On Fri, May 16, 2025 at 10:33:30AM +0100, Tvrtko Ursulin wrote: > On 24/04/2025 10:55, Philipp Stanner wrote: > > + * @kill_fence_context: kill the fence context belonging to this > > scheduler > > Which fence context would that be? ;)
There's one one per ring and a scheduler instance represents a single ring. So, what should be specified here? > Also, "fence context" would be a new terminology in gpu_scheduler.h API > level. You could call it ->sched_fini() or similar to signify at which point > in the API it gets called and then the fact it takes sched as parameter > would be natural. The driver should tear down the fence context in this callback, not the while scheduler. ->sched_fini() would hence be misleading. > We also probably want some commentary on the topic of indefinite (or very > long at least) blocking a thread exit / SIGINT/TERM/KILL time. You mean in case the driver does implement the callback, but does *not* properly tear down the fence context? So, you ask for describing potential consequences of drivers having bugs in the implementation of the callback? Or something else? > Is the idea to let drivers shoot themselves in the foot or what? Please abstain from such rhetorical questions, that's not a good way of having technical discussions. - Danilo
