[Info-vax] New VSI Roadmap (yipee!)

[David Froble]

Stephen Hoffman wrote:

> It'd also be interesting to instrument and figure out why the code was 
> so slow, and if there was a way to optimize it.   David: tried checking 
> for alignment faults (if Itanium) or other nasties?

It's been too long, I don't remember the details.

Randy Park (XENTIS developer) had a solution for the overhead of calling 
a subroutine.  I think it was the compiler /NOCHECK option.  The result 
was errors were not trapped, but the overhead for error trapping was not 
present, which reduced the overhead.

I personally never used the option.

But I'm aware that it's subroutine calling overhead.  Bill Judy has 
tested in one program with in-line code, and it changed the run time 
from 8 hours to 7 hours.  It's still doing the math, just not the 
subroutine call.

Regardless, what's a language doing, not having a rounding function?