syntax fixes for Solaris (David Dawes)

6 files changed, 726 insertions, 613 deletions

diff --git a/src/mesa/x86/assyntax.h b/src/mesa/x86/assyntax.h
index 442d21b2a7d..8acc6d031d9 100644
--- a/src/mesa/x86/assyntax.h
+++ b/src/mesa/x86/assyntax.h
@@ -1,4 +1,4 @@
-/* $Id: assyntax.h,v 1.17 2002/01/04 14:35:46 brianp Exp $ */
+/* $Id: assyntax.h,v 1.18 2002/03/07 21:40:08 brianp Exp $ */
 
 #ifndef __ASSYNTAX_H__
 #define __ASSYNTAX_H__
@@ -69,6 +69,11 @@
 
 #if !(defined(NASM_ASSEMBLER) || defined(MASM_ASSEMBLER))
 
+/* Default to ATT_ASSEMBLER when SVR4 or SYSV are defined */
+#if (defined(SVR4) || defined(SYSV)) && !defined(GNU_ASSEMBLER)
+#define ATT_ASSEMBLER
+#endif
+
 #if !defined(ATT_ASSEMBLER) && !defined(GNU_ASSEMBLER) && !defined(ACK_ASSEMBLER)
 #define GNU_ASSEMBLER
 #endif
@@ -200,6 +205,11 @@
 #define _STX6		%st(6)
 #define _STX7		%st(7)
 #define ST(x)		CONCAT(_STX,x)
+#ifdef GNU_ASSEMBLER
+#define ST0		%st(0)
+#else
+#define ST0		%st
+#endif
 /* MMX Registers */
 #define MM0		%mm0
 #define MM1		%mm1
@@ -858,6 +868,7 @@
 #if defined(NASM_ASSEMBLER)
 
 #define ST(n)		st ## n
+#define ST0		st0
 
 #define TBYTE_PTR	tword
 #define QWORD_PTR	qword
diff --git a/src/mesa/x86/common_x86_asm.S b/src/mesa/x86/common_x86_asm.S
index 8a0a6477afa..61ff15a1df8 100644
--- a/src/mesa/x86/common_x86_asm.S
+++ b/src/mesa/x86/common_x86_asm.S
@@ -1,4 +1,4 @@
-/* $Id: common_x86_asm.S,v 1.7 2001/03/29 06:46:16 gareth Exp $ */
+/* $Id: common_x86_asm.S,v 1.8 2002/03/07 21:40:08 brianp Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -34,6 +34,12 @@
  * Cleaned up and simplified by Gareth Hughes <[email protected]>
  */
 
+/*
+ * NOTE: Avoid using spaces in between '(' ')' and arguments, especially
+ * with macros like CONST, LLBL that expand to CONCAT(...).  Putting spaces
+ * in there will break the build on some platforms.
+ */
+
 #include "matypes.h"
 #include "common_x86_features.h"
 
@@ -87,7 +93,7 @@ GLNAME( _mesa_identify_x86_cpu_features ):
 	/* Verify the ID Flag bit has been written.
 	 */
 	CMP_L	( ECX, EAX )
-	JZ	( LLBL ( cpuid_done ) )
+	JZ	( LLBL (cpuid_done) )
 
 	/* Get the CPU vendor info.
 	 */
@@ -98,11 +104,11 @@ GLNAME( _mesa_identify_x86_cpu_features ):
 	 * "GenuineIntel" string in EBX, ECX and EDX.
 	 */
 	CMP_L	( CONST(GENU), EBX )
-	JNE	( LLBL( cpuid_amd ) )
+	JNE	( LLBL(cpuid_amd) )
 	CMP_L	( CONST(INEI), EDX )
-	JNE	( LLBL( cpuid_amd ) )
+	JNE	( LLBL(cpuid_amd) )
 	CMP_L	( CONST(NTEL), ECX )
-	JNE	( LLBL( cpuid_amd ) )
+	JNE	( LLBL(cpuid_amd) )
 
 	/* We have an Intel processor, so we can get the feature
 	 * information with an CPUID input value of 1.
@@ -110,19 +116,19 @@ GLNAME( _mesa_identify_x86_cpu_features ):
 	MOV_L	( CONST(0x1), EAX )
 	CPUID
 	MOV_L	( EDX, EAX )
-	JMP	( LLBL( cpuid_done ) )
+	JMP	( LLBL(cpuid_done) )
 
-LLBL( cpuid_amd ):
+LLBL(cpuid_amd):
 
 	/* Test for AMD processors.  We must look for the
 	 * "AuthenticAMD" string in EBX, ECX and EDX.
 	 */
 	CMP_L	( CONST(AUTH), EBX )
-	JNE	( LLBL( cpuid_other ) )
+	JNE	( LLBL(cpuid_other) )
 	CMP_L	( CONST(ENTI), EDX )
-	JNE	( LLBL( cpuid_other ) )
+	JNE	( LLBL(cpuid_other) )
 	CMP_L	( CONST(CAMD), ECX )
-	JNE	( LLBL( cpuid_other ) )
+	JNE	( LLBL(cpuid_other) )
 
 	/* We have an AMD processor, so we can get the feature
 	 * information after we verify that the extended functions are
@@ -131,19 +137,19 @@ LLBL( cpuid_amd ):
 	MOV_L	( CONST(0x80000000), EAX )
 	CPUID
 	TEST_L	( EAX, EAX )
-	JZ	( LLBL ( cpuid_failed ) )
+	JZ	( LLBL (cpuid_failed) )
 
 	MOV_L	( CONST(0x80000001), EAX )
 	CPUID
 	MOV_L	( EDX, EAX )
-	JMP	( LLBL ( cpuid_done ) )
+	JMP	( LLBL (cpuid_done) )
 
-LLBL( cpuid_other ):
+LLBL(cpuid_other):
 
 	/* Test for other processors here when required.
 	 */
 
-LLBL( cpuid_failed ):
+LLBL(cpuid_failed):
 
 	/* If we can't determine the feature information, we must
 	 * return zero to indicate that no platform-specific
@@ -151,7 +157,7 @@ LLBL( cpuid_failed ):
 	 */
 	MOV_L	( CONST(0), EAX )
 
-LLBL ( cpuid_done ):
+LLBL (cpuid_done):
 
 	POP_L	( EBX )
 	RET
diff --git a/src/mesa/x86/x86_cliptest.S b/src/mesa/x86/x86_cliptest.S
index 08e34ecbd6f..617eb2b6fea 100644
--- a/src/mesa/x86/x86_cliptest.S
+++ b/src/mesa/x86/x86_cliptest.S
@@ -1,4 +1,4 @@
-/* $Id: x86_cliptest.S,v 1.8 2001/05/21 16:33:41 gareth Exp $ */
+/* $Id: x86_cliptest.S,v 1.9 2002/03/07 21:40:08 brianp Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -24,23 +24,38 @@
  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
  */
 
+/*
+ * NOTE: Avoid using spaces in between '(' ')' and arguments, especially
+ * with macros like CONST, LLBL that expand to CONCAT(...).  Putting spaces
+ * in there will break the build on some platforms.
+ */
+
 #include "matypes.h"
 #include "clip_args.h"
 
-#define SRC(i)		REGOFF(i * 4, ESI)
-#define DST(i)		REGOFF(i * 4, EDI)
-#define MAT(i)		REGOFF(i * 4, EDX)
+#define SRC0		REGOFF(0, ESI)
+#define SRC1		REGOFF(4, ESI)
+#define SRC2		REGOFF(8, ESI)
+#define SRC3		REGOFF(12, ESI)
+#define DST0		REGOFF(0, EDI)
+#define DST1		REGOFF(4, EDI)
+#define DST2		REGOFF(8, EDI)
+#define DST3		REGOFF(12, EDI)
+#define MAT0		REGOFF(0, EDX)
+#define MAT1		REGOFF(4, EDX)
+#define MAT2		REGOFF(8, EDX)
+#define MAT3		REGOFF(12, EDX)
 
 
 /*
  * Table for clip test.
  *
- * 	bit6 = SRC(3) < 0
- * 	bit5 = SRC(2) < 0
+ * 	bit6 = SRC3 < 0
+ * 	bit5 = SRC2 < 0
  * 	bit4 = abs(S(2)) > abs(S(3))
- * 	bit3 = SRC(1) < 0
+ * 	bit3 = SRC1 < 0
  * 	bit2 = abs(S(1)) > abs(S(3))
- * 	bit1 = SRC(0) < 0
+ * 	bit1 = SRC0 < 0
  * 	bit0 = abs(S(0)) > abs(S(3))
  */
 
@@ -100,18 +115,18 @@ GLNAME( _mesa_x86_cliptest_points4 ):
 
 #ifdef ELFPIC
 	/* store pointer to clip_table on stack */
-	CALL( LLBL( ctp4_get_eip ) )
+	CALL( LLBL(ctp4_get_eip) )
 	ADD_L( CONST(_GLOBAL_OFFSET_TABLE_), EBX )
 	MOV_L( REGOFF(clip_table@GOT, EBX), EBX )
 	PUSH_L( EBX )
-	JMP( LLBL( ctp4_clip_table_ready ) )
+	JMP( LLBL(ctp4_clip_table_ready) )
 
-LLBL( ctp4_get_eip ):
+LLBL(ctp4_get_eip):
 	/* store eip in ebx */
 	MOV_L( REGIND(ESP), EBX )
 	RET
 
-LLBL( ctp4_clip_table_ready ):
+LLBL(ctp4_clip_table_ready):
 #endif
 
 	MOV_L( ARG_SOURCE, ESI )
@@ -141,16 +156,16 @@ LLBL( ctp4_clip_table_ready ):
 	MOV_B( REGIND(EBX), AL )
 	MOV_B( REGIND(EBP), AH )
 
-	JZ( LLBL( ctp4_finish ) )
+	JZ( LLBL(ctp4_finish) )
 
 ALIGNTEXT16
-LLBL( ctp4_top ):
+LLBL(ctp4_top):
 
 	FLD1				/* F3 */
-	FDIV_S( SRC(3) )		/* GH: don't care about div-by-zero */
+	FDIV_S( SRC3 )		/* GH: don't care about div-by-zero */
 
-	MOV_L( SRC(3), EBP )
-	MOV_L( SRC(2), EBX )
+	MOV_L( SRC3, EBP )
+	MOV_L( SRC2, EBX )
 
 	XOR_L( ECX, ECX )
 	ADD_L( EBP, EBP )	/* ebp = abs(S(3))*2 ; carry = sign of S(3) */
@@ -162,7 +177,7 @@ LLBL( ctp4_top ):
 	CMP_L( EBX, EBP )	/* carry = abs(S(2))*2 > abs(S(3))*2 */
 
 	ADC_L( ECX, ECX )
-	MOV_L( SRC(1), EBX )
+	MOV_L( SRC1, EBX )
 
 	ADD_L( EBX, EBX )	/* ebx = abs(S(1))*2 ; carry = sign of S(1) */
 
@@ -170,7 +185,7 @@ LLBL( ctp4_top ):
 	CMP_L( EBX, EBP )	/* carry = abs(S(1))*2 > abs(S(3))*2 */
 
 	ADC_L( ECX, ECX )
-	MOV_L( SRC(0), EBX )
+	MOV_L( SRC0, EBX )
 
 	ADD_L( EBX, EBX )	/* ebx = abs(S(0))*2 ; carry = sign of S(0) */
 
@@ -193,37 +208,37 @@ LLBL( ctp4_top ):
 	TEST_B( CL, CL )
 	MOV_B( CL, REGIND(EDX) )
 
-	JZ( LLBL( ctp4_proj ) )
+	JZ( LLBL(ctp4_proj) )
 
-LLBL( ctp4_noproj ):
+LLBL(ctp4_noproj):
 
 	FSTP( ST(0) )			/* */
 
-	MOV_L( CONST(0), DST(0) )
-	MOV_L( CONST(0), DST(1) )
-	MOV_L( CONST(0), DST(2) )
-	MOV_L( CONST(0x3f800000), DST(3) )
+	MOV_L( CONST(0), DST0 )
+	MOV_L( CONST(0), DST1 )
+	MOV_L( CONST(0), DST2 )
+	MOV_L( CONST(0x3f800000), DST3 )
 
-	JMP( LLBL( ctp4_next ) )
+	JMP( LLBL(ctp4_next) )
 
-LLBL( ctp4_proj ):
+LLBL(ctp4_proj):
 
-	FLD_S( SRC(0) )			/* F0 F3 */
-	FMUL2( ST(1), ST(0) )
+	FLD_S( SRC0 )			/* F0 F3 */
+	FMUL2( ST(1), ST0 )
 
-	FLD_S( SRC(1) )			/* F1 F0 F3 */
-	FMUL2( ST(2), ST(0) )
+	FLD_S( SRC1 )			/* F1 F0 F3 */
+	FMUL2( ST(2), ST0 )
 
-	FLD_S( SRC(2) )			/* F2 F1 F0 F3 */
-	FMUL2( ST(3), ST(0) )
+	FLD_S( SRC2 )			/* F2 F1 F0 F3 */
+	FMUL2( ST(3), ST0 )
 
 	FXCH( ST(2) )			/* F0 F1 F2 F3 */
-	FSTP_S( DST(0) )		/* F1 F2 F3 */
-	FSTP_S( DST(1) )		/* F2 F3 */
-	FSTP_S( DST(2) )		/* F3 */
-	FSTP_S( DST(3) )		/* */
+	FSTP_S( DST0 )		/* F1 F2 F3 */
+	FSTP_S( DST1 )		/* F2 F3 */
+	FSTP_S( DST2 )		/* F3 */
+	FSTP_S( DST3 )		/* */
 
-LLBL( ctp4_next ):
+LLBL(ctp4_next):
 
 	INC_L( EDX )
 	ADD_L( CONST(16), EDI )
@@ -231,7 +246,7 @@ LLBL( ctp4_next ):
 	ADD_L( ARG_SOURCE, ESI )
 	CMP_L( EDX, ARG_CLIP )
 
-	JNZ( LLBL( ctp4_top ) )
+	JNZ( LLBL(ctp4_top) )
 
 	MOV_L( ARG_OR, ECX )
 	MOV_L( ARG_AND, EDX )
@@ -239,7 +254,7 @@ LLBL( ctp4_next ):
 	MOV_B( AL, REGIND(ECX) )
 	MOV_B( AH, REGIND(EDX) )
 
-LLBL( ctp4_finish ):
+LLBL(ctp4_finish):
 
 	MOV_L( ARG_DEST, EAX )
 #ifdef ELFPIC
@@ -274,18 +289,18 @@ GLNAME( _mesa_x86_cliptest_points4_np ):
 
 #ifdef ELFPIC
 	/* store pointer to clip_table on stack */
-	CALL( LLBL( ctp4_np_get_eip ) )
+	CALL( LLBL(ctp4_np_get_eip) )
 	ADD_L( CONST(_GLOBAL_OFFSET_TABLE_), EBX )
 	MOV_L( REGOFF(clip_table@GOT, EBX), EBX )
 	PUSH_L( EBX )
-	JMP( LLBL( ctp4_np_clip_table_ready ) )
+	JMP( LLBL(ctp4_np_clip_table_ready) )
 
-LLBL( ctp4_np_get_eip ):
+LLBL(ctp4_np_get_eip):
 	/* store eip in ebx */
 	MOV_L( REGIND(ESP), EBX )
 	RET
 
-LLBL( ctp4_np_clip_table_ready ):
+LLBL(ctp4_np_clip_table_ready):
 #endif
 
 	MOV_L( ARG_SOURCE, ESI )
@@ -309,13 +324,13 @@ LLBL( ctp4_np_clip_table_ready ):
 	MOV_B( REGIND(EBX), AL )
 	MOV_B( REGIND(EBP), AH )
 
-	JZ( LLBL( ctp4_np_finish ) )
+	JZ( LLBL(ctp4_np_finish) )
 
 ALIGNTEXT16
-LLBL( ctp4_np_top ):
+LLBL(ctp4_np_top):
 
-	MOV_L( SRC(3), EBP )
-	MOV_L( SRC(2), EBX )
+	MOV_L( SRC3, EBP )
+	MOV_L( SRC2, EBX )
 
 	XOR_L( ECX, ECX )
 	ADD_L( EBP, EBP )	/* ebp = abs(S(3))*2 ; carry = sign of S(3) */
@@ -327,7 +342,7 @@ LLBL( ctp4_np_top ):
 	CMP_L( EBX, EBP )	/* carry = abs(S(2))*2 > abs(S(3))*2 */
 
 	ADC_L( ECX, ECX )
-	MOV_L( SRC(1), EBX )
+	MOV_L( SRC1, EBX )
 
 	ADD_L( EBX, EBX )	/* ebx = abs(S(1))*2 ; carry = sign of S(1) */
 
@@ -335,7 +350,7 @@ LLBL( ctp4_np_top ):
 	CMP_L( EBX, EBP )	/* carry = abs(S(1))*2 > abs(S(3))*2 */
 
 	ADC_L( ECX, ECX )
-	MOV_L( SRC(0), EBX )
+	MOV_L( SRC0, EBX )
 
 	ADD_L( EBX, EBX )	/* ebx = abs(S(0))*2 ; carry = sign of S(0) */
 
@@ -364,7 +379,7 @@ LLBL( ctp4_np_top ):
 	ADD_L( ARG_DEST, ESI )
 	CMP_L( EDX, EDI )
 
-	JNZ( LLBL( ctp4_np_top ) )
+	JNZ( LLBL(ctp4_np_top) )
 
 	MOV_L( ARG_OR, ECX )
 	MOV_L( ARG_AND, EDX )
@@ -372,7 +387,7 @@ LLBL( ctp4_np_top ):
 	MOV_B( AL, REGIND(ECX) )
 	MOV_B( AH, REGIND(EDX) )
 
-LLBL( ctp4_np_finish ):
+LLBL(ctp4_np_finish):
 
 	MOV_L( ARG_SOURCE, EAX )
 #ifdef ELFPIC
diff --git a/src/mesa/x86/x86_xform2.S b/src/mesa/x86/x86_xform2.S
index 0e46a8411ec..b9528e1653c 100644
--- a/src/mesa/x86/x86_xform2.S
+++ b/src/mesa/x86/x86_xform2.S
@@ -1,4 +1,4 @@
-/* $Id: x86_xform2.S,v 1.1 2001/03/29 06:46:27 gareth Exp $ */
+/* $Id: x86_xform2.S,v 1.2 2002/03/07 21:40:08 brianp Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -24,6 +24,12 @@
  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
  */
 
+/*
+ * NOTE: Avoid using spaces in between '(' ')' and arguments, especially
+ * with macros like CONST, LLBL that expand to CONCAT(...).  Putting spaces
+ * in there will break the build on some platforms.
+ */
+
 #include "matypes.h"
 #include "xform_args.h"
 
@@ -32,9 +38,30 @@
 #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)
+#define SRC0		REGOFF(0, ESI)
+#define SRC1		REGOFF(4, ESI)
+#define SRC2		REGOFF(8, ESI)
+#define SRC3		REGOFF(12, ESI)
+#define DST0		REGOFF(0, EDI)
+#define DST1		REGOFF(4, EDI)
+#define DST2		REGOFF(8, EDI)
+#define DST3		REGOFF(12, EDI)
+#define MAT0		REGOFF(0, EDX)
+#define MAT1		REGOFF(4, EDX)
+#define MAT2		REGOFF(8, EDX)
+#define MAT3		REGOFF(12, EDX)
+#define MAT4		REGOFF(16, EDX)
+#define MAT5		REGOFF(20, EDX)
+#define MAT6		REGOFF(24, EDX)
+#define MAT7		REGOFF(28, EDX)
+#define MAT8		REGOFF(32, EDX)
+#define MAT9		REGOFF(36, EDX)
+#define MAT10		REGOFF(40, EDX)
+#define MAT11		REGOFF(44, EDX)
+#define MAT12		REGOFF(48, EDX)
+#define MAT13		REGOFF(52, EDX)
+#define MAT14		REGOFF(56, EDX)
+#define MAT15		REGOFF(60, EDX)
 
 
 ALIGNTEXT16
@@ -52,7 +79,7 @@ GLNAME( _mesa_x86_transform_points2_general ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p2_gr_done ) )
+	JZ( LLBL(x86_p2_gr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -67,57 +94,57 @@ GLNAME( _mesa_x86_transform_points2_general ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p2_gr_loop ):
-
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
-	FLD_S( SRC(0) )			/* F5 F4 */
-	FMUL_S( MAT(1) )
-	FLD_S( SRC(0) )			/* F6 F5 F4 */
-	FMUL_S( MAT(2) )
-	FLD_S( SRC(0) )			/* F7 F6 F5 F4 */
-	FMUL_S( MAT(3) )
-
-	FLD_S( SRC(1) )			/* F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(4) )
-	FLD_S( SRC(1) )			/* F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(5) )
-	FLD_S( SRC(1) )			/* F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(6) )
-	FLD_S( SRC(1) )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(7) )
+LLBL(x86_p2_gr_loop):
+
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
+	FLD_S( SRC0 )			/* F5 F4 */
+	FMUL_S( MAT1 )
+	FLD_S( SRC0 )			/* F6 F5 F4 */
+	FMUL_S( MAT2 )
+	FLD_S( SRC0 )			/* F7 F6 F5 F4 */
+	FMUL_S( MAT3 )
+
+	FLD_S( SRC1 )			/* F0 F7 F6 F5 F4 */
+	FMUL_S( MAT4 )
+	FLD_S( SRC1 )			/* F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT5 )
+	FLD_S( SRC1 )			/* F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT6 )
+	FLD_S( SRC1 )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT7 )
 
 	FXCH( ST(3) )			/* F0 F2 F1 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
 	FXCH( ST(1) )			/* F1 F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F7 F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F2 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F7 F6 F5 F4 */
 
 	FXCH( ST(3) )			/* F4 F6 F5 F7 */
-	FADD_S( MAT(12) )
+	FADD_S( MAT12 )
 	FXCH( ST(2) )			/* F5 F6 F4 F7 */
-	FADD_S( MAT(13) )
+	FADD_S( MAT13 )
 	FXCH( ST(1) )			/* F6 F5 F4 F7 */
-	FADD_S( MAT(14) )
+	FADD_S( MAT14 )
 	FXCH( ST(3) )			/* F7 F5 F4 F6 */
-	FADD_S( MAT(15) )
+	FADD_S( MAT15 )
 
 	FXCH( ST(2) )			/* F4 F5 F7 F6 */
-	FSTP_S( DST(0) )		/* F5 F7 F6 */
-	FSTP_S( DST(1) )		/* F7 F6 */
+	FSTP_S( DST0 )			/* F5 F7 F6 */
+	FSTP_S( DST1 )			/* F7 F6 */
 	FXCH( ST(1) )			/* F6 F7 */
-	FSTP_S( DST(2) )		/* F7 */
-	FSTP_S( DST(3) )		/* */
+	FSTP_S( DST2 )			/* F7 */
+	FSTP_S( DST3 )			/* */
 
-LLBL( x86_p2_gr_skip ):
+LLBL(x86_p2_gr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p2_gr_loop ) )
+	JNE( LLBL(x86_p2_gr_loop) )
 
-LLBL( x86_p2_gr_done ):
+LLBL(x86_p2_gr_done):
 
 	POP_L( EDI )
 	POP_L( ESI )
@@ -143,7 +170,7 @@ GLNAME( _mesa_x86_transform_points2_perspective ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p2_pr_done ) )
+	JZ( LLBL(x86_p2_pr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -157,31 +184,31 @@ GLNAME( _mesa_x86_transform_points2_perspective ):
 	MOV_L( REGOFF(V4F_START, EDI), EDI )
 	ADD_L( EDI, ECX )
 
-	MOV_L( MAT(14), EBX )
+	MOV_L( MAT14, EBX )
 
 ALIGNTEXT16
-LLBL( x86_p2_pr_loop ):
+LLBL(x86_p2_pr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
 
-	FLD_S( SRC(1) )			/* F1 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F1 F4 */
+	FMUL_S( MAT5 )
 
 	FXCH( ST(1) )			/* F4 F1 */
-	FSTP_S( DST(0)   )		/* F1 */
-	FSTP_S( DST(1)   )		/* */
-	MOV_L( EBX, DST(2) )
-	MOV_L( CONST(FP_ZERO), DST(3) )
+	FSTP_S( DST0   )		/* F1 */
+	FSTP_S( DST1   )		/* */
+	MOV_L( EBX, DST2 )
+	MOV_L( CONST(FP_ZERO), DST3 )
 
-LLBL( x86_p2_pr_skip ):
+LLBL(x86_p2_pr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p2_pr_loop ) )
+	JNE( LLBL(x86_p2_pr_loop) )
 
-LLBL( x86_p2_pr_done ):
+LLBL(x86_p2_pr_done):
 
 	POP_L( EBX )
 	POP_L( EDI )
@@ -207,7 +234,7 @@ GLNAME( _mesa_x86_transform_points2_3d ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p2_3dr_done ) )
+	JZ( LLBL(x86_p2_3dr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_3), REGOFF(V4F_FLAGS, EDI) )
@@ -222,48 +249,48 @@ GLNAME( _mesa_x86_transform_points2_3d ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p2_3dr_loop ):
-
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
-	FLD_S( SRC(0) )			/* F5 F4 */
-	FMUL_S( MAT(1) )
-	FLD_S( SRC(0) )			/* F6 F5 F4 */
-	FMUL_S( MAT(2) )
-
-	FLD_S( SRC(1) )			/* F0 F6 F5 F4 */
-	FMUL_S( MAT(4) )
-	FLD_S( SRC(1) )			/* F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(5) )
-	FLD_S( SRC(1) )			/* F2 F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(6) )
+LLBL(x86_p2_3dr_loop):
+
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
+	FLD_S( SRC0 )			/* F5 F4 */
+	FMUL_S( MAT1 )
+	FLD_S( SRC0 )			/* F6 F5 F4 */
+	FMUL_S( MAT2 )
+
+	FLD_S( SRC1 )			/* F0 F6 F5 F4 */
+	FMUL_S( MAT4 )
+	FLD_S( SRC1 )			/* F1 F0 F6 F5 F4 */
+	FMUL_S( MAT5 )
+	FLD_S( SRC1 )			/* F2 F1 F0 F6 F5 F4 */
+	FMUL_S( MAT6 )
 
 	FXCH( ST(2) )			/* F0 F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F2 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F1 F2 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F2 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F6 F5 F4 */
 
 	FXCH( ST(2) )			/* F4 F5 F6 */
-	FADD_S( MAT(12) )
+	FADD_S( MAT12 )
 	FXCH( ST(1) )			/* F5 F4 F6 */
-	FADD_S( MAT(13) )
+	FADD_S( MAT13 )
 	FXCH( ST(2) )			/* F6 F4 F5 */
-	FADD_S( MAT(14) )
+	FADD_S( MAT14 )
 
 	FXCH( ST(1) )			/* F4 F6 F5 */
-	FSTP_S( DST(0) )		/* F6 F5 */
+	FSTP_S( DST0 )			/* F6 F5 */
 	FXCH( ST(1) )			/* F5 F6 */
-	FSTP_S( DST(1) )		/* F6 */
-	FSTP_S( DST(2) )		/* */
+	FSTP_S( DST1 )			/* F6 */
+	FSTP_S( DST2 )			/* */
 
-LLBL( x86_p2_3dr_skip ):
+LLBL(x86_p2_3dr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p2_3dr_loop ) )
+	JNE( LLBL(x86_p2_3dr_loop) )
 
-LLBL( x86_p2_3dr_done ):
+LLBL(x86_p2_3dr_done):
 
 	POP_L( EDI )
 	POP_L( ESI )
@@ -289,7 +316,7 @@ GLNAME( _mesa_x86_transform_points2_3d_no_rot ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p2_3dnrr_done ) )
+	JZ( LLBL(x86_p2_3dnrr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_3), REGOFF(V4F_FLAGS, EDI) )
@@ -303,35 +330,35 @@ GLNAME( _mesa_x86_transform_points2_3d_no_rot ):
 	MOV_L( REGOFF(V4F_START, EDI), EDI )
 	ADD_L( EDI, ECX )
 
-	MOV_L( MAT(14), EBX )
+	MOV_L( MAT14, EBX )
 
 ALIGNTEXT16
-LLBL( x86_p2_3dnrr_loop ):
+LLBL(x86_p2_3dnrr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
 
-	FLD_S( SRC(1) )			/* F1 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F1 F4 */
+	FMUL_S( MAT5 )
 
 	FXCH( ST(1) )			/* F4 F1 */
-	FADD_S( MAT(12) )
-	FLD_S( MAT(13) )		/* F5 F4 F1 */
+	FADD_S( MAT12 )
+	FLD_S( MAT13 )		/* F5 F4 F1 */
 	FXCH( ST(2) )			/* F1 F4 F5 */
-	FADDP( ST(0), ST(2) )		/* F4 F5 */
+	FADDP( ST0, ST(2) )		/* F4 F5 */
 
-	FSTP_S( DST(0) )		/* F5 */
-	FSTP_S( DST(1) )		/* */
-	MOV_L( EBX, DST(2) )
+	FSTP_S( DST0 )		/* F5 */
+	FSTP_S( DST1 )		/* */
+	MOV_L( EBX, DST2 )
 
-LLBL( x86_p2_3dnrr_skip ):
+LLBL(x86_p2_3dnrr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p2_3dnrr_loop ) )
+	JNE( LLBL(x86_p2_3dnrr_loop) )
 
-LLBL( x86_p2_3dnrr_done ):
+LLBL(x86_p2_3dnrr_done):
 
 	POP_L( EBX )
 	POP_L( EDI )
@@ -357,7 +384,7 @@ GLNAME( _mesa_x86_transform_points2_2d ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p2_2dr_done ) )
+	JZ( LLBL(x86_p2_2dr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_2), REGOFF(V4F_FLAGS, EDI) )
@@ -372,39 +399,39 @@ GLNAME( _mesa_x86_transform_points2_2d ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p2_2dr_loop ):
+LLBL(x86_p2_2dr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
-	FLD_S( SRC(0) )			/* F5 F4 */
-	FMUL_S( MAT(1) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
+	FLD_S( SRC0 )			/* F5 F4 */
+	FMUL_S( MAT1 )
 
-	FLD_S( SRC(1) )			/* F0 F5 F4 */
-	FMUL_S( MAT(4) )
-	FLD_S( SRC(1) )			/* F1 F0 F5 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F0 F5 F4 */
+	FMUL_S( MAT4 )
+	FLD_S( SRC1 )			/* F1 F0 F5 F4 */
+	FMUL_S( MAT5 )
 
 	FXCH( ST(1) )			/* F0 F1 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F1 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F5 F4 */
+	FADDP( ST0, ST(3) )		/* F1 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F5 F4 */
 
 	FXCH( ST(1) )			/* F4 F5 */
-	FADD_S( MAT(12) )
+	FADD_S( MAT12 )
 	FXCH( ST(1) )			/* F5 F4 */
-	FADD_S( MAT(13) )
+	FADD_S( MAT13 )
 
 	FXCH( ST(1) )			/* F4 F5 */
-	FSTP_S( DST(0) )		/* F5 */
-	FSTP_S( DST(1) )		/* */
+	FSTP_S( DST0 )		/* F5 */
+	FSTP_S( DST1 )		/* */
 
-LLBL( x86_p2_2dr_skip ):
+LLBL(x86_p2_2dr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p2_2dr_loop ) )
+	JNE( LLBL(x86_p2_2dr_loop) )
 
-LLBL( x86_p2_2dr_done ):
+LLBL(x86_p2_2dr_done):
 
 	POP_L( EDI )
 	POP_L( ESI )
@@ -429,7 +456,7 @@ GLNAME( _mesa_x86_transform_points2_2d_no_rot ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p2_2dnrr_done ) )
+	JZ( LLBL(x86_p2_2dnrr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_2), REGOFF(V4F_FLAGS, EDI) )
@@ -444,31 +471,31 @@ GLNAME( _mesa_x86_transform_points2_2d_no_rot ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p2_2dnrr_loop ):
+LLBL(x86_p2_2dnrr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
 
-	FLD_S( SRC(1) )			/* F1 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F1 F4 */
+	FMUL_S( MAT5 )
 
 	FXCH( ST(1) )			/* F4 F1 */
-	FADD_S( MAT(12) )
-	FLD_S( MAT(13) )		/* F5 F4 F1 */
+	FADD_S( MAT12 )
+	FLD_S( MAT13 )		/* F5 F4 F1 */
 	FXCH( ST(2) )			/* F1 F4 F5 */
-	FADDP( ST(0), ST(2) )		/* F4 F5 */
+	FADDP( ST0, ST(2) )		/* F4 F5 */
 
-	FSTP_S( DST(0)   )		/* F5 */
-	FSTP_S( DST(1)   )		/* */
+	FSTP_S( DST0   )		/* F5 */
+	FSTP_S( DST1   )		/* */
 
-LLBL( x86_p2_2dnrr_skip ):
+LLBL(x86_p2_2dnrr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p2_2dnrr_loop ) )
+	JNE( LLBL(x86_p2_2dnrr_loop) )
 
-LLBL( x86_p2_2dnrr_done ):
+LLBL(x86_p2_2dnrr_done):
 
 	POP_L( EDI )
 	POP_L( ESI )
@@ -494,7 +521,7 @@ GLNAME( _mesa_x86_transform_points2_identity ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p2_ir_done ) )
+	JZ( LLBL(x86_p2_ir_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_2), REGOFF(V4F_FLAGS, EDI) )
@@ -509,25 +536,25 @@ GLNAME( _mesa_x86_transform_points2_identity ):
 	ADD_L( EDI, ECX )
 
 	CMP_L( ESI, EDI )
-	JE( LLBL( x86_p2_ir_done ) )
+	JE( LLBL(x86_p2_ir_done) )
 
 ALIGNTEXT16
-LLBL( x86_p2_ir_loop ):
+LLBL(x86_p2_ir_loop):
 
-	MOV_L( SRC(0), EBX )
-	MOV_L( SRC(1), EDX )
+	MOV_L( SRC0, EBX )
+	MOV_L( SRC1, EDX )
 
-	MOV_L( EBX, DST(0) )
-	MOV_L( EDX, DST(1) )
+	MOV_L( EBX, DST0 )
+	MOV_L( EDX, DST1 )
 
-LLBL( x86_p2_ir_skip ):
+LLBL(x86_p2_ir_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p2_ir_loop ) )
+	JNE( LLBL(x86_p2_ir_loop) )
 
-LLBL( x86_p2_ir_done ):
+LLBL(x86_p2_ir_done):
 
 	POP_L( EBX )
 	POP_L( EDI )
diff --git a/src/mesa/x86/x86_xform3.S b/src/mesa/x86/x86_xform3.S
index 1f798744e81..d71a0fc3ba6 100644
--- a/src/mesa/x86/x86_xform3.S
+++ b/src/mesa/x86/x86_xform3.S
@@ -1,4 +1,4 @@
-/* $Id: x86_xform3.S,v 1.1 2001/03/29 06:46:27 gareth Exp $ */
+/* $Id: x86_xform3.S,v 1.2 2002/03/07 21:40:08 brianp Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -24,6 +24,12 @@
  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
  */
 
+/*
+ * NOTE: Avoid using spaces in between '(' ')' and arguments, especially
+ * with macros like CONST, LLBL that expand to CONCAT(...).  Putting spaces
+ * in there will break the build on some platforms.
+ */
+
 #include "matypes.h"
 #include "xform_args.h"
 
@@ -32,9 +38,30 @@
 #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)
+#define SRC0		REGOFF(0, ESI)
+#define SRC1		REGOFF(4, ESI)
+#define SRC2		REGOFF(8, ESI)
+#define SRC3		REGOFF(12, ESI)
+#define DST0		REGOFF(0, EDI)
+#define DST1		REGOFF(4, EDI)
+#define DST2		REGOFF(8, EDI)
+#define DST3		REGOFF(12, EDI)
+#define MAT0		REGOFF(0, EDX)
+#define MAT1		REGOFF(4, EDX)
+#define MAT2		REGOFF(8, EDX)
+#define MAT3		REGOFF(12, EDX)
+#define MAT4		REGOFF(16, EDX)
+#define MAT5		REGOFF(20, EDX)
+#define MAT6		REGOFF(24, EDX)
+#define MAT7		REGOFF(28, EDX)
+#define MAT8		REGOFF(32, EDX)
+#define MAT9		REGOFF(36, EDX)
+#define MAT10		REGOFF(40, EDX)
+#define MAT11		REGOFF(44, EDX)
+#define MAT12		REGOFF(48, EDX)
+#define MAT13		REGOFF(52, EDX)
+#define MAT14		REGOFF(56, EDX)
+#define MAT15		REGOFF(60, EDX)
 
 
 ALIGNTEXT16
@@ -52,7 +79,7 @@ GLNAME( _mesa_x86_transform_points3_general ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p3_gr_done ) )
+	JZ( LLBL(x86_p3_gr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -67,73 +94,73 @@ GLNAME( _mesa_x86_transform_points3_general ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p3_gr_loop ):
-
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
-	FLD_S( SRC(0) )			/* F5 F4 */
-	FMUL_S( MAT(1) )
-	FLD_S( SRC(0) )			/* F6 F5 F4 */
-	FMUL_S( MAT(2) )
-	FLD_S( SRC(0) )			/* F7 F6 F5 F4 */
-	FMUL_S( MAT(3) )
-
-	FLD_S( SRC(1) )			/* F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(4) )
-	FLD_S( SRC(1) )			/* F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(5) )
-	FLD_S( SRC(1) )			/* F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(6) )
-	FLD_S( SRC(1) )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(7) )
+LLBL(x86_p3_gr_loop):
+
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
+	FLD_S( SRC0 )			/* F5 F4 */
+	FMUL_S( MAT1 )
+	FLD_S( SRC0 )			/* F6 F5 F4 */
+	FMUL_S( MAT2 )
+	FLD_S( SRC0 )			/* F7 F6 F5 F4 */
+	FMUL_S( MAT3 )
+
+	FLD_S( SRC1 )			/* F0 F7 F6 F5 F4 */
+	FMUL_S( MAT4 )
+	FLD_S( SRC1 )			/* F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT5 )
+	FLD_S( SRC1 )			/* F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT6 )
+	FLD_S( SRC1 )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT7 )
 
 	FXCH( ST(3) )			/* F0 F2 F1 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
 	FXCH( ST(1) )			/* F1 F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F7 F6 F5 F4 */
-
-	FLD_S( SRC(2) )			/* F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(8) )
-	FLD_S( SRC(2) )			/* F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(9) )
-	FLD_S( SRC(2) )			/* F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(10) )
-	FLD_S( SRC(2) )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(11) )
+	FADDP( ST0, ST(5) )		/* F2 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F7 F6 F5 F4 */
+
+	FLD_S( SRC2 )			/* F0 F7 F6 F5 F4 */
+	FMUL_S( MAT8 )
+	FLD_S( SRC2 )			/* F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT9 )
+	FLD_S( SRC2 )			/* F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT10 )
+	FLD_S( SRC2 )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT11 )
 
 	FXCH( ST(3) )			/* F0 F2 F1 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
 	FXCH( ST(1) )			/* F1 F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F7 F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F2 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F7 F6 F5 F4 */
 
 	FXCH( ST(3) )			/* F4 F6 F5 F7 */
-	FADD_S( MAT(12) )
+	FADD_S( MAT12 )
 	FXCH( ST(2) )			/* F5 F6 F4 F7 */
-	FADD_S( MAT(13) )
+	FADD_S( MAT13 )
 	FXCH( ST(1) )			/* F6 F5 F4 F7 */
-	FADD_S( MAT(14) )
+	FADD_S( MAT14 )
 	FXCH( ST(3) )			/* F7 F5 F4 F6 */
-	FADD_S( MAT(15) )
+	FADD_S( MAT15 )
 
 	FXCH( ST(2) )			/* F4 F5 F7 F6 */
-	FSTP_S( DST(0) )		/* F5 F7 F6 */
-	FSTP_S( DST(1) )		/* F7 F6 */
+	FSTP_S( DST0 )		/* F5 F7 F6 */
+	FSTP_S( DST1 )		/* F7 F6 */
 	FXCH( ST(1) )			/* F6 F7 */
-	FSTP_S( DST(2) )		/* F7 */
-	FSTP_S( DST(3) )		/* */
+	FSTP_S( DST2 )		/* F7 */
+	FSTP_S( DST3 )		/* */
 
-LLBL( x86_p3_gr_skip ):
+LLBL(x86_p3_gr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p3_gr_loop ) )
+	JNE( LLBL(x86_p3_gr_loop) )
 
-LLBL( x86_p3_gr_done ):
+LLBL(x86_p3_gr_done):
 
 	POP_L( EDI )
 	POP_L( ESI )
@@ -159,7 +186,7 @@ GLNAME( _mesa_x86_transform_points3_perspective ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p3_pr_done ) )
+	JZ( LLBL(x86_p3_pr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -174,45 +201,45 @@ GLNAME( _mesa_x86_transform_points3_perspective ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p3_pr_loop ):
+LLBL(x86_p3_pr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
 
-	FLD_S( SRC(1) )			/* F5 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F5 F4 */
+	FMUL_S( MAT5 )
 
-	FLD_S( SRC(2) )			/* F0 F5 F4 */
-	FMUL_S( MAT(8) )
-	FLD_S( SRC(2) )			/* F1 F0 F5 F4 */
-	FMUL_S( MAT(9) )
-	FLD_S( SRC(2) )			/* F2 F1 F0 F5 F4 */
-	FMUL_S( MAT(10) )
+	FLD_S( SRC2 )			/* F0 F5 F4 */
+	FMUL_S( MAT8 )
+	FLD_S( SRC2 )			/* F1 F0 F5 F4 */
+	FMUL_S( MAT9 )
+	FLD_S( SRC2 )			/* F2 F1 F0 F5 F4 */
+	FMUL_S( MAT10 )
 
 	FXCH( ST(2) )			/* F0 F1 F2 F5 F4 */
-	FADDP( ST(0), ST(4) )		/* F1 F2 F5 F4 */
-	FADDP( ST(0), ST(2) )		/* F2 F5 F4 */
-	FLD_S( MAT(14) )		/* F6 F2 F5 F4 */
+	FADDP( ST0, ST(4) )		/* F1 F2 F5 F4 */
+	FADDP( ST0, ST(2) )		/* F2 F5 F4 */
+	FLD_S( MAT14 )		/* F6 F2 F5 F4 */
 	FXCH( ST(1) )			/* F2 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F6 F5 F4 */
 
-	MOV_L( SRC(2), EBX )
+	MOV_L( SRC2, EBX )
 	XOR_L( CONST(-2147483648), EBX )/* change sign */
 
 	FXCH( ST(2) )			/* F4 F5 F6 */
-	FSTP_S( DST(0) )		/* F5 F6 */
-	FSTP_S( DST(1) )		/* F6 */
-	FSTP_S( DST(2) )		/* */
-	MOV_L( EBX, DST(3) )
+	FSTP_S( DST0 )		/* F5 F6 */
+	FSTP_S( DST1 )		/* F6 */
+	FSTP_S( DST2 )		/* */
+	MOV_L( EBX, DST3 )
 
-LLBL( x86_p3_pr_skip ):
+LLBL(x86_p3_pr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p3_pr_loop ) )
+	JNE( LLBL(x86_p3_pr_loop) )
 
-LLBL( x86_p3_pr_done ):
+LLBL(x86_p3_pr_done):
 
 	POP_L( EBX )
 	POP_L( EDI )
@@ -238,7 +265,7 @@ GLNAME( _mesa_x86_transform_points3_3d ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p3_3dr_done ) )
+	JZ( LLBL(x86_p3_3dr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_3), REGOFF(V4F_FLAGS, EDI) )
@@ -253,60 +280,60 @@ GLNAME( _mesa_x86_transform_points3_3d ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p3_3dr_loop ):
-
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
-	FLD_S( SRC(0) )			/* F5 F4 */
-	FMUL_S( MAT(1) )
-	FLD_S( SRC(0) )			/* F6 F5 F4 */
-	FMUL_S( MAT(2) )
-
-	FLD_S( SRC(1) )			/* F0 F6 F5 F4 */
-	FMUL_S( MAT(4) )
-	FLD_S( SRC(1) )			/* F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(5) )
-	FLD_S( SRC(1) )			/* F2 F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(6) )
+LLBL(x86_p3_3dr_loop):
+
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
+	FLD_S( SRC0 )			/* F5 F4 */
+	FMUL_S( MAT1 )
+	FLD_S( SRC0 )			/* F6 F5 F4 */
+	FMUL_S( MAT2 )
+
+	FLD_S( SRC1 )			/* F0 F6 F5 F4 */
+	FMUL_S( MAT4 )
+	FLD_S( SRC1 )			/* F1 F0 F6 F5 F4 */
+	FMUL_S( MAT5 )
+	FLD_S( SRC1 )			/* F2 F1 F0 F6 F5 F4 */
+	FMUL_S( MAT6 )
 
 	FXCH( ST(2) )			/* F0 F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F2 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F1 F2 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F2 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F6 F5 F4 */
 
-	FLD_S( SRC(2) )			/* F0 F6 F5 F4 */
-	FMUL_S( MAT(8) )
-	FLD_S( SRC(2) )			/* F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(9) )
-	FLD_S( SRC(2) )			/* F2 F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(10) )
+	FLD_S( SRC2 )			/* F0 F6 F5 F4 */
+	FMUL_S( MAT8 )
+	FLD_S( SRC2 )			/* F1 F0 F6 F5 F4 */
+	FMUL_S( MAT9 )
+	FLD_S( SRC2 )			/* F2 F1 F0 F6 F5 F4 */
+	FMUL_S( MAT10 )
 
 	FXCH( ST(2) )			/* F0 F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F2 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F1 F2 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F2 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F6 F5 F4 */
 
 	FXCH( ST(2) )			/* F4 F5 F6 */
-	FADD_S( MAT(12) )
+	FADD_S( MAT12 )
 	FXCH( ST(1) )			/* F5 F4 F6 */
-	FADD_S( MAT(13) )
+	FADD_S( MAT13 )
 	FXCH( ST(2) )			/* F6 F4 F5 */
-	FADD_S( MAT(14) )
+	FADD_S( MAT14 )
 
 	FXCH( ST(1) )			/* F4 F6 F5 */
-	FSTP_S( DST(0)   )		/* F6 F5 */
+	FSTP_S( DST0   )		/* F6 F5 */
 	FXCH( ST(1) )			/* F5 F6 */
-	FSTP_S( DST(1)   )		/* F6 */
-	FSTP_S( DST(2)   )		/* */
+	FSTP_S( DST1   )		/* F6 */
+	FSTP_S( DST2   )		/* */
 
-LLBL( x86_p3_3dr_skip ):
+LLBL(x86_p3_3dr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p3_3dr_loop ) )
+	JNE( LLBL(x86_p3_3dr_loop) )
 
-LLBL( x86_p3_3dr_done ):
+LLBL(x86_p3_3dr_done):
 
 	POP_L( EDI )
 	POP_L( ESI )
@@ -332,7 +359,7 @@ GLNAME( _mesa_x86_transform_points3_3d_no_rot ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p3_3dnrr_done ) )
+	JZ( LLBL(x86_p3_3dnrr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_3), REGOFF(V4F_FLAGS, EDI) )
@@ -347,38 +374,38 @@ GLNAME( _mesa_x86_transform_points3_3d_no_rot ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p3_3dnrr_loop ):
+LLBL(x86_p3_3dnrr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
 
-	FLD_S( SRC(1) )			/* F1 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F1 F4 */
+	FMUL_S( MAT5 )
 
-	FLD_S( SRC(2) )			/* F2 F1 F4 */
-	FMUL_S( MAT(10) )
+	FLD_S( SRC2 )			/* F2 F1 F4 */
+	FMUL_S( MAT10 )
 
 	FXCH( ST(2) )			/* F4 F1 F2 */
-	FADD_S( MAT(12) )
-	FLD_S( MAT(13) )		/* F5 F4 F1 F2 */
+	FADD_S( MAT12 )
+	FLD_S( MAT13 )		/* F5 F4 F1 F2 */
 	FXCH( ST(2) )			/* F1 F4 F5 F2 */
-	FADDP( ST(0), ST(2) )		/* F4 F5 F2 */
-	FLD_S( MAT(14) )		/* F6 F4 F5 F2 */
+	FADDP( ST0, ST(2) )		/* F4 F5 F2 */
+	FLD_S( MAT14 )		/* F6 F4 F5 F2 */
 	FXCH( ST(3) )			/* F2 F4 F5 F6 */
-	FADDP( ST(0), ST(3) )		/* F4 F5 F6 */
+	FADDP( ST0, ST(3) )		/* F4 F5 F6 */
 
-	FSTP_S( DST(0)   )		/* F5 F6 */
-	FSTP_S( DST(1)   )		/* F6 */
-	FSTP_S( DST(2)   )		/* */
+	FSTP_S( DST0   )		/* F5 F6 */
+	FSTP_S( DST1   )		/* F6 */
+	FSTP_S( DST2   )		/* */
 
-LLBL( x86_p3_3dnrr_skip ):
+LLBL(x86_p3_3dnrr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p3_3dnrr_loop ) )
+	JNE( LLBL(x86_p3_3dnrr_loop) )
 
-LLBL( x86_p3_3dnrr_done ):
+LLBL(x86_p3_3dnrr_done):
 
 	POP_L( EDI )
 	POP_L( ESI )
@@ -404,7 +431,7 @@ GLNAME( _mesa_x86_transform_points3_2d ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p3_2dr_done ) )
+	JZ( LLBL(x86_p3_2dr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_3), REGOFF(V4F_FLAGS, EDI) )
@@ -419,42 +446,42 @@ GLNAME( _mesa_x86_transform_points3_2d ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p3_2dr_loop ):
+LLBL(x86_p3_2dr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
-	FLD_S( SRC(0) )			/* F5 F4 */
-	FMUL_S( MAT(1) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
+	FLD_S( SRC0 )			/* F5 F4 */
+	FMUL_S( MAT1 )
 
-	FLD_S( SRC(1) )			/* F0 F5 F4 */
-	FMUL_S( MAT(4) )
-	FLD_S( SRC(1) )			/* F1 F0 F5 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F0 F5 F4 */
+	FMUL_S( MAT4 )
+	FLD_S( SRC1 )			/* F1 F0 F5 F4 */
+	FMUL_S( MAT5 )
 
 	FXCH( ST(1) )			/* F0 F1 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F1 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F5 F4 */
+	FADDP( ST0, ST(3) )		/* F1 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F5 F4 */
 
 	FXCH( ST(1) )			/* F4 F5 */
-	FADD_S( MAT(12) )
+	FADD_S( MAT12 )
 	FXCH( ST(1) )			/* F5 F4 */
-	FADD_S( MAT(13) )
+	FADD_S( MAT13 )
 
-	MOV_L( SRC(2), EBX )
+	MOV_L( SRC2, EBX )
 
 	FXCH( ST(1) )			/* F4 F5 */
-	FSTP_S( DST(0)   )		/* F5 */
-	FSTP_S( DST(1)   )		/* */
-	MOV_L( EBX, DST(2) )
+	FSTP_S( DST0   )		/* F5 */
+	FSTP_S( DST1   )		/* */
+	MOV_L( EBX, DST2 )
 
-LLBL( x86_p3_2dr_skip ):
+LLBL(x86_p3_2dr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p3_2dr_loop ) )
+	JNE( LLBL(x86_p3_2dr_loop) )
 
-LLBL( x86_p3_2dr_done ):
+LLBL(x86_p3_2dr_done):
 
 	POP_L( EBX )
 	POP_L( EDI )
@@ -481,7 +508,7 @@ GLNAME( _mesa_x86_transform_points3_2d_no_rot ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p3_2dnrr_done ) )
+	JZ( LLBL(x86_p3_2dnrr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_3), REGOFF(V4F_FLAGS, EDI) )
@@ -496,35 +523,35 @@ GLNAME( _mesa_x86_transform_points3_2d_no_rot ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p3_2dnrr_loop ):
+LLBL(x86_p3_2dnrr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
 
-	FLD_S( SRC(1) )			/* F1 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F1 F4 */
+	FMUL_S( MAT5 )
 
 	FXCH( ST(1) )			/* F4 F1 */
-	FADD_S( MAT(12) )
-	FLD_S( MAT(13) )		/* F5 F4 F1 */
+	FADD_S( MAT12 )
+	FLD_S( MAT13 )		/* F5 F4 F1 */
 
 	FXCH( ST(2) )			/* F1 F4 F5 */
-	FADDP( ST(0), ST(2) )		/* F4 F5 */
+	FADDP( ST0, ST(2) )		/* F4 F5 */
 
-	MOV_L( SRC(2), EBX )
+	MOV_L( SRC2, EBX )
 
-	FSTP_S( DST(0) )		/* F5 */
-	FSTP_S( DST(1) )		/* */
-	MOV_L( EBX, DST(2) )
+	FSTP_S( DST0 )		/* F5 */
+	FSTP_S( DST1 )		/* */
+	MOV_L( EBX, DST2 )
 
-LLBL( x86_p3_2dnrr_skip ):
+LLBL(x86_p3_2dnrr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p3_2dnrr_loop ) )
+	JNE( LLBL(x86_p3_2dnrr_loop) )
 
-LLBL( x86_p3_2dnrr_done ):
+LLBL(x86_p3_2dnrr_done):
 
 	POP_L( EBX )
 	POP_L( EDI )
@@ -552,7 +579,7 @@ GLNAME(_mesa_x86_transform_points3_identity ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p3_ir_done ) )
+	JZ( LLBL(x86_p3_ir_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_3), REGOFF(V4F_FLAGS, EDI) )
@@ -567,37 +594,37 @@ GLNAME(_mesa_x86_transform_points3_identity ):
 	ADD_L( EDI, ECX )
 
 	CMP_L( ESI, EDI )
-	JE( LLBL( x86_p3_ir_done ) )
+	JE( LLBL(x86_p3_ir_done) )
 
 ALIGNTEXT16
-LLBL( x86_p3_ir_loop ):
+LLBL(x86_p3_ir_loop):
 
 #if 1
-	MOV_L( SRC(0), EBX )
-	MOV_L( SRC(1), EBP )
-	MOV_L( SRC(2), EDX )
+	MOV_L( SRC0, EBX )
+	MOV_L( SRC1, EBP )
+	MOV_L( SRC2, EDX )
 
-	MOV_L( EBX, DST(0) )
-	MOV_L( EBP, DST(1) )
-	MOV_L( EDX, DST(2) )
+	MOV_L( EBX, DST0 )
+	MOV_L( EBP, DST1 )
+	MOV_L( EDX, DST2 )
 #else
-	FLD_S( SRC(0) )
-	FLD_S( SRC(1) )
-	FLD_S( SRC(2) )
+	FLD_S( SRC0 )
+	FLD_S( SRC1 )
+	FLD_S( SRC2 )
 
-	FSTP_S( DST(2) )
-	FSTP_S( DST(1) )
-	FSTP_S( DST(0) )
+	FSTP_S( DST2 )
+	FSTP_S( DST1 )
+	FSTP_S( DST0 )
 #endif
 
-LLBL( x86_p3_ir_skip ):
+LLBL(x86_p3_ir_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p3_ir_loop ) )
+	JNE( LLBL(x86_p3_ir_loop) )
 
-LLBL( x86_p3_ir_done ):
+LLBL(x86_p3_ir_done):
 
 	POP_L( EBP )
 	POP_L( EBX )
diff --git a/src/mesa/x86/x86_xform4.S b/src/mesa/x86/x86_xform4.S
index 1b6a8d8aec0..a5c1c1c4b03 100644
--- a/src/mesa/x86/x86_xform4.S
+++ b/src/mesa/x86/x86_xform4.S
@@ -1,4 +1,4 @@
-/* $Id: x86_xform4.S,v 1.1 2001/03/29 06:46:27 gareth Exp $ */
+/* $Id: x86_xform4.S,v 1.2 2002/03/07 21:40:08 brianp Exp $ */
 
 /*
  * Mesa 3-D graphics library
@@ -24,6 +24,12 @@
  * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
  */
 
+/*
+ * NOTE: Avoid using spaces in between '(' ')' and arguments, especially
+ * with macros like CONST, LLBL that expand to CONCAT(...).  Putting spaces
+ * in there will break the build on some platforms.
+ */
+
 #include "matypes.h"
 #include "xform_args.h"
 
@@ -32,9 +38,30 @@
 #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)
+#define SRC0		REGOFF(0, ESI)
+#define SRC1		REGOFF(4, ESI)
+#define SRC2		REGOFF(8, ESI)
+#define SRC3		REGOFF(12, ESI)
+#define DST0		REGOFF(0, EDI)
+#define DST1		REGOFF(4, EDI)
+#define DST2		REGOFF(8, EDI)
+#define DST3		REGOFF(12, EDI)
+#define MAT0		REGOFF(0, EDX)
+#define MAT1		REGOFF(4, EDX)
+#define MAT2		REGOFF(8, EDX)
+#define MAT3		REGOFF(12, EDX)
+#define MAT4		REGOFF(16, EDX)
+#define MAT5		REGOFF(20, EDX)
+#define MAT6		REGOFF(24, EDX)
+#define MAT7		REGOFF(28, EDX)
+#define MAT8		REGOFF(32, EDX)
+#define MAT9		REGOFF(36, EDX)
+#define MAT10		REGOFF(40, EDX)
+#define MAT11		REGOFF(44, EDX)
+#define MAT12		REGOFF(48, EDX)
+#define MAT13		REGOFF(52, EDX)
+#define MAT14		REGOFF(56, EDX)
+#define MAT15		REGOFF(60, EDX)
 
 
 ALIGNTEXT16
@@ -52,7 +79,7 @@ GLNAME( _mesa_x86_transform_points4_general ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p4_gr_done ) )
+	JZ( LLBL(x86_p4_gr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -67,80 +94,80 @@ GLNAME( _mesa_x86_transform_points4_general ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p4_gr_loop ):
-
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
-	FLD_S( SRC(0) )			/* F5 F4 */
-	FMUL_S( MAT(1) )
-	FLD_S( SRC(0) )			/* F6 F5 F4 */
-	FMUL_S( MAT(2) )
-	FLD_S( SRC(0) )			/* F7 F6 F5 F4 */
-	FMUL_S( MAT(3) )
-
-	FLD_S( SRC(1) )			/* F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(4) )
-	FLD_S( SRC(1) )			/* F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(5) )
-	FLD_S( SRC(1) )			/* F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(6) )
-	FLD_S( SRC(1) )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(7) )
+LLBL(x86_p4_gr_loop):
+
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
+	FLD_S( SRC0 )			/* F5 F4 */
+	FMUL_S( MAT1 )
+	FLD_S( SRC0 )			/* F6 F5 F4 */
+	FMUL_S( MAT2 )
+	FLD_S( SRC0 )			/* F7 F6 F5 F4 */
+	FMUL_S( MAT3 )
+
+	FLD_S( SRC1 )			/* F0 F7 F6 F5 F4 */
+	FMUL_S( MAT4 )
+	FLD_S( SRC1 )			/* F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT5 )
+	FLD_S( SRC1 )			/* F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT6 )
+	FLD_S( SRC1 )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT7 )
 
 	FXCH( ST(3) )			/* F0 F2 F1 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
 	FXCH( ST(1) )			/* F1 F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F7 F6 F5 F4 */
-
-	FLD_S( SRC(2) )			/* F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(8) )
-	FLD_S( SRC(2) )			/* F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(9) )
-	FLD_S( SRC(2) )			/* F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(10) )
-	FLD_S( SRC(2) )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(11) )
+	FADDP( ST0, ST(5) )		/* F2 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F7 F6 F5 F4 */
+
+	FLD_S( SRC2 )			/* F0 F7 F6 F5 F4 */
+	FMUL_S( MAT8 )
+	FLD_S( SRC2 )			/* F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT9 )
+	FLD_S( SRC2 )			/* F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT10 )
+	FLD_S( SRC2 )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT11 )
 
 	FXCH( ST(3) )			/* F0 F2 F1 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
 	FXCH( ST(1) )			/* F1 F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F7 F6 F5 F4 */
-
-	FLD_S( SRC(3) )			/* F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(12) )
-	FLD_S( SRC(3) )			/* F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(13) )
-	FLD_S( SRC(3) )			/* F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(14) )
-	FLD_S( SRC(3) )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
-	FMUL_S( MAT(15) )
+	FADDP( ST0, ST(5) )		/* F2 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F7 F6 F5 F4 */
+
+	FLD_S( SRC3 )			/* F0 F7 F6 F5 F4 */
+	FMUL_S( MAT12 )
+	FLD_S( SRC3 )			/* F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT13 )
+	FLD_S( SRC3 )			/* F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT14 )
+	FLD_S( SRC3 )			/* F3 F2 F1 F0 F7 F6 F5 F4 */
+	FMUL_S( MAT15 )
 
 	FXCH( ST(3) )			/* F0 F2 F1 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(7) )		/* F2 F1 F3 F7 F6 F5 F4 */
 	FXCH( ST(1) )			/* F1 F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F2 F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F3 F7 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F7 F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F2 F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F3 F7 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F7 F6 F5 F4 */
 
 	FXCH( ST(3) )			/* F4 F6 F5 F7 */
-	FSTP_S( DST(0) )		/* F6 F5 F7 */
+	FSTP_S( DST0 )		/* F6 F5 F7 */
 	FXCH( ST(1) )			/* F5 F6 F7 */
-	FSTP_S( DST(1) )		/* F6 F7 */
-	FSTP_S( DST(2) )		/* F7 */
-	FSTP_S( DST(3) )		/* */
+	FSTP_S( DST1 )		/* F6 F7 */
+	FSTP_S( DST2 )		/* F7 */
+	FSTP_S( DST3 )		/* */
 
-LLBL( x86_p4_gr_skip ):
+LLBL(x86_p4_gr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p4_gr_loop ) )
+	JNE( LLBL(x86_p4_gr_loop) )
 
-LLBL( x86_p4_gr_done ):
+LLBL(x86_p4_gr_done):
 
 	POP_L( EDI )
 	POP_L( ESI )
@@ -166,7 +193,7 @@ GLNAME( _mesa_x86_transform_points4_perspective ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p4_pr_done ) )
+	JZ( LLBL(x86_p4_pr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -181,47 +208,47 @@ GLNAME( _mesa_x86_transform_points4_perspective ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p4_pr_loop ):
+LLBL(x86_p4_pr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
 
-	FLD_S( SRC(1) )			/* F5 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F5 F4 */
+	FMUL_S( MAT5 )
 
-	FLD_S( SRC(2) )			/* F0 F5 F4 */
-	FMUL_S( MAT(8) )
-	FLD_S( SRC(2) )			/* F1 F0 F5 F4 */
-	FMUL_S( MAT(9) )
-	FLD_S( SRC(2) )			/* F6 F1 F0 F5 F4 */
-	FMUL_S( MAT(10) )
+	FLD_S( SRC2 )			/* F0 F5 F4 */
+	FMUL_S( MAT8 )
+	FLD_S( SRC2 )			/* F1 F0 F5 F4 */
+	FMUL_S( MAT9 )
+	FLD_S( SRC2 )			/* F6 F1 F0 F5 F4 */
+	FMUL_S( MAT10 )
 
 	FXCH( ST(2) )			/* F0 F1 F6 F5 F4 */
-	FADDP( ST(0), ST(4) )		/* F1 F6 F5 F4 */
-	FADDP( ST(0), ST(2) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(4) )		/* F1 F6 F5 F4 */
+	FADDP( ST0, ST(2) )		/* F6 F5 F4 */
 
-	FLD_S( SRC(3) )			/* F2 F6 F5 F4 */
-	FMUL_S( MAT(14) )
+	FLD_S( SRC3 )			/* F2 F6 F5 F4 */
+	FMUL_S( MAT14 )
 
-	FADDP( ST(0), ST(1) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F6 F5 F4 */
 
-	MOV_L( SRC(2), EBX )
+	MOV_L( SRC2, EBX )
 	XOR_L( CONST(-2147483648), EBX )/* change sign */
 
 	FXCH( ST(2) )			/* F4 F5 F6 */
-	FSTP_S( DST(0) )		/* F5 F6 */
-	FSTP_S( DST(1) )		/* F6 */
-	FSTP_S( DST(2) )		/* */
-	MOV_L( EBX, DST(3) )
+	FSTP_S( DST0 )		/* F5 F6 */
+	FSTP_S( DST1 )		/* F6 */
+	FSTP_S( DST2 )		/* */
+	MOV_L( EBX, DST3 )
 
-LLBL( x86_p4_pr_skip ):
+LLBL(x86_p4_pr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p4_pr_loop ) )
+	JNE( LLBL(x86_p4_pr_loop) )
 
-LLBL( x86_p4_pr_done ):
+LLBL(x86_p4_pr_done):
 
 	POP_L( EBX )
 	POP_L( EDI )
@@ -248,7 +275,7 @@ GLNAME( _mesa_x86_transform_points4_3d ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p4_3dr_done ) )
+	JZ( LLBL(x86_p4_3dr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -263,67 +290,67 @@ GLNAME( _mesa_x86_transform_points4_3d ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p4_3dr_loop ):
-
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
-	FLD_S( SRC(0) )			/* F5 F4 */
-	FMUL_S( MAT(1) )
-	FLD_S( SRC(0) )			/* F6 F5 F4 */
-	FMUL_S( MAT(2) )
-
-	FLD_S( SRC(1) )			/* F0 F6 F5 F4 */
-	FMUL_S( MAT(4) )
-	FLD_S( SRC(1) )			/* F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(5) )
-	FLD_S( SRC(1) )			/* F2 F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(6) )
+LLBL(x86_p4_3dr_loop):
+
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
+	FLD_S( SRC0 )			/* F5 F4 */
+	FMUL_S( MAT1 )
+	FLD_S( SRC0 )			/* F6 F5 F4 */
+	FMUL_S( MAT2 )
+
+	FLD_S( SRC1 )			/* F0 F6 F5 F4 */
+	FMUL_S( MAT4 )
+	FLD_S( SRC1 )			/* F1 F0 F6 F5 F4 */
+	FMUL_S( MAT5 )
+	FLD_S( SRC1 )			/* F2 F1 F0 F6 F5 F4 */
+	FMUL_S( MAT6 )
 
 	FXCH( ST(2) )			/* F0 F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F2 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F1 F2 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F2 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F6 F5 F4 */
 
-	FLD_S( SRC(2) )			/* F0 F6 F5 F4 */
-	FMUL_S( MAT(8) )
-	FLD_S( SRC(2) )			/* F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(9) )
-	FLD_S( SRC(2) )			/* F2 F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(10) )
+	FLD_S( SRC2 )			/* F0 F6 F5 F4 */
+	FMUL_S( MAT8 )
+	FLD_S( SRC2 )			/* F1 F0 F6 F5 F4 */
+	FMUL_S( MAT9 )
+	FLD_S( SRC2 )			/* F2 F1 F0 F6 F5 F4 */
+	FMUL_S( MAT10 )
 
 	FXCH( ST(2) )			/* F0 F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F2 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F1 F2 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F2 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F6 F5 F4 */
 
-	FLD_S( SRC(3) )			/* F0 F6 F5 F4 */
-	FMUL_S( MAT(12) )
-	FLD_S( SRC(3) )			/* F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(13) )
-	FLD_S( SRC(3) )			/* F2 F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(14) )
+	FLD_S( SRC3 )			/* F0 F6 F5 F4 */
+	FMUL_S( MAT12 )
+	FLD_S( SRC3 )			/* F1 F0 F6 F5 F4 */
+	FMUL_S( MAT13 )
+	FLD_S( SRC3 )			/* F2 F1 F0 F6 F5 F4 */
+	FMUL_S( MAT14 )
 
 	FXCH( ST(2) )			/* F0 F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F2 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F1 F2 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F2 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F6 F5 F4 */
 
-	MOV_L( SRC(3), EBX )
+	MOV_L( SRC3, EBX )
 
 	FXCH( ST(2) )			/* F4 F5 F6 */
-	FSTP_S( DST(0) )		/* F5 F6 */
-	FSTP_S( DST(1) )		/* F6 */
-	FSTP_S( DST(2) )		/* */
-	MOV_L( EBX, DST(3) )
+	FSTP_S( DST0 )		/* F5 F6 */
+	FSTP_S( DST1 )		/* F6 */
+	FSTP_S( DST2 )		/* */
+	MOV_L( EBX, DST3 )
 
-LLBL( x86_p4_3dr_skip ):
+LLBL(x86_p4_3dr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p4_3dr_loop ) )
+	JNE( LLBL(x86_p4_3dr_loop) )
 
-LLBL( x86_p4_3dr_done ):
+LLBL(x86_p4_3dr_done):
 
 	POP_L( EBX )
 	POP_L( EDI )
@@ -350,7 +377,7 @@ GLNAME(_mesa_x86_transform_points4_3d_no_rot):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p4_3dnrr_done ) )
+	JZ( LLBL(x86_p4_3dnrr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -365,45 +392,45 @@ GLNAME(_mesa_x86_transform_points4_3d_no_rot):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p4_3dnrr_loop ):
+LLBL(x86_p4_3dnrr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
 
-	FLD_S( SRC(1) )			/* F5 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F5 F4 */
+	FMUL_S( MAT5 )
 
-	FLD_S( SRC(2) )			/* F6 F5 F4 */
-	FMUL_S( MAT(10) )
+	FLD_S( SRC2 )			/* F6 F5 F4 */
+	FMUL_S( MAT10 )
 
-	FLD_S( SRC(3) )			/* F0 F6 F5 F4 */
-	FMUL_S( MAT(12) )
-	FLD_S( SRC(3) )			/* F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(13) )
-	FLD_S( SRC(3) )			/* F2 F1 F0 F6 F5 F4 */
-	FMUL_S( MAT(14) )
+	FLD_S( SRC3 )			/* F0 F6 F5 F4 */
+	FMUL_S( MAT12 )
+	FLD_S( SRC3 )			/* F1 F0 F6 F5 F4 */
+	FMUL_S( MAT13 )
+	FLD_S( SRC3 )			/* F2 F1 F0 F6 F5 F4 */
+	FMUL_S( MAT14 )
 
 	FXCH( ST(2) )			/* F0 F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(5) )		/* F1 F2 F6 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F2 F6 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F6 F5 F4 */
+	FADDP( ST0, ST(5) )		/* F1 F2 F6 F5 F4 */
+	FADDP( ST0, ST(3) )		/* F2 F6 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F6 F5 F4 */
 
-	MOV_L( SRC(3), EBX )
+	MOV_L( SRC3, EBX )
 
 	FXCH( ST(2) )			/* F4 F5 F6 */
-	FSTP_S( DST(0)   )		/* F5 F6 */
-	FSTP_S( DST(1)   )		/* F6 */
-	FSTP_S( DST(2)   )		/* */
-	MOV_L( EBX, DST(3) )
+	FSTP_S( DST0   )		/* F5 F6 */
+	FSTP_S( DST1   )		/* F6 */
+	FSTP_S( DST2   )		/* */
+	MOV_L( EBX, DST3 )
 
-LLBL( x86_p4_3dnrr_skip ):
+LLBL(x86_p4_3dnrr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p4_3dnrr_loop ) )
+	JNE( LLBL(x86_p4_3dnrr_loop) )
 
-LLBL( x86_p4_3dnrr_done ):
+LLBL(x86_p4_3dnrr_done):
 
 	POP_L( EBX )
 	POP_L( EDI )
@@ -431,7 +458,7 @@ GLNAME( _mesa_x86_transform_points4_2d ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p4_2dr_done ) )
+	JZ( LLBL(x86_p4_2dr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -446,48 +473,48 @@ GLNAME( _mesa_x86_transform_points4_2d ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p4_2dr_loop ):
+LLBL(x86_p4_2dr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
-	FLD_S( SRC(0) )			/* F5 F4 */
-	FMUL_S( MAT(1) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
+	FLD_S( SRC0 )			/* F5 F4 */
+	FMUL_S( MAT1 )
 
-	FLD_S( SRC(1) )			/* F0 F5 F4 */
-	FMUL_S( MAT(4) )
-	FLD_S( SRC(1) )			/* F1 F0 F5 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F0 F5 F4 */
+	FMUL_S( MAT4 )
+	FLD_S( SRC1 )			/* F1 F0 F5 F4 */
+	FMUL_S( MAT5 )
 
 	FXCH( ST(1) )			/* F0 F1 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F1 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F5 F4 */
+	FADDP( ST0, ST(3) )		/* F1 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F5 F4 */
 
-	FLD_S( SRC(3) )			/* F0 F5 F4 */
-	FMUL_S( MAT(12) )
-	FLD_S( SRC(3) )			/* F1 F0 F5 F4 */
-	FMUL_S( MAT(13) )
+	FLD_S( SRC3 )			/* F0 F5 F4 */
+	FMUL_S( MAT12 )
+	FLD_S( SRC3 )			/* F1 F0 F5 F4 */
+	FMUL_S( MAT13 )
 
 	FXCH( ST(1) )			/* F0 F1 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F1 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F5 F4 */
+	FADDP( ST0, ST(3) )		/* F1 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F5 F4 */
 
-	MOV_L( SRC(2), EBX )
-	MOV_L( SRC(3), EBP )
+	MOV_L( SRC2, EBX )
+	MOV_L( SRC3, EBP )
 
 	FXCH( ST(1) )			/* F4 F5 */
-	FSTP_S( DST(0) )		/* F5 */
-	FSTP_S( DST(1) )		/* */
-	MOV_L( EBX, DST(2) )
-	MOV_L( EBP, DST(3) )
+	FSTP_S( DST0 )		/* F5 */
+	FSTP_S( DST1 )		/* */
+	MOV_L( EBX, DST2 )
+	MOV_L( EBP, DST3 )
 
-LLBL( x86_p4_2dr_skip ):
+LLBL(x86_p4_2dr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p4_2dr_loop ) )
+	JNE( LLBL(x86_p4_2dr_loop) )
 
-LLBL( x86_p4_2dr_done ):
+LLBL(x86_p4_2dr_done):
 
 	POP_L( EBP )
 	POP_L( EBX )
@@ -516,7 +543,7 @@ GLNAME( _mesa_x86_transform_points4_2d_no_rot ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p4_2dnrr_done ) )
+	JZ( LLBL(x86_p4_2dnrr_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -531,40 +558,40 @@ GLNAME( _mesa_x86_transform_points4_2d_no_rot ):
 	ADD_L( EDI, ECX )
 
 ALIGNTEXT16
-LLBL( x86_p4_2dnrr_loop ):
+LLBL(x86_p4_2dnrr_loop):
 
-	FLD_S( SRC(0) )			/* F4 */
-	FMUL_S( MAT(0) )
+	FLD_S( SRC0 )			/* F4 */
+	FMUL_S( MAT0 )
 
-	FLD_S( SRC(1) )			/* F5 F4 */
-	FMUL_S( MAT(5) )
+	FLD_S( SRC1 )			/* F5 F4 */
+	FMUL_S( MAT5 )
 
-	FLD_S( SRC(3) )			/* F0 F5 F4 */
-	FMUL_S( MAT(12) )
-	FLD_S( SRC(3) )			/* F1 F0 F5 F4 */
-	FMUL_S( MAT(13) )
+	FLD_S( SRC3 )			/* F0 F5 F4 */
+	FMUL_S( MAT12 )
+	FLD_S( SRC3 )			/* F1 F0 F5 F4 */
+	FMUL_S( MAT13 )
 
 	FXCH( ST(1) )			/* F0 F1 F5 F4 */
-	FADDP( ST(0), ST(3) )		/* F1 F5 F4 */
-	FADDP( ST(0), ST(1) )		/* F5 F4 */
+	FADDP( ST0, ST(3) )		/* F1 F5 F4 */
+	FADDP( ST0, ST(1) )		/* F5 F4 */
 
-	MOV_L( SRC(2), EBX )
-	MOV_L( SRC(3), EBP )
+	MOV_L( SRC2, EBX )
+	MOV_L( SRC3, EBP )
 
 	FXCH( ST(1) )			/* F4 F5 */
-	FSTP_S( DST(0)   )		/* F5 */
-	FSTP_S( DST(1)   )		/* */
-	MOV_L( EBX, DST(2) )
-	MOV_L( EBP, DST(3) )
+	FSTP_S( DST0   )		/* F5 */
+	FSTP_S( DST1   )		/* */
+	MOV_L( EBX, DST2 )
+	MOV_L( EBP, DST3 )
 
-LLBL( x86_p4_2dnrr_skip ):
+LLBL(x86_p4_2dnrr_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p4_2dnrr_loop ) )
+	JNE( LLBL(x86_p4_2dnrr_loop) )
 
-LLBL( x86_p4_2dnrr_done ):
+LLBL(x86_p4_2dnrr_done):
 
 	POP_L( EBP )
 	POP_L( EBX )
@@ -592,7 +619,7 @@ GLNAME( _mesa_x86_transform_points4_identity ):
 	MOV_L( REGOFF(V4F_COUNT, ESI), ECX )
 
 	TEST_L( ECX, ECX )
-	JZ( LLBL( x86_p4_ir_done ) )
+	JZ( LLBL(x86_p4_ir_done) )
 
 	MOV_L( REGOFF(V4F_STRIDE, ESI), EAX )
 	OR_L( CONST(VEC_SIZE_4), REGOFF(V4F_FLAGS, EDI) )
@@ -607,31 +634,31 @@ GLNAME( _mesa_x86_transform_points4_identity ):
 	ADD_L( EDI, ECX )
 
 	CMP_L( ESI, EDI )
-	JE( LLBL( x86_p4_ir_done ) )
+	JE( LLBL(x86_p4_ir_done) )
 
 ALIGNTEXT16
-LLBL( x86_p4_ir_loop ):
+LLBL(x86_p4_ir_loop):
 
-	MOV_L( SRC(0), EBX )
-	MOV_L( SRC(1), EDX )
+	MOV_L( SRC0, EBX )
+	MOV_L( SRC1, EDX )
 
-	MOV_L( EBX, DST(0) )
-	MOV_L( EDX, DST(1) )
+	MOV_L( EBX, DST0 )
+	MOV_L( EDX, DST1 )
 
-	MOV_L( SRC(2), EBX )
-	MOV_L( SRC(3), EDX )
+	MOV_L( SRC2, EBX )
+	MOV_L( SRC3, EDX )
 
-	MOV_L( EBX, DST(2) )
-	MOV_L( EDX, DST(3) )
+	MOV_L( EBX, DST2 )
+	MOV_L( EDX, DST3 )
 
-LLBL( x86_p4_ir_skip ):
+LLBL(x86_p4_ir_skip):
 
 	ADD_L( CONST(16), EDI )
 	ADD_L( EAX, ESI )
 	CMP_L( ECX, EDI )
-	JNE( LLBL( x86_p4_ir_loop ) )
+	JNE( LLBL(x86_p4_ir_loop) )
 
-LLBL( x86_p4_ir_done ):
+LLBL(x86_p4_ir_done):
 
 	POP_L( EBX )
 	POP_L( EDI )