Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 7 Jan 2002 20:40:16 +0100
From:      Wilko Bulte <wkb@freebie.xs4all.nl>
To:        John Baldwin <jhb@FreeBSD.org>
Cc:        cvs-all@FreeBSD.org, cvs-committers@FreeBSD.org, Sheldon Hearn <sheldonh@starjuice.net>
Subject:   Re: cvs commit: src/release/sysinstall install.c
Message-ID:  <20020107204016.A11458@freebie.xs4all.nl>
In-Reply-To: <XFMail.020107102644.jhb@FreeBSD.org>; from jhb@FreeBSD.org on Mon, Jan 07, 2002 at 10:26:44AM -0800
References:  <20020107140906.A9972@freebie.xs4all.nl> <XFMail.020107102644.jhb@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Jan 07, 2002 at 10:26:44AM -0800, John Baldwin wrote:
> 
> On 07-Jan-02 Wilko Bulte wrote:
> > On Mon, Jan 07, 2002 at 03:00:21PM +0200, Sheldon Hearn wrote:
> >> 
> >> 
> >> On Mon, 07 Jan 2002 04:44:29 PST, Sheldon Hearn wrote:
> >> 
> >> >   Modified files:        (Branch: RELENG_4)
> >> >     release/sysinstall   install.c 
> >> >   Log:
> >> >   MFC rev 1.316: revert to old newfs 8192/1024 block/frag sizes on the
> >> >   alpha
> >> 
> >> This is required to avoid triggering whatever bugs exist in the Alpha
> >> boot code that prevent it from using a root partition with a block size
> >> other than 8192.
> >> 
> >> If anyone has any great ideas for a low impact method of reducing the
> >> effect of this regression to just the root partition on Alphas for
> >> 4.5-RELEASE, I'm all ears. :-)
> > 
> > Strange.. 2 hours ago I tested my RC0 snapshot (so without this revert)
> > on a DS20 and it boots just fine after the install.
> > 
> > What errros do you get?
> 
> Humm, see the '-CURRENT boot1 broken?' thread in -alpha from a week or so ago. 
> Christan "naddy" <can't remember last name> noted that either boot1 or perhaps
> the loader blew up when he reinstalled current with 16384/2048.  Reformatting
> the disk as 8192/1024 fixed his problem.

OK... I got an extra disk and tried a 4.5RC0 install on my MiataGL here.

# dumpfs /
magic   11954   time    Mon Jan  7 19:31:24 2002
id      [ 3c39f491 5bc0397e ]
cylgrp  dynamic inodes  4.4BSD
nbfree  5229    ndir    78      nifree  14972   nffree  430
ncg     1       ncyl    65      size    65540   blocks  64499
bsize   16384   shift   14      mask    0xffffc000
fsize   2048    shift   11      mask    0xfffff800
frag    8       shift   3       fsbtodb 2
...

# uname -a
FreeBSD  4.5-RC1 FreeBSD 4.5-RC1 #0: Sun Jan  6 21:13:34 GMT 2002
root@ds10.wbnet:/usr/src/sys/compile/GENERIC  alpha

This is a MiataGL600. So, sofar I have not been able to reproduce
this phantom problem on DS20 and MiataGL.

I would appreciate it if the people who were seeing this would try
the RC snapshot I put at http://people.freebsd.org/~wilko/Alpha-snaps/
That one has the 16384/2048 default for the root fs. 

Sofar I see no reason to have 4.5R/alpha use a different block/frag
size than 4.5R/x86

Wilko

-- 
|   / o / /_  _   		email: 	wilko@FreeBSD.org
|/|/ / / /(  (_)  Bulte		Arnhem, The Netherlands	

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020107204016.A11458>