Re: [PATCH] fix storage leak on persistent connection

[Jeff Trawick <tr...@attglobal.net>]

Jeff Trawick <tr...@attglobal.net> writes:

> The core output filter uses c->pool to store a brigade representing
> response data from a request which won't be sent for a while.  The
> storage for that temporary brigade is not completely freed until
> c->pool is cleaned up.  This represents a storage leak which will
> become a problem if a bazillion requests are issued on that
> connection.

I'll plan on committing this after the next tarball.  Is anybody
opposed?  Does anybody mind if I don't try to keep up with whether or
not anything is in the subpool and simply call apr_pool_clear() in the
normal exit path (i.e., get rid of has_stuff)?  I'm afraid that future
updates to the function may not do the right thing w.r.t. has_stuff.

-- 
Jeff Trawick | trawick@attglobal.net | PGP public key at web site:
       http://www.geocities.com/SiliconValley/Park/9289/
             Born in Roswell... married an alien...