Removes trailing spaces. Removes tabs.

[vlc] / modules / video_chroma / i422_yuy2.c

diff --git a/modules/video_chroma/i422_yuy2.c b/modules/video_chroma/i422_yuy2.c
index 2119755a2899d2835fec05889336e84459b6acc1..ec1a90305da92b31fd5e811cb81131f2d05dc0fd 100644 (file)
--- a/modules/video_chroma/i422_yuy2.c
+++ b/modules/video_chroma/i422_yuy2.c
@@ -11,7 +11,7 @@
  * it under the terms of the GNU General Public License as published by
  * the Free Software Foundation; either version 2 of the License, or
  * (at your option) any later version.
- * 
+ *
  * This program is distributed in the hope that it will be useful,
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
@@ -25,8 +25,6 @@
 /*****************************************************************************
  * Preamble
  *****************************************************************************/
-#include <string.h>                                            /* strerror() */
-#include <stdlib.h>                                      /* malloc(), free() */
 
 #include <vlc/vlc.h>
 #include <vlc_vout.h>
@@ -70,7 +68,7 @@ vlc_module_begin();
 #elif defined (MODULE_NAME_IS_i422_yuy2_sse2)
     set_description( _("SSE2 conversions from " SRC_FOURCC " to " DEST_FOURCC) );
     set_capability( "chroma", 120 );
-    add_requirement( MMX );
+    add_requirement( SSE2 );
 #endif
     set_callbacks( Activate, NULL );
 vlc_module_end();
@@ -135,8 +133,7 @@ static int Activate( vlc_object_t *p_this )
         default:
             return -1;
     }
-    
-    return 0; 
+    return 0;
 }
 
 /* Following functions are local */
@@ -164,7 +161,7 @@ static void I422_YUY2( vout_thread_t *p_vout, picture_t *p_source,
 #if defined (MODULE_NAME_IS_i422_yuy2_sse2)
 
     if( 0 == (15 & (p_source->p[Y_PLANE].i_pitch|p_dest->p->i_pitch|
-        ((int)p_line|(int)p_y))) )
+        ((intptr_t)p_line|(intptr_t)p_y))) )
     {
         /* use faster SSE2 aligned fetch and store */
         for( i_y = p_vout->render.i_height ; i_y-- ; )
@@ -257,7 +254,7 @@ static void I422_YVYU( vout_thread_t *p_vout, picture_t *p_source,
 #if defined (MODULE_NAME_IS_i422_yuy2_sse2)
 
     if( 0 == (15 & (p_source->p[Y_PLANE].i_pitch|p_dest->p->i_pitch|
-        ((int)p_line|(int)p_y))) )
+        ((intptr_t)p_line|(intptr_t)p_y))) )
     {
         /* use faster SSE2 aligned fetch and store */
         for( i_y = p_vout->render.i_height ; i_y-- ; )
@@ -350,7 +347,7 @@ static void I422_UYVY( vout_thread_t *p_vout, picture_t *p_source,
 #if defined (MODULE_NAME_IS_i422_yuy2_sse2)
 
     if( 0 == (15 & (p_source->p[Y_PLANE].i_pitch|p_dest->p->i_pitch|
-        ((int)p_line|(int)p_y))) )
+        ((intptr_t)p_line|(intptr_t)p_y))) )
     {
         /* use faster SSE2 aligned fetch and store */
         for( i_y = p_vout->render.i_height ; i_y-- ; )
@@ -443,6 +440,61 @@ static void I422_cyuv( vout_thread_t *p_vout, picture_t *p_source,
 
     int i_x, i_y;
 
+    const int i_source_margin = p_source->p[0].i_pitch
+                                 - p_source->p[0].i_visible_pitch;
+    const int i_source_margin_c = p_source->p[1].i_pitch
+                                 - p_source->p[1].i_visible_pitch;
+    const int i_dest_margin = p_dest->p->i_pitch
+                               - p_dest->p->i_visible_pitch;
+
+#if defined (MODULE_NAME_IS_i422_yuy2_sse2)
+
+    if( 0 == (15 & (p_source->p[Y_PLANE].i_pitch|p_dest->p->i_pitch|
+        ((intptr_t)p_line|(intptr_t)p_y))) )
+    {
+        /* use faster SSE2 aligned fetch and store */
+        for( i_y = p_vout->render.i_height ; i_y-- ; )
+        {
+            p_line -= 2 * p_dest->p->i_pitch;
+
+            for( i_x = p_vout->render.i_width / 16 ; i_x-- ; )
+            {
+                SSE2_CALL( SSE2_YUV422_UYVY_ALIGNED );
+            }
+            for( i_x = ( p_vout->render.i_width % 16 ) / 2; i_x-- ; )
+            {
+                C_YUV422_UYVY( p_line, p_y, p_u, p_v );
+            }
+            p_y += i_source_margin;
+            p_u += i_source_margin_c;
+            p_v += i_source_margin_c;
+            p_line += i_dest_margin;
+        }
+    }
+    else {
+        /* use slower SSE2 unaligned fetch and store */
+        for( i_y = p_vout->render.i_height ; i_y-- ; )
+        {
+            p_line -= 2 * p_dest->p->i_pitch;
+
+            for( i_x = p_vout->render.i_width / 16 ; i_x-- ; )
+            {
+                SSE2_CALL( SSE2_YUV422_UYVY_UNALIGNED );
+            }
+            for( i_x = ( p_vout->render.i_width % 16 ) / 2; i_x-- ; )
+            {
+                C_YUV422_UYVY( p_line, p_y, p_u, p_v );
+            }
+            p_y += i_source_margin;
+            p_u += i_source_margin_c;
+            p_v += i_source_margin_c;
+            p_line += i_dest_margin;
+        }
+    }
+    SSE2_END;
+
+#else
+
     for( i_y = p_vout->render.i_height ; i_y-- ; )
     {
         for( i_x = p_vout->render.i_width / 8 ; i_x-- ; )
@@ -458,12 +510,18 @@ static void I422_cyuv( vout_thread_t *p_vout, picture_t *p_source,
             MMX_CALL( MMX_YUV422_UYVY );
 #endif
         }
+        p_y += i_source_margin;
+        p_u += i_source_margin_c;
+        p_v += i_source_margin_c;
+        p_line += i_dest_margin;
     }
 #if defined (MODULE_NAME_IS_i422_yuy2_mmx)
     MMX_END;
 #elif defined (MODULE_NAME_IS_i422_yuy2_sse2)
     SSE2_END;
 #endif
+
+#endif
 }
 
 /*****************************************************************************