source: trunk/src/gui/painting/qdrawhelper_armv6_rvct.inc@ 574

Last change on this file since 574 was 561, checked in by Dmitry A. Kuminov, 16 years ago

trunk: Merged in qt 4.6.1 sources.
  • Property svn:eol-style set to native
File size: 13.1 KB
Line 
1;****************************************************************************
2;**
3;** Copyright (C) 2009 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;
43; W A R N I N G
44; -------------
45;
46; This file is not part of the Qt API. It exists purely as an
47; implementation detail. This header file may change from version to
48; version without notice, or even be removed.
49;
50; We mean it.
51;
52
53;-----------------------------------------------------------------------------
54; Globals.
55; Earch marcro expects that caller has loaded 0x800080 to r14.
56;-----------------------------------------------------------------------------
57
58ComponentHalf EQU 0x800080
59
60;-----------------------------------------------------------------------------
61; ARM assembly implementations of accelerated graphics operations.
62;
63; Conventions:
64;
65; - r0 = Target buffer pointer
66; - r1 = Source buffer pointer
67; - r2 = Length of the buffer to blend
68; - r3 = Constant alpha for source buffer
69;
70;-----------------------------------------------------------------------------
71
72; A macro for transparently defining ARM functions
73 MACRO
74$func Function
75 AREA Function_$func, CODE
76 GLOBAL $func
77 ALIGN 4
78 CODE32
79$func
80 MEND
81
82
83;-----------------------------------------------------------------------------
84; Armv6 boosted implementation of BYTE_MUL(...) function found in qdrawhelper_p.h.
85;
86; @param dst Destination register where to store the result
87; @param x Value to multiply
88; @param a Multiplicator byte
89; @param r14 Component half 0x800080
90;
91; @note Trashes x, r8
92;-----------------------------------------------------------------------------
93 MACRO
94 ByteMul $dst, $x, $a
95
96 ; static inline uint BYTE_MUL(uint x, uint a)
97
98 ; uint r8 = (x & 0xff00ff) * a + 0x800080
99 uxtb16 r8, $x ; r8 = r8 & 0x00FF00FF
100 mla r8, r8, $a, r14
101
102 ; x = ((r >> 8) & 0xff00ff) * a + 0x800080
103 uxtb16 $x, $x, ror #8
104 mla $x, $x, $a, r14
105
106
107 ; r8 = (r8 + ((r8 >> 8) & 0xff00ff) ) >> 8
108 ; r8 &= 0xff00ff
109 uxtab16 r8, r8, r8, ror #8
110 uxtb16 r8, r8, ror #8
111
112 ; x = x + ((x >>8) & 0xff00ff)
113 uxtab16 $x, $x, $x, ror #8
114
115 ; x &= 0xff00ff00
116 ; x |= r8
117 uxtb16 $x, $x, ror #8
118 orr $dst, r8, $x, lsl #8
119
120 MEND
121
122;-----------------------------------------------------------------------------
123; Armv6 boosted implementation of INTERPOLATE_PIXEL_255(...) function found in
124; qdrawhelper_p.h.
125;
126; @param dst Destination register where to store the result
127; @param x First value to multiply
128; @param a Multiplicator byte for first value
129; @param y Second value to multiply
130; @param b Multiplicator byte for second value
131; @param r14 Component half 0x800080
132;
133;
134; @note Trashes x, r8, r14
135;-----------------------------------------------------------------------------
136 MACRO
137 InterpolatePixel255 $dst, $x, $a, $y, $b
138
139 ; static inline uint INTERPOLATE_PIXEL_255(uint x, uint a, uint y, uint b)
140
141 ; First calculate the parts where we need 0x800080
142
143 ; uint r8 = (((x & 0xff00ff) * a) + 0x800080)
144 uxtb16 r8, $x ; r8 = r8 & 0x00FF00FF
145 mla r8, r8, $a, r14
146
147 ; x = ((((x >> 8) & 0xff00ff) * a) + 0x800080)
148 uxtb16 $x, $x, ror #8
149 mla $x, $x, $a, r14
150
151 ; Now we are trashing r14 to free it for other purposes
152
153 ; uint r14 = (y & 0xff00ff) * b
154 uxtb16 r14, $y ; r14 = y & 0x00FF00FF
155 mul r14, r14, $b
156
157 ; r8 = r8 + r14
158 add r8, r8, r14
159
160 ; r8 = (r8 + ((r8 >> 8) & 0xff00ff) ) >> 8
161 ; r8 &= 0xff00ff
162 uxtab16 r8, r8, r8, ror #8
163 uxtb16 r8, r8, ror #8
164
165 ; r14 = ((y >> 8) & 0xff00ff) * b
166 uxtb16 r14, $y, ror #8 ; r14 = ((y >> 8) & 0xFF00FF)
167 mul r14, r14, $b
168
169 ; x = x + r14
170 add $x, $x, r14
171
172 ; x = x + ((x >>8) & 0xff00ff)
173 uxtab16 $x, $x, $x, ror #8
174
175 ; x &= 0xff00ff00
176 ; x |= r8
177 uxtb16 $x, $x, ror #8
178 orr $dst, r8, $x, lsl #8
179
180 MEND
181
182;-----------------------------------------------------------------------------
183;
184;-----------------------------------------------------------------------------
185 MACRO
186$label Blend4Pixels $BlendPixel
187
188 ; Blend first 4 pixels
189
190 ldmia r1!, {r4-r7}
191 ldm r0, {r9-r12}
192
193b4p1_$label $BlendPixel r9, r4, r3
194b4p2_$label $BlendPixel r10, r5, r3
195b4p3_$label $BlendPixel r11, r6, r3
196b4p4_$label $BlendPixel r12, r7, r3
197
198 stmia r0!, {r9-r12}
199
200 MEND
201
202;-----------------------------------------------------------------------------
203;
204;-----------------------------------------------------------------------------
205 MACRO
206$label Blend8Pixels $BlendPixel
207
208b8p1_$label Blend4Pixels $BlendPixel
209b8p2_$label Blend4Pixels $BlendPixel
210
211 MEND
212
213;-----------------------------------------------------------------------------
214;
215;-----------------------------------------------------------------------------
216 MACRO
217$label Blend16Pixels $BlendPixel
218
219b16p1_$label Blend8Pixels $BlendPixel
220b16p2_$label Blend8Pixels $BlendPixel
221
222 MEND
223
224;-----------------------------------------------------------------------------
225;
226;-----------------------------------------------------------------------------
227 MACRO
228$label Blend32Pixels $BlendPixel
229
230b32p1_$label Blend16Pixels $BlendPixel
231b32p2_$label Blend16Pixels $BlendPixel
232
233 MEND
234
235;-----------------------------------------------------------------------------
236; A macro for source over compositing one row of pixels and saving the results
237; to destination buffer.
238;
239; @param dest Destination buffer (r0)
240; @param src Source buffer (r1)
241; @param length Length (r2)
242; @param const_alpha Constant alpha (r3)
243; @param r14 Component Half (0x800080) (r14)
244;
245; @note Advances r0, r1
246; @note Trashes r2, r4-r12
247;-----------------------------------------------------------------------------
248 MACRO
249$label BlendRow $BlendPixel
250
251 pld [r1]
252
253bloop_$label
254 ; Blend 32 pixels per loop iteration
255 subs r2, r2, #32
256 bmi b_remaining_$label
257
258brp1_$label Blend32Pixels $BlendPixel
259
260 b bloop_$label
261
262b_remaining_$label
263
264 ; Remaining 31 pixels
265
266 addmi r2, r2, #32
267
268 ; Blend 16 pixels
269 tst r2, #16
270 beq b_remaining8_$label
271
272brp2_$label Blend16Pixels $BlendPixel
273
274b_remaining8_$label
275
276 ; Blend 8 pixels
277 tst r2, #8
278 beq b_remaining4_$label
279
280brp3_$label Blend8Pixels $BlendPixel
281
282b_remaining4_$label
283
284 ; Blend 4 pixels
285 tst r2, #4
286 beq b_remaining3_$label
287
288brp4_$label Blend4Pixels $BlendPixel
289
290b_remaining3_$label
291
292 ; Remaining 3 pixels
293
294 tst r2, #2
295 beq b_last_$label
296
297 ldmia r1!, {r4-r5}
298 ldm r0, {r9-r10}
299
300brp5_$label $BlendPixel r9, r4, r3
301brp6_$label $BlendPixel r10, r5, r3
302
303 stmia r0!, {r9-r10}
304
305b_last_$label
306
307 tst r2, #1
308 beq bexit_$label
309
310 ldr r4, [r1]
311 ldr r9, [r0]
312
313bpl_$label $BlendPixel r9, r4, r3
314
315 str r9, [r0]
316
317bexit_$label
318
319 MEND
320
321;-----------------------------------------------------------------------------
322; A macro for source over compositing one row of pixels and saving the results
323; to destination buffer. Restores all registers.
324;
325; @param dest Destination buffer (r0)
326; @param src Source buffer (r1)
327; @param length Length (r2)
328; @param const_alpha Constant alpha (r3)
329; @param r14 Component Half (0x800080) (r14)
330;
331; @note Advances r0, r1
332; @note Trashes r2, r4-r12
333;-----------------------------------------------------------------------------
334 MACRO
335$label BlendRowSafe $BlendPixel
336
337 stmfd sp!, {r0-r6} ; Preserves registers only up to r6
338
339brs_$label BlendRow $BlendPixel
340
341 ldmfd sp!, {r0-r6}
342
343 MEND
344
345
346;-----------------------------------------------------------------------------
347; Pix Copy.
348; NOTE! Cache line size of ARM1136JF-S and ARM1136J-S is 32 bytes (8 pixels).
349;
350; @param dst Destination pixels (r0)
351; @param src Source pixels (r1)
352; @param len Length (r2)
353;
354; @note Trashes r3-r10
355;-----------------------------------------------------------------------------
356 MACRO
357$label PixCpy $dst, $src, $len
358
359 pld [$src]
360
361pcpy_loop_$label
362 ; Copy 8 pixels per loop iteration
363 pld [$src, #96]
364 subs $len, $len, #8
365 ldmgeia $src!, {r3-r10}
366 stmgeia $dst!, {r3-r10}
367 bgt pcpy_loop_$label
368
369pcpy_remaining_$label
370
371 ; Copy up to 7 remaining pixels
372
373 ; Copy 4 pixels
374 tst $len, #4
375 ldmneia $src!, {r3-r6}
376 stmneia $dst!, {r3-r6}
377
378 tst $len, #2
379 ldmneia $src!, {r3-r4}
380 stmneia $dst!, {r3-r4}
381
382 tst $len, #1
383 ldrne r3, [$src]
384 strne r3, [$dst]
385
386 MEND
387
388;-----------------------------------------------------------------------------
389; General Pix Copy. Maximum 8 pixels at time. Restores all registers.
390;
391; @param dst Destination pixels (r0)
392; @param src Source pixels (r1)
393; @param len Length (r2)
394;
395; @note Trashes r3-r10
396;-----------------------------------------------------------------------------
397 MACRO
398$label PixCpySafe $dst, $src, $len
399
400 stmfd sp!, {r0-r6} ; Preserves registers only up to r6
401
402pcs_$label PixCpy $dst, $src, $len
403
404 ldmfd sp!, {r0-r6} ; pop
405
406 MEND
407
408
409;-----------------------------------------------------------------------------
410; A macro for source over compositing one pixel and saving the result to
411; dst register.
412;
413; @param dst Destination register, must contain destination pixel upon entry
414; @param src Source register, must contain source pixel upon entry
415; @param const_alpha Constant source alpha
416; @param r14 Component half 0x800080
417;
418; @note Trashes const_alpha, r8
419;-----------------------------------------------------------------------------
420 MACRO
421$label PixelSourceOver $dst, $src, $const_alpha
422
423 ; Negate src and extract alpha
424 mvn $const_alpha, $src ; bitwise not
425 uxtb $const_alpha, $const_alpha, ror #24 ; r3 = ((r3 & 0xFF000000) >> 24);
426
427 ;cmp $const_alpha, #255 ; test for full transparency ( negated )
428 ;beq exit_$label
429 cmp $const_alpha, #0 ; test for full opacity ( negated )
430 moveq $dst, $src
431 beq exit_$label
432
433 ByteMul $dst, $dst, $const_alpha
434 add $dst, $src, $dst
435
436exit_$label
437 MEND
438
439;-----------------------------------------------------------------------------
440; A macro for source over compositing one pixel and saving the result to
441; dst register.
442;
443; @param dst Destination register, must contain destination pixel upon entry
444; @param src Source register, must contain source pixel upon entry
445; @param const_alpha Constant source alpha
446; @param r14 Component half 0x800080
447;
448; @note Trashes src, const_alpha, r8
449;-----------------------------------------------------------------------------
450 MACRO
451$label PixelSourceOverConstAlpha $dst, $src, $const_alpha
452
453 ; store alpha because we are going to trash it
454 stmfd sp!, {$const_alpha}
455
456 ByteMul $src, $src, $const_alpha
457
458 ; Negate src and extract alpha
459 mvn $const_alpha, $src ; bitwise not
460 uxtb $const_alpha, $const_alpha, ror #24 ; r3 = ((r3 & 0xFF000000) >> 24);
461
462 ByteMul $dst, $dst, $const_alpha
463
464 add $dst, $src, $dst
465
466 ; recover alpha
467 ldmfd sp!, {$const_alpha}
468
469 MEND
470
471;-----------------------------------------------------------------------------
472; A macro for source over compositing one pixel and saving the result to
473; a register.
474;
475; @param dst Destination register, must contain destination pixel upon entry
476; @param src Source register, must contain source pixel upon entry
477; @param const_alpha Constant source alpha
478; @param r14 Component half 0x800080
479;
480; @note Trashes src, r8
481;-----------------------------------------------------------------------------
482 MACRO
483$label PixelSourceConstAlpha $dst, $src, $const_alpha
484
485 ; store r2 and r14 because we are going to trash them
486 stmfd sp!, {r2, r14}
487
488 rsb r2, $const_alpha, #255
489 InterpolatePixel255 $dst, $src, $const_alpha, $dst, r2
490
491 ; recover r2 and r14
492 ldmfd sp!, {r2, r14}
493
494 MEND
495
496 END ; File end
Note: See TracBrowser for help on using the repository browser.