inputs |= _TNL_BIT_TEX0;
 
       if (ctx->Texture.Unit[1]._ReallyEnabled)
-	 inputs |= VERT_BIT_TEX1;
+	 inputs |= _TNL_BIT_TEX1;
 
       if (ctx->Fog.Enabled)
-	 inputs |= VERT_BIT_FOG;
+	 inputs |= _TNL_BIT_FOG;
    }
 
    stage->inputs = inputs;
 }
 
 
-static void dtr( struct gl_pipeline_stage *stage )
+static void dtr( struct tnl_pipeline_stage *stage )
 {
    (void)stage;
 }
 
 
-const struct gl_pipeline_stage _r200_render_stage =
+const struct tnl_pipeline_stage _r200_render_stage =
 {
    "r200 render",
    (_DD_NEW_SEPARATE_SPECULAR |
@@ -1230,7 +1232,7 @@ r200PointsBitmap( GLcontext *ctx, GLint px, GLint py,
 
 void r200FlushVertices( GLcontext *ctx, GLuint flags )
 {
-   _tnl_flush_vertices( ctx, flags );
+   _tnl_FlushVertices( ctx, flags );
 
    if (flags & FLUSH_STORED_VERTICES)
       R200_NEWPRIM( R200_CONTEXT( ctx ) );
diff --git a/src/mesa/drivers/dri/r200/r200_tcl.c b/src/mesa/drivers/dri/r200/r200_tcl.c
index b02a7cc559b..60e281ffaf3 100644
--- a/src/mesa/drivers/dri/r200/r200_tcl.c
+++ b/src/mesa/drivers/dri/r200/r200_tcl.c
@@ -254,7 +254,7 @@ void r200TclPrimitive( GLcontext *ctx,
 /* TCL render.
  */
 static GLboolean r200_run_tcl_render( GLcontext *ctx,
-					struct gl_pipeline_stage *stage )
+				      struct tnl_pipeline_stage *stage )
 {
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
@@ -277,24 +277,19 @@ static GLboolean r200_run_tcl_render( GLcontext *ctx,
 
    rmesa->tcl.Elts = VB->Elts;
 
-   for (i = VB->FirstPrimitive ; !(flags & PRIM_LAST) ; i += length)
+   for (i = 0 ; i < VB->PrimitiveCount ; i++)
    {
-      flags = VB->Primitive[i];
-      length = VB->PrimitiveLength[i];
-
-      if (R200_DEBUG & DEBUG_PRIMS)
-	 fprintf(stderr, "%s: prim %s %d..%d\n", 
-		 __FUNCTION__,
-		 _mesa_lookup_enum_by_nr(flags & PRIM_MODE_MASK), 
-		 i, i+length);
+      GLuint prim = VB->Primitive[i].mode;
+      GLuint start = VB->Primitive[i].start;
+      GLuint length = VB->Primitive[i].count;
 
       if (!length)
 	 continue;
 
       if (rmesa->tcl.Elts)
-	 r200EmitEltPrimitive( ctx, i, i+length, flags );
+	 r200EmitEltPrimitive( ctx, start, start+length, prim );
       else
-	 r200EmitPrimitive( ctx, i, i+length, flags );
+	 r200EmitPrimitive( ctx, start, start+length, prim );
    }
 
    return GL_FALSE;		/* finished the pipe */
@@ -303,7 +298,7 @@ static GLboolean r200_run_tcl_render( GLcontext *ctx,
 
 
 static void r200_check_tcl_render( GLcontext *ctx,
-				     struct gl_pipeline_stage *stage )
+				   struct tnl_pipeline_stage *stage )
 {
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
    GLuint inputs = VERT_BIT_POS;
@@ -359,13 +354,13 @@ static void r200_check_tcl_render( GLcontext *ctx,
 }
 
 static void r200_init_tcl_render( GLcontext *ctx,
-				    struct gl_pipeline_stage *stage )
+				    struct tnl_pipeline_stage *stage )
 {
    stage->check = r200_check_tcl_render;
    stage->check( ctx, stage );
 }
 
-static void dtr( struct gl_pipeline_stage *stage )
+static void dtr( struct tnl_pipeline_stage *stage )
 {
    (void)stage;
 }
@@ -373,7 +368,7 @@ static void dtr( struct gl_pipeline_stage *stage )
 
 /* Initial state for tcl stage.  
  */
-const struct gl_pipeline_stage _r200_tcl_stage =
+const struct tnl_pipeline_stage _r200_tcl_stage =
 {
    "r200 render",
    (_DD_NEW_SEPARATE_SPECULAR |
diff --git a/src/mesa/drivers/dri/r200/r200_vtxfmt.c b/src/mesa/drivers/dri/r200/r200_vtxfmt.c
index a617d3f0290..2e742d4d04f 100644
--- a/src/mesa/drivers/dri/r200/r200_vtxfmt.c
+++ b/src/mesa/drivers/dri/r200/r200_vtxfmt.c
@@ -74,10 +74,10 @@ static void count_funcs( r200ContextPtr rmesa )
    count_func( "Vertex2fv", &rmesa->vb.dfn_cache.Vertex2fv );
    count_func( "Vertex3f", &rmesa->vb.dfn_cache.Vertex3f );
    count_func( "Vertex3fv", &rmesa->vb.dfn_cache.Vertex3fv );
-   count_func( "Color4ub", &rmesa->vb.dfn_cache.Color4ub );
-   count_func( "Color4ubv", &rmesa->vb.dfn_cache.Color4ubv );
-   count_func( "Color3ub", &rmesa->vb.dfn_cache.Color3ub );
-   count_func( "Color3ubv", &rmesa->vb.dfn_cache.Color3ubv );
+   count_func( "Color4ub", &rmesa->vb.dfn_cache.Color4ub ); 
+   count_func( "Color4ubv", &rmesa->vb.dfn_cache.Color4ubv ); 
+   count_func( "Color3ub", &rmesa->vb.dfn_cache.Color3ub ); 
+   count_func( "Color3ubv", &rmesa->vb.dfn_cache.Color3ubv ); 
    count_func( "Color4f", &rmesa->vb.dfn_cache.Color4f );
    count_func( "Color4fv", &rmesa->vb.dfn_cache.Color4fv );
    count_func( "Color3f", &rmesa->vb.dfn_cache.Color3f );
@@ -467,7 +467,10 @@ static void VFMT_FALLBACK( const char *caller )
       glNormal3fv( rmesa->vb.normalptr );
 
    if (VTX_COLOR(ind0, 0) == R200_VTX_PK_RGBA) 
-         glColor4ub( rmesa->vb.colorptr->red, rmesa->vb.colorptr->green, rmesa->vb.colorptr->blue, rmesa->vb.colorptr->alpha );
+         glColor4ub( rmesa->vb.colorptr->red,
+		     rmesa->vb.colorptr->green,
+		     rmesa->vb.colorptr->blue,
+		     rmesa->vb.colorptr->alpha );
    else if (VTX_COLOR(ind0, 0) == R200_VTX_FP_RGBA) 
       glColor4fv( rmesa->vb.floatcolorptr );
    else if (VTX_COLOR(ind0, 0) == R200_VTX_FP_RGB) {
@@ -481,7 +484,9 @@ static void VFMT_FALLBACK( const char *caller )
    }
 
    if (VTX_COLOR(ind0, 1) == R200_VTX_PK_RGBA) 
-      _glapi_Dispatch->SecondaryColor3ubEXT( rmesa->vb.specptr->red, rmesa->vb.specptr->green, rmesa->vb.specptr->blue ); 
+      _glapi_Dispatch->SecondaryColor3ubEXT( rmesa->vb.specptr->red, 
+					     rmesa->vb.specptr->green,
+					     rmesa->vb.specptr->blue ); 
 
    if (ind1 & (7 << R200_VTX_TEX0_COMP_CNT_SHIFT)) 
       glTexCoord2fv( rmesa->vb.texcoordptr[0] );
@@ -1005,13 +1010,14 @@ void r200VtxfmtInit( GLcontext *ctx, GLboolean useCodegen )
    vfmt->FogCoordfEXT = _mesa_noop_FogCoordfEXT;
    vfmt->EdgeFlag = _mesa_noop_EdgeFlag;
    vfmt->EdgeFlagv = _mesa_noop_EdgeFlagv;
-   vfmt->Indexi = _mesa_noop_Indexi;
-   vfmt->Indexiv = _mesa_noop_Indexiv;
+   vfmt->Indexf = _mesa_noop_Indexf;
+   vfmt->Indexfv = _mesa_noop_Indexfv;
 
 
    /* Active but unsupported -- fallback if we receive these:
     */
    vfmt->CallList = r200_fallback_CallList;
+   vfmt->CallLists = r200_fallback_CallLists;
    vfmt->EvalCoord1f = r200_fallback_EvalCoord1f;
    vfmt->EvalCoord1fv = r200_fallback_EvalCoord1fv;
    vfmt->EvalCoord2f = r200_fallback_EvalCoord2f;
diff --git a/src/mesa/drivers/dri/r200/r200_vtxfmt_c.c b/src/mesa/drivers/dri/r200/r200_vtxfmt_c.c
index efef751a616..6f38fffc276 100644
--- a/src/mesa/drivers/dri/r200/r200_vtxfmt_c.c
+++ b/src/mesa/drivers/dri/r200/r200_vtxfmt_c.c
@@ -826,16 +826,10 @@ void r200VtxfmtInitChoosers( GLvertexformat *vfmt )
 {
    vfmt->Color3f = choose_Color3f;
    vfmt->Color3fv = choose_Color3fv;
-   vfmt->Color3ub = choose_Color3ub;
-   vfmt->Color3ubv = choose_Color3ubv;
    vfmt->Color4f = choose_Color4f;
    vfmt->Color4fv = choose_Color4fv;
-   vfmt->Color4ub = choose_Color4ub;
-   vfmt->Color4ubv = choose_Color4ubv;
    vfmt->SecondaryColor3fEXT = choose_SecondaryColor3fEXT;
    vfmt->SecondaryColor3fvEXT = choose_SecondaryColor3fvEXT;
-   vfmt->SecondaryColor3ubEXT = choose_SecondaryColor3ubEXT;
-   vfmt->SecondaryColor3ubvEXT = choose_SecondaryColor3ubvEXT;
    vfmt->MultiTexCoord1fARB = choose_MultiTexCoord1fARB;
    vfmt->MultiTexCoord1fvARB = choose_MultiTexCoord1fvARB;
    vfmt->MultiTexCoord2fARB = choose_MultiTexCoord2fARB;
@@ -850,6 +844,17 @@ void r200VtxfmtInitChoosers( GLvertexformat *vfmt )
    vfmt->Vertex2fv = choose_Vertex2fv;
    vfmt->Vertex3f = choose_Vertex3f;
    vfmt->Vertex3fv = choose_Vertex3fv;
+
+   /* TODO: restore ubyte colors to vtxfmt.
+    */
+#if 0
+   vfmt->Color3ub = choose_Color3ub;
+   vfmt->Color3ubv = choose_Color3ubv;
+   vfmt->Color4ub = choose_Color4ub;
+   vfmt->Color4ubv = choose_Color4ubv;
+   vfmt->SecondaryColor3ubEXT = choose_SecondaryColor3ubEXT;
+   vfmt->SecondaryColor3ubvEXT = choose_SecondaryColor3ubvEXT;
+#endif
 }
 
 
diff --git a/src/mesa/drivers/dri/radeon/Makefile.X11 b/src/mesa/drivers/dri/radeon/Makefile.X11
index 19dabce262c..fd379e9a0d6 100644
--- a/src/mesa/drivers/dri/radeon/Makefile.X11
+++ b/src/mesa/drivers/dri/radeon/Makefile.X11
@@ -1,4 +1,3 @@
-# $Id: Makefile.X11,v 1.3 2003/10/21 06:05:49 jonsmirl Exp $
 
 # Mesa 3-D graphics library
 # Version:  5.0
diff --git a/src/mesa/drivers/dri/radeon/radeon_context.c b/src/mesa/drivers/dri/radeon/radeon_context.c
index 5604d9c8c21..02bfb23ca59 100644
--- a/src/mesa/drivers/dri/radeon/radeon_context.c
+++ b/src/mesa/drivers/dri/radeon/radeon_context.c
@@ -165,11 +165,11 @@ static const char * const card_extensions[] =
     NULL
 };
 
-extern const struct gl_pipeline_stage _radeon_texrect_stage;
-extern const struct gl_pipeline_stage _radeon_render_stage;
-extern const struct gl_pipeline_stage _radeon_tcl_stage;
+extern const struct tnl_pipeline_stage _radeon_texrect_stage;
+extern const struct tnl_pipeline_stage _radeon_render_stage;
+extern const struct tnl_pipeline_stage _radeon_tcl_stage;
 
-static const struct gl_pipeline_stage *radeon_pipeline[] = {
+static const struct tnl_pipeline_stage *radeon_pipeline[] = {
 
    /* Try and go straight to t&l
     */
@@ -412,6 +412,9 @@ radeonCreateContext( const __GLcontextModes *glVisual,
    radeonInitState( rmesa );
    radeonInitSwtcl( ctx );
 
+   _mesa_vector4f_alloc( &rmesa->tcl.ObjClean, 0, 
+			 ctx->Const.MaxArrayLockSize, 32 );
+
    fthrottle_mode = driQueryOptioni(&rmesa->optionCache, "fthrottle_mode");
    rmesa->iw.irq_seq = -1;
    rmesa->irqsEmitted = 0;
@@ -505,6 +508,8 @@ void radeonDestroyContext( __DRIcontextPrivate *driContextPriv )
       rmesa->glCtx->DriverCtx = NULL;
       _mesa_destroy_context( rmesa->glCtx );
 
+      _mesa_vector4f_free( &rmesa->tcl.ObjClean );
+
       if (rmesa->state.scissor.pClipRects) {
 	 FREE(rmesa->state.scissor.pClipRects);
 	 rmesa->state.scissor.pClipRects = 0;
diff --git a/src/mesa/drivers/dri/radeon/radeon_context.h b/src/mesa/drivers/dri/radeon/radeon_context.h
index 69808837ea4..b0c3c1919e1 100644
--- a/src/mesa/drivers/dri/radeon/radeon_context.h
+++ b/src/mesa/drivers/dri/radeon/radeon_context.h
@@ -57,6 +57,8 @@ typedef struct radeon_context *radeonContextPtr;
 #include "radeon_screen.h"
 #include "mm.h"
 
+#include "math/m_vector.h"
+
 /* Flags for software fallback cases */
 /* See correponding strings in radeon_swtcl.c */
 #define RADEON_FALLBACK_TEXTURE		0x0001
@@ -503,6 +505,11 @@ struct radeon_tcl_info {
    GLint last_offset;
    GLuint hw_primitive;
 
+   /* Temporary for cases where incoming vertex data is incompatible
+    * with maos code.
+    */
+   GLvector4f ObjClean;
+
    struct radeon_dma_region *aos_components[8];
    GLuint nr_aos_components;
 
@@ -699,12 +706,6 @@ struct radeon_context {
    GLuint Fallback;
    GLuint NewGLState;
 
-   
-   /* Temporaries for translating away float colors:
-    */
-   struct gl_client_array UbyteColor;
-   struct gl_client_array UbyteSecondaryColor;
-
    /* Vertex buffers
     */
    struct radeon_ioctl ioctl;
diff --git a/src/mesa/drivers/dri/radeon/radeon_maos_vbtmp.h b/src/mesa/drivers/dri/radeon/radeon_maos_vbtmp.h
index b379bad985f..c16234a9437 100644
--- a/src/mesa/drivers/dri/radeon/radeon_maos_vbtmp.h
+++ b/src/mesa/drivers/dri/radeon/radeon_maos_vbtmp.h
@@ -39,13 +39,14 @@ static void TAG(emit)( GLcontext *ctx,
 		       void *dest )
 {
    LOCALVARS
-   struct vertex_buffer *VB = &TNL_CONTEXT(ctx)->vb;
+      struct vertex_buffer *VB = &TNL_CONTEXT(ctx)->vb;
    GLuint (*tc0)[4], (*tc1)[4], (*tc2)[4];
+   GLfloat (*col)[4], (*spec)[4];
    GLfloat (*fog)[4];
    GLuint (*norm)[4];
-   GLubyte (*col)[4], (*spec)[4];
    GLuint tc0_stride, tc1_stride, col_stride, spec_stride, fog_stride;
    GLuint tc2_stride, norm_stride;
+   GLuint fill_tex = 0;
    GLuint (*coord)[4];
    GLuint coord_stride; /* object coordinates */
    GLubyte dummy[4];
@@ -56,36 +57,20 @@ static void TAG(emit)( GLcontext *ctx,
    if (RADEON_DEBUG & DEBUG_VERTS)
       fprintf(stderr, "%s\n", __FUNCTION__); 
 
-   /* The vertex code expects Obj to be clean to element 3.  To fix
-    * this, add more vertex code (for obj-2, obj-3) or preferably move
-    * to maos.  
-    */
-   if (VB->ObjPtr->size < 3) {
-      if (VB->ObjPtr->flags & VEC_NOT_WRITEABLE) {
-	 VB->import_data( ctx, VERT_BIT_POS, VEC_NOT_WRITEABLE );
-      }
-      _mesa_vector4f_clean_elem( VB->ObjPtr, VB->Count, 2 );
-   }
-
-   if (DO_W && VB->ObjPtr->size < 4) {
-      if (VB->ObjPtr->flags & VEC_NOT_WRITEABLE) {
-	 VB->import_data( ctx, VERT_BIT_POS, VEC_NOT_WRITEABLE );
-      }
-      _mesa_vector4f_clean_elem( VB->ObjPtr, VB->Count, 3 );
-   }
-
    coord = (GLuint (*)[4])VB->ObjPtr->data;
    coord_stride = VB->ObjPtr->stride;
 
    if (DO_TEX2) {
-      const GLuint t2 = GET_TEXSOURCE(2);
-      tc2 = (GLuint (*)[4])VB->TexCoordPtr[t2]->data;
-      tc2_stride = VB->TexCoordPtr[t2]->stride;
-      if (DO_PTEX && VB->TexCoordPtr[t2]->size < 4) {
-	 if (VB->TexCoordPtr[t2]->flags & VEC_NOT_WRITEABLE) {
-	    VB->import_data( ctx, VERT_BIT_TEX2, VEC_NOT_WRITEABLE );
+      if (VB->TexCoordPtr[2]) {
+	 const GLuint t2 = GET_TEXSOURCE(2);
+	 tc2 = (GLuint (*)[4])VB->TexCoordPtr[t2]->data;
+	 tc2_stride = VB->TexCoordPtr[t2]->stride;
+	 if (DO_PTEX && VB->TexCoordPtr[t2]->size < 4) {
+	    fill_tex |= (1<<2);
 	 }
-	 _mesa_vector4f_clean_elem( VB->TexCoordPtr[t2], VB->Count, 3 );
+      } else {
+	 tc2 = (GLuint (*)[4])&ctx->Current.Attrib[VERT_ATTRIB_TEX2];
+	 tc2_stride = 0;
       }
    }
 
@@ -95,13 +80,10 @@ static void TAG(emit)( GLcontext *ctx,
 	 tc1 = (GLuint (*)[4])VB->TexCoordPtr[t1]->data;
 	 tc1_stride = VB->TexCoordPtr[t1]->stride;
 	 if (DO_PTEX && VB->TexCoordPtr[t1]->size < 4) {
-	    if (VB->TexCoordPtr[t1]->flags & VEC_NOT_WRITEABLE) {
-	       VB->import_data( ctx, VERT_BIT_TEX1, VEC_NOT_WRITEABLE );
-	    }
-	    _mesa_vector4f_clean_elem( VB->TexCoordPtr[t1], VB->Count, 3 );
+	    fill_tex |= (1<<1);
 	 }
       } else {
-	 tc1 = (GLuint (*)[4])&ctx->Current.Attrib[VERT_ATTRIB_TEX1]; /* could be anything, really */
+	 tc1 = (GLuint (*)[4])&ctx->Current.Attrib[VERT_ATTRIB_TEX1];
 	 tc1_stride = 0;
       }
    }
@@ -112,13 +94,10 @@ static void TAG(emit)( GLcontext *ctx,
 	 tc0_stride = VB->TexCoordPtr[t0]->stride;
 	 tc0 = (GLuint (*)[4])VB->TexCoordPtr[t0]->data;
 	 if (DO_PTEX && VB->TexCoordPtr[t0]->size < 4) {
-	    if (VB->TexCoordPtr[t0]->flags & VEC_NOT_WRITEABLE) {
-	       VB->import_data( ctx, VERT_BIT_TEX0, VEC_NOT_WRITEABLE );
-	    }
-	    _mesa_vector4f_clean_elem( VB->TexCoordPtr[t0], VB->Count, 3 );
+	    fill_tex |= (1<<0);
 	 }
       } else {
-	 tc0 = (GLuint (*)[4])&ctx->Current.Attrib[VERT_ATTRIB_TEX0]; /* could be anything, really */
+	 tc0 = (GLuint (*)[4])&ctx->Current.Attrib[VERT_ATTRIB_TEX0];
 	 tc0_stride = 0;
       }
 	 
@@ -136,28 +115,20 @@ static void TAG(emit)( GLcontext *ctx,
 
    if (DO_RGBA) {
       if (VB->ColorPtr[0]) {
-	 /* This is incorrect when colormaterial is enabled:
-	  */
-	 if (VB->ColorPtr[0]->Type != GL_UNSIGNED_BYTE) {
-	    if (0) fprintf(stderr, "IMPORTING FLOAT COLORS\n");
-	    IMPORT_FLOAT_COLORS( ctx );
-	 }
-	 col = (GLubyte (*)[4])VB->ColorPtr[0]->Ptr;
-	 col_stride = VB->ColorPtr[0]->StrideB;
+	 col = VB->ColorPtr[0]->data;
+	 col_stride = VB->ColorPtr[0]->stride;
       } else {
-	 col = &dummy; /* any old memory is fine */
+	 col = (GLfloat (*)[4])ctx->Current.Attrib[VERT_ATTRIB_COLOR0];
 	 col_stride = 0;
       }
    }
 
    if (DO_SPEC) {
       if (VB->SecondaryColorPtr[0]) {
-	 if (VB->SecondaryColorPtr[0]->Type != GL_UNSIGNED_BYTE)
-	    IMPORT_FLOAT_SPEC_COLORS( ctx );
-	 spec = (GLubyte (*)[4])VB->SecondaryColorPtr[0]->Ptr;
-	 spec_stride = VB->SecondaryColorPtr[0]->StrideB;
+	 spec = VB->SecondaryColorPtr[0]->data;
+	 spec_stride = VB->SecondaryColorPtr[0]->stride;
       } else {
-	 spec = &dummy;
+	 spec = (GLfloat (*)[4])ctx->Current.Attrib[VERT_ATTRIB_COLOR1];
 	 spec_stride = 0;
       }
    }
@@ -173,33 +144,33 @@ static void TAG(emit)( GLcontext *ctx,
    }
    
    
-   if (VB->importable_data) {
-      if (start) {
-	 coord =  (GLuint (*)[4])((GLubyte *)coord + start * coord_stride);
-	 if (DO_TEX0)
-	    tc0 =  (GLuint (*)[4])((GLubyte *)tc0 + start * tc0_stride);
-	 if (DO_TEX1) 
-	    tc1 =  (GLuint (*)[4])((GLubyte *)tc1 + start * tc1_stride);
-	 if (DO_TEX2) 
-	    tc2 =  (GLuint (*)[4])((GLubyte *)tc2 + start * tc2_stride);
-	 if (DO_NORM) 
-	    norm =  (GLuint (*)[4])((GLubyte *)norm + start * norm_stride);
-	 if (DO_RGBA) 
-	    STRIDE_4UB(col, start * col_stride);
-	 if (DO_SPEC)
-	    STRIDE_4UB(spec, start * spec_stride);
-	 if (DO_FOG)
-	    fog =  (GLfloat (*)[4])((GLubyte *)fog + start * fog_stride);
-      }
+   if (start) {
+      coord =  (GLuint (*)[4])((GLubyte *)coord + start * coord_stride);
+      if (DO_TEX0)
+	 tc0 =  (GLuint (*)[4])((GLubyte *)tc0 + start * tc0_stride);
+      if (DO_TEX1) 
+	 tc1 =  (GLuint (*)[4])((GLubyte *)tc1 + start * tc1_stride);
+      if (DO_TEX2) 
+	 tc2 =  (GLuint (*)[4])((GLubyte *)tc2 + start * tc2_stride);
+      if (DO_NORM) 
+	 norm =  (GLuint (*)[4])((GLubyte *)norm + start * norm_stride);
+      if (DO_RGBA) 
+	 STRIDE_4F(col, start * col_stride);
+      if (DO_SPEC)
+	 STRIDE_4F(spec, start * spec_stride);
+      if (DO_FOG)
+	 STRIDE_4F(fog, start * fog_stride);
+   }
 
+
+   {
       for (i=start; i < end; i++) {
+	 
 	 v[0].ui = coord[0][0];
 	 v[1].ui = coord[0][1];
 	 v[2].ui = coord[0][2];
-	 if (TCL_DEBUG) fprintf(stderr, "%d: %.2f %.2f %.2f ", i, v[0].f, v[1].f, v[2].f);
 	 if (DO_W) {
 	    v[3].ui = coord[0][3];
-	    if (TCL_DEBUG) fprintf(stderr, "%.2f ", v[3].f);
 	    v += 4;
 	 } 
 	 else
@@ -210,26 +181,27 @@ static void TAG(emit)( GLcontext *ctx,
 	    v[0].ui = norm[0][0];
 	    v[1].ui = norm[0][1];
 	    v[2].ui = norm[0][2];
-	    if (TCL_DEBUG) fprintf(stderr, "norm: %.2f %.2f %.2f ", v[0].f, v[1].f, v[2].f);
 	    v += 3;
 	    norm =  (GLuint (*)[4])((GLubyte *)norm +  norm_stride);
 	 }
 	 if (DO_RGBA) {
-	    v[0].ui = LE32_TO_CPU(*(GLuint *)&col[0]);
-	    STRIDE_4UB(col, col_stride);
-	    if (TCL_DEBUG) fprintf(stderr, "%x ", v[0].ui);
+	    UNCLAMPED_FLOAT_TO_UBYTE(v[0].rgba.red, col[0][0]);
+	    UNCLAMPED_FLOAT_TO_UBYTE(v[0].rgba.green, col[0][1]);
+	    UNCLAMPED_FLOAT_TO_UBYTE(v[0].rgba.blue, col[0][2]);
+	    UNCLAMPED_FLOAT_TO_UBYTE(v[0].rgba.alpha, col[0][3]);
+	    STRIDE_4F(col, col_stride);
 	    v++;
 	 }
 	 if (DO_SPEC || DO_FOG) {
 	    if (DO_SPEC) {
-	       v[0].specular.red   = spec[0][0];
-	       v[0].specular.green = spec[0][1];
-	       v[0].specular.blue  = spec[0][2];
-	       STRIDE_4UB(spec, spec_stride);
+	       UNCLAMPED_FLOAT_TO_UBYTE(v[0].rgba.red, spec[0][0]);
+	       UNCLAMPED_FLOAT_TO_UBYTE(v[0].rgba.green, spec[0][1]);
+	       UNCLAMPED_FLOAT_TO_UBYTE(v[0].rgba.blue, spec[0][2]);
+	       STRIDE_4F(spec, spec_stride);
 	    }
 	    if (DO_FOG) {
-	       v[0].specular.alpha = fog[0][0] * 255.0;
-               fog = (GLfloat (*)[4])((GLubyte *)fog + fog_stride);
+	       UNCLAMPED_FLOAT_TO_UBYTE(v[0].rgba.alpha, fog[0][0]);
+	       fog = (GLfloat (*)[4])((GLubyte *)fog + fog_stride);
 	    }
 	    if (TCL_DEBUG) fprintf(stderr, "%x ", v[0].ui);
 	    v++;
@@ -239,7 +211,10 @@ static void TAG(emit)( GLcontext *ctx,
 	    v[1].ui = tc0[0][1];
 	    if (TCL_DEBUG) fprintf(stderr, "t0: %.2f %.2f ", v[0].f, v[1].f);
 	    if (DO_PTEX) {
-	       v[2].ui = tc0[0][3];
+	       if (fill_tex & (1<<0))
+		  v[2].f = 1.0;
+	       else
+		  v[2].ui = tc0[0][3];
 	       if (TCL_DEBUG) fprintf(stderr, "%.2f ", v[2].f);
 	       v += 3;
 	    } 
@@ -252,7 +227,10 @@ static void TAG(emit)( GLcontext *ctx,
 	    v[1].ui = tc1[0][1];
 	    if (TCL_DEBUG) fprintf(stderr, "t1: %.2f %.2f ", v[0].f, v[1].f);
 	    if (DO_PTEX) {
-	       v[2].ui = tc1[0][3];
+	       if (fill_tex & (1<<1))
+		  v[2].f = 1.0;
+	       else
+		  v[2].ui = tc1[0][3];
 	       if (TCL_DEBUG) fprintf(stderr, "%.2f ", v[2].f);
 	       v += 3;
 	    } 
@@ -264,7 +242,10 @@ static void TAG(emit)( GLcontext *ctx,
 	    v[0].ui = tc2[0][0];
 	    v[1].ui = tc2[0][1];
 	    if (DO_PTEX) {
-	       v[2].ui = tc2[0][3];
+	       if (fill_tex & (1<<2))
+		  v[2].f = 1.0;
+	       else
+		  v[2].ui = tc2[0][3];
 	       v += 3;
 	    } 
 	    else
@@ -273,71 +254,6 @@ static void TAG(emit)( GLcontext *ctx,
 	 } 
 	 if (TCL_DEBUG) fprintf(stderr, "\n");
       }
-   } else {
-      for (i=start; i < end; i++) {
-	 v[0].ui = coord[i][0];
-	 v[1].ui = coord[i][1];
-	 v[2].ui = coord[i][2];
-	 if (DO_W) {
-	    v[3].ui = coord[i][3];
-	    v += 4;
-	 } 
-	 else
-	    v += 3;
-
-	 if (DO_NORM) {
-	    v[0].ui = norm[i][0];
-	    v[1].ui = norm[i][1];
-	    v[2].ui = norm[i][2];
-	    v += 3;
-	 }
-	 if (DO_RGBA) {
-	    v[0].ui = LE32_TO_CPU(*(GLuint *)&col[i]);
-	    v++;
-	 }
-	 if (DO_SPEC || DO_FOG) {
-	    if (DO_SPEC) {
-	       v[0].specular.red   = spec[i][0];
-	       v[0].specular.green = spec[i][1];
-	       v[0].specular.blue  = spec[i][2];
-	    }
-	    if (DO_FOG) {
-               GLfloat *f = (GLfloat *) ((GLubyte *)fog + fog_stride);
-               v[0].specular.alpha = *f * 255.0;
-	    }
-	    v++;
-	 }
-	 if (DO_TEX0) {
-	    v[0].ui = tc0[i][0];
-	    v[1].ui = tc0[i][1];
-	    if (DO_PTEX) {
-	       v[2].ui = tc0[i][3];
-	       v += 3;
-	    } 
-	    else
-	       v += 2;
-	 }
-	 if (DO_TEX1) {
-	    v[0].ui = tc1[i][0];
-	    v[1].ui = tc1[i][1];
-	    if (DO_PTEX) {
-	       v[2].ui = tc1[i][3];
-	       v += 3;
-	    } 
-	    else
-	       v += 2;
-	 } 
-	 if (DO_TEX2) {
-	    v[0].ui = tc2[i][0];
-	    v[1].ui = tc2[i][1];
-	    if (DO_PTEX) {
-	       v[2].ui = tc2[i][3];
-	       v += 3;
-	    } 
-	    else
-	       v += 2;
-	 } 
-      }
    }
 }
 
diff --git a/src/mesa/drivers/dri/radeon/radeon_maos_verts.c b/src/mesa/drivers/dri/radeon/radeon_maos_verts.c
index b32fd624fa3..5e63e6e880f 100644
--- a/src/mesa/drivers/dri/radeon/radeon_maos_verts.c
+++ b/src/mesa/drivers/dri/radeon/radeon_maos_verts.c
@@ -40,7 +40,6 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 #include "array_cache/acache.h"
 #include "tnl/tnl.h"
 #include "tnl/t_pipeline.h"
-#include "tnl/t_imm_debug.h"
 
 #include "radeon_context.h"
 #include "radeon_state.h"
@@ -53,7 +52,7 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
 #define RADEON_TCL_MAX_SETUP 13
 
-union emit_union { float f; GLuint ui; radeon_color_t specular; };
+union emit_union { float f; GLuint ui; radeon_color_t rgba; };
 
 static struct {
    void   (*emit)( GLcontext *, GLuint, GLuint, void * );
@@ -308,6 +307,41 @@ void radeonEmitArrays( GLcontext *ctx, GLuint inputs )
 			      setup_tab[i].vertex_size * 4, 
 			      4);
 
+   /* The vertex code expects Obj to be clean to element 3.  To fix
+    * this, add more vertex code (for obj-2, obj-3) or preferably move
+    * to maos.  
+    */
+   if (VB->ObjPtr->size < 3 || 
+       (VB->ObjPtr->size == 3 && 
+	(setup_tab[i].vertex_format & RADEON_CP_VC_FRMT_W0))) {
+
+      _math_trans_4f( rmesa->tcl.ObjClean.data,
+		      VB->ObjPtr->data,
+		      VB->ObjPtr->stride,
+		      GL_FLOAT,
+		      VB->ObjPtr->size,
+		      0,
+		      VB->Count );
+
+      switch (VB->ObjPtr->size) {
+      case 1:
+	    _mesa_vector4f_clean_elem(&rmesa->tcl.ObjClean, VB->Count, 1);
+      case 2:
+	    _mesa_vector4f_clean_elem(&rmesa->tcl.ObjClean, VB->Count, 2);
+      case 3:
+	 if (setup_tab[i].vertex_format & RADEON_CP_VC_FRMT_W0) {
+	    _mesa_vector4f_clean_elem(&rmesa->tcl.ObjClean, VB->Count, 3);
+	 }
+      case 4:
+      default:
+	 break;
+      }
+
+      VB->ObjPtr = &rmesa->tcl.ObjClean;
+   }
+
+
+
    setup_tab[i].emit( ctx, 0, VB->Count, 
 		      rmesa->tcl.indexed_verts.address + 
 		      rmesa->tcl.indexed_verts.start );
diff --git a/src/mesa/drivers/dri/radeon/radeon_state.c b/src/mesa/drivers/dri/radeon/radeon_state.c
index 7b1bbe75fd2..c15f3417bb9 100644
--- a/src/mesa/drivers/dri/radeon/radeon_state.c
+++ b/src/mesa/drivers/dri/radeon/radeon_state.c
@@ -2115,10 +2115,30 @@ static void radeonInvalidateState( GLcontext *ctx, GLuint new_state )
    radeonVtxfmtInvalidate( ctx );
 }
 
+
+/* A hack.  Need a faster way to find this out.
+ */
+static GLboolean check_material( GLcontext *ctx )
+{
+   TNLcontext *tnl = TNL_CONTEXT(ctx);
+   GLint i;
+
+   for (i = _TNL_ATTRIB_MAT_FRONT_AMBIENT; 
+	i < _TNL_ATTRIB_MAT_BACK_INDEXES; 
+	i++)
+      if (tnl->vb.AttribPtr[i] &&
+	  tnl->vb.AttribPtr[i]->stride)
+	 return GL_TRUE;
+
+   return GL_FALSE;
+}
+      
+
 static void radeonWrapRunPipeline( GLcontext *ctx )
 {
    radeonContextPtr rmesa = RADEON_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
+   GLboolean has_material;
 
    if (0)
       fprintf(stderr, "%s, newstate: %x\n", __FUNCTION__, rmesa->NewGLState);
@@ -2128,7 +2148,9 @@ static void radeonWrapRunPipeline( GLcontext *ctx )
    if (rmesa->NewGLState)
       radeonValidateState( ctx );
 
-   if (tnl->vb.Material) {
+   has_material = (ctx->Light.Enabled && check_material( ctx ));
+
+   if (has_material) {
       TCL_FALLBACK( ctx, RADEON_TCL_FALLBACK_MATERIAL, GL_TRUE );
    }
 
@@ -2136,7 +2158,7 @@ static void radeonWrapRunPipeline( GLcontext *ctx )
     */ 
    _tnl_run_pipeline( ctx );
 
-   if (tnl->vb.Material) {
+   if (has_material) {
       TCL_FALLBACK( ctx, RADEON_TCL_FALLBACK_MATERIAL, GL_FALSE );
       radeonUpdateMaterial( ctx ); /* not needed any more? */
    }
diff --git a/src/mesa/drivers/dri/radeon/radeon_swtcl.c b/src/mesa/drivers/dri/radeon/radeon_swtcl.c
index 89db677392d..157a863e328 100644
--- a/src/mesa/drivers/dri/radeon/radeon_swtcl.c
+++ b/src/mesa/drivers/dri/radeon/radeon_swtcl.c
@@ -44,8 +44,8 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 #include "math/m_translate.h"
 #include "tnl/tnl.h"
 #include "tnl/t_context.h"
-#include "tnl/t_imm_exec.h"
 #include "tnl/t_pipeline.h"
+#include "tnl/t_vtx_api.h"	/* for _tnl_FlushVertices */
 
 #include "radeon_context.h"
 #include "radeon_ioctl.h"
@@ -521,7 +521,7 @@ static void VERT_FALLBACK( GLcontext *ctx,
    tnl->Driver.Render.PrimitiveNotify( ctx, flags & PRIM_MODE_MASK );
    tnl->Driver.Render.BuildVertices( ctx, start, count, ~0 );
    tnl->Driver.Render.PrimTabVerts[flags&PRIM_MODE_MASK]( ctx, start, count, flags );
-   RADEON_CONTEXT(ctx)->swtcl.SetupNewInputs = VERT_BIT_CLIP;
+   RADEON_CONTEXT(ctx)->swtcl.SetupNewInputs = VERT_BIT_POS;
 }
 
 static void ELT_FALLBACK( GLcontext *ctx,
@@ -533,7 +533,7 @@ static void ELT_FALLBACK( GLcontext *ctx,
    tnl->Driver.Render.PrimitiveNotify( ctx, flags & PRIM_MODE_MASK );
    tnl->Driver.Render.BuildVertices( ctx, start, count, ~0 );
    tnl->Driver.Render.PrimTabElts[flags&PRIM_MODE_MASK]( ctx, start, count, flags );
-   RADEON_CONTEXT(ctx)->swtcl.SetupNewInputs = VERT_BIT_CLIP;
+   RADEON_CONTEXT(ctx)->swtcl.SetupNewInputs = VERT_BIT_POS;
 }
 
 
@@ -621,7 +621,7 @@ do {									\
 
 
 static GLboolean radeon_run_render( GLcontext *ctx,
-				    struct gl_pipeline_stage *stage )
+				    struct tnl_pipeline_stage *stage )
 {
    radeonContextPtr rmesa = RADEON_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
@@ -646,18 +646,22 @@ static GLboolean radeon_run_render( GLcontext *ctx,
 	    return GL_TRUE;	/* too many vertices */
    }
 
-   for (i = 0 ; !(flags & PRIM_LAST) ; i += length)
+   for (i = 0 ; i < VB->PrimitiveCount ; i++)
    {
-      flags = VB->Primitive[i];
-      length = VB->PrimitiveLength[i];
+      GLuint prim = VB->Primitive[i].mode;
+      GLuint start = VB->Primitive[i].start;
+      GLuint length = VB->Primitive[i].count;
+
+      if (!length)
+	 continue;
 
       if (RADEON_DEBUG & DEBUG_PRIMS)
-	 fprintf(stderr, "radeon_render.c: prim %s %d..%d\n", 
-		 _mesa_lookup_enum_by_nr(flags & PRIM_MODE_MASK), 
-		 i, i+length);
+	 fprintf(stderr, "r200_render.c: prim %s %d..%d\n", 
+		 _mesa_lookup_enum_by_nr(prim & PRIM_MODE_MASK), 
+		 start, start+length);
 
       if (length)
-	 tab[flags & PRIM_MODE_MASK]( ctx, i, i + length, flags );
+	 tab[prim & PRIM_MODE_MASK]( ctx, start, start + length, flags );
    }
 
    tnl->Driver.Render.Finish( ctx );
@@ -668,9 +672,9 @@ static GLboolean radeon_run_render( GLcontext *ctx,
 
 
 static void radeon_check_render( GLcontext *ctx,
-				 struct gl_pipeline_stage *stage )
+				 struct tnl_pipeline_stage *stage )
 {
-   GLuint inputs = VERT_BIT_POS | VERT_BIT_CLIP | VERT_BIT_COLOR0;
+   GLuint inputs = VERT_BIT_POS | VERT_BIT_COLOR0;
 
    if (ctx->RenderMode == GL_RENDER) {
       if (ctx->_TriangleCaps & DD_SEPARATE_SPECULAR)
@@ -690,13 +694,13 @@ static void radeon_check_render( GLcontext *ctx,
 }
 
 
-static void dtr( struct gl_pipeline_stage *stage )
+static void dtr( struct tnl_pipeline_stage *stage )
 {
    (void)stage;
 }
 
 
-const struct gl_pipeline_stage _radeon_render_stage =
+const struct tnl_pipeline_stage _radeon_render_stage =
 {
    "radeon render",
    (_DD_NEW_SEPARATE_SPECULAR |
@@ -729,7 +733,7 @@ struct texrect_stage_data {
 
 
 static GLboolean run_texrect_stage( GLcontext *ctx,
-				    struct gl_pipeline_stage *stage )
+				    struct tnl_pipeline_stage *stage )
 {
    struct texrect_stage_data *store = TEXRECT_STAGE_DATA(stage);
    radeonContextPtr rmesa = RADEON_CONTEXT(ctx);
@@ -771,7 +775,7 @@ static GLboolean run_texrect_stage( GLcontext *ctx,
 /* Called the first time stage->run() is invoked.
  */
 static GLboolean alloc_texrect_data( GLcontext *ctx,
-				     struct gl_pipeline_stage *stage )
+				     struct tnl_pipeline_stage *stage )
 {
    struct vertex_buffer *VB = &TNL_CONTEXT(ctx)->vb;
    struct texrect_stage_data *store;
@@ -793,7 +797,7 @@ static GLboolean alloc_texrect_data( GLcontext *ctx,
 
 
 static void check_texrect( GLcontext *ctx,
-			   struct gl_pipeline_stage *stage )
+			   struct tnl_pipeline_stage *stage )
 {
    GLuint flags = 0;
 
@@ -809,7 +813,7 @@ static void check_texrect( GLcontext *ctx,
 }
 
 
-static void free_texrect_data( struct gl_pipeline_stage *stage )
+static void free_texrect_data( struct tnl_pipeline_stage *stage )
 {
    struct texrect_stage_data *store = TEXRECT_STAGE_DATA(stage);
    GLuint i;
@@ -824,7 +828,7 @@ static void free_texrect_data( struct gl_pipeline_stage *stage )
 }
 
 
-const struct gl_pipeline_stage _radeon_texrect_stage =
+const struct tnl_pipeline_stage _radeon_texrect_stage =
 {
    "radeon texrect stage",			/* name */
    _NEW_TEXTURE,	/* check_state */
@@ -1193,7 +1197,7 @@ void radeonFallback( GLcontext *ctx, GLuint bit, GLboolean mode )
 
 void radeonFlushVertices( GLcontext *ctx, GLuint flags )
 {
-   _tnl_flush_vertices( ctx, flags );
+   _tnl_FlushVertices( ctx, flags );
 
    if (flags & FLUSH_STORED_VERTICES)
       RADEON_NEWPRIM( RADEON_CONTEXT( ctx ) );
@@ -1242,13 +1246,4 @@ void radeonDestroySwtcl( GLcontext *ctx )
       rmesa->swtcl.verts = 0;
    }
 
-   if (rmesa->UbyteSecondaryColor.Ptr) {
-      ALIGN_FREE(rmesa->UbyteSecondaryColor.Ptr);
-      rmesa->UbyteSecondaryColor.Ptr = 0;
-   }
-
-   if (rmesa->UbyteColor.Ptr) {
-      ALIGN_FREE(rmesa->UbyteColor.Ptr);
-      rmesa->UbyteColor.Ptr = 0;
-   }
 }
diff --git a/src/mesa/drivers/dri/radeon/radeon_tcl.c b/src/mesa/drivers/dri/radeon/radeon_tcl.c
index 651194a8047..ffa09d15dac 100644
--- a/src/mesa/drivers/dri/radeon/radeon_tcl.c
+++ b/src/mesa/drivers/dri/radeon/radeon_tcl.c
@@ -277,7 +277,7 @@ void radeonTclPrimitive( GLcontext *ctx,
 /* TCL render.
  */
 static GLboolean radeon_run_tcl_render( GLcontext *ctx,
-					struct gl_pipeline_stage *stage )
+					struct tnl_pipeline_stage *stage )
 {
    radeonContextPtr rmesa = RADEON_CONTEXT(ctx);
    TNLcontext *tnl = TNL_CONTEXT(ctx);
@@ -297,24 +297,19 @@ static GLboolean radeon_run_tcl_render( GLcontext *ctx,
 
    rmesa->tcl.Elts = VB->Elts;
 
-   for (i = VB->FirstPrimitive ; !(flags & PRIM_LAST) ; i += length)
+   for (i = 0 ; i < VB->PrimitiveCount ; i++)
    {
-      flags = VB->Primitive[i];
-      length = VB->PrimitiveLength[i];
-
-      if (RADEON_DEBUG & DEBUG_PRIMS)
-	 fprintf(stderr, "%s: prim %s %d..%d\n", 
-		 __FUNCTION__,
-		 _mesa_lookup_enum_by_nr(flags & PRIM_MODE_MASK), 
-		 i, i+length);
+      GLuint prim = VB->Primitive[i].mode;
+      GLuint start = VB->Primitive[i].start;
+      GLuint length = VB->Primitive[i].count;
 
       if (!length)
 	 continue;
 
       if (rmesa->tcl.Elts)
-	 radeonEmitEltPrimitive( ctx, i, i+length, flags );
+	 radeonEmitEltPrimitive( ctx, start, start+length, prim );
       else
-	 radeonEmitPrimitive( ctx, i, i+length, flags );
+	 radeonEmitPrimitive( ctx, start, start+length, prim );
    }
 
    return GL_FALSE;		/* finished the pipe */
@@ -323,7 +318,7 @@ static GLboolean radeon_run_tcl_render( GLcontext *ctx,
 
 
 static void radeon_check_tcl_render( GLcontext *ctx,
-				     struct gl_pipeline_stage *stage )
+				     struct tnl_pipeline_stage *stage )
 {
    radeonContextPtr rmesa = RADEON_CONTEXT(ctx);
    GLuint inputs = VERT_BIT_POS;
@@ -374,13 +369,13 @@ static void radeon_check_tcl_render( GLcontext *ctx,
 }
 
 static void radeon_init_tcl_render( GLcontext *ctx,
-				    struct gl_pipeline_stage *stage )
+				    struct tnl_pipeline_stage *stage )
 {
    stage->check = radeon_check_tcl_render;
    stage->check( ctx, stage );
 }
 
-static void dtr( struct gl_pipeline_stage *stage )
+static void dtr( struct tnl_pipeline_stage *stage )
 {
    (void)stage;
 }
@@ -388,7 +383,7 @@ static void dtr( struct gl_pipeline_stage *stage )
 
 /* Initial state for tcl stage.  
  */
-const struct gl_pipeline_stage _radeon_tcl_stage =
+const struct tnl_pipeline_stage _radeon_tcl_stage =
 {
    "radeon render",
    (_DD_NEW_SEPARATE_SPECULAR |
diff --git a/src/mesa/drivers/dri/radeon/radeon_vtxfmt.c b/src/mesa/drivers/dri/radeon/radeon_vtxfmt.c
index 9af0942898d..4e27324807f 100644
--- a/src/mesa/drivers/dri/radeon/radeon_vtxfmt.c
+++ b/src/mesa/drivers/dri/radeon/radeon_vtxfmt.c
@@ -973,13 +973,14 @@ void radeonVtxfmtInit( GLcontext *ctx, GLboolean useCodegen )
    vfmt->FogCoordfEXT = _mesa_noop_FogCoordfEXT;
    vfmt->EdgeFlag = _mesa_noop_EdgeFlag;
    vfmt->EdgeFlagv = _mesa_noop_EdgeFlagv;
-   vfmt->Indexi = _mesa_noop_Indexi;
-   vfmt->Indexiv = _mesa_noop_Indexiv;
+   vfmt->Indexf = _mesa_noop_Indexf;
+   vfmt->Indexfv = _mesa_noop_Indexfv;
 
 
    /* Active but unsupported -- fallback if we receive these:
     */
    vfmt->CallList = radeon_fallback_CallList;
+   vfmt->CallLists = radeon_fallback_CallLists;
    vfmt->EvalCoord1f = radeon_fallback_EvalCoord1f;
    vfmt->EvalCoord1fv = radeon_fallback_EvalCoord1fv;
    vfmt->EvalCoord2f = radeon_fallback_EvalCoord2f;
diff --git a/src/mesa/drivers/dri/radeon/radeon_vtxfmt_c.c b/src/mesa/drivers/dri/radeon/radeon_vtxfmt_c.c
index 04cffb77729..0bdb9a8bbd6 100644
--- a/src/mesa/drivers/dri/radeon/radeon_vtxfmt_c.c
+++ b/src/mesa/drivers/dri/radeon/radeon_vtxfmt_c.c
@@ -831,16 +831,10 @@ void radeonVtxfmtInitChoosers( GLvertexformat *vfmt )
 {
    vfmt->Color3f = choose_Color3f;
    vfmt->Color3fv = choose_Color3fv;
-   vfmt->Color3ub = choose_Color3ub;
-   vfmt->Color3ubv = choose_Color3ubv;
    vfmt->Color4f = choose_Color4f;
    vfmt->Color4fv = choose_Color4fv;
-   vfmt->Color4ub = choose_Color4ub;
-   vfmt->Color4ubv = choose_Color4ubv;
    vfmt->SecondaryColor3fEXT = choose_SecondaryColor3fEXT;
    vfmt->SecondaryColor3fvEXT = choose_SecondaryColor3fvEXT;
-   vfmt->SecondaryColor3ubEXT = choose_SecondaryColor3ubEXT;
-   vfmt->SecondaryColor3ubvEXT = choose_SecondaryColor3ubvEXT;
    vfmt->MultiTexCoord1fARB = choose_MultiTexCoord1fARB;
    vfmt->MultiTexCoord1fvARB = choose_MultiTexCoord1fvARB;
    vfmt->MultiTexCoord2fARB = choose_MultiTexCoord2fARB;
@@ -855,6 +849,15 @@ void radeonVtxfmtInitChoosers( GLvertexformat *vfmt )
    vfmt->Vertex2fv = choose_Vertex2fv;
    vfmt->Vertex3f = choose_Vertex3f;
    vfmt->Vertex3fv = choose_Vertex3fv;
+
+#if 0
+   vfmt->Color3ub = choose_Color3ub;
+   vfmt->Color3ubv = choose_Color3ubv;
+   vfmt->Color4ub = choose_Color4ub;
+   vfmt->Color4ubv = choose_Color4ubv;
+   vfmt->SecondaryColor3ubEXT = choose_SecondaryColor3ubEXT;
+   vfmt->SecondaryColor3ubvEXT = choose_SecondaryColor3ubvEXT;
+#endif
 }
 
 
diff --git a/src/mesa/drivers/dri/sis/Makefile.X11 b/src/mesa/drivers/dri/sis/Makefile.X11
index d6cea8fa6d8..8cdf77f1147 100644
--- a/src/mesa/drivers/dri/sis/Makefile.X11
+++ b/src/mesa/drivers/dri/sis/Makefile.X11
@@ -1,4 +1,3 @@
-# $Id: Makefile.X11,v 1.2 2003/10/20 02:17:33 jonsmirl Exp $
 
 # Mesa 3-D graphics library
 # Version:  5.0
diff --git a/src/mesa/drivers/dri/sis/sis_vb.c b/src/mesa/drivers/dri/sis/sis_vb.c
index 6d56451d69e..5c17bd3801a 100644
--- a/src/mesa/drivers/dri/sis/sis_vb.c
+++ b/src/mesa/drivers/dri/sis/sis_vb.c
@@ -361,7 +361,7 @@ void sisBuildVertices( GLcontext *ctx,
    if (!newinputs)
       return;
 
-   if (newinputs & VERT_BIT_CLIP) {
+   if (newinputs & VERT_BIT_POS) {
       setup_tab[smesa->SetupIndex].emit( ctx, start, count, v, stride );
    } else {
       GLuint ind = 0;