Sander Striker wrote:
So, actually, there isn't much difference between the plain apr_pcalloc implementation and the apr_pcalloc macro? (other than code duplication)
The real solution seems to be eliminating apr_pcalloc calls when possible (and sensible). Maybe we can have another quantify run to identify the most prominent left? Brian?
I've done some profiling recently, and there aren't any obvious candidates for apr_pcalloc elimination left. We've fixed the big ones already, and the cases that remain are apr_pcallocs of relatively large structures where only a small subset of the fields are overwritten right after the pcalloc. The creation of request_recs is a good example.
--Brian
