Re: port-alpha/48697: Sysinst compute illegal disk geometry and newfs is bound to fail.

[Izumi Tsutsui <tsutsui%ceres.dti.ne.jp@localhost>]

The following reply was made to PR port-alpha/48697; it has been noted by GNATS.

From: Izumi Tsutsui <tsutsui%ceres.dti.ne.jp@localhost>
To: martin%duskware.de@localhost, gnats-bugs%NetBSD.org@localhost
Cc: tsutsui%ceres.dti.ne.jp@localhost
Subject: Re: port-alpha/48697: Sysinst compute illegal disk geometry and newfs
         is bound to fail.
Date: Thu, 3 Apr 2014 07:47:18 +0900

 >  Can you please boot into sysinst, then use the Utilities menu to start
 >  a shell and show us the output of
 >  
 >     disklabel sd0
 >  
 >  ?
 
 Do you read the whole PR?
 
 The PR already says that Tru64's disklabel structure is the same
 as BSD's one and it contains Tru64's disklabel output.
 
 >  I suppose the Tru64 installation left something on the disk that we
 >  misinterpret - and sysinst needs to validate and clean that.
 
 The PR already says that the "sectors/track" written by Tru64
 seems wrong and our sysinst blindly uses it to calclate dlsize
 in md_get_info() in src/distrib/utils/sysinst/arch/alpha/md.c.
 
 >  If you want a quick workaround:
 
 The PR proposes a patch for the md_get_info() to use
 disklabel.d_secperunit for dlsize.
 
 Could you review it first, before considering the GoS result?
 Otherwise we can't pullup fixes to netbsd-6 etc.
 
 ---
 Izumi Tstusui