Re: err_exit/err_return regression

[Bart Schaefer <schaefer@xxxxxxxxxxxxxxxx>]

On Oct 3,  5:26pm, Bart Schaefer wrote:
}
} I'm still not entirely happy with [36707], in case anyone has any
} better ideas.
} 
} [36766] seems [to fix] an obvious former thinko, in retrospect.  All
} the regular tests still pass.  Anyone see a problem?

In fact with 36766 in place, 36707 isn't necessary any longer, so that
seems to have been the root of the problem all along.

I'll back out 36707.