Re: L4Ka::Pistachio Development

Jan Stoess <[email protected]> Mon, 19 Dec 2011 15:48:59 +0100
Newsgroups gmane.comp.micro-kernel.l4.l4ka.general
Message-ID <[email protected]>
This is a multi-part message in MIME format.
--------------070407000507020307050909
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Hi Robert,

On 12/15/2011 10:31 PM, Robert Campbell wrote:
> I'm Robert Campbell. I'm recently started looking at operating system 
> development where I've taken an interest in the L4ka::Pistachio 
> microkernel.
>
> I've noticed that there is not active development on the github 
> project, so I have the following questions/concerns that I would very 
> much appreciate if you could address:
>
There's not much going on in terms of new functionality, correct. 
Pistachio is still maintained in the sense that you can find a 
reasonably modern piece of hardware plus a reasonably modern build 
environment and get it going there without the compilation and booting 
bailing out left and right because it's so 1990. Also there's OS efforts 
such as Genode's OS framework which actively support Pistachio as base 
micro-kernel.

> 1. Is there active research/development going on with regards to L4 v4 
> and in particular Pistachio?
As of now, there is no active development towards a new version (v4) in 
Karlsruhe.

> 2. Looking at the page http://www.l4ka.org/99.php, I noticed the 
> following:
> Missing Features / Known Bugs
> Lipc() not implemented (mapped to IPC)
>
> for x86-64

> and http://www.l4ka.org/94.php
> Missing Features / Known Bugs
> Lipc not implemented
LIPC is defined in the API but not implemented for any architecture.

> IPC does not return local thread IDs on intra-address space IPC
> Kernel crashes without warning on systems not supporting 4MB pages
> A post-Pentium kernel crashes without warning on Pentium systems
> Kernel-ptab synchronization has race on SMP systems
Those have been fixed imho.
> SMP support for pre-ACPI systems missing
Non-ACPI SMP system aren't supported, correct.
>
> Do these bugs/missing features still stand for the source in the 
> github repository?
>
Yes. See above.
> 3. Are there any plans to implement capability mechanisms in the 
> kernel as a basis for security like fiasco.OC, OKL4 etc...?
>
Not in Karlsruhe.
> 4. Are there any plans to implement an ARM kernel?
>
Not in Karlsruhe. There once was an ARM version of Pistachio, but it has 
been discontinued.

> 5. What are the technical advantages if any to the changes made in 
> OKL4 and Fiasco.OC respectively when compared to Pistachio as it 
> stands (This one might be a bit reaching so if you may skip it) ?
>
http://en.wikipedia.org/wiki/L4_microkernel_family gives a pretty good 
overview over L4 flavors and their goals.

-Jan

-- 
Dr. Jan Stoess, KIT System Architecture Group
Phone: +49 (721) 6084 4056

--------------070407000507020307050909
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Hi Robert,<br>
    <br>
    On 12/15/2011 10:31 PM, Robert Campbell wrote:
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div>I'm Robert Campbell. I'm recently started looking at
        operating system development where I've taken an interest in the
        L4ka::Pistachio microkernel.&nbsp;</div>
      <div><br>
      </div>
      <div>I've noticed that there is not active development on the
        github project, so I have the following questions/concerns that
        I would very much appreciate if you could address:</div>
      <div><br>
      </div>
    </blockquote>
    There's not much going on in terms of new functionality, correct.
    Pistachio is still maintained in the sense that you can find a
    reasonably modern piece of hardware plus a reasonably modern build
    environment and get it going there without the compilation and
    booting bailing out left and right because it's so 1990. Also
    there's OS efforts such as Genode's OS framework which actively
    support Pistachio as base micro-kernel.<br>
    <br>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div> </div>
      <div>1. Is there active research/development going on with regards
        to L4 v4 and in particular Pistachio?</div>
    </blockquote>
    As of now, there is no active development towards a new version (v4)
    in Karlsruhe. <br>
    <br>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div>2. Looking at the page <a moz-do-not-send="true"
          href="http://www.l4ka.org/99.php">http://www.l4ka.org/99.php</a>,
        I noticed the following:</div>
      <div>Missing Features / Known Bugs <br>
        Lipc() not implemented (mapped to IPC)</div>
      <div><br>
      </div>
      <div>for x86-64</div>
    </blockquote>
    <br>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div>and <a moz-do-not-send="true"
          href="http://www.l4ka.org/94.php">http://www.l4ka.org/94.php</a></div>
      <div> Missing Features / Known Bugs<br>
        Lipc not implemented <br>
      </div>
    </blockquote>
    LIPC is defined in the API but not implemented for any architecture.<br>
    <br>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div> IPC does not return local thread IDs on intra-address space
        IPC <br>
      </div>
    </blockquote>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div>Kernel crashes without warning on systems not supporting 4MB
        pages <br>
      </div>
    </blockquote>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div> A post-Pentium kernel crashes without warning on Pentium
        systems <br>
        Kernel-ptab synchronization has race on SMP systems <br>
      </div>
    </blockquote>
    Those have been fixed imho.
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div> SMP support for pre-ACPI systems missing</div>
    </blockquote>
    Non-ACPI SMP system aren't supported, correct.<br>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div><br>
      </div>
      <div>Do these bugs/missing features still stand for the source in
        the github repository?</div>
      <div><br>
      </div>
    </blockquote>
    Yes. See above.<br>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div> </div>
      <div>3. Are there any plans to implement capability mechanisms in
        the kernel as a basis for security like fiasco.OC, OKL4 etc...?</div>
      <div><br>
      </div>
    </blockquote>
    Not in Karlsruhe.<br>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div>4. Are there any plans to implement an ARM kernel?</div>
      <div><br>
      </div>
    </blockquote>
    Not in Karlsruhe. There once was an ARM version of Pistachio, but it
    has been discontinued.<br>
    <br>
    <blockquote
cite="mid:CABqF2Rer3atc5t5RsUKHyQwAtck9gkY_cydj8-V19v-CFbW4dA@mail.gmail.com"
      type="cite">
      <div> </div>
      <div>5. What are the technical advantages if any to the changes
        made in OKL4 and Fiasco.OC respectively when compared to
        Pistachio as it stands (This one might be a bit reaching so if
        you may skip it) ?</div>
      <br>
    </blockquote>
    &nbsp;<a href="http://en.wikipedia.org/wiki/L4_microkernel_family">http://en.wikipedia.org/wiki/L4_microkernel_family</a>
    gives a pretty good overview over L4 flavors and their goals.<br>
    <br>
    -Jan<br>
    <br>
    <div class="moz-signature">-- <br>
      <font size="0">Dr. Jan Stoess, KIT System Architecture Group <br>
        Phone: +49 (721) 6084 4056</font></div>
  </body>
</html>

--------------070407000507020307050909--