EFN$C_ENF Chalk or Cheese?
- From: "Richard Maher" <maher_rj@xxxxxxxxxxxxxxxxxx>
- Date: Sat, 25 Mar 2006 12:29:10 +0800
Hi,
How does EFN$C_ENF work?
My guess is it fires whenever *any* of the 128 event flags are set.
(Otherwise it would be just too clever for words and something to be weary
of. Like yeast or pool-cleaners/creepy-crawlies!) But you'll never get a
false negative so it is robust.
Now, if I do a $synch with efn$c_enf and my_iosb then CPU (and perhaps
paging) will be burned every time any event flag is set in the process? But
if I took the default event flag zero then there would now be conflict and
(judging by the docs) a spurious clear can lead to an erroneous wait state.
So, for my money we should all be using LIB$GET_EF whenever and wherever
possible. Or is the miniscule amount of resource wastage (especially in a
single-threaded application) not woth worrying about. Use efn$c_enf, protect
your iosb/lksb and you're laughing.
Or none of the above?
Regards Richard Maher
.
- Follow-Ups:
- Re: EFN$C_ENF Chalk or Cheese?
- From:
- Re: EFN$C_ENF Chalk or Cheese?
- From: Hein RMS van den Heuvel
- Re: EFN$C_ENF Chalk or Cheese?
- Prev by Date: Re: Is VMS Security being dumbed-down for Java?
- Next by Date: Just Ask The Vendor: Securing The Endpoint
- Previous by thread: Bootable disk copy of CD
- Next by thread: Re: EFN$C_ENF Chalk or Cheese?
- Index(es):