• Messages sorted by: [ date ][ thread ][ subject ][ author ]
• [ Next ]In reply to: Norman Hardy

What I like is the fix availability:
BSDI, Linux - Fix available by FTP
IBM(OS2), Microsoft, NCR, SCO, Sequent, SunSoft(Solaris), Unisys

• We're working on it. Novel - Programs must be approved so we don't need it.

Ah, the superior support available for commercial software.

I don't have a Pentium manual handy. Are any of the interrupts 0-6 frequently occurring? I don't think "invalid operation" is. If none of them are frequent, than the fix should not cost too much at runtime.

At 01:34 PM 11/15/97 -0800, Mark S. Miller wrote:
>>Date: Sat, 15 Nov 1997 13:26:37 -0800
>>To: jsshapiro@earthlink.net
>>From: Norman Hardy <norm@netcom.com>
>>Subject: Pentium erratum
>>Cc: eros-arch@eros.cis.upenn.edu
>>
>>See <http://support.intel.com/support/processors/pentium/ppiie/index.htm>
>>for a description of the Pentium bug. Their workaround is truly bizare. I
>>don't have my pentium manual near but I wonder about the impact of the
>>workaroung on timing of other interrupts.
>>
>>Norman Hardy <http://www.mediacity.com/~norm>
>
>

• [ Next ] In reply to: Norman Hardy