On Thu, Jan 20, 2011 at 12:53:01AM +0100, Denys Vlasenko wrote:
On Wednesday 19 January 2011 20:16, Chris Rees wrote:
Hey all,
Still plodding away at updating the FreeBSD port.
Got another patch; FreeBSD etc doesn't have sighandler_t, used in linedit.c.
Perhaps for platform.h?
On Sat, Jan 22, 2011 at 01:40:45PM +0100, Mikhail Gusarov wrote:
Twas brillig at 02:57:47 22.01.2011 UTC+01 when vda.li...@googlemail.com
did gyre and gimble:
DV But it may lead to a hang if serial line gets mangled: in some
DV situations tcdrain may wait forever.
DV I propose adding
On Sun, Jan 23, 2011 at 03:39:02PM +0100, Denys Vlasenko wrote:
On Sun, Jan 23, 2011 at 11:28 AM, Michal Simek mon...@monstr.eu wrote:
Give me what I should test and I will do it.
Insert sleep(1) before tcflush(0, TCIOFLUSH) in termios_init in getty.c,
and let us know whether the getty
On Mon, Jan 24, 2011 at 09:03:45AM +0100, Peter Korsgaard wrote:
Ok, what about:
+ /* Wait up to 5 seconds for the output buffer to drain */
+ signal(SIGALRM, record_signo);
+ alarm(5);
+ tcdrain(STDIN_FILENO);
+ alarm(0);
+ signal(SIGALRM, SIG_DFL);
On Tue, Jan 25, 2011 at 04:45:32PM -0500, rouble wrote:
Busybox gurus,
Is there anyway to get ash to execute /etc/profile by default without
having to use the -l flag?
If there isn't a configurable way to do this, any ideas on the best
way to change the code to do this are appreciated.
On Tue, Jan 25, 2011 at 11:22:38PM +0100, Denys Vlasenko wrote:
On Tuesday 25 January 2011 21:48, Chris Rees wrote:
Hi again,
I've got Busybox to compile on FreeBSD-9 now, but to do that I needed
to knock out all occurrences of utmp.h.
My solution was to make compilation of utmp.c
On Thu, Jan 27, 2011 at 06:18:25AM +0100, Denys Vlasenko wrote:
On Wednesday 26 January 2011 13:21, Baruch Siach wrote:
Hi busybox list,
I'm running the following command under strace (thanks Rob):
echo 56 /sys/class/gpio/export
and I see the following output:
write(1,
On Wed, Feb 02, 2011 at 08:24:28PM +0100, Denys Vlasenko wrote:
+static volatile int timed_out;
+
+static void alrm_catcher(int i)
+{
+ printf (\007timed out waiting for input: auto-logout\n);
+ timed_out = 1;
+ return;
+}
(1) printf in signal handler is a bad idea.
It's
On Wed, Feb 02, 2011 at 04:59:37PM -0600, Rob Landley wrote:
Yeah, factor of 20 is fairly significant. Of course mine didn't have
any command line option parsing overhead or similar, it was _just_
measuring the overhead of fork plus wait vs open plus close plus
unlink, and doing a
On Thu, Feb 03, 2011 at 10:45:00AM +, David Collier wrote:
Denys,
your suggestion above for fixing tar to work with stdin is wonderful -
and I'm sure it will work.
But it illustrates the different world-views that you and I have. In an
illuminating way.
You think/speak as if
On Mon, Feb 07, 2011 at 02:01:13PM +, bugzi...@busybox.net wrote:
https://bugs.busybox.net/show_bug.cgi?id=3097
Denys Vlasenko vda.li...@googlemail.com changed:
What|Removed |Added
On Mon, Feb 07, 2011 at 01:49:16PM +, bugzi...@busybox.net wrote:
https://bugs.busybox.net/show_bug.cgi?id=3091
--- Comment #1 from Denys Vlasenko vda.li...@googlemail.com ---
(In reply to comment #0)
Created attachment 2845 [details]
Patch to fix the issues
Busybox sed does not
On Mon, Feb 07, 2011 at 08:10:48PM +0100, Joakim Tjernlund wrote:
Relying on undef - 0 is dangerous. Much better build with -Wundef
and fixup the errors.
How is this dangerous? Undef - 0 is part of the C language at least
since C89, and perhaps before that. Nobody is using pre-ANSI C
compilers,
On Mon, Feb 07, 2011 at 11:13:03PM +0100, Joakim Tjernlund wrote:
On Mon, Feb 07, 2011 at 08:10:48PM +0100, Joakim Tjernlund wrote:
Relying on undef - 0 is dangerous. Much better build with -Wundef
and fixup the errors.
How is this dangerous? Undef - 0 is part of the C language at
On Tue, Feb 08, 2011 at 12:02:26AM +0100, Joakim Tjernlund wrote:
On Mon, Feb 07, 2011 at 11:13:03PM +0100, Joakim Tjernlund wrote:
On Mon, Feb 07, 2011 at 08:10:48PM +0100, Joakim Tjernlund wrote:
Relying on undef - 0 is dangerous. Much better build with -Wundef
and fixup the
On Tue, Feb 08, 2011 at 09:49:21PM +0100, Denys Vlasenko wrote:
Also, the ? character is not special in POSIX BRE, and \? is not a
conformant
substitute. The portable solution is to use egrep instead of grep.
Great, can you point out where exactly is that ?
On Wed, Feb 09, 2011 at 01:33:28AM +0100, Denys Vlasenko wrote:
Oh, sorry. In scripts/mkconfigs, there are two occurrances of:
grep '^#\? \?CONFIG_' $config \
they should read:
egrep '^#? ?CONFIG_' $config \
I see that this '?' thing is a can of worms.
I propose to sidestep it
On Sun, Jan 23, 2011 at 08:05:44PM -0600, Rob Landley wrote:
/* Optionally wait for CR or LF before writing /etc/issue */
if (option_mask32 F_WAITCRLF) {
char ch;
debug(waiting for cr-lf\n);
while (safe_read(STDIN_FILENO, ch, 1) == 1) {
debug(read
On Wed, Feb 09, 2011 at 12:26:41PM +0100, Ralf Friedl wrote:
Denys Vlasenko wrote:
On Wednesday 09 February 2011 00:56, Rich Felker wrote:
Oh, sorry. In scripts/mkconfigs, there are two occurrances of:
grep '^#\? \?CONFIG_' $config \
they should read:
egrep '^#? ?CONFIG_' $config
On Tue, Feb 22, 2011 at 11:16:46PM -0600, Rob Landley wrote:
So if you start with allnoconfig and look to see what applets you can
enable, things like mount don't exist for you unless you know to look
for them. This adds a new layer of black magic and a-priori knowledge
required in order to
On Wed, Feb 23, 2011 at 10:58:09PM +, Mark Turner wrote:
You wrote :
There is plenty of usefulness to Busybox that is not Linux-dependent
and that, contrary to what you wrote, is not trivial to reimplement.
One great example is Busybox ash, which is the only lightweight
POSIX-compatible
On Thu, Feb 24, 2011 at 09:39:16AM +0100, Harald Becker wrote:
Hallo David!
So the question is - can anyone explain/guess why it might have been
worthwhile to go and get proftpd, when ftpd seems to be there, and to
have been fine on the previous model?
And in addition: Are you sure that
On Wed, Mar 30, 2011 at 08:52:26AM -0700, Cathey, Jim wrote:
One system I used, DNIX, would push an unlinked executing binary
into swap space, then do the FS deletion. Was actually pretty nice,
it 'just worked'. Once the last executing process died the swap
space was released. There was
On Wed, Mar 30, 2011 at 11:51:37AM -0700, Cathey, Jim wrote:
In this (DNIX) system the file was not 'open' at the user level,
and as a resident of swap space it didn't really have a device or
an inode. (As data/bss segments did not. Just memory pages backed
to a blocking store.) Even if it
On Sat, Apr 02, 2011 at 08:40:08PM +0400, Sergey Naumov wrote:
It seems that it is problem of new mc. I have similar problems with mc
from Debian 6.0.0 and openssh. mc from debian 5.0.0 works fine with my
busybox-based system.
Parsing ls Output Considered Harmful.
Rich
On Thu, Feb 10, 2011 at 09:33:32AM +0100, Denys Vlasenko wrote:
While BusyBox supports it, the -r option to sed is nonstandard, so
little is improved. If you want to go this route, use:
sed -ne 's/\(^#\{0,1\} \{0,1\}CONFIG_.*\)$/\1\\n/gp' $config
But really I don't care which solution
On Wed, Apr 06, 2011 at 08:33:13PM +0100, Chris Rees wrote:
On 5 April 2011 16:43, Rich Felker dal...@aerifal.cx wrote:
On Thu, Feb 10, 2011 at 09:33:32AM +0100, Denys Vlasenko wrote:
While BusyBox supports it, the -r option to sed is nonstandard, so
little is improved. If you want to go
Hi,
I am the author of musl (www.etalabs.net/musl) and also follow busybox
development on and off. Recently one of my project's testers has been
experimenting with building a musl+busybox based mini-distro. I've
been working with him to resolve the build issues caused by things on
my end, but
On Sat, Apr 09, 2011 at 12:21:00AM -0400, Rich Felker wrote:
I am the author of musl (www.etalabs.net/musl) and also follow busybox
development on and off. Recently one of my project's testers has been
experimenting with building a musl+busybox based mini-distro. I've
been working with him
On Fri, Apr 15, 2011 at 07:17:22PM +0100, Chris Rees wrote:
Hi,
Is it vital that 'sed --' is used in scripts/gen-build-files.sh?
It uses the variable $srcdir, which is set in the Makefile to an
absolute path; although of course it would normally be sensible to use
-- when using a
On Sun, Apr 17, 2011 at 11:19:10PM +0200, Denys Vlasenko wrote:
Finally, one thing that's not in the patch, but should be: Busybox's
make config and make menuconfig require a some POSIX interfaces to
be provided by the standard C headers, so on musl, HOSTCFLAGS
containing
On Sun, Apr 17, 2011 at 11:19:10PM +0200, Denys Vlasenko wrote:
Finally, one thing that's not in the patch, but should be: Busybox's
make config and make menuconfig require a some POSIX interfaces to
be provided by the standard C headers, so on musl, HOSTCFLAGS
containing
On Sun, Apr 17, 2011 at 11:19:10PM +0200, Denys Vlasenko wrote:
Finally, one thing that's not in the patch, but should be: Busybox's
make config and make menuconfig require a some POSIX interfaces to
be provided by the standard C headers, so on musl, HOSTCFLAGS
containing
On Fri, Apr 29, 2011 at 03:30:44PM -0700, Ian Wienand wrote:
Hi,
666 seems too permissive for default pidfile permissions.
man umask
Rich
___
busybox mailing list
busybox@busybox.net
http://lists.busybox.net/mailman/listinfo/busybox
On Sun, May 01, 2011 at 02:00:27PM +0200, Wolfgang Denk wrote:
chains). This commit changes behaviour so -O2 gets used as fallback
when -Os is not available. This reduces the image size in above
test to 1.3 MiB. This is still 10...15% more then what we get with
-Os, but much better than
On Sun, May 01, 2011 at 07:38:27PM +0200, Marek Polacek wrote:
+int portmap_main(int argc, char **argv) MAIN_EXTERNALLY_VISIBLE;
+int portmap_main(int argc, char **argv)
+{
+ char *host = (char *)0.0.0.0;
This cast is a no-op.
This doesn't seem right. I think the `host' should be
On Mon, May 02, 2011 at 11:28:22AM +0200, Stefan Seyfried wrote:
On Sun, 1 May 2011 18:13:34 -0400
Rich Felker dal...@aerifal.cx wrote:
+ char *host = (char *)0.0.0.0;
This cast is a no-op.
I think it's not. 0.0.0.0 is a (const char *) IIUC.
Well you remember incorrectly. C
On Wed, May 04, 2011 at 09:11:35PM +0200, Denys Vlasenko wrote:
All this trouble just for the single one case where we store time?
It's much easier to just add statbuf-st_mtime = 0 ? statbuf-st_mtime : 0
there.
Agreed. Negative time is generally considered bogus. I don't see why
lots of code
On Thu, May 05, 2011 at 01:40:57AM +0200, Denys Vlasenko wrote:
config/mconf.c came from kernel tree. I don't plan to maintain a fork,
I plan to migrate to updated version from newer kernels.
In this case, might I suggest we go back and put the feature test
macros (_XOPEN_SOURCE and
On Mon, May 09, 2011 at 01:20:16AM +0200, Denys Vlasenko wrote:
Changing signal handling to a different scheme,
where they are caught and recorded, will somewhat help,
because now poll/read syscalls will be interruptible,
but this creates another problem: many other sycalls
may start failing
On Mon, May 09, 2011 at 03:28:15PM +0200, Harald Becker wrote:
I highly agree to this. It at least was the way it has been handled in
bash (at the time I dig deep into bash code, 10 years ago). Don't know
if this type of fd protection is still in there. Older shells even
limited user fds to
On Mon, May 09, 2011 at 09:58:05PM +0200, y...@klacno.sk wrote:
const char *alg[] = {
md5,
sha1,
sha224,
sha256,
sha384,
sha512
};
This is a very bloated way of doing strings. Note that the pointer
table is not even const! At least use const char *const
On Mon, May 09, 2011 at 11:16:00PM +0200, y...@klacno.sk wrote:
Thank you for suggestion, but this is only sample not suited for busybox.
First I want to get some criticism if this is good idea before continue
porting.
I think it's a very good idea, tho I'm not someone to make decisions
on
On Fri, May 13, 2011 at 03:45:06AM +0200, Denys Vlasenko wrote:
IOW: static i686 build, running on 64-bit kernel, consumes 46 mb of RAM
per 1000 busybox sleep processes, or 46 kbytes per process.
This is pretty good, but in theory it could be as low as 8k per
process (1 page data and 1 page
On Fri, May 13, 2011 at 11:48:40AM +0200, Denys Vlasenko wrote:
On Friday 13 May 2011 06:00, Rich Felker wrote:
On Fri, May 13, 2011 at 03:45:06AM +0200, Denys Vlasenko wrote:
IOW: static i686 build, running on 64-bit kernel, consumes 46 mb of RAM
per 1000 busybox sleep processes, or 46
On Fri, May 20, 2011 at 03:54:18AM +0200, Denys Vlasenko wrote:
On Thursday 05 May 2011 06:12, Rich Felker wrote:
On Thu, May 05, 2011 at 01:40:57AM +0200, Denys Vlasenko wrote:
config/mconf.c came from kernel tree. I don't plan to maintain a fork,
I plan to migrate to updated version
On Sun, May 22, 2011 at 03:51:01AM +0200, Denys Vlasenko wrote:
- return fnmatch(ap-pattern, fileName, 0) == 0;
+ return fnmatch(ap-pattern, fileName, (ap-ipath ? FNM_CASEFOLD : 0))
== 0;
FNM_CASEFOLD is a nonportable GNU feature. This should be made
optional, as this patch seems to
On Sun, May 22, 2011 at 04:35:11AM +0200, Denys Vlasenko wrote:
Also, note that it's nearly impossible to implement case folding
fnmatch without just implementing fnmatch in terms of regex, due to
the complexity of Unicode case folding, so we have very good reason
for not supporting this
On Wed, May 25, 2011 at 04:33:56PM +0200, Yann E. MORIN wrote:
echo myvar=my value | while read opt; do
[...]
That's expected. All that is to the right part of the pipe is running in a
sub-shell. So the evaluation is done in a sub-shell. So it is lost as soon
as this sub-shell ends.
The
On Mon, May 30, 2011 at 02:05:13PM -0700, Dan Fandrich wrote:
It's not easy compiling something against bionic using the
Android compiler outside the Android build system, but it can
at least now happen.
Signed-off-by: Dan Fandrich d...@coneharvesters.com
---
Conditionally compiling
On Mon, May 30, 2011 at 11:14:35PM -0700, Dan Fandrich wrote:
On Mon, May 30, 2011 at 07:48:05PM -0400, Rich Felker wrote:
Plain ttyname could be used, I think. The only reason ttyname_r is
used is to avoid enlarging bss for the nasty static buffer; it's not
for reentrancy. This could also
On Wed, Jun 01, 2011 at 01:54:14PM +0200, Tito wrote:
It does look like Walter Harms' suggestion of using sysconf() is the most
portable way of getting this value. The attached patch makes that change.
Why don't you just include #include asm/page.h
asm/* is absolutely not portable and
On Wed, Jun 01, 2011 at 09:21:27AM -0700, Dan Fandrich wrote:
On Wed, Jun 01, 2011 at 11:59:46AM +0200, walter harms wrote:
I have no idea how common _sc_pagesize is.
perhaps you should protect that _sc_pagesize like:
#ifndef PAGE_SIZE
#ifndef _sc_pagesize
# error no way to find your
On Wed, Jun 01, 2011 at 01:36:20PM -0700, Dan Fandrich wrote:
+#if defined(__GLIBC__) (__GLIBC__ 2 || __GLIBC_MINOR__ 1)
+# undef HAVE_NET_ETHERNET_H
+#endif
This is nonsense. glibc older than 2.1 is definitely not usable.
-#if (defined(__GLIBC__) __GLIBC__ = 2 __GLIBC_MINOR__ = 1) ||
On Wed, Jun 01, 2011 at 11:17:16PM -0700, Dan Fandrich wrote:
I suspect this might break support for anything that lacks
netpacket/packet.h, including musl. In general it's a very bad idea to
use the same HAVE_* macro for two completely separate headers.
In general, yes, but
On Thu, Jun 02, 2011 at 11:03:14AM -0700, Dan Fandrich wrote:
The only serious case I can think of that might break with this would be
alternate libcs on Linux, but given that most aim to be compatible with
glibc, which is it? If musl (or any others you're aware of) doesn't have
On Fri, Jun 03, 2011 at 09:29:26PM +0200, Denys Vlasenko wrote:
Moreover, PAGE_SIZE is only used for
mallopt(M_TRIM_THRESHOLD, 2 * PAGE_SIZE);
mallopt(M_MMAP_THRESHOLD, 8 * PAGE_SIZE - 256);
We can simply use fixed constants instead.
Agreed. The cleanest is:
#ifndef
On Thu, Apr 07, 2011 at 08:31:23PM +0100, Ed W wrote:
Obviously fixing iptables is desirable, but is it possible to improve
performance
I think your time would be much better spent adding a sane version of
iptables (without the hideous dlopen abomination) to Busybox. Then the
problem would be
On Mon, Jun 13, 2011 at 12:42:09PM -0700, Daniel Fandrich wrote:
This is a GNU extension that isn't found in other libcs.
[...]
--- a/include/platform.h
+++ b/include/platform.h
@@ -348,6 +348,7 @@ typedef unsigned smalluint;
#define HAVE_STRCHRNUL 1
#define HAVE_STRSEP 1
#define
On Mon, Jun 13, 2011 at 01:13:55PM -0700, Dan Fandrich wrote:
On Mon, Jun 13, 2011 at 03:47:39PM -0400, Rich Felker wrote:
If it's nonstandard and glibc-only, shouldn't it be *undefined* by
default and defined only if defined(__GLIBC__)!defined(__UCLIBC__)
or similar?
Practically all
On Mon, Jun 13, 2011 at 02:57:46PM -0700, Dan Fandrich wrote:
It's now used in ls -v, as of a few hours ago. The alternative would
be to just surround the -v support with HAVE_STRVERSCMP and not
provide a fallback implementation of strverscmp.
I would hope -v support would be a build-time
On Wed, Jun 15, 2011 at 01:49:30PM +0300, Timo Teräs wrote:
Hi all,
As recently discussed on the 'Significant performance problems with
modprobe', the config_* and xmalloc_fget* API has some performance issues:
1. It uses locking getc
2. It does malloc/realloc/free for each line read
Now
On Wed, Jun 15, 2011 at 03:28:16PM +0300, Timo Teräs wrote:
On 06/15/2011 03:14 PM, Rich Felker wrote:
On Wed, Jun 15, 2011 at 01:49:30PM +0300, Timo Teräs wrote:
Hi all,
As recently discussed on the 'Significant performance problems with
modprobe', the config_* and xmalloc_fget* API
On Wed, Jun 15, 2011 at 04:04:52PM +0300, Timo Teräs wrote:
Mmm.. yeah, uclibc needs fixing here. I'm still slightly doubtful how
much the performance difference would be actually. It probably depends
on the line lengths totally. On small lines, it's not much. On really
long lines, it could be
On Wed, Jun 15, 2011 at 04:47:39PM +0300, Timo Teräs wrote:
In any case, the difference is *very* small at least on x86 (including
SMP/multicore). The cost of the lock xchg vs if (threads1) is
something like 50 cycles, and if I remember right, making my
implementation always perform
On Thu, Jun 16, 2011 at 02:53:41PM +0100, Nuno Lucas wrote:
On 06/16/2011 05:51 AM, Timo Teräs wrote:
Notably, though, at least sed.c seems to rely on the fact that \n and \0
are both recognized, and jumps through many hoops to handle them both
properly. Apparently there's out some sed scripts
On Thu, Jun 16, 2011 at 10:26:53AM -0700, Cathey, Jim wrote:
There are exactly two characters (of an unsigned char)
that cannot be in a Unix filename: '/', which is the
path component separator, and '\0'. Everything else,
including newlines and spaces, is perfectly legal
in a filename,
On Thu, Jun 16, 2011 at 03:16:38PM -0400, Paul Smith wrote:
PS. Interestingly, there's a move afoot to remove newline from the set
of acceptable characters in a POSIX-compliant filename...
Do you have details on this? It's interesting because aside from the
'/' (which is granted special
On Thu, Jun 16, 2011 at 10:12:13PM +0200, Laurent Bercot wrote:
Sure -print0 required almost no change in the code for find, but it
requires significant change in any program using the output of find. I
maintain that shell quoting would have been a much better choice.
And it would have
On Fri, Jun 17, 2011 at 12:27:29PM +0200, Bernd Petrovitsch wrote:
Transformation in the other direction is virtually impossible using
the shell and standard utilities - the only way would be to pipe
I wouldn't call
echo -n $quoted | tr \000 \012
This is useless because you can no longer
On Sun, Jun 19, 2011 at 01:08:53PM +0200, Denys Vlasenko wrote:
On both machines speedup is not so big. Hash seems to help more then getline:
on F15, it reduced run time from 19 to 16 seconds, while getline won
~0.5 sec only.
For what it's worth, getline should make a much more dramatic
On Mon, Jun 20, 2011 at 10:21:02AM +0200, Denys Vlasenko wrote:
Since we don't use getdelim, maybe we should only implement getline?
How about just making getdelim static inline until if/when it's
needed? That would keep the code nicely factored and should eliminate
any cost to leaving it there.
On Mon, Jun 27, 2011 at 11:37:10AM -0700, Dan Fandrich wrote:
On Mon, Jun 20, 2011 at 10:21:02AM +0200, Denys Vlasenko wrote:
On Saturday 18 June 2011 17:56, Timo Teräs wrote:
Simple implementations. Need to check which platforms need the
wrapper.
Leaks line on error path.
I think you
On Wed, Jun 29, 2011 at 04:43:58AM +0200, Denys Vlasenko wrote:
Are you just waiting for an update on this patch to push it in?
The current code in git fails to compile on non-GNU systems because of
missing getline().
Applied the fix. Please try current git.
Wouldn't it be better to use
On Wed, Jun 29, 2011 at 10:54:34AM +0200, Denys Vlasenko wrote:
On Wednesday 29 June 2011 04:45, Rich Felker wrote:
On Wed, Jun 29, 2011 at 04:43:58AM +0200, Denys Vlasenko wrote:
Are you just waiting for an update on this patch to push it in?
The current code in git fails to compile
On Mon, Jul 04, 2011 at 06:55:42PM +0200, Gilles wrote:
On Mon, 04 Jul 2011 10:28:39 -0400, Chuck Meade
chuckme...@mindspring.com wrote:
http://www.busybox.net/FAQ.html#job_control
Thanks for the tip. Busybox's inittab doesn't mention /dev/console:
===
# Startup the system
On Fri, Jul 08, 2011 at 09:33:30AM +0200, Denys Vlasenko wrote:
What would cause TIOCGSERIAL to be undefined while __linux__ is
defined, or vice versa? Is that intended to support really old
kernels, broken system headers, or something else?
There is probably no deep wisdom behind it. I
On Sun, Jul 10, 2011 at 06:09:01AM +0100, Davide wrote:
Thirdly I somewhat disagree that using for instead of brace
expansion would make any shell script more readable:
mkdir -p /usr/local/people/{me,you,others}
mkdir -p /usr/local/people/me /usr/local/people/you /usr/local/people/others
On Sat, Jul 23, 2011 at 05:50:53PM -0500, Rob Landley wrote:
On 07/23/2011 05:29 PM, Douglas Mencken wrote:
Well, sed -i '/Include our own copy of struct
sysinfo/{NN;d}' ./include/libbb.h **doesn't work** as
expected (successful build): init/init.c:1010:18: error: storage
On Mon, Jul 25, 2011 at 08:28:21AM +0200, Laurent Bercot wrote:
Because they are Unix tools, i.e. they do one thing and do it well.
The netcat utility, as you say, has a 'server mode'. But clients and
servers aren't the same thing. Yup, I went there: netcat is bloated.
The real question
On Thu, Jul 28, 2011 at 10:29:01PM +0200, Matthias Andree wrote:
/* For example, name like .. can make adduser
* chown /home/.. recursively - NOT GOOD
*/
do {
if (*name == '-' || *name == '.')
continue;
skip:
if
On Fri, Jul 29, 2011 at 04:30:53PM +0200, Tito wrote:
Hi,
after more research about this topic I've found a few interesting things
about this can of worms I've unknowningly uncovered:
1) 3.426 User Name
A string that is used to identify a user; see also User Database. To be
portable
On Tue, Aug 02, 2011 at 03:37:35PM +0200, walter harms wrote:
granted, ntl it would be nice to have a better error that it does not work.
Perhaps (s-name) = position ? illegal character at position %d,(s-name) ?
wfprintf(stderr, illegal character %.1s\n, s);
I'm sure everyone will flame me for
On Wed, Aug 03, 2011 at 10:19:55AM +0200, Ralf Friedl wrote:
... just to throw that in. Beside this I like to use User Names with
German Umlauts (ÄÖÜäöü) as regular German names may contain those. And I
did this successfully on Linux/GNU systems for years ... but know about
incompatibility and
of the portable filename character set plus '@' and '$';
3) don't use isalnum as it will allow non-ASCII letters in legacy 8-bit
locales as pointed out by Rich Felker;
We use ASCII-only isalnum throughout bbox anyway. But ok.
Do you mean you have a special isalnum for ASCII-only, or you assume
On Tue, Aug 09, 2011 at 09:37:58AM +0200, Denys Vlasenko wrote:
#define isalnum(a) bb_ascii_isalnum(a)
static ALWAYS_INLINE int bb_ascii_isalnum(unsigned char a)
{
unsigned char b = a - '0';
if (b = 9)
return (b = 9);
b = (a|0x20) - 'a';
On Wed, Aug 17, 2011 at 03:13:09PM +0200, Denys Vlasenko wrote:
This means that the programs you start from the command line, inheriting the
shell's stdin/out/err can influence them, and by extension, influence all
further programs to be launched.
Yes, this is a bug in Unix API.
On Wed, Aug 17, 2011 at 10:09:52PM +0200, Laurent Bercot wrote:
This is by design and part of how multi-tool script programming works.
The Unix design is that scions (via fork() and execve()) inherit stuff
from their ancestors. If I'm running nice shoes and the nice program
sets stdin
On Thu, Aug 18, 2011 at 12:16:30AM +0200, Denys Vlasenko wrote:
On Wednesday 17 August 2011 23:20, Rich Felker wrote:
I hate it when fools who don't know what they're doing and don't
respect standards lobby for nonstandard behavior in Linux, thereby
making the portability landscape *worse
On Thu, Aug 18, 2011 at 12:45:57AM +0200, Laurent Bercot wrote:
Look at multitee: http://code.dogmap.org/fdtools/multitee/
Don't you agree that the *natural* way to write such a program would
be an asynchronous loop? Yet, for the very same reason, the author
jumped through hoops to avoid
On Thu, Aug 18, 2011 at 12:56:09AM +0200, Laurent Bercot wrote:
How can I do nonblocking read from fd 0 without
affecting other processes which might share it?
Assuming no other process will be reading at the same time, you simply
call poll first (or select, but select requires the fd #
On Thu, Aug 18, 2011 at 08:04:27AM +0200, Laurent Bercot wrote:
Nobody decided to accept/include gets. What was decided is the
general principle that POSIX is a superset of ISO/IEC 9899:1999, and
that in any areas where the two conflict, the conflict is
unintentional and the latter (C)
On Fri, Aug 19, 2011 at 12:36:32AM +0200, Denys Vlasenko wrote:
(very) pseudo-code of the good main application loop:
while (read(tqfd, event, sizeof event) == sizeof event)
switch (event) {
case keystroke: ...
case menu_item: ...
case
On Fri, Aug 19, 2011 at 05:03:03AM +0200, Laurent Bercot wrote:
that era) did not. (In single-CPU systems, which most DNIX boxes
were, LWP's are just a way to reorganize the problem and offer no
inherent performance or maintainability advantage over an AIO
approach, and might actually be
On Fri, Aug 19, 2011 at 11:36:43AM +0200, Bernd Petrovitsch wrote:
On Fre, 2011-08-19 at 06:55 +0200, Denys Vlasenko wrote:
On Thursday 18 August 2011 07:34, Laurent Bercot wrote:
It doesn't; it can't. Hence my ranting. A 100% safe
100% orthogonal, useful easy-to-use AIO mechanism
On Fri, Aug 19, 2011 at 10:03:40AM -0700, Cathey, Jim wrote:
Threads are also purely a question of taste. Rich likes them. Lots
of people like them. I don't. Alan Cox doesn't. YMMV.
I'm torn. Sometimes good, sometimes bad. We're actually having
very good luck using them in our product at
On Fri, Aug 19, 2011 at 05:17:52PM +0200, Harald Becker wrote:
Hallo Rich!
sigfillset(set);
sigprocmask(SIG_BLOCK, set, 0);
sigemptyset(set);
sigaddset(set, SIGTERM); /* and any others I want to handle */
sigprocmask(SIG_UNBLOCK, set, 0);
This is a lot more robust.
Robust yes
On Sat, Aug 20, 2011 at 01:37:23AM +0200, Bernd Petrovitsch wrote:
Multiprocessing can accomplish *some* of the same things, but not all,
What does multiprocessing exactly mean?
Writing a program that consists of more than one process running in
parallel.
- A library can't fork new
On Sat, Aug 20, 2011 at 06:06:04PM +0200, Bernd Petrovitsch wrote:
On Sam, 2011-08-20 at 10:00 -0400, Rich Felker wrote:
On Sat, Aug 20, 2011 at 01:37:23AM +0200, Bernd Petrovitsch wrote:
Multiprocessing can accomplish *some* of the same things, but not all,
[]
Note that just using
On Sat, Aug 20, 2011 at 11:26:30PM +0200, Denys Vlasenko wrote:
Guys, can you semd emails to lkml and advocate for recv/send to be enabled
for non-sockets?
Sorry, first you need to try to get this changed:
http://pubs.opengroup.org/onlinepubs/9699919799/functions/recv.html
ERRORS
The recv()
1 - 100 of 344 matches
Mail list logo