All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Tian, Kevin" <kevin.tian@intel.com>
To: Nadav Har'El <nyh@math.technion.ac.il>
Cc: "kvm@vger.kernel.org" <kvm@vger.kernel.org>,
	"gleb@redhat.com" <gleb@redhat.com>,
	"avi@redhat.com" <avi@redhat.com>
Subject: RE: [PATCH 07/31] nVMX: Introduce vmcs02: VMCS used to run L2
Date: Sun, 22 May 2011 10:00:58 +0800	[thread overview]
Message-ID: <625BA99ED14B2D499DC4E29D8138F1505C9BEEFFD8@shsmsx502.ccr.corp.intel.com> (raw)
In-Reply-To: <20110520203211.GA5519@fermat.math.technion.ac.il>

> From: Nadav Har'El [mailto:nyh@math.technion.ac.il]
> Sent: Saturday, May 21, 2011 4:32 AM
> 
> On Fri, May 20, 2011, Tian, Kevin wrote about "RE: [PATCH 07/31] nVMX:
> Introduce vmcs02: VMCS used to run L2":
> > btw, shouldn't you clear recycled VMCS and reset 'cpu' and 'launched' fields?
> 
> Well, I believe the answer is "no": As far as I understood, a host is allowed
> to take a VMCS that has been used once to launch a certain guest, and then
> modify all the VMCS's fields to define a completely different guest, and then
> VMRESUME it, without doing the regular VMCLEAR/VMLAUNCH, even though
> it's
> "a different guest". Is there something wrong in my assumption? Does VMX
> keep
> anything constant between successive VMRESUMEs?

Yes, you can reuse a VMCS with a completely different state if the VMCS is used
on the same processor, and you must ensure that VMCS not having a dirty state
on other processors. The SDM 3B (21.10.1) explicitly requires:

----

No VMCS should ever be active on more than one logical processor. If a VMCS is to be
"migrated" from one logical processor to another, the first logical processor should
execute VMCLEAR for the VMCS (to make it inactive on that logical processor and to
ensure that all VMCS data are in memory) before the other logical processor
executes VMPTRLD for the VMCS (to make it active on the second logical processor).
A VMCS that is made active on more than one logical processor may become
corrupted

----

Here the vmcs02 being overridden may have been run on another processor before
but is not vmclear-ed yet. When you resume this vmcs02 with new content on a 
separate processor, the risk of corruption exists.

> 
> > Have you tried SMP L2 guest?
> 
> It "sort of" works, but it *does* appear to still have a bug which I didn't
> yet have the time to hunt... In one case, for example, an 8-vcpu L2 on an
> 8-vcpu L1 seemed to work well (e.g., doing parallel make) for about a minute,
> and then hung with some sort of page fault in the kernel.
> 

See whether cleaning up above can help here.

Thanks
Kevin

  reply	other threads:[~2011-05-22  2:01 UTC|newest]

Thread overview: 118+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-05-16 19:43 [PATCH 0/31] nVMX: Nested VMX, v10 Nadav Har'El
2011-05-16 19:44 ` [PATCH 01/31] nVMX: Add "nested" module option to kvm_intel Nadav Har'El
2011-05-16 19:44 ` [PATCH 02/31] nVMX: Implement VMXON and VMXOFF Nadav Har'El
2011-05-20  7:58   ` Tian, Kevin
2011-05-16 19:45 ` [PATCH 03/31] nVMX: Allow setting the VMXE bit in CR4 Nadav Har'El
2011-05-16 19:45 ` [PATCH 04/31] nVMX: Introduce vmcs12: a VMCS structure for L1 Nadav Har'El
2011-05-16 19:46 ` [PATCH 05/31] nVMX: Implement reading and writing of VMX MSRs Nadav Har'El
2011-05-16 19:46 ` [PATCH 06/31] nVMX: Decoding memory operands of VMX instructions Nadav Har'El
2011-05-16 19:47 ` [PATCH 07/31] nVMX: Introduce vmcs02: VMCS used to run L2 Nadav Har'El
2011-05-20  8:04   ` Tian, Kevin
2011-05-20  8:48     ` Tian, Kevin
2011-05-20 20:32       ` Nadav Har'El
2011-05-22  2:00         ` Tian, Kevin [this message]
2011-05-22  7:22           ` Nadav Har'El
2011-05-24  0:54             ` Tian, Kevin
2011-05-22  8:29     ` Nadav Har'El
2011-05-24  1:03       ` Tian, Kevin
2011-05-16 19:48 ` [PATCH 08/31] nVMX: Fix local_vcpus_link handling Nadav Har'El
2011-05-17 13:19   ` Marcelo Tosatti
2011-05-17 13:35     ` Avi Kivity
2011-05-17 14:35       ` Nadav Har'El
2011-05-17 14:42         ` Marcelo Tosatti
2011-05-17 17:57           ` Nadav Har'El
2011-05-17 15:11         ` Avi Kivity
2011-05-17 18:11           ` Nadav Har'El
2011-05-17 18:43             ` Marcelo Tosatti
2011-05-17 19:30               ` Nadav Har'El
2011-05-17 19:52                 ` Marcelo Tosatti
2011-05-18  5:52                   ` Nadav Har'El
2011-05-18  8:31                     ` Avi Kivity
2011-05-18  9:02                       ` Nadav Har'El
2011-05-18  9:16                         ` Avi Kivity
2011-05-18 12:08                     ` Marcelo Tosatti
2011-05-18 12:19                       ` Nadav Har'El
2011-05-22  8:57                       ` Nadav Har'El
2011-05-23 15:49                         ` Avi Kivity
2011-05-23 16:17                           ` Gleb Natapov
2011-05-23 18:59                             ` Nadav Har'El
2011-05-23 19:03                               ` Gleb Natapov
2011-05-23 16:43                           ` Roedel, Joerg
2011-05-23 16:51                             ` Avi Kivity
2011-05-24  9:22                               ` Roedel, Joerg
2011-05-24  9:28                                 ` Nadav Har'El
2011-05-24  9:57                                   ` Roedel, Joerg
2011-05-24 10:08                                     ` Avi Kivity
2011-05-24 10:12                                     ` Nadav Har'El
2011-05-23 18:51                           ` Nadav Har'El
2011-05-24  2:22                             ` Tian, Kevin
2011-05-24  7:56                               ` Nadav Har'El
2011-05-24  8:20                                 ` Tian, Kevin
2011-05-24 11:05                                   ` Avi Kivity
2011-05-24 11:20                                     ` Tian, Kevin
2011-05-24 11:27                                       ` Avi Kivity
2011-05-24 11:30                                         ` Tian, Kevin
2011-05-24 11:36                                           ` Avi Kivity
2011-05-24 11:40                                             ` Tian, Kevin
2011-05-24 11:59                                               ` Nadav Har'El
2011-05-24  0:57                           ` Tian, Kevin
2011-05-18  8:29                   ` Avi Kivity
2011-05-16 19:48 ` [PATCH 09/31] nVMX: Add VMCS fields to the vmcs12 Nadav Har'El
2011-05-20  8:22   ` Tian, Kevin
2011-05-16 19:49 ` [PATCH 10/31] nVMX: Success/failure of VMX instructions Nadav Har'El
2011-05-16 19:49 ` [PATCH 11/31] nVMX: Implement VMCLEAR Nadav Har'El
2011-05-16 19:50 ` [PATCH 12/31] nVMX: Implement VMPTRLD Nadav Har'El
2011-05-16 19:50 ` [PATCH 13/31] nVMX: Implement VMPTRST Nadav Har'El
2011-05-16 19:51 ` [PATCH 14/31] nVMX: Implement VMREAD and VMWRITE Nadav Har'El
2011-05-16 19:51 ` [PATCH 15/31] nVMX: Move host-state field setup to a function Nadav Har'El
2011-05-16 19:52 ` [PATCH 16/31] nVMX: Move control field setup to functions Nadav Har'El
2011-05-16 19:52 ` [PATCH 17/31] nVMX: Prepare vmcs02 from vmcs01 and vmcs12 Nadav Har'El
2011-05-24  8:02   ` Tian, Kevin
2011-05-24  9:19     ` Nadav Har'El
2011-05-24 10:52       ` Tian, Kevin
2011-05-16 19:53 ` [PATCH 18/31] nVMX: Implement VMLAUNCH and VMRESUME Nadav Har'El
2011-05-24  8:45   ` Tian, Kevin
2011-05-24  9:45     ` Nadav Har'El
2011-05-24 10:54       ` Tian, Kevin
2011-05-25  8:00   ` Tian, Kevin
2011-05-25 13:26     ` Nadav Har'El
2011-05-26  0:42       ` Tian, Kevin
2011-05-16 19:53 ` [PATCH 19/31] nVMX: No need for handle_vmx_insn function any more Nadav Har'El
2011-05-16 19:54 ` [PATCH 20/31] nVMX: Exiting from L2 to L1 Nadav Har'El
2011-05-24 12:58   ` Tian, Kevin
2011-05-24 13:43     ` Nadav Har'El
2011-05-25  0:55       ` Tian, Kevin
2011-05-25  8:06         ` Nadav Har'El
2011-05-25  8:23           ` Tian, Kevin
2011-05-25  2:43   ` Tian, Kevin
2011-05-25 13:21     ` Nadav Har'El
2011-05-26  0:41       ` Tian, Kevin
2011-05-16 19:54 ` [PATCH 21/31] nVMX: vmcs12 checks on nested entry Nadav Har'El
2011-05-25  3:01   ` Tian, Kevin
2011-05-25  5:38     ` Nadav Har'El
2011-05-25  7:33       ` Tian, Kevin
2011-05-16 19:55 ` [PATCH 22/31] nVMX: Deciding if L0 or L1 should handle an L2 exit Nadav Har'El
2011-05-25  7:56   ` Tian, Kevin
2011-05-25 13:45     ` Nadav Har'El
2011-05-16 19:55 ` [PATCH 23/31] nVMX: Correct handling of interrupt injection Nadav Har'El
2011-05-25  8:39   ` Tian, Kevin
2011-05-25  8:45     ` Tian, Kevin
2011-05-25 10:56     ` Nadav Har'El
2011-05-25  9:18   ` Tian, Kevin
2011-05-25 12:33     ` Nadav Har'El
2011-05-25 12:55       ` Tian, Kevin
2011-05-16 19:56 ` [PATCH 24/31] nVMX: Correct handling of exception injection Nadav Har'El
2011-05-16 19:56 ` [PATCH 25/31] nVMX: Correct handling of idt vectoring info Nadav Har'El
2011-05-25 10:02   ` Tian, Kevin
2011-05-25 10:13     ` Nadav Har'El
2011-05-25 10:17       ` Tian, Kevin
2011-05-16 19:57 ` [PATCH 26/31] nVMX: Handling of CR0 and CR4 modifying instructions Nadav Har'El
2011-05-16 19:57 ` [PATCH 27/31] nVMX: Further fixes for lazy FPU loading Nadav Har'El
2011-05-16 19:58 ` [PATCH 28/31] nVMX: Additional TSC-offset handling Nadav Har'El
2011-05-16 19:58 ` [PATCH 29/31] nVMX: Add VMX to list of supported cpuid features Nadav Har'El
2011-05-16 19:59 ` [PATCH 30/31] nVMX: Miscellenous small corrections Nadav Har'El
2011-05-16 19:59 ` [PATCH 31/31] nVMX: Documentation Nadav Har'El
2011-05-25 10:33   ` Tian, Kevin
2011-05-25 11:54     ` Nadav Har'El
2011-05-25 12:11       ` Tian, Kevin
2011-05-25 12:13     ` Muli Ben-Yehuda

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=625BA99ED14B2D499DC4E29D8138F1505C9BEEFFD8@shsmsx502.ccr.corp.intel.com \
    --to=kevin.tian@intel.com \
    --cc=avi@redhat.com \
    --cc=gleb@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=nyh@math.technion.ac.il \
    /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.