Subject: Re: 2.0 and >2T filesystems
To: None <tech-kern@netbsd.org>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
List: tech-kern
Date: 11/21/2005 16:06:43
>>> It would not surprise me, in the older code in 2.0 and even in
>>> -current, if there were still some 32-bit related bugs or limits.
>> Isn't legacy code fun? :-
> You told that you were trying 2.0, check the changelog carefully,
> especially this entry:

> sys/ufs/ufs/ufs_bmap.c                          patch

> 	Fix a silent truncation problem that could cause corruption
> 	with large FFSv1 file systems.
> 	[mycroft, ticket #1035]

> I don't know what this means exactly, but I would be careful.

This is *probably* mycroft fixing the problems I ran into earlier, back
when we first set this machine up, with a >1T filesystem.  There were
problems; mycroft sent me a patch, which made the symptom, at least, go
away.  My guess is that the change you refer to was his fixing that in
the master source.

In any case, I am being careful; I'm creating a file 2600111222333
bytes long, big enough to guarantee that a lot of it is beyond the 2T
point, with nonrepeating but easy-to-compute content (just the counting
numbers: 1\n2\n3\n...9\n10\n11\n12\n...99\n100\n101\n...).  When it's
done, I'll regenerate it and compare against the file's contents, to
make sure it was stored correctly.

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B