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

Header Data

From: Tom Weinstein <tomw@cthulhu.engr.sgi.com>
To: perry@piermont.com
Message Hash: 0960fab4ee733b8709b5cdf84939b46a1dd0b7b4f67d52495525c03f8f4507a3
Message ID: <3096CA5A.41C6@engr.sgi.com>
Reply To: <199511010026.TAA22659@jekyll.piermont.com>
UTC Datetime: 1995-11-01 02:02:57 UTC
Raw Date: Wed, 1 Nov 1995 10:02:57 +0800

Raw message

From: Tom Weinstein <tomw@cthulhu.engr.sgi.com>
Date: Wed, 1 Nov 1995 10:02:57 +0800
To: perry@piermont.com
Subject: Re: /dev/random for FreeBSD [was: Re: /dev/random for Linux]
In-Reply-To: <199511010026.TAA22659@jekyll.piermont.com>
Message-ID: <3096CA5A.41C6@engr.sgi.com>
MIME-Version: 1.0
Content-Type: text/plain

Perry E. Metzger wrote:
> Tom Weinstein writes:
> > One problem with this scheme is that if multiple processes have
> > /dev/random open you can block unexpectedly.  If I try to avoid
> > blocking by first checking if entropy is available there's a race
> > condition if another process reads from the device.  Is there
> > another way to avoid blocking?
> Yeah. Use non-blocking I/O. Its in every version of Unix I've touched
> for over a decade.

I guess I wasn't clear.  The message I was replying to defined how
the driver decided whether to block.  Since I don't have the source
code, I was wondering whether non-blocking I/O worked for this driver.

Sure we spend a lot of money, but that doesn't mean | Tom Weinstein
we *do* anything.  --  Washington DC motto          | tomw@engr.sgi.com