xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Ian Jackson <ian.jackson@eu.citrix.com>
To: xen-devel@lists.xensource.com
Cc: George Dunlap <george.dunlap@eu.citrix.com>,
	Jim Fehlig <jfehlig@suse.com>,
	Ian Jackson <Ian.Jackson@eu.citrix.com>,
	Ian Campbell <Ian.Campbell@citrix.com>
Subject: [PATCH 07/18] libxl: fork: Provide ..._always_selective_reap
Date: Mon, 3 Feb 2014 16:14:40 +0000	[thread overview]
Message-ID: <1391444091-22796-8-git-send-email-ian.jackson@eu.citrix.com> (raw)
In-Reply-To: <1391444091-22796-1-git-send-email-ian.jackson@eu.citrix.com>

Applications exist which want to use libxl in an event-driven mode but
which do not integrate child termination into their event system, but
instead reap all their own children synchronously.

In such an application libxl must own SIGCHLD but avoid reaping any
children that don't belong to libxl.

Provide libxl_sigchld_owner_libxl_always_selective_reap which has this
behaviour.

Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
Cc: Jim Fehlig <jfehlig@suse.com>
Cc: Ian Campbell <Ian.Campbell@citrix.com>
Acked-by: Ian Campbell <Ian.Campbell@citrix.com>

---
v2: Document the new mode in the big "Subprocess handling" comment.
---
 tools/libxl/libxl_event.h |   11 +++++++++++
 tools/libxl/libxl_fork.c  |    7 +++++++
 2 files changed, 18 insertions(+)

diff --git a/tools/libxl/libxl_event.h b/tools/libxl/libxl_event.h
index 3c93955..824ac88 100644
--- a/tools/libxl/libxl_event.h
+++ b/tools/libxl/libxl_event.h
@@ -474,6 +474,12 @@ void libxl_osevent_occurred_timeout(libxl_ctx *ctx, void *for_libxl)
  *       and provides a callback to be notified of their exit
  *       statues.  The application must have only one libxl_ctx
  *       configured this way.
+ *
+ *     libxl_sigchld_owner_libxl_always_selective_reap:
+ *
+ *       The application expects to reap all of its own children
+ *       synchronously, and does not use SIGCHLD.  libxl is
+ *       to install a SIGCHLD handler.
  */
 
 
@@ -491,6 +497,11 @@ typedef enum {
     /* libxl owns SIGCHLD all the time, and the application is
      * relying on libxl's event loop for reaping its children too. */
     libxl_sigchld_owner_libxl_always,
+
+    /* libxl owns SIGCHLD all the time, but it must only reap its own
+     * children.  The application will reap its own children
+     * synchronously with waitpid, without the assistance of SIGCHLD. */
+    libxl_sigchld_owner_libxl_always_selective_reap,
 } libxl_sigchld_owner;
 
 typedef struct {
diff --git a/tools/libxl/libxl_fork.c b/tools/libxl/libxl_fork.c
index b2325e0..16e17f6 100644
--- a/tools/libxl/libxl_fork.c
+++ b/tools/libxl/libxl_fork.c
@@ -268,6 +268,7 @@ static bool chldmode_ours(libxl_ctx *ctx, bool creating)
     case libxl_sigchld_owner_mainloop:
         return 0;
     case libxl_sigchld_owner_libxl_always:
+    case libxl_sigchld_owner_libxl_always_selective_reap:
         return 1;
     }
     abort();
@@ -398,6 +399,12 @@ static void sigchld_selfpipe_handler(libxl__egc *egc, libxl__ev_fd *ev,
     int e = libxl__self_pipe_eatall(selfpipe);
     if (e) LIBXL__EVENT_DISASTER(egc, "read sigchld pipe", e, 0);
 
+    if (CTX->childproc_hooks->chldowner
+        == libxl_sigchld_owner_libxl_always_selective_reap) {
+        childproc_checkall(egc);
+        return;
+    }
+
     while (chldmode_ours(CTX, 0) /* in case the app changes the mode */) {
         int status;
         pid_t pid = checked_waitpid(egc, -1, &status);
-- 
1.7.10.4

  parent reply	other threads:[~2014-02-03 16:14 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-02-03 16:14 [PATCH 00/18 v3] libxl: fork and event fixes for libvirt and 4.4 Ian Jackson
2014-02-03 16:14 ` [PATCH 01/18] libxl: fork: Break out checked_waitpid Ian Jackson
2014-02-03 16:14 ` [PATCH 02/18] libxl: fork: Break out childproc_reaped_ours Ian Jackson
2014-02-03 16:14 ` [PATCH 03/18] libxl: fork: Clarify docs for libxl_sigchld_owner Ian Jackson
2014-02-03 16:14 ` [PATCH 04/18] libxl: fork: Document libxl_sigchld_owner_libxl better Ian Jackson
2014-02-03 16:14 ` [PATCH 05/18] libxl: fork: assert that chldmode is right Ian Jackson
2014-02-03 16:14 ` [PATCH 06/18] libxl: fork: Provide libxl_childproc_sigchld_occurred Ian Jackson
2014-02-03 16:14 ` Ian Jackson [this message]
2014-02-03 16:14 ` [PATCH 08/18] libxl: fork: Provide LIBXL_HAVE_SIGCHLD_SELECTIVE_REAP Ian Jackson
2014-02-03 16:14 ` [PATCH 09/18] libxl: fork: Rename sigchld handler functions Ian Jackson
2014-02-03 16:14 ` [PATCH 10/18] libxl: fork: Break out sigchld_installhandler_core Ian Jackson
2014-02-03 16:14 ` [PATCH 11/18] libxl: fork: Break out sigchld_sethandler_raw Ian Jackson
2014-02-06 13:53   ` Ian Campbell
2014-02-03 16:14 ` [PATCH 12/18] libxl: fork: Share SIGCHLD handler amongst ctxs Ian Jackson
2014-02-03 16:14 ` [PATCH 13/18] libxl: events: Break out libxl__pipe_nonblock, _close Ian Jackson
2014-02-03 16:14 ` [PATCH 14/18] libxl: fork: Make SIGCHLD self-pipe nonblocking Ian Jackson
2014-02-03 16:14 ` [PATCH 15/18] libxl: events: Makefile builds internal unit tests Ian Jackson
2014-02-06 14:00   ` Ian Campbell
2014-02-03 16:14 ` [PATCH 16/18] libxl: events: timedereg internal unit test Ian Jackson
2014-02-06 14:01   ` Ian Campbell
2014-02-03 16:14 ` [PATCH 17/18] libxl: timeouts: Break out time_occurs Ian Jackson
2014-02-06 14:02   ` Ian Campbell
2014-02-03 16:14 ` [PATCH 18/18] libxl: timeouts: Record deregistration when one occurs Ian Jackson
2014-02-06 14:04   ` Ian Campbell
2014-02-06 14:24     ` Ian Jackson
2014-02-06 14:27       ` Ian Campbell
2014-02-03 16:16 ` [PATCH 00/18 v3] libxl: fork and event fixes for libvirt and 4.4 Ian Jackson
2014-02-05  5:46   ` Jim Fehlig
2014-02-05 11:21     ` Ian Jackson
2014-02-05 14:10 ` George Dunlap
2014-02-05 15:03   ` Ian Jackson
2014-02-06 10:52     ` George Dunlap
2014-02-06 12:35       ` Ian Jackson
2014-02-06 14:07         ` Ian Campbell
2014-02-06 14:33           ` Ian Jackson
2014-02-07  4:17       ` Jim Fehlig

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1391444091-22796-8-git-send-email-ian.jackson@eu.citrix.com \
    --to=ian.jackson@eu.citrix.com \
    --cc=Ian.Campbell@citrix.com \
    --cc=george.dunlap@eu.citrix.com \
    --cc=jfehlig@suse.com \
    --cc=xen-devel@lists.xensource.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).