From: Richard Jones <rjones@redhat.com>
Date: Wed, 7 Jul 2010 14:01:38 +0000 (+0100)
Subject: fish: Don't fail if -m and --listen flags are both given (RHBZ#612178).
X-Git-Tag: 1.2.10~3
X-Git-Url: http://git.annexia.org/?a=commitdiff_plain;h=1bbd84b1ff8ed2a74522887499c8f2b5154eb363;p=libguestfs.git

fish: Don't fail if -m and --listen flags are both given (RHBZ#612178).

Testing this against a Fedora disk image:

$ ./fish/guestfish --ro -a F13.img -m /dev/sda1 --listen
export GUESTFISH_PID=6033
$ ./fish/guestfish --remote=6033 -- ping-daemon
$ ./fish/guestfish --remote=6033 -- ping-daemon
$ ./fish/guestfish --remote=6033 -- exit

Without this fix the first remote command would fail because
qemu would have already been killed.

For stable-1.2 branch:
 - Cherry picked from commit 8161ea9bb046c8450384b5b15e1f4b2f3dca582b
 - Rebased for older guestfish code.
---

diff --git a/fish/fish.c b/fish/fish.c
index a6db085..b962c3c 100644
--- a/fish/fish.c
+++ b/fish/fish.c
@@ -405,6 +405,17 @@ main (int argc, char *argv[])
 
   /* If we've got mountpoints, we must launch the guest and mount them. */
   if (mps != NULL) {
+    /* RHBZ#612178: If --listen flag is given, then we will fork into
+     * the background in rc_listen().  However you can't do this while
+     * holding a libguestfs handle open because the recovery process
+     * will think the main program has died and kill qemu.  Therefore
+     * don't use the recovery process for this case.  (A better
+     * solution would be to call launch () etc after the fork, but
+     * that greatly complicates the code here).
+     */
+    if (remote_control_listen)
+      guestfs_set_recovery_proc (g, 0);
+
     if (launch (g) == -1) exit (EXIT_FAILURE);
     mount_mps (mps);
   }