[racket-dev] Pre-Release Checklist for v5.2.1, second call

From: Matthew Flatt (mflatt at cs.utah.edu)
Date: Mon Jan 23 13:11:12 EST 2012

At Mon, 23 Jan 2012 02:16:06 +0400, Aleksej Saushev wrote:
> I'm testing racket-5.2.0.901 on NetBSD 5.99.59 i386.
> 
> After applying patch as attached, 

Thanks!

> Section(deep)
> [1]   Illegal instruction (core dumped) (cd collects/tests/racket && racket -f 
> quiet.rktl)
> 
> [...]
> 
> Any idea what can be wrong?

This looks like a problem with Racket's inference of the stack size.

In some BSD variants, enable pthread support makes the main thread have
a different size than the usual stack limit. Did you specifically
enable places, futures, or pthread when configuring? (I think they're
off by default for NetBSD, but I may be mistaken.) Does configuring so
that `-pthread' isn't used in the compiler command change the success
of the "deep.rktl" test?


Matthew


Posted on the dev mailing list.