Re: [PATCH] Potentially controversial BDB backend pool usage changes gstein will probably fuss about

Greg Stein wrote:
> Ha!
>
> The patch looks good except that a few of the scratch_pool values are
> not destroyed.

Ooh, good catch.

> Your second approach, of passing body functions a result_pool, might
> be nice to add as an option for future code. retry_txn2() would take a
> result_pool, construct its own scratch_pool (as it does today), and
> pass both into the body function (imo, as two args rather than one arg
> plus one trail member). As you work on code, you could selectively
> switch to txn2() usage for clarity and finer-grained memory usage.

Okay. I'll fix the missing scratch_pool removals, and commit. And then
I'll file an issue about the _txn2() improved API. Thanks for the review, man!

-- 
C. Michael Pilato <cmpilato_at_collab.net>
CollabNet   <>   www.collab.net   <>   Distributed Development On Demand
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=1615465