Page MenuHomeFreeBSD
Differential D32360 Diff 96932 sys/powerpc/powerpc/vm_machdep.c

Changeset View

Standalone View

View Options

sys/powerpc/powerpc/vm_machdep.c

Show First 20 Lines • Show All 262 Lines • ▼ Show 20 Lines
int int
cpu_procctl(struct thread *td __unused, int idtype __unused, id_t id __unused, cpu_procctl(struct thread *td __unused, int idtype __unused, id_t id __unused,
int com __unused, void *data __unused) int com __unused, void *data __unused)
{ {
return (EINVAL); return (EINVAL);
} }
void
cpu_core_sync(void)
{
isync();
jhibbitsUnsubmitted
Done
Inline Actions

Since this is only used in the smp_rendezvous case, never called directly, I think it can be a nop. 'rfi' and its many variants is a context synchronizing instruction.

jhibbits: Since this is only used in the smp_rendezvous case, never called directly, I think it can be a…
kibAuthorUnsubmitted
Done
Inline Actions

I did not rechecked it, but most likely it is what Linux does. Do you mean that our rendezvous on ppc already does isync or stronger barrier?

kib: I did not rechecked it, but most likely it is what Linux does. Do you mean that our rendezvous…
jhibbitsUnsubmitted
Done
Inline Actions

Yes. Rendezvous does the following:

atomic_add_acq_int() (add+isync)
action
atomic_add_rel_int(lwsync or sync + add)

So the rendezvous is already a heavyweight barrier on its own.

jhibbits: Yes. Rendezvous does the following: atomic_add_acq_int() (add+isync) action…
}