AnsweredAssumed Answered

VRf-Timeslicing_Userfunctions

Question asked by VRFuser on May 4, 1997
send q VRf-Timeslicing_Userfunctions ww vrf kennedy_d PDereux@aol.com

from: Greg Goebel / HP-MXD
      gvg@lvld.hp.com / 970-679-3030 / FAX 970-679-5971
to:   P-E Dereux
date: Monday, 05 May 1997 1029 MDT

> From: PDereux@aol.com
> To: hpvxd_xc@hpislsup.lvld.hp.com
>
> Hye !
>
> Some comments about point 5 (objects running simultaneously in VEE 4.00)
> and 1 question about Vee 4.0 :
>
> I've not yet tried Vee 4.00, but what I can tell some about Vee 3.21 and
> multithreads.
> In Vee 3.21, if you have several threads running simultaneously, and each
> calling the same user function at the same time, these function calls will be
> executed sequentially, on call after this other. And I hope it's unchanged in
> Vee 4.00.
> IS IT STILL TRUE IN VEE 4.0???

Yes, it is still true, though only if you call the *same* UserFunction.  They
are not reentrant.

However, if you are executing *different* UserFunctions, they *will*
timeslice.  This was not true in older versions of VEE.

This turns out to affect a few people in a negative fashion.  We are likely
to incorporate a flag in a future version of VEE that allows you to select
timeslicing or "atomic" operation.

> Note it's very usefull for "real time" features : if this function (and
> others ...)
> reads / writes a data, you could imagine the mess if all the function calls
> try to read /write the same data in the "same time" ...
> Hope this will help.
>
> P-E Dereux
> E S Lab
> Bell Helicopter Textron Canada

If you have more questions, let us know.

[<>] regards -- gvg

Outcomes