Revert "Remove Wextra warnings from vp9_sad.c"

This reverts commit 7ab9a9587b96db4edce6be916c1f02297a9555ff

Nightly test http://build.webmproject.org/jenkins/view/libvpx-nightly-tests/job/libvpx%20unit%20tests%20(valgrind-2)/arch=x86_64-linux-gcc,filter=-*VP8*:*Large.*/276/console

Failed 

This patch did not address all the assembly issues 
some of the vp8 assembly counts on 5 arguments being passed in to this function:   

one example : vp8_sad8x16_wmt

Please address or split this into vp9 and vp8 patches.

Change-Id: I78afcc171649894f887bb8ee3c66de24aaddc7ca

5 files changed, 132 insertions, 150 deletions

diff --git a/vp8/common/rtcd_defs.pl b/vp8/common/rtcd_defs.pl
index 88296f902..dcf5b8e26 100644
--- a/vp8/common/rtcd_defs.pl
+++ b/vp8/common/rtcd_defs.pl
@@ -316,23 +316,23 @@ $vp8_variance_halfpixvar16x16_hv_neon_asm=vp8_variance_halfpixvar16x16_hv_neon;
 #
 # Single block SAD
 #
-add_proto qw/unsigned int vp8_sad4x4/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride";
+add_proto qw/unsigned int vp8_sad4x4/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad";
 specialize qw/vp8_sad4x4 mmx sse2 neon/;
 $vp8_sad4x4_sse2=vp8_sad4x4_wmt;
 
-add_proto qw/unsigned int vp8_sad8x8/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride";
+add_proto qw/unsigned int vp8_sad8x8/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad";
 specialize qw/vp8_sad8x8 mmx sse2 neon/;
 $vp8_sad8x8_sse2=vp8_sad8x8_wmt;
 
-add_proto qw/unsigned int vp8_sad8x16/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride";
+add_proto qw/unsigned int vp8_sad8x16/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad";
 specialize qw/vp8_sad8x16 mmx sse2 neon/;
 $vp8_sad8x16_sse2=vp8_sad8x16_wmt;
 
-add_proto qw/unsigned int vp8_sad16x8/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride";
+add_proto qw/unsigned int vp8_sad16x8/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad";
 specialize qw/vp8_sad16x8 mmx sse2 neon/;
 $vp8_sad16x8_sse2=vp8_sad16x8_wmt;
 
-add_proto qw/unsigned int vp8_sad16x16/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride";
+add_proto qw/unsigned int vp8_sad16x16/, "const unsigned char *src_ptr, int src_stride, const unsigned char *ref_ptr, int ref_stride, unsigned int max_sad";
 specialize qw/vp8_sad16x16 mmx sse2 sse3 media neon/;
 $vp8_sad16x16_sse2=vp8_sad16x16_wmt;
 $vp8_sad16x16_media=vp8_sad16x16_armv6;
diff --git a/vp8/common/sad_c.c b/vp8/common/sad_c.c
index 65872028c..5f36fc96e 100644
--- a/vp8/common/sad_c.c
+++ b/vp8/common/sad_c.c
@@ -16,7 +16,7 @@
 
 static unsigned int sad_mx_n_c(const unsigned char *src_ptr, int src_stride,
                                const unsigned char *ref_ptr, int ref_stride,
-                               int m, int n)
+                               unsigned int max_sad, int m, int n)
 {
     int r, c;
     unsigned int sad = 0;
@@ -28,6 +28,9 @@ static unsigned int sad_mx_n_c(const unsigned char *src_ptr, int src_stride,
             sad += abs(src_ptr[c] - ref_ptr[c]);
         }
 
+        if (sad > max_sad)
+          break;
+
         src_ptr += src_stride;
         ref_ptr += ref_stride;
     }
@@ -40,199 +43,204 @@ static unsigned int sad_mx_n_c(const unsigned char *src_ptr, int src_stride,
  */
 
 unsigned int vp8_sad16x16_c(const unsigned char *src_ptr, int src_stride,
-                            const unsigned char *ref_ptr, int ref_stride)
+                            const unsigned char *ref_ptr, int ref_stride,
+                            unsigned int max_sad)
 {
-    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 16, 16);
+    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 16, 16);
 }
 
 unsigned int vp8_sad8x8_c(const unsigned char *src_ptr, int src_stride,
-                          const unsigned char *ref_ptr, int ref_stride)
+                          const unsigned char *ref_ptr, int ref_stride,
+                          unsigned int max_sad)
 {
-    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 8, 8);
+    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 8, 8);
 }
 
 unsigned int vp8_sad16x8_c(const unsigned char *src_ptr, int src_stride,
-                           const unsigned char *ref_ptr, int ref_stride)
+                           const unsigned char *ref_ptr, int ref_stride,
+                           unsigned int max_sad)
 {
-    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 16, 8);
+    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 16, 8);
 
 }
 
 unsigned int vp8_sad8x16_c(const unsigned char *src_ptr, int src_stride,
-                           const unsigned char *ref_ptr, int ref_stride)
+                           const unsigned char *ref_ptr, int ref_stride,
+                           unsigned int max_sad)
 {
-    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 8, 16);
+    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 8, 16);
 }
 
 unsigned int vp8_sad4x4_c(const unsigned char *src_ptr, int src_stride,
-                          const unsigned char *ref_ptr, int ref_stride)
+                          const unsigned char *ref_ptr, int ref_stride,
+                          unsigned int max_sad)
 {
-    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, 4, 4);
+    return sad_mx_n_c(src_ptr, src_stride, ref_ptr, ref_stride, max_sad, 4, 4);
 }
 
 void vp8_sad16x16x3_c(const unsigned char *src_ptr, int src_stride,
                       const unsigned char *ref_ptr, int ref_stride,
                       unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
+    sad_array[0] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
 }
 
 void vp8_sad16x16x8_c(const unsigned char *src_ptr, int src_stride,
                       const unsigned char *ref_ptr, int ref_stride,
                       unsigned short *sad_array)
 {
-    sad_array[0] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
-    sad_array[3] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 3, ref_stride);
-    sad_array[4] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 4, ref_stride);
-    sad_array[5] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 5, ref_stride);
-    sad_array[6] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 6, ref_stride);
-    sad_array[7] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 7, ref_stride);
+    sad_array[0] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
+    sad_array[3] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
+    sad_array[4] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
+    sad_array[5] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
+    sad_array[6] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
+    sad_array[7] = (unsigned short)vp8_sad16x16_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
 }
 
 void vp8_sad16x8x3_c(const unsigned char *src_ptr, int src_stride,
                      const unsigned char *ref_ptr, int ref_stride,
                      unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
+    sad_array[0] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
 }
 
 void vp8_sad16x8x8_c(const unsigned char *src_ptr, int src_stride,
                      const unsigned char *ref_ptr, int ref_stride,
                      unsigned short *sad_array)
 {
-    sad_array[0] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
-    sad_array[3] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 3, ref_stride);
-    sad_array[4] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 4, ref_stride);
-    sad_array[5] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 5, ref_stride);
-    sad_array[6] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 6, ref_stride);
-    sad_array[7] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 7, ref_stride);
+    sad_array[0] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
+    sad_array[3] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
+    sad_array[4] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
+    sad_array[5] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
+    sad_array[6] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
+    sad_array[7] = (unsigned short)vp8_sad16x8_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
 }
 
 void vp8_sad8x8x3_c(const unsigned char *src_ptr, int src_stride,
                     const unsigned char *ref_ptr, int ref_stride,
                     unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
+    sad_array[0] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
 }
 
 void vp8_sad8x8x8_c(const unsigned char *src_ptr, int src_stride,
                     const unsigned char *ref_ptr, int ref_stride,
                     unsigned short *sad_array)
 {
-    sad_array[0] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
-    sad_array[3] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 3, ref_stride);
-    sad_array[4] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 4, ref_stride);
-    sad_array[5] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 5, ref_stride);
-    sad_array[6] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 6, ref_stride);
-    sad_array[7] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 7, ref_stride);
+    sad_array[0] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
+    sad_array[3] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
+    sad_array[4] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
+    sad_array[5] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
+    sad_array[6] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
+    sad_array[7] = (unsigned short)vp8_sad8x8_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
 }
 
 void vp8_sad8x16x3_c(const unsigned char *src_ptr, int src_stride,
                      const unsigned char *ref_ptr, int ref_stride,
                      unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
+    sad_array[0] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
 }
 
 void vp8_sad8x16x8_c(const unsigned char *src_ptr, int src_stride,
                      const unsigned char *ref_ptr, int ref_stride,
                      unsigned short *sad_array)
 {
-    sad_array[0] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
-    sad_array[3] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 3, ref_stride);
-    sad_array[4] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 4, ref_stride);
-    sad_array[5] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 5, ref_stride);
-    sad_array[6] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 6, ref_stride);
-    sad_array[7] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 7, ref_stride);
+    sad_array[0] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
+    sad_array[3] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
+    sad_array[4] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
+    sad_array[5] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
+    sad_array[6] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
+    sad_array[7] = (unsigned short)vp8_sad8x16_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
 }
 
 void vp8_sad4x4x3_c(const unsigned char *src_ptr, int src_stride,
                     const unsigned char *ref_ptr, int ref_stride,
                     unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
+    sad_array[0] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
 }
 
 void vp8_sad4x4x8_c(const unsigned char *src_ptr, int src_stride,
                     const unsigned char *ref_ptr, int ref_stride,
                     unsigned short *sad_array)
 {
-    sad_array[0] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 0, ref_stride);
-    sad_array[1] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 1, ref_stride);
-    sad_array[2] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 2, ref_stride);
-    sad_array[3] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 3, ref_stride);
-    sad_array[4] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 4, ref_stride);
-    sad_array[5] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 5, ref_stride);
-    sad_array[6] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 6, ref_stride);
-    sad_array[7] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 7, ref_stride);
+    sad_array[0] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 0, ref_stride, UINT_MAX);
+    sad_array[1] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 1, ref_stride, UINT_MAX);
+    sad_array[2] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 2, ref_stride, UINT_MAX);
+    sad_array[3] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 3, ref_stride, UINT_MAX);
+    sad_array[4] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 4, ref_stride, UINT_MAX);
+    sad_array[5] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 5, ref_stride, UINT_MAX);
+    sad_array[6] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 6, ref_stride, UINT_MAX);
+    sad_array[7] = (unsigned short)vp8_sad4x4_c(src_ptr, src_stride, ref_ptr + 7, ref_stride, UINT_MAX);
 }
 
 void vp8_sad16x16x4d_c(const unsigned char *src_ptr, int src_stride,
                        const unsigned char * const ref_ptr[], int ref_stride,
                        unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[0], ref_stride);
-    sad_array[1] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[1], ref_stride);
-    sad_array[2] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[2], ref_stride);
-    sad_array[3] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[3], ref_stride);
+    sad_array[0] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
+    sad_array[3] = vp8_sad16x16_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
 }
 
 void vp8_sad16x8x4d_c(const unsigned char *src_ptr, int src_stride,
                       const unsigned char * const ref_ptr[], int ref_stride,
                       unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[0], ref_stride);
-    sad_array[1] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[1], ref_stride);
-    sad_array[2] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[2], ref_stride);
-    sad_array[3] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[3], ref_stride);
+    sad_array[0] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
+    sad_array[3] = vp8_sad16x8_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
 }
 
 void vp8_sad8x8x4d_c(const unsigned char *src_ptr, int src_stride,
                      const unsigned char * const ref_ptr[], int ref_stride,
                      unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[0], ref_stride);
-    sad_array[1] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[1], ref_stride);
-    sad_array[2] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[2], ref_stride);
-    sad_array[3] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[3], ref_stride);
+    sad_array[0] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
+    sad_array[3] = vp8_sad8x8_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
 }
 
 void vp8_sad8x16x4d_c(const unsigned char *src_ptr, int src_stride,
                       const unsigned char * const ref_ptr[], int ref_stride,
                       unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[0], ref_stride);
-    sad_array[1] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[1], ref_stride);
-    sad_array[2] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[2], ref_stride);
-    sad_array[3] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[3], ref_stride);
+    sad_array[0] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
+    sad_array[3] = vp8_sad8x16_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
 }
 
 void vp8_sad4x4x4d_c(const unsigned char *src_ptr, int src_stride,
                      const unsigned char * const ref_ptr[], int  ref_stride,
                      unsigned int *sad_array)
 {
-    sad_array[0] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[0], ref_stride);
-    sad_array[1] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[1], ref_stride);
-    sad_array[2] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[2], ref_stride);
-    sad_array[3] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[3], ref_stride);
+    sad_array[0] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[0], ref_stride, UINT_MAX);
+    sad_array[1] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[1], ref_stride, UINT_MAX);
+    sad_array[2] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[2], ref_stride, UINT_MAX);
+    sad_array[3] = vp8_sad4x4_c(src_ptr, src_stride, ref_ptr[3], ref_stride, UINT_MAX);
 }
 
 /* Copy 2 macroblocks to a buffer */
diff --git a/vp8/common/variance.h b/vp8/common/variance.h
index 765c19a22..89a32a722 100644
--- a/vp8/common/variance.h
+++ b/vp8/common/variance.h
@@ -22,7 +22,8 @@ typedef unsigned int(*vp8_sad_fn_t)(
     const unsigned char *src_ptr,
     int source_stride,
     const unsigned char *ref_ptr,
-    int ref_stride);
+    int ref_stride,
+    unsigned int max_sad);
 
 typedef void (*vp8_copy32xn_fn_t)(
     const unsigned char *src_ptr,
diff --git a/vp8/encoder/mcomp.c b/vp8/encoder/mcomp.c
index e7a1d7616..54abe76ac 100644
--- a/vp8/encoder/mcomp.c
+++ b/vp8/encoder/mcomp.c
@@ -898,7 +898,7 @@ int vp8_hex_search
     this_offset = base_offset + (br * (pre_stride)) + bc;
     this_mv.as_mv.row = br;
     this_mv.as_mv.col = bc;
-    bestsad = vfp->sdf(what, what_stride, this_offset, in_what_stride)
+    bestsad = vfp->sdf(what, what_stride, this_offset, in_what_stride, UINT_MAX)
             + mvsad_err_cost(&this_mv, &fcenter_mv, mvsadcost, sad_per_bit);
 
 #if CONFIG_MULTI_RES_ENCODING
@@ -923,7 +923,7 @@ int vp8_hex_search
             this_mv.as_mv.row = br + hex[i].row;
             this_mv.as_mv.col = bc + hex[i].col;
             this_offset = base_offset + (this_mv.as_mv.row * in_what_stride) + this_mv.as_mv.col;
-            thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride);
+            thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
             CHECK_BETTER
         }
     }else
@@ -934,7 +934,7 @@ int vp8_hex_search
             this_mv.as_mv.col = bc + hex[i].col;
             CHECK_POINT
             this_offset = base_offset + (this_mv.as_mv.row * in_what_stride) + this_mv.as_mv.col;
-            thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride);
+            thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
             CHECK_BETTER
         }
     }
@@ -960,8 +960,7 @@ int vp8_hex_search
                 this_mv.as_mv.row = br + next_chkpts[k][i].row;
                 this_mv.as_mv.col = bc + next_chkpts[k][i].col;
                 this_offset = base_offset + (this_mv.as_mv.row * (in_what_stride)) + this_mv.as_mv.col;
-                thissad = vfp->sdf(what, what_stride, this_offset,
-                                   in_what_stride);
+                thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
                 CHECK_BETTER
             }
         }else
@@ -972,8 +971,7 @@ int vp8_hex_search
                 this_mv.as_mv.col = bc + next_chkpts[k][i].col;
                 CHECK_POINT
                 this_offset = base_offset + (this_mv.as_mv.row * (in_what_stride)) + this_mv.as_mv.col;
-                thissad = vfp->sdf(what, what_stride, this_offset,
-                                   in_what_stride);
+                thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
                 CHECK_BETTER
             }
         }
@@ -1004,8 +1002,7 @@ cal_neighbors:
                 this_mv.as_mv.row = br + neighbors[i].row;
                 this_mv.as_mv.col = bc + neighbors[i].col;
                 this_offset = base_offset + (this_mv.as_mv.row * (in_what_stride)) + this_mv.as_mv.col;
-                thissad = vfp->sdf(what, what_stride, this_offset,
-                                   in_what_stride);
+                thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
                 CHECK_BETTER
             }
         }else
@@ -1016,8 +1013,7 @@ cal_neighbors:
                 this_mv.as_mv.col = bc + neighbors[i].col;
                 CHECK_POINT
                 this_offset = base_offset + (this_mv.as_mv.row * (in_what_stride)) + this_mv.as_mv.col;
-                thissad = vfp->sdf(what, what_stride, this_offset,
-                                   in_what_stride);
+                thissad = vfp->sdf(what, what_stride, this_offset, in_what_stride, bestsad);
                 CHECK_BETTER
             }
         }
@@ -1101,7 +1097,7 @@ int vp8_diamond_search_sad_c
     best_address = in_what;
 
     /* Check the starting position */
-    bestsad = fn_ptr->sdf(what, what_stride, in_what, in_what_stride)
+    bestsad = fn_ptr->sdf(what, what_stride, in_what, in_what_stride, UINT_MAX)
             + mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
 
     /* search_param determines the length of the initial step and hence
@@ -1126,8 +1122,7 @@ int vp8_diamond_search_sad_c
 
             {
                 check_here = ss[i].offset + best_address;
-                thissad = fn_ptr->sdf(what, what_stride, check_here,
-                                      in_what_stride);
+                thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride, bestsad);
 
                 if (thissad < bestsad)
                 {
@@ -1226,7 +1221,7 @@ int vp8_diamond_search_sadx4
     best_address = in_what;
 
     /* Check the starting position */
-    bestsad = fn_ptr->sdf(what, what_stride, in_what, in_what_stride)
+    bestsad = fn_ptr->sdf(what, what_stride, in_what, in_what_stride, UINT_MAX)
             + mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
 
     /* search_param determines the length of the initial step and hence the
@@ -1294,8 +1289,7 @@ int vp8_diamond_search_sadx4
                 (this_row_offset > x->mv_row_min) && (this_row_offset < x->mv_row_max))
                 {
                     check_here = ss[i].offset + best_address;
-                    thissad = fn_ptr->sdf(what, what_stride, check_here,
-                                          in_what_stride);
+                    thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride, bestsad);
 
                     if (thissad < bestsad)
                     {
@@ -1379,7 +1373,7 @@ int vp8_full_search_sad_c(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
 
     /* Baseline value at the centre */
     bestsad = fn_ptr->sdf(what, what_stride, bestaddress,
-                          in_what_stride)
+                          in_what_stride, UINT_MAX)
             + mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
 
     /* Apply further limits to prevent us looking using vectors that
@@ -1404,8 +1398,7 @@ int vp8_full_search_sad_c(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
 
         for (c = col_min; c < col_max; c++)
         {
-            thissad = fn_ptr->sdf(what, what_stride, check_here,
-                                  in_what_stride);
+            thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride, bestsad);
 
             this_mv.as_mv.col = c;
             thissad += mvsad_err_cost(&this_mv, &fcenter_mv,
@@ -1478,7 +1471,7 @@ int vp8_full_search_sadx3(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
 
     /* Baseline value at the centre */
     bestsad = fn_ptr->sdf(what, what_stride, bestaddress,
-                          in_what_stride)
+                          in_what_stride, UINT_MAX)
             + mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
 
     /* Apply further limits to prevent us looking using vectors that stretch
@@ -1534,8 +1527,7 @@ int vp8_full_search_sadx3(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
 
         while (c < col_max)
         {
-            thissad = fn_ptr->sdf(what, what_stride, check_here,
-                                  in_what_stride);
+            thissad = fn_ptr->sdf(what, what_stride, check_here, in_what_stride, bestsad);
 
             if (thissad < bestsad)
             {
@@ -1614,7 +1606,7 @@ int vp8_full_search_sadx8(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
 
     /* Baseline value at the centre */
     bestsad = fn_ptr->sdf(what, what_stride,
-                          bestaddress, in_what_stride)
+                          bestaddress, in_what_stride, UINT_MAX)
             + mvsad_err_cost(best_mv, &fcenter_mv, mvsadcost, sad_per_bit);
 
     /* Apply further limits to prevent us looking using vectors that stretch
@@ -1700,8 +1692,7 @@ int vp8_full_search_sadx8(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv,
 
         while (c < col_max)
         {
-            thissad = fn_ptr->sdf(what, what_stride, check_here,
-                                  in_what_stride);
+            thissad = fn_ptr->sdf(what, what_stride, check_here , in_what_stride, bestsad);
 
             if (thissad < bestsad)
             {
@@ -1760,7 +1751,7 @@ int vp8_refining_search_sad_c(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv
     fcenter_mv.as_mv.col = center_mv->as_mv.col >> 3;
 
     bestsad = fn_ptr->sdf(what, what_stride, best_address,
-                          in_what_stride)
+                          in_what_stride, UINT_MAX)
             + mvsad_err_cost(ref_mv, &fcenter_mv, mvsadcost, error_per_bit);
 
     for (i=0; i<search_range; i++)
@@ -1776,8 +1767,7 @@ int vp8_refining_search_sad_c(MACROBLOCK *x, BLOCK *b, BLOCKD *d, int_mv *ref_mv
             (this_row_offset > x->mv_row_min) && (this_row_offset < x->mv_row_max))
             {
                 check_here = (neighbors[j].row)*in_what_stride + neighbors[j].col + best_address;
-                thissad = fn_ptr->sdf(what, what_stride, check_here,
-                                      in_what_stride);
+                thissad = fn_ptr->sdf(what, what_stride, check_here , in_what_stride, bestsad);
 
                 if (thissad < bestsad)
                 {
@@ -1841,7 +1831,7 @@ int vp8_refining_search_sadx4(MACROBLOCK *x, BLOCK *b, BLOCKD *d,
     fcenter_mv.as_mv.col = center_mv->as_mv.col >> 3;
 
     bestsad = fn_ptr->sdf(what, what_stride, best_address,
-                          in_what_stride)
+                          in_what_stride, UINT_MAX)
             + mvsad_err_cost(ref_mv, &fcenter_mv, mvsadcost, error_per_bit);
 
     for (i=0; i<search_range; i++)
@@ -1892,8 +1882,7 @@ int vp8_refining_search_sadx4(MACROBLOCK *x, BLOCK *b, BLOCKD *d,
                 (this_row_offset > x->mv_row_min) && (this_row_offset < x->mv_row_max))
                 {
                     check_here = (neighbors[j].row)*in_what_stride + neighbors[j].col + best_address;
-                    thissad = fn_ptr->sdf(what, what_stride, check_here,
-                                          in_what_stride);
+                    thissad = fn_ptr->sdf(what, what_stride, check_here , in_what_stride, bestsad);
 
                     if (thissad < bestsad)
                     {
diff --git a/vp8/encoder/rdopt.c b/vp8/encoder/rdopt.c
index b27348374..387701c57 100644
--- a/vp8/encoder/rdopt.c
+++ b/vp8/encoder/rdopt.c
@@ -1686,25 +1686,16 @@ void vp8_cal_sad(VP8_COMP *cpi, MACROBLOCKD *xd, MACROBLOCK *x, int recon_yoffse
     }else if(xd->mb_to_top_edge==0)
     {   /* only has left MB for sad calculation. */
         near_sad[0] = near_sad[2] = INT_MAX;
-        near_sad[1] = cpi->fn_ptr[BLOCK_16X16].sdf(
-            src_y_ptr, b->src_stride, xd->dst.y_buffer - 16, xd->dst.y_stride);
+        near_sad[1] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - 16,xd->dst.y_stride, UINT_MAX);
     }else if(xd->mb_to_left_edge ==0)
     {   /* only has left MB for sad calculation. */
         near_sad[1] = near_sad[2] = INT_MAX;
-        near_sad[0] = cpi->fn_ptr[BLOCK_16X16].sdf(
-            src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16,
-            xd->dst.y_stride);
+        near_sad[0] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16,xd->dst.y_stride, UINT_MAX);
     }else
     {
-        near_sad[0] = cpi->fn_ptr[BLOCK_16X16].sdf(
-            src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16,
-            xd->dst.y_stride);
-        near_sad[1] = cpi->fn_ptr[BLOCK_16X16].sdf(
-            src_y_ptr, b->src_stride, xd->dst.y_buffer - 16, xd->dst.y_stride);
-        near_sad[2] = cpi->fn_ptr[BLOCK_16X16].sdf(
-            src_y_ptr,
-            b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16 -16,
-            xd->dst.y_stride);
+        near_sad[0] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16,xd->dst.y_stride, UINT_MAX);
+        near_sad[1] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - 16,xd->dst.y_stride, UINT_MAX);
+        near_sad[2] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, xd->dst.y_buffer - xd->dst.y_stride *16 -16,xd->dst.y_stride, UINT_MAX);
     }
 
     if(cpi->common.last_frame_type != KEY_FRAME)
@@ -1719,21 +1710,14 @@ void vp8_cal_sad(VP8_COMP *cpi, MACROBLOCKD *xd, MACROBLOCK *x, int recon_yoffse
         if(xd->mb_to_bottom_edge==0) near_sad[7] = INT_MAX;
 
         if(near_sad[4] != INT_MAX)
-            near_sad[4] = cpi->fn_ptr[BLOCK_16X16].sdf(
-                src_y_ptr, b->src_stride, pre_y_buffer - pre_y_stride *16,
-                pre_y_stride);
+            near_sad[4] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer - pre_y_stride *16, pre_y_stride, UINT_MAX);
         if(near_sad[5] != INT_MAX)
-            near_sad[5] = cpi->fn_ptr[BLOCK_16X16].sdf(
-                src_y_ptr, b->src_stride, pre_y_buffer - 16, pre_y_stride);
-        near_sad[3] = cpi->fn_ptr[BLOCK_16X16].sdf(
-            src_y_ptr, b->src_stride, pre_y_buffer, pre_y_stride);
+            near_sad[5] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer - 16, pre_y_stride, UINT_MAX);
+        near_sad[3] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer, pre_y_stride, UINT_MAX);
         if(near_sad[6] != INT_MAX)
-            near_sad[6] = cpi->fn_ptr[BLOCK_16X16].sdf(
-                src_y_ptr, b->src_stride, pre_y_buffer + 16, pre_y_stride);
+            near_sad[6] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer + 16, pre_y_stride, UINT_MAX);
         if(near_sad[7] != INT_MAX)
-            near_sad[7] = cpi->fn_ptr[BLOCK_16X16].sdf(
-                src_y_ptr, b->src_stride, pre_y_buffer + pre_y_stride *16,
-                pre_y_stride);
+            near_sad[7] = cpi->fn_ptr[BLOCK_16X16].sdf(src_y_ptr, b->src_stride, pre_y_buffer + pre_y_stride *16, pre_y_stride, UINT_MAX);
     }
 
     if(cpi->common.last_frame_type != KEY_FRAME)