Fix blending of RGB32 on RGB32 with partial opacity

The alpha channel of an RGB32 image was not properly ignored when doing
blending with partial opacity.

Now the alpha value is properly ignored, which is both more correct
and faster. This also makes SSE2 and AVX2 implementations match NEON
which was already doing the right thing (though had dead code for
doing it wrong).

Change-Id: I4613b8d70ed8c2e36ced10baaa7a4a55bd36a940
Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>

1 files changed, 3 insertions, 5 deletions

diff --git a/src/gui/painting/qdrawhelper_avx2.cpp b/src/gui/painting/qdrawhelper_avx2.cpp
index acc9bc7ba1..9c1335298e 100644
--- a/src/gui/painting/qdrawhelper_avx2.cpp
+++ b/src/gui/painting/qdrawhelper_avx2.cpp
@@ -267,11 +267,9 @@ void qt_blend_rgb32_on_rgb32_avx2(uchar *destPixels, int dbpl,
         // 2) interpolate pixels with AVX2
         for (; x < (w - 7); x += 8) {
             const __m256i srcVector = _mm256_lddqu_si256((const __m256i *)&src[x]);
-            if (!_mm256_testz_si256(srcVector, srcVector)) {
-                __m256i dstVector = _mm256_load_si256((__m256i *)&dst[x]);
-                INTERPOLATE_PIXEL_255_AVX2(srcVector, dstVector, constAlphaVector, oneMinusConstAlpha, colorMask, half);
-                _mm256_store_si256((__m256i *)&dst[x], dstVector);
-            }
+            __m256i dstVector = _mm256_load_si256((__m256i *)&dst[x]);
+            INTERPOLATE_PIXEL_255_AVX2(srcVector, dstVector, constAlphaVector, oneMinusConstAlpha, colorMask, half);
+            _mm256_store_si256((__m256i *)&dst[x], dstVector);
         }
 
         // 3) Epilogue