xfs
[Top] [All Lists]

Re: xfsdump looking for libhandle.la in /lib

To: Linux XFS Mailing List <linux-xfs@xxxxxxxxxxx>
Subject: Re: xfsdump looking for libhandle.la in /lib
From: Federico Sevilla III <jijo@xxxxxxxxxxxxxxxxxxxx>
Date: Thu, 26 Jul 2001 16:55:38 +0800 (PHT)
In-reply-to: <10107261746.ZM214654@xxxxxxxxxxxxxxxxxxxxxxxx>
Sender: owner-linux-xfs@xxxxxxxxxxx
On Thu, 26 Jul 2001 at 17:46, Nathan Scott wrote:
> Current xfsdump/configure.in in cvs looks in /usr/lib/libhandle.la -
> but even if it doesn't find this, it will fall back to -lhandle, so
> I'm not sure what the problem is?  (error message?)

I'm building a Debian package so I run "debian/rules build". It failed
saying it could not find /lib/libhandle.la.

> Perhaps you have some residual state from an earlier build & need to
> do a "make realclean"?  /usr/lib is where the .la and .a ended up.

There was no residual state. I have a copy of the CVS tree that I update,
and a copy I use to build kernels. I deleted the old tree that I worked
on, updated my "updated" copy, then copied that to a new working tree. :)

> There have been some reports of fsr causing kernel panics - I saw some
> mail fly by in the last couple of days stating that problem was on
> Steve's work-in-progress list.

I remember this, too, which is why I asked to find out if maybe someone
had an update. Perhaps we can wait for Steve to let us know about how his
work-in-progress is doing. :)

> I think that in the general case, the improvement from fsr is
> relatively small - but believe there are corner cases where it makes a
> big difference.

Okay. I wonder if someone has ever written a paper or something about the
fragmentation in XFS and hence the situations where fsr really helps.
There is a current filesystem fragmentation study being done comparing
JFS, XFS and ReiserFS by a student taking his masters, but I don't think
it's finished so we can wait for that, too. :)

 --> Jijo

--
Federico Sevilla III  :: jijo@xxxxxxxxxxxxxxxxxxxx
Network Administrator :: The Leather Collection, Inc.


<Prev in Thread] Current Thread [Next in Thread>