xref: /titanic_52/usr/src/lib/libc/i386/gen/strncat.s (revision 9512fe850e98fdd448c638ca63fdd92a8a510255)
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, Version 1.0 only
6 * (the "License").  You may not use this file except in compliance
7 * with the License.
8 *
9 * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
10 * or http://www.opensolaris.org/os/licensing.
11 * See the License for the specific language governing permissions
12 * and limitations under the License.
13 *
14 * When distributing Covered Code, include this CDDL HEADER in each
15 * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
16 * If applicable, add the following below this CDDL HEADER, with the
17 * fields enclosed by brackets "[]" replaced with your own identifying
18 * information: Portions Copyright [yyyy] [name of copyright owner]
19 *
20 * CDDL HEADER END
21 */
22/*
23 * Copyright 2004 Sun Microsystems, Inc.  All rights reserved.
24 * Use is subject to license terms.
25 */
26
27	.ident	"%Z%%M%	%I%	%E% SMI"
28
29	.file	"%M%"
30
31/
32/ strncat(s1, s2, n)
33/
34/ Concatenates s2 on the end of s1.  s1's space must be large enough.
35/ At most n characters are moved.
36/ Returns s1.
37/
38/ Fast assembly language version of the following C-program strncat
39/ which represents the `standard' for the C-library.
40/
41/	char *
42/	strncat(char *s1, const char *s2, size_t n)
43/	{
44/		char	*os1 = s1;
45/
46/		n++;
47/		while (*s1++)
48/			;
49/		--s1;
50/		while (*s1++ = *s2++)
51/			if (--n == 0) {
52/				s1[-1] = '\0';
53/				break;
54/			}
55/		return (os1);
56/	}
57/
58/ In this assembly language version, the following expression is used
59/ to check if a 32-bit word data contains a null byte or not:
60/	(((A & 0x7f7f7f7f) + 0x7f7f7f7f) | A) & 0x80808080
61/ If the above expression geneates a value other than 0x80808080,
62/ that means the 32-bit word data contains a null byte.
63/
64
65#include "SYS.h"
66
67	ENTRY(strncat)
68	pushl	%edi			/ save register variables
69	pushl	%esi
70	movl	12(%esp), %edi		/ %edi = destination string address
71	testl	$3, %edi		/ if %edi not word aligned
72	jnz	.L1			/ goto .L1
73	.align	4
74.L2:
75	movl	(%edi), %edx		/ move 1 word from (%edi) to %edx
76	movl	$0x7f7f7f7f, %ecx
77	andl	%edx, %ecx		/ %ecx = %edx & 0x7f7f7f7f
78	addl	$4, %edi		/ next word
79	addl	$0x7f7f7f7f, %ecx	/ %ecx += 0x7f7f7f7f
80	orl	%edx, %ecx		/ %ecx |= %edx
81	andl	$0x80808080, %ecx	/ %ecx &= 0x80808080
82	cmpl	$0x80808080, %ecx	/ if no null byte in this word
83	je	.L2			/ goto .L2
84	subl	$4, %edi		/ post-incremented
85.L1:
86	cmpb	$0, (%edi)		/ if a byte in (%edi) is null
87	je	.L3			/ goto .L3
88	incl	%edi			/ next byte
89	testl	$3, %edi		/ if %edi not word aligned
90	jnz	.L1			/ goto .L1
91	jmp	.L2			/ goto .L2 (%edi word aligned)
92	.align	4
93.L3:
94	/ %edi points to a null byte in destination string
95	movl	16(%esp), %eax		/ %eax = source string address
96	movl	20(%esp), %esi		/ %esi = number of bytes
97
98	testl	$3, %eax		/ if %eax not word aligned
99	jnz	.L4			/ goto .L4
100	cmpl	$4, %esi		/ if number of bytes < 4
101	jb	.L7			/ goto .L7
102	.align	4
103.L5:
104	movl	(%eax), %edx		/ move 1 word from (%eax) to %edx
105	movl	$0x7f7f7f7f, %ecx
106	andl	%edx, %ecx		/ %ecx = %edx & 0x7f7f7f7f
107	addl	$4, %eax		/ next word
108	addl	$0x7f7f7f7f, %ecx	/ %ecx += 0x7f7f7f7f
109	orl	%edx, %ecx		/ %ecx |= %edx
110	andl	$0x80808080, %ecx	/ %ecx &= 0x80808080
111	cmpl	$0x80808080, %ecx	/ if null byte in this word
112	jne	.L6			/ goto .L6
113	movl	%edx, (%edi)		/ copy this word to (%edi)
114	subl	$4, %esi		/ decrement number of bytes by 4
115	addl	$4, %edi		/ next word
116	cmpl	$4, %esi		/ if number of bytes >= 4
117	jae	.L5			/ goto .L5
118	jmp	.L7			/ goto .L7
119.L6:
120	subl	$4, %eax		/ post-incremented
121	.align	4
122.L7:
123	/ number of bytes < 4  or  a null byte found in the word
124	cmpl	$0, %esi		/ if number of bytes == 0
125	jz	.L8			/ goto .L8 (finished)
126	movb	(%eax), %dl		/ %dl = a byte in (%eax)
127	decl	%esi			/ decrement number of bytes by 1
128	movb	%dl, (%edi)		/ copy %dl to (%edi)
129	incl	%eax			/ next byte
130	incl	%edi			/ next byte
131	cmpb	$0, %dl			/ compare %dl with a null byte
132	je	.L9			/ if %dl is a null, goto .L9
133	jmp	.L7			/ goto .L7
134	.align	4
135
136.L4:
137	/ %eax not aligned
138	cmpl	$0, %esi		/ if number of bytes == 0
139	jz	.L8			/ goto .L8 (finished)
140	movb	(%eax), %dl		/ %dl = a byte in (%eax)
141	decl	%esi			/ decrement number of bytes by 1
142	movb	%dl, (%edi)		/ copy %dl to (%edi)
143	incl	%edi			/ next byte
144	incl	%eax			/ next byte
145	cmpb	$0, %dl			/ compare %dl with a null byte
146	je	.L9			/ if %dl is a null, goto .L9
147	jmp	.L4			/ goto .L4
148	.align	4
149.L8:
150	movb	$0, (%edi)		/ null termination
151.L9:
152	movl	12(%esp), %eax		/ return the destination address
153	popl	%esi			/ restore register variables
154	popl	%edi
155	ret
156	SET_SIZE(strncat)
157