oweals/busybox.git
21 years agoccording to this:
Eric Andersen [Mon, 20 Oct 2003 05:40:40 +0000 (05:40 -0000)]
ccording to this:
    http://sources.redhat.com/ml/binutils/2003-01/msg00290.html
The name R_390_GOTOFF was changed to R_390_GOTOFF32.
 -Erik

21 years agoManuel Novoa III writes:
Eric Andersen [Sat, 18 Oct 2003 01:59:46 +0000 (01:59 -0000)]
Manuel Novoa III writes:

Hello Rob,

Here's a patch to your bunzip-3.c file.  Nice work btw.

One minor bug fix... checking for error return when read()ing.
Some size/performance optimizations as well.  One instance of
memset() seems unnecssary.  You might want to take a look.

Anyway, on my machine, decompressing linux-2.6.0-test7.tar.bz2
to /dev/null gave the following times:

        bunzip-3.c    bzcat (system)   bunzip-3.c (patched)
real    0m24.420s     0m22.725s        0m20.701s
user    0m23.930s     0m22.170s        0m20.180s
sys     0m0.070s      0m0.080s         0m0.140s

Size of the patched version is comparable (slightly larger or
smaller depending on compiler flags).

Manuel

21 years agoRob Landley's new micro-bunzip version 3. Rob writes:
Eric Andersen [Sat, 18 Oct 2003 01:58:35 +0000 (01:58 -0000)]
Rob Landley's new micro-bunzip version 3.  Rob writes:

The API for using partial writes, as described in my last message, sucked.

So here's a patch against my last patch that changes things so that
write_bunzip_data calls read_bunzip_data itself behind the scenes whenever
necessary.  So usage is now just start_bunzip(), write_bunzip_data() until it
returns a negative number, and then the cleanup at the end of
uncompressStream.

It adds 32 bytes to the executable, but it should allow the caller (tar) to be
simplified enough to compensate.  Total -Os stripped exe size now 6856 bytes.

Rob

P.S.  I attached the whole C file so you don't have to keep incremental
patches straight if you don't want to. :)

P.S.  In the version I'm banging on now, I've simplified the license to just
LGPL.  I read the OSL a bit more closely and the patent termination clause
would have bit IBM in their counter-suit of SCO if the code in question had
been OSL instead of GPL, and I've decided I just don't want to beta-test
legal code right now.

21 years agoold gcc
Tim Riker [Fri, 17 Oct 2003 18:54:55 +0000 (18:54 -0000)]
old gcc

21 years agoPaul Mundt (lethal) writes:
Eric Andersen [Sat, 11 Oct 2003 18:47:20 +0000 (18:47 -0000)]
Paul Mundt (lethal) writes:

Erik,

The format for /proc/meminfo has changed between 2.4 and 2.6, quite considerably.
In addition to the removal of the two-line summary that was present in 2.4,
MemShared was also removed. Presently (at least in busybox CVS HEAD), top fails
to parse this correctly and spews forth a:

top: failed to read 'meminfo'

message. This patch switches around some of the semantics a little to do sane
parsing for both 2.4 and 2.6. Also, in the event that the summary gets yanked
from 2.4, this patch will deal with that as well. With this patch, I'm able
to run top correctly on 2.6.0-test7 (tested on sh).

Please apply.

 procps/top.c |   60 +++++++++++++++++++++++++++++++++++++++++++++++------------
  1 files changed, 48 insertions(+), 12 deletions(-)

21 years agoVodz last_patch_117, update options for new xargs
Glenn L McGrath [Fri, 10 Oct 2003 23:15:23 +0000 (23:15 -0000)]
Vodz last_patch_117, update options for new xargs

21 years agoVods versions of xargs
Glenn L McGrath [Fri, 10 Oct 2003 12:10:18 +0000 (12:10 -0000)]
Vods versions of xargs

21 years agoI don't need to have my email adder listed twice
Eric Andersen [Thu, 9 Oct 2003 21:19:21 +0000 (21:19 -0000)]
I don't need to have my email adder listed twice

21 years agoDo not use kernel headers
Eric Andersen [Thu, 9 Oct 2003 21:02:23 +0000 (21:02 -0000)]
Do not use kernel headers

21 years agoFix an ommision preventing building, patch by Tito.
Glenn L McGrath [Thu, 9 Oct 2003 13:40:17 +0000 (13:40 -0000)]
Fix an ommision preventing building, patch by Tito.

21 years agoexample devfs config file, patch from Tito.
Glenn L McGrath [Thu, 9 Oct 2003 13:25:00 +0000 (13:25 -0000)]
example devfs config file, patch from Tito.

21 years agoNew applet, devfsd, by Matteo Croce and Tito
Glenn L McGrath [Thu, 9 Oct 2003 11:46:23 +0000 (11:46 -0000)]
New applet, devfsd, by Matteo Croce and Tito

21 years agoPatch from Steven Scholz, fix some warnings
Glenn L McGrath [Thu, 9 Oct 2003 11:38:45 +0000 (11:38 -0000)]
Patch from Steven Scholz, fix some warnings

21 years agoBugfix for xargs
Glenn L McGrath [Thu, 9 Oct 2003 11:06:45 +0000 (11:06 -0000)]
Bugfix for xargs

21 years agoArnd Ben Otto writes:
Eric Andersen [Thu, 9 Oct 2003 09:43:18 +0000 (09:43 -0000)]
Arnd Ben Otto writes:

Hi Eric

I have written a small patch for the Busybox syslogd. With this patch
one can limit the size of the messagfile. As soon as the limit is
reached the syslogd can rotate or purge the messagefile(s) on his own.
There is no necessity to use an external rotatescript.

Even if logread does something similar, its very handy to have some
messagefile after your box crash.

I wrote this patch initial vor BB 0.6x where no cron daemon was avail.
Now I adapted it for the new Version and i hope it is still useful. At
least I still use it :-)

bye
Arnd

21 years agoWe did not have a safe_write, which is the analog to safe_read. Convert
Eric Andersen [Thu, 9 Oct 2003 08:35:42 +0000 (08:35 -0000)]
We did not have a safe_write, which is the analog to safe_read.  Convert
full_write to use safe_write internally, which is needed to guarantee proper
behavior, i.e. when writing to a pipe.

21 years agoFix some warnings that have crept in recently
Eric Andersen [Thu, 9 Oct 2003 08:18:36 +0000 (08:18 -0000)]
Fix some warnings that have crept in recently

21 years agoPatch from Tito, size optimisation and fix error in exit code of -V
Glenn L McGrath [Thu, 9 Oct 2003 07:28:22 +0000 (07:28 -0000)]
Patch from Tito, size optimisation and fix error in exit code of -V
option

21 years agoComaptability with gcc-2.95
Glenn L McGrath [Thu, 9 Oct 2003 07:22:59 +0000 (07:22 -0000)]
Comaptability with gcc-2.95

21 years agoVodz last_patch_105 without his xargs patch which doenst apply cleanly
Glenn L McGrath [Mon, 6 Oct 2003 13:23:06 +0000 (13:23 -0000)]
Vodz last_patch_105 without his xargs patch which doenst apply cleanly

21 years agoAdd the x, n, s and E options, remove -r as its expected behaviour.
Glenn L McGrath [Sat, 4 Oct 2003 14:44:27 +0000 (14:44 -0000)]
Add the x, n, s and E options, remove -r as its expected behaviour.

21 years agoPatch from Rob Landley;
Glenn L McGrath [Sat, 4 Oct 2003 05:27:56 +0000 (05:27 -0000)]
Patch from Rob Landley;
Moving on to building diffutils, busybox sed needs this patch to get
past the first problem.  (Passing it a multi-line command line argument
with -e works, but if you don't use -e it doesn't break up the multiple
lines...)

21 years agoWoops, the previous patch reversal wasnt complete, breaking everything.
Glenn L McGrath [Sat, 4 Oct 2003 00:05:47 +0000 (00:05 -0000)]
Woops, the previous patch reversal wasnt complete, breaking everything.

21 years agoReverse my previous changes and make a note about why its dont this way
Glenn L McGrath [Fri, 3 Oct 2003 13:21:10 +0000 (13:21 -0000)]
Reverse my previous changes and make a note about why its dont this way

21 years agoDont mix xarg options with utility option, fixes example of
Glenn L McGrath [Fri, 3 Oct 2003 13:15:44 +0000 (13:15 -0000)]
Dont mix xarg options with utility option, fixes example of
`echo "README" | xargs ls -al`
Dont specify a path for the default behaviour of echo
args allocated space for an extra ptr
Use defines for the different options

21 years agoVodz, last_patch_114
Glenn L McGrath [Fri, 3 Oct 2003 10:50:56 +0000 (10:50 -0000)]
Vodz, last_patch_114
- env vars CONTENT_TYPE, CONTENT_LENGTH, HTTPD_REFERER, REMOTE_USER and
AUTH_TYPE(Basic always).
- POST data pipied now (previous version have loading into memory may be
big size data and reducing with hardcoded limit)
- removed $CGI_foo environment variables, else my have rubbish
enviroment if POST data have big binary file

21 years agoPatch by Rob Landley, fix warning
Glenn L McGrath [Fri, 3 Oct 2003 08:28:59 +0000 (08:28 -0000)]
Patch by Rob Landley, fix warning

21 years agoargc has already been decremented
Glenn L McGrath [Fri, 3 Oct 2003 07:51:30 +0000 (07:51 -0000)]
argc has already been decremented

21 years agoPatch from Rob Landley, Simplify organisation of arguments.
Glenn L McGrath [Fri, 3 Oct 2003 03:25:30 +0000 (03:25 -0000)]
Patch from Rob Landley, Simplify organisation of arguments.

21 years agoFix -C option when creating tar files.
Glenn L McGrath [Thu, 2 Oct 2003 14:33:23 +0000 (14:33 -0000)]
Fix -C option when creating tar files.
Need to chdir after the tar file is opened, so make common tar filename
parsing and send the file descriptor rather than filename to
writeTarFile.
Modify the verboseFlag operation to determine wether to display on
stderr or stdout at display time, simpler than doing it in tar_main.

21 years agoThomas Lundquist, update for current structure
Glenn L McGrath [Wed, 1 Oct 2003 11:33:46 +0000 (11:33 -0000)]
Thomas Lundquist, update for current structure

21 years agoPatch from Rob Landley to fix backrefs
Glenn L McGrath [Wed, 1 Oct 2003 10:26:23 +0000 (10:26 -0000)]
Patch from Rob Landley to fix backrefs

21 years agoPatch by Rob Landley, fix "newline after edit command"
Glenn L McGrath [Wed, 1 Oct 2003 06:45:11 +0000 (06:45 -0000)]
Patch by Rob Landley, fix "newline after edit command"

21 years agoPatch by Rob Landley, work in progress update, fixes lots of bugs,
Glenn L McGrath [Wed, 1 Oct 2003 03:06:16 +0000 (03:06 -0000)]
Patch by Rob Landley, work in progress update, fixes lots of bugs,
introduces a few others (but they are being worked on)

21 years agoPatch by Guillaume Morin
Glenn L McGrath [Fri, 26 Sep 2003 10:45:55 +0000 (10:45 -0000)]
Patch by Guillaume Morin
Fix two race conditions, as described at.
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=212764

21 years agoPatch from Fillod Stephane
Glenn L McGrath [Fri, 26 Sep 2003 01:03:16 +0000 (01:03 -0000)]
Patch from Fillod Stephane
Add follow mode to logread, ala "tail -f /var/log/messages"
Note: output to a slow serial terminal can have side effects
on syslog because of the semaphore. In such case, define
RC_LOGREAD.

21 years agoadd size parameter to syslogd -C
Glenn L McGrath [Fri, 26 Sep 2003 00:49:05 +0000 (00:49 -0000)]
add size parameter to syslogd -C
Patch by Padraig, resubmitted by Fillod Stephane

21 years agoPatch by Fillod Stephane, declare variables as static to not clobber
Glenn L McGrath [Fri, 26 Sep 2003 00:33:18 +0000 (00:33 -0000)]
Patch by Fillod Stephane, declare variables as static to not clobber
busybox namespace

21 years agoUse a variable for the config files, makes it easier to syncronise with
Glenn L McGrath [Wed, 24 Sep 2003 15:48:29 +0000 (15:48 -0000)]
Use a variable for the config files, makes it easier to syncronise with
debian builds

21 years agoSyncronise with debian busybox-cvs package
Glenn L McGrath [Wed, 24 Sep 2003 15:30:44 +0000 (15:30 -0000)]
Syncronise with debian busybox-cvs package

21 years agoFix some typo's, remove some extra free statements
Glenn L McGrath [Wed, 24 Sep 2003 10:23:39 +0000 (10:23 -0000)]
Fix some typo's, remove some extra free statements

21 years agoAdd the -d option
Glenn L McGrath [Wed, 24 Sep 2003 05:00:29 +0000 (05:00 -0000)]
Add the -d option

21 years agoAdd the "install" applet, move get_ug_id to libbb as its used by chown,
Glenn L McGrath [Wed, 24 Sep 2003 03:22:57 +0000 (03:22 -0000)]
Add the "install" applet, move get_ug_id to libbb as its used by chown,
chgrp and install.

21 years agoPatch by Steinar H. Gunderson to fix debian bug #211675.
Glenn L McGrath [Sat, 20 Sep 2003 00:59:35 +0000 (00:59 -0000)]
Patch by Steinar H. Gunderson to fix debian bug #211675.
Linking to my_getgrnam from libpwdgrp wasnt working, instead it was
trying to use functionality from glibc, which pulled in libnss.

21 years agoPatch by Junio C Hamano to workaround a gcc compiler bug.
Glenn L McGrath [Wed, 17 Sep 2003 00:22:26 +0000 (00:22 -0000)]
Patch by Junio C Hamano to workaround a gcc compiler bug.

The construct certain vintages of GCC (the one I have trouble
with is 3.2.3) have trouble with looks like the following:

    static struct st a;
    static struct st *p = &a;
    struct st { int foo; };
    static void init(void) { a.foo = 0; }

The problem disappears if we move the struct declaration up to
let the compiler know the shape of the struct before the first
definition uses it, like this:

    struct st { int foo; }; /* this has been moved up */
    static struct st a;
    static struct st *p = &a;
    static void init(void) { a.foo = 0; }

21 years agoConfiguration option to define wether to follows GNU sed's behaviour
Glenn L McGrath [Tue, 16 Sep 2003 05:25:43 +0000 (05:25 -0000)]
Configuration option to define wether to follows GNU sed's behaviour
or the posix standard.
Put the cleanup code back the way it was.

21 years agoFix a bug that creapt in recently with substitution subprinting, and add
Glenn L McGrath [Tue, 16 Sep 2003 01:46:36 +0000 (01:46 -0000)]
Fix a bug that creapt in recently with substitution subprinting, and add
a test for it.

21 years agoCompile get_terminal_width_height
Glenn L McGrath [Tue, 16 Sep 2003 00:50:36 +0000 (00:50 -0000)]
Compile get_terminal_width_height

21 years agoPatch from Stephane Billiart to fix an unused variable warning.
Glenn L McGrath [Mon, 15 Sep 2003 15:00:43 +0000 (15:00 -0000)]
Patch from Stephane Billiart to fix an unused variable warning.

21 years agoPatch from Bastian Blank to fix a problem when runing find under ash.
Glenn L McGrath [Mon, 15 Sep 2003 14:42:39 +0000 (14:42 -0000)]
Patch from Bastian Blank to fix a problem when runing find under ash.

"If the shell is compiled with -DJOBS, this is all fine -- find wasn't
stopped (it was killed), so it correctly uses WTERMSIG instead of WSTOPSIG.
However, if the shell _isn't_ compiled with -DJOBS (which it isn't in d-i),
only WSTOPSIG is used, which extracts the high byte instead of the low
byte from the status code.  Since the status code is 13 (SIGPIPE), "st"
suddenly gets the value 0, which is equivalent to SIGEXIT. Thus, ash prints
out "EXIT" on find's exit."

21 years agoPatch from Tito, Reduces the size of busybox's strings applet from 1900 to
Glenn L McGrath [Mon, 15 Sep 2003 14:22:37 +0000 (14:22 -0000)]
Patch from Tito, Reduces the size of busybox's strings applet from 1900 to
1788 bytes (for strings.o).

21 years agoFix a simple mistake with pattern space, and add a test for it
Glenn L McGrath [Mon, 15 Sep 2003 12:07:48 +0000 (12:07 -0000)]
Fix a simple mistake with pattern space, and add a test for it

21 years agoPatch by Jean Wolter to fix a bug where a script wouldnt be executed
Glenn L McGrath [Mon, 15 Sep 2003 12:00:19 +0000 (12:00 -0000)]
Patch by Jean Wolter to fix a bug where a script wouldnt be executed
unless it had #!/bin/sh in the first line

"It correctly locates the script, tries to execute it via execve which
fails. After that it tries to hand it over to /bin/sh which fails too,
since ash

    - neither provides the absolute pathname to /bin/sh
    - nor tries to lookup the script via PATH if called as "sh script"
"

21 years agoFix some memory allocation problems
Glenn L McGrath [Mon, 15 Sep 2003 09:22:04 +0000 (09:22 -0000)]
Fix some memory allocation problems
----------------------------------------------------------------------

21 years agoBe entirely consistant when using ioctl(0, TIOCGWINSZ, &winsize)
Eric Andersen [Mon, 15 Sep 2003 08:33:45 +0000 (08:33 -0000)]
Be entirely consistant when using ioctl(0, TIOCGWINSZ, &winsize)
to ensure proper fallback behavior on, i.e. serial consoles.
 -Erik

21 years agofix function prototype
Eric Andersen [Mon, 15 Sep 2003 08:13:43 +0000 (08:13 -0000)]
fix function prototype

21 years agoNeeds prototype for close()
Eric Andersen [Mon, 15 Sep 2003 08:12:53 +0000 (08:12 -0000)]
Needs prototype for close()

21 years agocomparison was always false due to limited range of data types.
Eric Andersen [Mon, 15 Sep 2003 08:11:29 +0000 (08:11 -0000)]
comparison was always false due to limited range of data types.
Carefully cast to unsigned long long prior to multiply to get
the expected result.

21 years agoDo not shadow the global name 'accept'
Eric Andersen [Mon, 15 Sep 2003 08:06:15 +0000 (08:06 -0000)]
Do not shadow the global name 'accept'

21 years agoAdd a test for the 'P' command and fix current implementation so it
Glenn L McGrath [Mon, 15 Sep 2003 06:28:45 +0000 (06:28 -0000)]
Add a test for the 'P' command and fix current implementation so it
doesnt permanently modify the pattern space.

21 years agoA test and fix for the sed 'n' command
Glenn L McGrath [Mon, 15 Sep 2003 06:12:53 +0000 (06:12 -0000)]
A test and fix for the sed 'n' command

21 years agoTest for use of newline in regex's, this feature is used by most
Glenn L McGrath [Mon, 15 Sep 2003 05:53:28 +0000 (05:53 -0000)]
Test for use of newline in regex's, this feature is used by most
configure scripts.

21 years agoFix for the sed-append-next-line test
Glenn L McGrath [Mon, 15 Sep 2003 05:42:05 +0000 (05:42 -0000)]
Fix for the sed-append-next-line test

21 years agoTest the N command
Glenn L McGrath [Mon, 15 Sep 2003 05:35:47 +0000 (05:35 -0000)]
Test the N command

21 years agoFix recursion problem
Glenn L McGrath [Mon, 15 Sep 2003 04:55:29 +0000 (04:55 -0000)]
Fix recursion problem

21 years agoCheck sed doesnt go into an infinite loop (yes it does)
Glenn L McGrath [Mon, 15 Sep 2003 04:41:17 +0000 (04:41 -0000)]
Check sed doesnt go into an infinite loop (yes it does)

21 years agoUpdate Matteo Croce's email address
Glenn L McGrath [Mon, 15 Sep 2003 03:37:32 +0000 (03:37 -0000)]
Update Matteo Croce's email address

21 years agoMemory cleanups and fix for `echo "foo" | sed 's/foo/bar/;H;q'`
Glenn L McGrath [Sun, 14 Sep 2003 16:28:08 +0000 (16:28 -0000)]
Memory cleanups and fix for `echo "foo" | sed 's/foo/bar/;H;q'`

21 years agoCleanup memory usage
Glenn L McGrath [Sun, 14 Sep 2003 15:24:18 +0000 (15:24 -0000)]
Cleanup memory usage

21 years agoUpdate sed branch tests
Glenn L McGrath [Sun, 14 Sep 2003 11:10:08 +0000 (11:10 -0000)]
Update sed branch tests

21 years agoAdd two new tests, sed-recurses-properly should always work
Glenn L McGrath [Sun, 14 Sep 2003 09:38:24 +0000 (09:38 -0000)]
Add two new tests, sed-recurses-properly should always work

21 years agoThe previous fix for 's/a/1/;s/b/2/;t one;p;:one;p' broke the case of
Glenn L McGrath [Sun, 14 Sep 2003 08:52:53 +0000 (08:52 -0000)]
The previous fix for 's/a/1/;s/b/2/;t one;p;:one;p' broke the case of
echo fooba | ./busybox sed -n 's/foo//;s/bar/found/p'

I really need to start adding these tests to the testsuite.

keep the substituted and altered flags seperate

21 years agoPreserve substitution flag value within the current line.
Glenn L McGrath [Sun, 14 Sep 2003 07:59:28 +0000 (07:59 -0000)]
Preserve substitution flag value within the current line.
Fixed the following testcase
# cat strings |./busybox sed -n -f test3.sed
1
1
2
c
c
# cat strings
a
b
c

21 years agoFix branching commands.
Glenn L McGrath [Sun, 14 Sep 2003 06:01:14 +0000 (06:01 -0000)]
Fix branching commands.

If a label isnt specified, jump to end of script, not the last command
in the script.

Print an error and exit if you try and jump to a non-existant label

Works for the following testcase
# cat strings
a
b
c
d
e
f
g
# cat strings | ./busybox sed -n '/d/b;p'
a
b
c
e
f
g

21 years agoPatch from Rob Landley
Glenn L McGrath [Sun, 14 Sep 2003 04:06:12 +0000 (04:06 -0000)]
Patch from Rob Landley

Fixed a memory leak in add_cmd/add_cmd_str by moving the allocation
of sed_cmd down to where it's actually first needed.

In get_address, if index_of_next_unescaped_regexp_delim ever failed, we
wouldn't notice because the return value was added to idx, which was
already guaranteed to be > 0.  (This is buried in the changes made when
I redid get_address to be based on pointer arithmetic, because all the tests
were gratuitously dereferencing with a constant zero, which wasn't obvious.)

Comment in parse_regex_delim was wrong: 's' and 'y' both call it.

The reason "sed_cmd->num_backrefs = 0;" isn't needed is that sed_cmd was
allocated with cmalloc, which zeroes memory.

Different handling of space after \ in i...

Different handling of pattern "s/a/b s/c/d"

Cool, resursive reads don't cause a crash. :)

Fixed "sed -f blah filename - < filename" since GNU sed was handling
both - and filenames on the same line.  (You can even list - more than
once, although it's immediate EOF...)

21 years agoStupid typo
Glenn L McGrath [Sun, 14 Sep 2003 02:37:46 +0000 (02:37 -0000)]
Stupid typo

21 years agoFix some memory allocation problems
Glenn L McGrath [Sun, 14 Sep 2003 01:25:31 +0000 (01:25 -0000)]
Fix some memory allocation problems

21 years agoFix the following testcase by disabling global substitution if the regex
Glenn L McGrath [Sat, 13 Sep 2003 15:12:22 +0000 (15:12 -0000)]
Fix the following testcase by disabling global substitution if the regex
is anchored to the start of line, there can be only one subst.
echo "aah" | sed 's/^a/b/g'

21 years agoFix the following testcase by storing the state of the adress match with
Glenn L McGrath [Sat, 13 Sep 2003 06:57:39 +0000 (06:57 -0000)]
Fix the following testcase by storing the state of the adress match with
the command.
# cat strings
a
b
c
d
e
f
g
# ./busybox sed '1,2d;4,$d' <strings
c
# ./busybox sed '4,$d;1,2d' <strings
# sed '4,$d;1,2d' <strings
c
# sed '1,2d;4,$d' <strings
c

21 years agoFix compile error, Vodz, last_path_113
Glenn L McGrath [Fri, 12 Sep 2003 11:27:15 +0000 (11:27 -0000)]
Fix compile error, Vodz, last_path_113

21 years agoTypo.
Glenn L McGrath [Fri, 12 Sep 2003 10:58:54 +0000 (10:58 -0000)]
Typo.

21 years agoAs vodz just pointed out, I screwup up the call to bb_xasprintf! 1_00_pre3
Eric Andersen [Fri, 12 Sep 2003 08:39:05 +0000 (08:39 -0000)]
As vodz just pointed out, I screwup up the call to bb_xasprintf!

21 years agoRemove final \n
Eric Andersen [Fri, 12 Sep 2003 08:32:24 +0000 (08:32 -0000)]
Remove final \n

21 years agoFix obligitory typos
Eric Andersen [Fri, 12 Sep 2003 07:36:46 +0000 (07:36 -0000)]
Fix obligitory typos

21 years agoFinal changelog update
Eric Andersen [Fri, 12 Sep 2003 07:03:52 +0000 (07:03 -0000)]
Final changelog update

21 years agoRemove version #
Eric Andersen [Fri, 12 Sep 2003 07:03:14 +0000 (07:03 -0000)]
Remove version #

21 years agoI suppose it is now Sept 12...
Eric Andersen [Fri, 12 Sep 2003 06:59:52 +0000 (06:59 -0000)]
I suppose it is now Sept 12...

21 years agoUpdate webpage a bit
Eric Andersen [Fri, 12 Sep 2003 06:59:17 +0000 (06:59 -0000)]
Update webpage a bit

21 years agoFix handling of hardlinks when OLDGNU and GNU extensions arent enabled.
Glenn L McGrath [Fri, 12 Sep 2003 06:49:09 +0000 (06:49 -0000)]
Fix handling of hardlinks when OLDGNU and GNU extensions arent enabled.

21 years agoUse the typeflag to identify if its a hardlink on OLD and GNU posix
Glenn L McGrath [Fri, 12 Sep 2003 06:31:28 +0000 (06:31 -0000)]
Use the typeflag to identify if its a hardlink on OLD and GNU posix
modes, fixes a bug extracting hardlinks to symlinks.

21 years agoInvoke run-parts on /etc/network/if-*.d per the behavior of
Eric Andersen [Fri, 12 Sep 2003 05:59:53 +0000 (05:59 -0000)]
Invoke run-parts on /etc/network/if-*.d per the behavior of
the upstream ifupdown code...

21 years agoTeach rdate to timeout in 10 seconds to avoid blocking forever
Eric Andersen [Fri, 12 Sep 2003 05:50:51 +0000 (05:50 -0000)]
Teach rdate to timeout in 10 seconds to avoid blocking forever
with odd or broken networking setups

21 years agoconfig option CONFIG_FEATURE_SH_APPLETS_ALWAYS_WIN must depend
Eric Andersen [Fri, 12 Sep 2003 04:49:21 +0000 (04:49 -0000)]
config option CONFIG_FEATURE_SH_APPLETS_ALWAYS_WIN must depend
on CONFIG_FEATURE_SH_STANDALONE_SHELL.

21 years agoUpdate changelog for release
Eric Andersen [Fri, 12 Sep 2003 04:43:49 +0000 (04:43 -0000)]
Update changelog for release

21 years agoBusybox CVS version ipcalc have problems:
Glenn L McGrath [Fri, 12 Sep 2003 00:44:50 +0000 (00:44 -0000)]
Busybox CVS version ipcalc have problems:
- Can`t use 255.255.255.255 address
- typo: double check for valid ip address and uncheck for valid mask

httpd unseted SERVER_PORT (my bug from last_patch111).

last_patch_112 from Vladimir N. Oleynik

21 years agoUpdate definition of struct serial_struct, per linux-2.6.0-test5,
Eric Andersen [Thu, 11 Sep 2003 08:52:22 +0000 (08:52 -0000)]
Update definition of struct serial_struct, per linux-2.6.0-test5,
to avoid stack corruption problems on some 64bit architectures
when sizeof(void*) != sizeof(int).  Thanks to Atsushi Nemoto
for finding this problem.

21 years agoMarc A. Lehmann writes:
Eric Andersen [Thu, 11 Sep 2003 08:32:40 +0000 (08:32 -0000)]
Marc A. Lehmann writes:

The tar -x command in busybox does not restore the file mode correctly.

The reason is most probably this code in
archival/libunarachive/data_extract_all.c:

       chmod(file_header->name, file_header->mode);
       chown(file_header->name, file_header->uid, file_header->gid);

chown clears the set*id bits (on current versions of linux :). Flipping
the order around fixes the problem.

(tested with 1.00pre3 from cvs).

21 years agoPatch from Philip Blundell:
Eric Andersen [Thu, 11 Sep 2003 08:25:11 +0000 (08:25 -0000)]
Patch from Philip Blundell:

On Thu, 2003-09-11 at 01:09, Glenn McGrath wrote:
> Applied

Thanks.  Unfortunately there was one small bug in that last patch.
Could you apply this one as well?

p.

21 years agoadd ipv6 literal support to wget
Glenn L McGrath [Wed, 10 Sep 2003 23:52:15 +0000 (23:52 -0000)]
add ipv6 literal support to wget

21 years agoFix a bug where make clean complained about having both : and :: entries.
Glenn L McGrath [Wed, 10 Sep 2003 23:47:10 +0000 (23:47 -0000)]
Fix a bug where make clean complained about having both : and :: entries.
Remove tester.log on make clean
Patch by Arthur Othieno