The patch titled
Subject: Revert "kmod: handle UMH_WAIT_PROC from system unbound workqueue"
has been added to the -mm tree. Its filename is
revert-kmod-handle-umh_wait_proc-from-system-unbound-workqueue.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/revert-kmod-handle-umh_wait_proc-from-system-unbound-workqueue.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/revert-kmod-handle-umh_wait_proc-from-system-unbound-workqueue.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Oleg Nesterov <oleg@xxxxxxxxxx>
Subject: Revert "kmod: handle UMH_WAIT_PROC from system unbound workqueue"
This reverts bb304a5c6fc63d8506c ("kmod: handle UMH_WAIT_PROC from system
unbound workqueue") because this patch leads to kthread zombies.
call_usermodehelper_exec_sync() does fork() + wait() with "unignored"
SIGCHLD. What we have missed is that this worker thread can have other
children previously forked by call_usermodehelper_exec_work() without
UMH_WAIT_PROC. If such a child exits in between it becomes a zombie and
nobody can reap it (unless/until this worker thread exits too).
Signed-off-by: Oleg Nesterov <oleg@xxxxxxxxxx>
Cc: Frederic Weisbecker <fweisbec@xxxxxxxxx>
Cc: Rik van Riel <riel@xxxxxxxxxx>
Cc: Christoph Lameter <cl@xxxxxxxxx>
Cc: Tejun Heo <tj@xxxxxxxxxx>
Cc: Rusty Russell <rusty@xxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
kernel/kmod.c | 44 ++++++++++++++++++++++++--------------------
1 file changed, 24 insertions(+), 20 deletions(-)
diff -puN kernel/kmod.c~revert-kmod-handle-umh_wait_proc-from-system-unbound-workqueue kernel/kmod.c
--- a/kernel/kmod.c~revert-kmod-handle-umh_wait_proc-from-system-unbound-workqueue
+++ a/kernel/kmod.c
@@ -265,9 +265,15 @@ out:
do_exit(0);
}
-/* Handles UMH_WAIT_PROC. */
-static void call_usermodehelper_exec_sync(struct subprocess_info *sub_info)
+/*
+ * Handles UMH_WAIT_PROC. Our parent (unbound workqueue) might not be able to
+ * run enough instances to handle usermodehelper completions without blocking
+ * some other pending requests. That's why we use a kernel thread dedicated for
+ * that purpose.
+ */
+static int call_usermodehelper_exec_sync(void *data)
{
+ struct subprocess_info *sub_info = data;
pid_t pid;
/* If SIGCLD is ignored sys_wait4 won't populate the status. */
@@ -281,9 +287,9 @@ static void call_usermodehelper_exec_syn
* Normally it is bogus to call wait4() from in-kernel because
* wait4() wants to write the exit code to a userspace address.
* But call_usermodehelper_exec_sync() always runs as kernel
- * thread (workqueue) and put_user() to a kernel address works
- * OK for kernel threads, due to their having an mm_segment_t
- * which spans the entire address space.
+ * thread and put_user() to a kernel address works OK for kernel
+ * threads, due to their having an mm_segment_t which spans the
+ * entire address space.
*
* Thus the __user pointer cast is valid here.
*/
@@ -298,21 +304,19 @@ static void call_usermodehelper_exec_syn
sub_info->retval = ret;
}
- /* Restore default kernel sig handler */
- kernel_sigaction(SIGCHLD, SIG_IGN);
-
umh_complete(sub_info);
+ do_exit(0);
}
/*
- * We need to create the usermodehelper kernel thread from a task that is affine
+ * This function doesn't strictly needs to be called asynchronously. But we
+ * need to create the usermodehelper kernel threads from a task that is affine
* to an optimized set of CPUs (or nohz housekeeping ones) such that they
* inherit a widest affinity irrespective of call_usermodehelper() callers with
* possibly reduced affinity (eg: per-cpu workqueues). We don't want
* usermodehelper targets to contend a busy CPU.
*
- * Unbound workqueues provide such wide affinity and allow to block on
- * UMH_WAIT_PROC requests without blocking pending request (up to some limit).
+ * Unbound workqueues provide such wide affinity.
*
* Besides, workqueues provide the privilege level that caller might not have
* to perform the usermodehelper request.
@@ -322,18 +326,18 @@ static void call_usermodehelper_exec_wor
{
struct subprocess_info *sub_info =
container_of(work, struct subprocess_info, work);
+ pid_t pid;
- if (sub_info->wait & UMH_WAIT_PROC) {
- call_usermodehelper_exec_sync(sub_info);
- } else {
- pid_t pid;
-
+ if (sub_info->wait & UMH_WAIT_PROC)
+ pid = kernel_thread(call_usermodehelper_exec_sync, sub_info,
+ CLONE_FS | CLONE_FILES | SIGCHLD);
+ else
pid = kernel_thread(call_usermodehelper_exec_async, sub_info,
SIGCHLD);
- if (pid < 0) {
- sub_info->retval = pid;
- umh_complete(sub_info);
- }
+
+ if (pid < 0) {
+ sub_info->retval = pid;
+ umh_complete(sub_info);
}
}
_
Patches currently in -mm which might be from oleg@xxxxxxxxxx are
revert-kmod-handle-umh_wait_proc-from-system-unbound-workqueue.patch
mmoom-fix-potentially-killing-unrelated-process-fix.patch
mm-fix-the-racy-mm-locked_vm-change-in.patch
mm-add-the-struct-mm_struct-mm-local-into.patch
mm-oom_kill-remove-the-wrong-fatal_signal_pending-check-in-oom_kill_process.patch
mm-oom_kill-cleanup-the-kill-sharing-same-memory-loop.patch
mm-oom_kill-fix-the-wrong-task-mm-==-mm-checks-in-oom_kill_process.patch
change-current_is_single_threaded-to-use-for_each_thread.patch
signals-kill-block_all_signals-and-unblock_all_signals.patch
signal-turn-dequeue_signal_lock-into-kernel_dequeue_signal.patch
signal-introduce-kernel_signal_stop-to-fix-jffs2_garbage_collect_thread.patch
signal-remove-jffs2_garbage_collect_thread-allow_signalsigcont.patch
coredump-ensure-all-coredumping-tasks-have-signal_group_coredump.patch
coredump-change-zap_threads-and-zap_process-to-use-for_each_thread.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
