From: Wan-Teh Chang <wtc@google.com>
Date: Thu, 28 Sep 2023 16:26:58 +0000 (-0700)
Subject: Use vpx_sse instead of vpx_mse to compute SSE
X-Git-Tag: accepted/tizen/7.0/unified/20240521.012539~1^2~14
X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=0d5811e4effb62672ec60aed5bd21553af158791;p=platform%2Fupstream%2Flibvpx.git

Use vpx_sse instead of vpx_mse to compute SSE

Use vpx_sse and vpx_highbd_sse instead of vpx_mse16x16 and
vpx_highbd_8_mse16x16 respectively to compute SSE for PSNR
calculations. This solves an issue whereby vpx_highbd_8_mse16x16
was being used to calculate SSE for 10- and 12-bit input.

This is a port of the libaom CL
https://aomedia-review.googlesource.com/c/aom/+/175063
by Jonathan Wright <jonathan.wright@arm.com>.

Bug: webm:1819
Change-Id: I37e3ac72835e67ccb44ac89a4ed16df62c2169a7
(cherry picked from commit 7dfe343199381bddddc5eaa648e947876979b61b)
---

diff --git a/vpx_dsp/psnr.c b/vpx_dsp/psnr.c
index f0d4e92..4ee4130 100644
--- a/vpx_dsp/psnr.c
+++ b/vpx_dsp/psnr.c
@@ -45,14 +45,14 @@ static int64_t encoder_sse(const uint8_t *a, int a_stride, const uint8_t *b,
 }
 
 #if CONFIG_VP9_HIGHBITDEPTH
-static int64_t encoder_highbd_8_sse(const uint8_t *a8, int a_stride,
-                                    const uint8_t *b8, int b_stride, int w,
-                                    int h) {
+static int64_t encoder_highbd_sse(const uint8_t *a8, int a_stride,
+                                  const uint8_t *b8, int b_stride, int w,
+                                  int h) {
   int i, j;
   int64_t sse = 0;
 
-  uint16_t *a = CONVERT_TO_SHORTPTR(a8);
-  uint16_t *b = CONVERT_TO_SHORTPTR(b8);
+  const uint16_t *a = CONVERT_TO_SHORTPTR(a8);
+  const uint16_t *b = CONVERT_TO_SHORTPTR(b8);
 
   for (i = 0; i < h; i++) {
     for (j = 0; j < w; j++) {
@@ -88,10 +88,8 @@ static int64_t get_sse(const uint8_t *a, int a_stride, const uint8_t *b,
   for (y = 0; y < height / 16; ++y) {
     const uint8_t *pa = a;
     const uint8_t *pb = b;
-    unsigned int sse;
     for (x = 0; x < width / 16; ++x) {
-      vpx_mse16x16(pa, a_stride, pb, b_stride, &sse);
-      total_sse += sse;
+      total_sse += vpx_sse(pa, a_stride, pb, b_stride, 16, 16);
 
       pa += 16;
       pb += 16;
@@ -131,21 +129,19 @@ static int64_t highbd_get_sse(const uint8_t *a, int a_stride, const uint8_t *b,
   const int dw = width % 16;
   const int dh = height % 16;
   if (dw > 0) {
-    total_sse += encoder_highbd_8_sse(&a[width - dw], a_stride, &b[width - dw],
-                                      b_stride, dw, height);
+    total_sse += encoder_highbd_sse(&a[width - dw], a_stride, &b[width - dw],
+                                    b_stride, dw, height);
   }
   if (dh > 0) {
-    total_sse += encoder_highbd_8_sse(&a[(height - dh) * a_stride], a_stride,
-                                      &b[(height - dh) * b_stride], b_stride,
-                                      width - dw, dh);
+    total_sse += encoder_highbd_sse(&a[(height - dh) * a_stride], a_stride,
+                                    &b[(height - dh) * b_stride], b_stride,
+                                    width - dw, dh);
   }
   for (y = 0; y < height / 16; ++y) {
     const uint8_t *pa = a;
     const uint8_t *pb = b;
-    unsigned int sse;
     for (x = 0; x < width / 16; ++x) {
-      vpx_highbd_8_mse16x16(pa, a_stride, pb, b_stride, &sse);
-      total_sse += sse;
+      total_sse += vpx_highbd_sse(pa, a_stride, pb, b_stride, 16, 16);
       pa += 16;
       pb += 16;
     }