[Home] [Groups] - Message: [Prev in Group] [Next in Group]

nu.kanga.list.mud-dev

12182: Re: [MUD-Dev] Two threads forced to one CPU? (was: Collecting ideas for a MUD server...)

[Full Header] [Plain Text]
From: Hans-Henrik Staerfeldt <hhs@cbs.dtu.dk>
Newsgroups: nu.kanga.list.mud-dev
Date: Tue, 28 Dec 1999 10:22:02 +0100 (MET)
References: [1]
Organization: Kanga.Nu
On Fri, 24 Dec 1999 cg@ami-cg.GraySage.Edmonton.AB.CA wrote:

> [Greg Miller:]
>=20
> > BTW: Here's something else anyone doing MT should be aware of: Your
> > software may have two sets of thread bugs, one for single-CPU executi=
on,
> > and one for MP execution. There may or may not be overlap between the
> > two. That means you *cannot* assume your MT application runs on both,
> > unless you've tested and otherwise verified your code in both
> > situations.
>=20
> Excellent advice! At work, I've worked on threaded software on machines
> with 1 2, and 4 CPU's, all on Windows NT. Bugs show up at all levels of
> that, and some of them can be difficult to figure out, since they can b=
e
> timing dependent. You've got to be extremely careful with any locking,
> etc. that you need. And, you've got to make *no* assumptions about
> things getting done and the timing of them.
>=20

Yes, or keep assuming the worst possible. Assume that the thread can be=20
interrupted, after you have written down 16 of 32 bit of any long word=20
in the memory, making the 32-bit value non-valid, which is why you need
thread-safe system calls :-) I actually had this error, though on older=20
hardware.=20


Hans Henrik St=E6rfeldt   |    bombman@diku.dk    | work:  hhs@cbs.dtu.dk=
      |
address:                |___  +45 40383492    __|__       +45 45252425   =
  __|
 Dybendalsvej 74 2. th, | Scientific programmer at Center for Biological =
    |
 2720 Vanl=F8se, Danmark. |  Sequence Analysis, Technical University of D=
enmark|




_______________________________________________
MUD-Dev maillist  -  MUD-Dev@kanga.nu
http://www.kanga.nu/lists/listinfo/mud-dev