Commit Graph

105 Commits

Author SHA1 Message Date
d63dee4019 Add an option to make the shells not advertise their busybox nature 2001-10-19 00:22:23 +00:00
2d91deba45 Allow multiple shells to be enabled. 2001-08-01 17:21:35 +00:00
7467c8d3b6 Patch from vodz:
Changed email address
    cmdedit API change
    optimizations for traceroute and md5sum
    added a new shared create_icmp_socket() function
2001-07-12 20:26:32 +00:00
028b65b060 More thrashing about trying to make hush behave itself wrt job
control...
2001-06-28 01:10:11 +00:00
5c66d06104 Explicitly call setsid() 2001-06-26 23:16:31 +00:00
a90f20b1c3 Don't choke if /etc/profile is not available 2001-06-26 23:00:21 +00:00
78b0e379d7 Vladimir's last_patch_15 2001-06-26 02:06:08 +00:00
6c947d289f Fix the code for establishing a controlling terminal.
-Erik
2001-06-25 22:24:38 +00:00
52a97ca00c Some more cleanups. Of special importance, never free a pipe
that still has running jobs.  Instead, we ignore it and expect
it to be cleaned by the background job stuff.
 -Erik
2001-06-22 06:49:26 +00:00
c798b07762 Some cleanups for hush. Saves about 350bytes.
-Erik
2001-06-22 06:23:03 +00:00
04407e522b Another hush update from Larry:
Fixes the interaction between if/then/else/fi syntax and variables.
    I planned to do it right from the beginning, but my implementation
    was buggy.  Also adds the relevant test cases.  Also adds some old
    Matt Kraai variable test cases that got left out somehow.
2001-06-07 16:42:05 +00:00
817e73cb63 A patch from Larry to fix pathological things like '>""' 2001-06-06 17:56:09 +00:00
c616e53ca2 Don't close file descriptors when we are duplicating them. 2001-06-05 16:50:08 +00:00
4f6753e586 Cope gracefully when people type something, then hit ^C.
-Erik
2001-05-31 17:17:12 +00:00
a813afc24f Plug another memory leak 2001-05-24 16:19:36 +00:00
a15dc15f12 Plug another memory leak in hush. This one showed up when backticked
commands were run.  This one was sure a bugger to find.
 -Erik
2001-05-23 23:46:09 +00:00
bf7df04ec1 Fix a massive memory leak in the run_list_test() function.
Rename run_list_test() as free_pipe_list().
Rename run_pipe_test() as free_pipe().
 -Erik
2001-05-23 22:18:35 +00:00
bdd4eced84 Commit a patch by Larry Doolittle which fixes recursive calls to file_peek,
and improves some debugging messages.
2001-05-23 17:43:00 +00:00
57e6a49e5b Excellent. This patch from Larry fixes the behavior of hush
when builtins are included in pipes.
2001-05-22 22:34:51 +00:00
9978576f15 Fix this case. No, really this time.
unset FOO
    export FOO=bar
    FOO=baz
    echo "global env: " `env | grep ^FOO`
    echo "local env:  " `set | grep ^FOO`
 -Erik
2001-05-22 21:37:48 +00:00
aeb44c4da6 Fix a race. Sometimes by the time we got to checkjobs(), the
pipe struct had already been freed.  Return immediately if the
pipe is NULL.
2001-05-22 20:29:00 +00:00
816867858b Oops. Remove some debug noise I left in. 2001-05-22 19:23:35 +00:00
94ac244dea Updates from both Vladimir and Larry 2001-05-22 19:05:18 +00:00
004015e9c4 Patch from Lars Kellogg-Stedman <lars@larsshack.org> to fix bug
#1130 (i.e. When you turn on features it should always ADD features)
2001-05-21 20:30:51 +00:00
ada18ff0f5 The latest patch from Larry 2001-05-21 16:18:22 +00:00
9ffb7dd9a4 This is a patch from Vladimir:
> I rewrite *local_variable* function in hush.c with:
    > 1) remove many memory leaks
    > 2) add support read_only protect (require write builtin function for set this,
    > I write this special for variable HUSH_VERION=0.01)
    > 3) commad read set only local variable now
    > 4) remove many error messages if "set unset export" not defined variable
    > (bash syntax not put and set error code). Hmm, if I set result to -1, you hush
    > called waitpid and returned with error "no waitpid" ( i not found place this
    > error).
    > 5) destroy error in new version check xgetcwd()==NULL and set "(unknow)" -
    > this have error: crashe in next call `pwd`, but xgetcwd(not null) called
    > free(arg).
    > 6) next add integraion with libbb

Valdimir's patch missed two cases of local variable handling
    FOO=bar
    export FOO=baz
    unset FOO
and
    export FOO=bar
    FOO=baz
which were working before, so I fixed those two cases.
2001-05-19 03:00:46 +00:00
bfae2529b8 It turns out job control in both hush and lash was broken by the
signal handling in cmdedit.c.  Disabling it makes the shells behave
themselves again.  hush isn't quite there, but is getting close...
 -Erik
2001-05-17 00:14:27 +00:00
bc604a2f41 Move init of a bunch of globals into main so we are at least
a bit more reentrant.
 -Erik
2001-05-16 05:24:03 +00:00
f72f562b2f Fix the behavior of local shell variables to match that of bash and ash.
-Erik
2001-05-15 23:21:41 +00:00
15649c11f3 Oops. I left in some debug noise. 2001-05-15 17:28:26 +00:00
20a69a77d2 Write set_local_var() and fixup lookup_param() so you can now use
shell local variables.
 -Erik
2001-05-15 17:24:44 +00:00
78a7c99f7f This patch covers one big part of variable handling.
$ a=b foo
should be handled correctly.
$ a=b
is parsed OK, but the actual variable setting is not
yet written.  Except for some weird exceptions related
to quoting rules, this code passes (matches ash behavior)
all the tests I threw at it.

If someone now writes set_local_var(), and updates lookup_param()
to match, we can claim success!

       - Larry
2001-05-15 16:30:25 +00:00
cfa88ecb72 Fix up some silly warnings 2001-05-11 18:08:16 +00:00
5f265b755a Fix a segfault in lash, hush, and cmdedit. Each of these used
xgetcwd, but did not check the return for a NULL, and then continued
to call strlen on the NULL when the cwd had been removed from under it.
 -Erik
2001-05-11 16:58:46 +00:00
9d94deabd3 Use xgetcwd 2001-05-11 16:36:03 +00:00
1a6d39b259 A few more structural updates for job control.
-Erik
2001-05-08 05:11:54 +00:00
0a36de0580 Fix a stupid segfault caused by dereferencing a NULL ptr.
-Erik
2001-05-08 04:25:46 +00:00
45e92ba2ce This patch from Evin Robertson <nitfol@my-deja.com> fixes export so it works.
This way leaks memory, but oh well.  We will probably fix that when we get
around to doing local variables.
2001-05-07 23:53:09 +00:00
83a2ae2184 More hush updates from Larry:
Update some comments.  Generate partial placeholders for the missing
    builtins.  Write builtin_umask.  Properly treat exec without arguments
    as a means to open/close files within the running script.  Implement
    "4<&-" that encodes for file descriptor closure.
2001-05-07 17:59:25 +00:00
2dcfba7117 Larry's variant on Evin Robertson's fix to what I messed up
trying to make job control work.  This fix makes it so that
hush doesn't error out when init hasn't set up job control,
as in when init=/bin/sh
2001-05-04 22:13:37 +00:00
d24f878546 Don't reset parents context when we encounter an terminator. Fixes
if true; then if true; then true; fi; fi
2001-05-04 21:20:20 +00:00
e5ffb911e8 If it fails to run a builtin, let it fall through to
exec the real thing.
 -Erik
2001-05-04 16:30:29 +00:00
80abc45182 Move up waitpid error check so that errno isn't clobbered by tcsetpgrp. 2001-05-02 21:48:17 +00:00
0fcd447d59 Some more job control updates. It will now sucessfully background
stuff.  Good luck getting things back into the foreground though...
 -Erik
2001-05-02 20:12:03 +00:00
20a3069ec0 Don't treat newline as pipe separator when processing substitutions. 2001-05-02 17:52:49 +00:00
9f8caf134f Fix $(...) construction. Noted and patched by Larry Doolittle
<ldoolitt@recycle.lbl.gov>.  I'm just a patch whore.  :)
2001-05-02 16:26:12 +00:00
bafd94f154 First pass at getting job control working. This gets the infrastructure
in place, but something is still missing/wrong in there.  Testing with
'grep foo &' seems to put _hush_ into the background.  Pondering...
 -Erik
2001-05-02 16:11:59 +00:00
1c8a59ab95 Use the current pgrp, not pid, when resetting the foreground process group.
Fixes bizarre suspension when executing `echo`.
2001-05-02 15:37:09 +00:00
f162e7d09d Don't segfault if the first word is the empty string. 2001-05-02 14:48:48 +00:00
e67c3ce327 Another update from Larry:
Makes hush return code equal to that of the last command executed.
    Fixes the mode where input comes from a file.
2001-05-02 02:09:36 +00:00