xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Wei Wang <wei.wang2@amd.com>
To: Sander Eikelenboom <linux@eikelenboom.it>
Cc: Keir Fraser <keir.xen@gmail.com>,
	Santosh Jodh <Santosh.Jodh@citrix.com>,
	"xen-devel@lists.xen.org" <xen-devel@lists.xen.org>
Subject: Re: Using debug-key 'o:  Dump IOMMU p2m table, locks up machine
Date: Mon, 3 Sep 2012 17:20:55 +0200	[thread overview]
Message-ID: <5044CAD7.8030800@amd.com> (raw)
In-Reply-To: <217459398.20120902171441@eikelenboom.it>

On 09/02/2012 05:14 PM, Sander Eikelenboom wrote:
> Sunday, September 2, 2012, 4:58:58 PM, you wrote:
>
>> On 02/09/2012 09:43, "Sander Eikelenboom"<linux@eikelenboom.it>  wrote:
>
>>>> Quite simply, there likely needs to be more tracing on the IOMMU fault path.
>>>> That's a separate concern from your keyhandler of course, but just saying
>>>> I'd be looking for the former rather than the latter, for diagnosing
>>>> Sander's bug.
>>>
>>> Are there any printk's I could add to get more relevant info about the AMD-Vi:
>>> IO_PAGE_FAULT ?
>
>> No really straightforward one. I think we need a per-IOMMU-type handler to
>> walk the IOMMU page table for a given virtual address, and dump every
>> page-table-entry on the path. Like an IOMMU version of show_page_walk().
>> Personally I would suspect this is more useful than the dump-everything
>> handlers: just give a *full* *detailed* walk for the actually interesting
>> virtual address (the one faulted on).
>
>>> I have attached new output from xl dmesg, this time with iommu=debug on (the
>>> option changed from 4.1 to 4.2).
>
>> Not easy to glean any more from that, without extra tracing such as
>> described above, and/or digging into the guest to find what driver-side
>> actions are causing the faults.
>
> OK, too bad!
> With xen 4.1 i haven't experienced those page faults, but a diff between /xen/drivers/passthrough/amd in both trees show quite some changes :(

Did you also update xen tools accordingly? Sometime I also saw a few 
IO_PAGE_FAULTs came from nic if my tools version and HV version did not 
match. But using recent 4.2 and corresponding xl, my tests went well.
BTW: You could also try iommu=no-sharept to see if it helps.

Thanks,
Wei

>>   -- Keir
>
>>>
>>>
>>>>   -- Keir
>>>
>
>
>
>
>
>

  reply	other threads:[~2012-09-03 15:20 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-08-31 21:45 Using debug-key 'o: Dump IOMMU p2m table, locks up machine Sander Eikelenboom
2012-08-31 22:24 ` Santosh Jodh
2012-08-31 22:42   ` Sander Eikelenboom
2012-08-31 22:57     ` Santosh Jodh
2012-08-31 23:16       ` Sander Eikelenboom
2012-08-31 23:58         ` Santosh Jodh
2012-09-01  0:42         ` Santosh Jodh
2012-09-03  8:14           ` Jan Beulich
2012-09-01  2:01         ` Keir Fraser
2012-09-01 17:03           ` Santosh Jodh
2012-09-01 19:13             ` Keir Fraser
2012-09-02  2:08               ` Santosh Jodh
2012-09-02  7:13                 ` Keir Fraser
2012-09-02  7:19                   ` Keir Fraser
2012-09-02  8:43               ` Sander Eikelenboom
2012-09-02 14:58                 ` Keir Fraser
2012-09-02 15:14                   ` Sander Eikelenboom
2012-09-03 15:20                     ` Wei Wang [this message]
2012-09-04  8:21                       ` Sander Eikelenboom
2012-09-04 16:43                       ` Sander Eikelenboom
2012-09-05 10:14                         ` Jan Beulich
2012-09-05 10:25                           ` Sander Eikelenboom
2012-09-05 10:40                             ` Jan Beulich
2012-09-05 10:48                               ` Sander Eikelenboom
2012-09-05 11:41                                 ` Jan Beulich
2012-09-05 12:11                                   ` Sander Eikelenboom
2012-09-05 14:15                               ` Wei Wang
2012-09-05 15:05                                 ` Jan Beulich
2012-09-05 12:48                             ` Wei Wang
2012-09-05 12:30                         ` Wei Wang
2012-09-03  8:21                 ` Jan Beulich
2012-09-03  8:33                   ` Sander Eikelenboom
2012-09-03  9:05                     ` Jan Beulich
2012-09-04  7:08                   ` Sander Eikelenboom
2012-09-04  7:46                     ` Jan Beulich
2012-09-04  8:13                       ` Sander Eikelenboom
2012-09-04  9:26                         ` Jan Beulich
2012-10-02 20:09                           ` Konrad Rzeszutek Wilk
2012-10-03 13:12                             ` Jan Beulich
2012-10-02 20:54                           ` Matt Wilson
2012-09-20  8:08                         ` Jan Beulich
2012-09-28 14:08                           ` Sander Eikelenboom
2012-09-28 21:26                           ` Jeremy Fitzhardinge
2012-10-02 20:08                             ` Konrad Rzeszutek Wilk
2012-09-02  7:42 ` Keir Fraser
2012-09-04  6:35   ` Jan Beulich
2012-09-04  6:52     ` Sander Eikelenboom
2012-09-04  7:01       ` Keir Fraser
2012-09-04  6:59     ` Keir Fraser
2012-09-04  7:55       ` Jan Beulich
2012-09-04  8:04         ` Keir Fraser
2012-09-04  8:11           ` Keir Fraser
2012-09-04  8:20             ` Sander Eikelenboom
2012-09-04  8:38             ` Jan Beulich
2012-09-04  8:54               ` Keir Fraser
2012-09-04  9:40                 ` Jan Beulich
2012-09-04 13:29                   ` Andrew Cooper

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=5044CAD7.8030800@amd.com \
    --to=wei.wang2@amd.com \
    --cc=Santosh.Jodh@citrix.com \
    --cc=keir.xen@gmail.com \
    --cc=linux@eikelenboom.it \
    --cc=xen-devel@lists.xen.org \
    /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).