1 /*-
2 * SPDX-License-Identifier: BSD-2-Clause
3 *
4 * Copyright (c) 1993, David Greenman
5 * All rights reserved.
6 *
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
9 * are met:
10 * 1. Redistributions of source code must retain the above copyright
11 * notice, this list of conditions and the following disclaimer.
12 * 2. Redistributions in binary form must reproduce the above copyright
13 * notice, this list of conditions and the following disclaimer in the
14 * documentation and/or other materials provided with the distribution.
15 *
16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
20 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26 * SUCH DAMAGE.
27 */
28
29 #include <sys/param.h>
30 #include <sys/vnode.h>
31 #include <sys/proc.h>
32 #include <sys/sbuf.h>
33 #include <sys/systm.h>
34 #include <sys/sysproto.h>
35 #include <sys/exec.h>
36 #include <sys/imgact.h>
37 #include <sys/kernel.h>
38
39 #if BYTE_ORDER == LITTLE_ENDIAN
40 #define SHELLMAGIC 0x2123 /* #! */
41 #else
42 #define SHELLMAGIC 0x2321
43 #endif
44
45 /*
46 * At the time of this writing, MAXSHELLCMDLEN == PAGE_SIZE. This is
47 * significant because the caller has only mapped in one page of the
48 * file we're reading.
49 */
50 #if MAXSHELLCMDLEN > PAGE_SIZE
51 #error "MAXSHELLCMDLEN is larger than a single page!"
52 #endif
53
54 /*
55 * MAXSHELLCMDLEN must be at least MAXINTERP plus the size of the `#!'
56 * prefix and terminating newline.
57 */
58 CTASSERT(MAXSHELLCMDLEN >= MAXINTERP + 3);
59
60 /**
61 * Shell interpreter image activator. An interpreter name beginning at
62 * imgp->args->begin_argv is the minimal successful exit requirement.
63 *
64 * If the given file is a shell-script, then the first line will start
65 * with the two characters `#!' (aka SHELLMAGIC), followed by the name
66 * of the shell-interpreter to run, followed by zero or more tokens.
67 *
68 * The interpreter is then started up such that it will see:
69 * arg[0] -> The name of interpreter as specified after `#!' in the
70 * first line of the script. The interpreter name must
71 * not be longer than MAXSHELLCMDLEN bytes.
72 * arg[1] -> *If* there are any additional tokens on the first line,
73 * then we add a new arg[1], which is a copy of the rest of
74 * that line. The copy starts at the first token after the
75 * interpreter name. We leave it to the interpreter to
76 * parse the tokens in that value.
77 * arg[x] -> the full pathname of the script. This will either be
78 * arg[2] or arg[1], depending on whether or not tokens
79 * were found after the interpreter name.
80 * arg[x+1] -> all the arguments that were specified on the original
81 * command line.
82 *
83 * This processing is described in the execve(2) man page.
84 */
85
86 /*
87 * HISTORICAL NOTE: From 1993 to mid-2005, FreeBSD parsed out the tokens as
88 * found on the first line of the script, and setup each token as a separate
89 * value in arg[]. This extra processing did not match the behavior of other
90 * OS's, and caused a few subtle problems. For one, it meant the kernel was
91 * deciding how those values should be parsed (wrt characters for quoting or
92 * comments, etc), while the interpreter might have other rules for parsing.
93 * It also meant the interpreter had no way of knowing which arguments came
94 * from the first line of the shell script, and which arguments were specified
95 * by the user on the command line. That extra processing was dropped in the
96 * 6.x branch on May 28, 2005 (matching __FreeBSD_version 600029).
97 */
98 int
exec_shell_imgact(struct image_params * imgp)99 exec_shell_imgact(struct image_params *imgp)
100 {
101 const char *image_header = imgp->image_header;
102 const char *ihp, *interpb, *interpe, *maxp, *optb, *opte, *fname;
103 int error, offset;
104 size_t length;
105 struct vattr vattr;
106 struct sbuf *sname;
107
108 /* a shell script? */
109 if (((const short *)image_header)[0] != SHELLMAGIC)
110 return (-1);
111
112 /*
113 * Don't allow a shell script to be the shell for a shell
114 * script. :-)
115 */
116 if (imgp->interpreted & IMGACT_SHELL)
117 return (ENOEXEC);
118
119 imgp->interpreted |= IMGACT_SHELL;
120
121 /*
122 * At this point we have the first page of the file mapped.
123 * However, we don't know how far into the page the contents are
124 * valid -- the actual file might be much shorter than the page.
125 * So find out the file size.
126 */
127 error = VOP_GETATTR(imgp->vp, &vattr, imgp->proc->p_ucred);
128 if (error)
129 return (error);
130
131 /*
132 * Copy shell name and arguments from image_header into a string
133 * buffer.
134 */
135 maxp = &image_header[MIN(vattr.va_size, MAXSHELLCMDLEN)];
136 ihp = &image_header[2];
137
138 /*
139 * Find the beginning and end of the interpreter_name. If the
140 * line does not include any interpreter, or if the name which
141 * was found is too long, we bail out.
142 */
143 while (ihp < maxp && ((*ihp == ' ') || (*ihp == '\t')))
144 ihp++;
145 interpb = ihp;
146 while (ihp < maxp && ((*ihp != ' ') && (*ihp != '\t') && (*ihp != '\n')
147 && (*ihp != '\0')))
148 ihp++;
149 interpe = ihp;
150 if (interpb == interpe)
151 return (ENOEXEC);
152 if (interpe - interpb >= MAXINTERP)
153 return (ENAMETOOLONG);
154
155 /*
156 * Find the beginning of the options (if any), and the end-of-line.
157 * Then trim the trailing blanks off the value. Note that some
158 * other operating systems do *not* trim the trailing whitespace...
159 */
160 while (ihp < maxp && ((*ihp == ' ') || (*ihp == '\t')))
161 ihp++;
162 optb = ihp;
163 while (ihp < maxp && ((*ihp != '\n') && (*ihp != '\0')))
164 ihp++;
165 opte = ihp;
166 if (opte == maxp)
167 return (ENOEXEC);
168 while (--ihp > optb && ((*ihp == ' ') || (*ihp == '\t')))
169 opte = ihp;
170
171 if (imgp->args->fname != NULL) {
172 fname = imgp->args->fname;
173 sname = NULL;
174 } else {
175 sname = sbuf_new_auto();
176 sbuf_printf(sname, "/dev/fd/%d", imgp->args->fd);
177 sbuf_finish(sname);
178 fname = sbuf_data(sname);
179 }
180
181 /*
182 * We need to "pop" (remove) the present value of arg[0], and "push"
183 * either two or three new values in the arg[] list. To do this,
184 * we first shift all the other values in the `begin_argv' area to
185 * provide the exact amount of room for the values added. Set up
186 * `offset' as the number of bytes to be added to the `begin_argv'
187 * area, and 'length' as the number of bytes being removed.
188 */
189 offset = interpe - interpb + 1; /* interpreter */
190 if (opte > optb) /* options (if any) */
191 offset += opte - optb + 1;
192 offset += strlen(fname) + 1; /* fname of script */
193 length = (imgp->args->argc == 0) ? 0 :
194 strlen(imgp->args->begin_argv) + 1; /* bytes to delete */
195
196 error = exec_args_adjust_args(imgp->args, length, offset);
197 if (error != 0) {
198 if (sname != NULL)
199 sbuf_delete(sname);
200 return (error);
201 }
202
203 /*
204 * If there was no arg[0] when we started, then the interpreter_name
205 * is adding an argument (instead of replacing the arg[0] we started
206 * with). And we're always adding an argument when we include the
207 * full pathname of the original script.
208 */
209 if (imgp->args->argc == 0)
210 imgp->args->argc = 1;
211 imgp->args->argc++;
212
213 /*
214 * The original arg[] list has been shifted appropriately. Copy in
215 * the interpreter name and options-string.
216 */
217 length = interpe - interpb;
218 bcopy(interpb, imgp->args->begin_argv, length);
219 *(imgp->args->begin_argv + length) = '\0';
220 offset = length + 1;
221 if (opte > optb) {
222 length = opte - optb;
223 bcopy(optb, imgp->args->begin_argv + offset, length);
224 *(imgp->args->begin_argv + offset + length) = '\0';
225 offset += length + 1;
226 imgp->args->argc++;
227 }
228
229 /*
230 * Finally, add the filename onto the end for the interpreter to
231 * use and copy the interpreter's name to imgp->interpreter_name
232 * for exec to use.
233 */
234 error = copystr(fname, imgp->args->begin_argv + offset,
235 imgp->args->stringspace, NULL);
236
237 if (error == 0)
238 imgp->interpreter_name = imgp->args->begin_argv;
239
240 if (sname != NULL)
241 sbuf_delete(sname);
242 return (error);
243 }
244
245 /*
246 * Tell kern_execve.c about it, with a little help from the linker.
247 */
248 static struct execsw shell_execsw = {
249 .ex_imgact = exec_shell_imgact,
250 .ex_name = "#!"
251 };
252 EXEC_SET(shell, shell_execsw);
253