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

Re: backend abandon

At 03:37 AM 2001-12-31, Howard Chu wrote:
>I couldn't find this discussed before on -devel, so I'll ask (again? I
>vaguely recall discussing this before) now. slapd/abandon.c just sets the
>o_abandon flag for every operation associated with a given connection. There
>is a per-backend abandon function pointer provided in the backendInfo, but
>no slapd code ever references or executes that function. Is this abandon
>pointer obsolete, or just an oversight?

Abandon, from a protocol perspective, is quite broken... and really
needs to be replaced (see draft-zeilenga-ldap-cancel-xx.txt).

Anways, I think the intent is to have a entry point which
the frontend would call upon abandon.  Looks like it wasn't
implemented.  I have no problem with axing it.

>I note that most of the backends have empty stubs for their abandon entry

They really should just have NULL stubs instead of empty stubs...

>but back-shell actually tries to do something meaningful there. Of
>course, none of that code ever gets executed...
>  -- Howard Chu
>  Chief Architect, Symas Corp.       Director, Highland Sun
>  http://www.symas.com               http://highlandsun.com/hyc
>  Symas: Premier OpenSource Development and Support