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
|
/*
* This file contains assembly-language implementations
* of IP-style 1's complement checksum routines.
*
* Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org)
*
* 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.
*
* Severely hacked about by Paul Mackerras (paulus@cs.anu.edu.au).
*/
#include <linux/sys.h>
#include <asm/processor.h>
#include <asm/cache.h>
#include <asm/errno.h>
#include <asm/ppc_asm.h>
.text
/*
* computes the checksum of a memory block at buff, length len,
* and adds in "sum" (32-bit)
*
* csum_partial(buff, len, sum)
*/
_GLOBAL(csum_partial)
addic r0,r5,0
subi r3,r3,4
srwi. r6,r4,2
beq 3f /* if we're doing < 4 bytes */
andi. r5,r3,2 /* Align buffer to longword boundary */
beq+ 1f
lhz r5,4(r3) /* do 2 bytes to get aligned */
addi r3,r3,2
subi r4,r4,2
addc r0,r0,r5
srwi. r6,r4,2 /* # words to do */
beq 3f
1: mtctr r6
2: lwzu r5,4(r3) /* the bdnz has zero overhead, so it should */
adde r0,r0,r5 /* be unnecessary to unroll this loop */
bdnz 2b
andi. r4,r4,3
3: cmpwi 0,r4,2
blt+ 4f
lhz r5,4(r3)
addi r3,r3,2
subi r4,r4,2
adde r0,r0,r5
4: cmpwi 0,r4,1
bne+ 5f
lbz r5,4(r3)
slwi r5,r5,8 /* Upper byte of word */
adde r0,r0,r5
5: addze r3,r0 /* add in final carry */
blr
/*
* Computes the checksum of a memory block at src, length len,
* and adds in "sum" (32-bit), while copying the block to dst.
* If an access exception occurs on src or dst, it stores -EFAULT
* to *src_err or *dst_err respectively, and (for an error on
* src) zeroes the rest of dst.
*
* csum_partial_copy_generic(src, dst, len, sum, src_err, dst_err)
*/
#define CSUM_COPY_16_BYTES_WITHEX(n) \
8 ## n ## 0: \
lwz r7,4(r4); \
8 ## n ## 1: \
lwz r8,8(r4); \
8 ## n ## 2: \
lwz r9,12(r4); \
8 ## n ## 3: \
lwzu r10,16(r4); \
8 ## n ## 4: \
stw r7,4(r6); \
adde r12,r12,r7; \
8 ## n ## 5: \
stw r8,8(r6); \
adde r12,r12,r8; \
8 ## n ## 6: \
stw r9,12(r6); \
adde r12,r12,r9; \
8 ## n ## 7: \
stwu r10,16(r6); \
adde r12,r12,r10
#define CSUM_COPY_16_BYTES_EXCODE(n) \
.section __ex_table,"a"; \
.align 2; \
.long 8 ## n ## 0b,src_error; \
.long 8 ## n ## 1b,src_error; \
.long 8 ## n ## 2b,src_error; \
.long 8 ## n ## 3b,src_error; \
.long 8 ## n ## 4b,dst_error; \
.long 8 ## n ## 5b,dst_error; \
.long 8 ## n ## 6b,dst_error; \
.long 8 ## n ## 7b,dst_error; \
.text
.text
.stabs "arch/powerpc/lib/",N_SO,0,0,0f
.stabs "checksum_32.S",N_SO,0,0,0f
0:
CACHELINE_BYTES = L1_CACHE_BYTES
LG_CACHELINE_BYTES = L1_CACHE_SHIFT
CACHELINE_MASK = (L1_CACHE_BYTES-1)
_GLOBAL(csum_partial_copy_generic)
stwu r1,-16(r1)
stw r7,12(r1)
stw r8,8(r1)
andi. r0,r4,1 /* is destination address even ? */
cmplwi cr7,r0,0
addic r12,r6,0
addi r6,r4,-4
neg r0,r4
addi r4,r3,-4
andi. r0,r0,CACHELINE_MASK /* # bytes to start of cache line */
beq 58f
cmplw 0,r5,r0 /* is this more than total to do? */
blt 63f /* if not much to do */
andi. r8,r0,3 /* get it word-aligned first */
mtctr r8
beq+ 61f
li r3,0
70: lbz r9,4(r4) /* do some bytes */
addi r4,r4,1
slwi r3,r3,8
rlwimi r3,r9,0,24,31
71: stb r9,4(r6)
addi r6,r6,1
bdnz 70b
adde r12,r12,r3
61: subf r5,r0,r5
srwi. r0,r0,2
mtctr r0
beq 58f
72: lwzu r9,4(r4) /* do some words */
adde r12,r12,r9
73: stwu r9,4(r6)
bdnz 72b
58: srwi. r0,r5,LG_CACHELINE_BYTES /* # complete cachelines */
clrlwi r5,r5,32-LG_CACHELINE_BYTES
li r11,4
beq 63f
/* Here we decide how far ahead to prefetch the source */
li r3,4
cmpwi r0,1
li r7,0
ble 114f
li r7,1
#if MAX_COPY_PREFETCH > 1
/* Heuristically, for large transfers we prefetch
MAX_COPY_PREFETCH cachelines ahead. For small transfers
we prefetch 1 cacheline ahead. */
cmpwi r0,MAX_COPY_PREFETCH
ble 112f
li r7,MAX_COPY_PREFETCH
112: mtctr r7
111: dcbt r3,r4
addi r3,r3,CACHELINE_BYTES
bdnz 111b
#else
dcbt r3,r4
addi r3,r3,CACHELINE_BYTES
#endif /* MAX_COPY_PREFETCH > 1 */
114: subf r8,r7,r0
mr r0,r7
mtctr r8
53: dcbt r3,r4
54: dcbz r11,r6
/* the main body of the cacheline loop */
CSUM_COPY_16_BYTES_WITHEX(0)
#if L1_CACHE_BYTES >= 32
CSUM_COPY_16_BYTES_WITHEX(1)
#if L1_CACHE_BYTES >= 64
CSUM_COPY_16_BYTES_WITHEX(2)
CSUM_COPY_16_BYTES_WITHEX(3)
#if L1_CACHE_BYTES >= 128
CSUM_COPY_16_BYTES_WITHEX(4)
CSUM_COPY_16_BYTES_WITHEX(5)
CSUM_COPY_16_BYTES_WITHEX(6)
CSUM_COPY_16_BYTES_WITHEX(7)
#endif
#endif
#endif
bdnz 53b
cmpwi r0,0
li r3,4
li r7,0
bne 114b
63: srwi. r0,r5,2
mtctr r0
beq 64f
30: lwzu r0,4(r4)
adde r12,r12,r0
31: stwu r0,4(r6)
bdnz 30b
64: andi. r0,r5,2
beq+ 65f
40: lhz r0,4(r4)
addi r4,r4,2
41: sth r0,4(r6)
adde r12,r12,r0
addi r6,r6,2
65: andi. r0,r5,1
beq+ 66f
50: lbz r0,4(r4)
51: stb r0,4(r6)
slwi r0,r0,8
adde r12,r12,r0
66: addze r3,r12
addi r1,r1,16
beqlr+ cr7
rlwinm r3,r3,8,0,31 /* swap bytes for odd destination */
blr
/* read fault */
src_error:
lwz r7,12(r1)
addi r1,r1,16
cmpwi cr0,r7,0
beqlr
li r0,-EFAULT
stw r0,0(r7)
blr
/* write fault */
dst_error:
lwz r8,8(r1)
addi r1,r1,16
cmpwi cr0,r8,0
beqlr
li r0,-EFAULT
stw r0,0(r8)
blr
.section __ex_table,"a"
.align 2
.long 70b,src_error
.long 71b,dst_error
.long 72b,src_error
.long 73b,dst_error
.long 54b,dst_error
.text
/*
* this stuff handles faults in the cacheline loop and branches to either
* src_error (if in read part) or dst_error (if in write part)
*/
CSUM_COPY_16_BYTES_EXCODE(0)
#if L1_CACHE_BYTES >= 32
CSUM_COPY_16_BYTES_EXCODE(1)
#if L1_CACHE_BYTES >= 64
CSUM_COPY_16_BYTES_EXCODE(2)
CSUM_COPY_16_BYTES_EXCODE(3)
#if L1_CACHE_BYTES >= 128
CSUM_COPY_16_BYTES_EXCODE(4)
CSUM_COPY_16_BYTES_EXCODE(5)
CSUM_COPY_16_BYTES_EXCODE(6)
CSUM_COPY_16_BYTES_EXCODE(7)
#endif
#endif
#endif
.section __ex_table,"a"
.align 2
.long 30b,src_error
.long 31b,dst_error
.long 40b,src_error
.long 41b,dst_error
.long 50b,src_error
.long 51b,dst_error
|
