Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1761672AbYGOQSm (ORCPT ); Tue, 15 Jul 2008 12:18:42 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752105AbYGOQSf (ORCPT ); Tue, 15 Jul 2008 12:18:35 -0400 Received: from x346.tv-sign.ru ([89.108.83.215]:49224 "EHLO mail.screens.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751315AbYGOQSe (ORCPT ); Tue, 15 Jul 2008 12:18:34 -0400 Date: Tue, 15 Jul 2008 20:21:50 +0400 From: Oleg Nesterov To: Andrew Morton Cc: Roland McGrath , linux-kernel@vger.kernel.org Subject: Re: [PATCH] coredump: kill mm->core_done Message-ID: <20080715162150.GA30927@tv-sign.ru> References: <20080715140715.GA25824@tv-sign.ru> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20080715140715.GA25824@tv-sign.ru> User-Agent: Mutt/1.5.11 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3506 Lines: 113 On 07/15, Oleg Nesterov wrote: > > + for (;;) { > + set_task_state(tsk, TASK_UNINTERRUPTIBLE); > + if (!self.task) /* see coredump_finish() */ > + break; > + schedule(); > + } > + __set_task_state(tsk, TASK_UNINTERRUPTIBLE); ^^^^^^^^^^^^^^^^^^^^ Ugh, sorry, this should be TASK_RUNNING. Please find the fixed patch below. [PATCH] coredump: kill mm->core_done Now that we have core_state->dumper list we can use it to wake up the sub-threads waiting for the coredump completion. This uglifies the code and .text grows by 47 bytes, but otoh mm_struct lessens by sizeof(struct completion). Also, with this change we can decouple exit_mm() from the coredumping code. Signed-off-by: Oleg Nesterov include/linux/mm_types.h | 4 +--- kernel/exit.c | 8 +++++++- fs/exec.c | 25 ++++++++++++++++++++++--- 3 files changed, 30 insertions(+), 7 deletions(-) --- 26-rc2/include/linux/mm_types.h~5_KILL_CORE_DONE 2008-07-13 18:28:36.000000000 +0400 +++ 26-rc2/include/linux/mm_types.h 2008-07-15 17:06:58.000000000 +0400 @@ -229,9 +229,7 @@ struct mm_struct { unsigned long flags; /* Must use atomic bitops to access the bits */ - /* coredumping support */ - struct core_state *core_state; - struct completion core_done; + struct core_state *core_state; /* coredumping support */ /* aio bits */ rwlock_t ioctx_list_lock; /* aio lock */ --- 26-rc2/kernel/exit.c~5_KILL_CORE_DONE 2008-07-13 19:58:19.000000000 +0400 +++ 26-rc2/kernel/exit.c 2008-07-15 20:17:28.000000000 +0400 @@ -680,7 +680,13 @@ static void exit_mm(struct task_struct * if (atomic_dec_and_test(&core_state->nr_threads)) complete(&core_state->startup); - wait_for_completion(&mm->core_done); + for (;;) { + set_task_state(tsk, TASK_UNINTERRUPTIBLE); + if (!self.task) /* see coredump_finish() */ + break; + schedule(); + } + __set_task_state(tsk, TASK_RUNNING); down_read(&mm->mmap_sem); } atomic_inc(&mm->mm_count); --- 26-rc2/fs/exec.c~5_KILL_CORE_DONE 2008-07-13 18:43:39.000000000 +0400 +++ 26-rc2/fs/exec.c 2008-07-15 20:16:36.000000000 +0400 @@ -1597,7 +1597,6 @@ static int coredump_wait(int exit_code, struct completion *vfork_done; int core_waiters; - init_completion(&mm->core_done); init_completion(&core_state->startup); core_state->dumper.task = tsk; core_state->dumper.next = NULL; @@ -1623,6 +1622,27 @@ fail: return core_waiters; } +static void coredump_finish(struct mm_struct *mm) +{ + struct core_thread *curr, *next; + struct task_struct *task; + + next = mm->core_state->dumper.next; + while ((curr = next) != NULL) { + next = curr->next; + task = curr->task; + /* + * see exit_mm(), curr->task must not see + * ->task == NULL before we read ->next. + */ + smp_mb(); + curr->task = NULL; + wake_up_process(task); + } + + mm->core_state = NULL; +} + /* * set_dumpable converts traditional three-value dumpable to two flags and * stores them into mm->flags. It modifies lower two bits of mm->flags, but @@ -1807,8 +1827,7 @@ fail_unlock: argv_free(helper_argv); current->fsuid = fsuid; - complete_all(&mm->core_done); - mm->core_state = NULL; + coredump_finish(mm); fail: return retval; } -- 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/