ubiattach: add option -O

Some images do not have the default VID offset. The option -O must
be used to attach such images.

Signed-off-by: Micke Prag <micke.prag@telldus.se>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

telnetd: we only need CONFIG_UNIX98_PTYS from the Linux kernel

(re-sending this as it got ignored completey and the format of the
previous mail was probably not correct - please let me know if there's
anything else I can do to get this trivial fix applied)

for telnetd to work, we only need CONFIG_UNIX98_PTYS to be enabled
in the Linux kernel - DEVPTS_FS has been obsolete for some time

Signed-off-by: Martin Kaiser <lists@kaiser.cx>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

dpkg: update supported compression methods

Based on a patch by Ron Yorston <rmy@tigress.co.uk>

function                                             old     new   delta
get_header_tar_xz                                      -      60     +60
filter_accept_list_reassign                          128     188     +60
unpack_package                                       585     621     +36
init_archive_deb_control                              52      76     +24
------------------------------------------------------------------------------
(add/remove: 2/0 grow/shrink: 3/0 up/down: 180/0)             Total: 180 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

platform.h: mempcpy needs <string.h>

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ntpd: get rid of G.initial_poll_complete

This speeds up syncing - now happens only just
two replies from a peer. Especially useful for "ntpd -q".
Shouldn't have ill effects: if we chose a bad peer,
we will discover it later and switch to another one.

The code is even smaller this way.

Suggested by Miroslav Lichvar <mlichvar@redhat.com>

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

dhcpd: fix an improper widening conversion

We wanted to detect when tv_sec = unsigned1 - unsigned2
underflows by looking at tv_sec's sign. But if tv_sec
is long and it is wider than unsigned, we get unsigned -> long
conversion which is in this case never negative.

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ftpd: prevent zombies. closes 7832

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

platform.h: add compat mempcpy

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

depmod: simple memory optimization

function                                             old     new   delta
filename2modname                                      67      86     +19
parse_module                                         374     351     -23

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

trim extra newlines in help texts

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

modprobe: fix modprobe -r and parsing of /etc/modprobe.d

This fixes a regression introduced with commit 78854520 (modprobe:
revert checking for /, stop doing basename() on modprobe args,
2015-01-01) that broke modprobe options stored in /etc/modprobe.conf and
/etc/modprobe.d/*.conf.

This also fixes modprobe -r

Signed-off-by: Natanael Copa <ncopa@alpinelinux.org>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

modprobe: fix "modprobe devname:snd/timer" case (alias with a slash)

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

modprobe: fix recent breakage: modules.dep reading code needs to strip dirname

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

modprobe-small: code shrink

function                                             old     new   delta
pathname_matches_modname                              80      42     -38

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

modprobe-small: fix recent breakage: modprobe -r was using full path

"modprobe -r MODNAME", after it found the full module pathname,
should strip dirpath before trying to remove the module.

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

diff: fix a bug in diffing against stdin. Closes 7784

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

vi: fix reading of file after last line

If the :r command is used to read a file after the last line of the
buffer the last line of the buffer and the first line of the file
are joined.  An extra blank line appears at the end of the buffer.

   file 1
   file 1
   file 1file 2
   file 2
   file 2

~
~

The insertion point is normally at the start of the line following the
specified line.  When the specified line is the last one the next_line
function baulks at moving to the non-existent following line.

Signed-off-by: Ron Yorston <rmy@tigress.co.uk>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

vi: failure to open file is not an error when initialising buffer

Commit 32afd3a introduced these regressions on the master branch:

Starting vi with no filename on the command line gives the status message
"'(null)' Bad address" instead of "- No file 1/1 100%".

Starting vi with a non-existent file on the command line gives the status
message "'new.txt' No such file or directory" instead of "- new.txt 1/1 100%"

Signed-off-by: Ron Yorston <rmy@tigress.co.uk>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

depmod: fix recent breakage in modules.dep generation

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: line counting in error meesages is broken, remove it

Signed-off-by: Tito Ragusa <farmatito@tiscali.it>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

swaponoff: add support for -e

Signed-off-by: René Rhéaume <rene.rheaume@gmail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

mpstat: code shrink

function                                             old     new   delta
get_uptime                                           134     129      -5
get_irqs_from_stat                                   110     103      -7

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

free: handle "cached" value too

function                                             old     new   delta
free_main                                            356     481    +125

Signed-off-by: Guillermo Rodriguez <guille.rodriguez@gmail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libbb: code shrink by factoring out common update_utmp_DEAD_PROCESS

function                                             old     new   delta
update_utmp_DEAD_PROCESS                               -      17     +17
telnetd_main                                        1685    1674     -11
mark_terminated                                       56      45     -11
handle_sigchld                                        74      63     -11

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

deleting unused ntpd_simple.c

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: fixes suggested by Tito, comment tweaks

function                                             old     new   delta
bb_internal_getpwent_r                               100     121     +21
parse_common                                         202     203      +1

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ntpd: relicense ntpd applet to ISC-style license

OpenNTPd is licensed under ISC-style license so it's good idea to keep
ntpd applet under same license to avoid mess, instead of having
our changes to be under GPL.

Names of original code's authors are added.

Signed-off-by: Adam Tkac <vonsch@gmail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: another code shrink

function                                             old     new   delta
massage_data_for_r_func                                -     110    +110
bb_internal_getpwent_r                               173     100     -73
getXXnam_r                                           176      95     -81
------------------------------------------------------------------------------
(add/remove: 1/0 grow/shrink: 0/2 up/down: 110/-154)          Total: -44 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: code shrink

function                                             old     new   delta
parse_common                                         206     205      -1
getgrouplist_internal                                229     225      -4
bb_internal_getpwent_r                               182     173      -9
getXXnam                                             232     222     -10
getXXnam_r                                           216     176     -40
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 0/5 up/down: 0/-64)             Total: -64 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: getgrXXX: empty string is not a member name, do not store it

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

fix a SEGV in previous commit

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: fix counting of commas in group list

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: store getXXnam result in a single malloc block

This saves a bit of memory but more importantly, allows to create
xmalloc_getpwnam() API where result can be deleted simply using free().

function                                             old     new   delta
getXXnam                                             134     173     +39
parse_common                                         188     212     +24
convert_to_struct                                    277     290     +13
get_S                                                 90      88      -2
tokenize                                             129     126      -3
bb_internal_getpwent_r                               175     172      -3
getXXnam_r                                           208     198     -10
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 3/4 up/down: 76/-18)             Total: 58 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: make db->def[] one byte shorter

In the future I will need another uint8_t, want to fit it
w/o using another word on 32 bits.

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: use a better estimate of max struct size

Previous code's trick with bitwise OR was giving this on 32-bit x86:

sizeof(struct passwd):28
sizeof(struct group):16
sizeof(struct spwd):36
sizeof(struct_result):60

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: reinstate Tito's optional atexit freeing

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: fix a memory leak in getXXnam (we did not save address of string buf)

function                                             old     new   delta
convert_to_struct                                    261     269      +8
const_sp_db                                           20      24      +4
const_pw_db                                           20      24      +4
const_gr_db                                           20      24      +4
tokenize                                             144     147      +3
parse_common                                         185     188      +3
get_S                                                 82      85      +3
bb_internal_getpwent_r                               188     185      -3
gr_off                                                 4       -      -4
getXXnam                                             171     165      -6
pw_off                                                 7       -      -7
getgrouplist_internal                                237     229      -8
getXXnam_r                                           215     207      -8
sp_off                                                 9       -      -9
------------------------------------------------------------------------------
(add/remove: 0/3 grow/shrink: 7/4 up/down: 29/-45)            Total: -16 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: tweak comments, replace one xmalloc with xzalloc

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: use FAST_FUNC to make "matching ABI" optimization more likely to succeed

See bb_internal_get*nam_r size reduction:

function                                             old     new   delta
bb_internal_getpwent_r                               167     176      +9
getXXnam_r                                           204     206      +2
sulogin_main                                         326     325      -1
su_main                                              471     470      -1
read_line_input                                     3832    3831      -1
print_stat                                           865     864      -1
prepare_socket_fd                                    283     282      -1
load_crontab                                         777     776      -1
fork_job                                             456     455      -1
do_shm                                               884     883      -1
do_sem                                               637     636      -1
do_msg                                               783     782      -1
complete_username                                    124     123      -1
bb_internal_getgrouplist                              71      70      -1
xgetpwuid                                             27      25      -2
xgetpwnam                                             27      25      -2
xgetgrnam                                             27      25      -2
xgetgrgid                                             27      25      -2
uid2uname                                             18      16      -2
login_main                                           980     978      -2
gid2group                                             18      16      -2
get_shell_name                                        54      52      -2
change_identity                                       50      48      -2
bb_internal_initgroups                                50      48      -2
argstr                                              1261    1259      -2
print_perms                                          177     174      -3
inetd_main                                          2077    2074      -3
run_applet_no_and_exit                               446     442      -4
fileaction_setowngrp                                  89      85      -4
deluser_main                                         312     308      -4
bb_internal_getpwuid                                  19      15      -4
bb_internal_getpwnam                                  11       7      -4
bb_internal_getgrnam                                  14      10      -4
bb_internal_getgrgid                                  19      15      -4
adduser_main                                         865     861      -4
passwd_main                                          989     984      -5
get_passwd                                            97      92      -5
data_extract_all                                     887     882      -5
check_user_passwd                                    490     485      -5
get_groups                                            81      75      -6
ftpd_main                                           2178    2171      -7
bb_internal_getspnam_r                                42      18     -24
bb_internal_getpwnam_r                                39      15     -24
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/41 up/down: 11/-153)         Total: -142 bytes
   text    data     bss     dec     hex filename
 923167     928   17676  941771   e5ecb busybox_old
 923023     928   17676  941627   e5e3b busybox_unstripped

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libpwdgrp: rewritten to use malloced implementation

This removed buffer size limitations.

function                                             old     new   delta
convert_to_struct                                      -     269    +269
getXXnam_r                                             -     204    +204
parse_common                                           -     185    +185
getXXnam                                               -     164    +164
tokenize                                               -     126    +126
bb_internal_getpwent_r                               102     167     +65
get_S                                                 30      88     +58
getgrouplist_internal                                195     240     +45
const_sp_db                                            -      20     +20
const_pw_db                                            -      20     +20
const_gr_db                                            -      20     +20
bb_internal_endpwent                                  27      36      +9
bb_internal_endgrent                                  27      36      +9
decode_one_format                                    726     734      +8
bb_internal_setpwent                                  17      24      +7
volume_id_probe_iso9660                              319     322      +3
scriptreplay_main                                    204     207      +3
mkfs_minix_main                                     2684    2687      +3
id_main                                              478     480      +2
hash_find                                            233     235      +2
pstree_main                                          321     322      +1
gr_off                                                 3       4      +1
expand_one_var                                      1579    1578      -1
pwf                                                    4       -      -4
grf                                                    4       -      -4
pack_gzip                                           1787    1780      -7
addattr32                                             67      56     -11
buffer_fill_and_print                                191     178     -13
dpkg_main                                           2944    2927     -17
bb_internal_setgrent                                  17       -     -17
bb_internal_getpwuid                                  38      19     -19
bb_internal_getgrgid                                  44      19     -25
bb_internal_getpwnam                                  38      11     -27
bb_internal_getgrnam                                  44      14     -30
bb_internal_fgetpwent_r                               51       -     -51
bb_internal_fgetgrent_r                               51       -     -51
bb_internal_getspnam_r                               121      42     -79
bb_internal_getpwnam_r                               121      39     -82
bb_internal_getgrent_r                               102       -    -102
bb__parsepwent                                       110       -    -110
bb_internal_getpwuid_r                               113       -    -113
bb_internal_getgrgid_r                               113       -    -113
bb__parsespent                                       120       -    -120
bb_internal_getgrnam_r                               121       -    -121
bb__pgsreader                                        213       -    -213
bb__parsegrent                                       226       -    -226
------------------------------------------------------------------------------
(add/remove: 8/13 grow/shrink: 14/11 up/down: 1224/-1556)    Total: -332 bytes
   text    data     bss     dec     hex filename
 923471     928   17684  942083   e6003 busybox_old
 923167     928   17676  941771   e5ecb busybox_unstripped

Signed-off-by: Tito Ragusa <farmatito@tiscali.it>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

modprobe: revert checking for /, stop doing basename() on modprobe args

function                                             old     new   delta
process_module                                       726     719      -7
filename2modname                                      81      67     -14
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 0/2 up/down: 0/-21)             Total: -21 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

Makefile: fix cscope target

This target doesn't work with current directory layout. Just make cscope
index all .c and .h files.

While we're at it: add cscope output files to .gitignore.

Signed-off-by: Bartosz Golaszewski <bartekgola@gmail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ash: fix a SEGV in ${#1}

function                                             old     new   delta
varvalue                                             760     805     +45
evalvar                                              648     603     -45

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

udhcpd: account for script delay in lease

Prevent nasty surprises if script runs longer than lease time / 2.

Signed-off-by: John Schroeder <jschroed@gmail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

Start 1.24.0 development cycle

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

Bump version to 1.23.0

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libbb: add sanity check in bb_arg_max()

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

randomconfig fixes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

Make it possible to override LOGIN_FAIL_DELAY in CONFIG_EXTRA_CFLAGS

Signed-off-by: Peter Kümmel <syntheticpp@gmx.net>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

udhcpc: account for script run time

Based on the following user report:

I ran into an issue where I was seeing a long delay in the scripts called
in udhcp_run_script. I was using an old version of OpenWrt (kamikaze)
and a satellite modem. An NTP script was being called and the modem
would sometimes take a long time to respond to the DNS lookup when
it was offline.

This delay started affecting my lease time. The lease that I would
get from my satellite modem before it was online would be short:
only 60 seconds. The delay with NTP and the modem would typically
be about 18 seconds. This would cause the first DHCP renew request
from dhcpc to be a little late. Under certain circumstances,
I could even see the first DHCP renew to occur after the lease
had expired!

function                                             old     new   delta
udhcpc_main                                         2816    2837     +21

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

nanddump: don't show --bb in usage when disabled

The --bb options now depends on LONG_OPTS. Omit mentions of --bb from usage
text when LONG_OPTS is disabled.

Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

nandwrite: fix build when long options are disabled

The Required_argument macro is only defined when long options are enabled.
Fixes the following build error:

miscutils/nandwrite.c: In function 'nandwrite_main':
miscutils/nandwrite.c:120:10: error: expected ',' or ';' before 'Required_argument'

Reported-by: Christian Kästner <kaestner at cs.cmu.edu>
Signed-off-by: Baruch Siach <baruch at tkos.co.il>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

lineedit: don't block when looking for escape sequence in vi-mode

In vi-mode lineedit tries to detect some escape sequences.
After the ESC it reads the next character to check for certain
values.  This read should have a timeout or a user-entered ESC to
switch to command mode doesn't properly handle the next character.

Signed-off-by: Ron Yorston <rmy@tigress.co.uk>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

tar: fix "tar -cJ" ignoring -J option. closes 7706

function                                             old     new   delta
tar_main                                             895     938     +43
vfork_compressor                                     206     191     -15
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 1/1 up/down: 43/-15)             Total: 28 bytes
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

lineedit: don't fall back to simple line input if tty is in raw mode

Testcase: shell command input after python ^Z should still work

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

Config: select PLATFORM_LINUX if using sendfile()

Man entry for sendfile:

  Not specified in POSIX.1-2001, or other standards.

  Other UNIX systems implement sendfile() with different  semantics  and
  prototypes. It should not be used in portable programs.

Select PLATFORM_LINUX if enabling FEATURE_USE_SENDFILE.

Signed-off-by: Bartosz Golaszewski <bartekgola@gmail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

flashcp: change BUFSIZE to 4k

some flash partitions can be smaller than the existing BUFSIZE thus write
BUFSIZE will fail with "no space left on device"

Signed-off-by: Jacob Kjaergaard <jacob.kjaergaard@prevas.dk>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

make xmalloc_open_zipped_read_close result NUL terminated

Compat with xmalloc_open_read_close

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libarchive: add capability to unpack to mem.buffer

The performance and number of processes for a "depmod -a" with gzipped
modules was abysmal. This patch adds a code path without fork,
benefiting all users of xmalloc_open_zipped_read_close.

"modinfo radeon.ko.gz", a single-file reader, got 30% faster.
"depmod -a", which used to fork over 800 times, got 20% faster.

Heavily based on a patch by Lauri Kasanen <curaga@operamail.com>

function                                             old     new   delta
setup_transformer_on_fd                                -     159    +159
transformer_write                                      -     122    +122
fork_transformer                                       -     112    +112
xmalloc_open_zipped_read_close                        63     118     +55
read_bunzip                                         1866    1896     +30
xtransformer_write                                     -      19     +19
unzip_main                                          2449    2462     +13
bbunpack                                             755     766     +11
unpack_lzma_stream                                  2717    2723      +6
unpack_xz_stream                                    2393    2397      +4
unpack_Z_stream                                     1173    1175      +2
inflate_unzip                                        111     105      -6
check_signature16                                     70      63      -7
unpack_bz2_stream                                    359     349     -10
unpack_unxz                                           12       -     -12
unpack_unlzma                                         12       -     -12
unpack_uncompress                                     12       -     -12
unpack_gunzip                                         12       -     -12
unpack_bunzip2                                        12       -     -12
open_transformer                                     106      92     -14
inflate_unzip_internal                              1945    1916     -29
unpack_gz_stream                                     693     655     -38
open_zipped                                           89      47     -42
setup_unzip_on_fd                                    142      53     -89
------------------------------------------------------------------------------
(add/remove: 4/5 grow/shrink: 7/8 up/down: 533/-295)          Total: 238 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

Rename transformer_aux_data_t -> transformer_state_t

No code changes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

man: do not mangle $MANPATH in memory

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libbb: use sendfile() to copy data between file descriptors

Busybox already uses sendfile in httpd. This patch proposes to use it
globally to copy data between file descriptors.

It speeds up the copying on slow systems a lot - below are the times needed
to copy a 450Mb file with and without this option enabled on a BeagleBone
Black:

sendfile:
user    0m0.000s
sys     0m8.170s

read/write 4k:
user    0m0.470s
sys     0m16.300s

function                                             old     new   delta
bb_full_fd_action                                    394     474     +80

Signed-off-by: Bartosz Golaszewski <bartekgola@gmail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

man: accept a list of dirs in $MANPATH

function                                             old     new   delta
add_MANPATH                                            -     143    +143
man_main                                             852     731    -121
------------------------------------------------------------------------------
(add/remove: 1/0 grow/shrink: 0/1 up/down: 143/-121)           Total: 22 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

libbb: use ARG_MAX for bb_arg_max() only if it's 60k+

Sometimes ARG_MAX is small (like 32k) yet sysconf(_SC_ARG_MAX)
is big, and people prefer using the bigger value.

OTOH, with sufficiently large ARG_MAX, further wins from
sysconf(_SC_ARG_MAX) being bigger are exponentially smaller:
you can see 4 times fewer fork+execs when you run find, but
when each execed process already takes a thousand parameters
it's likely execution time is dominated by what that process
does with each parameter.

Thus, with this change ARG_MAX is used if it's sufficiently big,
otherwise sysconf(_SC_ARG_MAX) is used.

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

bbunit: fix WANT_TIMING compilation

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

gitignore: add testsuite/echo-ne

Signed-off-by: Bartosz Golaszewski <bartekgola at gmail.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

udhcpd: if a lease from lease file coincides with a static one, ignore it

function                                             old     new   delta
read_leases                                          269     328     +59

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

init: do not run shutdown/reexec actions from signal handler

this is racy wrt various libc functions such as syslog()

function                                             old     new   delta
check_delayed_sigs                                   182     352    +170
init_main                                            772     728     -44
restart_handler                                       74       -     -74
halt_reboot_pwoff                                     79       -     -79
------------------------------------------------------------------------------
(add/remove: 0/2 grow/shrink: 1/1 up/down: 170/-197)          Total: -27 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

modprobe,rmmod: reject module names with slashes

function                                             old     new   delta
add_probe                                             86     113     +27

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

typo fix in comment

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ash,hush: do not segfault on $((2**63 / -1))

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ash: fix handling of negative start value in ${v:start:len}

function                                             old     new   delta
subevalvar                                          1140    1168     +28

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

iproute: support onelink route option and print route flags

function                                             old     new   delta
print_route                                         1797    1865     +68
do_iproute                                          2097    2112     +15
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/0 up/down: 83/0)               Total: 83 bytes

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

exit with 127 if appled name wasn't found - it's more POSIXy

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

zcip: Add environment variable for overriding log functionality

function                                             old     new   delta
bb_logenv_override                                     -      70     +70
packed_usage                                       29969   30033     +64
zcip_main                                           1426    1431      +5
------------------------------------------------------------------------------
(add/remove: 2/0 grow/shrink: 2/0 up/down: 139/0)             Total: 139 bytes

Signed-off-by: Michel Stam <m.stam@fugro.nl>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

zcip: allow our own class B range to be used for ZeroConf

169.254 may already be used by a local network. This patch allows
specifying your own IP range.

Our particular use case is a mesh network, in which the nodes partaking
were using an IP range specifically assigned for the meshing purpose.
As the LAN side of the mesh node could use 169.254, this default range
was not an option.

function                                             old     new   delta
zcip_main                                           1342    1426     +84
pick_nip                                               -      40     +40
packed_usage                                       29974   29969      -5
pick                                                  34       -     -34
------------------------------------------------------------------------------
(add/remove: 1/1 grow/shrink: 1/1 up/down: 124/-39)            Total: 85 bytes

Signed-off-by: Michel Stam <m.stam@fugro.nl>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

udhcpd: add option for tweaking arpping

Some clients have a very short timeout for sending the DHCP
DISCOVER, shorter than the arpping timeout of 2000 milliseconds
that udhcpd uses by default.

This patch allows tweaking the timeout, or disabling of arpping
altogether, at the risk of handing out addresses which are
already in use.

function                                             old     new   delta
udhcpd_main                                         1460    1501     +41
udhcpc_main                                         2814    2851     +37
packed_usage                                       29957   29974     +17
arpping                                              477     493     +16
find_free_or_expired_nip                             161     174     +13
send_offer                                           285     292      +7
nobody_responds_to_arp                                85      89      +4
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 7/0 up/down: 135/0)             Total: 135 bytes

Signed-off-by: Michel Stam <m.stam@fugro.nl>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ntpd: calculate offset to jitter ratio before updating jitter

The offset to jitter ratio is now calculated before updating
jitter to make the test more sensitive.

function                                             old     new   delta
ntp_init                                             460     474     +14
update_local_clock                                   752     764     +12
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/0 up/down: 26/0)               Total: 26 bytes

Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

sed: fix "sed CMD -i nonexistent_file". Closes 7484

function                                             old     new   delta
sed_main                                             643     676     +33

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ntpd: don't stay at short polling interval

To avoid polling servers frequently slowly increase the interval up
to BIGPOLL when
- no replies are received from a peer
- no source can be selected
- peer claims to be unsynchronized (e.g. we are polling it too
  frequently)

When recv() returns with an error, drop code to try to continue
on network errors: I'm not convinced those cases happen in real life.

function                                             old     new   delta
recv_and_process_peer_pkt                            919     838     -81

Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

getty: set tty attrs so that control chars are shown as ^c

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ntpd: be less eager to use shorter poll intervals

* on step, poll interval drops to 8.5 mins instead of 32 seconds
* on total loss of all replies (no replies from any peer
  for last 8 requests), also drop poll interval to 8.5 mins
  instead of 32 seconds
* on send abd recv errors, RETRY_INTERVAL is now 32 seconds,
  not 5 seconds
* on timing out listening to reply, instead of unconditional
  shortening poll interval by x4, clamp it to NOREPLY_INTERVAL
  (512 seconds)
* if a largish offset is seen, clamp nexp poll interval
  to 128 seconds, not 64 seconds

function                                             old     new   delta
clamp_pollexp_and_set_MAXSTRAT                         -      37     +37
recv_and_process_peer_pkt                            861     869      +8
poll_interval                                         52      48      -4
update_local_clock                                   762     752     -10
ntpd_main                                           1063    1050     -13
------------------------------------------------------------------------------
(add/remove: 1/0 grow/shrink: 1/3 up/down: 45/-27)             Total: 18 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ubiupdatevol: fix -t to not require an option. Closes 7466

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

install: fix help text

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ntpd: split out poll adjusting code

Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ntpd: don't wait for good offset before disabling burst mode

The burst mode needs to be stopped even when no replies are received.

Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

install: implement -t DIR

Some packages want to install themselves using "-t" to specify
the directory (as supported by GNU coreutils). Add support for the option
for compatibility reasons.

function                                             old     new   delta
install_longopts                                      76      95     +19
install_main                                         769     777      +8
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/0 up/down: 27/0)               Total: 27 bytes

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

less: make -E work

function                                             old     new   delta
buffer_print                                          71     104     +33

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

less: trivial code shrink

function                                             old     new   delta
read_lines                                           715     695     -20

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

less: move "retry-on-EAGAIN" logic closer to read ops

This makes "G" (goto end of input) command work as well as
/search_for_nonexistent_string: both will read to EOF now
even from somewhat slow input (such as kernel's "git log").

function                                             old     new   delta
ndelay_on                                             35      43      +8
ndelay_off                                            35      43      +8
read_lines                                           695     691      -4
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/1 up/down: 16/-4)              Total: 12 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

less: disable "suppress empty wraparound" optimization

It is buggy wrt another use case...

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

df: fix -T option when ENABLE_FEATURE_HUMAN_READABLE=0

Signed-off-by: Ari Sundholm <ari@tuxera.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

false: make "false --help" exit with 1

function                                             old     new   delta
run_applet_no_and_exit                               447     445      -2

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

df: implement -T option

function                                             old     new   delta
df_main                                              863     998    +135
packed_usage                                       29827   29861     +34
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/0 up/down: 169/0)             Total: 169 bytes

Signed-off-by: Ari Sundholm <ari@tuxera.com>
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

sed: fix a buglet in s///NUM handling

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ftpd: make LIST command show dotfiles too

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

vi: fix incorrect memory access on brace matching. Closes 7256

While at it, fix brace matching to actually show the match
(missed fflush was causing cursor positioning to be buffered);
shorten brace matching code; remove unused macro indirection
in indicate_error().

Custom linker script 'busybox_ldscript' found, using it
function                                             old     new   delta
indicate_error                                         -      61     +61
mysleep                                               43      56     +13
char_insert                                          483     486      +3
find_pair                                            167     124     -43
Indicate_Error                                        61       -     -61
------------------------------------------------------------------------------
(add/remove: 1/1 grow/shrink: 2/1 up/down: 77/-104)           Total: -27 bytes

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

mdev: treat zero-length /dev/mdev.seq the same as "\n" one. Closes 7334

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ash: fix off-by-one in "jobs %4" handling. closes 7310

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

ash: factor out ASH_HELP config option

It used to be aliased to !FEATURE_SH_EXTRA_QUIET for ash,
while hush had it separate from FEATURE_SH_EXTRA_QUIET.

Bring ash in line with hush.

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>