Commit Graph

81 Commits

Author SHA1 Message Date
christos 6bc6175052 Don't record runlvl transitions before we have a read-write /var. Noticed
by lrou at x dot ua.
2006-09-28 15:20:14 +00:00
christos 6accae529c Error message consistency:
1. don't print the program name in the message [syslog does it]
2. no extraneous \n [syslog takes care of it]
3. quote strings as `%s' [we are using syslog]
4. use %m instead of error = %d
2006-09-25 19:42:04 +00:00
he fa8dee8628 Um, correct previous, so that current_state def/use is consistent. 2006-09-23 18:57:26 +00:00
he 54f712bd18 Since current_state is conditionally defined, make sure its uses are as well. 2006-09-23 14:04:26 +00:00
christos c34b6372d0 Try entertain who(1) because it has been sad since it was swallowed by the
Death Star: add utmpx entries for boot time, down time, run level, login
process.
2006-09-22 21:49:21 +00:00
elad 03917bd261 update comment, thanks freza@. 2006-07-30 20:01:26 +00:00
elad 8bc495196e CTLFLAG_READONLY1 was deprecated in earlier commit, remove it here too 2006-07-30 19:53:19 +00:00
elad 8ffd510349 remove CTLFLAG_READONLY1 usage. pointed out by cube@, thanks.. 2006-07-30 19:46:33 +00:00
salo b0b7590bc7 Add support for running multi-user in a chroot() environment.
How it works:

- after successful execution of /etc/rc, check the value of "init.root"
  sysctl node, if it's different than "/", chroot() into its value and run
  /etc/rc inside the chroot(),

- in single-user, return back to the original / file system.

Allows running with / file system on e.g., cgd(4), vnd(4) or ccd(4) volumes.

Idea first discussed with Matt Thomas, implemented by Jachym Holecek <freza
(at) liberouter.org> with some nitpicks by me.  Successfully used by me for
almost a year with / on a cgd(4) volume (for more information about the setup
check ftp://ftp.NetBSD.org/pub/NetBSD/misc/salo/init-chroot/ ).
2006-04-18 11:40:26 +00:00
rumble e948e1b17f Check for allocation failures in malloc, calloc, realloc, asprintf, and
vasprintf and try to handle them.
2006-03-17 15:53:46 +00:00
chris 4727225848 Fix case where argv wasn't returned or freed.
This fixes Coverity CID 2023.

While here also return NULL if the malloc fails.
2006-03-17 02:48:29 +00:00
christos 2c6eadc9ce Move WARNS=3 to the Makefile.inc, and add a little const to the remaining
programs that did not compile before.
2005-06-27 01:00:04 +00:00
xtraeme 83d6c074cb Bump the number of inodes to 1024 in the MFS case, needed to
create more devices in the ramdisk images, ok'ed by thorpej@.
2005-03-15 23:43:50 +00:00
dan 603344819a be explicit about the perms of mfs /dev 2004-10-12 10:08:09 +00:00
mycroft e0fc84bd24 Put back one of the sleep()s -- it doesn't affect us in the usual case. 2004-08-13 17:22:39 +00:00
mycroft f3b5a6849c Replace the sleep(2) in setctty() with a nanosleep(), with a default value of
250ms.
2004-08-13 17:19:15 +00:00
mycroft 40321c3f44 Remove a completely gratuitous 30 delay if you enter the wrong shell path. 2004-08-13 17:11:49 +00:00
christos e53725b6b7 clear_session_logs is not needed when we are small. 2004-06-06 01:42:20 +00:00
dsl 920dff4b7c If /dev/constty exists, use it instead of /dev/console for single user shell.
I'm not 100% certain it should call login_tty() for anything else.
2003-10-03 13:31:32 +00:00
dsl 58915e2274 If /dev/MAKEDEV is a symlink then preserve the symlink in the mfs /dev.
If /dev/MAKEDEV doesn't exist then try to run /etc/MAKEDEV instead.
Use the (new) '-n inodes' option to mount_mfs and base the size on the number
of inodes and the size of the MAKEDEV script.
2003-09-11 12:51:51 +00:00
agc 276d62f603 Move UCB-licensed code from 4-clause to 3-clause licence.
Patches provided by Joel Baker in PR 22308, verified by myself.
2003-08-07 10:04:22 +00:00
dsl 9b8239809b Exit child process if we fail to 'cd /dev' or 'execl sh MAKEDEV'.
Don't really want two processes in the rest of init!
2003-08-05 09:31:48 +00:00
itojun 09401b303a use asprintf, which is easier 2003-07-12 14:46:41 +00:00
christos 9dd592469e Avoid code duplication by factoring out the open/map open/write/unmap functions. 2003-07-01 16:44:48 +00:00
lukem b96069c077 Always compile in support for attempting to obtain the major device
number of the console from sysctl("machdep.console_device"), rather
than making in optional on the (unused) #ifdef DEBUG.
Certain platforms do not have the console at makedev(0,0) (including
many arm32 and some sh3 platforms), so the "mfs MAKEDEV" hack would
have failed on those.

Noted in private discussion with Quentin Barnes.
2003-07-01 15:24:22 +00:00
lukem 35c2bddeea If RESCUEDIR is defined, use that as an alternate path to sh and mount_mfs,
and prepend RESCUEDIR to the default PATH.
(This replaces my _PATH_ALTSHELL mods from rev 1.52 of init.c)
2003-05-26 09:34:55 +00:00
christos c1c48aec57 PR/5544: Anders Magnusson: Init segfaults when hupped in single user.
- fix so that session_db is not de-referenced when null, and other
  possible null pointer dereferences.
- ignore hup and tstp in single user mode.
2003-04-20 17:16:31 +00:00
fvdl 9f959f87e7 Call MAKEDEV with "init" 2003-04-17 18:04:47 +00:00
simonb d9e49feb59 White space nit. 2002-10-04 13:19:05 +00:00
lukem 9f3703ab9a - use _PATH_ALTSHELL (which defaults to _PATH_BSHELL) as the default ALTSHELL
(as opposed to _PATH_BSHELL), to allow the ALTSHELL to be changed.
- print the full path to the default shell in the ALTSHELL prompt.
2002-08-24 01:02:51 +00:00
christos 0f38e42a7a fix typos that prevented us from cleaning utmpx. Noted by wiz 2002-08-02 14:03:22 +00:00
christos 06596adfbd add utmpx/wtmpx processing support. 2002-07-27 23:49:47 +00:00
mycroft e84a828ff1 In single-user, after taking a signal and having the `shell' die with
SIGKILL, go into an infinite loop (like in the runcom SIGTERM case), on the
assumption that reboot(8) is in the middle of taking the system down.
This fixes `panic: init died' from sysinst.
2002-04-16 01:59:37 +00:00
lukem 447e6b3ec5 crank size of mfs from 512 sectors to 768 sectors (large i386 MAKEDEV) 2002-02-03 12:55:26 +00:00
christos 62a708c06d KNF: - use %m always instead of sometimes using strerror()
- space problems in casts
     - don't cast 0, use NULL
     - don't have long lines
     - don't cast malloc
     - use warn/warnx instead of fprintf
     - fix bug where a failed mount_mfs would cause the program to continue.
     - pass lint
2002-01-26 19:50:25 +00:00
lukem d369b2c6a5 fix compilation problems on my alpha 2002-01-23 01:45:41 +00:00
abs c48a08886b Rename MSDOSFS_ROOT to MFS_DEV_IF_NO_CONSOLE, and rework:
- Always attempt mfs dev if missing /dev/console
    - Save and run both MAKEDEV and MAKEDEV.local
    - After creating the mfs dev, mknod() and freopen() /dev/console.
    - If MAKEDEV is missing do not fail out early, but still create the mfs,
      /dev/console.
    - If we hit any errors fail out to single user.
This change _only_ affects systems booting without /dev/console.
2002-01-21 15:57:40 +00:00
abs 83c1ea626b Increase the size of the dev mfs to at least enough for 1.5.2 i386.
bin/15235 by Gavan Fantom.
2002-01-16 18:30:57 +00:00
lukem 684735b97b complete conversion to ANSI C 2001-06-18 01:38:05 +00:00
lukem 37a54a1e9c deprecate LOG_ODELAY; it's the default 2001-01-10 03:01:41 +00:00
wiz 4b4b4ba4ae Fix typo in v1.38. 2000-12-30 15:32:19 +00:00
wiz f81de5856b sprinkle some const's 2000-12-30 15:03:23 +00:00
wiz 369f36b84f ANSIfy and de-__P() 2000-12-30 15:01:42 +00:00
wiz fb109fb5fa Handle NULL return value by fgets after shell prompt better. Also
improve '\n' slightly. Addresses bin/11847.
2000-12-30 14:46:21 +00:00
soren 7c2de1e4cf Unbreak compilation with -DLETS_GET_SMALL as in PR bin/9639 by Patrick Welche. 2000-03-19 23:21:46 +00:00
mycroft f5d7100e26 Nuke `extern int errno;' in code we compile with -Wstrict-prototypes. We get
the correct definition from errno.h.
2000-01-21 17:08:33 +00:00
drochner 85cbf55d16 Since our gcc doesn't warn about NULL format strings anymore, we can
fix the incorrect err(1, "%s", "") et al.
Closes PR bin/7592 by cgd.
1999-11-09 15:06:30 +00:00
tls 566ff31cb7 When downgrading from 'highly secure' mode (securelevel >= 2) to 'insecure' mode (securelevel 0) and the root password is set, always require it before giving a shell on the console. Reasoning: if an attacker has access to the machine console, he doesn't necessarily have access to the hardware itself; on a 'highly secure' machine, we may as well make his life hard. 1998-11-14 07:17:58 +00:00
thorpej cad7a502b7 Use sigset_t for signal masks. 1998-09-18 22:00:46 +00:00
mycroft 4462053ab3 Fill in missing (default) mmap(2) flags. 1998-02-20 09:27:18 +00:00