Re: 2.6-WOLK based on 2.6.20*

On Friday, April 13, 2007, 12:59:01 PM, Marc-Christian Petersen wrote:
> I don't know if there are still users subscribed to this list but ... :)

Yo!

'sup Marc, you got time to burn? What happened, girlfriend dump you?? :-D

For  servers  we  switched  some  customers  to FreeBSD and for firewalls to
OpenBSD where possible and apppropriate. The Zarafa mail servers that run in
VMware's  ESX could use a we-behaved bug-free well-scaling no-b/s 2.6 kernel
though. So far I have been using http://linux.1wt.eu/kernel/2.4-wt/?M=D with
great  pleasure.  It's  2.4  alright, no O(1) scheduler or stuff... but rock
stable.

> I'm cooking up a 2.6-WOLK and I want to know what people expect from a 
> 2.6-WOLK, for example, what features do you want to have?
> Please tell me your suggestions. For now I only have grsecurity2 and some VM
> tweaks and scheduler tweaks applied, runs more smooth than plain 2.6.20 :)

I  still  maintain  some  boxes  with  2.6.16.xy where xy is now like > 200.
Patches I got in there, or want but don't have because of a ton of rejects:

 - OpenIPMI from openipmi.sourceforge.net because servers speak IPMI
   nowadays, not lm-sensors

 - Squashfs3

 - grsecurity because some programmers just won't... oh you have that
   already

 - ESFQ and IMQ for QoS

 - HTB hysteresis off option because most CPUs are fast enough and it
   messes up next patch
   
diff -Nurp linux-2.6.16.1-orig/net/sched/sch_htb.c linux-2.6.16.1/net/sched/sch_htb.c
--- linux-2.6.16.1-orig/net/sched/sch_htb.c     2006-03-28 08:49:02.000000000 +0200
+++ linux-2.6.16.1/net/sched/sch_htb.c  2006-03-30 11:16:22.000000000 +0200
@@ -73,7 +73,7 @@
 #define HTB_EWMAC 2    /* rate average over HTB_EWMAC*HTB_HSIZE sec */
 #undef HTB_DEBUG       /* compile debugging support (activated by tc tool) */
 #define HTB_RATECM 1    /* whether to use rate computer */
-#define HTB_HYSTERESIS 1/* whether to use mode hysteresis for speedup */
+#define HTB_HYSTERESIS 0/* whether to use mode hysteresis for speedup */
 #define HTB_QLOCK(S) spin_lock_bh(&(S)->dev->queue_lock)
 #define HTB_QUNLOCK(S) spin_unlock_bh(&(S)->dev->queue_lock)
 #define HTB_VER 0x30011        /* major must be matched with number suplied by TC as version */ 
 or see
 http://mailman.ds9a.nl/pipermail/lartc/2006q2/019058.html for patch and
 explanation

 - http://ace-host.stuart.id.au/russell/files/tc/tc-atm/

 - Tweaks for server and desktop usage (throughput vs latency)

Cool!
Stephan



-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
WOLK - Working Overloaded Linux Kernel
[email protected]
https://lists.sourceforge.net/lists/listinfo/wolk-devel