xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Tim Deegan <tim@xen.org>
Cc: Yang Zhang <yang.z.zhang@intel.com>,
	xiantao.zhang@intel.com, JBeulich@suse.com,
	xen-devel@lists.xen.org
Subject: Re: [PATCH] Don't track all memory when enabling log dirty to track vram
Date: Mon, 10 Feb 2014 10:42:05 +0000	[thread overview]
Message-ID: <52F8ACFD.7090801@citrix.com> (raw)
In-Reply-To: <20140210080314.GA758@deinos.phlegethon.org>

On 10/02/14 08:03, Tim Deegan wrote:
> At 14:14 +0800 on 10 Feb (1392038040), Yang Zhang wrote:
>> From: Yang Zhang <yang.z.zhang@Intel.com>
>>
>> When enabling log dirty mode, it sets all guest's memory to readonly.
>> And in HAP enabled domain, it modifies all EPT entries to clear write bit
>> to make sure it is readonly. This will cause problem if VT-d shares page
>> table with EPT: the device may issue a DMA write request, then VT-d engine
>> tells it the target memory is readonly and result in VT-d fault.
> So that's a problem even if only the VGA framebuffer is being tracked
> -- DMA from a passthrough device will either cause a spurious error or
> fail to update the dirt bitmap. 
>
> I think it would be better not to allow VT-d and EPT to share
> pagetables in cases where devices are passed through (i.e. all cases
> where VT-d is in use).
>
> Tim.

Sadly, this would make shared VT-d/EPT completely pointless as a
feature, causing extra memory overhead in Xen by having to maintain EPT
and IOMMU tables separately.

Any usecase which doesn't involve dirty vram tracking (e.g. headless VM
with SRIOV, PVH dom0) would be adversely affected.

~Andrew

  parent reply	other threads:[~2014-02-10 10:42 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-02-10  6:14 [PATCH] Don't track all memory when enabling log dirty to track vram Yang Zhang
2014-02-10  8:03 ` Tim Deegan
2014-02-10  8:15   ` Zhang, Yang Z
2014-02-11  9:02     ` Tim Deegan
2014-02-11 10:59       ` George Dunlap
2014-02-11 11:55         ` Tim Deegan
2014-02-11 12:57           ` Jan Beulich
2014-02-11 15:55             ` George Dunlap
2014-02-12  0:53               ` Zhang, Yang Z
2014-02-13 15:46                 ` George Dunlap
2014-02-13 15:55                   ` Jan Beulich
2014-02-13 16:20                     ` Tim Deegan
2014-02-13 16:25                       ` George Dunlap
2014-02-13 16:45                         ` Processed: " xen
2014-02-17 10:18                       ` Jan Beulich
2014-02-17 12:23                         ` George Dunlap
2014-02-17 12:37                           ` Jan Beulich
2014-02-17 14:51                         ` George Dunlap
2014-02-17 15:05                           ` Jan Beulich
2014-02-18  3:14                             ` Zhang, Yang Z
2014-02-18 10:26                               ` George Dunlap
2014-02-19  1:28                                 ` Zhang, Yang Z
2014-02-19  8:55                                   ` Jan Beulich
2014-02-19 11:03                                     ` George Dunlap
2014-02-19 11:13                                       ` Jan Beulich
2014-02-19 11:17                                         ` George Dunlap
2014-02-17 15:00                         ` Jan Beulich
2014-02-18  3:25                           ` Zhang, Yang Z
2014-02-18  8:45                             ` Jan Beulich
2014-02-18 11:46                             ` Jan Beulich
2014-02-18 15:28                               ` Jan Beulich
2014-02-19  6:40                                 ` Xu, Dongxiao
2014-02-19  1:17                               ` Zhang, Yang Z
2014-02-19  8:50                                 ` Jan Beulich
2014-02-18  8:30                           ` Jan Beulich
2014-05-19  7:48                       ` Zhang, Yang Z
2014-05-19  9:03                         ` Jan Beulich
2014-05-20  3:09                           ` Zhang, Yang Z
2014-05-20  7:17                             ` Jan Beulich
2014-05-19 13:27                         ` George Dunlap
2014-05-19 13:50                           ` Jan Beulich
2014-05-19 13:59                             ` George Dunlap
2014-05-19 14:19                               ` Jan Beulich
2014-05-20  3:13                           ` Zhang, Yang Z
2014-05-20  7:20                             ` Jan Beulich
2014-05-20 10:12                               ` George Dunlap
2014-05-20 10:46                                 ` Jan Beulich
2014-05-21  1:02                                   ` Zhang, Yang Z
2014-05-21  7:49                                     ` Jan Beulich
2014-05-21  8:37                                       ` Zhang, Yang Z
2014-05-21  9:58                                         ` Jan Beulich
2014-05-23  6:42                                         ` Jan Beulich
2014-05-26  8:16                                           ` Zhang, Yang Z
2014-05-26  9:04                                             ` Jan Beulich
2014-05-31  1:26                                               ` Nakajima, Jun
2014-06-02  6:55                                                 ` Jan Beulich
2014-06-02 14:06                                                   ` George Dunlap
2014-06-02 14:27                                                     ` Jan Beulich
2014-06-02 15:03                                                       ` George Dunlap
2014-02-10 10:42   ` Andrew Cooper [this message]
2014-02-10 16:13 ` George Dunlap
2014-02-10 16:30   ` Processed: " xen

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=52F8ACFD.7090801@citrix.com \
    --to=andrew.cooper3@citrix.com \
    --cc=JBeulich@suse.com \
    --cc=tim@xen.org \
    --cc=xen-devel@lists.xen.org \
    --cc=xiantao.zhang@intel.com \
    --cc=yang.z.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 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).