[Date Prev][Date Next] [Chronological] [Thread] [Top]

RE: Thread-local malloc discussion summary



The only way we can fix this without breaking API signatures is to
change the semantics of the computed attribute search rewriter so
that the front-end, rather than the plugin, is responsible for
freeing the original search filter. Obviously this is problematic
because only one plugin can rewrite the search filter without
leaking memory.

Another alternative would be to disable the thread-local malloc
entirely if SLAPI was enabled. We would prefer this. Comments?

-- Luke

>From: Luke Howard <lukeh@PADL.COM>
>Subject: RE: Thread-local malloc discussion summary
>To: lukeh@PADL.COM
>Cc: hyc@highlandsun.com, jongchoi@us.ibm.com, openldap-devel@OpenLDAP.org
>Date: Fri, 11 Apr 2003 12:57:19 +1000
>Organization: PADL Software Pty Ltd
>Versions: dmail (bsd44) 2.4c/makemail 2.9d
>
>
>The new thread-local malloc blows up SLAPI filter rewriters pretty
>badly. Hopefully there's a way to get this to work without breaking
>the API; I'm looking into it.
>
>-- Luke
>
>--
>Luke Howard | PADL Software Pty Ltd | www.padl.com

--
Luke Howard | PADL Software Pty Ltd | www.padl.com