Subject: Re: Longer "dmesg" buffer?
To: Mike Cheponis , Daniel Carosone <dan@geek.com.au>
From: Richard Rauch <rauch@rice.edu>
List: netbsd-help
Date: 10/25/2002 07:22:22
Re. http://mail-index.netbsd.org/port-i386/2002/10/25/0005.html
and http://mail-index.netbsd.org/netbsd-help/2002/10/14/0008.html


(I think that this question transends port-i386, which is why when I asked
this same question a week or so ago, I asked on netbsd-help.  I'm replying
there.  If you like, think of it as a followup to the other thread,
sparked by seeing the parallel thread on port-i386...(^&)

When I did this, I noticed that my laptop's disk did a *lot* more grinding
during boot (and I think that it was taking longer to boot).

It still did boot, and ran pretty much normally.  But after putting up
with that for a while, I removed the option.

(No, I don't think that the laptop was thrashing.  It has 128MB of RAM; it
shouldn't thrash during boot...(^&)

Also, the problem seemed to get worse as I increased the buffer size.  (I
went up as high as 32K, I think; for some reason, even *that* wasn't
enough space for the maximally-verbose laptop dmesg's to be entirely
captured and saved.  I actually did a "dmesg | wc" and found that I was
in fact getting (almost) the full 32K of used space---I assumed that the
missing bytes were due to buffer tracking overhead.))


  ``I probably don't know what I'm talking about.'' --rauch@math.rice.edu