summaryrefslogtreecommitdiffstats
path: root/arch/score/lib/checksum.S
blob: 706157edc7d54698ee7cd8748c2389f773edf512 (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
/*
 * arch/score/lib/csum_partial.S
 *
 * Score Processor version.
 *
 * Copyright (C) 2009 Sunplus Core Technology Co., Ltd.
 *  Lennox Wu <lennox.wu@sunplusct.com>
 *  Chen Liqin <liqin.chen@sunplusct.com>
 *
 * 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, see the file COPYING, or write
 * to the Free Software Foundation, Inc.,
 * 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 */
#include <linux/linkage.h>

#define ADDC(sum,reg)			\
	add	sum, sum, reg;		\
	cmp.c	reg, sum;		\
	bleu	9f;			\
	addi	sum, 0x1;		\
9:

#define CSUM_BIGCHUNK(src, offset, sum)		\
	lw	r8, [src, offset + 0x00];	\
	lw	r9, [src, offset + 0x04];	\
	lw	r10, [src, offset + 0x08];	\
	lw	r11, [src, offset + 0x0c];	\
	ADDC(sum, r8);				\
	ADDC(sum, r9);				\
	ADDC(sum, r10);				\
	ADDC(sum, r11);				\
	lw	r8, [src, offset + 0x10];	\
	lw	r9, [src, offset + 0x14];	\
	lw	r10, [src, offset + 0x18]; 	\
	lw	r11, [src, offset + 0x1c]; 	\
	ADDC(sum, r8);				\
	ADDC(sum, r9);				\
	ADDC(sum, r10);				\
	ADDC(sum, r11);				\

#define src r4
#define dest r5
#define sum r27

	.text
/* unknown src alignment and < 8 bytes to go */
small_csumcpy:
	mv	r5, r10
	ldi	r9, 0x0
	cmpi.c	r25, 0x1
	beq pass_small_set_t7	/*already set, jump to pass_small_set_t7*/
	andri.c	r25,r4 , 0x1	/*Is src 2 bytes aligned?*/

pass_small_set_t7:
	beq	aligned
	cmpi.c	r5, 0x0
	beq	fold
	lbu	r9, [src]
	slli	r9,r9, 0x8	/*Little endian*/
	ADDC(sum, r9)
	addi	src, 0x1
	subi.c	r5, 0x1

	/*len still a full word */
aligned:
	andri.c r8, r5, 0x4	/*Len >= 4?*/
	beq	len_less_4bytes

	/* Still a full word (4byte) to go,and the src is word aligned.*/
	andri.c	r8, src, 0x3	/*src is 4bytes aligned, so use LW!!*/
	beq	four_byte_aligned
	lhu 	r9, [src]
	addi	src, 2
	ADDC(sum, r9)
	lhu 	r9, [src]
	addi	src, 2
	ADDC(sum, r9)
	b len_less_4bytes

four_byte_aligned:		/* Len >=4 and four byte aligned */
	lw	r9, [src]
	addi	src, 4
	ADDC(sum, r9)

len_less_4bytes:		/* 2 byte aligned aligned and length<4B */
	andri.c r8, r5, 0x2
	beq	len_less_2bytes
	lhu	r9, [src]
	addi	src, 0x2	/* src+=2 */
	ADDC(sum, r9)

len_less_2bytes:		/* len = 1 */
	andri.c r8, r5, 0x1
	beq 	fold		/* less than 2 and not equal 1--> len=0 -> fold */
	lbu	r9, [src]

fold_ADDC:
	ADDC(sum, r9)
fold:
	/* fold checksum */
	slli	r26, sum, 16
	add	sum, sum, r26
	cmp.c	r26, sum
	srli	sum, sum, 16
	bleu 	1f 		/* if r26<=sum */
	addi	sum, 0x1 	/* r26>sum */
1:
	/* odd buffer alignment? r25 was set in csum_partial */
	cmpi.c	r25, 0x0
	beq	1f
	slli	r26, sum, 8
	srli	sum, sum, 8
	or	sum, sum, r26
	andi	sum, 0xffff
1:
	.set	optimize
	/* Add the passed partial csum. */
	ADDC(sum, r6)
	mv	r4, sum
	br	r3
	.set	volatile

	.align	5
ENTRY(csum_partial)
	ldi sum, 0
	ldi r25, 0
	mv r10, r5
	cmpi.c	r5, 0x8
	blt	small_csumcpy		/* < 8(singed) bytes to copy */
	cmpi.c	r5, 0x0
	beq	out
	andri.c	r25, src, 0x1		/* odd buffer? */

	beq	word_align
hword_align:				/* 1 byte */
	lbu	r8, [src]
	subi	r5, 0x1
	slli	r8, r8, 8
	ADDC(sum, r8)
	addi	src, 0x1

word_align:				/* 2 bytes */
	andri.c r8, src, 0x2		/* 4bytes(dword)_aligned? */
	beq	dword_align		/* not, maybe dword_align */
	lhu	r8, [src]
	subi	r5, 0x2
	ADDC(sum, r8)
	addi	src, 0x2

dword_align:				/* 4bytes */
	mv 	r26, r5			/* maybe useless when len >=56 */
	ldi 	r8, 56
	cmp.c	r8, r5
	bgtu	do_end_words		/* if a1(len)<t0(56) ,unsigned */
	andri.c	r26, src, 0x4
	beq	qword_align
	lw	r8, [src]
	subi	r5, 0x4
	ADDC(sum, r8)
	addi	src, 0x4

qword_align:				/* 8 bytes */
	andri.c r26, src, 0x8
	beq	oword_align
	lw	r8, [src, 0x0]
	lw	r9, [src, 0x4]
	subi	r5, 0x8			/* len-=0x8 */
	ADDC(sum, r8)
	ADDC(sum, r9)
	addi	src, 0x8

oword_align:				/* 16bytes */
	andri.c	r26, src, 0x10
	beq	begin_movement
	lw	r10, [src, 0x08]
	lw	r11, [src, 0x0c]
	lw	r8, [src, 0x00]
	lw	r9, [src, 0x04]
	ADDC(sum, r10)
	ADDC(sum, r11)
	ADDC(sum, r8)
	ADDC(sum, r9)
	subi	r5, 0x10
	addi	src, 0x10

begin_movement:
	srli.c	r26, r5, 0x7		/* len>=128? */
	beq	1f			/* len<128 */

/* r26 is the result that computed in oword_align */
move_128bytes:
	CSUM_BIGCHUNK(src, 0x00, sum)
	CSUM_BIGCHUNK(src, 0x20, sum)
	CSUM_BIGCHUNK(src, 0x40, sum)
	CSUM_BIGCHUNK(src, 0x60, sum)
	subi.c	r26, 0x01		/* r26 equals len/128 */
	addi	src, 0x80
	bne	move_128bytes

1:	/* len<128,we process 64byte here */
	andri.c	r10, r5, 0x40
	beq	1f

move_64bytes:
	CSUM_BIGCHUNK(src, 0x00, sum)
	CSUM_BIGCHUNK(src, 0x20, sum)
	addi	src, 0x40

1:					/* len<64 */
	andri	r26, r5, 0x1c		/* 0x1c=28 */
	andri.c	r10, r5, 0x20
	beq	do_end_words		/* decided by andri */

move_32bytes:
	CSUM_BIGCHUNK(src, 0x00, sum)
	andri	r26, r5, 0x1c
	addri	src, src, 0x20

do_end_words:				/* len<32 */
	/* r26 was set already in dword_align */
	cmpi.c	r26, 0x0
	beq	maybe_end_cruft		/* len<28 or len<56 */
	srli	r26, r26, 0x2

end_words:
	lw	r8, [src]
	subi.c	r26, 0x1		/* unit is 4 byte */
	ADDC(sum, r8)
	addi	src, 0x4
	cmpi.c	r26, 0x0
	bne	end_words		/* r26!=0 */

maybe_end_cruft:			/* len<4 */
	andri	r10, r5, 0x3

small_memcpy:
	mv	r5, r10
	j	small_csumcpy

out:
	mv	r4, sum
	br	r3

END(csum_partial)
OpenPOWER on IntegriCloud