From: Tim Deegan <tim@xen.org>
To: Andres Lagar-Cavilla <andres@lagarcavilla.org>
Cc: xen-devel@lists.xensource.com, ian.campbell@citrix.com,
andres@gridcentric.ca, olaf@apefle.de, keir.xen@gmail.com,
ian.jackson@citrix.com, adin@gridcentric.ca
Subject: Re: [PATCH] RFC: Use memops for mem paging, sharing, and access, instead of domctls
Date: Thu, 12 Jan 2012 13:01:20 +0000 [thread overview]
Message-ID: <20120112130120.GF47092@ocelot.phlegethon.org> (raw)
In-Reply-To: <ed4d429d7026a3cd2b33.1326307372@xdev.gridcentric.ca>
At 13:42 -0500 on 11 Jan (1326289372), Andres Lagar-Cavilla wrote:
> Per page operations in the paging, sharing, and access tracking subsystems are
> all implemented with domctls (e.g. a domctl to evict one page, or to share one
> page).
>
> Under heavy load, the domctl path reveals a lack of scalability. The domctl
> lock serializes dom0's vcpus in the hypervisor. When performing thousands of
> per-page operations on dozens of domains, these vcpus will spin in the
> hypervisor. Beyond the aggressive locking, an added inefficiency of blocking vcpus
> in the domctl lock is that dom0 is prevented from re-scheduling.
>
> In this proposal we retain the domctl interface for setting up and tearing down
> paging/sharing/mem access for a domain. But we migrate all the per page operations
> to use the memory_op hypercalls (e.g XENMEM_*).
>
> While we naturally welcome comments on the correctness of the approach, we are also
> concerned about the viability of this API change. With 4.2 coming, this is the right
> time to get an interface right, for the long run.
I'm happy with the API change but I'd like the other users of it to
comment.
Tim.
next prev parent reply other threads:[~2012-01-12 13:01 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-01-11 18:42 [PATCH] RFC: Use memops for mem paging, sharing, and access, instead of domctls Andres Lagar-Cavilla
2012-01-12 13:01 ` Tim Deegan [this message]
2012-01-12 14:43 ` Olaf Hering
2012-01-12 16:04 ` Andres Lagar-Cavilla
2012-01-26 12:57 ` Tim Deegan
2012-01-26 13:00 ` Andres Lagar-Cavilla
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=20120112130120.GF47092@ocelot.phlegethon.org \
--to=tim@xen.org \
--cc=adin@gridcentric.ca \
--cc=andres@gridcentric.ca \
--cc=andres@lagarcavilla.org \
--cc=ian.campbell@citrix.com \
--cc=ian.jackson@citrix.com \
--cc=keir.xen@gmail.com \
--cc=olaf@apefle.de \
--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 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).