Re: reiserfs blocks long on getdents64() during concurrent write

Oleg Drokin (green@namesys.com)
Mon, 5 Aug 2002 15:04:36 +0400


Hello!

On Mon, Aug 05, 2002 at 12:51:35PM +0200, Roland Kuhn wrote:

> > Unfortunatelly dirty flag for reiserfs gets set way to often than necessary,
> > we have some patches that should help this (from Chris Mason).
> > You can try these for yourself too, for example from here:
> > ftp://ftp.suse.com/pub/people/mason/patches/data-logging/02-commit_super-8-relocation.diff.gz
> I will try it immediately. And I will try to find documentation on how
> this all is working, so I can understand the implications. Does the
> "flushing" also happen, when the journal is full? Or is it possible to

Yes, if journal is full it is also flushed.

> begin a new journal while the old one is written out?

No, it is not possible to begin "new journal" in reiser3.

> > So you might try the patch I mentioned or you can mount with nodiratime mount
> > option to prevent updqting of directory atimes, but having atimes still to be
> > updated on regular files at the same time.
> Interesting mount option, though I think very few application actually use
> atime anyway, isn't it?

I know for sure that atime on mbox-style files used to determine which messages
are new, and which are not. (N and O status in mutt).
There may be other users as well.

Bye,
Oleg
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/