Hello,

"Linas Vepstas" <[EMAIL PROTECTED]> writes:

> I've been seeing all sorts of deadlocks in guile, and so I wrote a small
> debugging utility to try to track down the problems.  I'd like to see
> this patch included in future versions of guile.

FWIW, I'm reluctant to committing this patch for various reasons:
intrusion, non-genericity, bitrot of conditional code as Han-Wen noted,
non-C89ness, pollution outside of the `scm_' namespace, incorrect style.

Another reason is that pthread libraries are sometimes able to debug
deadlocks by themselves.  NetBSD's performs several run-time checks by
default [0].  Another solution is the explicit use of error-checking
mutexes, as with the patch at [1].  What do you think?

Thanks,
Ludo'.

[0] http://thread.gmane.org/gmane.lisp.guile.devel/6656
[1] http://thread.gmane.org/gmane.lisp.guile.devel/6733



Reply via email to