summaryrefslogtreecommitdiffstats
path: root/arch/arm/lib/csumpartialcopygeneric.S
blob: 4a4609c19095f9188005e4e2e9b68d235008b079 (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
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
/*
 *  linux/arch/arm/lib/csumpartialcopygeneric.S
 *
 *  Copyright (C) 1995-2001 Russell King
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 */

/*
 * unsigned int
 * csum_partial_copy_xxx(const char *src, char *dst, int len, int sum, )
 *  r0 = src, r1 = dst, r2 = len, r3 = sum
 *  Returns : r0 = checksum
 *
 * Note that 'tst' and 'teq' preserve the carry flag.
 */

src	.req	r0
dst	.req	r1
len	.req	r2
sum	.req	r3

.Lzero:		mov	r0, sum
		load_regs	ea

		/*
		 * Align an unaligned destination pointer.  We know that
		 * we have >= 8 bytes here, so we don't need to check
		 * the length.  Note that the source pointer hasn't been
		 * aligned yet.
		 */
.Ldst_unaligned:
		tst	dst, #1
		beq	.Ldst_16bit

		load1b	ip
		sub	len, len, #1
		adcs	sum, sum, ip, put_byte_1	@ update checksum
		strb	ip, [dst], #1
		tst	dst, #2
		moveq	pc, lr			@ dst is now 32bit aligned

.Ldst_16bit:	load2b	r8, ip
		sub	len, len, #2
		adcs	sum, sum, r8, put_byte_0
		strb	r8, [dst], #1
		adcs	sum, sum, ip, put_byte_1
		strb	ip, [dst], #1
		mov	pc, lr			@ dst is now 32bit aligned

		/*
		 * Handle 0 to 7 bytes, with any alignment of source and
		 * destination pointers.  Note that when we get here, C = 0
		 */
.Lless8:	teq	len, #0			@ check for zero count
		beq	.Lzero

		/* we must have at least one byte. */
		tst	dst, #1			@ dst 16-bit aligned
		beq	.Lless8_aligned

		/* Align dst */
		load1b	ip
		sub	len, len, #1
		adcs	sum, sum, ip, put_byte_1	@ update checksum
		strb	ip, [dst], #1
		tst	len, #6
		beq	.Lless8_byteonly

1:		load2b	r8, ip
		sub	len, len, #2
		adcs	sum, sum, r8, put_byte_0
		strb	r8, [dst], #1
		adcs	sum, sum, ip, put_byte_1
		strb	ip, [dst], #1
.Lless8_aligned:
		tst	len, #6
		bne	1b
.Lless8_byteonly:
		tst	len, #1
		beq	.Ldone
		load1b	r8
		adcs	sum, sum, r8, put_byte_0	@ update checksum
		strb	r8, [dst], #1
		b	.Ldone

FN_ENTRY
		mov	ip, sp
		save_regs
		sub	fp, ip, #4

		cmp	len, #8			@ Ensure that we have at least
		blo	.Lless8			@ 8 bytes to copy.

		adds	sum, sum, #0		@ C = 0
		tst	dst, #3			@ Test destination alignment
		blne	.Ldst_unaligned		@ align destination, return here

		/*
		 * Ok, the dst pointer is now 32bit aligned, and we know
		 * that we must have more than 4 bytes to copy.  Note
		 * that C contains the carry from the dst alignment above.
		 */

		tst	src, #3			@ Test source alignment
		bne	.Lsrc_not_aligned

		/* Routine for src & dst aligned */

		bics	ip, len, #15
		beq	2f

1:		load4l	r4, r5, r6, r7
		stmia	dst!, {r4, r5, r6, r7}
		adcs	sum, sum, r4
		adcs	sum, sum, r5
		adcs	sum, sum, r6
		adcs	sum, sum, r7
		sub	ip, ip, #16
		teq	ip, #0
		bne	1b

2:		ands	ip, len, #12
		beq	4f
		tst	ip, #8
		beq	3f
		load2l	r4, r5
		stmia	dst!, {r4, r5}
		adcs	sum, sum, r4
		adcs	sum, sum, r5
		tst	ip, #4
		beq	4f

3:		load1l	r4
		str	r4, [dst], #4
		adcs	sum, sum, r4

4:		ands	len, len, #3
		beq	.Ldone
		load1l	r4
		tst	len, #2
		mov	r5, r4, get_byte_0
		beq	.Lexit
		adcs	sum, sum, r4, push #16
		strb	r5, [dst], #1
		mov	r5, r4, get_byte_1
		strb	r5, [dst], #1
		mov	r5, r4, get_byte_2
.Lexit:		tst	len, #1
		strneb	r5, [dst], #1
		andne	r5, r5, #255
		adcnes	sum, sum, r5, put_byte_0

		/*
		 * If the dst pointer was not 16-bit aligned, we
		 * need to rotate the checksum here to get around
		 * the inefficient byte manipulations in the
		 * architecture independent code.
		 */
.Ldone:		adc	r0, sum, #0
		ldr	sum, [sp, #0]		@ dst
		tst	sum, #1
		movne	r0, r0, ror #8
		load_regs	ea

.Lsrc_not_aligned:
		adc	sum, sum, #0		@ include C from dst alignment
		and	ip, src, #3
		bic	src, src, #3
		load1l	r5
		cmp	ip, #2
		beq	.Lsrc2_aligned
		bhi	.Lsrc3_aligned
		mov	r4, r5, pull #8		@ C = 0
		bics	ip, len, #15
		beq	2f
1:		load4l	r5, r6, r7, r8
		orr	r4, r4, r5, push #24
		mov	r5, r5, pull #8
		orr	r5, r5, r6, push #24
		mov	r6, r6, pull #8
		orr	r6, r6, r7, push #24
		mov	r7, r7, pull #8
		orr	r7, r7, r8, push #24
		stmia	dst!, {r4, r5, r6, r7}
		adcs	sum, sum, r4
		adcs	sum, sum, r5
		adcs	sum, sum, r6
		adcs	sum, sum, r7
		mov	r4, r8, pull #8
		sub	ip, ip, #16
		teq	ip, #0
		bne	1b
2:		ands	ip, len, #12
		beq	4f
		tst	ip, #8
		beq	3f
		load2l	r5, r6
		orr	r4, r4, r5, push #24
		mov	r5, r5, pull #8
		orr	r5, r5, r6, push #24
		stmia	dst!, {r4, r5}
		adcs	sum, sum, r4
		adcs	sum, sum, r5
		mov	r4, r6, pull #8
		tst	ip, #4
		beq	4f
3:		load1l	r5
		orr	r4, r4, r5, push #24
		str	r4, [dst], #4
		adcs	sum, sum, r4
		mov	r4, r5, pull #8
4:		ands	len, len, #3
		beq	.Ldone
		mov	r5, r4, get_byte_0
		tst	len, #2
		beq	.Lexit
		adcs	sum, sum, r4, push #16
		strb	r5, [dst], #1
		mov	r5, r4, get_byte_1
		strb	r5, [dst], #1
		mov	r5, r4, get_byte_2
		b	.Lexit

.Lsrc2_aligned:	mov	r4, r5, pull #16
		adds	sum, sum, #0
		bics	ip, len, #15
		beq	2f
1:		load4l	r5, r6, r7, r8
		orr	r4, r4, r5, push #16
		mov	r5, r5, pull #16
		orr	r5, r5, r6, push #16
		mov	r6, r6, pull #16
		orr	r6, r6, r7, push #16
		mov	r7, r7, pull #16
		orr	r7, r7, r8, push #16
		stmia	dst!, {r4, r5, r6, r7}
		adcs	sum, sum, r4
		adcs	sum, sum, r5
		adcs	sum, sum, r6
		adcs	sum, sum, r7
		mov	r4, r8, pull #16
		sub	ip, ip, #16
		teq	ip, #0
		bne	1b
2:		ands	ip, len, #12
		beq	4f
		tst	ip, #8
		beq	3f
		load2l	r5, r6
		orr	r4, r4, r5, push #16
		mov	r5, r5, pull #16
		orr	r5, r5, r6, push #16
		stmia	dst!, {r4, r5}
		adcs	sum, sum, r4
		adcs	sum, sum, r5
		mov	r4, r6, pull #16
		tst	ip, #4
		beq	4f
3:		load1l	r5
		orr	r4, r4, r5, push #16
		str	r4, [dst], #4
		adcs	sum, sum, r4
		mov	r4, r5, pull #16
4:		ands	len, len, #3
		beq	.Ldone
		mov	r5, r4, get_byte_0
		tst	len, #2
		beq	.Lexit
		adcs	sum, sum, r4
		strb	r5, [dst], #1
		mov	r5, r4, get_byte_1
		strb	r5, [dst], #1
		tst	len, #1
		beq	.Ldone
		load1b	r5
		b	.Lexit

.Lsrc3_aligned:	mov	r4, r5, pull #24
		adds	sum, sum, #0
		bics	ip, len, #15
		beq	2f
1:		load4l	r5, r6, r7, r8
		orr	r4, r4, r5, push #8
		mov	r5, r5, pull #24
		orr	r5, r5, r6, push #8
		mov	r6, r6, pull #24
		orr	r6, r6, r7, push #8
		mov	r7, r7, pull #24
		orr	r7, r7, r8, push #8
		stmia	dst!, {r4, r5, r6, r7}
		adcs	sum, sum, r4
		adcs	sum, sum, r5
		adcs	sum, sum, r6
		adcs	sum, sum, r7
		mov	r4, r8, pull #24
		sub	ip, ip, #16
		teq	ip, #0
		bne	1b
2:		ands	ip, len, #12
		beq	4f
		tst	ip, #8
		beq	3f
		load2l	r5, r6
		orr	r4, r4, r5, push #8
		mov	r5, r5, pull #24
		orr	r5, r5, r6, push #8
		stmia	dst!, {r4, r5}
		adcs	sum, sum, r4
		adcs	sum, sum, r5
		mov	r4, r6, pull #24
		tst	ip, #4
		beq	4f
3:		load1l	r5
		orr	r4, r4, r5, push #8
		str	r4, [dst], #4
		adcs	sum, sum, r4
		mov	r4, r5, pull #24
4:		ands	len, len, #3
		beq	.Ldone
		mov	r5, r4, get_byte_0
		tst	len, #2
		beq	.Lexit
		strb	r5, [dst], #1
		adcs	sum, sum, r4
		load1l	r4
		mov	r5, r4, get_byte_0
		strb	r5, [dst], #1
		adcs	sum, sum, r4, push #24
		mov	r5, r4, get_byte_1
		b	.Lexit