lkml.org 
[lkml]   [2005]   [Aug]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 0/6] i386 virtualization patches, Set 3
Brian Gerst wrote:

> If you really want to test the math emu code, you can hack check_x87
> in head.S to always leave the fpu disabled. Then you can test it on
> any cpu, not just a 386.


That is a good idea, and while a valid point, it actually still requires
writing the code to actually test the FPU, specifically, using weird
prefixes, LDT based segments, and other oddities that don't get
generated from "normal" C code. I'm pretty sure the existing code works
for the 99% cases or else it wouldn't have gotten there in the first
place. But testing the corner cases here is even nastier than testing
the LDT corner cases - you would basically need to write a lot of hand
coded i387 assembler. Perhaps such a test might exist, but in all
honesty, without a comprehensive test, it is simply far too easy to
introduce a bug here, and far too likely it will either not be noticed
until it has caused someone a possibly undetected numerical error, or
I'm just wasting my time because noone is using this code anyways.
Fortunately, the Hubble telescope has been upgraded to a 486 ;)

Zach
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-08-16 05:36    [W:0.328 / U:2.576 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site