Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!clyde.concordia.ca!thunder.mcrcim.mcgill.edu!snorkelwacker.mit.edu!apple!usc!wuarchive!udel!haven!mimsy!chris
From: chris@mimsy.umd.edu (Chris Torek)
Newsgroups: comp.unix.questions
Subject: Re: determining size of physical memory
Keywords: core, physical memory, size of core
Message-ID: <28951@mimsy.umd.edu>
Date: 6 Jan 91 19:38:08 GMT
References: <1990Dec27.202715.27688@Neon.Stanford.EDU> <1368@prcrs.UUCP> <3318@mrsvr.UUCP>
Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742
Lines: 21

>In article <7325@plains.NoDak.edu>
(which, for some reason, is not included in the references line; apparently
the news software at mrsvr is broken)
>bakke@plains.NoDak.edu (Jeffrey P. Bakke) writes:
>>'wc -c < /dev/mem' [will often produce the answer]

In article <3318@mrsvr.UUCP> kohli@gemed (Jim Kohli) writes:
>While in theory this should work, there seems
>to be a problem with SunOS 4.01-- the /dev/mem
>driver gives you an early EOF indicating fewer
>bytes than there really are.

This is not a bug.  /dev/mem gives access to physical bus addresses.
The physical addresses of the memory in Sun 4s are not always
contiguous.  (That is why you can mix different capacity SIMMs on
these.) Note that on some machines physical memory does not start at
address 0; on these also straightforward reading of /dev/mem with `wc'
or `dd' will not work.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7163)
Domain:	chris@cs.umd.edu	Path:	uunet!mimsy!chris