Changed comments to standard /* ... */

Changed 'asm' to __asm__

CVS patchset: 2929
CVS date: 2002/10/22 05:51:04

1 files changed, 307 insertions, 307 deletions

diff --git a/src/libffmpeg/libavcodec/i386/mpegvideo_mmx.c b/src/libffmpeg/libavcodec/i386/mpegvideo_mmx.c
index 10efc173f..80cf233c8 100644
--- a/src/libffmpeg/libavcodec/i386/mpegvideo_mmx.c
+++ b/src/libffmpeg/libavcodec/i386/mpegvideo_mmx.c
@@ -38,7 +38,7 @@ extern UINT32 inverse[256];
 void MPV_frame_start(MpegEncContext *s)
 {
     if (s->pict_type == B_TYPE) {
-	__asm __volatile(
+	____asm____ volatile(
 	    "movl	(%1), %%eax\n\t"
 	    "movl	4(%1), %%edx\n\t"
 	    "movl	8(%1), %%ecx\n\t"
@@ -50,7 +50,7 @@ void MPV_frame_start(MpegEncContext *s)
 	    :"eax","edx","ecx","memory");
     } else {
             /* swap next and last */
-	__asm __volatile(
+	____asm____ volatile(
 	    "movl	(%1), %%eax\n\t"
 	    "movl	4(%1), %%edx\n\t"
 	    "movl	8(%1), %%ecx\n\t"
@@ -78,7 +78,7 @@ static void dct_unquantize_h263_mmx(MpegEncContext *s,
                                   DCTELEM *block, int n, int qscale)
 {
     int i, level, qmul, qadd, nCoeffs;
-    
+
     qmul = s->qscale << 1;
     if (s->h263_aic && s->mb_intra)
         qadd = 0;
@@ -95,7 +95,7 @@ static void dct_unquantize_h263_mmx(MpegEncContext *s,
         for(i=1; i<8; i++) {
             level = block[i];
             if (level) {
-    	        if (level < 0) {
+	        if (level < 0) {
                     level = level * qmul - qadd;
                 } else {
                     level = level * qmul + qadd;
@@ -108,53 +108,53 @@ static void dct_unquantize_h263_mmx(MpegEncContext *s,
         i = 0;
         nCoeffs= zigzag_end[ s->block_last_index[n] ];
     }
-//printf("%d %d  ", qmul, qadd);
-asm volatile(
-		"movd %1, %%mm6			\n\t" //qmul
-		"packssdw %%mm6, %%mm6		\n\t"
-		"packssdw %%mm6, %%mm6		\n\t"
-		"movd %2, %%mm5			\n\t" //qadd
-		"pxor %%mm7, %%mm7		\n\t"
-		"packssdw %%mm5, %%mm5		\n\t"
-		"packssdw %%mm5, %%mm5		\n\t"
-		"psubw %%mm5, %%mm7		\n\t"
-		"pxor %%mm4, %%mm4		\n\t"
-		".balign 16\n\t"
-		"1:				\n\t"
-		"movq (%0, %3), %%mm0		\n\t"
-		"movq 8(%0, %3), %%mm1		\n\t"
-
-		"pmullw %%mm6, %%mm0		\n\t"
-		"pmullw %%mm6, %%mm1		\n\t"
-
-		"movq (%0, %3), %%mm2		\n\t"
-		"movq 8(%0, %3), %%mm3		\n\t"
-
-		"pcmpgtw %%mm4, %%mm2		\n\t" // block[i] < 0 ? -1 : 0
-		"pcmpgtw %%mm4, %%mm3		\n\t" // block[i] < 0 ? -1 : 0
-
-		"pxor %%mm2, %%mm0		\n\t"
-		"pxor %%mm3, %%mm1		\n\t"
-
-		"paddw %%mm7, %%mm0		\n\t"
-		"paddw %%mm7, %%mm1		\n\t"
-
-		"pxor %%mm0, %%mm2		\n\t"
-		"pxor %%mm1, %%mm3		\n\t"
-
-		"pcmpeqw %%mm7, %%mm0		\n\t" // block[i] == 0 ? -1 : 0
-		"pcmpeqw %%mm7, %%mm1		\n\t" // block[i] == 0 ? -1 : 0
-
-		"pandn %%mm2, %%mm0		\n\t"
-		"pandn %%mm3, %%mm1		\n\t"
-
-		"movq %%mm0, (%0, %3)		\n\t"
-		"movq %%mm1, 8(%0, %3)		\n\t"
-
-		"addl $16, %3			\n\t"
-		"js 1b				\n\t"
-		::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(i-nCoeffs))
-		: "memory"
+/* printf("%d %d  ", qmul, qadd); */
+__asm__ volatile(
+	"movd %1, %%mm6			\n\t" /* qmul */
+	"packssdw %%mm6, %%mm6		\n\t"
+	"packssdw %%mm6, %%mm6		\n\t"
+	"movd %2, %%mm5			\n\t" /* qadd */
+	"pxor %%mm7, %%mm7		\n\t"
+	"packssdw %%mm5, %%mm5		\n\t"
+	"packssdw %%mm5, %%mm5		\n\t"
+	"psubw %%mm5, %%mm7		\n\t"
+	"pxor %%mm4, %%mm4		\n\t"
+	".balign 16\n\t"
+	"1:				\n\t"
+	"movq (%0, %3), %%mm0		\n\t"
+	"movq 8(%0, %3), %%mm1		\n\t"
+
+	"pmullw %%mm6, %%mm0		\n\t"
+	"pmullw %%mm6, %%mm1		\n\t"
+
+	"movq (%0, %3), %%mm2		\n\t"
+	"movq 8(%0, %3), %%mm3		\n\t"
+
+	"pcmpgtw %%mm4, %%mm2		\n\t" /* block[i] < 0 ? -1 : 0 */
+	"pcmpgtw %%mm4, %%mm3		\n\t" /* block[i] < 0 ? -1 : 0 */
+
+	"pxor %%mm2, %%mm0		\n\t"
+	"pxor %%mm3, %%mm1		\n\t"
+
+	"paddw %%mm7, %%mm0		\n\t"
+	"paddw %%mm7, %%mm1		\n\t"
+
+	"pxor %%mm0, %%mm2		\n\t"
+	"pxor %%mm1, %%mm3		\n\t"
+
+	"pcmpeqw %%mm7, %%mm0		\n\t" /*  block[i] == 0 ? -1 : 0 */
+	"pcmpeqw %%mm7, %%mm1		\n\t" /*  block[i] == 0 ? -1 : 0 */
+
+	"pandn %%mm2, %%mm0		\n\t"
+	"pandn %%mm3, %%mm1		\n\t"
+
+	"movq %%mm0, (%0, %3)		\n\t"
+	"movq %%mm1, 8(%0, %3)		\n\t"
+
+	"addl $16, %3			\n\t"
+	"js 1b				\n\t"
+	::"r" (block+nCoeffs), "g"(qmul), "g" (qadd), "r" (2*(i-nCoeffs))
+	: "memory"
 	);
 }
 
@@ -193,123 +193,123 @@ static void dct_unquantize_mpeg1_mmx(MpegEncContext *s,
 {
     int nCoeffs;
     const UINT16 *quant_matrix;
-    
+
     if(s->alternate_scan) nCoeffs= 64;
     else nCoeffs= zigzag_end[ s->block_last_index[n] ];
 
     if (s->mb_intra) {
         int block0;
-        if (n < 4) 
+        if (n < 4)
             block0 = block[0] * s->y_dc_scale;
         else
             block0 = block[0] * s->c_dc_scale;
         /* XXX: only mpeg1 */
         quant_matrix = s->intra_matrix;
-asm volatile(
-		"pcmpeqw %%mm7, %%mm7		\n\t"
-		"psrlw $15, %%mm7		\n\t"
-		"movd %2, %%mm6			\n\t"
-		"packssdw %%mm6, %%mm6		\n\t"
-		"packssdw %%mm6, %%mm6		\n\t"
+__asm__ volatile(
+	"pcmpeqw %%mm7, %%mm7		\n\t"
+	"psrlw $15, %%mm7		\n\t"
+	"movd %2, %%mm6			\n\t"
+	"packssdw %%mm6, %%mm6		\n\t"
+	"packssdw %%mm6, %%mm6		\n\t"
                 "movl %3, %%eax			\n\t"
-		".balign 16\n\t"
-		"1:				\n\t"
-		"movq (%0, %%eax), %%mm0	\n\t"
-		"movq 8(%0, %%eax), %%mm1	\n\t"
-		"movq (%1, %%eax), %%mm4	\n\t"
-		"movq 8(%1, %%eax), %%mm5	\n\t"
-		"pmullw %%mm6, %%mm4		\n\t" // q=qscale*quant_matrix[i]
-		"pmullw %%mm6, %%mm5		\n\t" // q=qscale*quant_matrix[i]
-		"pxor %%mm2, %%mm2		\n\t"
-		"pxor %%mm3, %%mm3		\n\t"
-		"pcmpgtw %%mm0, %%mm2		\n\t" // block[i] < 0 ? -1 : 0
-		"pcmpgtw %%mm1, %%mm3		\n\t" // block[i] < 0 ? -1 : 0
-		"pxor %%mm2, %%mm0		\n\t"
-		"pxor %%mm3, %%mm1		\n\t"
-		"psubw %%mm2, %%mm0		\n\t" // abs(block[i])
-		"psubw %%mm3, %%mm1		\n\t" // abs(block[i])
-		"pmullw %%mm4, %%mm0		\n\t" // abs(block[i])*q
-		"pmullw %%mm5, %%mm1		\n\t" // abs(block[i])*q
-		"pxor %%mm4, %%mm4		\n\t"
-		"pxor %%mm5, %%mm5		\n\t" // FIXME slow
-		"pcmpeqw (%0, %%eax), %%mm4	\n\t" // block[i] == 0 ? -1 : 0
-		"pcmpeqw 8(%0, %%eax), %%mm5	\n\t" // block[i] == 0 ? -1 : 0
-		"psraw $3, %%mm0		\n\t"
-		"psraw $3, %%mm1		\n\t"
-		"psubw %%mm7, %%mm0		\n\t"
-		"psubw %%mm7, %%mm1		\n\t"
-		"por %%mm7, %%mm0		\n\t"
-		"por %%mm7, %%mm1		\n\t"
-		"pxor %%mm2, %%mm0		\n\t"
-		"pxor %%mm3, %%mm1		\n\t"
-		"psubw %%mm2, %%mm0		\n\t"
-		"psubw %%mm3, %%mm1		\n\t"
-		"pandn %%mm0, %%mm4		\n\t"
-		"pandn %%mm1, %%mm5		\n\t"
-		"movq %%mm4, (%0, %%eax)	\n\t"
-		"movq %%mm5, 8(%0, %%eax)	\n\t"
-
-		"addl $16, %%eax		\n\t"
-		"js 1b				\n\t"
-		::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
-		: "%eax", "memory"
-	);    
+	".balign 16\n\t"
+	"1:				\n\t"
+	"movq (%0, %%eax), %%mm0	\n\t"
+	"movq 8(%0, %%eax), %%mm1	\n\t"
+	"movq (%1, %%eax), %%mm4	\n\t"
+	"movq 8(%1, %%eax), %%mm5	\n\t"
+	"pmullw %%mm6, %%mm4		\n\t" /*  q=qscale*quant_matrix[i] */
+	"pmullw %%mm6, %%mm5		\n\t" /*  q=qscale*quant_matrix[i] */
+	"pxor %%mm2, %%mm2		\n\t"
+	"pxor %%mm3, %%mm3		\n\t"
+	"pcmpgtw %%mm0, %%mm2		\n\t" /*  block[i] < 0 ? -1 : 0 */
+	"pcmpgtw %%mm1, %%mm3		\n\t" /*  block[i] < 0 ? -1 : 0 */
+	"pxor %%mm2, %%mm0		\n\t"
+	"pxor %%mm3, %%mm1		\n\t"
+	"psubw %%mm2, %%mm0		\n\t" /*  abs(block[i]) */
+	"psubw %%mm3, %%mm1		\n\t" /*  abs(block[i]) */
+	"pmullw %%mm4, %%mm0		\n\t" /*  abs(block[i])*q */
+	"pmullw %%mm5, %%mm1		\n\t" /*  abs(block[i])*q */
+	"pxor %%mm4, %%mm4		\n\t"
+	"pxor %%mm5, %%mm5		\n\t" /*  FIXME slow */
+	"pcmpeqw (%0, %%eax), %%mm4	\n\t" /*  block[i] == 0 ? -1 : 0 */
+	"pcmpeqw 8(%0, %%eax), %%mm5	\n\t" /*  block[i] == 0 ? -1 : 0 */
+	"psraw $3, %%mm0		\n\t"
+	"psraw $3, %%mm1		\n\t"
+	"psubw %%mm7, %%mm0		\n\t"
+	"psubw %%mm7, %%mm1		\n\t"
+	"por %%mm7, %%mm0		\n\t"
+	"por %%mm7, %%mm1		\n\t"
+	"pxor %%mm2, %%mm0		\n\t"
+	"pxor %%mm3, %%mm1		\n\t"
+	"psubw %%mm2, %%mm0		\n\t"
+	"psubw %%mm3, %%mm1		\n\t"
+	"pandn %%mm0, %%mm4		\n\t"
+	"pandn %%mm1, %%mm5		\n\t"
+	"movq %%mm4, (%0, %%eax)	\n\t"
+	"movq %%mm5, 8(%0, %%eax)	\n\t"
+
+	"addl $16, %%eax		\n\t"
+	"js 1b				\n\t"
+	::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
+	: "%eax", "memory"
+	);
         block[0]= block0;
 
         } else {
         quant_matrix = s->inter_matrix;
-asm volatile(
-		"pcmpeqw %%mm7, %%mm7		\n\t"
-		"psrlw $15, %%mm7		\n\t"
-		"movd %2, %%mm6			\n\t"
-		"packssdw %%mm6, %%mm6		\n\t"
-		"packssdw %%mm6, %%mm6		\n\t"
+__asm__ volatile(
+	"pcmpeqw %%mm7, %%mm7		\n\t"
+	"psrlw $15, %%mm7		\n\t"
+	"movd %2, %%mm6			\n\t"
+	"packssdw %%mm6, %%mm6		\n\t"
+	"packssdw %%mm6, %%mm6		\n\t"
                 "movl %3, %%eax			\n\t"
-		".balign 16\n\t"
-		"1:				\n\t"
-		"movq (%0, %%eax), %%mm0	\n\t"
-		"movq 8(%0, %%eax), %%mm1	\n\t"
-		"movq (%1, %%eax), %%mm4	\n\t"
-		"movq 8(%1, %%eax), %%mm5	\n\t"
-		"pmullw %%mm6, %%mm4		\n\t" // q=qscale*quant_matrix[i]
-		"pmullw %%mm6, %%mm5		\n\t" // q=qscale*quant_matrix[i]
-		"pxor %%mm2, %%mm2		\n\t"
-		"pxor %%mm3, %%mm3		\n\t"
-		"pcmpgtw %%mm0, %%mm2		\n\t" // block[i] < 0 ? -1 : 0
-		"pcmpgtw %%mm1, %%mm3		\n\t" // block[i] < 0 ? -1 : 0
-		"pxor %%mm2, %%mm0		\n\t"
-		"pxor %%mm3, %%mm1		\n\t"
-		"psubw %%mm2, %%mm0		\n\t" // abs(block[i])
-		"psubw %%mm3, %%mm1		\n\t" // abs(block[i])
-		"paddw %%mm0, %%mm0		\n\t" // abs(block[i])*2
-		"paddw %%mm1, %%mm1		\n\t" // abs(block[i])*2
-		"paddw %%mm7, %%mm0		\n\t" // abs(block[i])*2 + 1
-		"paddw %%mm7, %%mm1		\n\t" // abs(block[i])*2 + 1
-		"pmullw %%mm4, %%mm0		\n\t" // (abs(block[i])*2 + 1)*q
-		"pmullw %%mm5, %%mm1		\n\t" // (abs(block[i])*2 + 1)*q
-		"pxor %%mm4, %%mm4		\n\t"
-		"pxor %%mm5, %%mm5		\n\t" // FIXME slow
-		"pcmpeqw (%0, %%eax), %%mm4	\n\t" // block[i] == 0 ? -1 : 0
-		"pcmpeqw 8(%0, %%eax), %%mm5	\n\t" // block[i] == 0 ? -1 : 0
-		"psraw $4, %%mm0		\n\t"
-		"psraw $4, %%mm1		\n\t"
-		"psubw %%mm7, %%mm0		\n\t"
-		"psubw %%mm7, %%mm1		\n\t"
-		"por %%mm7, %%mm0		\n\t"
-		"por %%mm7, %%mm1		\n\t"
-		"pxor %%mm2, %%mm0		\n\t"
-		"pxor %%mm3, %%mm1		\n\t"
-		"psubw %%mm2, %%mm0		\n\t"
-		"psubw %%mm3, %%mm1		\n\t"
-		"pandn %%mm0, %%mm4		\n\t"
-		"pandn %%mm1, %%mm5		\n\t"
-		"movq %%mm4, (%0, %%eax)	\n\t"
-		"movq %%mm5, 8(%0, %%eax)	\n\t"
-
-		"addl $16, %%eax		\n\t"
-		"js 1b				\n\t"
-		::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
-		: "%eax", "memory"
+	".balign 16\n\t"
+	"1:				\n\t"
+	"movq (%0, %%eax), %%mm0	\n\t"
+	"movq 8(%0, %%eax), %%mm1	\n\t"
+	"movq (%1, %%eax), %%mm4	\n\t"
+	"movq 8(%1, %%eax), %%mm5	\n\t"
+	"pmullw %%mm6, %%mm4		\n\t" /*  q=qscale*quant_matrix[i] */
+	"pmullw %%mm6, %%mm5		\n\t" /*  q=qscale*quant_matrix[i] */
+	"pxor %%mm2, %%mm2		\n\t"
+	"pxor %%mm3, %%mm3		\n\t"
+	"pcmpgtw %%mm0, %%mm2		\n\t" /*  block[i] < 0 ? -1 : 0 */
+	"pcmpgtw %%mm1, %%mm3		\n\t" /*  block[i] < 0 ? -1 : 0 */
+	"pxor %%mm2, %%mm0		\n\t"
+	"pxor %%mm3, %%mm1		\n\t"
+	"psubw %%mm2, %%mm0		\n\t" /*  abs(block[i]) */
+	"psubw %%mm3, %%mm1		\n\t" /*  abs(block[i]) */
+	"paddw %%mm0, %%mm0		\n\t" /*  abs(block[i])*2 */
+	"paddw %%mm1, %%mm1		\n\t" /*  abs(block[i])*2 */
+	"paddw %%mm7, %%mm0		\n\t" /*  abs(block[i])*2 + 1 */
+	"paddw %%mm7, %%mm1		\n\t" /*  abs(block[i])*2 + 1 */
+	"pmullw %%mm4, %%mm0		\n\t" /*  (abs(block[i])*2 + 1)*q */
+	"pmullw %%mm5, %%mm1		\n\t" /*  (abs(block[i])*2 + 1)*q */
+	"pxor %%mm4, %%mm4		\n\t"
+	"pxor %%mm5, %%mm5		\n\t" /*  FIXME slow */
+	"pcmpeqw (%0, %%eax), %%mm4	\n\t" /*  block[i] == 0 ? -1 : 0 */
+	"pcmpeqw 8(%0, %%eax), %%mm5	\n\t" /*  block[i] == 0 ? -1 : 0 */
+	"psraw $4, %%mm0		\n\t"
+	"psraw $4, %%mm1		\n\t"
+	"psubw %%mm7, %%mm0		\n\t"
+	"psubw %%mm7, %%mm1		\n\t"
+	"por %%mm7, %%mm0		\n\t"
+	"por %%mm7, %%mm1		\n\t"
+	"pxor %%mm2, %%mm0		\n\t"
+	"pxor %%mm3, %%mm1		\n\t"
+	"psubw %%mm2, %%mm0		\n\t"
+	"psubw %%mm3, %%mm1		\n\t"
+	"pandn %%mm0, %%mm4		\n\t"
+	"pandn %%mm1, %%mm5		\n\t"
+	"movq %%mm4, (%0, %%eax)	\n\t"
+	"movq %%mm5, 8(%0, %%eax)	\n\t"
+
+	"addl $16, %%eax		\n\t"
+	"js 1b				\n\t"
+	::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
+	: "%eax", "memory"
 	);
     }
 }
@@ -319,115 +319,115 @@ static void dct_unquantize_mpeg2_mmx(MpegEncContext *s,
 {
     int nCoeffs;
     const UINT16 *quant_matrix;
-    
+
     if(s->alternate_scan) nCoeffs= 64;
     else nCoeffs= zigzag_end[ s->block_last_index[n] ];
 
     if (s->mb_intra) {
         int block0;
-        if (n < 4) 
+        if (n < 4)
             block0 = block[0] * s->y_dc_scale;
         else
             block0 = block[0] * s->c_dc_scale;
         quant_matrix = s->intra_matrix;
-asm volatile(
-		"pcmpeqw %%mm7, %%mm7		\n\t"
-		"psrlw $15, %%mm7		\n\t"
-		"movd %2, %%mm6			\n\t"
-		"packssdw %%mm6, %%mm6		\n\t"
-		"packssdw %%mm6, %%mm6		\n\t"
+__asm__ volatile(
+	"pcmpeqw %%mm7, %%mm7		\n\t"
+	"psrlw $15, %%mm7		\n\t"
+	"movd %2, %%mm6			\n\t"
+	"packssdw %%mm6, %%mm6		\n\t"
+	"packssdw %%mm6, %%mm6		\n\t"
                 "movl %3, %%eax			\n\t"
-		".balign 16\n\t"
-		"1:				\n\t"
-		"movq (%0, %%eax), %%mm0	\n\t"
-		"movq 8(%0, %%eax), %%mm1	\n\t"
-		"movq (%1, %%eax), %%mm4	\n\t"
-		"movq 8(%1, %%eax), %%mm5	\n\t"
-		"pmullw %%mm6, %%mm4		\n\t" // q=qscale*quant_matrix[i]
-		"pmullw %%mm6, %%mm5		\n\t" // q=qscale*quant_matrix[i]
-		"pxor %%mm2, %%mm2		\n\t"
-		"pxor %%mm3, %%mm3		\n\t"
-		"pcmpgtw %%mm0, %%mm2		\n\t" // block[i] < 0 ? -1 : 0
-		"pcmpgtw %%mm1, %%mm3		\n\t" // block[i] < 0 ? -1 : 0
-		"pxor %%mm2, %%mm0		\n\t"
-		"pxor %%mm3, %%mm1		\n\t"
-		"psubw %%mm2, %%mm0		\n\t" // abs(block[i])
-		"psubw %%mm3, %%mm1		\n\t" // abs(block[i])
-		"pmullw %%mm4, %%mm0		\n\t" // abs(block[i])*q
-		"pmullw %%mm5, %%mm1		\n\t" // abs(block[i])*q
-		"pxor %%mm4, %%mm4		\n\t"
-		"pxor %%mm5, %%mm5		\n\t" // FIXME slow
-		"pcmpeqw (%0, %%eax), %%mm4	\n\t" // block[i] == 0 ? -1 : 0
-		"pcmpeqw 8(%0, %%eax), %%mm5	\n\t" // block[i] == 0 ? -1 : 0
-		"psraw $3, %%mm0		\n\t"
-		"psraw $3, %%mm1		\n\t"
-		"pxor %%mm2, %%mm0		\n\t"
-		"pxor %%mm3, %%mm1		\n\t"
-		"psubw %%mm2, %%mm0		\n\t"
-		"psubw %%mm3, %%mm1		\n\t"
-		"pandn %%mm0, %%mm4		\n\t"
-		"pandn %%mm1, %%mm5		\n\t"
-		"movq %%mm4, (%0, %%eax)	\n\t"
-		"movq %%mm5, 8(%0, %%eax)	\n\t"
-
-		"addl $16, %%eax		\n\t"
-		"js 1b				\n\t"
-		::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
-		: "%eax", "memory"
-	);    
+	".balign 16\n\t"
+	"1:				\n\t"
+	"movq (%0, %%eax), %%mm0	\n\t"
+	"movq 8(%0, %%eax), %%mm1	\n\t"
+	"movq (%1, %%eax), %%mm4	\n\t"
+	"movq 8(%1, %%eax), %%mm5	\n\t"
+	"pmullw %%mm6, %%mm4		\n\t" /*  q=qscale*quant_matrix[i] */
+	"pmullw %%mm6, %%mm5		\n\t" /*  q=qscale*quant_matrix[i] */
+	"pxor %%mm2, %%mm2		\n\t"
+	"pxor %%mm3, %%mm3		\n\t"
+	"pcmpgtw %%mm0, %%mm2		\n\t" /*  block[i] < 0 ? -1 : 0 */
+	"pcmpgtw %%mm1, %%mm3		\n\t" /*  block[i] < 0 ? -1 : 0 */
+	"pxor %%mm2, %%mm0		\n\t"
+	"pxor %%mm3, %%mm1		\n\t"
+	"psubw %%mm2, %%mm0		\n\t" /*  abs(block[i]) */
+	"psubw %%mm3, %%mm1		\n\t" /*  abs(block[i]) */
+	"pmullw %%mm4, %%mm0		\n\t" /*  abs(block[i])*q */
+	"pmullw %%mm5, %%mm1		\n\t" /*  abs(block[i])*q */
+	"pxor %%mm4, %%mm4		\n\t"
+	"pxor %%mm5, %%mm5		\n\t" /*  FIXME slow */
+	"pcmpeqw (%0, %%eax), %%mm4	\n\t" /*  block[i] == 0 ? -1 : 0 */
+	"pcmpeqw 8(%0, %%eax), %%mm5	\n\t" /*  block[i] == 0 ? -1 : 0 */
+	"psraw $3, %%mm0		\n\t"
+	"psraw $3, %%mm1		\n\t"
+	"pxor %%mm2, %%mm0		\n\t"
+	"pxor %%mm3, %%mm1		\n\t"
+	"psubw %%mm2, %%mm0		\n\t"
+	"psubw %%mm3, %%mm1		\n\t"
+	"pandn %%mm0, %%mm4		\n\t"
+	"pandn %%mm1, %%mm5		\n\t"
+	"movq %%mm4, (%0, %%eax)	\n\t"
+	"movq %%mm5, 8(%0, %%eax)	\n\t"
+
+	"addl $16, %%eax		\n\t"
+	"js 1b				\n\t"
+	::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "g" (-2*nCoeffs)
+	: "%eax", "memory"
+	);
         block[0]= block0;
-        //Note, we dont do mismatch control for intra as errors cannot accumulate
+        /* Note, we dont do mismatch control for intra as errors cannot accumulate */
 
     } else {
         quant_matrix = s->inter_matrix;
-asm volatile(
-		"pcmpeqw %%mm7, %%mm7		\n\t"
+__asm__ volatile(
+	"pcmpeqw %%mm7, %%mm7		\n\t"
                 "psrlq $48, %%mm7		\n\t"
-		"movd %2, %%mm6			\n\t"
-		"packssdw %%mm6, %%mm6		\n\t"
-		"packssdw %%mm6, %%mm6		\n\t"
+	"movd %2, %%mm6			\n\t"
+	"packssdw %%mm6, %%mm6		\n\t"
+	"packssdw %%mm6, %%mm6		\n\t"
                 "movl %3, %%eax			\n\t"
-		".balign 16\n\t"
-		"1:				\n\t"
-		"movq (%0, %%eax), %%mm0	\n\t"
-		"movq 8(%0, %%eax), %%mm1	\n\t"
-		"movq (%1, %%eax), %%mm4	\n\t"
-		"movq 8(%1, %%eax), %%mm5	\n\t"
-		"pmullw %%mm6, %%mm4		\n\t" // q=qscale*quant_matrix[i]
-		"pmullw %%mm6, %%mm5		\n\t" // q=qscale*quant_matrix[i]
-		"pxor %%mm2, %%mm2		\n\t"
-		"pxor %%mm3, %%mm3		\n\t"
-		"pcmpgtw %%mm0, %%mm2		\n\t" // block[i] < 0 ? -1 : 0
-		"pcmpgtw %%mm1, %%mm3		\n\t" // block[i] < 0 ? -1 : 0
-		"pxor %%mm2, %%mm0		\n\t"
-		"pxor %%mm3, %%mm1		\n\t"
-		"psubw %%mm2, %%mm0		\n\t" // abs(block[i])
-		"psubw %%mm3, %%mm1		\n\t" // abs(block[i])
-		"paddw %%mm0, %%mm0		\n\t" // abs(block[i])*2
-		"paddw %%mm1, %%mm1		\n\t" // abs(block[i])*2
-		"pmullw %%mm4, %%mm0		\n\t" // abs(block[i])*2*q
-		"pmullw %%mm5, %%mm1		\n\t" // abs(block[i])*2*q
-		"paddw %%mm4, %%mm0		\n\t" // (abs(block[i])*2 + 1)*q
-		"paddw %%mm5, %%mm1		\n\t" // (abs(block[i])*2 + 1)*q
-		"pxor %%mm4, %%mm4		\n\t"
-		"pxor %%mm5, %%mm5		\n\t" // FIXME slow
-		"pcmpeqw (%0, %%eax), %%mm4	\n\t" // block[i] == 0 ? -1 : 0
-		"pcmpeqw 8(%0, %%eax), %%mm5	\n\t" // block[i] == 0 ? -1 : 0
-		"psrlw $4, %%mm0		\n\t"
-		"psrlw $4, %%mm1		\n\t"
-		"pxor %%mm2, %%mm0		\n\t"
-		"pxor %%mm3, %%mm1		\n\t"
-		"psubw %%mm2, %%mm0		\n\t"
-		"psubw %%mm3, %%mm1		\n\t"
-		"pandn %%mm0, %%mm4		\n\t"
-		"pandn %%mm1, %%mm5		\n\t"
+	".balign 16\n\t"
+	"1:				\n\t"
+	"movq (%0, %%eax), %%mm0	\n\t"
+	"movq 8(%0, %%eax), %%mm1	\n\t"
+	"movq (%1, %%eax), %%mm4	\n\t"
+	"movq 8(%1, %%eax), %%mm5	\n\t"
+	"pmullw %%mm6, %%mm4		\n\t" /*  q=qscale*quant_matrix[i] */
+	"pmullw %%mm6, %%mm5		\n\t" /*  q=qscale*quant_matrix[i] */
+	"pxor %%mm2, %%mm2		\n\t"
+	"pxor %%mm3, %%mm3		\n\t"
+	"pcmpgtw %%mm0, %%mm2		\n\t" /*  block[i] < 0 ? -1 : 0 */
+	"pcmpgtw %%mm1, %%mm3		\n\t" /*  block[i] < 0 ? -1 : 0 */
+	"pxor %%mm2, %%mm0		\n\t"
+	"pxor %%mm3, %%mm1		\n\t"
+	"psubw %%mm2, %%mm0		\n\t" /*  abs(block[i]) */
+	"psubw %%mm3, %%mm1		\n\t" /*  abs(block[i]) */
+	"paddw %%mm0, %%mm0		\n\t" /*  abs(block[i])*2 */
+	"paddw %%mm1, %%mm1		\n\t" /*  abs(block[i])*2 */
+	"pmullw %%mm4, %%mm0		\n\t" /*  abs(block[i])*2*q */
+	"pmullw %%mm5, %%mm1		\n\t" /*  abs(block[i])*2*q */
+	"paddw %%mm4, %%mm0		\n\t" /*  (abs(block[i])*2 + 1)*q */
+	"paddw %%mm5, %%mm1		\n\t" /*  (abs(block[i])*2 + 1)*q */
+	"pxor %%mm4, %%mm4		\n\t"
+	"pxor %%mm5, %%mm5		\n\t" /*  FIXME slow */
+	"pcmpeqw (%0, %%eax), %%mm4	\n\t" /*  block[i] == 0 ? -1 : 0 */
+	"pcmpeqw 8(%0, %%eax), %%mm5	\n\t" /*  block[i] == 0 ? -1 : 0 */
+	"psrlw $4, %%mm0		\n\t"
+	"psrlw $4, %%mm1		\n\t"
+	"pxor %%mm2, %%mm0		\n\t"
+	"pxor %%mm3, %%mm1		\n\t"
+	"psubw %%mm2, %%mm0		\n\t"
+	"psubw %%mm3, %%mm1		\n\t"
+	"pandn %%mm0, %%mm4		\n\t"
+	"pandn %%mm1, %%mm5		\n\t"
                 "pxor %%mm4, %%mm7		\n\t"
                 "pxor %%mm5, %%mm7		\n\t"
-		"movq %%mm4, (%0, %%eax)	\n\t"
-		"movq %%mm5, 8(%0, %%eax)	\n\t"
+	"movq %%mm4, (%0, %%eax)	\n\t"
+	"movq %%mm5, 8(%0, %%eax)	\n\t"
 
-		"addl $16, %%eax		\n\t"
-		"js 1b				\n\t"
+	"addl $16, %%eax		\n\t"
+	"js 1b				\n\t"
                 "movd 124(%0, %3), %%mm0	\n\t"
                 "movq %%mm7, %%mm6		\n\t"
                 "psrlq $32, %%mm7		\n\t"
@@ -439,14 +439,14 @@ asm volatile(
                 "psrlq $15, %%mm7		\n\t"
                 "pxor %%mm7, %%mm0		\n\t"
                 "movd %%mm0, 124(%0, %3)	\n\t"
-                
-		::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "r" (-2*nCoeffs)
-		: "%eax", "memory"
+
+	::"r" (block+nCoeffs), "r"(quant_matrix+nCoeffs), "g" (qscale), "r" (-2*nCoeffs)
+	: "%eax", "memory"
 	);
     }
 }
 
-/* draw the edges of width 'w' of an image of size width, height 
+/* draw the edges of width 'w' of an image of size width, height
    this mmx version can only handle w==8 || w==16 */
 static void draw_edges_mmx(UINT8 *buf, int wrap, int width, int height, int w)
 {
@@ -458,78 +458,78 @@ static void draw_edges_mmx(UINT8 *buf, int wrap, int width, int height, int w)
     ptr = buf;
     if(w==8)
     {
-	asm volatile(
-		"1:				\n\t"
-		"movd (%0), %%mm0		\n\t"
-		"punpcklbw %%mm0, %%mm0		\n\t" 
-		"punpcklwd %%mm0, %%mm0		\n\t"
-		"punpckldq %%mm0, %%mm0		\n\t"
-		"movq %%mm0, -8(%0)		\n\t"
-		"movq -8(%0, %2), %%mm1		\n\t"
-		"punpckhbw %%mm1, %%mm1		\n\t"
-		"punpckhwd %%mm1, %%mm1		\n\t"
-		"punpckhdq %%mm1, %%mm1		\n\t"
-		"movq %%mm1, (%0, %2)		\n\t"
-		"addl %1, %0			\n\t"
-		"cmpl %3, %0			\n\t"
-		" jb 1b				\n\t"
-		: "+r" (ptr)
-		: "r" (wrap), "r" (width), "r" (ptr + wrap*height)
+	__asm__ volatile(
+	"1:				\n\t"
+	"movd (%0), %%mm0		\n\t"
+	"punpcklbw %%mm0, %%mm0		\n\t"
+	"punpcklwd %%mm0, %%mm0		\n\t"
+	"punpckldq %%mm0, %%mm0		\n\t"
+	"movq %%mm0, -8(%0)		\n\t"
+	"movq -8(%0, %2), %%mm1		\n\t"
+	"punpckhbw %%mm1, %%mm1		\n\t"
+	"punpckhwd %%mm1, %%mm1		\n\t"
+	"punpckhdq %%mm1, %%mm1		\n\t"
+	"movq %%mm1, (%0, %2)		\n\t"
+	"addl %1, %0			\n\t"
+	"cmpl %3, %0			\n\t"
+	" jb 1b				\n\t"
+	: "+r" (ptr)
+	: "r" (wrap), "r" (width), "r" (ptr + wrap*height)
 	);
     }
     else
     {
-	asm volatile(
-		"1:				\n\t"
-		"movd (%0), %%mm0		\n\t"
-		"punpcklbw %%mm0, %%mm0		\n\t" 
-		"punpcklwd %%mm0, %%mm0		\n\t"
-		"punpckldq %%mm0, %%mm0		\n\t"
-		"movq %%mm0, -8(%0)		\n\t"
-		"movq %%mm0, -16(%0)		\n\t"
-		"movq -8(%0, %2), %%mm1		\n\t"
-		"punpckhbw %%mm1, %%mm1		\n\t"
-		"punpckhwd %%mm1, %%mm1		\n\t"
-		"punpckhdq %%mm1, %%mm1		\n\t"
-		"movq %%mm1, (%0, %2)		\n\t"
-		"movq %%mm1, 8(%0, %2)		\n\t"
-		"addl %1, %0			\n\t"
-		"cmpl %3, %0			\n\t"
-		" jb 1b				\n\t"		
-		: "+r" (ptr)
-		: "r" (wrap), "r" (width), "r" (ptr + wrap*height)
+	__asm__ volatile(
+	"1:				\n\t"
+	"movd (%0), %%mm0		\n\t"
+	"punpcklbw %%mm0, %%mm0		\n\t"
+	"punpcklwd %%mm0, %%mm0		\n\t"
+	"punpckldq %%mm0, %%mm0		\n\t"
+	"movq %%mm0, -8(%0)		\n\t"
+	"movq %%mm0, -16(%0)		\n\t"
+	"movq -8(%0, %2), %%mm1		\n\t"
+	"punpckhbw %%mm1, %%mm1		\n\t"
+	"punpckhwd %%mm1, %%mm1		\n\t"
+	"punpckhdq %%mm1, %%mm1		\n\t"
+	"movq %%mm1, (%0, %2)		\n\t"
+	"movq %%mm1, 8(%0, %2)		\n\t"
+	"addl %1, %0			\n\t"
+	"cmpl %3, %0			\n\t"
+	" jb 1b				\n\t"
+	: "+r" (ptr)
+	: "r" (wrap), "r" (width), "r" (ptr + wrap*height)
 	);
     }
-    
+
     for(i=0;i<w;i+=4) {
         /* top and bottom (and hopefully also the corners) */
 	ptr= buf - (i + 1) * wrap - w;
-	asm volatile(
-		"1:				\n\t"
-		"movq (%1, %0), %%mm0		\n\t"
-		"movq %%mm0, (%0)		\n\t"
-		"movq %%mm0, (%0, %2)		\n\t"
-		"movq %%mm0, (%0, %2, 2)	\n\t"
-		"movq %%mm0, (%0, %3)		\n\t"
-		"addl $8, %0			\n\t"
-		"cmpl %4, %0			\n\t"
-		" jb 1b				\n\t"
-		: "+r" (ptr)
-		: "r" ((int)buf - (int)ptr - w), "r" (-wrap), "r" (-wrap*3), "r" (ptr+width+2*w)
+	__asm__ volatile(
+	"1:				\n\t"
+	"movq (%1, %0), %%mm0		\n\t"
+	"movq %%mm0, (%0)		\n\t"
+	"movq %%mm0, (%0, %2)		\n\t"
+	"movq %%mm0, (%0, %2, 2)	\n\t"
+	"movq %%mm0, (%0, %3)		\n\t"
+	"addl $8, %0			\n\t"
+	"cmpl %4, %0			\n\t"
+	" jb 1b				\n\t"
+	: "+r" (ptr)
+	: "r" ((int)buf - (int)ptr - w), "r" (-wrap), "r" (-wrap*3), "r" (ptr+width+2*w)
 	);
 	ptr= last_line + (i + 1) * wrap - w;
-	asm volatile(
-		"1:				\n\t"
-		"movq (%1, %0), %%mm0		\n\t"
-		"movq %%mm0, (%0)		\n\t"
-		"movq %%mm0, (%0, %2)		\n\t"
-		"movq %%mm0, (%0, %2, 2)	\n\t"
-		"movq %%mm0, (%0, %3)		\n\t"
-		"addl $8, %0			\n\t"
-		"cmpl %4, %0			\n\t"
-		" jb 1b				\n\t"
-		: "+r" (ptr)
-		: "r" ((int)last_line - (int)ptr - w), "r" (wrap), "r" (wrap*3), "r" (ptr+width+2*w)
+	__asm__ volatile(
+	"1:				\n\t"
+	"movq (%1, %0), %%mm0		\n\t"
+	"movq %%mm0, (%0)		\n\t"
+	"movq %%mm0, (%0, %2)		\n\t"
+	"movq %%mm0, (%0, %2, 2)	\n\t"
+	"movq %%mm0, (%0, %3)		\n\t"
+	"addl $8, %0			\n\t"
+	"cmpl %4, %0			\n\t"
+	" jb 1b				\n\t"
+	: "+r" (ptr)
+	: "r" ((int)last_line - (int)ptr - w), "r" (wrap), "r" (wrap*3), "r" (ptr+width+2*w)
 	);
     }
 }