From mboxrd@z Thu Jan 1 00:00:00 1970 From: Patrick McHardy Subject: Re: ACPI/HT or Packet Scheduler BUG? Date: Sun, 17 Apr 2005 19:46:16 +0200 Message-ID: <4262A0E8.9020905@trash.net> References: <1113601029.4294.80.camel@localhost.localdomain> <1113601446.17859.36.camel@localhost.localdomain> <1113602052.4294.89.camel@localhost.localdomain> <20050415225422.GF4114@postel.suug.ch> <20050416014906.GA3291@gondor.apana.org.au> <20050416110639.GI4114@postel.suug.ch> <20050416111236.GA31550@gondor.apana.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Cc: Thomas Graf , Steven Rostedt , hadi@cyberus.ca, netdev , Tarhon-Onu Victor , kuznet@ms2.inr.ac.ru, devik@cdi.cz, linux-kernel@vger.kernel.org, "David S. Miller" Return-path: To: Herbert Xu In-Reply-To: <20050416111236.GA31550@gondor.apana.org.au> Sender: netdev-bounce@oss.sgi.com Errors-to: netdev-bounce@oss.sgi.com List-Id: netdev.vger.kernel.org Herbert Xu wrote: > On Sat, Apr 16, 2005 at 01:06:39PM +0200, Thomas Graf wrote: > >>qdisc_destroy can still be invoked without qdisc_tree_lock via the >>deletion of a class when it calls qdisc_destroy to destroy its >>leaf qdisc. > > Indeed. Fortuantely HTB seems to be safe as it calls sch_tree_lock > which is another name for qdisc_tree_lock. CBQ on the other hand > needs to have a little tweak. HTB also needs to be fixed. Destruction is usually defered by the refcnt until ->put(), htb_put() doesn't lock the tree. Same for HFSC and CBQ. Regards Patrick