lkml.org 
[lkml]   [2020]   [Aug]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH v3] mm: Fix kthread_use_mm() vs TLB invalidate
On Fri, Aug 21, 2020 at 11:09:51AM +0530, Aneesh Kumar K.V wrote:
> Peter Zijlstra <peterz@infradead.org> writes:
>
> > For SMP systems using IPI based TLB invalidation, looking at
> > current->active_mm is entirely reasonable. This then presents the
> > following race condition:
> >
> >
> > CPU0 CPU1
> >
> > flush_tlb_mm(mm) use_mm(mm)
> > <send-IPI>
> > tsk->active_mm = mm;
> > <IPI>
> > if (tsk->active_mm == mm)
> > // flush TLBs
> > </IPI>
> > switch_mm(old_mm,mm,tsk);
> >
> >
> > Where it is possible the IPI flushed the TLBs for @old_mm, not @mm,
> > because the IPI lands before we actually switched.
> >
> > Avoid this by disabling IRQs across changing ->active_mm and
> > switch_mm().
> >
> > [ There are all sorts of reasons this might be harmless for various
> > architecture specific reasons, but best not leave the door open at
> > all. ]
>
>
> Do we have similar race with exec_mmap()? I am looking at exec_mmap()
> runnning parallel to do_exit_flush_lazy_tlb(). We can get
>
> if (current->active_mm == mm) {
>
> true and if we don't disable irq around updating tsk->mm/active_mm we
> can end up doing mmdrop on wrong mm?

exec_mmap() is called after de_thread(), there should not be any mm
specific invalidations around I think.

Then again, CLONE_VM without CLONE_THREAD might still be possible, so
yeah, we probably want IRQs disabled there too, just for consistency and
general paranoia if nothing else.

\
 
 \ /
  Last update: 2020-08-21 15:05    [W:0.140 / U:0.524 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site