From: Dante Cinco <dantecinco@gmail.com>
To: Xen-devel <xen-devel@lists.xensource.com>
Subject: swiotlb=force in Konrad's xen-pcifront-0.8.2 pvops domU kernel with PCI passthrough
Date: Wed, 10 Nov 2010 17:16:14 -0800 [thread overview]
Message-ID: <AANLkTi=q_ZgxEm9gYB6LvWOcJNnozW3cYqaySowTnD2y@mail.gmail.com> (raw)
We have Fibre Channel HBA devices that we PCI passthrough to our pvops
domU kernel. Without swiotlb=force in the domU's kernel command line,
both domU and dom0 lock up after loading the kernel module drivers for
the HBA devices. With swiotlb=force, the domU and dom0 are stable
after loading the kernel module drivers but the I/O performance is at
least an order of magnitude worse than what we were seeing with the
HVM kernel. I see the following in /var/log/kern.log in the pvops
domU:
PCI-DMA: Using software bounce buffering for IO (SWIOTLB)
Placing 64MB software IO TLB between ffff880005800000 - ffff880009800000
software IO TLB at phys 0x5800000 - 0x9800000
Is swiotlb=force responsible for the I/O performance degradation? I
don't understand what swiotlb=force does so I would appreciate an
explanation or a pointer.
Thanks.
- Dante
next reply other threads:[~2010-11-11 1:16 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-11-11 1:16 Dante Cinco [this message]
2010-11-11 16:04 ` swiotlb=force in Konrad's xen-pcifront-0.8.2 pvops domU kernel with PCI passthrough Konrad Rzeszutek Wilk
2010-11-11 18:31 ` Dante Cinco
2010-11-11 19:03 ` Konrad Rzeszutek Wilk
2010-11-11 19:42 ` Lin, Ray
2010-11-12 15:56 ` Konrad Rzeszutek Wilk
2010-11-12 16:20 ` Lin, Ray
2010-11-12 16:55 ` Konrad Rzeszutek Wilk
2010-11-12 19:38 ` Lin, Ray
2010-11-12 22:33 ` Konrad Rzeszutek Wilk
2010-11-12 22:57 ` Lin, Ray
2010-11-16 17:07 ` Dante Cinco
2010-11-16 18:57 ` Konrad Rzeszutek Wilk
2010-11-16 19:43 ` Dante Cinco
2010-11-16 20:15 ` Konrad Rzeszutek Wilk
2010-11-18 1:09 ` Dante Cinco
2010-11-18 17:19 ` Konrad Rzeszutek Wilk
2010-11-18 17:28 ` Chris Mason
2010-11-18 17:54 ` Mathieu Desnoyers
2010-11-18 18:43 ` Dante Cinco
2010-11-18 18:52 ` Lin, Ray
2010-11-18 19:35 ` Dante Cinco
2010-11-18 21:20 ` Dan Magenheimer
2010-11-18 21:39 ` Lin, Ray
2010-11-19 0:20 ` Dan Magenheimer
2010-11-19 1:38 ` Dante Cinco
2010-11-19 17:10 ` Jeremy Fitzhardinge
2010-11-19 17:52 ` Dante Cinco
2010-11-19 17:58 ` Keir Fraser
2010-11-19 22:36 ` Dan Magenheimer
2010-11-20 0:13 ` Dante Cinco
2010-11-19 17:55 ` Lin, Ray
2010-11-12 18:29 ` Dante Cinco
2010-11-11 22:32 ` Dante Cinco
2010-11-12 1:02 ` Dante Cinco
2010-11-12 16:58 ` Konrad Rzeszutek Wilk
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='AANLkTi=q_ZgxEm9gYB6LvWOcJNnozW3cYqaySowTnD2y@mail.gmail.com' \
--to=dantecinco@gmail.com \
--cc=xen-devel@lists.xensource.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.