Re: [Chicken-users] Segfault - a hard one

[F. Wittenberger]

Am Donnerstag, den 07.08.2008, 12:03 -0400 schrieb John Cowan:
> Vincent Manis scripsit:
> 
> > I'd prefer to have the manual document that to the best of our knowledge
> > thread-terminate works `correctly', but that its use is fraught with
> > design and testing problems, and therefore it should be avoided.
> 
> I'd go even further:  KILL THE BEAST.

Do whatever you feel appropriate.

;-)

Since chicken has thread-signal! [now ;-) when some hacker, who's well
known on this list, suggested to use thread-terminate! to limit the real
time a thunk takes, it had not]  Hence I do not depend on
thread-terminate! anymore.

Though I know that the program I'm porting works fine, if
thread-terminate! does (at least) abandon all mutex's the thread holds
and removes it properly from any queue/port it might be waiting on etc.
I'm not yet sure, that it still works fine, when those threads get the
chance to properly clean up after themself - after all, they could not,
while the Scheme system easily can [though this is not trivial].

Nevertheless, I would not remove thread-terminate!  Maybe repeat in the
manual those warnings from SRFI-18 that this is not exactly a nice call
to make.  But why support half a srfi?

Furthermore: I do understand those issues about thread being left in
inconsistent state (not that bad when they are terminated anyway) and
worse inconsistent data structures.

Looking at compatibility issues, I can't see a SRFI-18 compatible way to
interrupt a thead except thread-terminate! .  That's basically the
challenge I'm facing: I have some user supplied code (which, being user
supplied, might endless loop), which I want to run for a limited amount
of time until I decide to terminate the computation.  How would I do
that?

}}8-[