From: Oleg Nesterov <oleg@tv-sign.ru>
To: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Andrew Morton <akpm@osdl.org>, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 01/23] tref: Implement task references.
Date: Fri, 03 Mar 2006 22:23:04 +0300 [thread overview]
Message-ID: <44089798.F004D18B@tv-sign.ru> (raw)
In-Reply-To: m1mzgidnr0.fsf@ebiederm.dsl.xmission.com
Eric W. Biederman wrote:
>
> +++ devel-akpm/kernel/task_ref.c 2006-02-27 20:28:59.000000000 -0800
> @@ -0,0 +1,131 @@
> +#include <linux/sched.h>
> +#include <linux/task_ref.h>
> +
> +struct task_ref init_tref = {
> + .count = ATOMIC_INIT(1),
> + .type = PIDTYPE_PID,
> + .pid = 0,
> + .task = NULL,
> +};
Make it static? Actually, I don't understand why init_tref is better
than NULL. Yes, NULL will add some checks into task_ref.c, but we can
avoid some costly atoimic ops.
> +void tref_put(struct task_ref *ref)
> +{
> + might_sleep();
> + if (atomic_dec_and_test(&ref->count)) {
> + struct task_struct *task;
> + BUG_ON(ref == &init_tref);
> + /* Carefully serialize against __detach_pid and tref_get_by_pid */
> + write_lock_irq(&tasklist_lock);
> + task = ref->task;
> + if (task)
> + task->pids[ref->type].ref = NULL;
> + write_unlock_irq(&tasklist_lock);
> + kfree(ref);
> + }
I think this is racy. Suppose ref->count == 1. What if another cpu does
tref_get_by_task() between atomic_dec_and_test() and write_lock_irq() ?
It takes tasklist_lock, increments ->count again, and returns the pointer
to the memory which will be freed soon.
> +struct task_ref *tref_get_by_pid(int pid, enum pid_type type)
> +{
> + struct task_struct *task;
> + struct task_ref *tref;
> +
> + /* Lookup the and pin the task */
> + read_lock(&tasklist_lock);
> + task = find_task_by_pid_type(type, pid);
> + if (task)
> + get_task_struct(task);
> + read_unlock(&tasklist_lock);
> +
> + /* Now get the tref */
> + if (task) {
> + tref = tref_get_by_task(task, type);
> + put_task_struct(task);
> + }
> + else
> + tref = tref_get(&init_tref);
> + return tref;
> +}
I beleive this could be simplified, we don't need to get/put task_struct,
rcu_read_lock();
task = find_task_by_pid_type(type, pid);
if (task)
tref = tref_get_by_task(task, type);
else
tref = tref_get(&init_tref);
rcu_read_unlock();
Oleg.
next prev parent reply other threads:[~2006-03-03 19:26 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-02-23 15:52 [PATCH 00/23] proc cleanup Eric W. Biederman
2006-02-23 15:54 ` [PATCH 01/23] tref: Implement task references Eric W. Biederman
2006-02-23 15:56 ` [PATCH 02/23] proc: Fix the .. inode number on /proc/<pid>/fd Eric W. Biederman
2006-02-23 15:57 ` [PATCH 03/23] proc: Remove useless BKL in proc_pid_readlink Eric W. Biederman
2006-02-23 15:58 ` [PATCH 04/23] proc: Remove unnecessary and misleading assignments from proc_pid_make_inode Eric W. Biederman
2006-02-23 16:00 ` [PATCH 05/23] proc: Simplify the ownership rules for /proc Eric W. Biederman
2006-02-23 16:02 ` Eric W. Biederman
2006-02-23 16:04 ` [PATCH 06/23] proc: Replace proc_inode.type with proc_inode.fd Eric W. Biederman
2006-02-23 16:05 ` [PATCH 07/23] proc: Remove bogus proc_task_permission Eric W. Biederman
2006-02-23 16:06 ` [PATCH 08/23] proc: Kill proc_mem_inode_operations Eric W. Biederman
2006-02-23 16:08 ` [PATCH 09/23] proc: Properly filter out files that are not visible to a process Eric W. Biederman
2006-02-23 16:10 ` [PATCH 10/23] proc: Fix the link count for /proc/<pid>/task Eric W. Biederman
2006-02-23 16:12 ` [PATCH 11/23] proc: Move proc_maps_operations into task_mmu.c Eric W. Biederman
2006-02-23 16:15 ` [PATCH 12/23] proc: Rewrite the proc dentry flush on exit optimization Eric W. Biederman
2006-02-23 16:16 ` [PATCH 13/23] proc: Close the race of a process dying durning lookup Eric W. Biederman
2006-02-23 16:18 ` [PATCH 14/23] proc: Make PROC_NUMBUF the buffer size for holding a integers as strings Eric W. Biederman
2006-02-23 16:20 ` [PATCH 15/23] proc: refactor reading directories of tasks Eric W. Biederman
2006-02-23 16:23 ` [PATCH 16/23] proc: Don't lock task_structs indefinitely Eric W. Biederman
2006-02-23 16:24 ` [PATCH 17/23] proc: Give the root directory a task Eric W. Biederman
2006-02-23 16:25 ` [PATCH 18/23] proc: Reorder the functions in base.c Eric W. Biederman
2006-02-23 16:27 ` [PATCH 19/23] proc: Modify proc_pident_lookup to be completely table driven Eric W. Biederman
2006-02-23 16:28 ` [PATCH 20/23] proc: Make the generation of the self symlink " Eric W. Biederman
2006-02-23 16:30 ` [PATCH 21/23] proc: Factor out an instantiate method from every lookup method Eric W. Biederman
2006-02-23 16:32 ` [PATCH 22/23] proc: Remove the hard coded inode numbers Eric W. Biederman
2006-02-23 16:34 ` [PATCH 23/23] proc: Merge proc_tid_attr and proc_tgid_attr Eric W. Biederman
2006-02-23 16:49 ` [PATCH 01/23] tref: Implement task references Eric W. Biederman
2006-03-02 19:16 ` Oleg Nesterov
2006-03-02 20:37 ` Oleg Nesterov
2006-03-02 22:19 ` Eric W. Biederman
2006-03-03 16:56 ` Oleg Nesterov
2006-03-03 17:48 ` Eric W. Biederman
2006-03-04 11:16 ` Eric W. Biederman
2006-03-04 12:31 ` Oleg Nesterov
2006-03-04 17:30 ` Oleg Nesterov
2006-03-06 21:06 ` Oleg Nesterov
2006-03-06 22:18 ` Eric W. Biederman
2006-03-07 20:44 ` Oleg Nesterov
2006-03-07 1:39 ` Eric W. Biederman
2006-03-07 20:38 ` Oleg Nesterov
2006-03-07 13:12 ` Eric W. Biederman
2006-03-07 21:02 ` Oleg Nesterov
2006-03-07 23:00 ` Eric W. Biederman
2006-03-03 19:23 ` Oleg Nesterov [this message]
2006-03-04 10:51 ` Eric W. Biederman
2006-02-25 12:27 ` [PATCH 00/23] proc cleanup Andrew Morton
2006-02-25 13:34 ` Eric W. Biederman
2006-02-25 15:20 ` Eric W. Biederman
2006-02-27 15:26 ` Serge E. Hallyn
2006-02-27 15:56 ` Eric W. Biederman
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=44089798.F004D18B@tv-sign.ru \
--to=oleg@tv-sign.ru \
--cc=akpm@osdl.org \
--cc=ebiederm@xmission.com \
--cc=linux-kernel@vger.kernel.org \
/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).