RE: Backport of 2.6.15 code to 2.6.12, 2.6.13 and 2.6.14

Wow, Andreas! Let me know your sourceforge ID and I will give you the commit
bit for the tree. We should add this to the devel tree as a branch so Kris
can  build test rpms. 

-----Original Message-----
From: openmosix-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
[mailto:openmosix-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org] On Behalf Of Andreas
Schäfer
Sent: Saturday, February 11, 2006 9:49 PM
To: openmosix-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
Subject: [Openmosix-devel] Backport of 2.6.15 code to 2.6.12, 2.6.13 and
2.6.14

Hi!

Short Version:
--------------
I've backported the newer 2.6.15 patch to 2.6.14, 2.6.13 and 2.6.12. The
curious ones can find the links to the patches below.

Long Version:
-------------
I was amazed by Tab's recent release of the 2.6.15 patch which can actually
do the migration from the home node away and back again w/o interruption
(@Tab: excellent work, really! ;-) ). As some of you know I'm still toying
around with this coLinux and openMosix stuff to create heterogeneous
clusters. My problem was that coLinux is still stuck at 2.6.12, so
originally I tried a update of their patch to 2.6.15. This, however proved
to be a bit more tricky since coLinux seems a bit touchy if you don't have
real kernel coding skills (I don't). 

So as they say "If the mountain will not come to Mohamet, Mohamet must go to
the mountain", I started a backport of the oM patch to 2.6.12. After some
kernel panics and some sleepless nights the three patches below turned out
to work sufficiently (read: they are not well tested, but a simple
test-program of mine migrated back and forth w/o crashing). The downside is
that my boot splash screen stopped working using the 2.6.14 and 2.6.13
patches (but this was harmless and didn't cause any crashed) and started to
cause kernel panics when using the
2.6.12 patch (which I fixed by turning the splash screen off). Porting the
patches was in essence not as difficult as I had been afraid which gives me
the impression, as it seems to be only loosely coupled to a given kernel
version, that this must be really good code (IMHO writing truly decoupled
code is very hard, especially for kernel coding).

I don't know if this is in deed interesting for anybody but I thought I'd do
no harm to let you guys know ;-) 


The Patches:
------------
Patch 1:
[http://cluster.inf-ra.uni-jena.de/~gentryx/patch-openmosix-2.6.14.bz2]
  Patch against kernel 2.6.14
  Tested with 2.6.14, Gentoo patches version 8 Patch 2:
[http://cluster.inf-ra.uni-jena.de/~gentryx/patch-openmosix-2.6.13.bz2]
  Patch against kernel 2.6.13
  Tested with 2.6.13, Gentoo patches version 7 Patch 3:
[http://cluster.inf-ra.uni-jena.de/~gentryx/patch-openmosix-2.6.12.bz2]
  Patch against kernel 2.6.12
  Tested with 2.6.12, Gentoo patches version 14

Have fun!
-Andreas


-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
_______________________________________________
openMosix-devel mailing list
openMosix-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/openmosix-devel



-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid3432&bid#0486&dat1642