Re: bin/20767

[Robert Elz <kre%munnari.OZ.AU@localhost>]

The following reply was made to PR bin/20767; it has been noted by GNATS.

From: Robert Elz <kre%munnari.OZ.AU@localhost>
To: jruohonen%iki.fi@localhost
Cc: gnats-bugs%netbsd.org@localhost
Subject: Re: bin/20767 
Date: Wed, 14 Oct 2009 19:08:14 +0700

     Date:        Wed, 14 Oct 2009 14:00:07 +0300
     From:        Jukka Ruohonen <jruohonen%iki.fi@localhost>
     Message-ID:  <20091014110007.GA28884%marx.net.bit@localhost>
 
   | Fair enough. Reducing the use of kmem sounds highly welcome, but I didn't
   | know that the usage of kvm is discouraged also in this particular scenario.
 
 Please don't think I speak for TNF, or any NetBSD developers ...
 (just in case that was not obvious).   Someone else may decide to ignore
 my opinion and commit your patch...
 
   | How about the use of cgd for swap devices, for an instance?
 
 I believe (from long distant past memory, notoriously unreliable) that
 swap avoids the first few blocks so stuff like labels remain safe.
 
 But cgd has to have some known data somewhere on the device anyway, I
 assume, otherwise it would have no way to tell is the key (password)
 given for the device is correct or not, and it is certainly able to tell
 that - so I'm assuming (guessing) that cdg already has at least one
 data block reserved for its own purposes.
 
 kre