[plt-scheme] Interesting Seg-Fault Interfacing With C Code

From: Will M Farr (farr at MIT.EDU)
Date: Tue Mar 24 13:08:07 EDT 2009


On Mar 24, 2009, at 12:46 PM, Matthew Flatt wrote:

> Did
> you try setting a breakpoint on garbage_collect() to see if it really
> is running during a qsort()?

No.  It's probably a good idea, but unfortunately, the qsort occupies  
something smaller than 1% of the computation time in the hours-long  
running of this program.  I would like to try this, but it's probably  
impractical for me to break on every gc, check the stack, and then  
continue when qsort doesn't appear for an hour or two.

Sorry not to be able to help pin this down more.  If I can find a  
quicker test-case (the old code would only break after an hour or two  
of running), I'll submit it as a true bug-report for you guys.

Thanks for the comments,

Posted on the users mailing list.