From: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
To: "Cinco, Dante" <Dante.Cinco@lsi.com>
Cc: "xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>,
Keir Fraser <keir.fraser@eu.citrix.com>,
Qing He <qing.he@intel.com>,
"xiantao.zhang@intel.com" <xiantao.zhang@intel.com>
Subject: Re: IRQ SMP affinity problems in domU with vcpus > 4 on HP ProLiant G6 with dual Xeon 5540 (Nehalem)
Date: Thu, 15 Oct 2009 21:40:30 -0400 [thread overview]
Message-ID: <20091016014030.GA6450@phenom.dumpdata.com> (raw)
In-Reply-To: <20091016000942.GA24471@phenom.dumpdata.com>
On Thu, Oct 15, 2009 at 08:09:42PM -0400, Konrad Rzeszutek Wilk wrote:
> On Wed, Oct 14, 2009 at 01:54:33PM -0600, Cinco, Dante wrote:
> > I switched over to Xen 3.5-unstable (changeset 20303) and pv_ops dom0 2.6.31.1 hoping that this would resolve the IRQ SMP affinity problem. I had to use pci-stub to hide the PCI devices since pciback wasn't working. With vcpus=16 (APIC routing is physical flat), the interrupts were working in domU and being routed to CPU0 with the default smp_affinity (ffff) but as soon as I changed it to any 16-bit one-hot value or even setting it to the same default value resulted in a complete loss of interrupts (even in the devices that didn't have any change to smp_affinity). With vcpus=4 (APIC routing is logical flat), I can see the interrupts being load balanced across all CPUs but as soon as I changed smp_affinity to any value, the interrupts stopped. This used to work reliably with the non-pv
_ops kernel. I attached the logs in case anyone wants to take a look.
> >
> > I did see the MSI message address/data change in both domU and dom0 (using "lspci -vv"):
> >
> > vcpus=16:
> >
> > domU MSI message address/data with default smp_affinity: Address: 00000000fee00000 Data: 40a9
> > domU MSI message address/data after smp_affinity=0010: Address: 00000000fee08000 Data: 40b1 (8 is APIC ID of CPU4)
>
> What does Xen tell you (hit Ctrl-A three times and then 'z'). Specifically look for vector 169 (a9) and 177 (b1).
> Do those values match with what you see in DomU and Dom0? Mainly that 177 has dest_id of 8.
> Oh, and also check the guest interrupt information, to see if those values match..
N/m. I was thinking that maybe your IOAPIC has those vectors programmed in it. But
that would not make any sense.
> >
> > dom0 MSI message address/data with default smp_affinity: Address: 00000000fee00000 Data: 4094
> > dom0 MSI message address/data after smp_affinity=0010: Address: 00000000fee00000 Data: 409c
next prev parent reply other threads:[~2009-10-16 1:40 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-08 0:08 IRQ SMP affinity problems in domU with vcpus > 4 on HP ProLiant G6 with dual Xeon 5540 (Nehalem) Cinco, Dante
2009-10-08 16:07 ` Bruce Edge
2009-10-08 18:05 ` Keir Fraser
2009-10-08 18:11 ` Cinco, Dante
2009-10-08 21:35 ` Keir Fraser
2009-10-09 9:07 ` Qing He
2009-10-09 15:59 ` Cinco, Dante
2009-10-09 23:39 ` Cinco, Dante
2009-10-10 9:43 ` Qing He
2009-10-10 10:10 ` Keir Fraser
2009-10-12 5:25 ` Cinco, Dante
2009-10-12 5:54 ` Qing He
2009-10-14 19:54 ` Cinco, Dante
2009-10-16 0:09 ` Konrad Rzeszutek Wilk
2009-10-16 1:40 ` Konrad Rzeszutek Wilk [this message]
-- strict thread matches above, loose matches on Subject: below --
2009-10-16 1:38 Cinco, Dante
2009-10-16 2:34 ` Qing He
2009-10-16 6:37 ` Keir Fraser
2009-10-16 7:32 ` Zhang, Xiantao
2009-10-16 8:24 ` Qing He
2009-10-16 8:22 ` Zhang, Xiantao
2009-10-16 8:34 ` Qing He
2009-10-16 8:35 ` Zhang, Xiantao
2009-10-16 9:01 ` Qing He
2009-10-16 9:42 ` Qing He
2009-10-16 9:49 ` Zhang, Xiantao
2009-10-16 14:54 ` Zhang, Xiantao
2009-10-16 18:24 ` Cinco, Dante
2009-10-17 0:59 ` Zhang, Xiantao
2009-10-20 0:19 ` Cinco, Dante
2009-10-20 5:46 ` Zhang, Xiantao
2009-10-20 7:51 ` Zhang, Xiantao
2009-10-20 17:26 ` Cinco, Dante
2009-10-21 1:10 ` Zhang, Xiantao
2009-10-22 1:00 ` Cinco, Dante
2009-10-22 1:58 ` Zhang, Xiantao
2009-10-22 2:42 ` Zhang, Xiantao
2009-10-22 6:25 ` Keir Fraser
2009-10-22 21:11 ` Jeremy Fitzhardinge
2009-10-22 5:10 ` Qing He
2009-10-23 0:10 ` Cinco, Dante
2009-10-22 6:46 ` Jan Beulich
2009-10-22 7:11 ` Zhang, Xiantao
2009-10-22 7:31 ` Jan Beulich
2009-10-22 8:41 ` Zhang, Xiantao
2009-10-22 9:42 ` Keir Fraser
2009-10-22 16:32 ` Zhang, Xiantao
2009-10-22 16:33 ` Cinco, Dante
2009-10-23 1:06 ` Zhang, Xiantao
2009-10-26 13:02 ` Zhang, Xiantao
2009-10-26 13:34 ` Keir Fraser
2009-10-16 9:41 ` Keir Fraser
2009-10-16 9:57 ` Qing He
2009-10-16 9:58 ` Zhang, Xiantao
2009-10-16 10:21 ` Jan Beulich
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20091016014030.GA6450@phenom.dumpdata.com \
--to=konrad.wilk@oracle.com \
--cc=Dante.Cinco@lsi.com \
--cc=keir.fraser@eu.citrix.com \
--cc=qing.he@intel.com \
--cc=xen-devel@lists.xensource.com \
--cc=xiantao.zhang@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.