1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
|
;
; Copyright (c) 2010 The WebM project authors. All Rights Reserved.
;
; Use of this source code is governed by a BSD-style license
; that can be found in the LICENSE file in the root of the source
; tree. An additional intellectual property rights grant can be found
; in the file PATENTS. All contributing project authors may
; be found in the AUTHORS file in the root of the source tree.
;
%include "vpx_ports/x86_abi_support.asm"
;int vp8_fast_quantize_b_impl_mmx(short *coeff_ptr, short *zbin_ptr,
; short *qcoeff_ptr,short *dequant_ptr,
; short *scan_mask, short *round_ptr,
; short *quant_ptr, short *dqcoeff_ptr);
global sym(vp8_fast_quantize_b_impl_mmx)
sym(vp8_fast_quantize_b_impl_mmx):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 8
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;coeff_ptr
movq mm0, [rsi]
mov rax, arg(1) ;zbin_ptr
movq mm1, [rax]
movq mm3, mm0
psraw mm0, 15
pxor mm3, mm0
psubw mm3, mm0 ; abs
movq mm2, mm3
pcmpgtw mm1, mm2
pandn mm1, mm2
movq mm3, mm1
mov rdx, arg(6) ;quant_ptr
movq mm1, [rdx]
mov rcx, arg(5) ;round_ptr
movq mm2, [rcx]
paddw mm3, mm2
pmulhuw mm3, mm1
pxor mm3, mm0
psubw mm3, mm0 ;gain the sign back
mov rdi, arg(2) ;qcoeff_ptr
movq mm0, mm3
movq [rdi], mm3
mov rax, arg(3) ;dequant_ptr
movq mm2, [rax]
pmullw mm3, mm2
mov rax, arg(7) ;dqcoeff_ptr
movq [rax], mm3
; next 8
movq mm4, [rsi+8]
mov rax, arg(1) ;zbin_ptr
movq mm5, [rax+8]
movq mm7, mm4
psraw mm4, 15
pxor mm7, mm4
psubw mm7, mm4 ; abs
movq mm6, mm7
pcmpgtw mm5, mm6
pandn mm5, mm6
movq mm7, mm5
movq mm5, [rdx+8]
movq mm6, [rcx+8]
paddw mm7, mm6
pmulhuw mm7, mm5
pxor mm7, mm4
psubw mm7, mm4;gain the sign back
mov rdi, arg(2) ;qcoeff_ptr
movq mm1, mm7
movq [rdi+8], mm7
mov rax, arg(3) ;dequant_ptr
movq mm6, [rax+8]
pmullw mm7, mm6
mov rax, arg(7) ;dqcoeff_ptr
movq [rax+8], mm7
; next 8
movq mm4, [rsi+16]
mov rax, arg(1) ;zbin_ptr
movq mm5, [rax+16]
movq mm7, mm4
psraw mm4, 15
pxor mm7, mm4
psubw mm7, mm4 ; abs
movq mm6, mm7
pcmpgtw mm5, mm6
pandn mm5, mm6
movq mm7, mm5
movq mm5, [rdx+16]
movq mm6, [rcx+16]
paddw mm7, mm6
pmulhuw mm7, mm5
pxor mm7, mm4
psubw mm7, mm4;gain the sign back
mov rdi, arg(2) ;qcoeff_ptr
movq mm1, mm7
movq [rdi+16], mm7
mov rax, arg(3) ;dequant_ptr
movq mm6, [rax+16]
pmullw mm7, mm6
mov rax, arg(7) ;dqcoeff_ptr
movq [rax+16], mm7
; next 8
movq mm4, [rsi+24]
mov rax, arg(1) ;zbin_ptr
movq mm5, [rax+24]
movq mm7, mm4
psraw mm4, 15
pxor mm7, mm4
psubw mm7, mm4 ; abs
movq mm6, mm7
pcmpgtw mm5, mm6
pandn mm5, mm6
movq mm7, mm5
movq mm5, [rdx+24]
movq mm6, [rcx+24]
paddw mm7, mm6
pmulhuw mm7, mm5
pxor mm7, mm4
psubw mm7, mm4;gain the sign back
mov rdi, arg(2) ;qcoeff_ptr
movq mm1, mm7
movq [rdi+24], mm7
mov rax, arg(3) ;dequant_ptr
movq mm6, [rax+24]
pmullw mm7, mm6
mov rax, arg(7) ;dqcoeff_ptr
movq [rax+24], mm7
mov rdi, arg(4) ;scan_mask
mov rsi, arg(2) ;qcoeff_ptr
pxor mm5, mm5
pxor mm7, mm7
movq mm0, [rsi]
movq mm1, [rsi+8]
movq mm2, [rdi]
movq mm3, [rdi+8];
pcmpeqw mm0, mm7
pcmpeqw mm1, mm7
pcmpeqw mm6, mm6
pxor mm0, mm6
pxor mm1, mm6
psrlw mm0, 15
psrlw mm1, 15
pmaddwd mm0, mm2
pmaddwd mm1, mm3
movq mm5, mm0
paddd mm5, mm1
movq mm0, [rsi+16]
movq mm1, [rsi+24]
movq mm2, [rdi+16]
movq mm3, [rdi+24];
pcmpeqw mm0, mm7
pcmpeqw mm1, mm7
pcmpeqw mm6, mm6
pxor mm0, mm6
pxor mm1, mm6
psrlw mm0, 15
psrlw mm1, 15
pmaddwd mm0, mm2
pmaddwd mm1, mm3
paddd mm5, mm0
paddd mm5, mm1
movq mm0, mm5
psrlq mm5, 32
paddd mm0, mm5
; eob adjustment begins here
movd rcx, mm0
and rcx, 0xffff
xor rdx, rdx
sub rdx, rcx ; rdx=-rcx
bsr rax, rcx
inc rax
sar rdx, 31
and rax, rdx
; Substitute the sse assembly for the old mmx mixed assembly/C. The
; following is kept as reference
; movd rcx, mm0
; bsr rax, rcx
;
; mov eob, rax
; mov eee, rcx
;
;if(eee==0)
;{
; eob=-1;
;}
;else if(eee<0)
;{
; eob=15;
;}
;d->eob = eob+1;
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
;int vp8_fast_quantize_b_impl_sse(short *coeff_ptr, short *zbin_ptr,
; short *qcoeff_ptr,short *dequant_ptr,
; short *scan_mask, short *round_ptr,
; short *quant_ptr, short *dqcoeff_ptr);
global sym(vp8_fast_quantize_b_impl_sse)
sym(vp8_fast_quantize_b_impl_sse):
push rbp
mov rbp, rsp
SHADOW_ARGS_TO_STACK 8
push rsi
push rdi
; end prolog
mov rsi, arg(0) ;coeff_ptr
movdqa xmm0, [rsi]
mov rax, arg(1) ;zbin_ptr
movdqa xmm1, [rax]
movdqa xmm3, xmm0
psraw xmm0, 15
pxor xmm3, xmm0
psubw xmm3, xmm0 ; abs
movdqa xmm2, xmm3
pcmpgtw xmm1, xmm2
pandn xmm1, xmm2
movdqa xmm3, xmm1
mov rdx, arg(6) ; quant_ptr
movdqa xmm1, [rdx]
mov rcx, arg(5) ; round_ptr
movdqa xmm2, [rcx]
paddw xmm3, xmm2
pmulhuw xmm3, xmm1
pxor xmm3, xmm0
psubw xmm3, xmm0 ;gain the sign back
mov rdi, arg(2) ;qcoeff_ptr
movdqa xmm0, xmm3
movdqa [rdi], xmm3
mov rax, arg(3) ;dequant_ptr
movdqa xmm2, [rax]
pmullw xmm3, xmm2
mov rax, arg(7) ;dqcoeff_ptr
movdqa [rax], xmm3
; next 8
movdqa xmm4, [rsi+16]
mov rax, arg(1) ;zbin_ptr
movdqa xmm5, [rax+16]
movdqa xmm7, xmm4
psraw xmm4, 15
pxor xmm7, xmm4
psubw xmm7, xmm4 ; abs
movdqa xmm6, xmm7
pcmpgtw xmm5, xmm6
pandn xmm5, xmm6
movdqa xmm7, xmm5
movdqa xmm5, [rdx+16]
movdqa xmm6, [rcx+16]
paddw xmm7, xmm6
pmulhuw xmm7, xmm5
pxor xmm7, xmm4
psubw xmm7, xmm4;gain the sign back
mov rdi, arg(2) ;qcoeff_ptr
movdqa xmm1, xmm7
movdqa [rdi+16], xmm7
mov rax, arg(3) ;dequant_ptr
movdqa xmm6, [rax+16]
pmullw xmm7, xmm6
mov rax, arg(7) ;dqcoeff_ptr
movdqa [rax+16], xmm7
mov rdi, arg(4) ;scan_mask
pxor xmm7, xmm7
movdqa xmm2, [rdi]
movdqa xmm3, [rdi+16];
pcmpeqw xmm0, xmm7
pcmpeqw xmm1, xmm7
pcmpeqw xmm6, xmm6
pxor xmm0, xmm6
pxor xmm1, xmm6
psrlw xmm0, 15
psrlw xmm1, 15
pmaddwd xmm0, xmm2
pmaddwd xmm1, xmm3
movq xmm2, xmm0
movq xmm3, xmm1
psrldq xmm0, 8
psrldq xmm1, 8
paddd xmm0, xmm1
paddd xmm2, xmm3
paddd xmm0, xmm2
movq xmm1, xmm0
psrldq xmm0, 4
paddd xmm1, xmm0
movd rcx, xmm1
and rcx, 0xffff
xor rdx, rdx
sub rdx, rcx
bsr rax, rcx
inc rax
sar rdx, 31
and rax, rdx
; begin epilog
pop rdi
pop rsi
UNSHADOW_ARGS
pop rbp
ret
|