shithub: libvpx

Download patch

ref: e4e08556db4f964bc44f3c4fdb3af8898dcb8522
parent: 6ae8f8dbe80b4b8dce1ac010e34d7f2ee0ade55d
author: Johann <[email protected]>
date: Thu Jul 6 05:16:02 EDT 2017 

sad neon: avg for 64x[32,64]

BUG=webm:1425

Change-Id: Id84d97807a6a0fbcc889c4dfe11929d54f85493d


--- a/test/sad_test.cc
+++ b/test/sad_test.cc
@@ -658,6 +658,8 @@
 INSTANTIATE_TEST_CASE_P(NEON, SADTest, ::testing::ValuesIn(neon_tests));
 
 const SadMxNAvgParam avg_neon_tests[] = {
+  SadMxNAvgParam(64, 64, &vpx_sad64x64_avg_neon),
+  SadMxNAvgParam(64, 32, &vpx_sad64x32_avg_neon),
   SadMxNAvgParam(32, 64, &vpx_sad32x64_avg_neon),
   SadMxNAvgParam(32, 32, &vpx_sad32x32_avg_neon),
   SadMxNAvgParam(32, 16, &vpx_sad32x16_avg_neon),
--- a/vpx_dsp/arm/sad_neon.c
+++ b/vpx_dsp/arm/sad_neon.c
@@ -297,10 +297,61 @@
   }
 }
 
+static INLINE uint32x4_t sad64x_avg(const uint8_t *a, int a_stride,
+                                    const uint8_t *b, int b_stride,
+                                    const uint8_t *c, const int height) {
+  int i;
+  uint16x8_t abs_0 = vdupq_n_u16(0);
+  uint16x8_t abs_1 = vdupq_n_u16(0);
+
+  for (i = 0; i < height; ++i) {
+    const uint8x16_t a_0 = vld1q_u8(a);
+    const uint8x16_t a_1 = vld1q_u8(a + 16);
+    const uint8x16_t a_2 = vld1q_u8(a + 32);
+    const uint8x16_t a_3 = vld1q_u8(a + 48);
+    const uint8x16_t b_0 = vld1q_u8(b);
+    const uint8x16_t b_1 = vld1q_u8(b + 16);
+    const uint8x16_t b_2 = vld1q_u8(b + 32);
+    const uint8x16_t b_3 = vld1q_u8(b + 48);
+    const uint8x16_t c_0 = vld1q_u8(c);
+    const uint8x16_t c_1 = vld1q_u8(c + 16);
+    const uint8x16_t c_2 = vld1q_u8(c + 32);
+    const uint8x16_t c_3 = vld1q_u8(c + 48);
+    const uint8x16_t avg_0 = vrhaddq_u8(b_0, c_0);
+    const uint8x16_t avg_1 = vrhaddq_u8(b_1, c_1);
+    const uint8x16_t avg_2 = vrhaddq_u8(b_2, c_2);
+    const uint8x16_t avg_3 = vrhaddq_u8(b_3, c_3);
+    a += a_stride;
+    b += b_stride;
+    c += 64;
+    abs_0 = vabal_u8(abs_0, vget_low_u8(a_0), vget_low_u8(avg_0));
+    abs_0 = vabal_u8(abs_0, vget_high_u8(a_0), vget_high_u8(avg_0));
+    abs_0 = vabal_u8(abs_0, vget_low_u8(a_1), vget_low_u8(avg_1));
+    abs_0 = vabal_u8(abs_0, vget_high_u8(a_1), vget_high_u8(avg_1));
+    abs_1 = vabal_u8(abs_1, vget_low_u8(a_2), vget_low_u8(avg_2));
+    abs_1 = vabal_u8(abs_1, vget_high_u8(a_2), vget_high_u8(avg_2));
+    abs_1 = vabal_u8(abs_1, vget_low_u8(a_3), vget_low_u8(avg_3));
+    abs_1 = vabal_u8(abs_1, vget_high_u8(a_3), vget_high_u8(avg_3));
+  }
+
+  {
+    const uint32x4_t sum = vpaddlq_u16(abs_0);
+    return vpadalq_u16(sum, abs_1);
+  }
+}
+
 #define sad64xN(n)                                                      \
   uint32_t vpx_sad64x##n##_neon(const uint8_t *src, int src_stride,     \
                                 const uint8_t *ref, int ref_stride) {   \
     const uint32x4_t abs = sad64x(src, src_stride, ref, ref_stride, n); \
+    return horizontal_add_32x4(abs);                                    \
+  }                                                                     \
+                                                                        \
+  uint32_t vpx_sad64x##n##_avg_neon(const uint8_t *src, int src_stride, \
+                                    const uint8_t *ref, int ref_stride, \
+                                    const uint8_t *second_pred) {       \
+    const uint32x4_t abs =                                              \
+        sad64x_avg(src, src_stride, ref, ref_stride, second_pred, n);   \
     return horizontal_add_32x4(abs);                                    \
   }
 
--- a/vpx_dsp/vpx_dsp_rtcd_defs.pl
+++ b/vpx_dsp/vpx_dsp_rtcd_defs.pl
@@ -780,10 +780,10 @@
 }  # CONFIG_VP9_ENCODER
 
 add_proto qw/unsigned int vpx_sad64x64_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
-specialize qw/vpx_sad64x64_avg avx2 msa sse2 vsx/;
+specialize qw/vpx_sad64x64_avg neon avx2 msa sse2 vsx/;
 
 add_proto qw/unsigned int vpx_sad64x32_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
-specialize qw/vpx_sad64x32_avg avx2 msa sse2 vsx/;
+specialize qw/vpx_sad64x32_avg neon avx2 msa sse2 vsx/;
 
 add_proto qw/unsigned int vpx_sad32x64_avg/, "const uint8_t *src_ptr, int src_stride, const uint8_t *ref_ptr, int ref_stride, const uint8_t *second_pred";
 specialize qw/vpx_sad32x64_avg neon avx2 msa sse2 vsx/;