[gclist] Re: [m-dev.] memory leaks & Boehm collector

[Kragen Sitaker]

Someone wrote:
> In fact the problem turned out to be one of those blindingly simple
> stupid mistakes... in the above-mentioned code to zero out the Mercury
> stacks, I had put the arguments to memset() in the wrong order :-(
> Instead of setting a lot of bytes to zero, I set zero bytes.
> Thus the Mercury stacks weren't being cleared at all, which
> explained the leak.

That's terrific, and I'm delighted to hear about the tracing tool.

memset() is almost always used to set things to zero.  Why, oh why,
couldn't they have standardized bzero() instead?  One less parameter,
one less potential bug to track down.

-- 
<kragen@pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
Mon Oct 04 1999
36 days until the Internet stock bubble bursts on Monday, 1999-11-08.
<URL:http://www.pobox.com/~kragen/bubble.html>