Messages in this thread | | | Date | Wed, 4 Sep 2019 14:03:37 +0200 | From | Oleg Nesterov <> | Subject | Re: [RFC PATCH 1/2] Fix: sched/membarrier: p->mm->membarrier_state racy load |
| |
On 09/04, Peter Zijlstra wrote: > > + struct task_struct *g, *t; > + > + read_lock(&tasklist_lock); > + do_each_thread(g, t) {
for_each_process_thread() looks better
> + if (t->mm == mm) { > + atomic_or(MEMBARRIER_STATE_GLOBAL_EXPEDITED, > + &t->membarrier_state); > + }
then you also need to change dup_task_struct(), it should clear ->membarrier_state unless CLONE_VM.
And probably unuse_mm() should clear current->membarrier_state too.
Hmm. And it can race with copy_process() anyway, tasklist_lock can't really help. So copy_process() needs to do
write_lock_irq(&tasklist_lock); ...
if (clone_flags & CLONE_VM) p->membarrier_state = current->membarrier_state; else p->membarrier_state = 0;
Oleg.
| |