Background processes receive SIGINT even though they shouldn't

2006-01-11 Thread Ingemar Nilsson 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1


Configuration Information [Automatically generated, do not change]:
Machine: i386
OS: linux-gnu
Compiler: i386-redhat-linux-gcc
Compilation CFLAGS:  -DPROGRAM='bash' -DCONF_HOSTTYPE='i386'
- -DCONF_OSTYPE='linux-gnu' -DCONF_MACHTYPE='i386-redhat-linux-gnu'
- -DCONF_VENDOR='redhat' -DLOCALEDIR='/usr/share/locale' -DPACKAGE='bash'
- -DSHELL -DHAVE_CONFIG_H  -I.  -I. -I./include -I./lib  -D_GNU_SOURCE
- -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -O2 -g -pipe
- -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -m32 -march=i386 -mtune=pentium4
- -fasynchronous-unwind-tables
uname output: Linux tyrannosaurus 2.6.14-1.1656_FC4 #1 Thu Jan 5
22:13:22 EST 2006 i686 i686 i386 GNU/Linux
Machine Type: i386-redhat-linux-gnu

Bash Version: 3.0
Patch Level: 16
Release Status: release

Description:
In some cases, background processes receive SIGINT from the
terminal where it was started even though the manual states
that such processes are immune to keyboard-generated signals.
This is best explained by trying out the example below.

There is a reason for the seemingly complex sleep arrangement
in test2. This is really a condensed version of a much larger
script, but it still demonstrates the problem well.

Repeat-By:
1) Create two shell scripts, test1 and test2, which are
supplied   below:

test1:

### START SCRIPT ###
#!/bin/bash

test2 &

while /bin/true; do
echo test1 sleeping
sleep 5
done
### END SCRIPT ###

test2:

### START SCRIPT ###
#!/bin/bash

while /bin/true; do
echo test2 sleeping
sleep 5 &
SLEEP_PID=$!
wait $SLEEP_PID
unset SLEEP_PID
done
### END SCRIPT ###

2) Run test1 from the terminal
3) Interrupt it with CTRL-C
4) Notice that test2 was also interrupted. This can also be
   done by finding the PID of test2 and letting strace attach
   to the process before interrupting test1. Strace will report
   that a SIGINT was caught.

Fix:
N/A

-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFDw4016sVa+fPBo9YRAj0qAJ49V67E1gVRq+1sfKlx9ndYacEGkwCgrRoA
LvtXYfV3IZaOdFfiEypResw=
=gAR/
-END PGP SIGNATURE-


___
Bug-bash mailing list
Bug-bash@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-bash

Re: Background processes receive SIGINT even though they shouldn't

2006-01-13 Thread Ingemar Nilsson 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Chet Ramey wrote:

> I think you're slightly misunderstanding how job control and process
> groups interact.

Okay.

> When you run test1 from an interactive shell with job control enabled,
> it is placed into a separate process group.  The terminal's foreground
> process group is set to test1's process group, which means that group
> receives keyboard-generated signals like SIGINT.
> 
> The shell started to run test1 is not interactive, so job control is
> disabled.  This means that all processes started by test1 and its
> descendants are placed into the same process group.  Since this is
> the terminal's foreground process group, these processes receive the
> SIGINT.

I see, thanks. Then I have another question: Why does test2 below fail
to trap SIGINT when started by test1? It works when I run test2 directly.

test1 (for reference):

### START SCRIPT ###
#!/bin/bash

test2 &

while /bin/true; do
echo test1 sleeping
sleep 5
done
### END SCRIPT ###

test2 (updated):

### START SCRIPT ###
#!/bin/bash

function shutdown {
echo "Shutting down test2..."
exit 0
}

trap shutdown SIGINT

while /bin/true; do
echo test2 sleeping
sleep 5 &
SLEEP_PID=$!
wait $SLEEP_PID
unset SLEEP_PID
done
### END SCRIPT ###

Thanks in advance.
Regards
Ingemar
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFDx3Fd6sVa+fPBo9YRAmUoAJ9ARnxH4rirGcfegpg20/QxWozyBgCfa98o
UE1JbSi4kq6jDM9Q8iOZNLY=
=pxe6
-END PGP SIGNATURE-


___
Bug-bash mailing list
Bug-bash@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-bash

Re: Background processes receive SIGINT even though they shouldn't

2006-01-30 Thread Ingemar Nilsson 
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Chet Ramey wrote:

> I am looking at the apparent contradiction between this and the
> previous answer.  It might be that the shell begun to run test2
> undoes the signal ignoring performed when it is initally forked.
> It seems like somewhere along the line the SIGINT handling is being
> changed.

Any progress yet? I found a way to solve the problem anyway, using the
setsid command to run test2 from test1.

Regards
Ingemar
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iD8DBQFD3hE46sVa+fPBo9YRAgOXAJ9chAfPqOCYvZrIQOOUWdYuyJoRVQCgo8IQ
93kZCfq4IW/pVetyWJ7VFWU=
=kmdf
-END PGP SIGNATURE-


___
Bug-bash mailing list
Bug-bash@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-bash