[Libguestfs] [PATCH 2/8] New APIs: set-pgroup, get-pgroup

Richard W.M. Jones rjones at redhat.com
Fri Jul 15 14:22:55 UTC 2011


-- 
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
virt-top is 'top' for virtual machines.  Tiny program with many
powerful monitoring features, net stats, disk stats, logging, etc.
http://et.redhat.com/~rjones/virt-top
-------------- next part --------------
>From f173543fd207bdc254a5eb75180d82ef25eacae9 Mon Sep 17 00:00:00 2001
From: "Richard W.M. Jones" <rjones at redhat.com>
Date: Fri, 15 Jul 2011 11:38:21 +0100
Subject: [PATCH 2/8] New APIs: set-pgroup, get-pgroup

If the pgroup flag is set in the handle, then the qemu and recovery
subprocesses are placed in separate process groups.  The default is
false.

The purpose for setting up a process group is that ^C will not be
passed from the main process down to these processes (killing them).
This allows ^C and other keyboard events to be caught and handled in
the main process.
---
 generator/generator_actions.ml |   19 +++++++++++++++++++
 src/guestfs-internal.h         |    2 ++
 src/guestfs.c                  |   13 +++++++++++++
 src/launch.c                   |   17 +++++++++++------
 4 files changed, 45 insertions(+), 6 deletions(-)

diff --git a/generator/generator_actions.ml b/generator/generator_actions.ml
index 1ccacdb..87c934a 100644
--- a/generator/generator_actions.ml
+++ b/generator/generator_actions.ml
@@ -1526,6 +1526,25 @@ advice before using trademarks in applications.
 
 =back");
 
+  ("set_pgroup", (RErr, [Bool "pgroup"], []), -1, [FishAlias "pgroup"],
+   [],
+   "set process group flag",
+   "\
+If C<pgroup> is true, child processes are placed into
+their own process group.
+
+The practical upshot of this is that signals like C<SIGINT> (from
+users pressing C<^C>) won't be received by the child process.
+
+The default for this flag is false, because usually you want
+C<^C> to kill the subprocess.");
+
+  ("get_pgroup", (RBool "pgroup", [], []), -1, [],
+   [],
+   "get process group flag",
+   "\
+This returns the process group flag.");
+
 ]
 
 (* daemon_functions are any functions which cause some action
diff --git a/src/guestfs-internal.h b/src/guestfs-internal.h
index 96f8152..e2ffdf3 100644
--- a/src/guestfs-internal.h
+++ b/src/guestfs-internal.h
@@ -170,6 +170,8 @@ struct guestfs_h
 
   int selinux;                  /* selinux enabled? */
 
+  int pgroup;                   /* Create process group for children? */
+
   char *last_error;
   int last_errnum;              /* errno, or 0 if there was no errno */
 
diff --git a/src/guestfs.c b/src/guestfs.c
index b02bdb9..e2b7159 100644
--- a/src/guestfs.c
+++ b/src/guestfs.c
@@ -801,6 +801,19 @@ guestfs__get_attach_method (guestfs_h *g)
   return ret;
 }
 
+int
+guestfs__set_pgroup (guestfs_h *g, int v)
+{
+  g->pgroup = !!v;
+  return 0;
+}
+
+int
+guestfs__get_pgroup (guestfs_h *g)
+{
+  return g->pgroup;
+}
+
 /* Note the private data area is allocated lazily, since the vast
  * majority of callers will never use it.  This means g->pda is
  * likely to be NULL.
diff --git a/src/launch.c b/src/launch.c
index 0b15ce9..1a47363 100644
--- a/src/launch.c
+++ b/src/launch.c
@@ -643,12 +643,9 @@ launch_appliance (guestfs_h *g)
       close (rfd[1]);
     }
 
-#if 0
-    /* Set up a new process group, so we can signal this process
-     * and all subprocesses (eg. if qemu is really a shell script).
-     */
-    setpgid (0, 0);
-#endif
+    /* Put qemu in a new process group. */
+    if (g->pgroup)
+      setpgid (0, 0);
 
     setenv ("LC_ALL", "C", 1);
 
@@ -677,6 +674,14 @@ launch_appliance (guestfs_h *g)
       pid_t qemu_pid = g->pid;
       pid_t parent_pid = getppid ();
 
+      /* It would be nice to be able to put this in the same process
+       * group as qemu (ie. setpgid (0, qemu_pid)).  However this is
+       * not possible because we don't have any guarantee here that
+       * the qemu process has started yet.
+       */
+      if (g->pgroup)
+        setpgid (0, 0);
+
       /* Writing to argv is hideously complicated and error prone.  See:
        * http://anoncvs.postgresql.org/cvsweb.cgi/pgsql/src/backend/utils/misc/ps_status.c?rev=1.33.2.1;content-type=text%2Fplain
        */
-- 
1.7.5.2



More information about the Libguestfs mailing list