Re: segfault in completion for configure

[Bart Schaefer <schaefer@xxxxxxxxxxxxxxxx>]

On Mar 21, 10:56am, Peter Stephenson wrote:
}
} Are you simply worried that a pattern match that happens
} asynchronously might not leave behind $MATCH / $match etc. values
} after return owing to other asynchronous goings on with pattern
} matching?

The point is that a pattern match that happens asynchronously might not
leave behind $MATCH / $match etc. values owing to other pattern matches
going on, period, whether synchronous or async.  Two signal handlers in
the example was just to set up a way to get into the situation; a glob
that takes several seconds to complete might just as easily do it, but
is hard to invent for purposes of example code.

I'm the tiniest bit worried that there's a way $MATCH et al. might be
left pointing at garbage memory, but I don't really think so.

} I don't think that's an issue.

I don't think it's an issue either, but it's going to do the potentially
unexpected thing more consistently now, so I wondered if it should be
documented.