[4296] in linux-scsi channel archive
Upgraded to 2.1.107 ... almost succesfully.
daemon@ATHENA.MIT.EDU (Brian Schau)
Mon Jul 6 12:28:19 1998
From: Brian Schau <bsc@fleggaard.dk>
To: "'linux-kernel@vger.rutgers.edu'" <linux-kernel@vger.rutgers.edu>,
"'linux-scsi@vger.rutgers.edu'" <linux-scsi@vger.rutgers.edu>
Date: Mon, 6 Jul 1998 18:21:17 +0200
Hi,
To all who responded to my request for help on upgrading to 2.1.107.
Especially thanks to Alan - your reply rocks, it wasn't necessary to go
all the way!
I just compiled and rebooted my new 2.1.107 kernel. It almost went
smooth ...
1) MODULES: When doing a 'depmod -a' it complains about 'unresolved
symbols'. I guess upgrading to modutils-2.1.something will help?
(Ok, silly question).
2) IO-APIC: I found this in the boot-log:
Jul 6 17:46:01 enjoy linux: ENABLING IO-APIC IRQs
Jul 6 17:46:01 enjoy linux: init IO_APIC IRQs
Jul 6 17:46:01 enjoy linux: IO-APIC pin 0broken BIOS, changing pin 11
to edge
Jul 6 17:46:01 enjoy linux: broken BIOS, changing pin 12 to edge
Jul 6 17:46:01 enjoy linux: broken BIOS, changing pin 15 to edge
Jul 6 17:46:01 enjoy linux: not connected.
... I've never seen IO-APIC's reports before, so I am a bit worried
about 'broken BIOS' & 'not connected'. What do they mean? Do I need
to worry?
3) SCSI: I also get these on 2.0.34:
Jul 6 17:46:01 enjoy linux: (scsi0:-1:-1:-1) Bad scbptr 255 during
SELTO.
Jul 6 17:46:01 enjoy linux: (scsi0:-1:-1:-1) Referenced SCB 255 not
valid during SELTO.
Jul 6 17:46:01 enjoy linux: SCSISEQ = 0x12 SEQADDR = 0x9 SSTAT0
= 0x15 SSTAT1 = 0x88
Jul 6 17:46:01 enjoy linux: (scsi1:-1:-1:-1) Bad scbptr 255 during
SELTO.
Jul 6 17:46:01 enjoy linux: (scsi1:-1:-1:-1) Referenced SCB 255 not
valid during SELTO.
Jul 6 17:46:01 enjoy linux: SCSISEQ = 0x12 SEQADDR = 0x9 SSTAT0
= 0x15 SSTAT1 = 0x88
... but this time the server doesn't crash - T H A N K S! (BTW: what
does the errors mean? They make the kernel panic ... Is it bad
hardware?)
4) RPC: nfs/mount needs to be upgraded. But why does RPC complain?
Jul 6 17:46:01 enjoy linux: nfs warning: mount version older than
kernel
Jul 6 17:46:01 enjoy linux: portmap: RPC call returned error 111
Jul 6 17:46:01 enjoy linux: RPC: task of released request still queued!
Jul 6 17:46:01 enjoy linux: RPC: (task is on xprt_pending)
Jul 6 17:46:01 enjoy linux: portmap: RPC call returned error 111
Jul 6 17:46:01 enjoy linux: RPC: task of released request still queued!
Jul 6 17:46:01 enjoy linux: RPC: (task is on xprt_pending)
Jul 6 17:46:01 enjoy linux: lockd_up: makesock failed, error=-111
Jul 6 17:46:01 enjoy linux: portmap: RPC call returned error 111
Jul 6 17:46:01 enjoy linux: RPC: task of released request still queued!
Jul 6 17:46:01 enjoy linux: RPC: (task is on xprt_pending)
Jul 6 17:46:10 enjoy fspool[290]: started
Jul 6 17:46:10 enjoy fspool[290]: faxmachine: rlis3.fleggaard.dk
Jul 6 17:46:10 enjoy fspool[290]: root: /home
Jul 6 17:46:10 enjoy fspool[290]: running
5) In 2.0.33-days I had this line in my /etc/login.defs, which allows me
to access files > 1Gb:
ULIMIT 4194304
... if I put this in my /etc/login.defs, the fs complains when copying
even small files (<1K!) from filesystem to filesystem. Setting the
ULIMIT to the default-value (2097152) "almost" solves the problem - I
can create/copy files up to 1Gb, but I can't create/copy files >1Gb.
How do I express 'unlimited' in ULIMIT?
Rodger Donaldson: you've heard about Unifix. When 2.1.107 boots, it
tells you in nice letter: "POSIX conformance testing by UNIFIX" ...
it's the same firm ;o)
Best regards,
Brian ("crossing my fingers that the upgrade to 4 CPU's will be as
smooth as the upgrade to 2.1.107 ..." ;o)
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu
