17 files changed, 122 insertions, 109 deletions

diff --git a/src/mesa/drivers/dri/r200/Makefile b/src/mesa/drivers/dri/r200/Makefile
index 25ecbef609..42635bf9d9 100644
--- a/src/mesa/drivers/dri/r200/Makefile
+++ b/src/mesa/drivers/dri/r200/Makefile
@@ -18,17 +18,18 @@ CS_SOURCES = radeon_cs_space_drm.c
 endif
 
 RADEON_COMMON_SOURCES = \
-	radeon_texture.c \
+	radeon_bo_legacy.c \
 	radeon_common_context.c \
 	radeon_common.c \
+	radeon_cs_legacy.c \
 	radeon_dma.c \
+	radeon_debug.c \
+	radeon_fbo.c \
 	radeon_lock.c \
-	radeon_bo_legacy.c \
-	radeon_cs_legacy.c \
 	radeon_mipmap_tree.c \
+	radeon_queryobj.c \
 	radeon_span.c \
-	radeon_fbo.c \
-	radeon_queryobj.c
+	radeon_texture.c
 
 
 DRIVER_SOURCES = r200_context.c \
diff --git a/src/mesa/drivers/dri/r200/r200_cmdbuf.c b/src/mesa/drivers/dri/r200/r200_cmdbuf.c
index 1fe68c2b4c..1d1bea6f5f 100644
--- a/src/mesa/drivers/dri/r200/r200_cmdbuf.c
+++ b/src/mesa/drivers/dri/r200/r200_cmdbuf.c
@@ -126,9 +126,9 @@ void r200EmitVbufPrim( r200ContextPtr rmesa,
    
    radeonEmitState(&rmesa->radeon);
    
-   if (R200_DEBUG & (DEBUG_IOCTL|DEBUG_PRIMS))
-      fprintf(stderr, "%s cmd_used/4: %d prim %x nr %d\n", __FUNCTION__,
-	      rmesa->store.cmd_used/4, primitive, vertex_nr);
+   radeon_print(RADEON_RENDER|RADEON_SWRENDER,RADEON_VERBOSE,
+           "%s cmd_used/4: %d prim %x nr %d\n", __FUNCTION__,
+           rmesa->store.cmd_used/4, primitive, vertex_nr);
  
    BEGIN_BATCH(3);
    OUT_BATCH_PACKET3_CLIP(R200_CP_CMD_3D_DRAW_VBUF_2, 0);
@@ -175,8 +175,7 @@ void r200FlushElts(GLcontext *ctx)
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
    int nr, elt_used = rmesa->tcl.elt_used;
 
-   if (R200_DEBUG & (DEBUG_IOCTL|DEBUG_PRIMS))
-     fprintf(stderr, "%s %x %d\n", __FUNCTION__, rmesa->tcl.hw_primitive, elt_used);
+   radeon_print(RADEON_RENDER, RADEON_VERBOSE, "%s %x %d\n", __FUNCTION__, rmesa->tcl.hw_primitive, elt_used);
 
    assert( rmesa->radeon.dma.flush == r200FlushElts );
    rmesa->radeon.dma.flush = NULL;
@@ -193,8 +192,8 @@ void r200FlushElts(GLcontext *ctx)
    if (R200_ELT_BUF_SZ > elt_used)
      radeonReturnDmaRegion(&rmesa->radeon, R200_ELT_BUF_SZ - elt_used);
 
-   if (R200_DEBUG & DEBUG_SYNC) {
-      fprintf(stderr, "%s: Syncing\n", __FUNCTION__);
+   if (radeon_is_debug_enabled(RADEON_SYNC, RADEON_CRITICAL)) {
+      radeon_print(RADEON_SYNC, RADEON_NORMAL, "%s: Syncing\n", __FUNCTION__);
       radeonFinish( rmesa->radeon.glCtx );
    }
 }
@@ -206,8 +205,7 @@ GLushort *r200AllocEltsOpenEnded( r200ContextPtr rmesa,
 {
    GLushort *retval;
 
-   if (R200_DEBUG & DEBUG_IOCTL)
-      fprintf(stderr, "%s %d prim %x\n", __FUNCTION__, min_nr, primitive);
+   radeon_print(RADEON_RENDER, RADEON_VERBOSE, "%s %d prim %x\n", __FUNCTION__, min_nr, primitive);
 
    assert((primitive & R200_VF_PRIM_WALK_IND));
    
@@ -220,10 +218,6 @@ GLushort *r200AllocEltsOpenEnded( r200ContextPtr rmesa,
    radeon_bo_map(rmesa->radeon.tcl.elt_dma_bo, 1);
    retval = rmesa->radeon.tcl.elt_dma_bo->ptr + rmesa->radeon.tcl.elt_dma_offset;
    
-   if (R200_DEBUG & DEBUG_PRIMS)
-      fprintf(stderr, "%s: header prim %x \n",
-	      __FUNCTION__, primitive);
-
    assert(!rmesa->radeon.dma.flush);
    rmesa->radeon.glCtx->Driver.NeedFlush |= FLUSH_STORED_VERTICES;
    rmesa->radeon.dma.flush = r200FlushElts;
@@ -231,7 +225,17 @@ GLushort *r200AllocEltsOpenEnded( r200ContextPtr rmesa,
    return retval;
 }
 
+void r200EmitMaxVtxIndex(r200ContextPtr rmesa, int count)
+{
+   BATCH_LOCALS(&rmesa->radeon);
 
+   if (rmesa->radeon.radeonScreen->kernel_mm) {
+	   BEGIN_BATCH_NO_AUTOSTATE(2);
+	   OUT_BATCH(CP_PACKET0(R200_SE_VF_MAX_VTX_INDX, 0));
+	   OUT_BATCH(count);
+	   END_BATCH();
+   }
+}
 
 void r200EmitVertexAOS( r200ContextPtr rmesa,
 			GLuint vertex_size,
@@ -240,8 +244,7 @@ void r200EmitVertexAOS( r200ContextPtr rmesa,
 {
    BATCH_LOCALS(&rmesa->radeon);
 
-   if (R200_DEBUG & (DEBUG_PRIMS|DEBUG_IOCTL))
-      fprintf(stderr, "%s:  vertex_size 0x%x offset 0x%x \n",
+   radeon_print(RADEON_SWRENDER, RADEON_VERBOSE, "%s:  vertex_size 0x%x offset 0x%x \n",
 	      __FUNCTION__, vertex_size, offset);
 
 
@@ -260,9 +263,9 @@ void r200EmitAOS(r200ContextPtr rmesa, GLuint nr, GLuint offset)
    int sz = 1 + (nr >> 1) * 3 + (nr & 1) * 2;
    int i;
    
-   if (RADEON_DEBUG & DEBUG_VERTS)
-      fprintf(stderr, "%s: nr=%d, ofs=0x%08x\n", __FUNCTION__, nr,
-	      offset);
+   radeon_print(RADEON_RENDER, RADEON_VERBOSE,
+           "%s: nr=%d, ofs=0x%08x\n",
+           __FUNCTION__, nr, offset);
 
    BEGIN_BATCH(sz+2+ (nr*2));
    OUT_BATCH_PACKET3(R200_CP_CMD_3D_LOAD_VBPNTR, sz - 1);
diff --git a/src/mesa/drivers/dri/r200/r200_context.c b/src/mesa/drivers/dri/r200/r200_context.c
index 0898617a3d..3ddb5bf7d6 100644
--- a/src/mesa/drivers/dri/r200/r200_context.c
+++ b/src/mesa/drivers/dri/r200/r200_context.c
@@ -221,26 +221,6 @@ static void r200InitDriverFuncs( struct dd_function_table *functions )
     functions->GetString		= r200GetString;
 }
 
-static const struct dri_debug_control debug_control[] =
-{
-    { "fall",  DEBUG_FALLBACKS },
-    { "tex",   DEBUG_TEXTURE },
-    { "ioctl", DEBUG_IOCTL },
-    { "prim",  DEBUG_PRIMS },
-    { "vert",  DEBUG_VERTS },
-    { "state", DEBUG_STATE },
-    { "code",  DEBUG_CODEGEN },
-    { "vfmt",  DEBUG_VFMT },
-    { "vtxf",  DEBUG_VFMT },
-    { "verb",  DEBUG_VERBOSE },
-    { "dri",   DEBUG_DRI },
-    { "dma",   DEBUG_DMA },
-    { "san",   DEBUG_SANITY },
-    { "sync",  DEBUG_SYNC },
-    { "pix",   DEBUG_PIXEL },
-    { "mem",   DEBUG_MEMORY },
-    { NULL,    0 }
-};
 
 static void r200_get_lock(radeonContextPtr radeon)
 {
@@ -498,13 +478,6 @@ GLboolean r200CreateContext( const __GLcontextModes *glVisual,
    rmesa->prefer_gart_client_texturing = 
       (getenv("R200_GART_CLIENT_TEXTURES") != 0);
 
-#if DO_DEBUG
-   R200_DEBUG  = driParseDebugString( getenv( "R200_DEBUG" ),
-				      debug_control );
-   R200_DEBUG |= driParseDebugString( getenv( "RADEON_DEBUG" ),
-				      debug_control );
-#endif
-
    tcl_mode = driQueryOptioni(&rmesa->radeon.optionCache, "tcl_mode");
    if (driQueryOptionb(&rmesa->radeon.optionCache, "no_rast")) {
       fprintf(stderr, "disabling 3D acceleration\n");
diff --git a/src/mesa/drivers/dri/r200/r200_context.h b/src/mesa/drivers/dri/r200/r200_context.h
index c5dccf0a75..246f98c6dc 100644
--- a/src/mesa/drivers/dri/r200/r200_context.h
+++ b/src/mesa/drivers/dri/r200/r200_context.h
@@ -479,6 +479,11 @@ struct r200_texture_state {
 #define R200_QUERYOBJ_DATA_0 1
 #define R200_QUERYOBJ_CMDSIZE  2
 
+#define STP_CMD_0 0
+#define STP_DATA_0 1
+#define STP_CMD_1 2
+#define STP_STATE_SIZE 35
+
 struct r200_hw_state {
    /* Hardware state, stored as cmdbuf commands:  
     *   -- Need to doublebuffer for
@@ -521,6 +526,7 @@ struct r200_hw_state {
    struct radeon_state_atom atf;
    struct radeon_state_atom spr;
    struct radeon_state_atom ptp;
+   struct radeon_state_atom stp;
 };
 
 struct r200_state {
diff --git a/src/mesa/drivers/dri/r200/r200_ioctl.c b/src/mesa/drivers/dri/r200/r200_ioctl.c
index 4dbda39eb9..b238adb972 100644
--- a/src/mesa/drivers/dri/r200/r200_ioctl.c
+++ b/src/mesa/drivers/dri/r200/r200_ioctl.c
@@ -190,7 +190,7 @@ static void r200Clear( GLcontext *ctx, GLbitfield mask )
    GLuint color_mask = 0;
    GLuint orig_mask = mask;
 
-   if ( R200_DEBUG & DEBUG_IOCTL ) {
+   if ( R200_DEBUG & RADEON_IOCTL ) {
 	   if (rmesa->radeon.sarea)
 	       fprintf( stderr, "r200Clear %x %d\n", mask, rmesa->radeon.sarea->pfCurrentPage);
 	   else
@@ -229,7 +229,7 @@ static void r200Clear( GLcontext *ctx, GLbitfield mask )
    }
 
    if ( mask ) {
-      if (R200_DEBUG & DEBUG_FALLBACKS)
+      if (R200_DEBUG & RADEON_FALLBACKS)
 	 fprintf(stderr, "%s: swrast clear, mask: %x\n", __FUNCTION__, mask);
       _swrast_Clear( ctx, mask );
    }
@@ -274,7 +274,7 @@ void *r200AllocateMemoryMESA(__DRIscreen *screen, GLsizei size,
    drm_radeon_mem_alloc_t alloc;
    int ret;
 
-   if (R200_DEBUG & DEBUG_IOCTL)
+   if (R200_DEBUG & RADEON_IOCTL)
       fprintf(stderr, "%s sz %d %f/%f/%f\n", __FUNCTION__, size, readfreq,
 	      writefreq, priority);
 
@@ -314,7 +314,7 @@ void r200FreeMemoryMESA(__DRIscreen *screen, GLvoid *pointer)
    drm_radeon_mem_free_t memfree;
    int ret;
 
-   if (R200_DEBUG & DEBUG_IOCTL)
+   if (R200_DEBUG & RADEON_IOCTL)
       fprintf(stderr, "%s %p\n", __FUNCTION__, pointer);
 
    if (!ctx || !(rmesa = R200_CONTEXT(ctx)) || !rmesa->radeon.radeonScreen->gartTextures.map) {
@@ -370,7 +370,7 @@ GLboolean r200IsGartMemory( r200ContextPtr rmesa, const GLvoid *pointer,
 		offset >= 0 &&
 		offset + size < rmesa->radeon.radeonScreen->gartTextures.size);
 
-   if (R200_DEBUG & DEBUG_IOCTL)
+   if (R200_DEBUG & RADEON_IOCTL)
       fprintf(stderr, "r200IsGartMemory( %p ) : %d\n", pointer, valid );
 
    return valid;
diff --git a/src/mesa/drivers/dri/r200/r200_ioctl.h b/src/mesa/drivers/dri/r200/r200_ioctl.h
index 9f06d23b38..8d51aefa04 100644
--- a/src/mesa/drivers/dri/r200/r200_ioctl.h
+++ b/src/mesa/drivers/dri/r200/r200_ioctl.h
@@ -44,6 +44,7 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 #include "drm.h"
 #include "radeon_drm.h"
 
+extern void r200EmitMaxVtxIndex(r200ContextPtr rmesa, int count);
 extern void r200EmitVertexAOS( r200ContextPtr rmesa,
 			       GLuint vertex_size,
 			       struct radeon_bo *bo,
diff --git a/src/mesa/drivers/dri/r200/r200_pixel.c b/src/mesa/drivers/dri/r200/r200_pixel.c
index 7201ddad2d..95773871e0 100644
--- a/src/mesa/drivers/dri/r200/r200_pixel.c
+++ b/src/mesa/drivers/dri/r200/r200_pixel.c
@@ -53,14 +53,14 @@ check_color( const GLcontext *ctx, GLenum type, GLenum format,
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
    GLuint cpp = rmesa->radeon.radeonScreen->cpp;
 
-   if (R200_DEBUG & DEBUG_PIXEL)
+   if (R200_DEBUG & RADEON_PIXEL)
       fprintf(stderr, "%s\n", __FUNCTION__);
 
    if (	(pitch & 63) ||
 	ctx->_ImageTransferState ||
 	packing->SwapBytes ||
 	packing->LsbFirst) {
-      if (R200_DEBUG & DEBUG_PIXEL)
+      if (R200_DEBUG & RADEON_PIXEL)
 	 fprintf(stderr, "%s: failed 1\n", __FUNCTION__);
       return GL_FALSE;
    }
@@ -68,12 +68,12 @@ check_color( const GLcontext *ctx, GLenum type, GLenum format,
    if ( type == GL_UNSIGNED_INT_8_8_8_8_REV &&
 	cpp == 4 &&
 	format == GL_BGRA ) {
-      if (R200_DEBUG & DEBUG_PIXEL)
+      if (R200_DEBUG & RADEON_PIXEL)
 	 fprintf(stderr, "%s: passed 2\n", __FUNCTION__);
       return GL_TRUE;
    }
 
-   if (R200_DEBUG & DEBUG_PIXEL)
+   if (R200_DEBUG & RADEON_PIXEL)
       fprintf(stderr, "%s: failed\n", __FUNCTION__);
 
    return GL_FALSE;
@@ -159,14 +159,14 @@ r200TryReadPixels( GLcontext *ctx,
    GLuint cpp = rmesa->radeon.radeonScreen->cpp;
    GLint size = width * height * cpp;
 
-   if (R200_DEBUG & DEBUG_PIXEL)
+   if (R200_DEBUG & RADEON_PIXEL)
       fprintf(stderr, "%s\n", __FUNCTION__);
 
    /* Only accelerate reading to GART buffers.
     */
    if ( !r200IsGartMemory(rmesa, pixels,
 			 pitch * height * rmesa->radeon.radeonScreen->cpp ) ) {
-      if (R200_DEBUG & DEBUG_PIXEL)
+      if (R200_DEBUG & RADEON_PIXEL)
 	 fprintf(stderr, "%s: dest not GART\n", __FUNCTION__);
    }
 
@@ -174,7 +174,7 @@ r200TryReadPixels( GLcontext *ctx,
     * blitter:
     */
    if (!pack->Invert) {
-      if (R200_DEBUG & DEBUG_PIXEL)
+      if (R200_DEBUG & RADEON_PIXEL)
 	 fprintf(stderr, "%s: MESA_PACK_INVERT not set\n", __FUNCTION__);
       return GL_FALSE;
    }
@@ -207,7 +207,7 @@ r200TryReadPixels( GLcontext *ctx,
    if (!clip_pixelrect(ctx, ctx->ReadBuffer, &x, &y, &width, &height,
 		       &size)) {
       UNLOCK_HARDWARE( &rmesa->radeon );
-      if (R200_DEBUG & DEBUG_PIXEL)
+      if (R200_DEBUG & RADEON_PIXEL)
 	 fprintf(stderr, "%s totally clipped -- nothing to do\n",
 		 __FUNCTION__);
       return GL_TRUE;
@@ -232,7 +232,7 @@ r200TryReadPixels( GLcontext *ctx,
       y += dPriv->y;
 
 
-      if (R200_DEBUG & DEBUG_PIXEL)
+      if (R200_DEBUG & RADEON_PIXEL)
 	 fprintf(stderr, "readpixel blit src_pitch %d dst_pitch %d\n",
 		 src_pitch, dst_pitch);
 
@@ -275,7 +275,7 @@ r200ReadPixels( GLcontext *ctx,
 		 const struct gl_pixelstore_attrib *pack,
 		 GLvoid *pixels )
 {
-   if (R200_DEBUG & DEBUG_PIXEL)
+   if (R200_DEBUG & RADEON_PIXEL)
       fprintf(stderr, "%s\n", __FUNCTION__);
 
    if (!r200TryReadPixels( ctx, x, y, width, height, format, type, pack,
@@ -293,7 +293,7 @@ static void do_draw_pix( GLcontext *ctx,
 			 const void *pixels,
 			 GLuint planemask)
 {
-   if (R200_DEBUG & DEBUG_PIXEL)
+   if (R200_DEBUG & RADEON_PIXEL)
       fprintf(stderr, "%s\n", __FUNCTION__);
 
 #if 0
@@ -388,7 +388,7 @@ r200TryDrawPixels( GLcontext *ctx,
    GLuint cpp = rmesa->radeon.radeonScreen->cpp;
    GLint size = height * pitch * cpp;
 
-   if (R200_DEBUG & DEBUG_PIXEL)
+   if (R200_DEBUG & RADEON_PIXEL)
       fprintf(stderr, "%s\n", __FUNCTION__);
 
    /* check that we're drawing to exactly one color buffer */
@@ -414,7 +414,7 @@ r200TryDrawPixels( GLcontext *ctx,
       /* Can't do conversions on GART reads/draws.
        */
       if ( !r200IsGartMemory( rmesa, pixels, size ) ) {
-	 if (R200_DEBUG & DEBUG_PIXEL)
+	 if (R200_DEBUG & RADEON_PIXEL)
 	    fprintf(stderr, "%s: not GART memory\n", __FUNCTION__);
 	 return GL_FALSE;
       }
@@ -457,7 +457,7 @@ r200DrawPixels( GLcontext *ctx,
 		 const struct gl_pixelstore_attrib *unpack,
 		 const GLvoid *pixels )
 {
-   if (R200_DEBUG & DEBUG_PIXEL)
+   if (R200_DEBUG & RADEON_PIXEL)
       fprintf(stderr, "%s\n", __FUNCTION__);
 
    if (!r200TryDrawPixels( ctx, x, y, width, height, format, type,
diff --git a/src/mesa/drivers/dri/r200/r200_sanity.c b/src/mesa/drivers/dri/r200/r200_sanity.c
index 36530c224e..1241a926ba 100644
--- a/src/mesa/drivers/dri/r200/r200_sanity.c
+++ b/src/mesa/drivers/dri/r200/r200_sanity.c
@@ -48,11 +48,11 @@ USE OR OTHER DEALINGS IN THE SOFTWARE.
 #define MORE_VERBOSE 1
 
 #if MORE_VERBOSE
-#define VERBOSE (R200_DEBUG & DEBUG_VERBOSE)
+#define VERBOSE (R200_DEBUG & RADEON_VERBOSE)
 #define NORMAL  (1)
 #else
 #define VERBOSE 0
-#define NORMAL  (R200_DEBUG & DEBUG_VERBOSE)
+#define NORMAL  (R200_DEBUG & RADEON_VERBOSE)
 #endif
 
 
diff --git a/src/mesa/drivers/dri/r200/r200_state.c b/src/mesa/drivers/dri/r200/r200_state.c
index ab1ba1f283..76852e315c 100644
--- a/src/mesa/drivers/dri/r200/r200_state.c
+++ b/src/mesa/drivers/dri/r200/r200_state.c
@@ -1028,7 +1028,7 @@ void r200UpdateMaterial( GLcontext *ctx )
    if (ctx->Light.ColorMaterialEnabled)
       mask &= ~ctx->Light.ColorMaterialBitmask;
 
-   if (R200_DEBUG & DEBUG_STATE)
+   if (R200_DEBUG & RADEON_STATE)
       fprintf(stderr, "%s\n", __FUNCTION__);
 
    if (mask & MAT_BIT_FRONT_EMISSION) {
@@ -1790,7 +1790,7 @@ static void r200Enable( GLcontext *ctx, GLenum cap, GLboolean state )
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
    GLuint p, flag;
 
-   if ( R200_DEBUG & DEBUG_STATE )
+   if ( R200_DEBUG & RADEON_STATE )
       fprintf( stderr, "%s( %s = %s )\n", __FUNCTION__,
 	       _mesa_lookup_enum_by_nr( cap ),
 	       state ? "GL_TRUE" : "GL_FALSE" );
@@ -2174,7 +2174,7 @@ void r200LightingSpaceChange( GLcontext *ctx )
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
    GLboolean tmp;
 
-   if (R200_DEBUG & DEBUG_STATE)
+   if (R200_DEBUG & RADEON_STATE)
       fprintf(stderr, "%s %d BEFORE %x\n", __FUNCTION__, ctx->_NeedEyeCoords,
 	      rmesa->hw.tcl.cmd[TCL_LIGHT_MODEL_CTL_0]);
 
@@ -2190,7 +2190,7 @@ void r200LightingSpaceChange( GLcontext *ctx )
       rmesa->hw.tcl.cmd[TCL_LIGHT_MODEL_CTL_0] &= ~R200_RESCALE_NORMALS;
    }
 
-   if (R200_DEBUG & DEBUG_STATE)
+   if (R200_DEBUG & RADEON_STATE)
       fprintf(stderr, "%s %d AFTER %x\n", __FUNCTION__, ctx->_NeedEyeCoords,
 	      rmesa->hw.tcl.cmd[TCL_LIGHT_MODEL_CTL_0]);
 }
@@ -2233,7 +2233,7 @@ static void update_texturematrix( GLcontext *ctx )
    GLuint compsel = rmesa->hw.vtx.cmd[VTX_TCL_OUTPUT_COMPSEL];
    int unit;
 
-   if (R200_DEBUG & DEBUG_STATE)
+   if (R200_DEBUG & RADEON_STATE)
       fprintf(stderr, "%s before COMPSEL: %x\n", __FUNCTION__,
 	      rmesa->hw.vtx.cmd[VTX_TCL_OUTPUT_COMPSEL]);
 
@@ -2291,7 +2291,7 @@ static GLboolean r200ValidateBuffers(GLcontext *ctx)
    struct radeon_dma_bo *dma_bo;
    int i, ret;
 
-	if (RADEON_DEBUG & DEBUG_IOCTL)
+	if (RADEON_DEBUG & RADEON_IOCTL)
 		fprintf(stderr, "%s\n", __FUNCTION__);
    radeon_cs_space_reset_bos(rmesa->radeon.cmdbuf.cs);
 
@@ -2466,6 +2466,21 @@ static void r200WrapRunPipeline( GLcontext *ctx )
 }
 
 
+static void r200PolygonStipple( GLcontext *ctx, const GLubyte *mask )
+{
+   r200ContextPtr r200 = R200_CONTEXT(ctx);
+   GLint i;
+
+   radeon_firevertices(&r200->radeon);
+
+   R200_STATECHANGE(r200, stp);
+
+   /* Must flip pattern upside down.
+    */
+   for ( i = 31 ; i >= 0; i--) {
+     r200->hw.stp.cmd[3 + i] = ((GLuint *) mask)[i];
+   }
+}
 /* Initialize the driver's state functions.
  */
 void r200InitStateFuncs( struct dd_function_table *functions, GLboolean dri2 )
@@ -2503,7 +2518,7 @@ void r200InitStateFuncs( struct dd_function_table *functions, GLboolean dri2 )
    functions->PolygonMode		= r200PolygonMode;
    functions->PolygonOffset		= r200PolygonOffset;
    if (dri2)
-      functions->PolygonStipple		= radeonPolygonStipple;
+      functions->PolygonStipple		= r200PolygonStipple;
    else
       functions->PolygonStipple		= radeonPolygonStipplePreKMS;
    functions->PointParameterfv		= r200PointParameter;
diff --git a/src/mesa/drivers/dri/r200/r200_state_init.c b/src/mesa/drivers/dri/r200/r200_state_init.c
index 4c484d067e..7697306d88 100644
--- a/src/mesa/drivers/dri/r200/r200_state_init.c
+++ b/src/mesa/drivers/dri/r200/r200_state_init.c
@@ -271,7 +271,7 @@ CHECK( pix_zero, !ctx->ATIFragmentShader._Enabled, 0 )
 CHECK( afs_pass1, (ctx->ATIFragmentShader._Enabled && (ctx->ATIFragmentShader.Current->NumPasses > 1)), 0 )
 CHECK( afs, ctx->ATIFragmentShader._Enabled, 0 )
 CHECK( tex_cube, rmesa->state.texture.unit[atom->idx].unitneeded & TEXTURE_CUBE_BIT, 3 + 3*5 - CUBE_STATE_SIZE )
-CHECK( tex_cube_cs, rmesa->state.texture.unit[atom->idx].unitneeded & TEXTURE_CUBE_BIT, 2 + 2*5 - CUBE_STATE_SIZE )
+CHECK( tex_cube_cs, rmesa->state.texture.unit[atom->idx].unitneeded & TEXTURE_CUBE_BIT, 2 + 4*5 - CUBE_STATE_SIZE )
 TCL_CHECK( tcl_fog, ctx->Fog.Enabled, 0 )
 TCL_CHECK( tcl_fog_add4, ctx->Fog.Enabled, 4 )
 TCL_CHECK( tcl, GL_TRUE, 0 )
@@ -764,7 +764,7 @@ static void cube_emit_cs(GLcontext *ctx, struct radeon_state_atom *atom)
      for (j = 1; j <= 5; j++) {
        OUT_BATCH(CP_PACKET0(R200_PP_CUBIC_OFFSET_F1_0 + (24*i) + (4 * (j-1)), 0));
        OUT_BATCH_RELOC(lvl->faces[j].offset, t->mt->bo, lvl->faces[j].offset,
-			RADEON_GEM_DOMAIN_VRAM, 0, 0);
+			RADEON_GEM_DOMAIN_GTT|RADEON_GEM_DOMAIN_VRAM, 0, 0);
      }
    }
    END_BATCH();
@@ -885,6 +885,10 @@ void r200InitState( r200ContextPtr rmesa )
          }
       }
    }
+   /* polygon stipple is done with irq for non-kms */
+   if (rmesa->radeon.radeonScreen->kernel_mm) {
+       ALLOC_STATE( stp, always, STP_STATE_SIZE, "STP/stp", 0 );
+   }
 
    for (i = 0; i < 6; i++)
       if (rmesa->radeon.radeonScreen->kernel_mm)
@@ -1117,6 +1121,11 @@ void r200InitState( r200ContextPtr rmesa )
    rmesa->hw.sci.cmd[SCI_CMD_2] = CP_PACKET0(R200_RE_WIDTH_HEIGHT, 0);
 
    if (rmesa->radeon.radeonScreen->kernel_mm) {
+
+	rmesa->hw.stp.cmd[STP_CMD_0] = CP_PACKET0(RADEON_RE_STIPPLE_ADDR, 0);
+	rmesa->hw.stp.cmd[STP_DATA_0] = 0;
+	rmesa->hw.stp.cmd[STP_CMD_1] = CP_PACKET0_ONE(RADEON_RE_STIPPLE_DATA, 31);
+
         rmesa->hw.mtl[0].emit = mtl_emit;
         rmesa->hw.mtl[1].emit = mtl_emit;
 
diff --git a/src/mesa/drivers/dri/r200/r200_swtcl.c b/src/mesa/drivers/dri/r200/r200_swtcl.c
index 0e7d24e7e9..240fb45078 100644
--- a/src/mesa/drivers/dri/r200/r200_swtcl.c
+++ b/src/mesa/drivers/dri/r200/r200_swtcl.c
@@ -203,7 +203,7 @@ static void r200SetVertexFormat( GLcontext *ctx )
 
 static void r200_predict_emit_size( r200ContextPtr rmesa )
 {
-   if (RADEON_DEBUG & DEBUG_VERTS)
+   if (RADEON_DEBUG & RADEON_VERTS)
       fprintf(stderr, "%s\n", __func__);
    const int vertex_array_size = 7;
    const int prim_size = 3;
@@ -225,7 +225,7 @@ static void r200_predict_emit_size( r200ContextPtr rmesa )
 static void r200RenderStart( GLcontext *ctx )
 {
    r200SetVertexFormat( ctx );
-   if (RADEON_DEBUG & DEBUG_VERTS)
+   if (RADEON_DEBUG & RADEON_VERTS)
       fprintf(stderr, "%s\n", __func__);
 }
 
@@ -290,7 +290,7 @@ void r200ChooseVertexState( GLcontext *ctx )
 void r200_swtcl_flush(GLcontext *ctx, uint32_t current_offset)
 {
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
-   if (RADEON_DEBUG & DEBUG_VERTS)
+   if (RADEON_DEBUG & RADEON_VERTS)
       fprintf(stderr, "%s\n", __func__);
 
 
@@ -377,7 +377,7 @@ static void* r200_alloc_verts( r200ContextPtr rmesa, GLuint n, GLuint size)
    const char *r200verts = (char *)rmesa->radeon.swtcl.verts;
 #define VERT(x) (radeonVertex *)(r200verts + ((x) * vertsize * sizeof(int)))
 #define VERTEX radeonVertex
-#define DO_DEBUG_VERTS (1 && (R200_DEBUG & DEBUG_VERTS))
+#define DO_DEBUG_VERTS (1 && (R200_DEBUG & RADEON_VERTS))
 
 #undef TAG
 #define TAG(x) r200_##x
@@ -688,7 +688,7 @@ void r200Fallback( GLcontext *ctx, GLuint bit, GLboolean mode )
 	 TCL_FALLBACK( ctx, R200_TCL_FALLBACK_RASTER, GL_TRUE );
 	 _swsetup_Wakeup( ctx );
 	 rmesa->radeon.swtcl.RenderIndex = ~0;
-         if (R200_DEBUG & DEBUG_FALLBACKS) {
+         if (R200_DEBUG & RADEON_FALLBACKS) {
             fprintf(stderr, "R200 begin rasterization fallback: 0x%x %s\n",
                     bit, getFallbackString(bit));
          }
@@ -720,7 +720,7 @@ void r200Fallback( GLcontext *ctx, GLuint bit, GLboolean mode )
 	    r200ChooseVertexState( ctx );
 	    r200ChooseRenderState( ctx );
 	 }
-         if (R200_DEBUG & DEBUG_FALLBACKS) {
+         if (R200_DEBUG & RADEON_FALLBACKS) {
             fprintf(stderr, "R200 end rasterization fallback: 0x%x %s\n",
                     bit, getFallbackString(bit));
          }
diff --git a/src/mesa/drivers/dri/r200/r200_tcl.c b/src/mesa/drivers/dri/r200/r200_tcl.c
index 0f35d4d5b1..c702910ef2 100644
--- a/src/mesa/drivers/dri/r200/r200_tcl.c
+++ b/src/mesa/drivers/dri/r200/r200_tcl.c
@@ -161,6 +161,7 @@ static GLushort *r200AllocElts( r200ContextPtr rmesa, GLuint nr )
       r200EmitAOS( rmesa,
 		   rmesa->radeon.tcl.aos_count, 0 );
 
+      r200EmitMaxVtxIndex(rmesa, rmesa->radeon.tcl.aos[0].count);
       return r200AllocEltsOpenEnded( rmesa, rmesa->tcl.hw_primitive, nr );
    }
 }
@@ -416,8 +417,9 @@ static GLuint r200EnsureEmitSize( GLcontext * ctx , GLubyte* vimap_rev )
     }
   }
 
-  if (RADEON_DEBUG & DEBUG_PRIMS)
-     fprintf(stderr, "%s space %u, aos %d\n", __func__, space_required, AOS_BUFSZ(nr_aos) );
+  radeon_print(RADEON_RENDER,RADEON_VERBOSE,
+      "%s space %u, aos %d\n",
+      __func__, space_required, AOS_BUFSZ(nr_aos) );
   /* flush the buffer in case we need more than is left. */
   if (rcommonEnsureCmdBufSpace(&rmesa->radeon, space_required + state_size, __FUNCTION__))
     return space_required + radeonCountStateEmitSize( &rmesa->radeon );
@@ -453,8 +455,7 @@ static GLboolean r200_run_tcl_render( GLcontext *ctx,
    if (rmesa->radeon.TclFallback)
       return GL_TRUE;	/* fallback to software t&l */
 
-   if (R200_DEBUG & DEBUG_PRIMS)
-      fprintf(stderr, "%s\n", __FUNCTION__);
+   radeon_print(RADEON_RENDER, RADEON_NORMAL, "%s\n", __FUNCTION__);
 
    if (VB->Count == 0)
       return GL_FALSE;
@@ -650,7 +651,7 @@ static void transition_to_hwtnl( GLcontext *ctx )
    rmesa->hw.vte.cmd[VTE_SE_VTE_CNTL] &= ~(R200_VTX_XY_FMT|R200_VTX_Z_FMT);
    rmesa->hw.vte.cmd[VTE_SE_VTE_CNTL] |= R200_VTX_W0_FMT;
 
-   if (R200_DEBUG & DEBUG_FALLBACKS) 
+   if (R200_DEBUG & RADEON_FALLBACKS)
       fprintf(stderr, "R200 end tcl fallback\n");
 }
 
@@ -692,7 +693,7 @@ void r200TclFallback( GLcontext *ctx, GLuint bit, GLboolean mode )
    if (mode) {
       rmesa->radeon.TclFallback |= bit;
       if (oldfallback == 0) {
-	 if (R200_DEBUG & DEBUG_FALLBACKS) 
+	 if (R200_DEBUG & RADEON_FALLBACKS)
 	    fprintf(stderr, "R200 begin tcl fallback %s\n",
 		    getFallbackString( bit ));
 	 transition_to_swtnl( ctx );
@@ -701,7 +702,7 @@ void r200TclFallback( GLcontext *ctx, GLuint bit, GLboolean mode )
    else {
       rmesa->radeon.TclFallback &= ~bit;
       if (oldfallback == bit) {
-	 if (R200_DEBUG & DEBUG_FALLBACKS) 
+	 if (R200_DEBUG & RADEON_FALLBACKS)
 	    fprintf(stderr, "R200 end tcl fallback %s\n",
 		    getFallbackString( bit ));
 	 transition_to_hwtnl( ctx );
diff --git a/src/mesa/drivers/dri/r200/r200_tex.c b/src/mesa/drivers/dri/r200/r200_tex.c
index 9f79157915..36d9e37d87 100644
--- a/src/mesa/drivers/dri/r200/r200_tex.c
+++ b/src/mesa/drivers/dri/r200/r200_tex.c
@@ -287,7 +287,7 @@ static void r200TexEnv( GLcontext *ctx, GLenum target,
    GLuint unit = ctx->Texture.CurrentUnit;
    struct gl_texture_unit *texUnit = &ctx->Texture.Unit[unit];
 
-   if ( R200_DEBUG & DEBUG_STATE ) {
+   if ( R200_DEBUG & RADEON_STATE ) {
       fprintf( stderr, "%s( %s )\n",
 	       __FUNCTION__, _mesa_lookup_enum_by_nr( pname ) );
    }
@@ -359,7 +359,7 @@ static void r200TexParameter( GLcontext *ctx, GLenum target,
 {
    radeonTexObj* t = radeon_tex_obj(texObj);
 
-   if ( R200_DEBUG & (DEBUG_STATE|DEBUG_TEXTURE) ) {
+   if ( R200_DEBUG & (RADEON_STATE|RADEON_TEXTURE) ) {
       fprintf( stderr, "%s( %s )\n", __FUNCTION__,
 	       _mesa_lookup_enum_by_nr( pname ) );
    }
@@ -409,7 +409,7 @@ static void r200DeleteTexture(GLcontext * ctx, struct gl_texture_object *texObj)
    r200ContextPtr rmesa = R200_CONTEXT(ctx);
    radeonTexObj* t = radeon_tex_obj(texObj);
 
-   if (RADEON_DEBUG & (DEBUG_STATE | DEBUG_TEXTURE)) {
+   if (RADEON_DEBUG & (RADEON_STATE | RADEON_TEXTURE)) {
       fprintf(stderr, "%s( %p (target = %s) )\n", __FUNCTION__,
 	      (void *)texObj,
 	      _mesa_lookup_enum_by_nr(texObj->Target));
@@ -470,7 +470,7 @@ static struct gl_texture_object *r200NewTextureObject(GLcontext * ctx,
    radeonTexObj* t = CALLOC_STRUCT(radeon_tex_obj);
 
 
-   if (RADEON_DEBUG & (DEBUG_STATE | DEBUG_TEXTURE)) {
+   if (RADEON_DEBUG & (RADEON_STATE | RADEON_TEXTURE)) {
      fprintf(stderr, "%s( %p (target = %s) )\n", __FUNCTION__,
 	     t, _mesa_lookup_enum_by_nr(target));
    }
diff --git a/src/mesa/drivers/dri/r200/r200_texstate.c b/src/mesa/drivers/dri/r200/r200_texstate.c
index 4e53672aee..c94834752e 100644
--- a/src/mesa/drivers/dri/r200/r200_texstate.c
+++ b/src/mesa/drivers/dri/r200/r200_texstate.c
@@ -321,7 +321,7 @@ static GLboolean r200UpdateTextureEnv( GLcontext *ctx, int unit, int slot, GLuin
    assert( (texUnit->_ReallyEnabled == 0)
 	   || (texUnit->_Current != NULL) );
 
-   if ( R200_DEBUG & DEBUG_TEXTURE ) {
+   if ( R200_DEBUG & RADEON_TEXTURE ) {
       fprintf( stderr, "%s( %p, %d )\n", __FUNCTION__, (void *)ctx, unit );
    }
 
@@ -1084,6 +1084,8 @@ static void disable_tex_obj_state( r200ContextPtr rmesa,
    R200_STATECHANGE( rmesa, vtx );
    rmesa->hw.vtx.cmd[VTX_TCL_OUTPUT_VTXFMT_1] &= ~(7 << (unit * 3));
 
+   R200_STATECHANGE( rmesa, ctx );
+   rmesa->hw.ctx.cmd[CTX_PP_CNTL] &= ~(R200_TEX_0_ENABLE << unit);
    if (rmesa->radeon.TclFallback & (R200_TCL_FALLBACK_TEXGEN_0<<unit)) {
       TCL_FALLBACK( rmesa->radeon.glCtx, (R200_TCL_FALLBACK_TEXGEN_0<<unit), GL_FALSE);
    }
@@ -1276,7 +1278,7 @@ static GLboolean r200_validate_texgen( GLcontext *ctx, GLuint unit )
    }
 
    if (mixed_fallback) {
-      if (R200_DEBUG & DEBUG_FALLBACKS)
+      if (R200_DEBUG & RADEON_FALLBACKS)
 	 fprintf(stderr, "fallback mixed texgen, 0x%x (0x%x 0x%x 0x%x 0x%x)\n",
 		 texUnit->TexGenEnabled, texUnit->GenS.Mode, texUnit->GenT.Mode,
 		 texUnit->GenR.Mode, texUnit->GenQ.Mode);
@@ -1302,7 +1304,7 @@ static GLboolean r200_validate_texgen( GLcontext *ctx, GLuint unit )
                                                   texUnit->GenR.ObjectPlane,
                                                   texUnit->GenQ.ObjectPlane );
       if (needtgenable & (S_BIT | T_BIT)) {
-	 if (R200_DEBUG & DEBUG_FALLBACKS)
+	 if (R200_DEBUG & RADEON_FALLBACKS)
 	 fprintf(stderr, "fallback mixed texgen / obj plane, 0x%x\n",
 		 texUnit->TexGenEnabled);
 	 return GL_FALSE;
@@ -1330,7 +1332,7 @@ static GLboolean r200_validate_texgen( GLcontext *ctx, GLuint unit )
                                                   texUnit->GenR.EyePlane,
                                                   texUnit->GenQ.EyePlane );
       if (needtgenable & (S_BIT | T_BIT)) {
-	 if (R200_DEBUG & DEBUG_FALLBACKS)
+	 if (R200_DEBUG & RADEON_FALLBACKS)
 	 fprintf(stderr, "fallback mixed texgen / eye plane, 0x%x\n",
 		 texUnit->TexGenEnabled);
 	 return GL_FALSE;
@@ -1380,7 +1382,7 @@ static GLboolean r200_validate_texgen( GLcontext *ctx, GLuint unit )
    default:
       /* Unsupported mode, fallback:
        */
-      if (R200_DEBUG & DEBUG_FALLBACKS)
+      if (R200_DEBUG & RADEON_FALLBACKS)
 	 fprintf(stderr, "fallback unsupported texgen, %d\n",
 		 texUnit->GenS.Mode);
       return GL_FALSE;
diff --git a/src/mesa/drivers/dri/r200/r200_vertprog.c b/src/mesa/drivers/dri/r200/r200_vertprog.c
index 620f29b5c6..11405d7cae 100644
--- a/src/mesa/drivers/dri/r200/r200_vertprog.c
+++ b/src/mesa/drivers/dri/r200/r200_vertprog.c
@@ -423,7 +423,7 @@ static GLboolean r200_translate_vertex_program(GLcontext *ctx, struct r200_verte
       ~(VERT_BIT_POS | VERT_BIT_NORMAL | VERT_BIT_COLOR0 | VERT_BIT_COLOR1 |
       VERT_BIT_FOG | VERT_BIT_TEX0 | VERT_BIT_TEX1 | VERT_BIT_TEX2 |
       VERT_BIT_TEX3 | VERT_BIT_TEX4 | VERT_BIT_TEX5)) != 0) {
-      if (R200_DEBUG & DEBUG_FALLBACKS) {
+      if (R200_DEBUG & RADEON_FALLBACKS) {
 	 fprintf(stderr, "can't handle vert prog inputs 0x%x\n",
 	    mesa_vp->Base.InputsRead);
       }
@@ -436,7 +436,7 @@ static GLboolean r200_translate_vertex_program(GLcontext *ctx, struct r200_verte
       (1 << VERT_RESULT_FOGC) | (1 << VERT_RESULT_TEX0) | (1 << VERT_RESULT_TEX1) |
       (1 << VERT_RESULT_TEX2) | (1 << VERT_RESULT_TEX3) | (1 << VERT_RESULT_TEX4) |
       (1 << VERT_RESULT_TEX5) | (1 << VERT_RESULT_PSIZ))) != 0) {
-      if (R200_DEBUG & DEBUG_FALLBACKS) {
+      if (R200_DEBUG & RADEON_FALLBACKS) {
 	 fprintf(stderr, "can't handle vert prog outputs 0x%x\n",
 	    mesa_vp->Base.OutputsWritten);
       }
@@ -551,7 +551,7 @@ static GLboolean r200_translate_vertex_program(GLcontext *ctx, struct r200_verte
       if (mesa_vp->Base.InputsRead & (1 << i)) {
 	 array_count++;
 	 if (array_count > 12) {
-	    if (R200_DEBUG & DEBUG_FALLBACKS) {
+	    if (R200_DEBUG & RADEON_FALLBACKS) {
 	       fprintf(stderr, "more than 12 attribs used in vert prog\n");
 	    }
 	    return GL_FALSE;
@@ -571,13 +571,13 @@ static GLboolean r200_translate_vertex_program(GLcontext *ctx, struct r200_verte
    }
 
    if (!(mesa_vp->Base.OutputsWritten & (1 << VERT_RESULT_HPOS))) {
-      if (R200_DEBUG & DEBUG_FALLBACKS) {
+      if (R200_DEBUG & RADEON_FALLBACKS) {
 	 fprintf(stderr, "can't handle vert prog without position output\n");
       }
       return GL_FALSE;
    }
    if (free_inputs & 1) {
-      if (R200_DEBUG & DEBUG_FALLBACKS) {
+      if (R200_DEBUG & RADEON_FALLBACKS) {
 	 fprintf(stderr, "can't handle vert prog without position input\n");
       }
       return GL_FALSE;
@@ -1070,7 +1070,7 @@ else {
 	    mesa_vp->Base.NumTemporaries + u_temp_used;
       }
       if ((mesa_vp->Base.NumTemporaries + u_temp_used) > R200_VSF_MAX_TEMPS) {
-	 if (R200_DEBUG & DEBUG_FALLBACKS) {
+	 if (R200_DEBUG & RADEON_FALLBACKS) {
 	    fprintf(stderr, "Ran out of temps, num temps %d, us %d\n", mesa_vp->Base.NumTemporaries, u_temp_used);
 	 }
 	 return GL_FALSE;
@@ -1078,7 +1078,7 @@ else {
       u_temp_i = R200_VSF_MAX_TEMPS - 1;
       if(o_inst - vp->instr >= R200_VSF_MAX_INST) {
 	 mesa_vp->Base.NumNativeInstructions = 129;
-	 if (R200_DEBUG & DEBUG_FALLBACKS) {
+	 if (R200_DEBUG & RADEON_FALLBACKS) {
 	    fprintf(stderr, "more than 128 native instructions\n");
 	 }
 	 return GL_FALSE;
diff --git a/src/mesa/drivers/dri/r200/radeon_debug.c b/src/mesa/drivers/dri/r200/radeon_debug.c
new file mode 120000
index 0000000000..c98c2e074c
--- /dev/null
+++ b/src/mesa/drivers/dri/r200/radeon_debug.c
@@ -0,0 +1 @@
+../radeon/radeon_debug.c
\ No newline at end of file
diff --git a/src/mesa/drivers/dri/r200/radeon_debug.h b/src/mesa/drivers/dri/r200/radeon_debug.h
new file mode 120000
index 0000000000..bd8aa28e89
--- /dev/null
+++ b/src/mesa/drivers/dri/r200/radeon_debug.h
@@ -0,0 +1 @@
+../radeon/radeon_debug.h
\ No newline at end of file