1 | /**************************************************************************** |
2 | ** |
3 | ** Copyright (C) 2016 The Qt Company Ltd. |
4 | ** Contact: https://www.qt.io/licensing/ |
5 | ** |
6 | ** This file is part of the QtGui module of the Qt Toolkit. |
7 | ** |
8 | ** $QT_BEGIN_LICENSE:LGPL$ |
9 | ** Commercial License Usage |
10 | ** Licensees holding valid commercial Qt licenses may use this file in |
11 | ** accordance with the commercial license agreement provided with the |
12 | ** Software or, alternatively, in accordance with the terms contained in |
13 | ** a written agreement between you and The Qt Company. For licensing terms |
14 | ** and conditions see https://www.qt.io/terms-conditions. For further |
15 | ** information use the contact form at https://www.qt.io/contact-us. |
16 | ** |
17 | ** GNU Lesser General Public License Usage |
18 | ** Alternatively, this file may be used under the terms of the GNU Lesser |
19 | ** General Public License version 3 as published by the Free Software |
20 | ** Foundation and appearing in the file LICENSE.LGPL3 included in the |
21 | ** packaging of this file. Please review the following information to |
22 | ** ensure the GNU Lesser General Public License version 3 requirements |
23 | ** will be met: https://www.gnu.org/licenses/lgpl-3.0.html. |
24 | ** |
25 | ** GNU General Public License Usage |
26 | ** Alternatively, this file may be used under the terms of the GNU |
27 | ** General Public License version 2.0 or (at your option) the GNU General |
28 | ** Public license version 3 or any later version approved by the KDE Free |
29 | ** Qt Foundation. The licenses are as published by the Free Software |
30 | ** Foundation and appearing in the file LICENSE.GPL2 and LICENSE.GPL3 |
31 | ** included in the packaging of this file. Please review the following |
32 | ** information to ensure the GNU General Public License requirements will |
33 | ** be met: https://www.gnu.org/licenses/gpl-2.0.html and |
34 | ** https://www.gnu.org/licenses/gpl-3.0.html. |
35 | ** |
36 | ** $QT_END_LICENSE$ |
37 | ** |
38 | ****************************************************************************/ |
39 | |
40 | #include <private/qdrawhelper_p.h> |
41 | #include <private/qdrawingprimitive_sse2_p.h> |
42 | #include <private/qpaintengine_raster_p.h> |
43 | #include <private/qpixellayout_p.h> |
44 | |
45 | #if defined(QT_COMPILER_SUPPORTS_SSE4_1) |
46 | |
47 | QT_BEGIN_NAMESPACE |
48 | |
49 | #ifndef __AVX2__ |
50 | template<bool RGBA> |
51 | static void convertARGBToARGB32PM_sse4(uint *buffer, const uint *src, int count) |
52 | { |
53 | int i = 0; |
54 | const __m128i alphaMask = _mm_set1_epi32(0xff000000); |
55 | const __m128i rgbaMask = _mm_setr_epi8(2, 1, 0, 3, 6, 5, 4, 7, 10, 9, 8, 11, 14, 13, 12, 15); |
56 | const __m128i shuffleMask = _mm_setr_epi8(6, 7, 6, 7, 6, 7, 6, 7, 14, 15, 14, 15, 14, 15, 14, 15); |
57 | const __m128i half = _mm_set1_epi16(0x0080); |
58 | const __m128i zero = _mm_setzero_si128(); |
59 | |
60 | for (; i < count - 3; i += 4) { |
61 | __m128i srcVector = _mm_loadu_si128((const __m128i *)&src[i]); |
62 | if (!_mm_testz_si128(srcVector, alphaMask)) { |
63 | if (!_mm_testc_si128(srcVector, alphaMask)) { |
64 | if (RGBA) |
65 | srcVector = _mm_shuffle_epi8(srcVector, rgbaMask); |
66 | __m128i src1 = _mm_unpacklo_epi8(srcVector, zero); |
67 | __m128i src2 = _mm_unpackhi_epi8(srcVector, zero); |
68 | __m128i alpha1 = _mm_shuffle_epi8(src1, shuffleMask); |
69 | __m128i alpha2 = _mm_shuffle_epi8(src2, shuffleMask); |
70 | src1 = _mm_mullo_epi16(src1, alpha1); |
71 | src2 = _mm_mullo_epi16(src2, alpha2); |
72 | src1 = _mm_add_epi16(src1, _mm_srli_epi16(src1, 8)); |
73 | src2 = _mm_add_epi16(src2, _mm_srli_epi16(src2, 8)); |
74 | src1 = _mm_add_epi16(src1, half); |
75 | src2 = _mm_add_epi16(src2, half); |
76 | src1 = _mm_srli_epi16(src1, 8); |
77 | src2 = _mm_srli_epi16(src2, 8); |
78 | src1 = _mm_blend_epi16(src1, alpha1, 0x88); |
79 | src2 = _mm_blend_epi16(src2, alpha2, 0x88); |
80 | srcVector = _mm_packus_epi16(src1, src2); |
81 | _mm_storeu_si128((__m128i *)&buffer[i], srcVector); |
82 | } else { |
83 | if (RGBA) |
84 | _mm_storeu_si128((__m128i *)&buffer[i], _mm_shuffle_epi8(srcVector, rgbaMask)); |
85 | else if (buffer != src) |
86 | _mm_storeu_si128((__m128i *)&buffer[i], srcVector); |
87 | } |
88 | } else { |
89 | _mm_storeu_si128((__m128i *)&buffer[i], zero); |
90 | } |
91 | } |
92 | |
93 | SIMD_EPILOGUE(i, count, 3) { |
94 | uint v = qPremultiply(src[i]); |
95 | buffer[i] = RGBA ? RGBA2ARGB(v) : v; |
96 | } |
97 | } |
98 | |
99 | template<bool RGBA> |
100 | static void convertARGBToRGBA64PM_sse4(QRgba64 *buffer, const uint *src, int count) |
101 | { |
102 | int i = 0; |
103 | const __m128i alphaMask = _mm_set1_epi32(0xff000000); |
104 | const __m128i rgbaMask = _mm_setr_epi8(2, 1, 0, 3, 6, 5, 4, 7, 10, 9, 8, 11, 14, 13, 12, 15); |
105 | const __m128i shuffleMask = _mm_setr_epi8(6, 7, 6, 7, 6, 7, 6, 7, 14, 15, 14, 15, 14, 15, 14, 15); |
106 | const __m128i zero = _mm_setzero_si128(); |
107 | |
108 | for (; i < count - 3; i += 4) { |
109 | __m128i srcVector = _mm_loadu_si128((const __m128i *)&src[i]); |
110 | if (!_mm_testz_si128(srcVector, alphaMask)) { |
111 | bool cf = _mm_testc_si128(srcVector, alphaMask); |
112 | |
113 | if (!RGBA) |
114 | srcVector = _mm_shuffle_epi8(srcVector, rgbaMask); |
115 | const __m128i src1 = _mm_unpacklo_epi8(srcVector, srcVector); |
116 | const __m128i src2 = _mm_unpackhi_epi8(srcVector, srcVector); |
117 | if (!cf) { |
118 | __m128i alpha1 = _mm_shuffle_epi8(src1, shuffleMask); |
119 | __m128i alpha2 = _mm_shuffle_epi8(src2, shuffleMask); |
120 | __m128i dst1 = _mm_mulhi_epu16(src1, alpha1); |
121 | __m128i dst2 = _mm_mulhi_epu16(src2, alpha2); |
122 | // Map 0->0xfffe to 0->0xffff |
123 | dst1 = _mm_add_epi16(dst1, _mm_srli_epi16(dst1, 15)); |
124 | dst2 = _mm_add_epi16(dst2, _mm_srli_epi16(dst2, 15)); |
125 | // correct alpha value: |
126 | dst1 = _mm_blend_epi16(dst1, src1, 0x88); |
127 | dst2 = _mm_blend_epi16(dst2, src2, 0x88); |
128 | _mm_storeu_si128((__m128i *)&buffer[i], dst1); |
129 | _mm_storeu_si128((__m128i *)&buffer[i + 2], dst2); |
130 | } else { |
131 | _mm_storeu_si128((__m128i *)&buffer[i], src1); |
132 | _mm_storeu_si128((__m128i *)&buffer[i + 2], src2); |
133 | } |
134 | } else { |
135 | _mm_storeu_si128((__m128i *)&buffer[i], zero); |
136 | _mm_storeu_si128((__m128i *)&buffer[i + 2], zero); |
137 | } |
138 | } |
139 | |
140 | SIMD_EPILOGUE(i, count, 3) { |
141 | const uint s = RGBA ? RGBA2ARGB(src[i]) : src[i]; |
142 | buffer[i] = QRgba64::fromArgb32(s).premultiplied(); |
143 | } |
144 | } |
145 | #endif // __AVX2__ |
146 | |
147 | static inline __m128 Q_DECL_VECTORCALL reciprocal_mul_ps(__m128 a, float mul) |
148 | { |
149 | __m128 ia = _mm_rcp_ps(a); // Approximate 1/a |
150 | // Improve precision of ia using Newton-Raphson |
151 | ia = _mm_sub_ps(_mm_add_ps(ia, ia), _mm_mul_ps(ia, _mm_mul_ps(ia, a))); |
152 | ia = _mm_mul_ps(ia, _mm_set1_ps(mul)); |
153 | return ia; |
154 | } |
155 | |
156 | template<bool RGBA, bool RGBx> |
157 | static inline void convertARGBFromARGB32PM_sse4(uint *buffer, const uint *src, int count) |
158 | { |
159 | int i = 0; |
160 | if ((_MM_GET_EXCEPTION_MASK() & _MM_MASK_INVALID) == 0) { |
161 | for (; i < count; ++i) { |
162 | uint v = qUnpremultiply(src[i]); |
163 | if (RGBx) |
164 | v = 0xff000000 | v; |
165 | if (RGBA) |
166 | v = ARGB2RGBA(v); |
167 | buffer[i] = v; |
168 | } |
169 | return; |
170 | } |
171 | const __m128i alphaMask = _mm_set1_epi32(0xff000000); |
172 | const __m128i rgbaMask = _mm_setr_epi8(2, 1, 0, 3, 6, 5, 4, 7, 10, 9, 8, 11, 14, 13, 12, 15); |
173 | const __m128i zero = _mm_setzero_si128(); |
174 | |
175 | for (; i < count - 3; i += 4) { |
176 | __m128i srcVector = _mm_loadu_si128((const __m128i *)&src[i]); |
177 | if (!_mm_testz_si128(srcVector, alphaMask)) { |
178 | if (!_mm_testc_si128(srcVector, alphaMask)) { |
179 | __m128i srcVectorAlpha = _mm_srli_epi32(srcVector, 24); |
180 | if (RGBA) |
181 | srcVector = _mm_shuffle_epi8(srcVector, rgbaMask); |
182 | const __m128 a = _mm_cvtepi32_ps(srcVectorAlpha); |
183 | const __m128 ia = reciprocal_mul_ps(a, 255.0f); |
184 | __m128i src1 = _mm_unpacklo_epi8(srcVector, zero); |
185 | __m128i src3 = _mm_unpackhi_epi8(srcVector, zero); |
186 | __m128i src2 = _mm_unpackhi_epi16(src1, zero); |
187 | __m128i src4 = _mm_unpackhi_epi16(src3, zero); |
188 | src1 = _mm_unpacklo_epi16(src1, zero); |
189 | src3 = _mm_unpacklo_epi16(src3, zero); |
190 | __m128 ia1 = _mm_shuffle_ps(ia, ia, _MM_SHUFFLE(0, 0, 0, 0)); |
191 | __m128 ia2 = _mm_shuffle_ps(ia, ia, _MM_SHUFFLE(1, 1, 1, 1)); |
192 | __m128 ia3 = _mm_shuffle_ps(ia, ia, _MM_SHUFFLE(2, 2, 2, 2)); |
193 | __m128 ia4 = _mm_shuffle_ps(ia, ia, _MM_SHUFFLE(3, 3, 3, 3)); |
194 | src1 = _mm_cvtps_epi32(_mm_mul_ps(_mm_cvtepi32_ps(src1), ia1)); |
195 | src2 = _mm_cvtps_epi32(_mm_mul_ps(_mm_cvtepi32_ps(src2), ia2)); |
196 | src3 = _mm_cvtps_epi32(_mm_mul_ps(_mm_cvtepi32_ps(src3), ia3)); |
197 | src4 = _mm_cvtps_epi32(_mm_mul_ps(_mm_cvtepi32_ps(src4), ia4)); |
198 | src1 = _mm_packus_epi32(src1, src2); |
199 | src3 = _mm_packus_epi32(src3, src4); |
200 | src1 = _mm_packus_epi16(src1, src3); |
201 | // Handle potential alpha == 0 values: |
202 | __m128i srcVectorAlphaMask = _mm_cmpeq_epi32(srcVectorAlpha, zero); |
203 | src1 = _mm_andnot_si128(srcVectorAlphaMask, src1); |
204 | // Fixup alpha values: |
205 | if (RGBx) |
206 | srcVector = _mm_or_si128(src1, alphaMask); |
207 | else |
208 | srcVector = _mm_blendv_epi8(src1, srcVector, alphaMask); |
209 | _mm_storeu_si128((__m128i *)&buffer[i], srcVector); |
210 | } else { |
211 | if (RGBA) |
212 | _mm_storeu_si128((__m128i *)&buffer[i], _mm_shuffle_epi8(srcVector, rgbaMask)); |
213 | else if (buffer != src) |
214 | _mm_storeu_si128((__m128i *)&buffer[i], srcVector); |
215 | } |
216 | } else { |
217 | if (RGBx) |
218 | _mm_storeu_si128((__m128i *)&buffer[i], alphaMask); |
219 | else |
220 | _mm_storeu_si128((__m128i *)&buffer[i], zero); |
221 | } |
222 | } |
223 | |
224 | SIMD_EPILOGUE(i, count, 3) { |
225 | uint v = qUnpremultiply_sse4(src[i]); |
226 | if (RGBx) |
227 | v = 0xff000000 | v; |
228 | if (RGBA) |
229 | v = ARGB2RGBA(v); |
230 | buffer[i] = v; |
231 | } |
232 | } |
233 | |
234 | template<bool RGBA> |
235 | static inline void convertARGBFromRGBA64PM_sse4(uint *buffer, const QRgba64 *src, int count) |
236 | { |
237 | int i = 0; |
238 | if ((_MM_GET_EXCEPTION_MASK() & _MM_MASK_INVALID) == 0) { |
239 | for (; i < count; ++i) { |
240 | const QRgba64 v = src[i].unpremultiplied(); |
241 | buffer[i] = RGBA ? toRgba8888(v) : toArgb32(v); |
242 | } |
243 | return; |
244 | } |
245 | const __m128i alphaMask = _mm_set1_epi64x(qint64(Q_UINT64_C(0xffff) << 48)); |
246 | const __m128i alphaMask32 = _mm_set1_epi32(0xff000000); |
247 | const __m128i rgbaMask = _mm_setr_epi8(2, 1, 0, 3, 6, 5, 4, 7, 10, 9, 8, 11, 14, 13, 12, 15); |
248 | const __m128i zero = _mm_setzero_si128(); |
249 | |
250 | for (; i < count - 3; i += 4) { |
251 | __m128i srcVector1 = _mm_loadu_si128((const __m128i *)&src[i]); |
252 | __m128i srcVector2 = _mm_loadu_si128((const __m128i *)&src[i + 2]); |
253 | bool transparent1 = _mm_testz_si128(srcVector1, alphaMask); |
254 | bool opaque1 = _mm_testc_si128(srcVector1, alphaMask); |
255 | bool transparent2 = _mm_testz_si128(srcVector2, alphaMask); |
256 | bool opaque2 = _mm_testc_si128(srcVector2, alphaMask); |
257 | |
258 | if (!(transparent1 && transparent2)) { |
259 | if (!(opaque1 && opaque2)) { |
260 | __m128i srcVector1Alpha = _mm_srli_epi64(srcVector1, 48); |
261 | __m128i srcVector2Alpha = _mm_srli_epi64(srcVector2, 48); |
262 | __m128i srcVectorAlpha = _mm_packus_epi32(srcVector1Alpha, srcVector2Alpha); |
263 | const __m128 a = _mm_cvtepi32_ps(srcVectorAlpha); |
264 | // Convert srcVectorAlpha to final 8-bit alpha channel |
265 | srcVectorAlpha = _mm_add_epi32(srcVectorAlpha, _mm_set1_epi32(128)); |
266 | srcVectorAlpha = _mm_sub_epi32(srcVectorAlpha, _mm_srli_epi32(srcVectorAlpha, 8)); |
267 | srcVectorAlpha = _mm_srli_epi32(srcVectorAlpha, 8); |
268 | srcVectorAlpha = _mm_slli_epi32(srcVectorAlpha, 24); |
269 | const __m128 ia = reciprocal_mul_ps(a, 255.0f); |
270 | __m128i src1 = _mm_unpacklo_epi16(srcVector1, zero); |
271 | __m128i src2 = _mm_unpackhi_epi16(srcVector1, zero); |
272 | __m128i src3 = _mm_unpacklo_epi16(srcVector2, zero); |
273 | __m128i src4 = _mm_unpackhi_epi16(srcVector2, zero); |
274 | __m128 ia1 = _mm_shuffle_ps(ia, ia, _MM_SHUFFLE(0, 0, 0, 0)); |
275 | __m128 ia2 = _mm_shuffle_ps(ia, ia, _MM_SHUFFLE(1, 1, 1, 1)); |
276 | __m128 ia3 = _mm_shuffle_ps(ia, ia, _MM_SHUFFLE(2, 2, 2, 2)); |
277 | __m128 ia4 = _mm_shuffle_ps(ia, ia, _MM_SHUFFLE(3, 3, 3, 3)); |
278 | src1 = _mm_cvtps_epi32(_mm_mul_ps(_mm_cvtepi32_ps(src1), ia1)); |
279 | src2 = _mm_cvtps_epi32(_mm_mul_ps(_mm_cvtepi32_ps(src2), ia2)); |
280 | src3 = _mm_cvtps_epi32(_mm_mul_ps(_mm_cvtepi32_ps(src3), ia3)); |
281 | src4 = _mm_cvtps_epi32(_mm_mul_ps(_mm_cvtepi32_ps(src4), ia4)); |
282 | src1 = _mm_packus_epi32(src1, src2); |
283 | src3 = _mm_packus_epi32(src3, src4); |
284 | // Handle potential alpha == 0 values: |
285 | __m128i srcVector1AlphaMask = _mm_cmpeq_epi64(srcVector1Alpha, zero); |
286 | __m128i srcVector2AlphaMask = _mm_cmpeq_epi64(srcVector2Alpha, zero); |
287 | src1 = _mm_andnot_si128(srcVector1AlphaMask, src1); |
288 | src3 = _mm_andnot_si128(srcVector2AlphaMask, src3); |
289 | src1 = _mm_packus_epi16(src1, src3); |
290 | // Fixup alpha values: |
291 | src1 = _mm_blendv_epi8(src1, srcVectorAlpha, alphaMask32); |
292 | // Fix RGB order |
293 | if (!RGBA) |
294 | src1 = _mm_shuffle_epi8(src1, rgbaMask); |
295 | _mm_storeu_si128((__m128i *)&buffer[i], src1); |
296 | } else { |
297 | __m128i src1 = _mm_unpacklo_epi16(srcVector1, zero); |
298 | __m128i src2 = _mm_unpackhi_epi16(srcVector1, zero); |
299 | __m128i src3 = _mm_unpacklo_epi16(srcVector2, zero); |
300 | __m128i src4 = _mm_unpackhi_epi16(srcVector2, zero); |
301 | src1 = _mm_add_epi32(src1, _mm_set1_epi32(128)); |
302 | src2 = _mm_add_epi32(src2, _mm_set1_epi32(128)); |
303 | src3 = _mm_add_epi32(src3, _mm_set1_epi32(128)); |
304 | src4 = _mm_add_epi32(src4, _mm_set1_epi32(128)); |
305 | src1 = _mm_sub_epi32(src1, _mm_srli_epi32(src1, 8)); |
306 | src2 = _mm_sub_epi32(src2, _mm_srli_epi32(src2, 8)); |
307 | src3 = _mm_sub_epi32(src3, _mm_srli_epi32(src3, 8)); |
308 | src4 = _mm_sub_epi32(src4, _mm_srli_epi32(src4, 8)); |
309 | src1 = _mm_srli_epi32(src1, 8); |
310 | src2 = _mm_srli_epi32(src2, 8); |
311 | src3 = _mm_srli_epi32(src3, 8); |
312 | src4 = _mm_srli_epi32(src4, 8); |
313 | src1 = _mm_packus_epi32(src1, src2); |
314 | src3 = _mm_packus_epi32(src3, src4); |
315 | src1 = _mm_packus_epi16(src1, src3); |
316 | if (!RGBA) |
317 | src1 = _mm_shuffle_epi8(src1, rgbaMask); |
318 | _mm_storeu_si128((__m128i *)&buffer[i], src1); |
319 | } |
320 | } else { |
321 | _mm_storeu_si128((__m128i *)&buffer[i], zero); |
322 | } |
323 | } |
324 | |
325 | SIMD_EPILOGUE(i, count, 3) { |
326 | buffer[i] = qConvertRgba64ToRgb32_sse4<RGBA ? PixelOrderRGB : PixelOrderBGR>(src[i]); |
327 | } |
328 | } |
329 | |
330 | #ifndef __AVX2__ |
331 | void QT_FASTCALL convertARGB32ToARGB32PM_sse4(uint *buffer, int count, const QList<QRgb> *) |
332 | { |
333 | convertARGBToARGB32PM_sse4<false>(buffer, buffer, count); |
334 | } |
335 | |
336 | void QT_FASTCALL convertRGBA8888ToARGB32PM_sse4(uint *buffer, int count, const QList<QRgb> *) |
337 | { |
338 | convertARGBToARGB32PM_sse4<true>(buffer, buffer, count); |
339 | } |
340 | |
341 | const QRgba64 * QT_FASTCALL convertARGB32ToRGBA64PM_sse4(QRgba64 *buffer, const uint *src, int count, |
342 | const QList<QRgb> *, QDitherInfo *) |
343 | { |
344 | convertARGBToRGBA64PM_sse4<false>(buffer, src, count); |
345 | return buffer; |
346 | } |
347 | |
348 | const QRgba64 * QT_FASTCALL convertRGBA8888ToRGBA64PM_sse4(QRgba64 *buffer, const uint *src, int count, |
349 | const QList<QRgb> *, QDitherInfo *) |
350 | { |
351 | convertARGBToRGBA64PM_sse4<true>(buffer, src, count); |
352 | return buffer; |
353 | } |
354 | |
355 | const uint *QT_FASTCALL fetchARGB32ToARGB32PM_sse4(uint *buffer, const uchar *src, int index, int count, |
356 | const QList<QRgb> *, QDitherInfo *) |
357 | { |
358 | convertARGBToARGB32PM_sse4<false>(buffer, reinterpret_cast<const uint *>(src) + index, count); |
359 | return buffer; |
360 | } |
361 | |
362 | const uint *QT_FASTCALL fetchRGBA8888ToARGB32PM_sse4(uint *buffer, const uchar *src, int index, int count, |
363 | const QList<QRgb> *, QDitherInfo *) |
364 | { |
365 | convertARGBToARGB32PM_sse4<true>(buffer, reinterpret_cast<const uint *>(src) + index, count); |
366 | return buffer; |
367 | } |
368 | |
369 | const QRgba64 *QT_FASTCALL fetchARGB32ToRGBA64PM_sse4(QRgba64 *buffer, const uchar *src, int index, int count, |
370 | const QList<QRgb> *, QDitherInfo *) |
371 | { |
372 | convertARGBToRGBA64PM_sse4<false>(buffer, reinterpret_cast<const uint *>(src) + index, count); |
373 | return buffer; |
374 | } |
375 | |
376 | const QRgba64 *QT_FASTCALL fetchRGBA8888ToRGBA64PM_sse4(QRgba64 *buffer, const uchar *src, int index, int count, |
377 | const QList<QRgb> *, QDitherInfo *) |
378 | { |
379 | convertARGBToRGBA64PM_sse4<true>(buffer, reinterpret_cast<const uint *>(src) + index, count); |
380 | return buffer; |
381 | } |
382 | #endif // __AVX2__ |
383 | |
384 | void QT_FASTCALL storeRGB32FromARGB32PM_sse4(uchar *dest, const uint *src, int index, int count, |
385 | const QList<QRgb> *, QDitherInfo *) |
386 | { |
387 | uint *d = reinterpret_cast<uint *>(dest) + index; |
388 | convertARGBFromARGB32PM_sse4<false,true>(d, src, count); |
389 | } |
390 | |
391 | void QT_FASTCALL storeARGB32FromARGB32PM_sse4(uchar *dest, const uint *src, int index, int count, |
392 | const QList<QRgb> *, QDitherInfo *) |
393 | { |
394 | uint *d = reinterpret_cast<uint *>(dest) + index; |
395 | convertARGBFromARGB32PM_sse4<false,false>(d, src, count); |
396 | } |
397 | |
398 | void QT_FASTCALL storeRGBA8888FromARGB32PM_sse4(uchar *dest, const uint *src, int index, int count, |
399 | const QList<QRgb> *, QDitherInfo *) |
400 | { |
401 | uint *d = reinterpret_cast<uint *>(dest) + index; |
402 | convertARGBFromARGB32PM_sse4<true,false>(d, src, count); |
403 | } |
404 | |
405 | void QT_FASTCALL storeRGBXFromARGB32PM_sse4(uchar *dest, const uint *src, int index, int count, |
406 | const QList<QRgb> *, QDitherInfo *) |
407 | { |
408 | uint *d = reinterpret_cast<uint *>(dest) + index; |
409 | convertARGBFromARGB32PM_sse4<true,true>(d, src, count); |
410 | } |
411 | |
412 | template<QtPixelOrder PixelOrder> |
413 | void QT_FASTCALL storeA2RGB30PMFromARGB32PM_sse4(uchar *dest, const uint *src, int index, int count, |
414 | const QList<QRgb> *, QDitherInfo *) |
415 | { |
416 | uint *d = reinterpret_cast<uint *>(dest) + index; |
417 | for (int i = 0; i < count; ++i) |
418 | d[i] = qConvertArgb32ToA2rgb30_sse4<PixelOrder>(src[i]); |
419 | } |
420 | |
421 | #if QT_CONFIG(raster_64bit) |
422 | void QT_FASTCALL destStore64ARGB32_sse4(QRasterBuffer *rasterBuffer, int x, int y, const QRgba64 *buffer, int length) |
423 | { |
424 | uint *dest = (uint*)rasterBuffer->scanLine(y) + x; |
425 | convertARGBFromRGBA64PM_sse4<false>(dest, buffer, length); |
426 | } |
427 | |
428 | void QT_FASTCALL destStore64RGBA8888_sse4(QRasterBuffer *rasterBuffer, int x, int y, const QRgba64 *buffer, int length) |
429 | { |
430 | uint *dest = (uint*)rasterBuffer->scanLine(y) + x; |
431 | convertARGBFromRGBA64PM_sse4<true>(dest, buffer, length); |
432 | } |
433 | #endif |
434 | |
435 | void QT_FASTCALL storeARGB32FromRGBA64PM_sse4(uchar *dest, const QRgba64 *src, int index, int count, |
436 | const QList<QRgb> *, QDitherInfo *) |
437 | { |
438 | uint *d = (uint*)dest + index; |
439 | convertARGBFromRGBA64PM_sse4<false>(d, src, count); |
440 | } |
441 | |
442 | void QT_FASTCALL storeRGBA8888FromRGBA64PM_sse4(uchar *dest, const QRgba64 *src, int index, int count, |
443 | const QList<QRgb> *, QDitherInfo *) |
444 | { |
445 | uint *d = (uint*)dest + index; |
446 | convertARGBFromRGBA64PM_sse4<true>(d, src, count); |
447 | } |
448 | |
449 | template |
450 | void QT_FASTCALL storeA2RGB30PMFromARGB32PM_sse4<PixelOrderBGR>(uchar *dest, const uint *src, int index, int count, |
451 | const QList<QRgb> *, QDitherInfo *); |
452 | template |
453 | void QT_FASTCALL storeA2RGB30PMFromARGB32PM_sse4<PixelOrderRGB>(uchar *dest, const uint *src, int index, int count, |
454 | const QList<QRgb> *, QDitherInfo *); |
455 | |
456 | QT_END_NAMESPACE |
457 | |
458 | #endif |
459 | |