I had goofed around with this on friday.  Initally I had installed it on 
a AMD A6 based asus board.  (which had 30-40us with rtai)  It ran around 
283us (Micheal thought there may be some tweeking needed for amd)

I took the harddrive out and threw it into a Pentium(R) Dual-Core CPU 
E5800 @ 3.20GHz × 2
Booted right up and ran xenomai at around 20us.

http://imagebin.org/index.php?mode=image&id=231774

I then booted the linuxcnc livecd and got this

http://imagebin.org/index.php?mode=image&id=231777

so very similar.  Micheal had mentioned I should try loading the 
systems.  I have not tried that yet.  He said that loading the rtai 
caused system issues vs the xenomai which just took it :)

sam


On 10/14/2012 06:21 AM, Michael Haberler wrote:
> as a prerequisite for the xenomai-integration branch work, Amit and myself 
> have been working on producing xenomai x86 kernels and support libraries.
>
> They seem to work fine, and first results are encouraging wrt latency.
>
> Two kernel versions are available: one based on 2.6.38.8, and one based on 
> 3.2.21 as supported by the latest xenomai patch set. We would be very 
> interested in any feedback. Both run fine for me on atoms and stock x86 
> boxes, with 10.04 or 12.04.
>
> Note this is the Xenomai x86 kernels only, patches for LinuxCNC are not yet 
> available.
>
> Debian packages are in http://static.mah.priv.at/public/xenomai-debs/  for a 
> trial.
>
> --
>
> To install, fetch the .deb's, and install them with 'dpkg -i *.deb'.
>
> latency test: can be found in /usr/lib/xenomai/latency
>
> To run tests after installing , cd to 
> /usr/share/libxenomai-dev/examples/native and make. If you do this on 
> precise, the stricter linker option ordering requirements suggest explicit 
> rules for sigdebug and trivial-periodic like so in the Makefile:
>
> rtprint: rtprint.c
>          $(CC) $(CFLAGS) $? $(LDFLAGS) -o $@
>
> sigdebug: sigdebug.c
>          $(CC) $(CFLAGS) $? $(LDFLAGS) -o $@
>
> trivial-periodic: trivial-periodic.c
>          $(CC) $(CFLAGS) $? $(LDFLAGS) -o $@
>
> ---
> All the work has been recorded here, and scripts for rebuilding off these 
> repos are available:
>
> http://git.mah.priv.at/gitweb/linuxcnc-kernel.git
> http://git.mah.priv.at/gitweb/xenomai-linuxcnc.git
>
> fetch-and-rebuild scripts are here in case you want to work on kernel options:
>
> http://git.mah.priv.at/gitweb/linuxcnc-kernel.git/blob/c7569797fffb424b128bf035e4ad3adbcf049086:/linuxcnc/fetch-and-build.sh
> http://git.mah.priv.at/gitweb/linuxcnc-kernel.git/blob/22054b4e6d9a2c800be911a6b361a62189774171:/linuxcnc/fetch-and-build-2.6.sh
> http://git.mah.priv.at/gitweb/xenomai-linuxcnc.git/blob/refs/heads/linuxcnc-v2.6.1:/linuxcnc/fetch-and-build.sh
>
> ----
>
> The packaging still leaves to be desired and rough edges remain - experienced 
> hands willing to help polish this are very welcome.
>
> All build steps are documented in the linuxcnc subdirectories in the above 
> repos.
>
>
> - Michael
>
>
>
>
> ------------------------------------------------------------------------------
> Don't let slow site performance ruin your business. Deploy New Relic APM
> Deploy New Relic app performance management and know exactly
> what is happening inside your Ruby, Python, PHP, Java, and .NET app
> Try New Relic at no cost today and get our sweet Data Nerd shirt too!
> http://p.sf.net/sfu/newrelic-dev2dev
> _______________________________________________
> Emc-developers mailing list
> Emc-developers@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-developers
>



------------------------------------------------------------------------------
Don't let slow site performance ruin your business. Deploy New Relic APM
Deploy New Relic app performance management and know exactly
what is happening inside your Ruby, Python, PHP, Java, and .NET app
Try New Relic at no cost today and get our sweet Data Nerd shirt too!
http://p.sf.net/sfu/newrelic-dev2dev
_______________________________________________
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers

Reply via email to