Hi,
> - doing this over dma-buf means we can only use this for the ww_mutx
> dance, not for anything else. Which means we need another layer on top
> for shared execbuf utils (which Gerd has started looking into, but I
> felt like not quite the right approach yet in his first draft series)
On Wed, Oct 09, 2019 at 03:21:06PM +0200, Christian König wrote:
> Am 08.10.19 um 17:16 schrieb Daniel Vetter:
> > On Wed, Sep 18, 2019 at 07:55:23PM +0200, Christian König wrote:
> > > The ww_mutex framework allows for detecting deadlocks when multiple
> > > threads try to acquire the same set of
Am 08.10.19 um 17:16 schrieb Daniel Vetter:
On Wed, Sep 18, 2019 at 07:55:23PM +0200, Christian König wrote:
The ww_mutex framework allows for detecting deadlocks when multiple
threads try to acquire the same set of locks in different order.
The problem is that handling those deadlocks was the
On Wed, Sep 18, 2019 at 07:55:23PM +0200, Christian König wrote:
> The ww_mutex framework allows for detecting deadlocks when multiple
> threads try to acquire the same set of locks in different order.
>
> The problem is that handling those deadlocks was the burden of the user of
> the ww_mutex im
The ww_mutex framework allows for detecting deadlocks when multiple
threads try to acquire the same set of locks in different order.
The problem is that handling those deadlocks was the burden of the user of
the ww_mutex implementation and at least some users didn't got that right
on the first try