Quoting Kumar Valsan, Prathap (2019-10-24 00:26:06) > On Wed, Oct 23, 2019 at 01:21:51PM +0100, Chris Wilson wrote: > > + > > + /* > > + * If the user has disabled hangchecking, we can not be sure that > > + * the batches will ever complete after the context is closed, > > + * keeping the context and all resources pinned forever. So in this > > + * case we opt to forcibly kill off all remaining requests on > > + * context close. > > + */ > > + if (!i915_modparams.enable_hangcheck) > > + kill_context(ctx); > > Why not killing the context always when a context is closed? Because we historically have not and so desktop userspace has come to depend on that behaviour (think one client handing over a framebuffer to the display server with pending rendering). > When hang_check is enabled, how would it know the context is closed and > we should release its resources, unless and untill the context has > hanged? Exactly. The contexts persist until complete. The same dos prevention rules apply to the outstanding work as applied when the context was open. -Chris _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
