1 /*- 2 * Copyright (c) 2003-2004 Sean M. Kelly <smkelly@FreeBSD.org> 3 * All rights reserved. 4 * 5 * Redistribution and use in source and binary forms, with or without 6 * modification, are permitted provided that the following conditions 7 * are met: 8 * 1. Redistributions of source code must retain the above copyright 9 * notice, this list of conditions and the following disclaimer. 10 * 2. Redistributions in binary form must reproduce the above copyright 11 * notice, this list of conditions and the following disclaimer in the 12 * documentation and/or other materials provided with the distribution. 13 * 14 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 15 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 16 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 17 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 18 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 19 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 20 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 21 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 22 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 23 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 24 * SUCH DAMAGE. 25 */ 26 27 /* 28 * Software watchdog daemon. 29 */ 30 31 #include <sys/types.h> 32 __FBSDID("$FreeBSD$"); 33 34 #include <sys/mman.h> 35 #include <sys/param.h> 36 #include <sys/rtprio.h> 37 #include <sys/stat.h> 38 #include <sys/time.h> 39 #include <sys/watchdog.h> 40 41 #include <err.h> 42 #include <errno.h> 43 #include <fcntl.h> 44 #include <libutil.h> 45 #include <math.h> 46 #include <paths.h> 47 #include <signal.h> 48 #include <stdio.h> 49 #include <stdlib.h> 50 #include <string.h> 51 #include <sysexits.h> 52 #include <unistd.h> 53 54 static void parseargs(int, char *[]); 55 static void sighandler(int); 56 static void watchdog_loop(void); 57 static int watchdog_init(void); 58 static int watchdog_onoff(int onoff); 59 static int watchdog_patpat(u_int timeout); 60 static void usage(void); 61 62 static int debugging = 0; 63 static int end_program = 0; 64 static const char *pidfile = _PATH_VARRUN "watchdogd.pid"; 65 static u_int timeout = WD_TO_16SEC; 66 static u_int passive = 0; 67 static int is_daemon = 0; 68 static int fd = -1; 69 static int nap = 1; 70 static char *test_cmd = NULL; 71 72 /* 73 * Periodically pat the watchdog, preventing it from firing. 74 */ 75 int 76 main(int argc, char *argv[]) 77 { 78 struct rtprio rtp; 79 struct pidfh *pfh; 80 pid_t otherpid; 81 82 if (getuid() != 0) 83 errx(EX_SOFTWARE, "not super user"); 84 85 parseargs(argc, argv); 86 87 rtp.type = RTP_PRIO_REALTIME; 88 rtp.prio = 0; 89 if (rtprio(RTP_SET, 0, &rtp) == -1) 90 err(EX_OSERR, "rtprio"); 91 92 if (watchdog_init() == -1) 93 errx(EX_SOFTWARE, "unable to initialize watchdog"); 94 95 if (is_daemon) { 96 if (watchdog_onoff(1) == -1) 97 err(EX_OSERR, "patting the dog"); 98 99 pfh = pidfile_open(pidfile, 0600, &otherpid); 100 if (pfh == NULL) { 101 if (errno == EEXIST) { 102 errx(EX_SOFTWARE, "%s already running, pid: %d", 103 getprogname(), otherpid); 104 } 105 warn("Cannot open or create pidfile"); 106 } 107 108 if (debugging == 0 && daemon(0, 0) == -1) { 109 watchdog_onoff(0); 110 pidfile_remove(pfh); 111 err(EX_OSERR, "daemon"); 112 } 113 114 signal(SIGHUP, SIG_IGN); 115 signal(SIGINT, sighandler); 116 signal(SIGTERM, sighandler); 117 118 pidfile_write(pfh); 119 if (madvise(0, 0, MADV_PROTECT) != 0) 120 warn("madvise failed"); 121 122 watchdog_loop(); 123 124 /* exiting */ 125 pidfile_remove(pfh); 126 return (EX_OK); 127 } else { 128 if (passive) 129 timeout |= WD_PASSIVE; 130 else 131 timeout |= WD_ACTIVE; 132 if (watchdog_patpat(timeout) < 0) 133 err(EX_OSERR, "patting the dog"); 134 return (EX_OK); 135 } 136 } 137 138 /* 139 * Catch signals and begin shutdown process. 140 */ 141 static void 142 sighandler(int signum) 143 { 144 145 if (signum == SIGINT || signum == SIGTERM) 146 end_program = 1; 147 } 148 149 /* 150 * Open the watchdog device. 151 */ 152 static int 153 watchdog_init(void) 154 { 155 156 fd = open("/dev/" _PATH_WATCHDOG, O_RDWR); 157 if (fd >= 0) 158 return (0); 159 warn("Could not open watchdog device"); 160 return (-1); 161 } 162 163 /* 164 * Main program loop which is iterated every second. 165 */ 166 static void 167 watchdog_loop(void) 168 { 169 struct stat sb; 170 int failed; 171 172 while (end_program != 2) { 173 failed = 0; 174 175 if (test_cmd != NULL) 176 failed = system(test_cmd); 177 else 178 failed = stat("/etc", &sb); 179 180 if (failed == 0) 181 watchdog_patpat(timeout|WD_ACTIVE); 182 sleep(nap); 183 184 if (end_program != 0) { 185 if (watchdog_onoff(0) == 0) { 186 end_program = 2; 187 } else { 188 warnx("Could not stop the watchdog, not exitting"); 189 end_program = 0; 190 } 191 } 192 } 193 } 194 195 /* 196 * Reset the watchdog timer. This function must be called periodically 197 * to keep the watchdog from firing. 198 */ 199 static int 200 watchdog_patpat(u_int t) 201 { 202 203 return ioctl(fd, WDIOCPATPAT, &t); 204 } 205 206 /* 207 * Toggle the kernel's watchdog. This routine is used to enable and 208 * disable the watchdog. 209 */ 210 static int 211 watchdog_onoff(int onoff) 212 { 213 214 if (onoff) 215 return watchdog_patpat((timeout|WD_ACTIVE)); 216 else 217 return watchdog_patpat(0); 218 } 219 220 /* 221 * Tell user how to use the program. 222 */ 223 static void 224 usage(void) 225 { 226 if (is_daemon) 227 fprintf(stderr, "usage: watchdogd [-d] [-e cmd] [-I file] [-s sleep] [-t timeout]\n"); 228 else 229 fprintf(stderr, "usage: watchdog [-d] [-t timeout]\n"); 230 exit(EX_USAGE); 231 } 232 233 /* 234 * Handle the few command line arguments supported. 235 */ 236 static void 237 parseargs(int argc, char *argv[]) 238 { 239 int c; 240 char *p; 241 double a; 242 243 c = strlen(argv[0]); 244 if (argv[0][c - 1] == 'd') 245 is_daemon = 1; 246 while ((c = getopt(argc, argv, 247 is_daemon ? "I:de:s:t:?" : "dt:?")) != -1) { 248 switch (c) { 249 case 'I': 250 pidfile = optarg; 251 break; 252 case 'd': 253 debugging = 1; 254 break; 255 case 'e': 256 test_cmd = strdup(optarg); 257 break; 258 #ifdef notyet 259 case 'p': 260 passive = 1; 261 break; 262 #endif 263 case 's': 264 p = NULL; 265 errno = 0; 266 nap = strtol(optarg, &p, 0); 267 if ((p != NULL && *p != '\0') || errno != 0) 268 errx(EX_USAGE, "-s argument is not a number"); 269 break; 270 case 't': 271 p = NULL; 272 errno = 0; 273 a = strtod(optarg, &p); 274 if ((p != NULL && *p != '\0') || errno != 0) 275 errx(EX_USAGE, "-t argument is not a number"); 276 if (a < 0) 277 errx(EX_USAGE, "-t argument must be positive"); 278 if (a == 0) 279 timeout = WD_TO_NEVER; 280 else 281 timeout = 1.0 + log(a * 1e9) / log(2.0); 282 if (debugging) 283 printf("Timeout is 2^%d nanoseconds\n", 284 timeout); 285 break; 286 case '?': 287 default: 288 usage(); 289 /* NOTREACHED */ 290 } 291 } 292 if (argc != optind) 293 errx(EX_USAGE, "extra arguments."); 294 if (is_daemon && timeout < WD_TO_1SEC) 295 errx(EX_USAGE, "-t argument is less than one second."); 296 } 297