158

I have detached a process from my terminal, like this:

$ process &

That terminal is now long closed, but process is still running, and I want to send some commands to that process's stdin. Is that possible?

Rogach
  • 6,303

5 Answers5

153

Yes, it is. First, create a pipe: mkfifo /tmp/fifo. Use gdb to attach to the process: gdb -p PID

Then close stdin: call close (0); and open it again: call open ("/tmp/fifo", 0600)

Finally, write away (from a different terminal, as gdb will probably hang):

echo blah > /tmp/fifo

  • 10
    Very impressive! – Samuel Edwin Ward Feb 16 '12 at 18:59
  • 1
    Can I do something similar to redirect process stdout to a file? – rustyx Aug 25 '14 at 15:28
  • @rustyx: Untested, but this should work: create a file rather than a pipe, touch /tmp/thefile. Stdout is 1, so call close (1); also, use the correct permissions for writing: call open ("/tmp/thefile", 0400). The echo… is, of course, not needed. – Ansgar Esztermann Aug 26 '14 at 11:25
  • This is great! I'm using this to send "y" or "n" responses to certain processes that have been completely detached. The detached process has its stdout to a separate window. When I do this trick however, I can see that it does not "receive" the 'y' or 'n' as soon as I echo it, I must quit gdb and detach it and then it receives all of the echos accordingly, so is there a way to perform this without needing to quit gdb before the process receives the input from the fifo? – krb686 Jul 27 '15 at 13:50
  • unfortunately it just seem hung on call open("/tmp/fifo", 0600). any help would be greatly appreciated – archer Dec 22 '18 at 13:34
  • i think you should use continue command of gdb to continue its process after doing the commands given to it – AshkanVZ Jun 03 '19 at 15:32
  • 2
    is this still valid? unfortunately does not work for me (ubuntu 18) – d9ngle Apr 09 '20 at 19:29
  • I believe this should be call open ("/tmp/fifo", 0) (Open for reading.) 0600 would be appropriate for a third argument to open if the second argument has the O_CREAT (0100) flag. Using 0600 as a second argument, it's O_RDONLY|O_EXCL|O_NOCTTY on Linux, which seems to be OK, but O_EXCL without O_CREAT is undefined (unless the file in a block device). – Mark Plotnick Dec 01 '21 at 20:49
58

When original terminal is no longer accessible...

reptyr might be what you want, see https://serverfault.com/a/284795/187998

Quote from there:

Have a look at reptyr, which does exactly that. The github page has all the information.

reptyr - A tool for "re-ptying" programs.

reptyr is a utility for taking an existing running program and attaching it to a new terminal. Started a long-running process over ssh, but have to leave and don't want to interrupt it? Just start a screen, use reptyr to grab it, and then kill the ssh session and head on home.

USAGE

reptyr PID

"reptyr PID" will grab the process with id PID and attach it to your current terminal.

After attaching, the process will take input from and write output to the new terminal, including ^C and ^Z. (Unfortunately, if you background it, you will still have to run "bg" or "fg" in the old terminal. This is likely impossible to fix in a reasonable way without patching your shell.)

Edit claims that "reptyr cannot grab a process which has subprocesses. Or the subprocess (reptyr version 0.6.2)." Limited support does exist Issue, Issue

NiKiZe
  • 699
  • 5
  • 11
  • If it's not installed, try sudo yum install -y reptyr. This worked great for me to take over a vi editing session I lost when my SSH connection dropped (just had to press ^L after taking it back over to see all my changes and continue editing). – Gary Jan 29 '20 at 01:29
  • 3
    I did not have this installed, and could not use the package manager because the process I was disconnected from was dpkg. I downloaded the deb (https://pkgs.org/download/reptyr), did a dpkg-deb --extract reptyr_0.6.2-1.2_amd64.deb reptyr && reptyr/usr/bin/reptyr $PID_OF_WANTED_PROCESS – miigotu Apr 03 '20 at 05:59
  • Did essentially the same as @miigotu: downloaded (what I hoped was) the right package from https://pkgs.org/download/reptyr and extracted reptyr... was able to save a file from a detached VI session. – TripeHound Apr 13 '20 at 12:57
  • 1
    In my scenario, I was able to use the newer -T argument like reptyr -T <pid> to reconnect (I was running as root already btw) – Mike Dearman Dec 08 '22 at 21:10
  • @miigotu : Thank you for this!! Just had to perform the same emergency maneuver due to nouveau driver locking up the display on a system running apt-get + dpkg-reconfigure on thousands of packages during an upgrade. Very inopportune time for nouveau kernel module to crash, but now I can SSH in and use reptyr to make sure the upgrade is complete before rebooting the system. – TrinitronX May 23 '23 at 22:55
14

I am quite sure you can not.

Check using ps x. If a process has a ? as controlling tty, you can not send input to it any more.

9942 ?        S      0:00 tail -F /var/log/messages
9947 pts/1    S      0:00 tail -F /var/log/messages

In this example, you can send input to 9947 doing something like echo "test" > /dev/pts/1. The other process (9942) is not reachable.

Next time, you could use screen or tmux to avoid this situation.

andcoz
  • 17,130
8

Just ending the command line with & will not completely detach the process, it will just run it in the background. (With zsh you can use &! to actually detach it, otherwise you have do disown it later).

When a process runs in the background, it won't receive input from its controlling terminal anymore. But you can send it back into the foreground with fg and then it will read input again.

Otherwise, it's not possible to externally change its filedescriptors (including stdin) or to reattach a lost controlling terminal… unless you use debugging tools (see Ansgar's answer, or have a look at the retty command).

7

EDIT : As Stephane Gimenez said, it's not that simple. It's only allowing you to print to a different terminal.

You can try to write to this process using /proc. It should be located in /proc/pid/fd/0, so a simple :

echo "hello" > /proc/PID/fd/0

should do it. I have not tried it, but it should work, as long as this process still has a valid stdin file descriptor. You can check it with ls -l on /proc/pid/fd/.

  • if it's a link to /dev/null => it's closed
  • if it's a link to /dev/pts/X or a socket => it's open

See nohup for more details about how to keep processes running.

Coren
  • 5,010
  • 3
    It's not that simple. For example, If stdin is linked to a terminal, echoing something to the terminal device will just print what you wrote on the terminal, it will not be transmitted to the process. – Stéphane Gimenez Feb 16 '12 at 16:16