[patch] clone-detached-2.5.31-A1

From: Ingo Molnar (mingo@elte.hu)
Date: Tue Aug 13 2002 - 13:58:35 EST


the attached patch implements the per-CPU thread-structure cache to do
detached exit, if the parent does not want to be notified of child exit
via a signal. It works wondefully. (tested patch against BK-current.)

i'm worried about one codepath though, in sys_wait4() we have this:

                                        p->parent = p->real_parent;
                                        add_parent(p, p->parent);
                                        do_notify_parent(p, SIGCHLD);

can this actually trigger for a detached process as well? In any case i've
put an assert in to make sure.

        Ingo

--- linux/include/linux/sched.h.orig Tue Aug 13 20:27:23 2002
+++ linux/include/linux/sched.h Tue Aug 13 20:50:18 2002
@@ -47,6 +47,7 @@
 #define CLONE_SYSVSEM 0x00040000 /* share system V SEM_UNDO semantics */
 #define CLONE_SETTLS 0x00080000 /* create a new TLS for the child */
 #define CLONE_SETTID 0x00100000 /* write the TID back to userspace */
+#define CLONE_DETACHED 0x00200000 /* parent wants no child-exit signal */
 
 #define CLONE_SIGNAL (CLONE_SIGHAND | CLONE_THREAD)
 
--- linux/kernel/exit.c.orig Tue Aug 13 19:55:04 2002
+++ linux/kernel/exit.c Tue Aug 13 20:52:46 2002
@@ -56,10 +56,11 @@
 
 static void release_task(struct task_struct * p)
 {
- if (p == current)
+ if (p->state != TASK_ZOMBIE)
                 BUG();
 #ifdef CONFIG_SMP
- wait_task_inactive(p);
+ if (p != current)
+ wait_task_inactive(p);
 #endif
         atomic_dec(&p->user->processes);
         security_ops->task_free_security(p);
@@ -67,10 +68,12 @@
         unhash_process(p);
 
         release_thread(p);
- current->cmin_flt += p->min_flt + p->cmin_flt;
- current->cmaj_flt += p->maj_flt + p->cmaj_flt;
- current->cnswap += p->nswap + p->cnswap;
- sched_exit(p);
+ if (p != current) {
+ current->cmin_flt += p->min_flt + p->cmin_flt;
+ current->cmaj_flt += p->maj_flt + p->cmaj_flt;
+ current->cnswap += p->nswap + p->cnswap;
+ sched_exit(p);
+ }
         put_task_struct(p);
 }
 
@@ -479,14 +482,15 @@
 
         write_lock_irq(&tasklist_lock);
         current->state = TASK_ZOMBIE;
- do_notify_parent(current, current->exit_signal);
+ if (current->exit_signal != -1)
+ do_notify_parent(current, current->exit_signal);
         while ((p = eldest_child(current))) {
                 list_del_init(&p->sibling);
                 p->ptrace = 0;
 
                 p->parent = p->real_parent;
                 list_add_tail(&p->sibling,&p->parent->children);
- if (p->state == TASK_ZOMBIE)
+ if (p->state == TASK_ZOMBIE && p->exit_signal != -1)
                         do_notify_parent(p, p->exit_signal);
                 /*
                  * process group orphan check
@@ -555,6 +559,9 @@
 
         tsk->exit_code = code;
         exit_notify();
+ preempt_disable();
+ if (current->exit_signal == -1)
+ release_task(current);
         schedule();
         BUG();
 /*
--- linux/kernel/signal.c.orig Tue Aug 13 19:55:07 2002
+++ linux/kernel/signal.c Tue Aug 13 20:53:43 2002
@@ -768,12 +768,15 @@
 /*
  * Let a parent know about a status change of a child.
  */
-
 void do_notify_parent(struct task_struct *tsk, int sig)
 {
         struct siginfo info;
         int why, status;
 
+ /* is the thread detached? */
+ if (sig == -1 || tsk->exit_signal == -1)
+ BUG();
+
         info.si_signo = sig;
         info.si_errno = 0;
         info.si_pid = tsk->pid;
@@ -823,9 +826,11 @@
 void
 notify_parent(struct task_struct *tsk, int sig)
 {
- read_lock(&tasklist_lock);
- do_notify_parent(tsk, sig);
- read_unlock(&tasklist_lock);
+ if (sig != -1) {
+ read_lock(&tasklist_lock);
+ do_notify_parent(tsk, sig);
+ read_unlock(&tasklist_lock);
+ }
 }
 
 #ifndef HAVE_ARCH_GET_SIGNAL_TO_DELIVER
--- linux/kernel/fork.c.orig Tue Aug 13 19:55:07 2002
+++ linux/kernel/fork.c Tue Aug 13 20:47:03 2002
@@ -50,6 +50,31 @@
 
 rwlock_t tasklist_lock __cacheline_aligned = RW_LOCK_UNLOCKED; /* outer */
 
+/*
+ * A per-CPU task cache - this relies on the fact that
+ * the very last portion of sys_exit() is executed with
+ * preemption turned off.
+ */
+static task_t *task_cache[NR_CPUS] __cacheline_aligned;
+
+void __put_task_struct(struct task_struct *tsk)
+{
+ if (tsk != current) {
+ free_thread_info(tsk->thread_info);
+ kmem_cache_free(task_struct_cachep,tsk);
+ } else {
+ int cpu = smp_processor_id();
+
+ tsk = task_cache[cpu];
+ if (tsk) {
+ free_thread_info(tsk->thread_info);
+ kmem_cache_free(task_struct_cachep,tsk);
+ }
+ task_cache[cpu] = current;
+ }
+}
+
+/* Protects next_safe and last_pid. */
 void add_wait_queue(wait_queue_head_t *q, wait_queue_t * wait)
 {
         unsigned long flags;
@@ -123,13 +148,6 @@
         return tsk;
 }
 
-void __put_task_struct(struct task_struct *tsk)
-{
- free_thread_info(tsk->thread_info);
- kmem_cache_free(task_struct_cachep,tsk);
-}
-
-/* Protects next_safe and last_pid. */
 spinlock_t lastpid_lock = SPIN_LOCK_UNLOCKED;
 
 static int get_pid(unsigned long flags)
@@ -737,7 +755,10 @@
 
         /* ok, now we should be set up.. */
         p->swappable = 1;
- p->exit_signal = clone_flags & CSIGNAL;
+ if (clone_flags & CLONE_DETACHED)
+ p->exit_signal = -1;
+ else
+ p->exit_signal = clone_flags & CSIGNAL;
         p->pdeath_signal = 0;
 
         /*

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Thu Aug 15 2002 - 22:00:33 EST