home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.ms-windows.misc
- Path: sparky!uunet!inmos!titan.inmos.co.uk!lion!jeremy
- From: jeremy@lion.inmos.co.uk (Jeremy Thorp)
- Subject: Re: 32 bit multiplication not reliable ???
- Message-ID: <1992Nov9.154812.5168@titan.inmos.co.uk>
- Keywords: 386/16, processor faults
- Sender: jeremy@lion (Jeremy Thorp)
- Organization: INMOS Limited, Bristol, UK
- References: <kaustell.80@viikki.Helsinki.FI> <1992Nov6.162509.28780@news.columbia.edu>
- Date: Mon, 9 Nov 1992 15:48:12 GMT
- Lines: 20
-
- In article <1992Nov6.162509.28780@news.columbia.edu>, kra@cunixa.cc.columbia.edu (Kenneth R Aron) writes:
- |> This is just a guess, but I have a vague recollection that there
- |> are problems with a math co-pro faster than the CPU. Your post
- |> says you have a 16mhz CPU and a 25mhz co-pro. Is this conflict
- |> Windows-specific? If not, this may be your problem.
-
- When the 386 first came out, there was at least one bug in the chip. 32 bit multiplies did
- not always work. Windows is checking for this (presumably) so that you don't
- get bizarre behaviour should you run any software that uses 32-bit multiplies.
-
- I don't know if/how you can turn the warning off, though doing so strikes me as
- potentially dangerous.
-
- --
- Jeremy Thorp, INMOS Ltd, | jeremy@inmos.com
- 1000 Aztec West, Almondsbury, | jeremy@inmos.co.uk
- Bristol, BS12 4SQ, UK. |
- Tel. +44 454 616616 Fax. +44 454 617910
- The above is the personal opinion of the author,
- and does not necessarily reflect INMOS policy.
-