Subject: initial UFS2 support committed
To: None <current-users@netbsd.org>
From: Frank van der Linden <fvdl@wasabisystems.com>
List: current-users
Date: 04/02/2003 12:56:06
I just committed the first part of the UFS2 code, derived from the
FreeBSD code. I've been porting this over the last month or so
as my daytime job for Wasabi.

UFS2 is an extension to FFS. It adds 64 bit block pointers (breaking
the 1T barrier), support for extended file storage, and a few other
things. Since a few structures were renamed, and the superblock
layout for FFS changed, this touches quite a lot of files.

UFS2 is not (yet) the default type for FFS filesystems. newfs(8)
will create a normal FFS filesystem by default. If you want an
UFS2 fileystem, specify "-O 2" as an option.

No additional kernel options are needed for UFS2 support, it's
contained within the FFS code.

I have tested the code in various circumstances (with and without
softdep, with different byteorder, etc), and tested all filesystem
utilities at least once. Should you run in to filesystem problems
you had not seen before, please use send-pr to file a PR, or send
me a message.

Please note that older fsck binaries will complain a bit about
filesystems if you boot a new kernel, because of some superblock
changes. This is harmless. However, if you have 1.6 fsck binaries,
they will signal a fatal superblock mismatch with the first
alternate, because they compare too many fields (even ones that
aren't useful). This is annoying, and I'd advise peole to
upgrade their fsck_ffs binary before using a new kernel.
1.6.1 will have an fsck that is forward compatible.
Again, none of this signals actual filesystem damage, but it's
still annoying.

- Frank

-- 
Frank van der Linden                                   fvdl@wasabisystems.com
-----------------------http://www.wasabisystems.com/--------------------------
NetBSD development                                     Embedded, Storage, other