From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Ayyappan Veeraiyan" Subject: Re: [ANNOUNCE] new driver ixgbe for Intel(R) 10GbE PCI Express adapters. Date: Wed, 13 Jun 2007 17:06:57 -0700 Message-ID: References: <20070612234417.5102.29147.stgit@localhost.localdomain> <20070613230513.GB22521@electric-eye.fr.zoreil.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, jeff@garzik.org, auke-jan.h.kok@intel.com, arjan@linux.intel.com, akpm@linux-foundation.org To: "Francois Romieu" Return-path: Received: from ag-out-0708.google.com ([72.14.246.243]:38784 "EHLO ag-out-0708.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755070AbXFNAG5 (ORCPT ); Wed, 13 Jun 2007 20:06:57 -0400 Received: by ag-out-0708.google.com with SMTP id 35so928aga for ; Wed, 13 Jun 2007 17:06:57 -0700 (PDT) In-Reply-To: <20070613230513.GB22521@electric-eye.fr.zoreil.com> Content-Disposition: inline Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On 6/13/07, Francois Romieu wrote: Please ignore my previous response. Forgot to reply_all.. > > o ixgbe_clean_tx_irq() is issued from the NAPI poll handler (as the > name suggests...). Its locking will probably cause a deadlock if > it is interrupted and a call to netpoll_send_skb follows. > netpoll_send_skb should not deadlock because ixgbe_xmit_frame should bail out because of this... if (!spin_trylock_irqsave(&tx_ring->tx_lock, flags)) /* Collision - tell upper layer to requeue */ return NETDEV_TX_LOCKED; Right? The LLTX and tx_lock handling in the driver is not clean and I am working on to fix it. Also will make sure we are handling the net_poll case correctly... Thanks for the feedback.. Ayyappan