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
|
#include <linux/types.h>
#include <linux/string.h>
#include <linux/sched.h>
#include <asm/i387.h>
#include <asm/hardirq.h>
/*
* MMX 3DNow! library helper functions
*
* To do:
* We can use MMX just for prefetch in IRQ's. This may be a win.
* (reported so on K6-III)
* We should use a better code neutral filler for the short jump
* leal ebx. [ebx] is apparently best for K6-2, but Cyrix ??
* We also want to clobber the filler register so we dont get any
* register forwarding stalls on the filler.
*
* Add *user handling. Checksums are not a win with MMX on any CPU
* tested so far for any MMX solution figured.
*
* 22/09/2000 - Arjan van de Ven
* Improved for non-egineering-sample Athlons
*
*/
void *_mmx_memcpy(void *to, const void *from, size_t len)
{
void *p;
int i;
if (in_interrupt())
return __memcpy(to, from, len);
p = to;
i = len >> 6; /* len/64 */
kernel_fpu_begin();
__asm__ __volatile__ (
"1: prefetch (%0)\n" /* This set is 28 bytes */
" prefetch 64(%0)\n"
" prefetch 128(%0)\n"
" prefetch 192(%0)\n"
" prefetch 256(%0)\n"
"2: \n"
".section .fixup, \"ax\"\n"
"3: movw $0x1AEB, 1b\n" /* jmp on 26 bytes */
" jmp 2b\n"
".previous\n"
".section __ex_table,\"a\"\n"
" .align 4\n"
" .long 1b, 3b\n"
".previous"
: : "r" (from) );
for(; i>0; i--)
{
__asm__ __volatile__ (
"1: prefetch 320(%0)\n"
"2: movq (%0), %%mm0\n"
" movq 8(%0), %%mm1\n"
" movq 16(%0), %%mm2\n"
" movq 24(%0), %%mm3\n"
" movq %%mm0, (%1)\n"
" movq %%mm1, 8(%1)\n"
" movq %%mm2, 16(%1)\n"
" movq %%mm3, 24(%1)\n"
" movq 32(%0), %%mm0\n"
" movq 40(%0), %%mm1\n"
" movq 48(%0), %%mm2\n"
" movq 56(%0), %%mm3\n"
" movq %%mm0, 32(%1)\n"
" movq %%mm1, 40(%1)\n"
" movq %%mm2, 48(%1)\n"
" movq %%mm3, 56(%1)\n"
".section .fixup, \"ax\"\n"
"3: movw $0x05EB, 1b\n" /* jmp on 5 bytes */
" jmp 2b\n"
".previous\n"
".section __ex_table,\"a\"\n"
" .align 4\n"
" .long 1b, 3b\n"
".previous"
: : "r" (from), "r" (to) : "memory");
from+=64;
to+=64;
}
/*
* Now do the tail of the block
*/
__memcpy(to, from, len&63);
kernel_fpu_end();
return p;
}
static void fast_clear_page(void *page)
{
int i;
kernel_fpu_begin();
__asm__ __volatile__ (
" pxor %%mm0, %%mm0\n" : :
);
for(i=0;i<4096/64;i++)
{
__asm__ __volatile__ (
" movntq %%mm0, (%0)\n"
" movntq %%mm0, 8(%0)\n"
" movntq %%mm0, 16(%0)\n"
" movntq %%mm0, 24(%0)\n"
" movntq %%mm0, 32(%0)\n"
" movntq %%mm0, 40(%0)\n"
" movntq %%mm0, 48(%0)\n"
" movntq %%mm0, 56(%0)\n"
: : "r" (page) : "memory");
page+=64;
}
/* since movntq is weakly-ordered, a "sfence" is needed to become
* ordered again.
*/
__asm__ __volatile__ (
" sfence \n" : :
);
kernel_fpu_end();
}
static void fast_copy_page(void *to, void *from)
{
int i;
kernel_fpu_begin();
/* maybe the prefetch stuff can go before the expensive fnsave...
* but that is for later. -AV
*/
__asm__ __volatile__ (
"1: prefetch (%0)\n"
" prefetch 64(%0)\n"
" prefetch 128(%0)\n"
" prefetch 192(%0)\n"
" prefetch 256(%0)\n"
"2: \n"
".section .fixup, \"ax\"\n"
"3: movw $0x1AEB, 1b\n" /* jmp on 26 bytes */
" jmp 2b\n"
".previous\n"
".section __ex_table,\"a\"\n"
" .align 4\n"
" .long 1b, 3b\n"
".previous"
: : "r" (from) );
for(i=0; i<4096/64; i++)
{
__asm__ __volatile__ (
"1: prefetch 320(%0)\n"
"2: movq (%0), %%mm0\n"
" movntq %%mm0, (%1)\n"
" movq 8(%0), %%mm1\n"
" movntq %%mm1, 8(%1)\n"
" movq 16(%0), %%mm2\n"
" movntq %%mm2, 16(%1)\n"
" movq 24(%0), %%mm3\n"
" movntq %%mm3, 24(%1)\n"
" movq 32(%0), %%mm4\n"
" movntq %%mm4, 32(%1)\n"
" movq 40(%0), %%mm5\n"
" movntq %%mm5, 40(%1)\n"
" movq 48(%0), %%mm6\n"
" movntq %%mm6, 48(%1)\n"
" movq 56(%0), %%mm7\n"
" movntq %%mm7, 56(%1)\n"
".section .fixup, \"ax\"\n"
"3: movw $0x05EB, 1b\n" /* jmp on 5 bytes */
" jmp 2b\n"
".previous\n"
".section __ex_table,\"a\"\n"
" .align 4\n"
" .long 1b, 3b\n"
".previous"
: : "r" (from), "r" (to) : "memory");
from+=64;
to+=64;
}
/* since movntq is weakly-ordered, a "sfence" is needed to become
* ordered again.
*/
__asm__ __volatile__ (
" sfence \n" : :
);
kernel_fpu_end();
}
/*
* Favour MMX for page clear and copy.
*/
static void slow_zero_page(void * page)
{
int d0, d1;
__asm__ __volatile__( \
"cld\n\t" \
"rep ; stosl" \
: "=&c" (d0), "=&D" (d1)
:"a" (0),"1" (page),"0" (1024)
:"memory");
}
void mmx_clear_page(void * page)
{
if(in_interrupt())
slow_zero_page(page);
else
fast_clear_page(page);
}
static void slow_copy_page(void *to, void *from)
{
int d0, d1, d2;
__asm__ __volatile__( \
"cld\n\t" \
"rep ; movsl" \
: "=&c" (d0), "=&D" (d1), "=&S" (d2) \
: "0" (1024),"1" ((long) to),"2" ((long) from) \
: "memory");
}
void mmx_copy_page(void *to, void *from)
{
if(in_interrupt())
slow_copy_page(to, from);
else
fast_copy_page(to, from);
}
|
