Commit Graph

2448 Commits

Author SHA1 Message Date
chs 83fd12c4fb add "uaddr" and improve text for "laddr". 2005-11-11 17:06:22 +00:00
chs 58c15fb5d1 add a keyword for the uarea address. very useful with crashdumps. 2005-11-05 19:27:24 +00:00
christos c952ef6b81 PR/31923: Sergey Svishchev: pax-as-tar ignores -k, overwrites existing files
Fix from Onno van der Linden
2005-11-01 23:43:27 +00:00
jschauma 0386433d98 Somehow missed this in the last commit, where I only changed the commentary
to reflect the actual change (plus minor nits).  So here's the actual change:

Rather than '-h' suppressing '-s' when using '-l', make it display
the blocks used in human readable form, as suggested by Alan Barrett
and Daniel Carosone.

(Missing bits spotted, as usual, by wiz@.)
2005-10-31 14:13:33 +00:00
elad 2fc1513987 The emulation for these should be "native", not "netbsd".
Noted by Ray Lai.
2005-10-26 21:17:23 +00:00
jschauma ea274fd69c Rather than '-h' suppressing '-s' when using '-l', make it display
the blocks used in human readable form, as suggested by Alan Barrett
and Daniel Carosone.
2005-10-26 02:24:22 +00:00
kleink fd4c891ff8 Remove the "" -> "." target conversion kludge. 2005-10-24 12:59:07 +00:00
jschauma 3478ebe108 document that -h makes -l suppress -s.
(Ie 'ls -lhs' == 'ls -lh')
2005-10-23 20:52:47 +00:00
elad c9981d258c PR/27140: "sleep time" doesn't error out.
Apply fix from Dheeraj Reddy.
2005-10-17 10:11:46 +00:00
christos 509c72ef94 - Don't output 'src -> dst' when -v and there was an error.
- Mark usage as noreturn, and remove extraneous exit.
- Pass lint.
Reported by Liam Foy.
2005-10-15 18:22:18 +00:00
jschauma ff54312f75 Honor the COLUMNS variable even when producing output not to a terminal,
as long as a specific format (that could use COLUMNS) was requested.
This makes

env COLUMNS=50 ls -C > file

not surprise the user who clearly wanted some specific output in the file.
2005-10-14 16:02:26 +00:00
christos 8d6ffce2ad PR/31548: Alexander Rigbo: $PWD not set when shell starts up, before pwd
or cd get executed.
2005-10-10 21:14:42 +00:00
wiz 803ccc7cb0 Bump date for previous. 2005-10-10 20:46:17 +00:00
jschauma 98615dcf47 Document that '-g' overrides -1, -C, -m, and -x, but is _always_
overridden by '-l', even if '-g' was specified last.
2005-10-10 20:30:34 +00:00
christos c05266af8a fix setmode error handling. 2005-10-01 20:23:54 +00:00
christos 16547af71a Fix setmode error handling. 2005-10-01 20:22:29 +00:00
christos 037f0a7b03 Now that setmode() is fixed, handle errors with err instead errx. 2005-10-01 20:09:18 +00:00
tv 964f98e703 Make slightly more portable without extra conditionals needed:
- Use getpgrp() rather than getpgid(0).
  (getpgrp() is already used elsewhere in this same file.)

- Use waitpid(-1, a, b) rather than wait3(a, b, NULL).
2005-09-30 18:57:55 +00:00
dsl 469d38dbae Don't leave arcn->org_name pointing to ftent->fts_path, if we being cpio
then ftent is freed just below.  Take a copy of the name and point org_name
at the copy.
Should fix PR/30627 (the fix in the PR will break pax and tar!)
2005-09-24 17:05:21 +00:00
christos 88ee54d8c2 Check for another race reported by chuq. 2005-09-18 12:15:41 +00:00
christos b39d90de0c mkdir(2) can return EISDIR for / 2005-09-17 06:05:04 +00:00
christos 5b36787563 spelling fixes from Joerg Sonnenberger 2005-09-16 17:05:14 +00:00
christos b60b306ef2 Another missing mkdir() race pointed out by chuq. Factor out the mkdir code. 2005-09-16 16:48:18 +00:00
christos 0743a9e845 use stat and not lstat to detect if we have a directory. refactor the mkpath()
code.
2005-09-13 20:09:55 +00:00
christos 929821126f when creating directories, check if the failure occured because someone
else created the directory before we did to avoid races. From chuq.
2005-09-13 15:50:17 +00:00
wiz 49c862f011 Use better section header. Sort sections. From YOMURA Masanori in private mail. 2005-09-11 23:21:17 +00:00
wiz 553e3c8477 Use better section header. From YOMURA Masanori in private mail. 2005-09-11 23:13:03 +00:00
christos a74ea5fcd7 Process window size changes after xread(). [change window size]
echo $COLUMS $LINES; will now print the new sizes immediately, instead
of after the second echo. From otto@OpenBSD
2005-09-11 22:23:42 +00:00
christos 57c8db64ce The following should print \"; from OpenBSD
$ cat << EOF
\"
EOF
2005-09-11 22:16:00 +00:00
wiz 2ea3362a10 In mdoc, use .Pp for new paragraphs, not .br. 2005-09-10 22:09:43 +00:00
wiz d8b6049e03 Sort options. Simplify macro usage. Use standard EXIT STATUS section. 2005-09-10 21:51:57 +00:00
elad d8a0e22d8a Adjust comment. Noted by Rui Paulo. 2005-09-07 19:15:49 +00:00
elad f777d75b5c Catch up with 64 bit inode changes. 2005-09-07 19:04:57 +00:00
elad e6a2d9e984 Simplify man-page a bit.
Patch from Patrick Welche.
2005-09-03 11:44:45 +00:00
elad 3752840791 Add a new keyword, ``topdir'', that grants access only if the file is
in a hierarchy below the specified path.
2005-08-24 19:09:03 +00:00
dsl 8c772bc707 Don't apply CDPATH if the the first component of the target is "." or "..".
Fixes PR/30973 and applies the principle of least surprise.
Update documentation to match (including date).
(matches behaviour of pdksh - if not it's documentation)
2005-08-20 21:07:42 +00:00
elad 76ed05e8b1 Add and document -N: When used with -p, it will not copy the file flags.
PR 13645.
2005-08-15 17:13:35 +00:00
elad 2d28a8933f Document -L correctly. PR #30430. 2005-08-13 15:59:28 +00:00
elad 3df38a6c89 Make inpath work like strstr again. :) 2005-08-10 21:53:01 +00:00
elad 9d164b22bc Add comments. 2005-08-10 21:33:36 +00:00
elad e4088913b5 Separate constructing the path to the policy away from systrace_addpolicy() and
into systrace_getpolicyfilename().
2005-08-10 18:19:21 +00:00
peter fa141339e0 More minor cleanup:
- remove unnecessary prototype for main()
- use symbolic exit code
- KNF & wrap a long line
2005-07-22 14:27:08 +00:00
hubertf 15bc25dc45 Minor code cleanup:
* remove unneeded cast
 * use err() instead of perror() + exit()
 * use symbolic exit code

Patch contributed by Liam J. Foy <liamfoy@sepulcrum.org> in private mail.
2005-07-22 11:06:24 +00:00
wiz 980a2652ee Aspell, fix an Xref, drop trailing whitespace. 2005-07-15 22:33:48 +00:00
christos 0df4f0ca31 Only do certain actions when we are trully interactive, not just when we
are connected to a tty:
	1. enable editor
	2. print directory on directory change.
	3. print job messages
2005-07-15 17:49:43 +00:00
christos b0d51313b4 If we have a script and we are interactive, set interactive to 2. This
allows us to run scripts with -i without printing the prompt like the
other shells do.
2005-07-15 17:46:54 +00:00
christos e996b4625d Allow trap to work on ignored signals when the shell is interactive. 2005-07-15 17:23:48 +00:00
christos ab760d2891 make setsig() return sig_t 2005-07-11 02:37:31 +00:00
christos 3564ac65d4 Don't hard ignore signals that were ignored by our environment, because
when we try to set a trap on them it will not work. Also while I am here:
1. don't change the action status if the signal system call failed.
2. don't try to sigignore it if signal failed.
3. clear the signal mask in case our parent blocked it for us.
2005-07-11 02:37:05 +00:00
elad 4e11de6548 Further correct handling of `inpath'.
The path in the policy is not normalized, and shouldn't be. We accept
only an absolute path, possibly with one trailing slash. Make note of
that in the manpage.
2005-07-04 16:32:30 +00:00
elad 931e7a5f52 Make `inpath' work as expected. Closes PR 29677.
Reported by Christian Biere, based on usenet post by John Wong.
2005-07-01 17:12:41 +00:00
elad 1d53f8bfa8 Use strcmp() when comparing strings in systrace.
Reported by Christian Biere in PR29676.
2005-07-01 16:12:11 +00:00
christos d7df71ba0e Don't compile in cpio usage, if we are SMALL. 2005-06-29 02:21:27 +00:00
elad 1b7133edad From marius@openbsd:
Fix an issue when scripts are executed under systrace where the argv[0]
would be normalized, and hence break scripts that depend on how they were
called.
2005-06-27 17:11:20 +00:00
christos 10a02919cc Add some more const so that this compiles again. 2005-06-27 00:46:04 +00:00
elad 8ebc2e9bee From marius@openbsd:
Add an exec message so that whenever a set-uid/gid process executes a new
image which we may control, the exec does not go by unnoticed.
2005-06-26 19:58:29 +00:00
christos 6310b59691 sprinkle a little const, and now everything compiles with WARNS=3 2005-06-26 19:10:48 +00:00
christos a397ec1f77 warns=3
- remove param names from function decls
- sprinkle const
- rename shadowed variables
- XXX: there is a lot of fishy __UNCONST going on.
2005-06-26 19:09:00 +00:00
elad 03d49eb07f From provos, via OpenBSD:
Prevent dereferencing possibly freed pointer.
2005-06-25 21:48:11 +00:00
elad 5cd24996d5 From marius@openbsd: (PR 3140)
If a string considered a filename is too long, don't exit but just let the
syscall fail, it might not actually be a filename.
2005-06-25 18:51:03 +00:00
elad b53ff10682 From rohee@openbsd:
Use strlcpy() instead of snprintf() where possible.
2005-06-25 18:47:42 +00:00
elad ddb0410ecc From deraadt@openbsd:
Wrap bind() to AF_UNIX with umask() to avoid race.
2005-06-25 18:41:30 +00:00
christos 671011e703 Simplify and fix the case where we have both known and unknown flags. 2005-06-25 14:37:33 +00:00
elad 8b57c1a70c Add translation for mmap()'s prot values. 2005-06-25 12:53:32 +00:00
elad cc0def8774 Add translation for mprotect()'s prot values. 2005-06-25 12:22:43 +00:00
elad 4d20624823 Improve readability. Order syscalls, add numbers where possible. 2005-06-25 12:17:57 +00:00
christos 250ff65369 Const poisoning. 2005-06-24 23:21:09 +00:00
christos b61c2197f1 Revert part of the previous commit. We cannot fix the problem by not waiting.
The problem is that the subshell code is not doing redirections properly.
2005-06-23 23:05:29 +00:00
christos 306d33d145 Don't wait for a background job in a subshell when we are set to EV_EXIT.
While I am here, call forkshell() explicitly FORK_FOO flags instead of
depending in FORK_FG == 0 and FORK_BG == 1.
2005-06-22 23:32:45 +00:00
hira f999ca1abd Sort options (description). 2005-06-17 14:40:47 +00:00
hira a2ed3bbe80 - Add -h to usage.
- Use set/getprogname(3).
2005-06-17 14:36:16 +00:00
hira dfa6df8e5a - Whitespace nit.
- Wrap long line.
2005-06-17 14:27:18 +00:00
he 0a5aea28f2 Initialize ipid in two functions to NULL, to appease -Wuninitialized.
Marked with XXXGCC for dreamcast (found while compiling for it).

Reviewed (in slightly different form) by lukem.
2005-06-07 09:00:18 +00:00
he 1d55d88929 Appease -Wuninitialized by initializing val and uval to 0.
Marked with XXXGCC for hpcarm (found while building for it).

Reviewed (in a slightly different form) by lukem.
2005-06-07 08:53:13 +00:00
chs 7bbdd188e1 appease gcc -Wuninitialized on hp700. 2005-06-05 19:08:28 +00:00
hubertf 2d4e6a01c0 the case of waitpid() returning an error was already tested, and errno
printed for that case. the tests for the return code don't have errno
set, so don't try to print it.

By Liam J. Foy <liamfoy sepulcrum.org>, posted to tech-userlevel@.
2005-06-03 13:55:04 +00:00
lukem 5166671bc3 appease gcc -Wuninitialized 2005-06-01 15:41:19 +00:00
lukem 50cde64b47 Mark temp_path volatile so that it won't get clobbered after longjmp.
(Also appeases gcc -Wuninitialized.)
2005-06-01 15:38:32 +00:00
lukem cfdf050ef3 appease gcc -Wuninitialized 2005-06-01 15:30:33 +00:00
lukem aab85a0fd4 appease gcc -Wuninitialized 2005-06-01 15:25:51 +00:00
lukem 5760da6393 appease gcc -Wuninitialized 2005-06-01 15:12:18 +00:00
wiz 6599ec1bca Bump date for -no-recursion. 2005-05-24 00:05:01 +00:00
wiz 24bf3d8ac3 Whitespace and punctuation fixes. 2005-05-24 00:03:52 +00:00
rillig 42627098be Fixed wrong use of the <ctype.h> functions by adding (unsigned char) casts. 2005-05-23 08:03:25 +00:00
rillig cb5d6856d5 Fixed another instance of isspace(char). 2005-05-23 08:01:17 +00:00
christos 4e0ae89b44 Don't even declare fdochroot() if we are in tool mode. 2005-05-22 17:41:50 +00:00
dsl d85801918d Set (and report) the new time (not the old one).
I broke it in rev 1.38, someone has finally noticed....
2005-05-15 18:39:39 +00:00
christos a1484a9949 be more lenient on the comparison to argv[0] to determine our personality. 2005-05-15 02:04:30 +00:00
christos 398a081644 Allow writing symlinks that point outside our tree, but not traversing
them to write other files.
2005-05-14 18:49:51 +00:00
pooka 2cf84fc283 Clarify the description for -s to include a pointer to an explanation
on how BLOCKSIZE affects the parameter.
2005-05-11 09:10:49 +00:00
christos c3b8cf7d21 Only add the function when we initialize. Restore the binding of tab to
ed-insert if tabcomplete is cleared.
2005-05-09 11:35:19 +00:00
christos 8dcba5679e PR/30167: J.T. Conklin: NetBSD tar does not support GNU tar --no-recursion flag 2005-05-07 22:27:33 +00:00
dsl 82537bf8e8 If 'set -o tabcomplete' it set, then bind <tab> to the libedit filename
completion function.
Note that the libedit code will probably want fine-tuning!
While editing the man page, add a note that non-whitespace IFS chars are
terminators and can generate null arguments.
2005-05-07 19:52:17 +00:00
provos d6497b197c provide translation for fcntl cmd names; requested by Nicolai Johannes 2005-05-07 15:11:02 +00:00
jmc 6ae4f78197 Wrap call to fchroot in CONFIG tests so this doesn't blow up on cross-tools 2005-05-06 16:49:25 +00:00
christos 0a6f2d385f PR/30132: Juan RP: tar --chroot refuses to extract files.
fchroot() changes the effective path, so we need to call updatepath().
2005-05-05 14:54:49 +00:00
christos 33cefcde7e PR/29005: Jesse Off: nbpax requires sys/mtio.h which doesn't exist on Interix
I added an ifdef called HAVE_MTIO_H and fixed the code to use it. It is
always defined now, but you can change this for Interix.
2005-05-01 02:59:28 +00:00
christos 13d863367b Pax longopts were not empty element terminated. 2005-04-24 03:39:26 +00:00
christos 14f615efd9 If we cannot resolve a path, that probably means that we have not created
directories for that path yet. So, do the check for each component recursively
and succeed if none of the components fall outside our current working
directory.
2005-04-24 03:36:54 +00:00
christos bc5cd54e43 PR/20228: Simon Burge: pax has problems reading a particular cpio archive
The problem here is that the archive is too short (< 512 bytes). The
buffer routines, try to read at least 512 bytes, even when we try to determine
what format file we have, which is wrong.
2005-04-24 03:26:03 +00:00