Running a bash script in the background with job control enabled and stdin closed will exit the PARENT shell. How can that happen?
To demonstrate make this background_bash_script:
#!/bin/bash
set -m
ruby -e "puts :here"
Then run it in bash - it will exit the shell you ran it in. The ruby command does not matter although it appears it must be a command and not a bash built-in (for example awk --version works but true does not). To get a better look I've been running it in yet another instance of bash. A full session looks like this.
parent: PS1='child: ' bash
child: ./background_bash_script <&- &
[1] 3893
child: here
exit
parent:
Confusing!
What seems like is happening is that after set -m is run in the script, the next command that is run is forced to be in the foreground process group, which takes the original shell out of the foreground process group. Once that process exits, the shell running the script is now in the foreground process group, but once that shell exits, the original shell doesn't put itself back into the foreground process group because it ran the script in the background. So you now have an interactive shell that is in a background process group.
You can see some weird behavior here if you put a sleep at the end of your script so that it doesn't exit immediately. When you run the script in the background you get the terminal prompt back, but now your interactive shell isn't in the foreground process group! As soon as you try to type anything the shell exits. I'm not sure exactly what mechanism causes the exit. Since the shell is in the background, any attempts to read or write characters to the terminal should result in SIGTTIN OR SIGTTOU, but these signals don't cause the shell to exit in my tests.
Related
I have two scenarios:
#!/usr/bin/env bash
sleep infinity
# When I type Ctrl-C here, "sleep" command and script are stopped so I didn't see "End"
echo End
#!/usr/bin/env bash
docker exec container-id sleep infinity
# When I type Ctrl-C here, "docker exec" command is stopped but script continued so I saw "End"
echo End
Why the difference in behaviour?
That's how bash behaves when its process group receives a SIGINT but the program currently running on the foreground terminates normally.
The rationale for this behavior is given here as follows:
The basic idea is that the user intends a keyboard-generated SIGINT to go
to the foreground process; that process gets to decide how to handle it;
and bash reacts accordingly. If the process dies to due SIGINT, bash acts
as if it received the SIGINT; if it does not, bash assumes the process
handled it and effectively ignores it.
Consider a process (emacs is the usual example) that uses SIGINT for its
own purposes as a normal part of operation. If you run that program in a
script, you don't want the shell aborting the script unexpectedly as a
result.
I'm trying to achieve the following:
from a fish script, open a PDF reader as a background job. Once it is opened, spawn another fish process (that runs an infinite while loop), also as a background job.
Next, open an editor (neovim) and allow it to take control of the running terminal. Once neovim terminates, also suspend the previous 2 background jobs (mupdf and the other fish process).
My current attempt looks something along the lines of:
mupdf $pdfpath &
set pid_mupdf $last_pid
fish -c "while inotifywait ...; [logic to rebuild the pdf file..]; end" &
set pid_sub $last_pid
nvim $mdpath && kill -2 $pid_mudf $pid_sub
First I open mupdf as a background job and save its PID in a variable. Next I spawn the other fish process, also as a background job, and I save its PID as well.
Next I run nvim (but not as a background job, as I intend to actually control it), and after it is terminated by the user, I gracefully kill the previous 2 background jobs.
However this doesn't work as intended.
mupdf and the second fish process open successfully, and so does nvim, but it quickly closes after around half a second, after which I get the following in the controlling terminal window: image (bote is just the filename of the script from which the lines above originate)
The 2 background processes stay running after that and I have to kill them manually.
I understand that the script is sent a SIGHUP because the controlling terminal now executes another application (neovim), but why does neovim close after that?
I also tried disowning the background processes after they're spawned but that didn't help.
How would I solve this issue?
The problem is that $last_pid, in fish 3, and %last, in fish 2, doesn't work by default in scripts. See https://github.com/fish-shell/fish-shell/issues/5036. You can "fix" this by putting status job-control full at the top of the script or using the (jobs -lp) hack that Glenn mentioned.
Regarding the background process remaining running... I can't reproduce that. It works for me. However, note that your nvim && kill will only run the kill if nvim exits with a status of zero. If you always want the kill to be run you should just unconditionally execute it. Also, your use of signal two (SIGINT) should produce the desired result but is unusual. You should use kill -15 or just omit the signal in which case it defaults to 15 (SIGTERM).
You're getting the PID incorrectly. The $pid_mudf and $pid_sub variables are empty. You want
set pid_mupdf (jobs -lp)
When a process is started with "&",
bash places the new process to foreground group for a short time
and then places
itself back to foreground group.
As far as I know, when we run any external
command without the "&" background operator, the shell creates a
new process group for the command and notifies the terminal that this process group is now
in the foreground. When we run with "&", foreground process group MUST NOT change.
I used a C program on linux which just prints foreground process group and
own process group on start and then prints the same after a sleep.
I get the following results:
tcgetpgrp=19981 getpgrp=19981
sleep(5)
tcgetpgrp=11996 getpgrp=19981
In this example 11996 is the pid of bash.
I only noticed this behavior when run as (./test&) and under heavy load (never as ./test&).
Is bash behaving the right way?
Bash version: 4.4.12
EDIT
Added pid value as suggested by #Costi:
tcgetpgrp=29148 getpgrp=29148 getpid=29149
sleep
tcgetpgrp=28566 getpgrp=29148 getpid=29149
28566 is pid of bash
(...) creates a subshell in bash. As per bash manual:
(list) list is executed in a subshell environment (see COMMAND EXECUTION ENVIRONMENT below)
My guess is you are seeing the PID of that subshell as the foreground process, not the PID of your test command. And the subshell exits quite quickly (right after firing your command in background).
To validate this, add the value returned by getpid to the mix and check if that is the same as your getpgrp value when run in a subshell.
I am using bash on Ubuntu. I would like to have a shell script open a program and continue on to the next line of the shell script, even though the program has not terminated.
Adding an & to a command places it in background.
example:
/path/to/foo
/path/to/bar # not executed untill foo is done
/path/to/foo & # in background
/path/to/bar & # executes as soon as foo is started
Read more about job-control here and here
Use something like this (my-long-running-process &) . This will launch your script as a separate process in the background.
You must run the process in the background, but you must enable job-control first. Otherwise, you cannot kill or bring the process to foreground if desired.
To enable job-control, execute:
set -m
To run some task in the background, execute:
task &
To manipulate the background task, use the jobspec syntax (%[n]). For example, to kill the last launched process, execute:
kill %
Note that enabling job-control is required only if you're actually running a script (as stated in the question). If running interactively, job-control is already enabled by default.
The manpage for bash has much more information in the JOB CONTROL section.
http://ubuntuforums.org/showthread.php?t=1657602
It looks like all you have to do is add a & at the end of the line.
I'd like to run a script every time I close a Bash session.
I use XFCE and Terminal 0.4.5 (Xfce Terminal Emulator), I would like to run a script every time I close a tab in Terminal including the last one (when I close Terminal).
Something like .bashrc but running at the end of every session.
.bash_logout doesn't work
You use trap (see man bash):
trap /u1/myuser/on_exit_script.sh EXIT
The command can be added to your .profile/.login
This works whether you exit the shell normally (e.g. via exit command) or simply kill the terminal window/tab, since the shell gets the EXIT signal either way - I just tested by exiting my putty window.
My answer is similar to DVK's answer but you have to use a command or function, not a file.
$ man bash
[...]
trap [-lp] [[arg] sigspec ...]
The command arg is to be read and executed when the shell
receives signal(s) sigspec.
[...]
If a sigspec is EXIT (0) the command arg is executed on
exit from the shell.
So, you can add to your .bashrc something like the following code:
finish() {
# Your code here
}
trap finish EXIT
Write you script in "~/.bash_logout". It executed by bash(1) when login shell exits.
If you close your session with "exit", might be able to something like
alias endbash="./runscript;exit" and just exit by entering endbash. I'm not entirely sure this works, as I'm running windows at the moment.
edit: DVK has a better answer.