Subject: Re: IEEE math on Alpha
To: None <email@example.com>
From: Andrei A. Dergatchev <A.Dergatchev@tn.utwente.nl>
Date: 12/31/2000 19:21:45
Hi, A Happy New Year :-) !
> again i'm working to get PostgreSQL to pass on all regression tests on the
> Alpha. however, i'm running into this problem with NetBSD failing on 'inf'
> and 'nan' with SIGFPE (kernel reports user traps). if compiled with -mieee
> the attached code will cause the user trap.
> i've taken the -mieee flag out of the Makefile in pgsql, and
> recompiled, but the program keeps dumping user trap errors and it's
> always something related to NAN. any suggestion is appreciated.
Indeed just a suggestion here.
fp traps are a notorious issue on Alpha. Very few another issues
caused so much traffic in Alpha lists, such as axp-redhat. I hope
you'll find something useful if you'll browse their archive, at
If I'm correct, it requires a special functionality of the kernel
to handle them properly, which is *not always present*. Indeed,
I recall quite a few discussions about code which caused traps
even with -mieee. Is I recall correctly, there were ways and successful
stories to deal with them, so you may find it useful to look it up.