On 10 April 2015 at 19:03, Jacob, Jerin <[email protected]> wrote:
> > I have query on ODP timer, > > Who(application/implementation) is responsible to free the odp_timeout > event object created by > "odp_timer_set_*" API by specifying tmo_ev == NULL ? > The timer implementation does not allocate or free any timeout events. This is the responsibility of the application. > IMO, It should be freed by the application by calling "odp_timeout_free" > Yes. The idea is that all timeout management (alloc and free) is done by the user. > > But it looks like linux generic implementation is freeing "implementation" > created > odp_timeout event by "odp_timer_free". Is it true ? > No I don't think so. odp_timer_free() returns any existing timeout event (i.e. when (cancelling and) freeing an active timer) to the caller. What makes you think odp_timer_free() frees the timeout event? Perhaps the documentation can be improved if the code is not clear enough. > If its true then It will lead to "double free" case when the system has > both implementation and application > created odp_timeout objects armed to timer as there is no way distinguish > between them. > > /Jerin > > > > > > > _______________________________________________ > lng-odp mailing list > [email protected] > https://lists.linaro.org/mailman/listinfo/lng-odp >
_______________________________________________ lng-odp mailing list [email protected] https://lists.linaro.org/mailman/listinfo/lng-odp
