On Thu, 2025-10-02 at 11:14 +0200, Nam Cao wrote:
> Gabriele Monaco <[email protected]> writes:
> > +/*
> > + * da_get_task - return the task associated to the monitor
> > + */
> > +static inline struct task_struct *da_get_task(struct da_monitor *da_mon)
> > +{
> > + return container_of(da_mon, struct task_struct,
> > rv[task_mon_slot].da_mon);
> > +}
>
> This function is not used? Maybe later in the series, let me keep looking..
Right, this doesn't belong here, I should add it later..
>
> > static inline bool da_handle_start_event(enum events event)
> > {
> > - struct da_monitor *da_mon;
> > -
> > if (!da_monitor_enabled())
> > return 0;
>
> Can't this part be shared between different monitor types?
>
> Same for the other handle functions.
Mmh good point, I left it separate because the per-object monitors (later in the
series) do a bit more before calling the __da_handle_* functions (like some
potential allocation).
But it's probably not a big deal to let them do that in any case, considering
da_monitor_enabled() is false only during teardown or with monitoring off, and
in neither case a superfluous allocation should cause problems.
Thanks,
Gabriele
