Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 29 Aug 1995 19:31:38 -0700 (PDT)
From:      "Rodney W. Grimes" <rgrimes@gndrsh.aac.dev.com>
To:        bde@zeta.org.au (Bruce Evans)
Cc:        leo@lisa.rur.com, freebsd-hackers@FreeBSD.ORG, jbryant@argus.iadfw.net, rashid@haven.ios.com
Subject:   Re: S.O.S -2.1Stable and ASUSP54TP4
Message-ID:  <199508300231.TAA05628@gndrsh.aac.dev.com>
In-Reply-To: <199508292149.HAA02386@godzilla.zeta.org.au> from "Bruce Evans" at Aug 30, 95 07:49:56 am

next in thread | previous in thread | raw e-mail | index | archive | help
> 
> >And to state my reason for agreement that parity is a ``itsy-bitsy comfort'',
> >think about the fact that 80% of your memory access are going to a L2
> >cache that has never had parity on it, yet has a same FIT rate as the
> >main memory system.  Basically your more likely today to take a single
> >bit error in your cache as you are in main memory :-(.
> 
> Is there anything to detect or correct errors in the registers or
> control logic?

The Pentium data books are not specific on this issue.  I can not publically
disclose that information due to NDA's.  I can say that the description of
signal IERR# would lead you to reasonable conclusions:

IERR#, Type Output:  The internal error pin is used to indicate two types of
errors, internal parity errors and functional redundancy errors.  If a parity
error occurs on a read from an internal array, the Pentium processor will
assert the IERR# pin for one clock and then shutdown.  If the Pentium
processor is configured as a checker and a mismatch occurs between the value
sampled on the pins and the corresponding value computed internally, the
Pentium process will assert IERR# two clocks after the mismatched value
is returned.

Source:  ``Pentium(TM) Processors and Related Products'', Intel Corporation,
1995.  Order #241732, ISBN 1-55512-239-6.

For full detection of register and control logic errors you run 2 Pentiums
in fucntional redudancy mode.  I am aware of products that do just that.

-- 
Rod Grimes                                      rgrimes@gndrsh.aac.dev.com
Accurate Automation Company                 Reliable computers for FreeBSD



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199508300231.TAA05628>