kvm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Avi Kivity <avi@redhat.com>
To: "Nadav Har'El" <nyh@il.ibm.com>
Cc: kvm@vger.kernel.org, gleb@redhat.com
Subject: Re: [PATCH 08/29] nVMX: Fix local_vcpus_link handling
Date: Sun, 30 Jan 2011 12:08:20 +0200	[thread overview]
Message-ID: <4D453894.4040706@redhat.com> (raw)
In-Reply-To: <201101270833.p0R8Xv9G002487@rice.haifa.ibm.com>

On 01/27/2011 10:33 AM, Nadav Har'El wrote:
> In VMX, before we bring down a CPU we must VMCLEAR all VMCSs loaded on it
> because (at least in theory) the processor might not have written all of its
> content back to memory. Since a patch from June 26, 2008, this is done using
> a per-cpu "vcpus_on_cpu" linked list of vcpus loaded on each CPU.
>
> The problem is that with nested VMX, we no longer have the concept of a
> vcpu being loaded on a cpu: A vcpu has multiple VMCSs (one for L1, others for
> each L2), and each of those may be have been last loaded on a different cpu.
>
> This trivial patch changes the code to keep on vcpus_on_cpu only L1 VMCSs.
> This fixes crashes on L1 shutdown caused by incorrectly maintaing the linked
> lists.
>
> It is not a complete solution, though. It doesn't flush the inactive L1 or L2
> VMCSs loaded on a CPU which is being shutdown. Doing this correctly will
> probably require replacing the vcpu linked list by a link list of "saved_vcms"
> objects (VMCS, cpu and launched), and it is left as a TODO.
>

It looks like the right thing is a structure that represents the common 
things between 02 and 02 vmcses:

- pointer to memory
- cpu
- linked list entries for vcpus_on_vcpu (to be renamed vmcses_on_cpu)

You could then use vcpu_clear() in the previous patch.

-- 
error compiling committee.c: too many arguments to function


  reply	other threads:[~2011-01-30 10:08 UTC|newest]

Thread overview: 47+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-01-27  8:29 [PATCH 0/29] nVMX: Nested VMX, v8 Nadav Har'El
2011-01-27  8:30 ` [PATCH 01/29] nVMX: Add "nested" module option to vmx.c Nadav Har'El
2011-01-27  8:30 ` [PATCH 02/29] nVMX: Implement VMXON and VMXOFF Nadav Har'El
2011-01-27  8:31 ` [PATCH 03/29] nVMX: Allow setting the VMXE bit in CR4 Nadav Har'El
2011-01-27  8:31 ` [PATCH 04/29] nVMX: Introduce vmcs12: a VMCS structure for L1 Nadav Har'El
2011-01-27  8:32 ` [PATCH 05/29] nVMX: Implement reading and writing of VMX MSRs Nadav Har'El
2011-01-30  9:52   ` Avi Kivity
2011-01-31  8:57     ` Nadav Har'El
2011-01-31  9:01       ` Avi Kivity
2011-01-27  8:32 ` [PATCH 06/29] nVMX: Decoding memory operands of VMX instructions Nadav Har'El
2011-01-27  8:33 ` [PATCH 07/29] nVMX: Hold a vmcs02 for each vmcs12 Nadav Har'El
2011-01-30 10:02   ` Avi Kivity
2011-01-31  9:26     ` Nadav Har'El
2011-01-31  9:41       ` Avi Kivity
2011-02-03 12:57     ` Nadav Har'El
2011-02-06  9:16       ` Avi Kivity
2011-02-13 13:04         ` Nadav Har'El
2011-02-13 14:58           ` Avi Kivity
2011-02-13 20:07             ` Nadav Har'El
2011-01-27  8:33 ` [PATCH 08/29] nVMX: Fix local_vcpus_link handling Nadav Har'El
2011-01-30 10:08   ` Avi Kivity [this message]
2011-01-27  8:34 ` [PATCH 09/29] nVMX: Add VMCS fields to the vmcs12 Nadav Har'El
2011-01-30 10:10   ` Avi Kivity
2011-01-27  8:34 ` [PATCH 10/29] nVMX: Success/failure of VMX instructions Nadav Har'El
2011-01-27  8:35 ` [PATCH 11/29] nVMX: Implement VMCLEAR Nadav Har'El
2011-01-30 12:07   ` Avi Kivity
2011-01-27  8:35 ` [PATCH 12/29] nVMX: Implement VMPTRLD Nadav Har'El
2011-01-27  8:36 ` [PATCH 13/29] nVMX: Implement VMPTRST Nadav Har'El
2011-01-27  8:37 ` [PATCH 14/29] nVMX: Implement VMREAD and VMWRITE Nadav Har'El
2011-01-27  8:37 ` [PATCH 15/29] nVMX: Prepare vmcs02 from vmcs01 and vmcs12 Nadav Har'El
2011-01-27  8:38 ` [PATCH 16/29] nVMX: Move register-syncing to a function Nadav Har'El
2011-01-27  8:38 ` [PATCH 17/29] nVMX: Implement VMLAUNCH and VMRESUME Nadav Har'El
2011-01-27  8:39 ` [PATCH 18/29] nVMX: No need for handle_vmx_insn function any more Nadav Har'El
2011-01-27  8:39 ` [PATCH 19/29] nVMX: Exiting from L2 to L1 Nadav Har'El
2011-01-27  8:40 ` [PATCH 20/29] nVMX: Deciding if L0 or L1 should handle an L2 exit Nadav Har'El
2011-01-27  8:40 ` [PATCH 21/29] nVMX: Correct handling of interrupt injection Nadav Har'El
2011-01-27  8:41 ` [PATCH 22/29] nVMX: Correct handling of exception injection Nadav Har'El
2011-01-27  8:41 ` [PATCH 23/29] nVMX: Correct handling of idt vectoring info Nadav Har'El
2011-01-27  8:42 ` [PATCH 24/29] nVMX: Handling of CR0 and CR4 modifying instructions Nadav Har'El
2011-01-27  8:42 ` [PATCH 25/29] nVMX: Further fixes for lazy FPU loading Nadav Har'El
2011-01-27  8:43 ` [PATCH 26/29] nVMX: Additional TSC-offset handling Nadav Har'El
2011-01-27  8:43 ` [PATCH 27/29] nVMX: Add VMX to list of supported cpuid features Nadav Har'El
2011-01-27  8:44 ` [PATCH 28/29] nVMX: Miscellenous small corrections Nadav Har'El
2011-01-27  8:44 ` [PATCH 29/29] nVMX: Documentation Nadav Har'El
2011-01-28  8:41 ` [PATCH 0/29] nVMX: Nested VMX, v8 Juerg Haefliger
2011-01-28 17:16   ` Nadav Har'El
2011-01-31 10:07   ` Nadav Har'El

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=4D453894.4040706@redhat.com \
    --to=avi@redhat.com \
    --cc=gleb@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=nyh@il.ibm.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).