On Mon, Oct 17, 2016 at 05:44:48PM -0200, Gustavo Padovan wrote:
> 2016-10-17 Chris Wilson :
>
> > On Mon, Oct 17, 2016 at 02:59:52PM -0400, Rob Clark wrote:
> > > On Mon, Oct 17, 2016 at 2:52 PM, Gustavo Padovan
> > > wrote:
> > > > 2016-10-17 Rob Clark :
> > > >
> > > >> Currently with fence-a
On Tue, Oct 18, 2016 at 3:41 AM, Daniel Vetter wrote:
> On Mon, Oct 17, 2016 at 05:44:48PM -0200, Gustavo Padovan wrote:
>> 2016-10-17 Chris Wilson :
>>
>> > On Mon, Oct 17, 2016 at 02:59:52PM -0400, Rob Clark wrote:
>> > > On Mon, Oct 17, 2016 at 2:52 PM, Gustavo Padovan > > > padovan.org> wrote:
On Mon, Oct 17, 2016 at 02:59:52PM -0400, Rob Clark wrote:
> On Mon, Oct 17, 2016 at 2:52 PM, Gustavo Padovan
> wrote:
> > 2016-10-17 Rob Clark :
> >
> >> Currently with fence-array, we have a potential deadlock situation. If we
> >> fence_add_callback() on an array-fence, the array-fence's lock
2016-10-17 Chris Wilson :
> On Mon, Oct 17, 2016 at 02:59:52PM -0400, Rob Clark wrote:
> > On Mon, Oct 17, 2016 at 2:52 PM, Gustavo Padovan
> > wrote:
> > > 2016-10-17 Rob Clark :
> > >
> > >> Currently with fence-array, we have a potential deadlock situation. If
> > >> we
> > >> fence_add_cal
2016-10-17 Rob Clark :
> On Mon, Oct 17, 2016 at 2:52 PM, Gustavo Padovan
> wrote:
> > 2016-10-17 Rob Clark :
> >
> >> Currently with fence-array, we have a potential deadlock situation. If we
> >> fence_add_callback() on an array-fence, the array-fence's lock is acquired
> >> first, and in it'
2016-10-17 Rob Clark :
> Currently with fence-array, we have a potential deadlock situation. If we
> fence_add_callback() on an array-fence, the array-fence's lock is acquired
> first, and in it's ->enable_signaling() callback, it will install cb's on
> it's array-member fences, so the array-memb
On Mon, Oct 17, 2016 at 2:52 PM, Gustavo Padovan wrote:
> 2016-10-17 Rob Clark :
>
>> Currently with fence-array, we have a potential deadlock situation. If we
>> fence_add_callback() on an array-fence, the array-fence's lock is acquired
>> first, and in it's ->enable_signaling() callback, it wil
Currently with fence-array, we have a potential deadlock situation. If we
fence_add_callback() on an array-fence, the array-fence's lock is acquired
first, and in it's ->enable_signaling() callback, it will install cb's on
it's array-member fences, so the array-member's lock is acquired second.
B
