[Top] [All Lists]

Re: [PATCH 00/31] KVM/MIPS: Implement hardware virtualization via the MI

To: Sanjay Lal <>
Subject: Re: [PATCH 00/31] KVM/MIPS: Implement hardware virtualization via the MIPS-VZ extensions.
From: David Daney <>
Date: Mon, 10 Jun 2013 10:14:27 -0700
Cc:,,,, David Daney <>
Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed;; s=20120113; h=message-id:date:from:user-agent:mime-version:to:cc:subject :references:in-reply-to:content-type:content-transfer-encoding; bh=4EYNFZo9lNFTIj9qk9m5ALZHpGtPbv55aY5xOXVXWHQ=; b=a7uMAU2nX7hXBpg8gFHdd7/mlva4uyTfkfTQGYS95eyAZ+CBzjdwzWSKT0DJSETjfI kQS0EF+JBx4vLWKHGFx3MoTn5TSTEjVJ640Ms3nvAF5uG9dK4Nl8smO5XgVfQJjBVyCE etl6bGS+rLDrdY+iKHVjTtDPqmunkblUQadQDE3082KusduKkHSAZkxQKRhzxEpKJsbk huYZVUjMsSOnvdWwtQ8d+Mm5h837kMyjyRlE1sCU9NYfoeTuhlM0CsqSeb/miAJYKUrI RR4BLwCa/WRWOZbqV6/bRSG1ydNAtEHuRWkd7o2mP8TF/0A6I8Pzu5nlmnNFuykyLG+o ygMQ==
In-reply-to: <>
List-archive: <>
List-help: <>
List-id: linux-mips <>
List-owner: <>
List-post: <>
List-software: Ecartis version 1.0.0
List-subscribe: <>
List-unsubscribe: <>
Original-recipient: rfc822;
References: <> <>
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130311 Thunderbird/17.0.4
On 06/10/2013 09:43 AM, Sanjay Lal wrote:

On Jun 7, 2013, at 4:03 PM, David Daney wrote:

From: David Daney <>

These patches take a somewhat different approach to MIPS
virtualization via the MIPS-VZ extensions than the patches previously
sent by Sanjay Lal.

Several facts about the code:

o Currently probably only usable on the OCTEON III CPU model, as some
  MIPS-VZ implementation-defined behaviors were assumed to have the
  OCTEON III behavior.

I've only briefly gone over the patches, but I was wondering if the Cavium 
implementation has support for GuestIDs, which are optional in the VZ-ASE?

No, OCTEON III does not support this optional behavior. For the most part this only impacts TLB management. I think in the context of KVM, you cannot leave foreign Guest's TLB entries present in the Guest TLB anyhow, so the feature is of little use.

Since MIPS TLBs are managed by software, it is valid for a guest to populate the TLB in any way it desires. To have the hypervisor (KVM) come and randomly invalidate the TLB, via the GuestID mechanism, would both be detectable by the guest, and potentially make the guest operate incorrectly.

David Daney


<Prev in Thread] Current Thread [Next in Thread>