From: Marcelo Tosatti <mtosatti@redhat.com>
To: "Mao, Junjie" <junjie.mao@intel.com>
Cc: Eric Northup <digitaleric@google.com>,
"kvm@vger.kernel.org" <kvm@vger.kernel.org>
Subject: Re: [PATCH] KVM: x86: Implement PCID/INVPCID for guests with EPT
Date: Tue, 15 May 2012 00:27:31 -0300 [thread overview]
Message-ID: <20120515032731.GC19552@amt.cnet> (raw)
In-Reply-To: <EF5A1D57CFBD5A4BA5EB3ED985B6DC6E067043@SHSMSX101.ccr.corp.intel.com>
On Tue, May 15, 2012 at 03:28:13AM +0000, Mao, Junjie wrote:
> > -----Original Message-----
> > From: Marcelo Tosatti [mailto:mtosatti@redhat.com]
> > Sent: Tuesday, May 15, 2012 10:20 AM
> > To: Mao, Junjie
> > Cc: Eric Northup; kvm@vger.kernel.org
> > Subject: Re: [PATCH] KVM: x86: Implement PCID/INVPCID for guests with EPT
> >
> > On Mon, May 14, 2012 at 07:15:18AM +0000, Mao, Junjie wrote:
> > > > On Wed, May 9, 2012 at 5:32 PM, Mao, Junjie <junjie.mao@intel.com>
> > wrote:
> > > > > This patch handles PCID/INVPCID for guests.
> > > > >
> > > > > Process-context identifiers (PCIDs) are a facility by which a logical
> > processor may cache information for multiple linear-address spaces so that the
> > processor may retain cached information when software switches to a
> > different linear-address space. Refer to section 4.10.1 in IA32 Intel Software
> > Developer's Manual Volume 3A for details.
> > > > >
> > > > > For guests with EPT, the PCID feature is enabled and INVPCID behaves as
> > running natively.
> > > > > For guests without EPT, the PCID feature is disabled and INVPCID triggers
> > #UD.
> > > > >
> > > > Do I understand correctly that this means it is impossible to migrate a
> > guest which is using PCID from a host with EPT to a host without EPT (by
> > passing enable_ept=0 to the module, for example) ?
> > >
> > > I think you are right. Guests using PCID/INVPCID cannot migrate to a host
> > without it, and EPT is a precondition of this feature.
> >
> > Are there processors that support PCI/INVPCID for the host but lack support
> > for SECONDARY_EXEC_ENABLE_INVPCID?
>
> There're some with PCID but without INVPCID. For those processors support INVPCID, SECONDARY_EXEC_ENABLE_INVPCID support should be present. As PCID has little benefits without INVPCID, these two features are exposed/hidden as a whole.
That means migration control should check INVPCID host support before deciding
whether to migrate.
Thanks
prev parent reply other threads:[~2012-05-15 3:32 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-05-10 0:32 [PATCH] KVM: x86: Implement PCID/INVPCID for guests with EPT Mao, Junjie
2012-05-10 11:48 ` Avi Kivity
2012-05-11 5:58 ` Mao, Junjie
2012-05-13 10:02 ` Avi Kivity
2012-05-14 7:18 ` Mao, Junjie
2012-05-10 11:49 ` Avi Kivity
[not found] ` <CAG7+5M2XSOoHqqpbp0YbjgNNfa6DwrfP+88TwRUbhBDUDH6q6A@mail.gmail.com>
2012-05-14 7:15 ` Mao, Junjie
2012-05-15 2:20 ` Marcelo Tosatti
2012-05-15 3:28 ` Mao, Junjie
2012-05-15 3:27 ` Marcelo Tosatti [this message]
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=20120515032731.GC19552@amt.cnet \
--to=mtosatti@redhat.com \
--cc=digitaleric@google.com \
--cc=junjie.mao@intel.com \
--cc=kvm@vger.kernel.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 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.