From mboxrd@z Thu Jan  1 00:00:00 1970
From: Kiyoshi MATSUI <kmatsui@t3.rim.or.jp>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Re: Cannot boot with kfs error
Message-Id: <20030607021337.0e381ce3.kmatsui@t3.rim.or.jp>
In-Reply-To: <20030505005647.6015e0ce.kmatsui@t3.rim.or.jp>
References: <20030430160006.21083.98722.Mailman@psuvax1.cse.psu.edu>
	<20030501230715.551fdc8a.kmatsui@t3.rim.or.jp>
	<3EB250A7.8030506@plan9.escet.urjc.es>
	<20030505005647.6015e0ce.kmatsui@t3.rim.or.jp>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-2022-JP
Content-Transfer-Encoding: 7bit
Date: Sat,  7 Jun 2003 02:13:37 +0900
Topicbox-Message-UUID: c727d9fe-eacb-11e9-9e20-41e7f4b1d025

A sequel to the story of kfs i/o error at Plan9 bootstrap.

I upgraded the machine, including CPU (K6/200MHz => AthlonXP/2.0GHz),
motherboard and memory. I added a new IDE hard drive. I am still using,
however, the old ncr53C875 SCSI card.

I tried again to install Plan9 on one of the SCSI drives only to get
again the kfs i/o error at booting.

Then I installed on the IDE drive. Plan9 booted successfully!

Maybe my SCSI system has some instabilities. Maybe the Plan9's SCSI
driver has some short-comings to cope with the problem?

Anyway, I can use Plan9 now. A problem, however, remains -- long pause
of 10 minutes in booting process.

I have edited /rc/bin/termrc as follows.

***************
*** 53,59 ****
  case NCR* 'AT&TNSX'* generic* _MP_* 'alpha apc'*
    for(i in H w f t m v L S P U '$' 裡)
      /bin/bind -a '#'^$i /dev >/dev/null >[2=1]
!     for(disk in /dev/sd??) {
        if(test -f $disk/data && test -f $disk/ctl)
          disk/fdisk -p $disk/data >$disk/ctl >[2]/dev/null
        for(part in $disk/plan9*)
--- 53,61 ----
  case NCR* 'AT&TNSX'* generic* _MP_* 'alpha apc'*
    for(i in H w f t m v L S P U '$' 裡)
      /bin/bind -a '#'^$i /dev >/dev/null >[2=1]
!     echo 'testing /dev/sdC?'              # [1] Appears promptly
!     for(disk in /dev/sdC?) {
!       echo testing $disk                  # [2] Appears 10 minutes later
        if(test -f $disk/data && test -f $disk/ctl)
          disk/fdisk -p $disk/data >$disk/ctl >[2]/dev/null
        for(part in $disk/plan9*)
***************

[1] appears promptly, but [2] appears 10 minutes later. Why?

When I edited the '/dev/sdC?' to '/dev/sdC0', rio started promptly and
the screen went black out displaying only an arrow cursor. Why? The
machine has only one IDE drive i.e. /dev/sdC0, neither /dev/sdC1 nor
/dev/sdC2.

Whether probe ethernet or not makes little difference of 10 minutes
pause. Speed of CPU also makes little difference.

Anyway, I am used to wait Plan9's bootstrap reading the Plan9 documents
on another machine's Linux.

                                                            kmatsui