From: <rolf@diesing.net> Date: Thu, 17 Jan 2002 23:24:33 +0100 Message-ID: <BCEMLCNMANGFBIDLIOAJAENGCAAA.rolf@diesing.net> Subject: RE: [suse-sparc] SuSE Kernel Source
Good News! Good News! Good News! Good News! Good News! Good News! Good News!
Welcome to SuSE Linux 7.3 (SPARC) - Kernel 2.4.17-for-rolf (ttyS0).
sunpc login:
---- sunpc:~ # cat /proc/version Linux version 2.4.17-for-rolf (root@sparc85) (gcc version egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)) #1 SMP Wed Jan 16 23:49:53 CET 2002---- sunpc:~ # cat /proc/cpuinfo cpu : ROSS HyperSparc RT625 or RT626 fpu : ROSS HyperSparc combined IU/FPU promlib : Version 3 Revision 2 prom : 2.25 type : sun4m ncpus probed : 2 ncpus active : 2 Cpu0Bogo : 104.85 Cpu1Bogo : 104.85 MMU type : ROSS HyperSparc contexts : 4096 nocache total : 1048576 nocache used : 184320 CPU0 : online CPU1 : online sunpc:~ #
The Kernel from Ingo is working slow but fine at this Time! Now i try if it is stable and how to compile a kernel on my own Box. I found a "Missconfiguration" witch it makes the Kernel "unstartable". It is important to fix this _before_ using _this_ Kernel.
In normal SuSE-Installations i'm seting IP_FORWARD in rc.config to "no" (the boot-script use this to disable routing between Networks) and on starting the SuSE-Firewall it may be enabled again. But with IP_FORWARD="no" the Kernel dies and you will have to switch the Box. Switching this Flag per Hand later in the Shell like echo "0" > /proc/sys/net/ipv4/ip_forward is ok and have no more bad consequences. The Problem may depends on the position in the startscripts (/etc/rc.d/boot), i think it is possible to move this in a later position (starting network devices) without loosing the functionality. Not tested yet.
The second "Problem" is a Problem on _my_ Box but other Boxes (mostly Server) may have this Problem too. I use no VGA/Keyboard at my Box and acessing it with ttyS0/ttya. In this Consequence there are no console-tty's at this Box. Some Scripts and Programms unses per default tty's for output and/or buffering output and this sounds like a BIG Problem. I think, SuSE know this and they made a smal fix for s390 in the Bootscript at /etc/rc.d/boot like # # s390 is 'like' serial console(better: no console ttys at all) # and we don't have access to the HW Clock # if test "$HOSTTYPE" = "s390" ; then SERIAL_CONSOLE=yes HWCLOCK_ACCESS=no fi
But the SERIAL_CONSOLE may be set on Sparcs and PA-Risc's (HP9000) too! (not only for fsck, many things depend on this like entrys in inittab or syslog.conf)
For all PC-Freaks, i386-Linuxer, and other Intel-People - ->Some Boxes didn't use Grafikboards, Keyboards, F-Key-Console-tty's, Mousepads,...<- (I think, PC'users will never imagine this! ;-) )
Some Scripts may have "devicedepended" outputs at the described way and need fixing by Hand. Using console-tty's on Boxes without them may crash the Box. A boot.log-Example:
Running /etc/init.d/boot.local done Creating /var/log/boot.msg done wrong console number
Rolf
-- To unsubscribe, e-mail: suse-sparc-unsubscribe@suse.com For additional commands, e-mail: suse-sparc-help@suse.com
This archive was generated by hypermail 2.1.0 : Thu Jan 17 2002 - 14:24:29 PST