On Mon, Jun 04, 2001 at 10:13:30AM -0700, Ian Holsman wrote:
> couldn't we have it so that the 'sub-handlers' request pool is joined with/the same
>as the main request's pool,
> (this is different to the 'connection' pool right?)
>
> so that sub-requests live for the life of the request...
> It looks like that is what the function apr_pool_join does in 'debug' mode
No... The whole point of having a subrequest pool is so that we can trash it
when the subrequest is done. If a request ran 1000 subrequests (it can
happen with a large directory processed by mod_autoindex), then you would
end up with a HUGE waste of memory in the request pool.
Passing a pool to setaside() should allow us to migrate a bucket from one
pool/lifetime (the subrequest) to another pool/lifetime (the request or the
connection depending on who does a setaside and where they want to put it).
Cheers,
-g
--
Greg Stein, http://www.lyra.org/
