develooper Front page | perl.perl5.porters | Postings from August 2001

bigintpm.t + VMS = kaboom

Thread Next
From:
Craig A. Berry
Date:
August 6, 2001 16:17
Subject:
bigintpm.t + VMS = kaboom
Message ID:
5.1.0.14.0.20010806162728.02bfd5f8@exchi01
Even a syntax check of this test fails with:

$ perl -c [-.lib.math.bigint.t]bigintpm.t
1..1424
%SYSTEM-F-HPARITH, high performance arithmetic trap, Imask=00000000, Fmask=00000001, summary=08, PC=0000000000116DFC, PS=0000001B
-SYSTEM-F-FLTOVF, arithmetic trap, floating overflow at PC=0000000000116DFC, PS=0000001B

<register dump snipped>

This was a -des build of perl@11597 with Compaq C V6.4-005 on OpenVMS Alpha 
V7.2-1.  I'm going to build a debug version and see if I can figure out 
where this is happening, but it would help to know what the new expectations 
are surrounding infinity and such.

1) Is the assumption being made that IEEE floats are in use on all platforms?  

2) Is the assumption being made that things like divide by zero and floating 
point overflow are not exceptions or are non-fatal exceptions?  

These two behaviors are independently optional on recent versions of OpenVMS 
Alpha, but neither is the default, and they are not available on OpenVMS 
VAX.  Obviously we need to at least be able to compile the test script even 
if we end up skipping parts of it.



Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About