aboutsummaryrefslogblamecommitdiffstats
path: root/include/asm-i386/xor.h
blob: 23c86cef3b258a8611e765bd168e1c89054a19f6 (plain) (tree)
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
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537























































































































































































































































































































































































































































































































































                                                                        
                                                 
                                                 



                                                 





                                                 


                                                 
                                                 
                                                 
                                                 
                                                 

































































































































































































































































































































                                                                            
/*
 * include/asm-i386/xor.h
 *
 * Optimized RAID-5 checksumming functions for MMX and SSE.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2, or (at your option)
 * any later version.
 *
 * You should have received a copy of the GNU General Public License
 * (for example /usr/src/linux/COPYING); if not, write to the Free
 * Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */

/*
 * High-speed RAID5 checksumming functions utilizing MMX instructions.
 * Copyright (C) 1998 Ingo Molnar.
 */

#define LD(x,y)		"       movq   8*("#x")(%1), %%mm"#y"   ;\n"
#define ST(x,y)		"       movq %%mm"#y",   8*("#x")(%1)   ;\n"
#define XO1(x,y)	"       pxor   8*("#x")(%2), %%mm"#y"   ;\n"
#define XO2(x,y)	"       pxor   8*("#x")(%3), %%mm"#y"   ;\n"
#define XO3(x,y)	"       pxor   8*("#x")(%4), %%mm"#y"   ;\n"
#define XO4(x,y)	"       pxor   8*("#x")(%5), %%mm"#y"   ;\n"

#include <asm/i387.h>

static void
xor_pII_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
{
	unsigned long lines = bytes >> 7;

	kernel_fpu_begin();

	__asm__ __volatile__ (
#undef BLOCK
#define BLOCK(i) \
	LD(i,0)					\
		LD(i+1,1)			\
			LD(i+2,2)		\
				LD(i+3,3)	\
	XO1(i,0)				\
	ST(i,0)					\
		XO1(i+1,1)			\
		ST(i+1,1)			\
			XO1(i+2,2)		\
			ST(i+2,2)		\
				XO1(i+3,3)	\
				ST(i+3,3)

	" .align 32			;\n"
  	" 1:                            ;\n"

	BLOCK(0)
	BLOCK(4)
	BLOCK(8)
	BLOCK(12)

	"       addl $128, %1         ;\n"
	"       addl $128, %2         ;\n"
	"       decl %0               ;\n"
	"       jnz 1b                ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2)
	:
	: "memory");

	kernel_fpu_end();
}

static void
xor_pII_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
	      unsigned long *p3)
{
	unsigned long lines = bytes >> 7;

	kernel_fpu_begin();

	__asm__ __volatile__ (
#undef BLOCK
#define BLOCK(i) \
	LD(i,0)					\
		LD(i+1,1)			\
			LD(i+2,2)		\
				LD(i+3,3)	\
	XO1(i,0)				\
		XO1(i+1,1)			\
			XO1(i+2,2)		\
				XO1(i+3,3)	\
	XO2(i,0)				\
	ST(i,0)					\
		XO2(i+1,1)			\
		ST(i+1,1)			\
			XO2(i+2,2)		\
			ST(i+2,2)		\
				XO2(i+3,3)	\
				ST(i+3,3)

	" .align 32			;\n"
	" 1:                            ;\n"

	BLOCK(0)
	BLOCK(4)
	BLOCK(8)
	BLOCK(12)

	"       addl $128, %1         ;\n"
	"       addl $128, %2         ;\n"
	"       addl $128, %3         ;\n"
	"       decl %0               ;\n"
	"       jnz 1b                ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2), "+r" (p3)
	:
	: "memory");

	kernel_fpu_end();
}

static void
xor_pII_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
	      unsigned long *p3, unsigned long *p4)
{
	unsigned long lines = bytes >> 7;

	kernel_fpu_begin();

	__asm__ __volatile__ (
#undef BLOCK
#define BLOCK(i) \
	LD(i,0)					\
		LD(i+1,1)			\
			LD(i+2,2)		\
				LD(i+3,3)	\
	XO1(i,0)				\
		XO1(i+1,1)			\
			XO1(i+2,2)		\
				XO1(i+3,3)	\
	XO2(i,0)				\
		XO2(i+1,1)			\
			XO2(i+2,2)		\
				XO2(i+3,3)	\
	XO3(i,0)				\
	ST(i,0)					\
		XO3(i+1,1)			\
		ST(i+1,1)			\
			XO3(i+2,2)		\
			ST(i+2,2)		\
				XO3(i+3,3)	\
				ST(i+3,3)

	" .align 32			;\n"
	" 1:                            ;\n"

	BLOCK(0)
	BLOCK(4)
	BLOCK(8)
	BLOCK(12)

	"       addl $128, %1         ;\n"
	"       addl $128, %2         ;\n"
	"       addl $128, %3         ;\n"
	"       addl $128, %4         ;\n"
	"       decl %0               ;\n"
	"       jnz 1b                ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2), "+r" (p3), "+r" (p4)
	:
	: "memory");

	kernel_fpu_end();
}


static void
xor_pII_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
	      unsigned long *p3, unsigned long *p4, unsigned long *p5)
{
	unsigned long lines = bytes >> 7;

	kernel_fpu_begin();

	/* Make sure GCC forgets anything it knows about p4 or p5,
	   such that it won't pass to the asm volatile below a
	   register that is shared with any other variable.  That's
	   because we modify p4 and p5 there, but we can't mark them
	   as read/write, otherwise we'd overflow the 10-asm-operands
	   limit of GCC < 3.1.  */
	__asm__ ("" : "+r" (p4), "+r" (p5));

	__asm__ __volatile__ (
#undef BLOCK
#define BLOCK(i) \
	LD(i,0)					\
		LD(i+1,1)			\
			LD(i+2,2)		\
				LD(i+3,3)	\
	XO1(i,0)				\
		XO1(i+1,1)			\
			XO1(i+2,2)		\
				XO1(i+3,3)	\
	XO2(i,0)				\
		XO2(i+1,1)			\
			XO2(i+2,2)		\
				XO2(i+3,3)	\
	XO3(i,0)				\
		XO3(i+1,1)			\
			XO3(i+2,2)		\
				XO3(i+3,3)	\
	XO4(i,0)				\
	ST(i,0)					\
		XO4(i+1,1)			\
		ST(i+1,1)			\
			XO4(i+2,2)		\
			ST(i+2,2)		\
				XO4(i+3,3)	\
				ST(i+3,3)

	" .align 32			;\n"
	" 1:                            ;\n"

	BLOCK(0)
	BLOCK(4)
	BLOCK(8)
	BLOCK(12)

	"       addl $128, %1         ;\n"
	"       addl $128, %2         ;\n"
	"       addl $128, %3         ;\n"
	"       addl $128, %4         ;\n"
	"       addl $128, %5         ;\n"
	"       decl %0               ;\n"
	"       jnz 1b                ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2), "+r" (p3)
	: "r" (p4), "r" (p5) 
	: "memory");

	/* p4 and p5 were modified, and now the variables are dead.
	   Clobber them just to be sure nobody does something stupid
	   like assuming they have some legal value.  */
	__asm__ ("" : "=r" (p4), "=r" (p5));

	kernel_fpu_end();
}

#undef LD
#undef XO1
#undef XO2
#undef XO3
#undef XO4
#undef ST
#undef BLOCK

static void
xor_p5_mmx_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
{
	unsigned long lines = bytes >> 6;

	kernel_fpu_begin();

	__asm__ __volatile__ (
	" .align 32	             ;\n"
	" 1:                         ;\n"
	"       movq   (%1), %%mm0   ;\n"
	"       movq  8(%1), %%mm1   ;\n"
	"       pxor   (%2), %%mm0   ;\n"
	"       movq 16(%1), %%mm2   ;\n"
	"       movq %%mm0,   (%1)   ;\n"
	"       pxor  8(%2), %%mm1   ;\n"
	"       movq 24(%1), %%mm3   ;\n"
	"       movq %%mm1,  8(%1)   ;\n"
	"       pxor 16(%2), %%mm2   ;\n"
	"       movq 32(%1), %%mm4   ;\n"
	"       movq %%mm2, 16(%1)   ;\n"
	"       pxor 24(%2), %%mm3   ;\n"
	"       movq 40(%1), %%mm5   ;\n"
	"       movq %%mm3, 24(%1)   ;\n"
	"       pxor 32(%2), %%mm4   ;\n"
	"       movq 48(%1), %%mm6   ;\n"
	"       movq %%mm4, 32(%1)   ;\n"
	"       pxor 40(%2), %%mm5   ;\n"
	"       movq 56(%1), %%mm7   ;\n"
	"       movq %%mm5, 40(%1)   ;\n"
	"       pxor 48(%2), %%mm6   ;\n"
	"       pxor 56(%2), %%mm7   ;\n"
	"       movq %%mm6, 48(%1)   ;\n"
	"       movq %%mm7, 56(%1)   ;\n"
	
	"       addl $64, %1         ;\n"
	"       addl $64, %2         ;\n"
	"       decl %0              ;\n"
	"       jnz 1b               ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2)
	:
	: "memory");

	kernel_fpu_end();
}

static void
xor_p5_mmx_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
	     unsigned long *p3)
{
	unsigned long lines = bytes >> 6;

	kernel_fpu_begin();

	__asm__ __volatile__ (
	" .align 32,0x90             ;\n"
	" 1:                         ;\n"
	"       movq   (%1), %%mm0   ;\n"
	"       movq  8(%1), %%mm1   ;\n"
	"       pxor   (%2), %%mm0   ;\n"
	"       movq 16(%1), %%mm2   ;\n"
	"       pxor  8(%2), %%mm1   ;\n"
	"       pxor   (%3), %%mm0   ;\n"
	"       pxor 16(%2), %%mm2   ;\n"
	"       movq %%mm0,   (%1)   ;\n"
	"       pxor  8(%3), %%mm1   ;\n"
	"       pxor 16(%3), %%mm2   ;\n"
	"       movq 24(%1), %%mm3   ;\n"
	"       movq %%mm1,  8(%1)   ;\n"
	"       movq 32(%1), %%mm4   ;\n"
	"       movq 40(%1), %%mm5   ;\n"
	"       pxor 24(%2), %%mm3   ;\n"
	"       movq %%mm2, 16(%1)   ;\n"
	"       pxor 32(%2), %%mm4   ;\n"
	"       pxor 24(%3), %%mm3   ;\n"
	"       pxor 40(%2), %%mm5   ;\n"
	"       movq %%mm3, 24(%1)   ;\n"
	"       pxor 32(%3), %%mm4   ;\n"
	"       pxor 40(%3), %%mm5   ;\n"
	"       movq 48(%1), %%mm6   ;\n"
	"       movq %%mm4, 32(%1)   ;\n"
	"       movq 56(%1), %%mm7   ;\n"
	"       pxor 48(%2), %%mm6   ;\n"
	"       movq %%mm5, 40(%1)   ;\n"
	"       pxor 56(%2), %%mm7   ;\n"
	"       pxor 48(%3), %%mm6   ;\n"
	"       pxor 56(%3), %%mm7   ;\n"
	"       movq %%mm6, 48(%1)   ;\n"
	"       movq %%mm7, 56(%1)   ;\n"
      
	"       addl $64, %1         ;\n"
	"       addl $64, %2         ;\n"
	"       addl $64, %3         ;\n"
	"       decl %0              ;\n"
	"       jnz 1b               ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2), "+r" (p3)
	:
	: "memory" );

	kernel_fpu_end();
}

static void
xor_p5_mmx_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
	     unsigned long *p3, unsigned long *p4)
{
	unsigned long lines = bytes >> 6;

	kernel_fpu_begin();

	__asm__ __volatile__ (
	" .align 32,0x90             ;\n"
	" 1:                         ;\n"
	"       movq   (%1), %%mm0   ;\n"
	"       movq  8(%1), %%mm1   ;\n"
	"       pxor   (%2), %%mm0   ;\n"
	"       movq 16(%1), %%mm2   ;\n"
	"       pxor  8(%2), %%mm1   ;\n"
	"       pxor   (%3), %%mm0   ;\n"
	"       pxor 16(%2), %%mm2   ;\n"
	"       pxor  8(%3), %%mm1   ;\n"
	"       pxor   (%4), %%mm0   ;\n"
	"       movq 24(%1), %%mm3   ;\n"
	"       pxor 16(%3), %%mm2   ;\n"
	"       pxor  8(%4), %%mm1   ;\n"
	"       movq %%mm0,   (%1)   ;\n"
	"       movq 32(%1), %%mm4   ;\n"
	"       pxor 24(%2), %%mm3   ;\n"
	"       pxor 16(%4), %%mm2   ;\n"
	"       movq %%mm1,  8(%1)   ;\n"
	"       movq 40(%1), %%mm5   ;\n"
	"       pxor 32(%2), %%mm4   ;\n"
	"       pxor 24(%3), %%mm3   ;\n"
	"       movq %%mm2, 16(%1)   ;\n"
	"       pxor 40(%2), %%mm5   ;\n"
	"       pxor 32(%3), %%mm4   ;\n"
	"       pxor 24(%4), %%mm3   ;\n"
	"       movq %%mm3, 24(%1)   ;\n"
	"       movq 56(%1), %%mm7   ;\n"
	"       movq 48(%1), %%mm6   ;\n"
	"       pxor 40(%3), %%mm5   ;\n"
	"       pxor 32(%4), %%mm4   ;\n"
	"       pxor 48(%2), %%mm6   ;\n"
	"       movq %%mm4, 32(%1)   ;\n"
	"       pxor 56(%2), %%mm7   ;\n"
	"       pxor 40(%4), %%mm5   ;\n"
	"       pxor 48(%3), %%mm6   ;\n"
	"       pxor 56(%3), %%mm7   ;\n"
	"       movq %%mm5, 40(%1)   ;\n"
	"       pxor 48(%4), %%mm6   ;\n"
	"       pxor 56(%4), %%mm7   ;\n"
	"       movq %%mm6, 48(%1)   ;\n"
	"       movq %%mm7, 56(%1)   ;\n"
      
	"       addl $64, %1         ;\n"
	"       addl $64, %2         ;\n"
	"       addl $64, %3         ;\n"
	"       addl $64, %4         ;\n"
	"       decl %0              ;\n"
	"       jnz 1b               ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2), "+r" (p3), "+r" (p4)
	:
	: "memory");

	kernel_fpu_end();
}

static void
xor_p5_mmx_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
	     unsigned long *p3, unsigned long *p4, unsigned long *p5)
{
	unsigned long lines = bytes >> 6;

	kernel_fpu_begin();

	/* Make sure GCC forgets anything it knows about p4 or p5,
	   such that it won't pass to the asm volatile below a
	   register that is shared with any other variable.  That's
	   because we modify p4 and p5 there, but we can't mark them
	   as read/write, otherwise we'd overflow the 10-asm-operands
	   limit of GCC < 3.1.  */
	__asm__ ("" : "+r" (p4), "+r" (p5));

	__asm__ __volatile__ (
	" .align 32,0x90             ;\n"
	" 1:                         ;\n"
	"       movq   (%1), %%mm0   ;\n"
	"       movq  8(%1), %%mm1   ;\n"
	"       pxor   (%2), %%mm0   ;\n"
	"       pxor  8(%2), %%mm1   ;\n"
	"       movq 16(%1), %%mm2   ;\n"
	"       pxor   (%3), %%mm0   ;\n"
	"       pxor  8(%3), %%mm1   ;\n"
	"       pxor 16(%2), %%mm2   ;\n"
	"       pxor   (%4), %%mm0   ;\n"
	"       pxor  8(%4), %%mm1   ;\n"
	"       pxor 16(%3), %%mm2   ;\n"
	"       movq 24(%1), %%mm3   ;\n"
	"       pxor   (%5), %%mm0   ;\n"
	"       pxor  8(%5), %%mm1   ;\n"
	"       movq %%mm0,   (%1)   ;\n"
	"       pxor 16(%4), %%mm2   ;\n"
	"       pxor 24(%2), %%mm3   ;\n"
	"       movq %%mm1,  8(%1)   ;\n"
	"       pxor 16(%5), %%mm2   ;\n"
	"       pxor 24(%3), %%mm3   ;\n"
	"       movq 32(%1), %%mm4   ;\n"
	"       movq %%mm2, 16(%1)   ;\n"
	"       pxor 24(%4), %%mm3   ;\n"
	"       pxor 32(%2), %%mm4   ;\n"
	"       movq 40(%1), %%mm5   ;\n"
	"       pxor 24(%5), %%mm3   ;\n"
	"       pxor 32(%3), %%mm4   ;\n"
	"       pxor 40(%2), %%mm5   ;\n"
	"       movq %%mm3, 24(%1)   ;\n"
	"       pxor 32(%4), %%mm4   ;\n"
	"       pxor 40(%3), %%mm5   ;\n"
	"       movq 48(%1), %%mm6   ;\n"
	"       movq 56(%1), %%mm7   ;\n"
	"       pxor 32(%5), %%mm4   ;\n"
	"       pxor 40(%4), %%mm5   ;\n"
	"       pxor 48(%2), %%mm6   ;\n"
	"       pxor 56(%2), %%mm7   ;\n"
	"       movq %%mm4, 32(%1)   ;\n"
	"       pxor 48(%3), %%mm6   ;\n"
	"       pxor 56(%3), %%mm7   ;\n"
	"       pxor 40(%5), %%mm5   ;\n"
	"       pxor 48(%4), %%mm6   ;\n"
	"       pxor 56(%4), %%mm7   ;\n"
	"       movq %%mm5, 40(%1)   ;\n"
	"       pxor 48(%5), %%mm6   ;\n"
	"       pxor 56(%5), %%mm7   ;\n"
	"       movq %%mm6, 48(%1)   ;\n"
	"       movq %%mm7, 56(%1)   ;\n"
      
	"       addl $64, %1         ;\n"
	"       addl $64, %2         ;\n"
	"       addl $64, %3         ;\n"
	"       addl $64, %4         ;\n"
	"       addl $64, %5         ;\n"
	"       decl %0              ;\n"
	"       jnz 1b               ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2), "+r" (p3)
	: "r" (p4), "r" (p5)
	: "memory");

	/* p4 and p5 were modified, and now the variables are dead.
	   Clobber them just to be sure nobody does something stupid
	   like assuming they have some legal value.  */
	__asm__ ("" : "=r" (p4), "=r" (p5));

	kernel_fpu_end();
}

static struct xor_block_template xor_block_pII_mmx = {
	.name = "pII_mmx",
	.do_2 = xor_pII_mmx_2,
	.do_3 = xor_pII_mmx_3,
	.do_4 = xor_pII_mmx_4,
	.do_5 = xor_pII_mmx_5,
};

static struct xor_block_template xor_block_p5_mmx = {
	.name = "p5_mmx",
	.do_2 = xor_p5_mmx_2,
	.do_3 = xor_p5_mmx_3,
	.do_4 = xor_p5_mmx_4,
	.do_5 = xor_p5_mmx_5,
};

/*
 * Cache avoiding checksumming functions utilizing KNI instructions
 * Copyright (C) 1999 Zach Brown (with obvious credit due Ingo)
 */

#define XMMS_SAVE do {				\
	preempt_disable();			\
	cr0 = read_cr0();			\
	clts();					\
	__asm__ __volatile__ ( 			\
		"movups %%xmm0,(%0)	;\n\t"	\
		"movups %%xmm1,0x10(%0)	;\n\t"	\
		"movups %%xmm2,0x20(%0)	;\n\t"	\
		"movups %%xmm3,0x30(%0)	;\n\t"	\
		:				\
		: "r" (xmm_save) 		\
		: "memory");			\
} while(0)

#define XMMS_RESTORE do {			\
	__asm__ __volatile__ ( 			\
		"sfence			;\n\t"	\
		"movups (%0),%%xmm0	;\n\t"	\
		"movups 0x10(%0),%%xmm1	;\n\t"	\
		"movups 0x20(%0),%%xmm2	;\n\t"	\
		"movups 0x30(%0),%%xmm3	;\n\t"	\
		:				\
		: "r" (xmm_save)		\
		: "memory");			\
	write_cr0(cr0);				\
	preempt_enable();			\
} while(0)

#define ALIGN16 __attribute__((aligned(16)))

#define OFFS(x)		"16*("#x")"
#define PF_OFFS(x)	"256+16*("#x")"
#define	PF0(x)		"	prefetchnta "PF_OFFS(x)"(%1)		;\n"
#define LD(x,y)		"       movaps   "OFFS(x)"(%1), %%xmm"#y"	;\n"
#define ST(x,y)		"       movaps %%xmm"#y",   "OFFS(x)"(%1)	;\n"
#define PF1(x)		"	prefetchnta "PF_OFFS(x)"(%2)		;\n"
#define PF2(x)		"	prefetchnta "PF_OFFS(x)"(%3)		;\n"
#define PF3(x)		"	prefetchnta "PF_OFFS(x)"(%4)		;\n"
#define PF4(x)		"	prefetchnta "PF_OFFS(x)"(%5)		;\n"
#define PF5(x)		"	prefetchnta "PF_OFFS(x)"(%6)		;\n"
#define XO1(x,y)	"       xorps   "OFFS(x)"(%2), %%xmm"#y"	;\n"
#define XO2(x,y)	"       xorps   "OFFS(x)"(%3), %%xmm"#y"	;\n"
#define XO3(x,y)	"       xorps   "OFFS(x)"(%4), %%xmm"#y"	;\n"
#define XO4(x,y)	"       xorps   "OFFS(x)"(%5), %%xmm"#y"	;\n"
#define XO5(x,y)	"       xorps   "OFFS(x)"(%6), %%xmm"#y"	;\n"


static void
xor_sse_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
{
        unsigned long lines = bytes >> 8;
	char xmm_save[16*4] ALIGN16;
	int cr0;

	XMMS_SAVE;

        __asm__ __volatile__ (
#undef BLOCK
#define BLOCK(i) \
		LD(i,0)					\
			LD(i+1,1)			\
		PF1(i)					\
				PF1(i+2)		\
				LD(i+2,2)		\
					LD(i+3,3)	\
		PF0(i+4)				\
				PF0(i+6)		\
		XO1(i,0)				\
			XO1(i+1,1)			\
				XO1(i+2,2)		\
					XO1(i+3,3)	\
		ST(i,0)					\
			ST(i+1,1)			\
				ST(i+2,2)		\
					ST(i+3,3)	\


		PF0(0)
				PF0(2)

	" .align 32			;\n"
        " 1:                            ;\n"

		BLOCK(0)
		BLOCK(4)
		BLOCK(8)
		BLOCK(12)

        "       addl $256, %1           ;\n"
        "       addl $256, %2           ;\n"
        "       decl %0                 ;\n"
        "       jnz 1b                  ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2)
	:
        : "memory");

	XMMS_RESTORE;
}

static void
xor_sse_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
	  unsigned long *p3)
{
        unsigned long lines = bytes >> 8;
	char xmm_save[16*4] ALIGN16;
	int cr0;

	XMMS_SAVE;

        __asm__ __volatile__ (
#undef BLOCK
#define BLOCK(i) \
		PF1(i)					\
				PF1(i+2)		\
		LD(i,0)					\
			LD(i+1,1)			\
				LD(i+2,2)		\
					LD(i+3,3)	\
		PF2(i)					\
				PF2(i+2)		\
		PF0(i+4)				\
				PF0(i+6)		\
		XO1(i,0)				\
			XO1(i+1,1)			\
				XO1(i+2,2)		\
					XO1(i+3,3)	\
		XO2(i,0)				\
			XO2(i+1,1)			\
				XO2(i+2,2)		\
					XO2(i+3,3)	\
		ST(i,0)					\
			ST(i+1,1)			\
				ST(i+2,2)		\
					ST(i+3,3)	\


		PF0(0)
				PF0(2)

	" .align 32			;\n"
        " 1:                            ;\n"

		BLOCK(0)
		BLOCK(4)
		BLOCK(8)
		BLOCK(12)

        "       addl $256, %1           ;\n"
        "       addl $256, %2           ;\n"
        "       addl $256, %3           ;\n"
        "       decl %0                 ;\n"
        "       jnz 1b                  ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r"(p2), "+r"(p3)
	:
        : "memory" );

	XMMS_RESTORE;
}

static void
xor_sse_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
	  unsigned long *p3, unsigned long *p4)
{
        unsigned long lines = bytes >> 8;
	char xmm_save[16*4] ALIGN16;
	int cr0;

	XMMS_SAVE;

        __asm__ __volatile__ (
#undef BLOCK
#define BLOCK(i) \
		PF1(i)					\
				PF1(i+2)		\
		LD(i,0)					\
			LD(i+1,1)			\
				LD(i+2,2)		\
					LD(i+3,3)	\
		PF2(i)					\
				PF2(i+2)		\
		XO1(i,0)				\
			XO1(i+1,1)			\
				XO1(i+2,2)		\
					XO1(i+3,3)	\
		PF3(i)					\
				PF3(i+2)		\
		PF0(i+4)				\
				PF0(i+6)		\
		XO2(i,0)				\
			XO2(i+1,1)			\
				XO2(i+2,2)		\
					XO2(i+3,3)	\
		XO3(i,0)				\
			XO3(i+1,1)			\
				XO3(i+2,2)		\
					XO3(i+3,3)	\
		ST(i,0)					\
			ST(i+1,1)			\
				ST(i+2,2)		\
					ST(i+3,3)	\


		PF0(0)
				PF0(2)

	" .align 32			;\n"
        " 1:                            ;\n"

		BLOCK(0)
		BLOCK(4)
		BLOCK(8)
		BLOCK(12)

        "       addl $256, %1           ;\n"
        "       addl $256, %2           ;\n"
        "       addl $256, %3           ;\n"
        "       addl $256, %4           ;\n"
        "       decl %0                 ;\n"
        "       jnz 1b                  ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2), "+r" (p3), "+r" (p4)
	:
        : "memory" );

	XMMS_RESTORE;
}

static void
xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
	  unsigned long *p3, unsigned long *p4, unsigned long *p5)
{
        unsigned long lines = bytes >> 8;
	char xmm_save[16*4] ALIGN16;
	int cr0;

	XMMS_SAVE;

	/* Make sure GCC forgets anything it knows about p4 or p5,
	   such that it won't pass to the asm volatile below a
	   register that is shared with any other variable.  That's
	   because we modify p4 and p5 there, but we can't mark them
	   as read/write, otherwise we'd overflow the 10-asm-operands
	   limit of GCC < 3.1.  */
	__asm__ ("" : "+r" (p4), "+r" (p5));

        __asm__ __volatile__ (
#undef BLOCK
#define BLOCK(i) \
		PF1(i)					\
				PF1(i+2)		\
		LD(i,0)					\
			LD(i+1,1)			\
				LD(i+2,2)		\
					LD(i+3,3)	\
		PF2(i)					\
				PF2(i+2)		\
		XO1(i,0)				\
			XO1(i+1,1)			\
				XO1(i+2,2)		\
					XO1(i+3,3)	\
		PF3(i)					\
				PF3(i+2)		\
		XO2(i,0)				\
			XO2(i+1,1)			\
				XO2(i+2,2)		\
					XO2(i+3,3)	\
		PF4(i)					\
				PF4(i+2)		\
		PF0(i+4)				\
				PF0(i+6)		\
		XO3(i,0)				\
			XO3(i+1,1)			\
				XO3(i+2,2)		\
					XO3(i+3,3)	\
		XO4(i,0)				\
			XO4(i+1,1)			\
				XO4(i+2,2)		\
					XO4(i+3,3)	\
		ST(i,0)					\
			ST(i+1,1)			\
				ST(i+2,2)		\
					ST(i+3,3)	\


		PF0(0)
				PF0(2)

	" .align 32			;\n"
        " 1:                            ;\n"

		BLOCK(0)
		BLOCK(4)
		BLOCK(8)
		BLOCK(12)

        "       addl $256, %1           ;\n"
        "       addl $256, %2           ;\n"
        "       addl $256, %3           ;\n"
        "       addl $256, %4           ;\n"
        "       addl $256, %5           ;\n"
        "       decl %0                 ;\n"
        "       jnz 1b                  ;\n"
	: "+r" (lines),
	  "+r" (p1), "+r" (p2), "+r" (p3)
	: "r" (p4), "r" (p5)
	: "memory");

	/* p4 and p5 were modified, and now the variables are dead.
	   Clobber them just to be sure nobody does something stupid
	   like assuming they have some legal value.  */
	__asm__ ("" : "=r" (p4), "=r" (p5));

	XMMS_RESTORE;
}

static struct xor_block_template xor_block_pIII_sse = {
        .name = "pIII_sse",
        .do_2 =  xor_sse_2,
        .do_3 =  xor_sse_3,
        .do_4 =  xor_sse_4,
        .do_5 = xor_sse_5,
};

/* Also try the generic routines.  */
#include <asm-generic/xor.h>

#undef XOR_TRY_TEMPLATES
#define XOR_TRY_TEMPLATES				\
	do {						\
		xor_speed(&xor_block_8regs);		\
		xor_speed(&xor_block_8regs_p);		\
		xor_speed(&xor_block_32regs);		\
		xor_speed(&xor_block_32regs_p);		\
	        if (cpu_has_xmm)			\
			xor_speed(&xor_block_pIII_sse);	\
	        if (cpu_has_mmx) {			\
	                xor_speed(&xor_block_pII_mmx);	\
	                xor_speed(&xor_block_p5_mmx);	\
	        }					\
	} while (0)

/* We force the use of the SSE xor block because it can write around L2.
   We may also be able to load into the L1 only depending on how the cpu
   deals with a load to a line that is being prefetched.  */
#define XOR_SELECT_TEMPLATE(FASTEST) \
	(cpu_has_xmm ? &xor_block_pIII_sse : FASTEST)