| 1 | /****************************************************************************
|
|---|
| 2 | **
|
|---|
| 3 | ** Copyright (C) 2011 Nokia Corporation and/or its subsidiary(-ies).
|
|---|
| 4 | ** All rights reserved.
|
|---|
| 5 | ** Contact: Nokia Corporation ([email protected])
|
|---|
| 6 | **
|
|---|
| 7 | ** This file is part of the QtGui module of the Qt Toolkit.
|
|---|
| 8 | **
|
|---|
| 9 | ** $QT_BEGIN_LICENSE:LGPL$
|
|---|
| 10 | ** Commercial Usage
|
|---|
| 11 | ** Licensees holding valid Qt Commercial licenses may use this file in
|
|---|
| 12 | ** accordance with the Qt Commercial License Agreement provided with the
|
|---|
| 13 | ** Software or, alternatively, in accordance with the terms contained in
|
|---|
| 14 | ** a written agreement between you and Nokia.
|
|---|
| 15 | **
|
|---|
| 16 | ** GNU Lesser General Public License Usage
|
|---|
| 17 | ** Alternatively, this file may be used under the terms of the GNU Lesser
|
|---|
| 18 | ** General Public License version 2.1 as published by the Free Software
|
|---|
| 19 | ** Foundation and appearing in the file LICENSE.LGPL included in the
|
|---|
| 20 | ** packaging of this file. Please review the following information to
|
|---|
| 21 | ** ensure the GNU Lesser General Public License version 2.1 requirements
|
|---|
| 22 | ** will be met: http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html.
|
|---|
| 23 | **
|
|---|
| 24 | ** In addition, as a special exception, Nokia gives you certain additional
|
|---|
| 25 | ** rights. These rights are described in the Nokia Qt LGPL Exception
|
|---|
| 26 | ** version 1.1, included in the file LGPL_EXCEPTION.txt in this package.
|
|---|
| 27 | **
|
|---|
| 28 | ** GNU General Public License Usage
|
|---|
| 29 | ** Alternatively, this file may be used under the terms of the GNU
|
|---|
| 30 | ** General Public License version 3.0 as published by the Free Software
|
|---|
| 31 | ** Foundation and appearing in the file LICENSE.GPL included in the
|
|---|
| 32 | ** packaging of this file. Please review the following information to
|
|---|
| 33 | ** ensure the GNU General Public License version 3.0 requirements will be
|
|---|
| 34 | ** met: http://www.gnu.org/copyleft/gpl.html.
|
|---|
| 35 | **
|
|---|
| 36 | ** If you have questions regarding the use of this file, please contact
|
|---|
| 37 | ** Nokia at [email protected].
|
|---|
| 38 | ** $QT_END_LICENSE$
|
|---|
| 39 | **
|
|---|
| 40 | ****************************************************************************/
|
|---|
| 41 |
|
|---|
| 42 | #include "qdrawhelper_arm_simd_p.h"
|
|---|
| 43 |
|
|---|
| 44 | #include <private/qpaintengine_raster_p.h>
|
|---|
| 45 | #include <private/qblendfunctions_p.h>
|
|---|
| 46 |
|
|---|
| 47 | #ifdef QT_HAVE_ARM_SIMD
|
|---|
| 48 |
|
|---|
| 49 | #if defined(Q_OS_SYMBIAN)
|
|---|
| 50 | #if !defined(__SWITCH_TO_ARM)
|
|---|
| 51 | #ifdef __MARM_THUMB__
|
|---|
| 52 | #ifndef __ARMCC__
|
|---|
| 53 | #define __SWITCH_TO_ARM asm("push {r0} ");\
|
|---|
| 54 | asm("add r0, pc, #4 ");\
|
|---|
| 55 | asm("bx r0 ");\
|
|---|
| 56 | asm("nop ");\
|
|---|
| 57 | asm(".align 2 ");\
|
|---|
| 58 | asm(".code 32 ");\
|
|---|
| 59 | asm("ldr r0, [sp], #4 ")
|
|---|
| 60 | #define __END_ARM asm(".code 16 ")
|
|---|
| 61 | #else
|
|---|
| 62 | #define __SWITCH_TO_ARM asm(".code 32 ");
|
|---|
| 63 | #define __END_ARM
|
|---|
| 64 | #endif // __ARMCC__
|
|---|
| 65 | #else
|
|---|
| 66 | #define __SWITCH_TO_ARM
|
|---|
| 67 | #define __END_ARM
|
|---|
| 68 | #endif //__MARM_THUMB__
|
|---|
| 69 | #endif
|
|---|
| 70 | #endif
|
|---|
| 71 |
|
|---|
| 72 | #if defined(Q_OS_SYMBIAN) && defined(Q_CC_RVCT)
|
|---|
| 73 | __asm void qt_blend_argb32_on_argb32_arm_simd(uchar *destPixels, int dbpl,
|
|---|
| 74 | const uchar *srcPixels, int sbpl,
|
|---|
| 75 | int w, int h,
|
|---|
| 76 | int const_alpha)
|
|---|
| 77 | {
|
|---|
| 78 | #ifndef __ARMCC__
|
|---|
| 79 | __SWITCH_TO_ARM;
|
|---|
| 80 | #else
|
|---|
| 81 | CODE32
|
|---|
| 82 | #endif // __ARMCC__
|
|---|
| 83 |
|
|---|
| 84 | stmfd sp!, {r4-r12, r14}
|
|---|
| 85 |
|
|---|
| 86 | // read arguments off the stack
|
|---|
| 87 | add r8, sp, #10 * 4
|
|---|
| 88 | ldmia r8, {r4-r6}
|
|---|
| 89 |
|
|---|
| 90 | // adjust dbpl and sbpl
|
|---|
| 91 | mov r14, #4
|
|---|
| 92 | mul r14, r4, r14
|
|---|
| 93 | sub r1, r1, r14
|
|---|
| 94 | sub r3, r3, r14
|
|---|
| 95 |
|
|---|
| 96 | // load 0xFF00FF00 to r12
|
|---|
| 97 | mov r12, #0xFF000000
|
|---|
| 98 | add r12, r12, #0xFF00
|
|---|
| 99 |
|
|---|
| 100 | // load 0x800080 to r14
|
|---|
| 101 | mov r14, #0x800000
|
|---|
| 102 | add r14, r14, #0x80
|
|---|
| 103 |
|
|---|
| 104 | /*
|
|---|
| 105 | Registers:
|
|---|
| 106 | r0 dst
|
|---|
| 107 | r1 dbpl
|
|---|
| 108 | r2 src
|
|---|
| 109 | r3 sbpl
|
|---|
| 110 | r4 w
|
|---|
| 111 | r5 h
|
|---|
| 112 | r6 const_alpha
|
|---|
| 113 | r12 0xFF0000
|
|---|
| 114 | r14 0x800080
|
|---|
| 115 | */
|
|---|
| 116 |
|
|---|
| 117 | cmp r6, #256 //test if we have fully opaque constant alpha value
|
|---|
| 118 | bne argb32constalpha // branch if not
|
|---|
| 119 |
|
|---|
| 120 | argb32_next_row
|
|---|
| 121 |
|
|---|
| 122 | mov r7, r4
|
|---|
| 123 |
|
|---|
| 124 | argb32_next_pixel
|
|---|
| 125 |
|
|---|
| 126 | ldr r8, [r2], #4 // load src pixel
|
|---|
| 127 |
|
|---|
| 128 | // Negate r8 and extract src alpha
|
|---|
| 129 | mvn r11, r8 // bitwise not
|
|---|
| 130 | uxtb r11, r11, ror #24
|
|---|
| 131 |
|
|---|
| 132 | cmp r11, #0 // test for full src opacity (negated)
|
|---|
| 133 | beq argb32_no_blend
|
|---|
| 134 |
|
|---|
| 135 | cmp r11, #255 // test for full src transparency (negated)
|
|---|
| 136 | addeq r0, #4
|
|---|
| 137 | beq argb32_nop
|
|---|
| 138 |
|
|---|
| 139 | ldr r9, [r0] // load dst pixel
|
|---|
| 140 |
|
|---|
| 141 | // blend
|
|---|
| 142 | uxtb16 r10, r9
|
|---|
| 143 | uxtb16 r6, r9, ror #8
|
|---|
| 144 | mla r10, r11, r10, r14
|
|---|
| 145 | mla r9, r6, r11, r14
|
|---|
| 146 | uxtab16 r10, r10, r10, ror #8
|
|---|
| 147 | uxtab16 r9, r9, r9, ror #8
|
|---|
| 148 | and r9, r9, r12
|
|---|
| 149 | uxtab16 r10, r9, r10, ror #8
|
|---|
| 150 |
|
|---|
| 151 | uqadd8 r8, r10, r8
|
|---|
| 152 |
|
|---|
| 153 | argb32_no_blend
|
|---|
| 154 |
|
|---|
| 155 | str r8, [r0], #4
|
|---|
| 156 |
|
|---|
| 157 | argb32_nop
|
|---|
| 158 |
|
|---|
| 159 | subs r7, r7, #1
|
|---|
| 160 | bgt argb32_next_pixel
|
|---|
| 161 |
|
|---|
| 162 | add r0, r0, r1 // dest = dest + dbpl
|
|---|
| 163 | add r2, r2, r3 // src = src + sbpl
|
|---|
| 164 |
|
|---|
| 165 | subs r5, r5, #1
|
|---|
| 166 | bgt argb32_next_row
|
|---|
| 167 |
|
|---|
| 168 | b argb32_blend_exit
|
|---|
| 169 |
|
|---|
| 170 | argb32constalpha
|
|---|
| 171 |
|
|---|
| 172 | cmp r6, #0
|
|---|
| 173 | beq argb32_blend_exit
|
|---|
| 174 |
|
|---|
| 175 | ; const_alpha = (const_alpha * 255) >> 8;
|
|---|
| 176 | mov r11, #255
|
|---|
| 177 | mul r6, r6, r11
|
|---|
| 178 | mov r11, r6, lsr #8
|
|---|
| 179 |
|
|---|
| 180 | argb32constalpha_next_row
|
|---|
| 181 |
|
|---|
| 182 | mov r7, r4
|
|---|
| 183 |
|
|---|
| 184 | argb32constalpha_next_pixel
|
|---|
| 185 |
|
|---|
| 186 | ldr r9, [r2], #4 // load src pixel
|
|---|
| 187 |
|
|---|
| 188 | // blend
|
|---|
| 189 | uxtb16 r10, r9
|
|---|
| 190 | uxtb16 r6, r9, ror #8
|
|---|
| 191 | mla r10, r11, r10, r14
|
|---|
| 192 | mla r9, r6, r11, r14
|
|---|
| 193 | uxtab16 r10, r10, r10, ror #8
|
|---|
| 194 | uxtab16 r9, r9, r9, ror #8
|
|---|
| 195 | and r9, r9, r12
|
|---|
| 196 | uxtab16 r8, r9, r10, ror #8
|
|---|
| 197 |
|
|---|
| 198 | ldr r9, [r0] // load dst pixel
|
|---|
| 199 |
|
|---|
| 200 | // blend
|
|---|
| 201 | uxtb16 r10, r9
|
|---|
| 202 | uxtb16 r6, r9, ror #8
|
|---|
| 203 |
|
|---|
| 204 | // Negate r8 and extract src alpha
|
|---|
| 205 | mvn r9, r8 // bitwise not
|
|---|
| 206 | uxtb r9, r9, ror #24
|
|---|
| 207 |
|
|---|
| 208 | mla r10, r9, r10, r14
|
|---|
| 209 | mla r9, r6, r9, r14
|
|---|
| 210 | uxtab16 r10, r10, r10, ror #8
|
|---|
| 211 | uxtab16 r9, r9, r9, ror #8
|
|---|
| 212 | and r9, r9, r12
|
|---|
| 213 | uxtab16 r10, r9, r10, ror #8
|
|---|
| 214 |
|
|---|
| 215 | uqadd8 r8, r10, r8
|
|---|
| 216 |
|
|---|
| 217 | str r8, [r0], #4
|
|---|
| 218 |
|
|---|
| 219 | subs r7, r7, #1
|
|---|
| 220 | bgt argb32constalpha_next_pixel
|
|---|
| 221 |
|
|---|
| 222 | add r0, r0, r1 // dest = dest + dbpl
|
|---|
| 223 | add r2, r2, r3 // src = src + sbpl
|
|---|
| 224 |
|
|---|
| 225 | subs r5, r5, #1
|
|---|
| 226 | bgt argb32constalpha_next_row
|
|---|
| 227 |
|
|---|
| 228 | argb32_blend_exit
|
|---|
| 229 |
|
|---|
| 230 | // Restore registers
|
|---|
| 231 | ldmfd sp!, {r4-r12, lr}
|
|---|
| 232 | bx lr
|
|---|
| 233 |
|
|---|
| 234 | __END_ARM
|
|---|
| 235 | }
|
|---|
| 236 |
|
|---|
| 237 | void qt_blend_rgb32_on_rgb32_arm_simd(uchar *destPixels, int dbpl,
|
|---|
| 238 | const uchar *srcPixels, int sbpl,
|
|---|
| 239 | int w, int h,
|
|---|
| 240 | int const_alpha)
|
|---|
| 241 | {
|
|---|
| 242 | if (const_alpha != 256) {
|
|---|
| 243 | qt_blend_argb32_on_argb32_arm_simd(destPixels, dbpl, srcPixels, sbpl, w, h, const_alpha);
|
|---|
| 244 | return;
|
|---|
| 245 | }
|
|---|
| 246 |
|
|---|
| 247 | const uint *src = (const uint *) srcPixels;
|
|---|
| 248 | uint *dst = (uint *) destPixels;
|
|---|
| 249 | if (w <= 64) {
|
|---|
| 250 | for (int y=0; y<h; ++y) {
|
|---|
| 251 | qt_memconvert(dst, src, w);
|
|---|
| 252 | dst = (quint32 *)(((uchar *) dst) + dbpl);
|
|---|
| 253 | src = (const quint32 *)(((const uchar *) src) + sbpl);
|
|---|
| 254 | }
|
|---|
| 255 | } else {
|
|---|
| 256 | int len = w * 4;
|
|---|
| 257 | for (int y=0; y<h; ++y) {
|
|---|
| 258 | memcpy(dst, src, len);
|
|---|
| 259 | dst = (quint32 *)(((uchar *) dst) + dbpl);
|
|---|
| 260 | src = (const quint32 *)(((const uchar *) src) + sbpl);
|
|---|
| 261 | }
|
|---|
| 262 | }
|
|---|
| 263 | }
|
|---|
| 264 |
|
|---|
| 265 | #else // defined(Q_OS_SYMBIAN) && defined(Q_CC_RVCT)
|
|---|
| 266 |
|
|---|
| 267 | // TODO: add GNU assembler instructions and support for other platforms.
|
|---|
| 268 | // Default to C code for now
|
|---|
| 269 |
|
|---|
| 270 | void qt_blend_argb32_on_argb32_arm_simd(uchar *destPixels, int dbpl,
|
|---|
| 271 | const uchar *srcPixels, int sbpl,
|
|---|
| 272 | int w, int h,
|
|---|
| 273 | int const_alpha)
|
|---|
| 274 | {
|
|---|
| 275 | const uint *src = (const uint *) srcPixels;
|
|---|
| 276 | uint *dst = (uint *) destPixels;
|
|---|
| 277 | if (const_alpha == 256) {
|
|---|
| 278 | for (int y=0; y<h; ++y) {
|
|---|
| 279 | for (int x=0; x<w; ++x) {
|
|---|
| 280 | uint s = src[x];
|
|---|
| 281 | if (s >= 0xff000000)
|
|---|
| 282 | dst[x] = s;
|
|---|
| 283 | else if (s != 0)
|
|---|
| 284 | dst[x] = s + BYTE_MUL(dst[x], qAlpha(~s));
|
|---|
| 285 | }
|
|---|
| 286 | dst = (quint32 *)(((uchar *) dst) + dbpl);
|
|---|
| 287 | src = (const quint32 *)(((const uchar *) src) + sbpl);
|
|---|
| 288 | }
|
|---|
| 289 | } else if (const_alpha != 0) {
|
|---|
| 290 | const_alpha = (const_alpha * 255) >> 8;
|
|---|
| 291 | for (int y=0; y<h; ++y) {
|
|---|
| 292 | for (int x=0; x<w; ++x) {
|
|---|
| 293 | uint s = BYTE_MUL(src[x], const_alpha);
|
|---|
| 294 | dst[x] = s + BYTE_MUL(dst[x], qAlpha(~s));
|
|---|
| 295 | }
|
|---|
| 296 | dst = (quint32 *)(((uchar *) dst) + dbpl);
|
|---|
| 297 | src = (const quint32 *)(((const uchar *) src) + sbpl);
|
|---|
| 298 | }
|
|---|
| 299 | }
|
|---|
| 300 | }
|
|---|
| 301 |
|
|---|
| 302 | void qt_blend_rgb32_on_rgb32_arm_simd(uchar *destPixels, int dbpl,
|
|---|
| 303 | const uchar *srcPixels, int sbpl,
|
|---|
| 304 | int w, int h,
|
|---|
| 305 | int const_alpha)
|
|---|
| 306 | {
|
|---|
| 307 | if (const_alpha != 256) {
|
|---|
| 308 | qt_blend_argb32_on_argb32_arm_simd(destPixels, dbpl, srcPixels, sbpl, w, h, const_alpha);
|
|---|
| 309 | return;
|
|---|
| 310 | }
|
|---|
| 311 |
|
|---|
| 312 | const uint *src = (const uint *) srcPixels;
|
|---|
| 313 | uint *dst = (uint *) destPixels;
|
|---|
| 314 | if (w <= 64) {
|
|---|
| 315 | for (int y=0; y<h; ++y) {
|
|---|
| 316 | qt_memconvert(dst, src, w);
|
|---|
| 317 | dst = (quint32 *)(((uchar *) dst) + dbpl);
|
|---|
| 318 | src = (const quint32 *)(((const uchar *) src) + sbpl);
|
|---|
| 319 | }
|
|---|
| 320 | } else {
|
|---|
| 321 | int len = w * 4;
|
|---|
| 322 | for (int y=0; y<h; ++y) {
|
|---|
| 323 | memcpy(dst, src, len);
|
|---|
| 324 | dst = (quint32 *)(((uchar *) dst) + dbpl);
|
|---|
| 325 | src = (const quint32 *)(((const uchar *) src) + sbpl);
|
|---|
| 326 | }
|
|---|
| 327 | }
|
|---|
| 328 | }
|
|---|
| 329 |
|
|---|
| 330 | #endif
|
|---|
| 331 |
|
|---|
| 332 | #endif // QT_HAVE_ARMV_SIMD
|
|---|