Re: Error reporting in ikev2_ike_sa_alive (was: Improve error reporting in pfkey_sa_last_used)

2020-06-03 Thread matthew j weaver
On Sun, May 31, 2020, at 17:27, Tobias Heider wrote: > I don't think this is a good idea > With your diff the log gets spammed with 'Undefined error: 0' for child SAs > that have never been used. > Also log_warn seems a bit too much as those errors are rarely serious. Thank you for having a

Re: Error reporting in ikev2_ike_sa_alive (was: Improve error reporting in pfkey_sa_last_used)

2020-05-31 Thread Tobias Heider
On Tue, May 26, 2020 at 12:08:08PM -0400, matthew j weaver wrote: > During childsa last use checks, iked debug logs results, per SA, after a > successful pfkey_sa_last_used call. > > This patch makes logging behavior more closely match that, on error. > > I chose log_warn instead of log_debug

Error reporting in ikev2_ike_sa_alive (was: Improve error reporting in pfkey_sa_last_used)

2020-05-26 Thread matthew j weaver
During childsa last use checks, iked debug logs results, per SA, after a successful pfkey_sa_last_used call. This patch makes logging behavior more closely match that, on error. I chose log_warn instead of log_debug since iked will complain about the nonzero errno after pfkey_reply: