[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: [Libevent-users] Fwd: leaking munmap



On Tue, Jul 30, 2013 at 11:09 AM, Nick Mathewson <nickm@xxxxxxxxxxxxx> wrote:
 [...]
> I think that right now, the best approach is probably to merge the
> obvious patch and move on.  Anybody feeling like writing that up as in
> "git format-patch" format (or as a branch I can merge somewhere), or
> should I stick it on my queue?

I've merged something as d4095146aff755f23df579cf0a5a376480ef2b5d; it
appears to work for me, though other examination would rock.

I'm skipping the &(page_size-1) trick and the sysconf cacheing for
now; I'd be glad to apply them if there's evidence showing that they
matter.  (I would be a bit surprised if you could measure a
significant overhead of sysconf and % next to the mmap() and munmap()
calls -- but I am surprised on a regular basis, so who knows.)

best wishes, and thanks again,
-- 
Nick
***********************************************************************
To unsubscribe, send an e-mail to majordomo@xxxxxxxxxxxxx with
unsubscribe libevent-users    in the body.