[Swan-dev] dpd vs liveness code suggests possibly missing code for liveness

[Paul Wouters]

I was cleaning up some dpd.h includes while trying to clean up some
ikev1 vs ikev2 code. I noticed a difference in the use of
delete_dpd_event() and delete_liveness_event():

paul at thinkpad:~/git/libreswan/programs/pluto (master *)$ grep
delete_dpd_event *.c
dpd.c:				delete_dpd_event(st);
dpd.c:			delete_dpd_event(p1st);
dpd.c:			delete_dpd_event(p1st);
dpd.c:				delete_dpd_event(p1st);
dpd.c:		delete_dpd_event(p1st);
dpd.c:		delete_dpd_event(p1st);
state.c:		delete_dpd_event(st);
state.c:	delete_dpd_event(this);
state.c:				delete_dpd_event(st);
timer.c:		delete_dpd_event(st);
timer.c:void attributed_delete_dpd_event(struct state *st, const char *file, int lineno)

paul at thinkpad:~/git/libreswan/programs/pluto (master *)$ grep delete_liveness_event *.c
timer.c:static void delete_liveness_event(struct state *st);	/* forward */
timer.c:		delete_liveness_event(st);
timer.c:		delete_liveness_event(st);
timer.c:static void delete_liveness_event(struct state *st)

I think that indicates some things that still need to be added? For
example, the delete_dpd_event() calls in state.c might also need a
delete_liveness_event() call?

I noticed mostly because delete_liveness_event() is static to timer.c but
delete_dpd_event() cannot be static (or just be limited to dpd.c and
timer.c)

Matt, let me know if you can have a look at this, since you've been
active around this code in the last few weeks. If not, I'll have a look
at it next week.

Paul