summaryrefslogtreecommitdiff
path: root/arch/mips/mm/pg-sb1.c
blob: adb37d0a30ea8d28ef90a8743844fc220dac396c (plain)
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
/*
 * Copyright (C) 1996 David S. Miller (dm@engr.sgi.com)
 * Copyright (C) 1997, 2001 Ralf Baechle (ralf@gnu.org)
 * Copyright (C) 2000 SiByte, Inc.
 * Copyright (C) 2005 Thiemo Seufer
 *
 * Written by Justin Carlson of SiByte, Inc.
 *         and Kip Walker of Broadcom Corp.
 *
 *
 * 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
 * of the License, or (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
 */
#include <linux/module.h>
#include <linux/sched.h>
#include <linux/smp.h>

#include <asm/io.h>
#include <asm/sibyte/sb1250.h>
#include <asm/sibyte/sb1250_regs.h>
#include <asm/sibyte/sb1250_dma.h>

#ifdef CONFIG_SB1_PASS_1_WORKAROUNDS
#define SB1_PREF_LOAD_STREAMED_HINT "0"
#define SB1_PREF_STORE_STREAMED_HINT "1"
#else
#define SB1_PREF_LOAD_STREAMED_HINT "4"
#define SB1_PREF_STORE_STREAMED_HINT "5"
#endif

static inline void clear_page_cpu(void *page)
{
	unsigned char *addr = (unsigned char *) page;
	unsigned char *end = addr + PAGE_SIZE;

	/*
	 * JDCXXX - This should be bottlenecked by the write buffer, but these
	 * things tend to be mildly unpredictable...should check this on the
	 * performance model
	 *
	 * We prefetch 4 lines ahead.  We're also "cheating" slightly here...
	 * since we know we're on an SB1, we force the assembler to take
	 * 64-bit operands to speed things up
	 */
	__asm__ __volatile__(
	"	.set	push		\n"
	"	.set	mips4		\n"
	"	.set	noreorder	\n"
#ifdef CONFIG_CPU_HAS_PREFETCH
	"	daddiu	%0, %0, 128	\n"
	"	pref	" SB1_PREF_STORE_STREAMED_HINT ", -128(%0)  \n"
					     /* Prefetch the first 4 lines */
	"	pref	" SB1_PREF_STORE_STREAMED_HINT ",  -96(%0)  \n"
	"	pref	" SB1_PREF_STORE_STREAMED_HINT ",  -64(%0)  \n"
	"	pref	" SB1_PREF_STORE_STREAMED_HINT ",  -32(%0)  \n"
	"1:	sd	$0, -128(%0)	\n"  /* Throw out a cacheline of 0's */
	"	sd	$0, -120(%0)	\n"
	"	sd	$0, -112(%0)	\n"
	"	sd	$0, -104(%0)	\n"
	"	daddiu	%0, %0, 32	\n"
	"	bnel	%0, %1, 1b	\n"
	"	 pref	" SB1_PREF_STORE_STREAMED_HINT ",  -32(%0)  \n"
	"	daddiu	%0, %0, -128	\n"
#endif
	"	sd	$0, 0(%0)	\n"  /* Throw out a cacheline of 0's */
	"1:	sd	$0, 8(%0)	\n"
	"	sd	$0, 16(%0)	\n"
	"	sd	$0, 24(%0)	\n"
	"	daddiu	%0, %0, 32	\n"
	"	bnel	%0, %1, 1b	\n"
	"	 sd	$0, 0(%0)	\n"
	"	.set	pop		\n"
	: "+r" (addr)
	: "r" (end)
	: "memory");
}

static inline void copy_page_cpu(void *to, void *from)
{
	unsigned char *src = (unsigned char *)from;
	unsigned char *dst = (unsigned char *)to;
	unsigned char *end = src + PAGE_SIZE;

	/*
	 * The pref's used here are using "streaming" hints, which cause the
	 * copied data to be kicked out of the cache sooner.  A page copy often
	 * ends up copying a lot more data than is commonly used, so this seems
	 * to make sense in terms of reducing cache pollution, but I've no real
	 * performance data to back this up
	 */
	__asm__ __volatile__(
	"	.set	push		\n"
	"	.set	mips4		\n"
	"	.set	noreorder	\n"
#ifdef CONFIG_CPU_HAS_PREFETCH
	"	daddiu	%0, %0, 128	\n"
	"	daddiu	%1, %1, 128	\n"
	"	pref	" SB1_PREF_LOAD_STREAMED_HINT  ", -128(%0)\n"
					     /* Prefetch the first 4 lines */
	"	pref	" SB1_PREF_STORE_STREAMED_HINT ", -128(%1)\n"
	"	pref	" SB1_PREF_LOAD_STREAMED_HINT  ",  -96(%0)\n"
	"	pref	" SB1_PREF_STORE_STREAMED_HINT ",  -96(%1)\n"
	"	pref	" SB1_PREF_LOAD_STREAMED_HINT  ",  -64(%0)\n"
	"	pref	" SB1_PREF_STORE_STREAMED_HINT ",  -64(%1)\n"
	"	pref	" SB1_PREF_LOAD_STREAMED_HINT  ",  -32(%0)\n"
	"1:	pref	" SB1_PREF_STORE_STREAMED_HINT ",  -32(%1)\n"
# ifdef CONFIG_64BIT
	"	ld	$8, -128(%0)	\n"  /* Block copy a cacheline */
	"	ld	$9, -120(%0)	\n"
	"	ld	$10, -112(%0)	\n"
	"	ld	$11, -104(%0)	\n"
	"	sd	$8, -128(%1)	\n"
	"	sd	$9, -120(%1)	\n"
	"	sd	$10, -112(%1)	\n"
	"	sd	$11, -104(%1)	\n"
# else
	"	lw	$2, -128(%0)	\n"  /* Block copy a cacheline */
	"	lw	$3, -124(%0)	\n"
	"	lw	$6, -120(%0)	\n"
	"	lw	$7, -116(%0)	\n"
	"	lw	$8, -112(%0)	\n"
	"	lw	$9, -108(%0)	\n"
	"	lw	$10, -104(%0)	\n"
	"	lw	$11, -100(%0)	\n"
	"	sw	$2, -128(%1)	\n"
	"	sw	$3, -124(%1)	\n"
	"	sw	$6, -120(%1)	\n"
	"	sw	$7, -116(%1)	\n"
	"	sw	$8, -112(%1)	\n"
	"	sw	$9, -108(%1)	\n"
	"	sw	$10, -104(%1)	\n"
	"	sw	$11, -100(%1)	\n"
# endif
	"	daddiu	%0, %0, 32	\n"
	"	daddiu	%1, %1, 32	\n"
	"	bnel	%0, %2, 1b	\n"
	"	 pref	" SB1_PREF_LOAD_STREAMED_HINT  ",  -32(%0)\n"
	"	daddiu	%0, %0, -128	\n"
	"	daddiu	%1, %1, -128	\n"
#endif
#ifdef CONFIG_64BIT
	"	ld	$8, 0(%0)	\n"  /* Block copy a cacheline */
	"1:	ld	$9, 8(%0)	\n"
	"	ld	$10, 16(%0)	\n"
	"	ld	$11, 24(%0)	\n"
	"	sd	$8, 0(%1)	\n"
	"	sd	$9, 8(%1)	\n"
	"	sd	$10, 16(%1)	\n"
	"	sd	$11, 24(%1)	\n"
#else
	"	lw	$2, 0(%0)	\n"  /* Block copy a cacheline */
	"1:	lw	$3, 4(%0)	\n"
	"	lw	$6, 8(%0)	\n"
	"	lw	$7, 12(%0)	\n"
	"	lw	$8, 16(%0)	\n"
	"	lw	$9, 20(%0)	\n"
	"	lw	$10, 24(%0)	\n"
	"	lw	$11, 28(%0)	\n"
	"	sw	$2, 0(%1)	\n"
	"	sw	$3, 4(%1)	\n"
	"	sw	$6, 8(%1)	\n"
	"	sw	$7, 12(%1)	\n"
	"	sw	$8, 16(%1)	\n"
	"	sw	$9, 20(%1)	\n"
	"	sw	$10, 24(%1)	\n"
	"	sw	$11, 28(%1)	\n"
#endif
	"	daddiu	%0, %0, 32	\n"
	"	daddiu	%1, %1, 32	\n"
	"	bnel	%0, %2, 1b	\n"
#ifdef CONFIG_64BIT
	"	 ld	$8, 0(%0)	\n"
#else
	"	 lw	$2, 0(%0)	\n"
#endif
	"	.set	pop		\n"
	: "+r" (src), "+r" (dst)
	: "r" (end)
#ifdef CONFIG_64BIT
	: "$8","$9","$10","$11","memory");
#else
	: "$2","$3","$6","$7","$8","$9","$10","$11","memory");
#endif
}


#ifdef CONFIG_SIBYTE_DMA_PAGEOPS

/*
 * Pad descriptors to cacheline, since each is exclusively owned by a
 * particular CPU.
 */
typedef struct dmadscr_s {
	u64 dscr_a;
	u64 dscr_b;
	u64 pad_a;
	u64 pad_b;
} dmadscr_t;

static dmadscr_t page_descr[DM_NUM_CHANNELS]
	__attribute__((aligned(SMP_CACHE_BYTES)));

void sb1_dma_init(void)
{
	int i;

	for (i = 0; i < DM_NUM_CHANNELS; i++) {
		const u64 base_val = CPHYSADDR(&page_descr[i]) |
				     V_DM_DSCR_BASE_RINGSZ(1);
		void *base_reg = IOADDR(A_DM_REGISTER(i, R_DM_DSCR_BASE));

		__raw_writeq(base_val, base_reg);
		__raw_writeq(base_val | M_DM_DSCR_BASE_RESET, base_reg);
		__raw_writeq(base_val | M_DM_DSCR_BASE_ENABL, base_reg);
	}
}

void clear_page(void *page)
{
	u64 to_phys = CPHYSADDR(page);
	unsigned int cpu = smp_processor_id();

	/* if the page is not in KSEG0, use old way */
	if ((long)KSEGX(page) != (long)CKSEG0)
		return clear_page_cpu(page);

	page_descr[cpu].dscr_a = to_phys | M_DM_DSCRA_ZERO_MEM |
				 M_DM_DSCRA_L2C_DEST | M_DM_DSCRA_INTERRUPT;
	page_descr[cpu].dscr_b = V_DM_DSCRB_SRC_LENGTH(PAGE_SIZE);
	__raw_writeq(1, IOADDR(A_DM_REGISTER(cpu, R_DM_DSCR_COUNT)));

	/*
	 * Don't really want to do it this way, but there's no
	 * reliable way to delay completion detection.
	 */
	while (!(__raw_readq(IOADDR(A_DM_REGISTER(cpu, R_DM_DSCR_BASE_DEBUG)))
		 & M_DM_DSCR_BASE_INTERRUPT))
		;
	__raw_readq(IOADDR(A_DM_REGISTER(cpu, R_DM_DSCR_BASE)));
}

void copy_page(void *to, void *from)
{
	u64 from_phys = CPHYSADDR(from);
	u64 to_phys = CPHYSADDR(to);
	unsigned int cpu = smp_processor_id();

	/* if any page is not in KSEG0, use old way */
	if ((long)KSEGX(to) != (long)CKSEG0
	    || (long)KSEGX(from) != (long)CKSEG0)
		return copy_page_cpu(to, from);

	page_descr[cpu].dscr_a = to_phys | M_DM_DSCRA_L2C_DEST |
				 M_DM_DSCRA_INTERRUPT;
	page_descr[cpu].dscr_b = from_phys | V_DM_DSCRB_SRC_LENGTH(PAGE_SIZE);
	__raw_writeq(1, IOADDR(A_DM_REGISTER(cpu, R_DM_DSCR_COUNT)));

	/*
	 * Don't really want to do it this way, but there's no
	 * reliable way to delay completion detection.
	 */
	while (!(__raw_readq(IOADDR(A_DM_REGISTER(cpu, R_DM_DSCR_BASE_DEBUG)))
		 & M_DM_DSCR_BASE_INTERRUPT))
		;
	__raw_readq(IOADDR(A_DM_REGISTER(cpu, R_DM_DSCR_BASE)));
}

#else /* !CONFIG_SIBYTE_DMA_PAGEOPS */

void clear_page(void *page)
{
	return clear_page_cpu(page);
}

void copy_page(void *to, void *from)
{
	return copy_page_cpu(to, from);
}

#endif /* !CONFIG_SIBYTE_DMA_PAGEOPS */

EXPORT_SYMBOL(clear_page);
EXPORT_SYMBOL(copy_page);