public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Anthony Liguori <aliguori-NZpS4cJIG2HvQtjrzfazuQ@public.gmane.org>
To: Ingo Molnar <mingo-X9Un+BFzKDI@public.gmane.org>
Cc: kvm-devel <kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org>
Subject: Re: [RFT] mmu optimizations branch
Date: Tue, 02 Jan 2007 21:05:46 -0600	[thread overview]
Message-ID: <459B1D8A.6040604@cs.utexas.edu> (raw)
In-Reply-To: <20070103022241.GA13840-X9Un+BFzKDI@public.gmane.org>

Ingo Molnar wrote:
> * Avi Kivity <avi-atKUWr5tajBWk0Htik3J/w@public.gmane.org> wrote:
>
>   
>>> lat_ctx -s 0 [zero memory footprint]:
>>>
>>>  -------------------------------------------------
>>>    #tasks    native    kvm-r4204    kvm-r4232(mmu)
>>>  -------------------------------------------------
>>>        2:      2.02       180.91         9.19
>>>       20:      4.04       183.21        10.01
>>>       50:      4.30       185.95        11.27
>>>
>>> so here it's a /massive/, almost 20 times speedup!
>>>
>>>       
>> Excellent. 10us is approximately the vmexit overhead on intel (we
>> regularly see 100-120k exits/sec), so it means a context switch is
>> exactly one exit.  Hard to beat without nested page tables.
>>     
>
> actually, the VM entry+exit cost on this CPU is around 3-4 microseconds,
> so it's still 2 VM exits per context switch.
>
> I debugged this a bit, and what happens is that when Linux does a
> task-switch it does a cr3 load /and/ a write (look at __flush_tlb()) -
> and both are causing a vm exit!
>
> I have started paravirtualizing the Linux kernel for KVM. I have
> eliminated the cr3 load from the Linux kernel via paravirtualization and
> that way lat_ctx shows a ~5-6 usecs context-switch cost. That's pretty
> good i think, compared to the 2-3 usecs of native. I'll send patches for
> this tomorrow.
>   

This should be hookable via arch_{enter,leave}_cpu_mode() via 
paravirt_ops.  I was actually just looking at this myself (although I 
was focusing on lazy mmu hooks).  I've taken the route of using a VMI 
ROM to actually hook it (instead of implementing a custom paravirt_ops 
for KVM).

I can post the ROM if you're interested in this sort of approach.  I 
like using VMI as we can access most of the things hookable by 
paravirt_ops without having to change a kernel binary.

Regards,

Anthony Liguori

> 	Ingo
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share your
> opinions on IT & business topics through brief surveys - and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________
> kvm-devel mailing list
> kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/kvm-devel
>   


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

  parent reply	other threads:[~2007-01-03  3:05 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-01-01 10:32 [RFT] mmu optimizations branch Avi Kivity
     [not found] ` <4598E33B.608-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-01-02 16:11   ` Ingo Molnar
     [not found]     ` <20070102161117.GA3306-X9Un+BFzKDI@public.gmane.org>
2007-01-02 16:32       ` Avi Kivity
     [not found]         ` <459A8909.7020600-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-01-02 16:49           ` Ingo Molnar
     [not found]             ` <20070102164912.GA25271-X9Un+BFzKDI@public.gmane.org>
2007-01-02 17:07               ` Avi Kivity
2007-01-02 17:01           ` Michael Riepe
     [not found]             ` <459A8FE0.2030202-0QoEqw4nQxo@public.gmane.org>
2007-01-03  8:49               ` Avi Kivity
2007-01-02 17:02           ` Ingo Molnar
     [not found]             ` <20070102170212.GB25271-X9Un+BFzKDI@public.gmane.org>
2007-01-02 17:09               ` Avi Kivity
2007-01-03  2:22           ` Ingo Molnar
     [not found]             ` <20070103022241.GA13840-X9Un+BFzKDI@public.gmane.org>
2007-01-03  3:05               ` Anthony Liguori [this message]
     [not found]                 ` <459B1D8A.6040604-NZpS4cJIG2HvQtjrzfazuQ@public.gmane.org>
2007-01-03  8:35                   ` Avi Kivity
2007-01-03  8:32               ` Avi Kivity
     [not found]                 ` <459B6A15.4010208-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-01-03 10:02                   ` Ingo Molnar
     [not found]                     ` <20070103100235.GA17168-X9Un+BFzKDI@public.gmane.org>
2007-01-03 10:16                       ` Avi Kivity
     [not found]                         ` <459B8267.5080000-atKUWr5tajBWk0Htik3J/w@public.gmane.org>
2007-01-03 11:30                           ` Ingo Molnar

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=459B1D8A.6040604@cs.utexas.edu \
    --to=aliguori-nzps4cjig2hvqtjrzfazuq@public.gmane.org \
    --cc=kvm-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org \
    --cc=mingo-X9Un+BFzKDI@public.gmane.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox