Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757013Ab1FUU0T (ORCPT ); Tue, 21 Jun 2011 16:26:19 -0400 Received: from mx1.redhat.com ([209.132.183.28]:4064 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756679Ab1FUU0S (ORCPT ); Tue, 21 Jun 2011 16:26:18 -0400 Date: Tue, 21 Jun 2011 22:23:30 +0200 From: Oleg Nesterov To: Tejun Heo Cc: linux-kernel@vger.kernel.org, akpm@linux-foundation.org, torvalds@linux-foundation.org, hch@infradead.org Subject: Re: [PATCH 6/7] ptrace: kill detah tracehooks Message-ID: <20110621202330.GA6738@redhat.com> References: <1308322240-8247-1-git-send-email-tj@kernel.org> <1308322240-8247-7-git-send-email-tj@kernel.org> <20110620193929.GA15228@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20110620193929.GA15228@redhat.com> User-Agent: Mutt/1.5.18 (2008-05-17) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 5347 Lines: 154 On 06/20, Oleg Nesterov wrote: > > Note! This looks very ugly, I wanted to cleanup this a long ago. > I think we should never change ->exit_signal, and do_notify_parent() > should return a boolean. I'll try to make the patches tomorrow. > Then _perhaps_ this patch will become simpler. Damn. Tomorrow (I promise ;) I'll send the full series. IMHO, we really need to cleanup the do_notify_parent/task_detached logic, and exit_signal should be "const" But, to remove this tracehook, we only need the patch below and then exit_notify() should do: if (unlikely(tsk->ptrace)) { int sig = ptrace_reparented(tsk) || task_detached(tsk) ? SIGCHLD : tsk->exit_signal; autoreap = do_notify_parent(tsk, sig); WARN_ON(autoreap); } else if (thread_group_leader(tsk)) { autoreap = thread_group_empty(tsk) && do_notify_parent(tsk, tsk->exit_signal); } else { autoreap = true; } tsk->exit_state = autoreap ? EXIT_DEAD : EXIT_ZOMBIE; Note the WARN_ON() above, perhaps instead we should autoreap if the tracer is the real parent and it ignores SIGCHLD. And. ptrace_reparented() should be changed to check same_thread_group(), I think. This also looks fine for wait_task_zombie(). Oleg. ----------------------------------------------------------------------------- [PATCH 3/XXX] make do_notify_parent() return bool - change do_notify_parent() to return a boolean, true if the task should be reaped because its parent ignores SIGCHLD. - update the only caller which checks the returned value, exit_notify(), to rely DEATH_REAP only if we do not call do_notify_parent(). This temporary uglifies exit_notify() even more, will be cleanuped by the next change. Signed-off-by: Oleg Nesterov --- --- ptrace/include/linux/sched.h~2_do_notify_parent_bool 2011-06-17 20:12:29.000000000 +0200 +++ ptrace/include/linux/sched.h 2011-06-21 19:18:00.000000000 +0200 @@ -2146,7 +2146,7 @@ static inline int dequeue_signal_lock(st spin_unlock_irqrestore(&tsk->sighand->siglock, flags); return ret; -} +} extern void block_all_signals(int (*notifier)(void *priv), void *priv, sigset_t *mask); @@ -2161,7 +2161,7 @@ extern int kill_pid_info_as_uid(int, str extern int kill_pgrp(struct pid *pid, int sig, int priv); extern int kill_pid(struct pid *pid, int sig, int priv); extern int kill_proc_info(int, struct siginfo *, pid_t); -extern int do_notify_parent(struct task_struct *, int); +extern bool do_notify_parent(struct task_struct *, int); extern void __wake_up_parent(struct task_struct *p, struct task_struct *parent); extern void force_sig(int, struct task_struct *); extern int send_sig(int, struct task_struct *, int); --- ptrace/kernel/signal.c~2_do_notify_parent_bool 2011-06-20 20:40:49.000000000 +0200 +++ ptrace/kernel/signal.c 2011-06-21 20:29:57.000000000 +0200 @@ -1577,15 +1577,15 @@ ret: * Let a parent know about the death of a child. * For a stopped/continued status change, use do_notify_parent_cldstop instead. * - * Returns -1 if our parent ignored us and so we've switched to - * self-reaping, or else @sig. + * Returns true if our parent ignored us and so we've switched to + * self-reaping. */ -int do_notify_parent(struct task_struct *tsk, int sig) +bool do_notify_parent(struct task_struct *tsk, int sig) { struct siginfo info; unsigned long flags; struct sighand_struct *psig; - int ret = sig; + bool autoreap = false; BUG_ON(sig == -1); @@ -1649,16 +1649,17 @@ int do_notify_parent(struct task_struct * is implementation-defined: we do (if you don't want * it, just use SIG_IGN instead). */ - ret = tsk->exit_signal = -1; + autoreap = true; + tsk->exit_signal = -1; if (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN) - sig = -1; + sig = 0; } - if (valid_signal(sig) && sig > 0) + if (valid_signal(sig) && sig) __group_send_sig_info(sig, &info, tsk->parent); __wake_up_parent(tsk, tsk->parent); spin_unlock_irqrestore(&psig->siglock, flags); - return ret; + return autoreap; } /** --- ptrace/kernel/exit.c~2_do_notify_parent_bool 2011-06-21 18:36:24.000000000 +0200 +++ ptrace/kernel/exit.c 2011-06-21 20:37:24.000000000 +0200 @@ -821,6 +821,7 @@ static void forget_original_parent(struc static void exit_notify(struct task_struct *tsk, int group_dead) { int signal; + bool autoreap; void *cookie; /* @@ -859,9 +860,11 @@ static void exit_notify(struct task_stru signal = tracehook_notify_death(tsk, &cookie, group_dead); if (signal >= 0) - signal = do_notify_parent(tsk, signal); + autoreap = do_notify_parent(tsk, signal); + else + autoreap = (signal == DEATH_REAP); - tsk->exit_state = signal == DEATH_REAP ? EXIT_DEAD : EXIT_ZOMBIE; + tsk->exit_state = autoreap ? EXIT_DEAD : EXIT_ZOMBIE; /* mt-exec, de_thread() is waiting for group leader */ if (unlikely(tsk->signal->notify_count < 0)) @@ -869,7 +872,7 @@ static void exit_notify(struct task_stru write_unlock_irq(&tasklist_lock); /* If the process is dead, release it - nobody will wait for it */ - if (signal == DEATH_REAP) + if (autoreap) release_task(tsk); } -- 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/