Re: auth_user_hash_pointer leak? (2.5 Bug #910)

From: Henrik Nordstrom <hno@dont-contact.us>
Date: Thu, 19 Feb 2004 11:50:58 +0100 (CET)

On Thu, 19 Feb 2004, Robert Collins wrote:

> On Thu, 2004-02-19 at 03:04, Henrik Nordstrom wrote:
>
> challenge-response caching: if the challenge given by the helper is the
> same, and the response is the same, it's a valid login a priori.

Ok. The use of this then obviously should be disabled if challenge-reuse
is disabled.

> It will be growing a lot because it's not currently trimmed before the
> user details are purged, and the new code will be never hitting.
>
> My opinion: blow it away.

For challenge-reuse configurations it is still good but needs to purge old
entries when the challenge is changed. Will look into this.

Regards
Henrik
Received on Thu Feb 19 2004 - 03:51:03 MST

This archive was generated by hypermail pre-2.1.9 : Mon Mar 01 2004 - 12:00:04 MST