diff options
author | Allan Sandfeld Jensen <allan.jensen@qt.io> | 2016-12-20 16:17:58 +0100 |
---|---|---|
committer | Allan Sandfeld Jensen <allan.jensen@qt.io> | 2017-01-31 00:14:11 +0000 |
commit | 85468f7bccb276c2be5801481a6ce10f07581cdb (patch) | |
tree | f96ef309303ed0caf91b0c37cabee4d295cb19d3 /src/gui/painting/qdrawhelper_avx2.cpp | |
parent | ad4f7b59ead6c4eb17e787bce25a7211b866063f (diff) |
Manually vectorize ARGB32toARGB32PM for SSE4.1 and NEON
Manually vectorizing is significantly faster because we can optimize
for common cases like long stretches of opaque or transparent pixels.
This is both smaller and faster than the auto-vectorized version, it is
also much faster than the autovectorized version for AVX2 which then can
be removed.
Change-Id: I0fa80ce273a8387cc6cd084879822ad9bade385c
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Diffstat (limited to 'src/gui/painting/qdrawhelper_avx2.cpp')
-rw-r--r-- | src/gui/painting/qdrawhelper_avx2.cpp | 13 |
1 files changed, 0 insertions, 13 deletions
diff --git a/src/gui/painting/qdrawhelper_avx2.cpp b/src/gui/painting/qdrawhelper_avx2.cpp index 9c1335298e..5e17e8abec 100644 --- a/src/gui/painting/qdrawhelper_avx2.cpp +++ b/src/gui/painting/qdrawhelper_avx2.cpp @@ -44,19 +44,6 @@ QT_BEGIN_NAMESPACE -// Autovectorized premultiply functions: -const uint *QT_FASTCALL convertARGB32ToARGB32PM_avx2(uint *buffer, const uint *src, int count, - const QVector<QRgb> *, QDitherInfo *) -{ - return qt_convertARGB32ToARGB32PM(buffer, src, count); -} - -const uint *QT_FASTCALL convertRGBA8888ToARGB32PM_avx2(uint *buffer, const uint *src, int count, - const QVector<QRgb> *, QDitherInfo *) -{ - return qt_convertRGBA8888ToARGB32PM(buffer, src, count); -} - // Vectorized blend functions: // See BYTE_MUL_SSE2 for details. |