From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pavel Tatashin Subject: Re: [net-next 3/6] ixgbe: release lock for the duration of ixgbe_suspend_close() Date: Fri, 18 May 2018 07:37:32 -0400 Message-ID: <9ee4abce-0841-374d-5395-0c5170880eb1@oracle.com> References: <20180517163732.30910-1-jeffrey.t.kirsher@intel.com> <20180517163732.30910-4-jeffrey.t.kirsher@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Cc: netdev@vger.kernel.org, nhorman@redhat.com, sassmann@redhat.com, jogreene@redhat.com To: Sergei Shtylyov , Jeff Kirsher , davem@davemloft.net Return-path: Received: from aserp2120.oracle.com ([141.146.126.78]:49206 "EHLO aserp2120.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750888AbeERLhy (ORCPT ); Fri, 18 May 2018 07:37:54 -0400 In-Reply-To: Content-Language: en-US Sender: netdev-owner@vger.kernel.org List-ID:       * parallelized this function, so drop lock for the > >     Parallelizing? Else the sentence doesn't parse for me. :-) Hi Sergei, In a separate series I parallelized device_shutdown(), see: http://lkml.kernel.org/r/20180516024004.28977-1-pasha.tatashin@oracle.com But, this particular patch should be dropped, as discussed in this thread: http://lkml.kernel.org/r/20180503035931.22439-2-pasha.tatashin@oracle.com Alexander Duyck, made a point that a generic RTNL scalability fix should be done. This particular patch might introduce a race, since it relies on assumption that RTNL is not needed in this place because ixgbe_close() does not have it, but Alexander Duyck, says that the callers of ixgbe_close() are assumed to own this lock. Thank you, Pavel