[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: Bug in passwd program

mark@mmi.netconnect.com.au wrote:
> I've not been too current on the PAM list recently so please excuse me if
> this is an already known problem and also for not looking at the source
> myself (I'm a bit busy with my own source at the moment!)... or if it's
> my dodgy programming that's the problem! ;-)   However if I have a lock
> on a file it shouldn't matter how long I hold it, the lock should be
> valid until I unlock it or the holding process terminates... IMO.

The rationale for this is as follows:

	The lock is to prevent two people doing something at the same time

	The something should not take too long. (less than a few secs!)

	Blocking indefinitely leads to very simple denial of service

The compromise is to not block indefinitely...

Ideally, all security sensitive edit-actions should be atomic and short.
Non-blocking locks with timeouts are an alternative that "seems to work".

This is not thought of as a "bug" but a "design feature"..

Hope that helped,

               Linux-PAM, libpwdb, Orange-Linux and Linux-GSS
       [ For those that prefer FTP  ---  ftp://ftp.lalug.org/morgan ]

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index] []