Hi Thomas, > Especially when working with core dumps it proved useful to have > charon's log messages prefixed with the thread id instead of the > human readable thread number.
Changing the current human-readable thread number to the actual thread ID makes the logs quite a bit harder to read. Not only are the IDs longer than the numbers, their range also varies significantly between individual runs. But I definitely see the benefit of knowing those IDs when debugging code. There are several possible ways to implement this. We could, for instance, log the thread ID only when a new thread is added to the thread pool, this would then allow to map the number to the ID if needed. Or we could add a new setting to strongswan.conf, which would allow users to configure whether the thread number or the thread ID is used during logging (defaulting to the number). The question is essentially whether it is required that the actual thread ID is logged at the beginning of every line or if listing them once during startup is enough. @Andreas, @Martin: What are your thoughts on this? Regards, Tobias _______________________________________________ Dev mailing list [email protected] https://lists.strongswan.org/mailman/listinfo/dev
