linux-kernel - Re: [PATCH 0/3] v4 Improve task->comm locking situation

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-Id: <201105180527.p4I5RHGl032657@www262.sakura.ne.jp>
Date:	Wed, 18 May 2011 14:27:17 +0900
From:	Tetsuo Handa <penguin-kernel@...ove.SAKURA.ne.jp>
To:	john.stultz@...aro.org
Cc:	tytso@....edu, mina86@...a86.com, jirislaby@...il.com,
	kosaki.motohiro@...fujitsu.com, rientjes@...gle.com,
	dave@...ux.vnet.ibm.com, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 0/3] v4 Improve task->comm locking situation

John Stultz wrote:
> So I think this approach is definitely worth trying. However, I think
> converting to RCU will be much easier once we've first converted all
> current comm users to making use of the get_task_comm accessor
> functions. In this way my hope is my current proposal will serve as a
> cleanup step before further optimizations can be done.

OK. But I guess converting all current comm users to making use of the
get_task_comm accessor is not easy because using get_task_comm() changes
locking dependency.

I think replacing

  "%s", tsk->comm

with

  "%ptc", tsk

(without using get_task_comm() inside pointer() because introducing a new lock
 might cause problem since callers can be inside a delicate condition based on
 the assumption that vsnprintf() will not block) should come first in order to
minimize direct ->comm users who may not be ready for this change.

After that, trying to replace

  task->comm

with

  char buf[TASK_COMM_LEN];
  get_task_comm(task, buf)

(with cautions for already held locks) may be considered.



By the way,

  sprintf(current->comm, "drbd%d_receiver", minor);

in drivers/block/drbd/drbd_receiver.c may be off-by-one
because DRBD_MINOR_COUNT_MAX is 256.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/