Commit Graph

2076 Commits

Author SHA1 Message Date
Denys Vlasenko
fec2b1f79d ash: stage backported LINENO support as a separate patch
Looks biggish and not particularly useful, but may be easier to just eat
the impact if future backports from dash would be otherwise increasingly
difficult.

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-17 16:43:33 +02:00
Denys Vlasenko
048491fbdc hush: trivial code shrink in builtin_getopts
function                                             old     new   delta
builtin_getopts                                      368     363      -5

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-17 12:36:39 +02:00
Denys Vlasenko
9832bbaba9 ash: unset OPTARG if getopts exits 1, support OPTERR=0 behavior
function                                             old     new   delta
getoptscmd                                           522     547     +25

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-15 15:44:41 +02:00
Denys Vlasenko
4476c70301 ash,hush: comment and debug tweaks, no code changes
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-15 15:27:41 +02:00
Denys Vlasenko
0485b677d2 ash: one "current line = 1" might be missing, fix that
I'm not sure this is necessary, but dash has this init here.
Just in case, do it too.

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-14 19:46:56 +02:00
Denys Vlasenko
c2aea025bb ash: update testsuite (we now error out on ${#=})
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-14 15:01:32 +02:00
Denys Vlasenko
452cc1d9bd ash: [PARSER] Catch variable length expansions on non-existant specials
Upstream commit:

    Date: Thu, 30 Oct 2014 11:53:35 +0800
    [PARSER] Catch variable length expansions on non-existant specials

    Currently we only check special variable names that follow directly
    after $ or ${.  So errors such as ${#&} are not caught.  This patch
    fixes that by moving the is_special check to just before we print out
    the special variable name.

    Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

function                                             old     new   delta
readtoken1                                          2630    2635      +5

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-14 14:23:45 +02:00
Denys Vlasenko
007ce9f807 shell: tweak getopts tests, no code changes
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-13 02:59:00 +02:00
Denys Vlasenko
419db0391e hush: implement "silent" optstrings of ":opts"
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-11 17:21:14 +02:00
Denys Vlasenko
129e1ce72c hush: add a test which fails due to uclibc bug in getopt()
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-11 17:00:39 +02:00
Denys Vlasenko
d16e612c93 hush: fix redirect code (was using uninitialized variables)
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-11 15:41:39 +02:00
Denys Vlasenko
9a7d0a0191 shell: add OPTARG poisoning to getopt_optarg.tests
ash fails this!

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-11 02:37:48 +02:00
Denys Vlasenko
81f962f3df hush: teach getopts to set/unset OPTARG
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-11 02:05:21 +02:00
Denys Vlasenko
74d4058928 hush: getopts builtin
function                                             old     new   delta
builtin_getopts                                        -     271    +271
bltins1                                              372     384     +12
------------------------------------------------------------------------------
(add/remove: 1/0 grow/shrink: 1/0 up/down: 283/0)             Total: 283 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-11 01:32:46 +02:00
Denys Vlasenko
4628945cd8 ash: fix "unset OPTIND" throwing an error message
Added test was failing quite severely. Now only one subtest fails
(OPTERR=0 has no effect).

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-11 00:59:36 +02:00
Denys Vlasenko
11f2e99c13 hush: optional times builtin
function                                             old     new   delta
builtin_times                                          -     108    +108
bltins1                                              360     372     +12
static.times_tbl                                       -       9      +9
------------------------------------------------------------------------------
(add/remove: 2/0 grow/shrink: 1/0 up/down: 129/0)             Total: 129 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-10 16:34:03 +02:00
Denys Vlasenko
cde46f75cb shell: more efficient check for EOL in read
function                                             old     new   delta
shell_builtin_read                                  1334    1320     -14

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-09 14:04:07 +02:00
Denys Vlasenko
1f41c885fc hush: implement -d DELIM option for 'read'
The POSIX standard only requires the 'read' builtin to handle '-r':
http://pubs.opengroup.org/onlinepubs/9699919799/utilities/read.html

However, Bash introduced the option '-d <DELIM>' to override IFS for
just one invocation, and it is quite useful.

We already support this in ash, let's add it to hush, too.

function                                             old     new   delta
builtin_read                                         263     284     +21
.rodata                                           163587  163589      +2
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/0 up/down: 23/0)               Total: 23 bytes

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-09 13:52:36 +02:00
Johannes Schindelin
3bef5d89b0 ash: implement -d DELIM option for read
The POSIX standard only requires the read builtin to handle -r:
http://pubs.opengroup.org/onlinepubs/9699919799/utilities/read.html

However, Bash introduced the option -d <DELIM> to override IFS for
just one invocation, and it is quite useful.

It is also super easy to implement in BusyBox' ash, so let's do that.

The motivation: This option is used by Git's test suite.

function                                             old     new   delta
.rodata                                           163505  163587     +82
shell_builtin_read                                  1244    1289     +45
readcmd                                              233     259     +26
builtin_read                                         258     263      +5
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 4/0 up/down: 158/0)             Total: 158 bytes

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-09 13:52:17 +02:00
Denys Vlasenko
80e8e3cc05 noexec: consolidate code
function                                             old     new   delta
run_noexec_applet_and_exit                             -      61     +61
find_applet_by_name                                  128     124      -4
run_applet_no_and_exit                               441     434      -7
tryexec                                              169     152     -17
pseudo_exec_argv                                     338     321     -17
------------------------------------------------------------------------------
(add/remove: 1/0 grow/shrink: 0/6 up/down: 61/-48)             Total: 13 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-07 19:24:57 +02:00
Denys Vlasenko
c9c1ccc4ed noexec: do GETOPT_RESET() before entering APPLET_main()
hush -c 'yes | head -1' was not happy.

function                                             old     new   delta
tryexec                                              159     169     +10
pseudo_exec_argv                                     328     338     +10

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-07 18:59:35 +02:00
Denys Vlasenko
f2cf1cc716 noexec: set comm field for noexecs
function                                             old     new   delta
set_task_comm                                          -      18     +18
tryexec                                              152     159      +7
pseudo_exec_argv                                     321     328      +7
main                                                 106      97      -9
------------------------------------------------------------------------------
(add/remove: 1/0 grow/shrink: 3/2 up/down: 34/-13)             Total: 23 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-07 18:45:33 +02:00
Denys Vlasenko
248a67fb75 free,stat: make NOEXEC
pkill/pgrep/pidof uncovered another quirk: what about noexec's _process names_?

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-07 18:18:09 +02:00
Denys Vlasenko
0f14f41e72 ash: do not set a signal to SIG_DFL if it already is
function                                             old     new   delta
setsignal                                            312     338     +26

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-06 20:06:19 +02:00
Denys Vlasenko
5c527dc57e make 17 state-changing execing applets (ex: "nice PROG ARGS") noexec
The applets with "<applet> [opts] PROG ARGS" API very quickly exec
another program, noexec is okay for them:

 chpst/envdir/envuidgid/softlimit/setuidgid
 chroot
 chrt
 ionice
 nice
 nohup
 setarch/linux32/linux64
 taskset
 cttyhack

"reset" and "sulogin" applets don't have this form, but also exec
another program at once, thus made noexec too.

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-04 19:55:01 +02:00
Denys Vlasenko
f8cdc7a2bc ash: BASH_XTRACEFD bashism
Based on patch by Johannes Schindelin <johannes.schindelin@gmx.de>

function                                             old     new   delta
evalcommand                                         1447    1500     +53

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-04 15:24:49 +02:00
Johannes Schindelin
7344755823 ash: remove no-longer-used variable
As of 035486c75 (ash: significant overhaul of redirect saving logic,
2017-07-31), the sv_pos variable is no longer used (just assigned to,
with no further effect).

Let's just remove it.

Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-04 14:52:08 +02:00
Denys Vlasenko
d329e34c96 ash: INT_OFF/INT_ON around run_nofork_applet()
function                                             old     new   delta
evalcommand                                         1441    1447      +6

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-04 14:50:03 +02:00
Denys Vlasenko
49e6bf2db9 sheel: improve comments on signal handling
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-04 14:28:16 +02:00
Denys Vlasenko
cfd392bea9 ash: add a fixme comment at run_nofork_applet
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-03 19:56:29 +02:00
Denys Vlasenko
19c9f31af1 nofork: fix a bug uncovered by hush testsuite (forgotten fflush)
function                                             old     new   delta
run_nofork_applet                                    280     287      +7

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-03 19:52:47 +02:00
Denys Vlasenko
39701204cf hush: do not accept "if() { echo; }" function def
function                                             old     new   delta
parse_stream                                        2634    2692     +58
msg_and_die_if_script                                  -      21     +21
syntax_error_unexpected_ch                            41      46      +5
syntax_error_at                                       14      18      +4
die_if_script                                         31      28      -3
setup_redirects                                      319     308     -11
------------------------------------------------------------------------------
(add/remove: 1/0 grow/shrink: 3/2 up/down: 88/-14)             Total: 74 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-02 19:46:14 +02:00
Denys Vlasenko
84ea60ed65 line editing: make read_line_input() not take timeout param
It's almost always -1.

function                                             old     new   delta
read_line_input                                     3902    3912     +10
new_line_input_t                                      24      31      +7
pgetc                                                583     585      +2
save_command_ps_at_cur_history                        80      78      -2
read_line                                             76      74      -2
fgetc_interactive                                    246     244      -2
addLines                                              84      82      -2
doCommands                                          2226    2222      -4
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 3/5 up/down: 19/-12)              Total: 7 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-02 17:27:28 +02:00
Denys Vlasenko
dd4b446f76 hush: make SIGINT handling visually less confusing
$ echo $$
18448
$ echo $? <wait here, run "kill -INT 18448" in other shell><press enter>
   <=== NOTHING??
$

That empty line does not look right. After this patch:

$ echo $$
18448
$ echo $? <wait here, run "kill -INT 18448" in other shell><press enter>
^C
$

function                                             old     new   delta
fgetc_interactive                                    245     246      +1

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-02 16:52:12 +02:00
Denys Vlasenko
7c40ddd950 NOFORK fixes
"rm -i FILE" and "yes" can now be interrupted by ^C in hush.
This also now works:

$ usleep 19999999
^C
$ echo $?
130

function                                             old     new   delta
run_pipe                                            1668    1711     +43
pseudo_exec_argv                                     312     321      +9
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/0 up/down: 52/0)               Total: 52 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-02 16:37:39 +02:00
Denys Vlasenko
95f7953f2c do not use `a' quoting style in comments
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-08-02 14:26:33 +02:00
Denys Vlasenko
ec05df13b0 ash: align --login code with dash
Upstream commit:

    Date: Sun, 13 Jul 2008 22:34:50 +0800
    [OPTIONS] Added support for -l

    This patch adds support for the -l option (login shell) as required
    by the LSB.

    Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

It's a bit bigger, but gets rid of one global variable

function                                             old     new   delta
options                                              554     576     +22

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 19:43:47 +02:00
Denys Vlasenko
cee603d921 hush: remove redundant "G_flag_return_in_progress = -1"
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 18:06:07 +02:00
Denys Vlasenko
5b3d2eb327 hush: fix "true | func_with_return" not allowing return.
function                                             old     new   delta
pseudo_exec_argv                                     305     312      +7

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 18:02:28 +02:00
Denys Vlasenko
d0fff9155b ash: fix display of ">&-" redirect in job strings
function                                             old     new   delta
cmdtxt                                               558     569     +11

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 14:32:18 +02:00
Denys Vlasenko
75481d3634 hush: functions have priority over builtins (!)
function                                             old     new   delta
pseudo_exec_argv                                     291     305     +14
run_pipe                                            1560    1555      -5
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 1/1 up/down: 14/-5)               Total: 9 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 05:27:09 +02:00
Denys Vlasenko
bf1c344dfd hush: if STANDALONE, close interactive fd for NOEXECed children
function                                             old     new   delta
pseudo_exec_argv                                     291     305     +14

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 04:54:53 +02:00
Denys Vlasenko
32fdf2f9fc ash,hush: ">&10" redirects to script/tty fds should not work
The fact that shell has open fds to tty and/or scripts should be
unobservable, if possible. In particular, if redirect tries to dup
one of them via ">&script_fd", it's better to pretend that script_fd
is closed, and thus redirect fails with EBADF.

Fixes these two testcase failures:
ash-redir/redir_to_bad_fd.tests
hush-redir/redir_to_bad_fd3.tests

function                                             old     new   delta
redirect                                            1018    1129    +111
setup_redirects                                      250     359    +109
readtoken1                                          2651    2655      +4
cmdloop                                              185     187      +2
changepath                                           194     195      +1
save_fd_on_redirect                                  203     194      -9
evaltree                                             501     484     -17
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 5/2 up/down: 227/-26)           Total: 201 bytes
   text    data     bss     dec     hex filename
 914553     485    6848  921886   e111e busybox_old
 914754     485    6848  922087   e11e7 busybox_unstripped

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 04:35:18 +02:00
Denys Vlasenko
035486c750 ash: significant overhaul of redirect saving logic
New code is similar to what hush is doing.
Make CLOSED to -1: same as dash.
popredir() loses "restore" parameter: same as dash.
COPYFD_RESTORE bit is no longer necessary.

This change fixes this interactive bug:

	$ ls -l /proc/$$/fd 10>&-
	ash: can't set tty process group: Bad file descriptor
	ash: can't set tty process group: Bad file descriptor
	[1]+  Done(2)                    ls -l /proc/${\$}/fd 10>&4294967295

function                                             old     new   delta
unwindredir                                           29      27      -2
tryexec                                              154     152      -2
evaltree                                             503     501      -2
evalcommand                                         1369    1367      -2
cmdloop                                              187     185      -2
redirect                                            1029    1018     -11
popredir                                             153     123     -30
need_to_remember                                      36       -     -36
is_hidden_fd                                          68       -     -68
------------------------------------------------------------------------------
(add/remove: 0/2 grow/shrink: 0/7 up/down: 0/-155)           Total: -155 bytes
   text    data     bss     dec     hex filename
 914572     485    6848  921905   e1131 busybox_old
 914553     485    6848  921886   e111e busybox_unstripped

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 04:21:46 +02:00
Denys Vlasenko
657e9005a9 hush: massage redirect code to be slightly more like ash
function                                             old     new   delta
save_fd_on_redirect                                    -     203    +203
xdup_CLOEXEC_and_close                                 -      75     +75
setup_redirects                                      245     250      +5
xdup_and_close                                        72       -     -72
save_fds_on_redirect                                 221       -    -221
------------------------------------------------------------------------------
(add/remove: 2/2 grow/shrink: 1/0 up/down: 283/-293)          Total: -10 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 04:08:09 +02:00
Denys Vlasenko
d07a15bd1b ash: remove REDIR_SAVEFD2
function                                             old     new   delta
evalcommand                                         1364    1369      +5
redirect                                            1055    1014     -41
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 1/1 up/down: 5/-41)             Total: -36 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-31 04:05:51 +02:00
Denys Vlasenko
5f0a75f24b ash: if !ENABLE_ASH_EXPAND_PRMT, disable PSSYNTAX code
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-29 22:58:44 +02:00
Denys Vlasenko
1c79aeb6a8 ash: [REDIR] Fix popredir on abnormal exit from built-in
Upstream commit:

    Date: Thu, 27 May 2010 15:03:46 +0800
    [REDIR] Fix popredir on abnormal exit from built-in

    Just like the poplocalvar problem recently fixed, redirections
    can also be leaked in case of an abnormal exit.  This patch fixes
    it using the same method as poplocalvar, by storing the previous
    redirection state and restoring to that point.

    Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-29 22:51:52 +02:00
Denys Vlasenko
469998015f ash: [PARSER] Add FAKEEOFMARK for expandstr
Upstream commit:

    Date: Thu, 27 Dec 2007 13:54:16 +1100
    [PARSER] Add FAKEEOFMARK for expandstr

    Previously expandstr used the string "" to indicate that it needs to be
    treated just like a here-doc except that there is no terminator.  However,
    the string "" is in fact a valid here-doc terminator so now that we deal
    with it correctly expandstr no longer works in the presence of new-lines
    in the prompt.

    This patch introduces the FAKEEOFMARK macro which does not equal any
    real EOF marker but is distinct from the NULL pointer which is used to
    indicate non-here-doc contexts.

    Thanks to Markus Triska for reporting this regression.

    Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>

Unfortunately, I did not find the failing example for this old fix.

I also tweaked the code which was added by this commit:
"
Date: Mon Sep 24 18:30:02 2007 +0000
ash: fix prompt expansion (Natanael Copa <natanael.copa@gmail.com>)
"
since other parts of code do expect expandstr() to use DQSYNTAX, not PSSYNTAX.

function                                             old     new   delta
parse_stream                                        2609    2634     +25
setprompt_if                                         128     133      +5
read_profile                                          32      37      +5
evalcommand                                         1334    1339      +5
expandstr                                            122     120      -2
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 4/1 up/down: 40/-2)              Total: 38 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-29 21:12:29 +02:00
Denys Vlasenko
0f018b3070 hush: fix handling of empty heredoc EOF marker
function                                             old     new   delta
parse_stream                                        2609    2634     +25

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2017-07-29 20:43:26 +02:00