[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20070725.171723.45714101.davem@davemloft.net>
Date: Wed, 25 Jul 2007 17:17:23 -0700 (PDT)
From: David Miller <davem@...emloft.net>
To: latten@...tin.ibm.com
Cc: netdev@...r.kernel.org, linux-audit@...hat.com, sgrubb@...hat.com
Subject: Re: [PATCH]: 2nd revision of make xfrm_audit_log more generic
From: Joy Latten <latten@...tin.ibm.com>
Date: Wed, 25 Jul 2007 14:21:43 -0500
> This is 2nd revision of patch to modify xfrm_audit_log() such
> that it can accomodate auditing other ipsec events
> besides add/delete of an SA or SPD entry.
>
> 2nd revision includes new define for all IPsec
> events in audit.h and introduces "op=" entry
> in logfile as well as add a hyphen in description
> for report parsing.
>
> This is a small change to accomodate updating
> ipsec protocol to RFCs 4301, 4302 and 4303 which
> require auditing some ipsec events if auditing
> is available. Please let me know if ok.
>
> Signed-off-by: Joy Latten <latten@...tin.ibm.com>
I like very much how the implementation of xfrm_audit_log() got
simplified.
But _TEN_ function call arguments, good grief!
That's at least twice as many as most cpus can pass in registers.
Let's try an alternative where you have specialized
xfrm_audit_log_foo() routines that take a user policy pointer, or
whatever the main object is.
If internally this just unpacks the needed bits and calls some
do_xfrm_audit_log() thing inside of the auditing code that takes lots
of arguments, that's fine, but let's not expand all of that argument
setup code in the main IPSEC code paths.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists