kern/115572: [gbde] gbde partitions fail at 28bit/48bit LBA addressing boundary

[Mark Linimon]

The following reply was made to PR kern/115572; it has been noted by GNATS.

From: linimon at lonesome.com (Mark Linimon)
To: bug-followup at FreeBSD.org
Cc:  
Subject: Re: kern/115572: [gbde] gbde partitions fail at 28bit/48bit LBA addressing boundary
Date: Fri, 17 Aug 2007 17:07:23 -0500

 ----- Forwarded message from Graham <gbradley at rocketmail.com> -----
 
 Sorry, but my only solution at the moment is to use geli !
 
 We should not guess at the cause, I know, but I am writing chunks in
 succession to a newly gbde initialised drive, checking access after
 each shot, with the intent of narrowing down to the specific write
 which causes the damage --- i.e. the next read and write at the 28/48
 boundary fails.
 
 This may take a little time, but I'd like to find a small specific
 write block address which does the damage.  It will then be easier
 (with a specific case) to work through the block address mapping which
 ghappens in gbde.  Then maybe a fix....hope this strategy is ok.
 
 On an optimistic note --- if we can pin this down, it may help explain
 some of the other gbde corruption cases in the bug list.
 
 I will keep you posted as soon as I can find a small, specific case
 which causes the problem.
 
 regards
 Graham
 
 ----- End forwarded message -----