From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:56082) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dAIe7-0007c8-M6 for qemu-devel@nongnu.org; Mon, 15 May 2017 12:14:57 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dAIe2-0003jB-Jo for qemu-devel@nongnu.org; Mon, 15 May 2017 12:14:55 -0400 Received: from mx1.redhat.com ([209.132.183.28]:48902) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1dAIe2-0003j6-AG for qemu-devel@nongnu.org; Mon, 15 May 2017 12:14:50 -0400 Date: Mon, 15 May 2017 17:14:46 +0100 From: "Dr. David Alan Gilbert" Message-ID: <20170515161445.GA2324@work-vm> References: <20170511123639.GE2078@work-vm> <225AC0F4-5672-456F-9032-F2C1C86342E5@daynix.com> <20170515092251.GB2089@work-vm> <6F78534C-7A67-43F4-A9E8-E532BCD16976@daynix.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <6F78534C-7A67-43F4-A9E8-E532BCD16976@daynix.com> Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] e1000e migration List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Dmitry Fleytman Cc: qemu-devel@nongnu.org * Dmitry Fleytman (dmitry@daynix.com) wrote: > Hello Dave, >=20 > It looks like we identified the problem. >=20 > We are working on fix and will send it as soon as it is ready. Thanks! Dave > ~Dmitry. >=20 > Sent from my iPhone >=20 > > On 15 May 2017, at 12:22, Dr. David Alan Gilbert wrote: > >=20 > > * Dmitry Fleytman (dmitry@daynix.com) wrote: > >> Hello Dave, > >=20 > > Hi Dmitry, > > Thanks for the reply. > >=20 > >> We are trying to reproduce this issue on our systems but with no luc= k so far=E2=80=A6 > >=20 > > Note our QE hit this with both a Win8.1 and a win2012r2 guest - altho= ugh > > the 2012r2 is reported to have recoverd after a few minutes. > > 2016 apparently works OK. > >=20 > >> From what you describe it looks like some bit in ICR is not being cl= eared by the driver. > >> This usually means that this bit should never be set in that specifi= c interrupt mode. > >>=20 > >> Could you please check which bit is not cleared and who sets it? > >=20 > > The full set of e1000e_irq_pending_interrupts after migration is: > > 23004@1494519346.673905:e1000e_irq_pending_interrupts ICR PENDING: 0x= 100000 (ICR: 0x80100082, IMS: 0x1f00004) > > 23004@1494519346.674787:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x80100082, IMS: 0x1e00004) > > 23004@1494519346.674946:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x80100082, IMS: 0x1e00004) > > 23004@1494519346.675119:e1000e_irq_pending_interrupts ICR PENDING: 0x= 200000 (ICR: 0x80300082, IMS: 0x1e00004) > > 23004@1494519346.675302:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x80100082, IMS: 0x1c00004) > > > > 23004@1494519346.716279:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x80300082, IMS: 0x1c00004) > > 23004@1494519346.716380:e1000e_irq_pending_interrupts ICR PENDING: 0x= 1000000 (ICR: 0x813000c2, IMS: 0x1c00004) > > 23004@1494519346.717040:e1000e_irq_pending_interrupts ICR PENDING: 0x= 1000000 (ICR: 0x813000c2, IMS: 0x1400004) > > 23004@1494519346.717276:e1000e_irq_pending_interrupts ICR PENDING: 0x= 1000000 (ICR: 0x813000c2, IMS: 0x1000004) > > 23004@1494519346.717443:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x813000c2, IMS: 0x4) > > 23004@1494519346.717567:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x813000c2, IMS: 0x4) > > 23004@1494519346.717782:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x813000c2, IMS: 0x4) > > 23004@1494519346.717918:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x813000c2, IMS: 0x4) > > 23004@1494519346.718319:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x813000c2, IMS: 0x4) > > 23004@1494519346.718523:e1000e_irq_pending_interrupts ICR PENDING: 0x= 200000 (ICR: 0x813000c2, IMS: 0xa00004) > > 23004@1494519346.718684:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0x4) > > 23004@1494519346.718890:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0x4) > > 23004@1494519346.719034:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0xa00004) > > 23004@1494519346.719130:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0xa00004) > > > > 23004@1494519346.722699:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0xa00004) > > 23004@1494519346.722868:e1000e_irq_pending_interrupts ICR PENDING: 0x= 200000 (ICR: 0x813000c2, IMS: 0xa00004) > > 23004@1494519346.723068:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0x800004) > > > > 23004@1494519346.731198:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x813000c2, IMS: 0x800004) > > 23004@1494519346.731422:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x813000c2, IMS: 0x4) > > 23004@1494519346.731930:e1000e_irq_pending_interrupts ICR PENDING: 0x= 200000 (ICR: 0x813000c2, IMS: 0xa00004) > > 23004@1494519346.732082:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0x4) > > 23004@1494519346.732274:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0x4) > > 23004@1494519346.732404:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0xa00004) > > 23004@1494519346.732504:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x811000c2, IMS: 0xa00004) > > 23004@1494519346.784150:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x815000c2, IMS: 0xa00004) > > 23004@1494519346.786506:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x815000c2, IMS: 0xa00004) > > 23004@1494519346.786534:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x815000c2, IMS: 0xa00004) > > 23004@1494519346.789644:e1000e_irq_pending_interrupts ICR PENDING: 0x= 1000000 (ICR: 0x815000c2, IMS: 0x1a00004) > > 23004@1494519346.789864:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x815000c2, IMS: 0xa00004) > > 23004@1494519346.789992:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x815000c2, IMS: 0xa00004) > > 23004@1494519346.790413:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x815000c2, IMS: 0xa00004) > > 23004@1494519346.790539:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x815000c2, IMS: 0xa00004) > > 23004@1494519346.792593:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x815000c2, IMS: 0xa00004) > > 23004@1494519346.792620:e1000e_irq_pending_interrupts ICR PENDING: 0x= 0 (ICR: 0x815000c2, IMS: 0xa00004) > > 23004@1494519346.795943:e1000e_irq_pending_interrupts ICR PENDING: 0x= 1000000 (ICR: 0x815000c2, IMS: 0x1a00004) > >=20 > > and then I think we get stuck in this cycle of this one always being = the > > one that fires repeatedly. I think that's the 'other' firing, I thin= k > > because of the receive-overrun. One thing I've not > > figured out is why the receive overrun happens - is that because we > > really have a very heavy packet rate or is it because something has > > stopped receiving them. > > The network I'm testing on does have a fair amount of broadcast traff= ic > > on. > >=20 > > Dave > >=20 > >> Regards, > >> Dmitry > >>=20 > >>> On 11 May 2017, at 15:36 PM, Dr. David Alan Gilbert wrote: > >>>=20 > >>> Hi Dmitry, > >>> Have you seen any problems with e1000e migration under windows? > >>> I've got a repeatable case where after migration with e1000e window= s > >>> hangs/almost hangs. > >>> I'm seeing the e1000e generate interrupts at a very very high > >>> rate (maybe ~1000 second ish?) after migration. > >>>=20 > >>> Some versions of qemu do it and some dont, but my attempts > >>> at bisection lead me to code that should be irrelevant. > >>>=20 > >>> Prior to migration I see: > >>>=20 > >>> 36461@1494504466.711929:e1000e_irq_pending_interrupts ICR PENDING: = 0x100000 (ICR: 0x80100082, IMS: 0x1f00004) > >>> 36461@1494504466.711992:e1000e_irq_pending_interrupts ICR PENDING: = 0x0 (ICR: 0x80000082, IMS: 0x1a00004) > >>> 36461@1494504466.712076:e1000e_irq_pending_interrupts ICR PENDING: = 0x0 (ICR: 0x80000082, IMS: 0x1f00004) > >>> 36461@1494504466.712245:e1000e_irq_pending_interrupts ICR PENDING: = 0x0 (ICR: 0x80000082, IMS: 0x1a00004) > >>> 36461@1494504466.712332:e1000e_irq_pending_interrupts ICR PENDING: = 0x0 (ICR: 0x80000082, IMS: 0x1f00004) > >>>=20 > >>> which I think the ICR means: > >>> 31 - int asserted > >>> 20 - RxQ0 - receive queue 0 interrupt > >>> 7 - RXT0 - receiver timer interrupt > >>> 1 - TXQE - Transmit Queue empty > >>>=20 > >>> after migration it varies more, I'm seeing mostly: > >>> 21977@1494504516.320707:e1000e_irq_pending_interrupts ICR PENDING: = 0x1000000 (ICR: 0x815000c2, IMS: 0x1a00004) > >>> 31 - int asserted > >>> 24 - 'Other' > >>> 22 - TxQ0 interrupt > >>> 20 - RxQ0 interrupt > >>> 07 - RXT0 Receiver timer interrupt > >>> 06 - RX0 - Receiver overrun > >>> 01 - TXQE - Transmit queue empty > >>>=20 > >>> For reference this is https://bugzilla.redhat.com/show_bug.cgi?id=3D= 1447935 > >>>=20 > >>> Dave > >>> -- > >>> Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK > >>=20 > > -- > > Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK