--- SDL-1.2.7/src/video/mmx.h.orig	2003-08-22 07:51:19.000000000 +0200
+++ SDL-1.2.7/src/video/mmx.h	2003-11-09 13:49:35.270218296 +0100
@@ -246,15 +246,37 @@
 		printf(#op "_i2r(" #imm "=0x%08x%08x, ", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
 		__asm__ __volatile__ ("movq %%" #reg ", %0" \
-				      : "=X" (mmx_trace) \
+				      : "=m" (mmx_trace) \
 				      : /* nothing */ ); \
 		printf(#reg "=0x%08x%08x) => ", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
 		__asm__ __volatile__ (#op " %0, %%" #reg \
 				      : /* nothing */ \
-				      : "X" (imm)); \
+				      : "i" (imm)); \
 		__asm__ __volatile__ ("movq %%" #reg ", %0" \
-				      : "=X" (mmx_trace) \
+				      : "=m" (mmx_trace) \
+				      : /* nothing */ ); \
+		printf(#reg "=0x%08x%08x\n", \
+			mmx_trace.d[1], mmx_trace.d[0]); \
+	}
+
+/* 32 bits or less */
+#define	mmx_m2r32(op, mem, reg) \
+	{ \
+		mmx_t mmx_trace; \
+		mmx_trace = (mem); \
+		printf(#op "_m2r(" #mem "=0x%08x%08x, ", \
+			mmx_trace.d[1], mmx_trace.d[0]); \
+		__asm__ __volatile__ ("movq %%" #reg ", %0" \
+				      : "=m" (mmx_trace) \
+				      : /* nothing */ ); \
+		printf(#reg "=0x%08x%08x) => ", \
+			mmx_trace.d[1], mmx_trace.d[0]); \
+		__asm__ __volatile__ (#op " %0, %%" #reg \
+				      : /* nothing */ \
+				      : "mr" (mem)); \
+		__asm__ __volatile__ ("movq %%" #reg ", %0" \
+				      : "=m" (mmx_trace) \
 				      : /* nothing */ ); \
 		printf(#reg "=0x%08x%08x\n", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
@@ -267,25 +289,45 @@
 		printf(#op "_m2r(" #mem "=0x%08x%08x, ", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
 		__asm__ __volatile__ ("movq %%" #reg ", %0" \
-				      : "=X" (mmx_trace) \
+				      : "=m" (mmx_trace) \
 				      : /* nothing */ ); \
 		printf(#reg "=0x%08x%08x) => ", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
 		__asm__ __volatile__ (#op " %0, %%" #reg \
 				      : /* nothing */ \
-				      : "X" (mem)); \
+				      : "m" (mem)); \
 		__asm__ __volatile__ ("movq %%" #reg ", %0" \
-				      : "=X" (mmx_trace) \
+				      : "=m" (mmx_trace) \
 				      : /* nothing */ ); \
 		printf(#reg "=0x%08x%08x\n", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
 	}
 
+/* 32 bits or less */
+#define	mmx_r2m32(op, reg, mem) \
+	{ \
+		mmx_t mmx_trace; \
+		__asm__ __volatile__ ("movq %%" #reg ", %0" \
+				      : "=m" (mmx_trace) \
+				      : /* nothing */ ); \
+		printf(#op "_r2m(" #reg "=0x%08x%08x, ", \
+			mmx_trace.d[1], mmx_trace.d[0]); \
+		mmx_trace = (mem); \
+		printf(#mem "=0x%08x%08x) => ", \
+			mmx_trace.d[1], mmx_trace.d[0]); \
+		__asm__ __volatile__ (#op " %%" #reg ", %0" \
+				      : "=mr" (mem) \
+				      : /* nothing */ ); \
+		mmx_trace = (mem); \
+		printf(#mem "=0x%08x%08x\n", \
+			mmx_trace.d[1], mmx_trace.d[0]); \
+	}
+
 #define	mmx_r2m(op, reg, mem) \
 	{ \
 		mmx_t mmx_trace; \
 		__asm__ __volatile__ ("movq %%" #reg ", %0" \
-				      : "=X" (mmx_trace) \
+				      : "=m" (mmx_trace) \
 				      : /* nothing */ ); \
 		printf(#op "_r2m(" #reg "=0x%08x%08x, ", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
@@ -293,7 +335,7 @@
 		printf(#mem "=0x%08x%08x) => ", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
 		__asm__ __volatile__ (#op " %%" #reg ", %0" \
-				      : "=X" (mem) \
+				      : "=m" (mem) \
 				      : /* nothing */ ); \
 		mmx_trace = (mem); \
 		printf(#mem "=0x%08x%08x\n", \
@@ -304,18 +346,18 @@
 	{ \
 		mmx_t mmx_trace; \
 		__asm__ __volatile__ ("movq %%" #regs ", %0" \
-				      : "=X" (mmx_trace) \
+				      : "=m" (mmx_trace) \
 				      : /* nothing */ ); \
 		printf(#op "_r2r(" #regs "=0x%08x%08x, ", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
 		__asm__ __volatile__ ("movq %%" #regd ", %0" \
-				      : "=X" (mmx_trace) \
+				      : "=m" (mmx_trace) \
 				      : /* nothing */ ); \
 		printf(#regd "=0x%08x%08x) => ", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
 		__asm__ __volatile__ (#op " %" #regs ", %" #regd); \
 		__asm__ __volatile__ ("movq %%" #regd ", %0" \
-				      : "=X" (mmx_trace) \
+				      : "=m" (mmx_trace) \
 				      : /* nothing */ ); \
 		printf(#regd "=0x%08x%08x\n", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
@@ -333,8 +375,8 @@
 		__asm__ __volatile__ ("movq %0, %%mm0\n\t" \
 				      #op " %1, %%mm0\n\t" \
 				      "movq %%mm0, %0" \
-				      : "=X" (memd) \
-				      : "X" (mems)); \
+				      : "=my" (memd) \
+				      : "my" (mems)); \
 		mmx_trace = (memd); \
 		printf(#memd "=0x%08x%08x\n", \
 			mmx_trace.d[1], mmx_trace.d[0]); \
@@ -348,16 +390,28 @@
 #define	mmx_i2r(op, imm, reg) \
 	__asm__ __volatile__ (#op " %0, %%" #reg \
 			      : /* nothing */ \
-			      : "X" (imm) )
+			      : "i" (imm) )
+
+/* 32 bits or less */
+#define	mmx_m2r32(op, mem, reg) \
+	__asm__ __volatile__ (#op " %0, %%" #reg \
+			      : /* nothing */ \
+			      : "mr" (mem))
 
 #define	mmx_m2r(op, mem, reg) \
 	__asm__ __volatile__ (#op " %0, %%" #reg \
 			      : /* nothing */ \
 			      : "m" (mem))
+
+/* 32 bits or less */
+#define	mmx_r2m32(op, reg, mem) \
+	__asm__ __volatile__ (#op " %%" #reg ", %0" \
+			      : "=mr" (mem) \
+			      : /* nothing */ )
 
 #define	mmx_r2m(op, reg, mem) \
 	__asm__ __volatile__ (#op " %%" #reg ", %0" \
-			      : "=X" (mem) \
+			      : "=m" (mem) \
 			      : /* nothing */ )
 
 #define	mmx_r2r(op, regs, regd) \
@@ -367,8 +421,8 @@
 	__asm__ __volatile__ ("movq %0, %%mm0\n\t" \
 			      #op " %1, %%mm0\n\t" \
 			      "movq %%mm0, %0" \
-			      : "=X" (memd) \
-			      : "X" (mems))
+			      : "=my" (memd) \
+			      : "my" (mems))
 
 #endif
 
@@ -383,8 +437,8 @@
 #define	movq(vars, vard) \
 	__asm__ __volatile__ ("movq %1, %%mm0\n\t" \
 			      "movq %%mm0, %0" \
-			      : "=X" (vard) \
-			      : "X" (vars))
+			      : "=my" (vard) \
+			      : "my" (vars))
 
 
 /*	1x32 MOVe Doubleword
@@ -392,14 +446,14 @@
 	 but is most useful for moving things between
 	 mmx registers and ordinary registers)
 */
-#define	movd_m2r(var, reg)	mmx_m2r(movd, var, reg)
-#define	movd_r2m(reg, var)	mmx_r2m(movd, reg, var)
+#define	movd_m2r(var, reg)	mmx_m2r32(movd, var, reg)
+#define	movd_r2m(reg, var)	mmx_r2m32(movd, reg, var)
 #define	movd_r2r(regs, regd)	mmx_r2r(movd, regs, regd)
 #define	movd(vars, vard) \
 	__asm__ __volatile__ ("movd %1, %%mm0\n\t" \
 			      "movd %%mm0, %0" \
-			      : "=X" (vard) \
-			      : "X" (vars))
+			      : "=mr" (vard) \
+			      : "mr" (vars))
 
 
 /*	2x32, 4x16, and 8x8 Parallel ADDs