xref: /titanic_51/usr/src/lib/libc/sparcv9/gen/strcpy.s (revision a1d92fe4831144630aa0b262cde1629785c37f23)
1/*
2 * CDDL HEADER START
3 *
4 * The contents of this file are subject to the terms of the
5 * Common Development and Distribution License (the "License").
6 * You may not use this file except in compliance with the License.
7 *
8 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
9 * or http://www.opensolaris.org/os/licensing.
10 * See the License for the specific language governing permissions
11 * and limitations under the License.
12 *
13 * When distributing Covered Code, include this CDDL HEADER in each
14 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
15 * If applicable, add the following below this CDDL HEADER, with the
16 * fields enclosed by brackets "[]" replaced with your own identifying
17 * information: Portions Copyright [yyyy] [name of copyright owner]
18 *
19 * CDDL HEADER END
20 */
21
22/*
23 * Copyright 2008 Sun Microsystems, Inc.  All rights reserved.
24 * Use is subject to license terms.
25 */
26
27#pragma ident	"%Z%%M%	%I%	%E% SMI"
28
29	.file	"%M%"
30
31/*
32 * strcpy(s1, s2)
33 *
34 * Copy string s2 to s1.  s1 must be large enough. Return s1.
35 *
36 * Fast assembler language version of the following C-program strcpy
37 * which represents the `standard' for the C-library.
38 *
39 *	char *
40 *	strcpy(s1, s2)
41 *	register char *s1;
42 *	register const char *s2;
43 *	{
44 *		char *os1 = s1;
45 *
46 *		while(*s1++ = *s2++)
47 *			;
48 *		return(os1);
49 *	}
50 *
51 */
52
53#include <sys/asm_linkage.h>
54
55	! This implementation of strcpy works by first checking the
56	! source alignment and copying byte, half byte, or word
57	! quantities until the source ptr is aligned at an extended
58	! word boundary.  Once this has occurred, the string is copied,
59	! checking for zero bytes, depending upon its dst ptr alignment.
60	! (methods for xword, word, half-word, and byte copies are present)
61
62	ENTRY(strcpy)
63
64	.align 32
65
66	sub	%o1, %o0, %o3		! src - dst
67	andcc	%o1, 7, %o4		! dword aligned ?
68	bz,pn	%ncc, .srcaligned	! yup
69	mov	%o0, %o2		! save dst
70
71.chkbyte:
72	andcc	%o1, 1, %g0		! need to copy byte ?
73	bz,pn	%ncc, .chkhalfword	! nope, maybe halfword
74	sub	%g0, %o1, %g1		! %g1<2:0> = # of unaligned bytes
75	ldub	[%o2 + %o3], %o5	! src[0]
76	tst	%o5			! src[0] == 0 ?
77	stb	%o5, [%o2]		! dst[0] = src[0]
78	bz,pn	%ncc, .done		! yup, done
79	inc	%o2			! src++, dst++
80
81.chkhalfword:
82	andcc	%g1, 2, %g0		! need to copy half-word ?
83	bz,pn	%ncc, .chkword		! nope, maybe word
84	nop				!
85	lduh	[%o2 + %o3], %o5	! load src halfword
86	srl	%o5, 8, %o4		! extract first byte
87	tst	%o4			! first byte == 0 ?
88	bz,pn	%ncc, .done		! yup, done
89	stb	%o4, [%o2]		! store first byte
90	andcc	%o5, 0xff, %g0		! extract second byte
91	stb 	%o5, [%o2 + 1]		! store second byte
92	bz,pn	%ncc, .done		! yup, 2nd byte zero, done
93	add	%o2, 2, %o2		! src += 2
94
95.chkword:
96	andcc	%g1, 4, %g0		! need to copy word ?
97	bz,pn	%ncc, .srcaligned	! nope
98	nop				!
99	lduw	[%o2 + %o3], %o5	! load src word
100	srl	%o5, 24, %o4		! extract first byte
101	tst	%o4			! is first byte zero ?
102	bz,pn	%ncc, .done		! yup, done
103	stb	%o4, [%o2]		! store first byte
104	srl	%o5, 16, %o4		! extract second byte
105	andcc	%o4, 0xff, %g0		! is second byte zero ?
106	bz,pn	%ncc, .done		! yup, done
107	stb	%o4, [%o2 + 1]		! store second byte
108	srl	%o5, 8, %o4		! extract third byte
109	andcc	%o4, 0xff, %g0		! third byte zero ?
110	bz,pn	%ncc, .done		! yup, done
111	stb	%o4, [%o2 + 2]		! store third byte
112	andcc	%o5, 0xff, %g0		! fourth byte zero ?
113	stb 	%o5, [%o2 + 3]		! store fourth byte
114	bz,pn	%ncc, .done		! yup, fourth byte zero, done
115	add	%o2, 4, %o2		! src += 2
116
117.srcaligned:
118	sethi	%hi(0x01010101), %o4	! Alan Mycroft's magic1
119	or	%o4, %lo(0x01010101),%o4!  finish loading magic1
120	sllx	%o4, 32, %o1		! spread magic1
121	and	%o2, 3, %g4		! dst<1:0> to examine offset
122	or	%o4, %o1, %o4		!   to all 64 bits
123	cmp	%g4, 1			! dst offset of 1 or 5
124	sllx	%o4, 7, %o5		!  Alan Mycroft's magic2
125	be,pn	%ncc, .storebyte1241	! store 1, 2, 4, 1 bytes
126	cmp	%g4, 3			! dst offset of 3 or 7
127	be,pn	%ncc, .storebyte1421	! store 1, 4, 2, 1 bytes
128	cmp	%g4, 2			! dst halfword aligned ?
129	be,pn	%ncc, .storehalfword	! yup, store half-word wise
130	andcc	%o2, 7, %g0		! dst word aligned ?
131	bnz,pn	%ncc, .storeword2	! yup, store word wise
132	.empty
133
134.storedword:
135	ldx	[%o2 + %o3], %o1	! src dword
136	add	%o2, 8, %o2		! src += 8, dst += 8
137	andn	%o5, %o1, %g1		! ~dword & 0x8080808080808080
138	sub	%o1, %o4, %g4		! dword - 0x0101010101010101
139	andcc	%g4, %g1, %g0		! ((dword - 0x0101010101010101) & ~dword & 0x8080808080808080)
140	bz,a,pt	%ncc, .storedword	! no zero byte if magic expression == 0
141	stx	%o1, [%o2 - 8]		! store word to dst (address pre-incremented)
142
143.zerobyte:
144	orn	%o4, %g0, %o4		! 0xffffffffffffffff
145	sllx	%o4, 56, %o4		! 0xff00000000000000
146	srlx	%o1, 56, %o3		! %o3<7:0> = first byte
147	andcc	%o1, %o4, %g0		! first byte zero?
148	bz,pn	%ncc, .done		! yup, done
149	stb	%o3, [%o2 - 8]		! store first byte
150	srlx	%o4, 8, %o4		! 0x00ff000000000000
151	srlx	%o1, 48, %o3		! %o3<7:0> = second byte
152	andcc	%o1, %o4, %g0		! second byte zero?
153	bz,pn	%ncc, .done		! yup, done
154	stb	%o3, [%o2 - 7]		! store second byte
155	srlx	%o4, 8, %o4		! 0x0000ff0000000000
156	srlx	%o1, 40, %o3		! %o3<7:0> = third byte
157	andcc	%o1, %o4, %g0		! third byte zero?
158	bz,pn	%ncc, .done		! yup, done
159	stb	%o3, [%o2 - 6]		! store third byte
160	srlx	%o4, 8, %o4		! 0x000000ff00000000
161	srlx	%o1, 32, %o3		! %o3<7:0> = fourth byte
162	andcc	%o1, %o4, %g0		! fourth byte zero?
163	bz,pn	%ncc, .done		! yup, done
164	stb	%o3, [%o2 - 5]		! store fourth byte
165	srlx	%o4, 8, %o4		! 0x00000000ff000000
166	srlx	%o1, 24, %o3		! %o3<7:0> = fifth byte
167	andcc	%o1, %o4, %g0		! fifth byte zero?
168	bz,pn	%ncc, .done		! yup, done
169	stb	%o3, [%o2 - 4]		! store fifth byte
170	srlx	%o4, 8, %o4		! 0x0000000000ff0000
171	srlx	%o1, 16, %o3		! %o3<7:0> = sixth byte
172	andcc	%o1, %o4, %g0		! sixth byte zero?
173	bz,pn	%ncc, .done		! yup, done
174	stb	%o3, [%o2 - 3]		! store sixth byte
175	srlx	%o4, 8, %o4		! 0x000000000000ff00
176	andcc	%o1, %o4, %g0		! seventh byte zero?
177	srlx	%o1, 8, %o3		! %o3<7:0> = seventh byte
178	bz,pn	%ncc, .done		! yup, done
179	stb	%o3, [%o2 - 2]		! store seventh byte
180	stb	%o1, [%o2 - 1]		! store eigth byte
181.done:
182	retl				! done with leaf function
183
184	nop				! ensure following loop 16-byte aligned
185
186.storebyte1421:
187	ldx	[%o2 + %o3], %o1	! x = src[]
188	add	%o2, 8, %o2		! src += 8, dst += 8
189	andn	%o5, %o1, %g1		! ~x & 0x8080808080808080
190	sub	%o1, %o4, %g4		! x - 0x0101010101010101
191	andcc	%g4, %g1, %g0		! ((x - 0x0101010101010101) & ~x & 0x8080808080808080)
192	bnz,pn	%ncc, .zerobyte		! x has zero byte, handle end cases
193	srlx	%o1, 56, %g1		! %g1<7:0> = first byte; word aligned now
194	stb	%g1, [%o2 - 8]		! store first byte
195	srlx	%o1, 24, %g1		! %g1<31:0> = bytes 2, 3, 4, 5
196	stw	%g1, [%o2 - 7]		! store bytes 2, 3, 4, 5
197	srlx	%o1, 8, %g1		! %g1<15:0> = bytes 6, 7
198	sth	%g1, [%o2 - 3]		! store bytes 6, 7
199	ba	.storebyte1421		! next dword
200	stb	%o1, [%o2 - 1]		! store eigth byte
201
202	nop				! ensure following loop 16-byte aligned
203	nop				! ensure following loop 16-byte aligned
204
205.storebyte1241:
206	ldx	[%o2 + %o3], %o1	! x = src[]
207	add	%o2, 8, %o2		! src += 8, dst += 8
208	andn	%o5, %o1, %g1		! ~x & 0x8080808080808080
209	sub	%o1, %o4, %g4		! x - 0x0101010101010101
210	andcc	%g4, %g1, %g0		! ((x - 0x0101010101010101) & ~x & 0x8080808080808080)
211	bnz,pn	%ncc, .zerobyte		! x has zero byte, handle end cases
212	srlx	%o1, 56, %g1		! %g1<7:0> = first byte; word aligned now
213	stb	%g1, [%o2 - 8]		! store first byte
214	srlx	%o1, 40, %g1		! %g1<15:0> = bytes 2, 3
215	sth	%g1, [%o2 - 7]		! store bytes 2, 3
216	srlx	%o1, 8, %g1		! %g1<31:0> = bytes 4, 5, 6, 7
217	stw	%g1, [%o2 - 5]		! store bytes 4, 5, 6, 7
218	ba	.storebyte1241		! next dword
219	stb	%o1, [%o2 - 1]		! store eigth byte
220
221	nop				! ensure following loop 16-byte aligned
222	nop				! ensure following loop 16-byte aligned
223
224.storehalfword:
225	ldx	[%o2 + %o3], %o1	! x = src[]
226	add	%o2, 8, %o2		! src += 8, dst += 8
227	andn	%o5, %o1, %g1		! ~x & 0x8080808080808080
228	sub	%o1, %o4, %g4		! x - 0x0101010101010101
229	andcc	%g4, %g1, %g0		! ((x - 0x0101010101010101) & ~x & 0x8080808080808080)
230	bnz,pn	%ncc, .zerobyte		! x has zero byte, handle end cases
231	srlx	%o1, 48, %g1		! get first and second byte
232	sth	%g1, [%o2 - 8]		! store first and second byte; word aligned now
233	srlx	%o1, 16, %g1		! %g1<31:0> = bytes 3, 4, 5, 6
234	stw	%g1, [%o2 - 6]		! store bytes 3, 4, 5, 6
235	ba	.storehalfword		! next word
236	sth	%o1, [%o2 - 2]		! store seventh and eigth byte
237
238.storeword:
239	ldx	[%o2 + %o3], %o1	! x = src[]
240.storeword2:
241	add	%o2, 8, %o2		! src += 8, dst += 8
242	andn	%o5, %o1, %g1		! ~x & 0x0x8080808080808080
243	sub	%o1, %o4, %g4		! x - 0x0101010101010101
244	andcc	%g4, %g1, %g0		! ((x - 0x0101010101010101) & ~x & 0x8080808080808080)
245	bnz,pn	%ncc, .zerobyte		! x has zero byte, handle end cases
246	srlx	%o1, 32, %g1		! get bytes 1,2,3,4
247	stw	%g1, [%o2 - 8]		! store bytes 1,2,3,4 (address is pre-incremented)
248	ba	.storeword		! no zero byte if magic expression == 0
249	stw	%o1, [%o2 - 4]		! store bytes 5,6,7,8
250
251	nop				! padding, do not remove!!!
252	nop				! padding, do not remove!!!
253	SET_SIZE(strcpy)
254
255