From: Alex Williamson <alex.williamson@redhat.com>
To: Shashidhar Patil <shashidhar.patil@gmail.com>
Cc: kvm@vger.kernel.org, Jan Kiszka <jan.kiszka@web.de>
Subject: Re: [PATCH 0/9] pci-assign: 64bit MMIO + better MSI-X table support
Date: Thu, 09 Feb 2012 10:23:06 -0700 [thread overview]
Message-ID: <1328808186.2954.28.camel@ul30vt> (raw)
In-Reply-To: <CADve3d5WueXs4SC60stJL985D8a9YqX0Wz=3G47Tgvvanx6z_w@mail.gmail.com>
On Thu, 2012-02-09 at 21:53 +0530, Shashidhar Patil wrote:
> Hi Alex,
> I tested your code changes for two problems
> 1. MSIX vectors allocation for FreeBSD Guest
> Tested the allocation of MSI-X allocation happening for both
> FreeBSD-9.0 (also tested 8.2) and Linux guest.
> Also tried multiple restart of the guests to see if the hosts
> still maintains those allocations. With Linux
> as guest I could pass traffic. Though interrupt allocation worked
> the FreeBSD could to pass traffic. The FreeBSD
> driver seems to be broken(atleast when running in guest).
Which device/driver? I was able to use an 82576, both physical device
and sr-iov virtual function) with the freebsd 9.0 igb driver.
> 2. Guest MSI-X smp_affinity fix.
> The smp_affinity configured in guest is honoured. The traffic
> distribution is uniform. When smp affinity is configured
> in host and guest properly this fix provides multi queue and
> hence multi core traffic distribution in both host and
> guest. The result is good packet throughput.
>
> Thanks for the the much needed fixes.
Thanks for testing!
Alex
> -Shashidhar
>
> On Mon, Feb 6, 2012 at 10:59 PM, Alex Williamson
> <alex.williamson@redhat.com> wrote:
> > On Mon, 2012-02-06 at 21:25 +0530, Shashidhar Patil wrote:
> >> HI Alex,
> >> I can give it a try. Please send me all patches as one file if possible.
> >> I was about to report yet another problem with Guest MSI-X smp affinity
> >> not being honoured by KVM while injecting interrupts. Its again in the context
> >> of 82599 device (tried with Linux as guest). I had a hunch that the MSI-X vector
> >> data which gets modified when SMP affinity is not tracked by the qemu_kvm.
> >> So the interrupt injection was done in load balancing fashion. Because of this
> >> interrupt LB in guest VCPUs I saw that packets for all the queues were processed
> >> by one VCPU instead 2 or 4 available.
> >> Anyway I will test the the MSI-X affinity fix also and get back with results.
> >> And I hope 82599 send/receive works. Thanks for your help.
> >
> > I pushed the v2 version to github, so you can grab it from here for
> > testing:
> >
> > git://github.com/awilliam/qemu-kvm.git
> >
> > Branch pci-assign-msix. Thanks,
> >
> > Alex
> >
next prev parent reply other threads:[~2012-02-09 17:23 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-01-28 14:21 [PATCH 0/9] pci-assign: 64bit MMIO + better MSI-X table support Alex Williamson
2012-01-28 14:21 ` [PATCH 1/9] pci-assign: Optionally enable 64bit BARs in guest Alex Williamson
2012-01-31 12:40 ` Avi Kivity
2012-01-31 12:45 ` Jan Kiszka
2012-01-31 12:51 ` Avi Kivity
2012-01-31 12:57 ` Jan Kiszka
2012-01-31 13:10 ` Avi Kivity
2012-01-31 13:21 ` Jan Kiszka
2012-01-31 13:33 ` Avi Kivity
2012-01-31 21:08 ` Alex Williamson
2012-01-31 21:14 ` Michael S. Tsirkin
2012-02-01 9:03 ` Avi Kivity
2012-02-01 10:03 ` Michael S. Tsirkin
2012-02-01 13:55 ` Alex Williamson
2012-02-01 15:18 ` Michael S. Tsirkin
2012-02-01 15:24 ` Alex Williamson
2012-01-28 14:21 ` [PATCH 2/9] pci-assign: Fix warnings with DEBUG enabled Alex Williamson
2012-01-28 14:21 ` [PATCH 3/9] pci-assign: Update MSI-X MMIO to Memory API Alex Williamson
2012-01-31 12:45 ` Avi Kivity
2012-01-31 21:13 ` Alex Williamson
2012-02-01 4:22 ` Alex Williamson
2012-02-01 9:04 ` Avi Kivity
2012-02-01 13:56 ` Alex Williamson
2012-01-28 14:21 ` [PATCH 4/9] pci-assign: Use struct for MSI-X table Alex Williamson
2012-01-31 17:40 ` Michael S. Tsirkin
2012-01-31 19:05 ` Alex Williamson
2012-01-31 20:00 ` Michael S. Tsirkin
2012-01-31 21:17 ` Alex Williamson
2012-01-31 21:24 ` Michael S. Tsirkin
2012-01-31 21:30 ` Alex Williamson
2012-01-28 14:22 ` [PATCH 5/9] pci-assign: Only calculate maximum MSI-X vector entries once Alex Williamson
2012-01-31 20:18 ` Michael S. Tsirkin
2012-01-31 20:31 ` Alex Williamson
2012-01-31 20:56 ` Michael S. Tsirkin
2012-01-28 14:22 ` [PATCH 6/9] pci-assign: Proper initialization for MSI-X table Alex Williamson
2012-01-31 17:40 ` Michael S. Tsirkin
2012-01-31 19:07 ` Alex Williamson
2012-01-31 19:12 ` Michael S. Tsirkin
2012-01-31 19:16 ` Jan Kiszka
2012-01-31 20:19 ` Michael S. Tsirkin
2012-01-31 21:06 ` Alex Williamson
2012-01-28 14:22 ` [PATCH 7/9] pci-assign: Allocate entries for all MSI-X vectors Alex Williamson
2012-01-28 14:22 ` [PATCH 8/9] pci-assign: Use MSIX_PAGE_SIZE Alex Williamson
2012-01-28 14:22 ` [PATCH 9/9] pci-assign: Update MSI-X config based on table writes Alex Williamson
2012-01-31 12:50 ` Avi Kivity
2012-01-30 10:11 ` [PATCH 0/9] pci-assign: 64bit MMIO + better MSI-X table support Jan Kiszka
2012-01-30 13:44 ` Alex Williamson
2012-01-31 12:52 ` Avi Kivity
2012-01-31 12:56 ` Jan Kiszka
2012-02-06 15:55 ` Shashidhar Patil
2012-02-06 17:29 ` Alex Williamson
2012-02-09 16:23 ` Shashidhar Patil
2012-02-09 17:23 ` Alex Williamson [this message]
2012-02-12 16:30 ` Shashidhar Patil
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=1328808186.2954.28.camel@ul30vt \
--to=alex.williamson@redhat.com \
--cc=jan.kiszka@web.de \
--cc=kvm@vger.kernel.org \
--cc=shashidhar.patil@gmail.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).