1995-11-01 - Re: /dev/random for FreeBSD [was: Re: /dev/random for Linux]

Header Data

From: Mark Murray <mark@grondar.za>
To: shields@tembel.org (Michael Shields)
Message Hash: 837677053d2301cad2940147109a99f8de39c65234f2811057ecef2dc828da0c
Message ID: <199511010614.IAA08682@grumble.grondar.za>
Reply To: N/A
UTC Datetime: 1995-11-01 06:34:59 UTC
Raw Date: Wed, 1 Nov 1995 14:34:59 +0800

Raw message

From: Mark Murray <mark@grondar.za>
Date: Wed, 1 Nov 1995 14:34:59 +0800
To: shields@tembel.org (Michael Shields)
Subject: Re: /dev/random for FreeBSD [was: Re: /dev/random for Linux]
Message-ID: <199511010614.IAA08682@grumble.grondar.za>
MIME-Version: 1.0
Content-Type: text/plain


> In article <199510311715.TAA05821@grumble.grondar.za>,
> Mark Murray <mark@grondar.za> wrote:
> > forever {
> > 	cat /dev/random > /dev/null
> > }
> > 
> > Severely limiting most decent folk's chance at getting PGP to work.
> 
> Ideally, if two processes are trying to read /dev/random at the same time,
> both would get data at half-speed.  Doesn't it work that way already?

Ideally, yes. However most processes won't swamp (and deplete) /dev/random
like this will. Most (well-behaved) processes will (should) just take what
they need. The above loop tries quite hard to take all that is there,
so any process asking for randomness will be sharing with the above loop
on an almost byte-by-byte basis, like you suggest.

The above won't leave a "pool of randomness" to act as a buffer for user
requests, so will cause a nasty slowdown.

M
--
Mark Murray
46 Harvey Rd, Claremont, Cape Town 7700, South Africa
+27 21 61-3768 GMT+0200
Finger mark@grumble.grondar.za for PGP key





Thread