1 /****************************************************************************
3 ** Copyright (C) 2011 Nokia Corporation and/or its subsidiary(-ies).
4 ** All rights reserved.
5 ** Contact: Nokia Corporation (qt-info@nokia.com)
7 ** This file is part of the QtGui module of the Qt Toolkit.
9 ** $QT_BEGIN_LICENSE:LGPL$
10 ** GNU Lesser General Public License Usage
11 ** This file may be used under the terms of the GNU Lesser General Public
12 ** License version 2.1 as published by the Free Software Foundation and
13 ** appearing in the file LICENSE.LGPL included in the packaging of this
14 ** file. Please review the following information to ensure the GNU Lesser
15 ** General Public License version 2.1 requirements will be met:
16 ** http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html.
18 ** In addition, as a special exception, Nokia gives you certain additional
19 ** rights. These rights are described in the Nokia Qt LGPL Exception
20 ** version 1.1, included in the file LGPL_EXCEPTION.txt in this package.
22 ** GNU General Public License Usage
23 ** Alternatively, this file may be used under the terms of the GNU General
24 ** Public License version 3.0 as published by the Free Software Foundation
25 ** and appearing in the file LICENSE.GPL included in the packaging of this
26 ** file. Please review the following information to ensure the GNU General
27 ** Public License version 3.0 requirements will be met:
28 ** http://www.gnu.org/copyleft/gpl.html.
31 ** Alternatively, this file may be used in accordance with the terms and
32 ** conditions contained in a signed written agreement between you and Nokia.
40 ****************************************************************************/
42 #ifndef QDRAWHELPER_SSE_P_H
43 #define QDRAWHELPER_SSE_P_H
49 // This file is not part of the Qt API. It exists purely as an
50 // implementation detail. This header file may change from version to
51 // version without notice, or even be removed.
56 #include <private/qdrawhelper_mmx_p.h>
61 // this is an evil hack - the posix_memalign declaration in LSB
62 // is wrong - see http://bugs.linuxbase.org/show_bug.cgi?id=2431
63 # define posix_memalign _lsb_hack_posix_memalign
64 # include <xmmintrin.h>
65 # undef posix_memalign
67 # include <xmmintrin.h>
73 #define _MM_SHUFFLE(fp3,fp2,fp1,fp0) \
74 (((fp3) << 6) | ((fp2) << 4) | ((fp1) << 2) | (fp0))
77 struct QSSEIntrinsics : public QMMXIntrinsics
79 static inline m64 alpha(m64 x) {
80 return _mm_shuffle_pi16 (x, _MM_SHUFFLE(3, 3, 3, 3));
83 static inline m64 _load_alpha(uint x, const m64 &mmx_0x0000) {
84 m64 t = _mm_unpacklo_pi8(_mm_cvtsi32_si64(x), mmx_0x0000);
85 return _mm_shuffle_pi16 (t, _MM_SHUFFLE(0, 0, 0, 0));
90 inline void qt_memfill32_sse_template(quint32 *dest, quint32 value, int count)
94 case 6: *dest++ = value;
95 case 5: *dest++ = value;
96 case 4: *dest++ = value;
97 case 3: *dest++ = value;
98 case 2: *dest++ = value;
99 case 1: *dest = value;
104 __m64 *dst64 = reinterpret_cast<__m64*>(dest);
105 const __m64 value64 = _mm_set_pi32(value, value);
106 int count64 = count / 2;
108 int n = (count64 + 3) / 4;
109 switch (count64 & 0x3) {
110 case 0: do { _mm_stream_pi(dst64++, value64);
111 case 3: _mm_stream_pi(dst64++, value64);
112 case 2: _mm_stream_pi(dst64++, value64);
113 case 1: _mm_stream_pi(dst64++, value64);
118 dest[count - 1] = value;
124 inline void qt_bitmapblit16_sse_template(QRasterBuffer *rasterBuffer,
128 int width, int height, int stride)
130 const quint16 c = qt_colorConvert<quint16, quint32>(color, 0);
131 quint16 *dest = reinterpret_cast<quint16*>(rasterBuffer->scanLine(y)) + x;
132 const int destStride = rasterBuffer->bytesPerLine() / sizeof(quint16);
134 const __m64 c64 = _mm_set1_pi16(c);
136 # pragma warning(disable: 4309) // truncation of constant value
138 const __m64 maskmask1 = _mm_set_pi16(0x1010, 0x2020, 0x4040, 0x8080);
139 const __m64 maskadd1 = _mm_set_pi16(0x7070, 0x6060, 0x4040, 0x0000);
142 const __m64 maskmask2 = _mm_set_pi16(0x0101, 0x0202, 0x0404, 0x0808);
143 const __m64 maskadd2 = _mm_set_pi16(0x7f7f, 0x7e7e, 0x7c7c, 0x7878);
146 for (int x = 0; x < width; x += 8) {
147 const quint8 s = src[x >> 3];
150 __m64 mask1 = _mm_set1_pi8(s);
152 mask1 = _m_pand(mask1, maskmask1);
153 mask1 = _mm_add_pi16(mask1, maskadd1);
154 _mm_maskmove_si64(c64, mask1, (char*)(dest + x));
155 mask2 = _m_pand(mask2, maskmask2);
156 mask2 = _mm_add_pi16(mask2, maskadd2);
157 _mm_maskmove_si64(c64, mask2, (char*)(dest + x + 4));
164 const quint8 s = *src;
166 __m64 mask1 = _mm_set1_pi8(s);
167 mask1 = _m_pand(mask1, maskmask1);
168 mask1 = _mm_add_pi16(mask1, maskadd1);
169 _mm_maskmove_si64(c64, mask1, (char*)(dest));
181 #endif // QT_HAVE_SSE
182 #endif // QDRAWHELPER_SSE_P_H