Merge commit '5b4dfbff'

* commit '5b4dfbff':
  x86: ABS1: port to cpuflags
  v210x: cosmetics, reformat
Merged-by: Michael Niedermayer <michaelni@gmx.at>

libavcodec/v210x.c

@@ -26,12 +26,12 @@
 
 static av_cold int decode_init(AVCodecContext *avctx)
 {
-    if(avctx->width & 1){
+    if (avctx->width & 1) {
         av_log(avctx, AV_LOG_ERROR, "v210x needs even width\n");
         return AVERROR(EINVAL);
     }
     avctx->pix_fmt             = AV_PIX_FMT_YUV422P16;
-    avctx->bits_per_raw_sample= 10;
+    avctx->bits_per_raw_sample = 10;
 
     avctx->coded_frame= avcodec_alloc_frame();
     if (!avctx->coded_frame)
@@ -43,82 +43,82 @@ static av_cold int decode_init(AVCodecContext *avctx)
 static int decode_frame(AVCodecContext *avctx, void *data, int *got_frame,
                         AVPacket *avpkt)
 {
-    int y=0;
-    int width= avctx->width;
-    AVFrame *pic= avctx->coded_frame;
-    const uint32_t *src= (const uint32_t *)avpkt->data;
+    const uint32_t *src = (const uint32_t *)avpkt->data;
+    AVFrame *pic        = avctx->coded_frame;
+    int width           = avctx->width;
+    int y               = 0;
     uint16_t *ydst, *udst, *vdst, *yend;
     int ret;
 
-    if(pic->data[0])
+    if (pic->data[0])
         avctx->release_buffer(avctx, pic);
 
-    if(avpkt->size < avctx->width * avctx->height * 8 / 3){
+    if (avpkt->size < avctx->width * avctx->height * 8 / 3) {
         av_log(avctx, AV_LOG_ERROR, "Packet too small\n");
         return AVERROR_INVALIDDATA;
     }
 
-    if(avpkt->size > avctx->width * avctx->height * 8 / 3){
+    if (avpkt->size > avctx->width * avctx->height * 8 / 3) {
         av_log_ask_for_sample(avctx, "Probably padded data\n");
     }
 
-    pic->reference= 0;
+    pic->reference = 0;
     if ((ret = ff_get_buffer(avctx, pic)) < 0)
         return ret;
 
-    ydst= (uint16_t *)pic->data[0];
-    udst= (uint16_t *)pic->data[1];
-    vdst= (uint16_t *)pic->data[2];
-    yend= ydst + width;
-    pic->pict_type= AV_PICTURE_TYPE_I;
-    pic->key_frame= 1;
-
-    for(;;){
-        uint32_t v= av_be2ne32(*src++);
-        *udst++= (v>>16) & 0xFFC0;
-        *ydst++= (v>>6 ) & 0xFFC0;
-        *vdst++= (v<<4 ) & 0xFFC0;
-
-        v= av_be2ne32(*src++);
-        *ydst++= (v>>16) & 0xFFC0;
-
-        if(ydst >= yend){
-            ydst+= pic->linesize[0]/2 - width;
-            udst+= pic->linesize[1]/2 - width/2;
-            vdst+= pic->linesize[2]/2 - width/2;
-            yend= ydst + width;
-            if(++y >= avctx->height)
+    ydst = (uint16_t *)pic->data[0];
+    udst = (uint16_t *)pic->data[1];
+    vdst = (uint16_t *)pic->data[2];
+    yend = ydst + width;
+    pic->pict_type = AV_PICTURE_TYPE_I;
+    pic->key_frame = 1;
+
+    for (;;) {
+        uint32_t v = av_be2ne32(*src++);
+        *udst++ = (v >> 16) & 0xFFC0;
+        *ydst++ = (v >> 6 ) & 0xFFC0;
+        *vdst++ = (v << 4 ) & 0xFFC0;
+
+        v       = av_be2ne32(*src++);
+        *ydst++ = (v >> 16) & 0xFFC0;
+
+        if (ydst >= yend) {
+            ydst += pic->linesize[0] / 2 - width;
+            udst += pic->linesize[1] / 2 - width / 2;
+            vdst += pic->linesize[2] / 2 - width / 2;
+            yend = ydst + width;
+            if (++y >= avctx->height)
                 break;
         }
 
-        *udst++= (v>>6 ) & 0xFFC0;
-        *ydst++= (v<<4 ) & 0xFFC0;
+        *udst++ = (v >> 6 ) & 0xFFC0;
+        *ydst++ = (v << 4 ) & 0xFFC0;
 
-        v= av_be2ne32(*src++);
-        *vdst++= (v>>16) & 0xFFC0;
-        *ydst++= (v>>6 ) & 0xFFC0;
+        v = av_be2ne32(*src++);
+        *vdst++ = (v >> 16) & 0xFFC0;
+        *ydst++ = (v >> 6 ) & 0xFFC0;
 
-        if(ydst >= yend){
-            ydst+= pic->linesize[0]/2 - width;
-            udst+= pic->linesize[1]/2 - width/2;
-            vdst+= pic->linesize[2]/2 - width/2;
-            yend= ydst + width;
-            if(++y >= avctx->height)
+        if (ydst >= yend) {
+            ydst += pic->linesize[0] / 2 - width;
+            udst += pic->linesize[1] / 2 - width / 2;
+            vdst += pic->linesize[2] / 2 - width / 2;
+            yend  = ydst + width;
+            if (++y >= avctx->height)
                 break;
         }
 
-        *udst++= (v<<4 ) & 0xFFC0;
-
-        v= av_be2ne32(*src++);
-        *ydst++= (v>>16) & 0xFFC0;
-        *vdst++= (v>>6 ) & 0xFFC0;
-        *ydst++= (v<<4 ) & 0xFFC0;
-        if(ydst >= yend){
-            ydst+= pic->linesize[0]/2 - width;
-            udst+= pic->linesize[1]/2 - width/2;
-            vdst+= pic->linesize[2]/2 - width/2;
-            yend= ydst + width;
-            if(++y >= avctx->height)
+        *udst++ = (v << 4 ) & 0xFFC0;
+
+        v = av_be2ne32(*src++);
+        *ydst++ = (v >> 16) & 0xFFC0;
+        *vdst++ = (v >> 6 ) & 0xFFC0;
+        *ydst++ = (v << 4 ) & 0xFFC0;
+        if (ydst >= yend) {
+            ydst += pic->linesize[0] / 2 - width;
+            udst += pic->linesize[1] / 2 - width / 2;
+            vdst += pic->linesize[2] / 2 - width / 2;
+            yend  = ydst + width;
+            if (++y >= avctx->height)
                 break;
         }
     }

libavcodec/x86/dsputilenc.asm

@@ -257,11 +257,9 @@ hadamard8_16_wrapper 0, 14
 %endmacro
 
 INIT_MMX mmx
-%define ABS1 ABS1_MMX
 HADAMARD8_DIFF
 
 INIT_MMX mmxext
-%define ABS1 ABS1_MMXEXT
 HADAMARD8_DIFF
 
 INIT_XMM sse2

libavutil/x86/x86util.asm

@@ -172,11 +172,19 @@
     psignw     %1, %2
 %endmacro
 
-%macro ABS1_MMX 2    ; a, tmp
+%macro ABS1 2
+%if cpuflag(ssse3)
+    pabsw   %1, %1
+%elif cpuflag(mmxext) ; a, tmp
+    pxor    %2, %2
+    psubw   %2, %1
+    pmaxsw  %1, %2
+%else ; a, tmp
     pxor       %2, %2
     pcmpgtw    %2, %1
     pxor       %1, %2
     psubw      %1, %2
+%endif
 %endmacro
 
 %macro ABS2_MMX 4    ; a, b, tmp0, tmp1
@@ -190,12 +198,6 @@
     psubw      %2, %4
 %endmacro
 
-%macro ABS1_MMXEXT 2 ; a, tmp
-    pxor    %2, %2
-    psubw   %2, %1
-    pmaxsw  %1, %2
-%endmacro
-
 %macro ABS2_MMXEXT 4 ; a, b, tmp0, tmp1
     pxor    %3, %3
     pxor    %4, %4
@@ -205,10 +207,6 @@
     pmaxsw  %2, %4
 %endmacro
 
-%macro ABS1_SSSE3 2
-    pabsw   %1, %1
-%endmacro
-
 %macro ABS2_SSSE3 4
     pabsw   %1, %1
     pabsw   %2, %2
@@ -254,7 +252,6 @@
     ABS2 %3, %4, %5, %6
 %endmacro
 
-%define ABS1 ABS1_MMX
 %define ABS2 ABS2_MMX
 %define ABSB ABSB_MMX
 %define ABSB2 ABSB2_MMX