Initial commit of cliptest work. More to come shortly.

 - Add debug, benchmark code.
 - Change linux/x86 FAST_MATH code to GCC/x86, and clear FP exceptions
   before exiting the fast math block.
 - Remove divide-by-zero test in x86 cliptest, and set clipped vertices
   to [0,0,0,1] instead of leaving them uninitialized.

10 files changed, 411 insertions, 46 deletions

diff --git a/src/mesa/Makefile.X11 b/src/mesa/Makefile.X11
index 7a7d18a2366..0950925a17c 100644
--- a/src/mesa/Makefile.X11
+++ b/src/mesa/Makefile.X11
@@ -1,4 +1,4 @@
-# $Id: Makefile.X11,v 1.51 2001/03/29 06:46:15 gareth Exp $
+# $Id: Makefile.X11,v 1.52 2001/05/21 16:33:41 gareth Exp $
 
 # Mesa 3-D graphics library
 # Version:  3.5
@@ -109,6 +109,7 @@ CORE_SOURCES = \
 	X86/common_x86.c \
 	X86/3dnow.c \
 	X86/sse.c \
+	math/m_debug_clip.c \
 	math/m_debug_norm.c \
 	math/m_debug_vertex.c \
 	math/m_debug_xform.c \
diff --git a/src/mesa/main/Makefile.X11 b/src/mesa/main/Makefile.X11
index 7a7d18a2366..0950925a17c 100644
--- a/src/mesa/main/Makefile.X11
+++ b/src/mesa/main/Makefile.X11
@@ -1,4 +1,4 @@
-# $Id: Makefile.X11,v 1.51 2001/03/29 06:46:15 gareth Exp $
+# $Id: Makefile.X11,v 1.52 2001/05/21 16:33:41 gareth Exp $
 
 # Mesa 3-D graphics library
 # Version:  3.5
@@ -109,6 +109,7 @@ CORE_SOURCES = \
 	X86/common_x86.c \
 	X86/3dnow.c \
 	X86/sse.c \
+	math/m_debug_clip.c \
 	math/m_debug_norm.c \
 	math/m_debug_vertex.c \
 	math/m_debug_xform.c \
diff --git a/src/mesa/math/m_clip_tmp.h b/src/mesa/math/m_clip_tmp.h
index 5f27d82959a..b4426f5609b 100644
--- a/src/mesa/math/m_clip_tmp.h
+++ b/src/mesa/math/m_clip_tmp.h
@@ -1,4 +1,4 @@
-/* $Id: m_clip_tmp.h,v 1.6 2001/03/12 00:48:41 gareth Exp $ */
+/* $Id: m_clip_tmp.h,v 1.7 2001/05/21 16:33:41 gareth Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -109,8 +109,8 @@ static GLvector4f * _XFORMAPI TAG(cliptest_np_points4)( GLvector4f *clip_vec,
 							GLubyte *andMask )
 {
    const GLuint stride = clip_vec->stride;
-   const GLfloat *from = (GLfloat *)clip_vec->start;
    const GLuint count = clip_vec->count;
+   const GLfloat *from = (GLfloat *)clip_vec->start;
    GLuint c = 0;
    GLubyte tmpAndMask = *andMask;
    GLubyte tmpOrMask = *orMask;
diff --git a/src/mesa/math/m_debug.h b/src/mesa/math/m_debug.h
index 8bb2a1976d3..21a8b044353 100644
--- a/src/mesa/math/m_debug.h
+++ b/src/mesa/math/m_debug.h
@@ -1,4 +1,4 @@
-/* $Id: m_debug.h,v 1.3 2001/03/12 00:48:41 gareth Exp $ */
+/* $Id: m_debug.h,v 1.4 2001/05/21 16:33:41 gareth Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -32,6 +32,10 @@
 
 extern void _math_test_all_transform_functions( char *description );
 extern void _math_test_all_normal_transform_functions( char *description );
+extern void _math_test_all_cliptest_functions( char *description );
+
+/* Deprecated?
+ */
 extern void _math_test_all_vertex_functions( char *description );
 
 extern char *mesa_profile;
diff --git a/src/mesa/math/m_debug_clip.c b/src/mesa/math/m_debug_clip.c
new file mode 100644
index 00000000000..b6557665a65
--- /dev/null
+++ b/src/mesa/math/m_debug_clip.c
@@ -0,0 +1,365 @@
+/* $Id: m_debug_clip.c,v 1.1 2001/05/21 16:33:41 gareth Exp $ */
+
+/*
+ * Mesa 3-D graphics library
+ * Version:  3.5
+ *
+ * Copyright (C) 1999-2001  Brian Paul   All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * BRIAN PAUL BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors:
+ *    Gareth Hughes <[email protected]>
+ */
+
+#include "glheader.h"
+#include "context.h"
+#include "macros.h"
+#include "mem.h"
+
+#include "m_matrix.h"
+#include "m_xform.h"
+
+#include "m_debug.h"
+#include "m_debug_util.h"
+
+#ifdef DEBUG  /* This code only used for debugging */
+
+static clip_func *clip_tab[2] = {
+   _mesa_clip_tab,
+   _mesa_clip_np_tab
+};
+static char *cnames[2] = {
+   "_mesa_clip_tab",
+   "_mesa_clip_np_tab"
+};
+static char *cstrings[2] = {
+   "clip, perspective divide",
+   "clip, no divide"
+};
+
+
+/* =============================================================
+ * Reference cliptests
+ */
+
+static GLvector4f *ref_cliptest_points4( GLvector4f *clip_vec,
+					 GLvector4f *proj_vec,
+					 GLubyte clipMask[],
+					 GLubyte *orMask,
+					 GLubyte *andMask )
+{
+   const GLuint stride = clip_vec->stride;
+   const GLuint count = clip_vec->count;
+   const GLfloat *from = (GLfloat *)clip_vec->start;
+   GLuint c = 0;
+   GLfloat (*vProj)[4] = (GLfloat (*)[4])proj_vec->start;
+   GLubyte tmpAndMask = *andMask;
+   GLubyte tmpOrMask = *orMask;
+   GLuint i;
+   for ( i = 0 ; i < count ; i++, STRIDE_F(from, stride) ) {
+      const GLfloat cx = from[0];
+      const GLfloat cy = from[1];
+      const GLfloat cz = from[2];
+      const GLfloat cw = from[3];
+      GLubyte mask = 0;
+      if ( -cx + cw < 0 ) mask |= CLIP_RIGHT_BIT;
+      if (  cx + cw < 0 ) mask |= CLIP_LEFT_BIT;
+      if ( -cy + cw < 0 ) mask |= CLIP_TOP_BIT;
+      if (  cy + cw < 0 ) mask |= CLIP_BOTTOM_BIT;
+      if ( -cz + cw < 0 ) mask |= CLIP_FAR_BIT;
+      if (  cz + cw < 0 ) mask |= CLIP_NEAR_BIT;
+      clipMask[i] = mask;
+      if ( mask ) {
+	 c++;
+	 tmpAndMask &= mask;
+	 tmpOrMask |= mask;
+	 vProj[i][0] = 0;
+	 vProj[i][1] = 0;
+	 vProj[i][2] = 0;
+	 vProj[i][3] = 1;
+      } else {
+	 GLfloat oow = 1.0F / cw;
+	 vProj[i][0] = cx * oow;
+	 vProj[i][1] = cy * oow;
+	 vProj[i][2] = cz * oow;
+	 vProj[i][3] = oow;
+      }
+   }
+
+   *orMask = tmpOrMask;
+   *andMask = (GLubyte) (c < count ? 0 : tmpAndMask);
+
+   proj_vec->flags |= VEC_SIZE_4;
+   proj_vec->size = 4;
+   proj_vec->count = clip_vec->count;
+   return proj_vec;
+}
+
+/* Keep these here for now, even though we don't use them...
+ */
+static GLvector4f *ref_cliptest_points3( GLvector4f *clip_vec,
+					 GLvector4f *proj_vec,
+					 GLubyte clipMask[],
+					 GLubyte *orMask,
+					 GLubyte *andMask )
+{
+   const GLuint stride = clip_vec->stride;
+   const GLuint count = clip_vec->count;
+   const GLfloat *from = (GLfloat *)clip_vec->start;
+
+   GLubyte tmpOrMask = *orMask;
+   GLubyte tmpAndMask = *andMask;
+   GLuint i;
+   for ( i = 0 ; i < count ; i++, STRIDE_F(from, stride) ) {
+      const GLfloat cx = from[0], cy = from[1], cz = from[2];
+      GLubyte mask = 0;
+      if ( cx >  1.0 )		mask |= CLIP_RIGHT_BIT;
+      else if ( cx < -1.0 )	mask |= CLIP_LEFT_BIT;
+      if ( cy >  1.0 )		mask |= CLIP_TOP_BIT;
+      else if ( cy < -1.0 )	mask |= CLIP_BOTTOM_BIT;
+      if ( cz >  1.0 )		mask |= CLIP_FAR_BIT;
+      else if ( cz < -1.0 )	mask |= CLIP_NEAR_BIT;
+      clipMask[i] = mask;
+      tmpOrMask |= mask;
+      tmpAndMask &= mask;
+   }
+
+   *orMask = tmpOrMask;
+   *andMask = tmpAndMask;
+   return clip_vec;
+}
+
+static GLvector4f * ref_cliptest_points2( GLvector4f *clip_vec,
+					  GLvector4f *proj_vec,
+					  GLubyte clipMask[],
+					  GLubyte *orMask,
+					  GLubyte *andMask )
+{
+   const GLuint stride = clip_vec->stride;
+   const GLuint count = clip_vec->count;
+   const GLfloat *from = (GLfloat *)clip_vec->start;
+
+   GLubyte tmpOrMask = *orMask;
+   GLubyte tmpAndMask = *andMask;
+   GLuint i;
+   for ( i = 0 ; i < count ; i++, STRIDE_F(from, stride) ) {
+      const GLfloat cx = from[0], cy = from[1];
+      GLubyte mask = 0;
+      if ( cx >  1.0 )		mask |= CLIP_RIGHT_BIT;
+      else if ( cx < -1.0 )	mask |= CLIP_LEFT_BIT;
+      if ( cy >  1.0 )		mask |= CLIP_TOP_BIT;
+      else if ( cy < -1.0 )	mask |= CLIP_BOTTOM_BIT;
+      clipMask[i] = mask;
+      tmpOrMask |= mask;
+      tmpAndMask &= mask;
+   }
+
+   *orMask = tmpOrMask;
+   *andMask = tmpAndMask;
+   return clip_vec;
+}
+
+static clip_func ref_cliptest[5] = {
+   0,
+   0,
+   ref_cliptest_points2,
+   ref_cliptest_points3,
+   ref_cliptest_points4
+};
+
+
+/* =============================================================
+ * Cliptest tests
+ */
+
+static GLfloat s[TEST_COUNT][4] ALIGN16;
+static GLfloat d[TEST_COUNT][4] ALIGN16;
+static GLfloat r[TEST_COUNT][4] ALIGN16;
+
+static int test_cliptest_function( clip_func func, int np,
+				   int psize, long *cycles )
+{
+   GLvector4f source[1], dest[1], ref[1];
+   GLubyte dm[TEST_COUNT], dco, dca;
+   GLubyte rm[TEST_COUNT], rco, rca;
+   int i, j;
+#ifdef  RUN_DEBUG_BENCHMARK
+   int cycle_i;                /* the counter for the benchmarks we run */
+#endif
+
+   (void) cycles;
+
+   if ( psize > 4 ) {
+      _mesa_problem( NULL, "test_cliptest_function called with psize > 4\n" );
+      return 0;
+   }
+
+   for ( i = 0 ; i < TEST_COUNT ; i++) {
+      ASSIGN_4V( d[i], 0.0, 0.0, 0.0, 1.0 );
+      ASSIGN_4V( s[i], 0.0, 0.0, 0.0, 1.0 );
+      for ( j = 0 ; j < psize ; j++ )
+         s[i][j] = rnd();
+   }
+
+   source->data = (GLfloat(*)[4])s;
+   source->start = (GLfloat *)s;
+   source->count = TEST_COUNT;
+   source->stride = sizeof(s[0]);
+   source->size = 4;
+   source->flags = 0;
+
+   dest->data = (GLfloat(*)[4])d;
+   dest->start = (GLfloat *)d;
+   dest->count = TEST_COUNT;
+   dest->stride = sizeof(float[4]);
+   dest->size = 0;
+   dest->flags = 0;
+
+   ref->data = (GLfloat(*)[4])r;
+   ref->start = (GLfloat *)r;
+   ref->count = TEST_COUNT;
+   ref->stride = sizeof(float[4]);
+   ref->size = 0;
+   ref->flags = 0;
+
+   dco = rco = 0;
+   dca = rca = CLIP_ALL_BITS;
+
+   ref_cliptest[psize]( source, ref, rm, &rco, &rca );
+
+   if ( mesa_profile ) {
+      BEGIN_RACE( *cycles );
+      func( source, dest, dm, &dco, &dca );
+      END_RACE( *cycles );
+   }
+   else {
+      func( source, dest, dm, &dco, &dca );
+   }
+
+   if ( dco != rco ) {
+      printf( "\n-----------------------------\n" );
+      printf( "dco = 0x%02x   rco = 0x%02x\n", dco, rco );
+      return 0;
+   }
+   if ( dca != rca ) {
+      printf( "\n-----------------------------\n" );
+      printf( "dca = 0x%02x   rca = 0x%02x\n", dca, rca );
+      return 0;
+   }
+   for ( i = 0 ; i < TEST_COUNT ; i++ ) {
+      if ( dm[i] != rm[i] ) {
+	 printf( "\n-----------------------------\n" );
+	 printf( "(i = %i)\n", i );
+	 printf( "dm = 0x%02x   rm = 0x%02x\n", dm[i], rm[i] );
+	 return 0;
+      }
+   }
+
+   /* Only verify output on projected points4 case.  FIXME: Do we need
+    * to test other cases?
+    */
+   if ( np || psize < 4 )
+      return 1;
+
+   for ( i = 0 ; i < TEST_COUNT ; i++ ) {
+      for ( j = 0 ; j < 4 ; j++ ) {
+         if ( significand_match( d[i][j], r[i][j] ) < REQUIRED_PRECISION ) {
+            printf( "\n-----------------------------\n" );
+            printf( "(i = %i, j = %i)  dm = 0x%02x   rm = 0x%02x\n",
+		    i, j, dm[i], rm[i] );
+            printf( "%f \t %f \t [diff = %e - %i bit missed]\n",
+		    d[i][0], r[i][0], r[i][0]-d[i][0],
+		    MAX_PRECISION - significand_match( d[i][0], r[i][0] ) );
+            printf( "%f \t %f \t [diff = %e - %i bit missed]\n",
+		    d[i][1], r[i][1], r[i][1]-d[i][1],
+		    MAX_PRECISION - significand_match( d[i][1], r[i][1] ) );
+            printf( "%f \t %f \t [diff = %e - %i bit missed]\n",
+		    d[i][2], r[i][2], r[i][2]-d[i][2],
+		    MAX_PRECISION - significand_match( d[i][2], r[i][2] ) );
+            printf( "%f \t %f \t [diff = %e - %i bit missed]\n",
+		    d[i][3], r[i][3], r[i][3]-d[i][3],
+		    MAX_PRECISION - significand_match( d[i][3], r[i][3] ) );
+            return 0;
+         }
+      }
+   }
+
+   return 1;
+}
+
+void _math_test_all_cliptest_functions( char *description )
+{
+   int np, psize;
+   long benchmark_tab[2][4];
+   static int first_time = 1;
+
+   if ( first_time ) {
+      first_time = 0;
+      mesa_profile = getenv( "MESA_PROFILE" );
+   }
+
+#ifdef RUN_DEBUG_BENCHMARK
+   if ( mesa_profile ) {
+      if ( !counter_overhead ) {
+	 INIT_COUNTER();
+	 printf( "counter overhead: %ld cycles\n\n", counter_overhead );
+      }
+      printf( "cliptest results after hooking in %s functions:\n", description );
+   }
+#endif
+
+#ifdef RUN_DEBUG_BENCHMARK
+   if ( mesa_profile ) {
+      printf( "\n\t" );
+      for ( psize = 2 ; psize <= 4 ; psize++ ) {
+	 printf( " p%d\t", psize );
+      }
+      printf( "\n--------------------------------------------------------\n\t" );
+   }
+#endif
+
+   for ( np = 0 ; np < 2 ; np++ ) {
+      for ( psize = 2 ; psize <= 4 ; psize++ ) {
+	 clip_func func = clip_tab[np][psize];
+	 long *cycles = &(benchmark_tab[np][psize-1]);
+
+	 if ( test_cliptest_function( func, np, psize, cycles ) == 0 ) {
+	    char buf[100];
+	    sprintf( buf, "%s[%d] failed test (%s)",
+		     cnames[np], psize, description );
+	    _mesa_problem( NULL, buf );
+	 }
+#ifdef RUN_DEBUG_BENCHMARK
+	 if ( mesa_profile )
+	    printf( " %li\t", benchmark_tab[np][psize-1] );
+#endif
+      }
+#ifdef RUN_DEBUG_BENCHMARK
+      if ( mesa_profile )
+	 printf( " | [%s]\n\t", cstrings[np] );
+#endif
+   }
+#ifdef RUN_DEBUG_BENCHMARK
+   if ( mesa_profile )
+      printf( "\n" );
+#endif
+}
+
+
+#endif /* DEBUG */
diff --git a/src/mesa/math/m_xform.c b/src/mesa/math/m_xform.c
index 6c938ea38b8..69430848be2 100644
--- a/src/mesa/math/m_xform.c
+++ b/src/mesa/math/m_xform.c
@@ -1,4 +1,4 @@
-/* $Id: m_xform.c,v 1.12 2001/03/30 14:44:43 gareth Exp $ */
+/* $Id: m_xform.c,v 1.13 2001/05/21 16:33:41 gareth Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -200,6 +200,7 @@ _math_init_transformation( void )
 #ifdef DEBUG
    _math_test_all_transform_functions( "default" );
    _math_test_all_normal_transform_functions( "default" );
+   _math_test_all_cliptest_functions( "default" );
 #endif
 
 #ifdef USE_X86_ASM
diff --git a/src/mesa/tnl/t_pipeline.c b/src/mesa/tnl/t_pipeline.c
index bc2b6682253..acca9d85b6a 100644
--- a/src/mesa/tnl/t_pipeline.c
+++ b/src/mesa/tnl/t_pipeline.c
@@ -1,4 +1,4 @@
-/* $Id: t_pipeline.c,v 1.18 2001/05/10 12:18:38 keithw Exp $ */
+/* $Id: t_pipeline.c,v 1.19 2001/05/21 16:33:41 gareth Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -134,18 +134,18 @@ void _tnl_run_pipeline( GLcontext *ctx )
    /* Done elsewhere.
     */
    ASSERT(pipe->build_state_changes == 0);
-   
+
    START_FAST_MATH(__tmp);
 
    /* If something changes in the pipeline, tag all subsequent stages
     * using this value for recalculation.  Inactive stages have their
     * state and inputs examined to try to keep cached data alive over
-    * state-changes.  
+    * state-changes.
     */
    for ( ; s->run ; s++) {
       s->changed_inputs |= s->inputs & changed_inputs;
 
-      if (s->run_state & changed_state) 
+      if (s->run_state & changed_state)
 	 s->changed_inputs = s->inputs;
 
       if (s->active && running) {
diff --git a/src/mesa/x86/clip_args.h b/src/mesa/x86/clip_args.h
index eeb10236b3b..9d415ea0149 100644
--- a/src/mesa/x86/clip_args.h
+++ b/src/mesa/x86/clip_args.h
@@ -1,4 +1,4 @@
-/* $Id: clip_args.h,v 1.3 2001/03/28 20:44:43 gareth Exp $ */
+/* $Id: clip_args.h,v 1.4 2001/05/21 16:33:41 gareth Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -38,8 +38,8 @@
 /*
  * Offsets for clip_func arguments
  *
- * typedef GLvector4f *(*clip_func)( GLvector4f *vClip,
- *	                             GLvector4f *vProj,
+ * typedef GLvector4f *(*clip_func)( GLvector4f *clip_vec,
+ *	                             GLvector4f *proj_vec,
  *	                             GLubyte clipMask[],
  *	                             GLubyte *orMask,
  *	                             GLubyte *andMask );
diff --git a/src/mesa/x86/x86.c b/src/mesa/x86/x86.c
index b2b442c87e2..03104d5370e 100644
--- a/src/mesa/x86/x86.c
+++ b/src/mesa/x86/x86.c
@@ -1,4 +1,4 @@
-/* $Id: x86.c,v 1.22 2001/05/13 18:55:15 gareth Exp $ */
+/* $Id: x86.c,v 1.23 2001/05/21 16:33:41 gareth Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -90,6 +90,7 @@ void _mesa_init_x86_transform_asm( void )
 
 #ifdef DEBUG
    _math_test_all_transform_functions( "x86" );
+   _math_test_all_cliptest_functions( "x86" );
 #endif
 #endif
 }
@@ -100,7 +101,7 @@ void _mesa_init_x86_vertex_asm( void )
    _mesa_xform_points3_v16_general = _mesa_v16_x86_general_xform;
    _mesa_cliptest_points4_v16 = _mesa_v16_x86_cliptest_points4;
 
-#ifdef DEBUG
+#ifdef 0 /* DEBUG */
    _math_test_all_vertex_functions( "x86" );
 #endif
 #endif
diff --git a/src/mesa/x86/x86_cliptest.S b/src/mesa/x86/x86_cliptest.S
index 76f34d5d5c2..08e34ecbd6f 100644
--- a/src/mesa/x86/x86_cliptest.S
+++ b/src/mesa/x86/x86_cliptest.S
@@ -1,4 +1,4 @@
-/* $Id: x86_cliptest.S,v 1.7 2001/05/13 18:55:15 gareth Exp $ */
+/* $Id: x86_cliptest.S,v 1.8 2001/05/21 16:33:41 gareth Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -27,9 +27,6 @@
 #include "matypes.h"
 #include "clip_args.h"
 
-#define FP_ONE		1065353216
-#define FP_ZERO		0
-
 #define SRC(i)		REGOFF(i * 4, ESI)
 #define DST(i)		REGOFF(i * 4, EDI)
 #define MAT(i)		REGOFF(i * 4, EDX)
@@ -50,22 +47,22 @@
 	SEG_DATA
 
 clip_table:
-	D_BYTE 0, 1, 0, 2, 4, 5, 4, 6
-	D_BYTE 0, 1, 0, 2, 8, 9, 8, 10
-	D_BYTE 32, 33, 32, 34, 36, 37, 36, 38
-	D_BYTE 32, 33, 32, 34, 40, 41, 40, 42
-	D_BYTE 0, 1, 0, 2, 4, 5, 4, 6
-	D_BYTE 0, 1, 0, 2, 8, 9, 8, 10
-	D_BYTE 16, 17, 16, 18, 20, 21, 20, 22
-	D_BYTE 16, 17, 16, 18, 24, 25, 24, 26
-	D_BYTE 63, 61, 63, 62, 55, 53, 55, 54
-	D_BYTE 63, 61, 63, 62, 59, 57, 59, 58
-	D_BYTE 47, 45, 47, 46, 39, 37, 39, 38
-	D_BYTE 47, 45, 47, 46, 43, 41, 43, 42
-	D_BYTE 63, 61, 63, 62, 55, 53, 55, 54
-	D_BYTE 63, 61, 63, 62, 59, 57, 59, 58
-	D_BYTE 31, 29, 31, 30, 23, 21, 23, 22
-	D_BYTE 31, 29, 31, 30, 27, 25, 27, 26
+	D_BYTE 0x00, 0x01, 0x00, 0x02, 0x04, 0x05, 0x04, 0x06
+	D_BYTE 0x00, 0x01, 0x00, 0x02, 0x08, 0x09, 0x08, 0x0a
+	D_BYTE 0x20, 0x21, 0x20, 0x22, 0x24, 0x25, 0x24, 0x26
+	D_BYTE 0x20, 0x21, 0x20, 0x22, 0x28, 0x29, 0x28, 0x2a
+	D_BYTE 0x00, 0x01, 0x00, 0x02, 0x04, 0x05, 0x04, 0x06
+	D_BYTE 0x00, 0x01, 0x00, 0x02, 0x08, 0x09, 0x08, 0x0a
+	D_BYTE 0x10, 0x11, 0x10, 0x12, 0x14, 0x15, 0x14, 0x16
+	D_BYTE 0x10, 0x11, 0x10, 0x12, 0x18, 0x19, 0x18, 0x1a
+	D_BYTE 0x3f, 0x3d, 0x3f, 0x3e, 0x37, 0x35, 0x37, 0x36
+	D_BYTE 0x3f, 0x3d, 0x3f, 0x3e, 0x3b, 0x39, 0x3b, 0x3a
+	D_BYTE 0x2f, 0x2d, 0x2f, 0x2e, 0x27, 0x25, 0x27, 0x26
+	D_BYTE 0x2f, 0x2d, 0x2f, 0x2e, 0x2b, 0x29, 0x2b, 0x2a
+	D_BYTE 0x3f, 0x3d, 0x3f, 0x3e, 0x37, 0x35, 0x37, 0x36
+	D_BYTE 0x3f, 0x3d, 0x3f, 0x3e, 0x3b, 0x39, 0x3b, 0x3a
+	D_BYTE 0x1f, 0x1d, 0x1f, 0x1e, 0x17, 0x15, 0x17, 0x16
+	D_BYTE 0x1f, 0x1d, 0x1f, 0x1e, 0x1b, 0x19, 0x1b, 0x1a
 
 
 	SEG_TEXT
@@ -149,13 +146,8 @@ LLBL( ctp4_clip_table_ready ):
 ALIGNTEXT16
 LLBL( ctp4_top ):
 
-	MOV_L( SRC(3), ECX )
 	FLD1				/* F3 */
-
-	TEST_L( ECX, ECX )		/* GH: If src[3] is zero, abort */
-	JZ( LLBL( ctp4_noproj ) )
-
-	FDIV_S( SRC(3) )
+	FDIV_S( SRC(3) )		/* GH: don't care about div-by-zero */
 
 	MOV_L( SRC(3), EBP )
 	MOV_L( SRC(2), EBX )
@@ -205,12 +197,12 @@ LLBL( ctp4_top ):
 
 LLBL( ctp4_noproj ):
 
-	FLDZ				/* F0 F3 */
+	FSTP( ST(0) )			/* */
 
-	FST_S( DST(0) )			/* F0 F3 */
-	FST_S( DST(1) )			/* F0 F3 */
-	FSTP_S( DST(2) )		/* F3 */
-	FSTP_S( DST(3) )		/* */
+	MOV_L( CONST(0), DST(0) )
+	MOV_L( CONST(0), DST(1) )
+	MOV_L( CONST(0), DST(2) )
+	MOV_L( CONST(0x3f800000), DST(3) )
 
 	JMP( LLBL( ctp4_next ) )