Quoting Joonas Lahtinen (2018-06-13 13:54:01) > Quoting Chris Wilson (2018-06-12 13:51:35) > > For symmetry, simplicity and ensuring the request is always truly idle > > upon its completion, always emit the closing flush prior to emitting the > > request breadcrumb. Previously, we would only emit the flush if we had > > started a user batch, but this just leaves all the other paths open to > > speculation (do they affect the GPU caches or not?) With mm switching, a > > key requirement is that the GPU is flushed and invalidated before hand, > > so for absolute safety, we want that closing flush be mandatory. > > > > Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> > > No word about perf impact? The patch description matches what it does. There's no change to the normal execution paths. There may be a change on interrupted execution paths, but they are already returning to userspace empty handed and not as relevant to performance. We are just making sure the odd ones don't have any odd side effects. -Chris _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
