Quoting Chris Wilson (2019-01-31 16:20:37) > Quoting Chris Wilson (2019-01-31 13:39:50) > > Quoting Tvrtko Ursulin (2019-01-31 13:19:31) > > > > > > On 30/01/2019 02:19, Chris Wilson wrote: > > > > Having introduced per-context seqno, we now have a means to identity > > > > progress across the system without feel of rollback as befell the > > > > global_seqno. That is we can program a MI_SEMAPHORE_WAIT operation in > > > > advance of submission safe in the knowledge that our target seqno and > > > > address is stable. > > > > > > > > However, since we are telling the GPU to busy-spin on the target address > > > > until it matches the signaling seqno, we only want to do so when we are > > > > sure that busy-spin will be completed quickly. To achieve this we only > > > > submit the request to HW once the signaler is itself executing (modulo > > > > preemption causing us to wait longer), and we only do so for default and > > > > above priority requests (so that idle priority tasks never themselves > > > > hog the GPU waiting for others). > > > > > > It could be milliseconds though. I think apart from media-bench saying > > > this is faster, we would need to look at performance per Watt as well. > > > > All throughput measurements are substantially faster, as you would > > expect, and inter-engine latency decreased. I would hope it would > > powergate/rc6 the EU while the CS was spinning, but I don't know :) > > Fwiw, it's about the power cost of simply spinning with the CS without > any additional cost of utilizing the engine. Another interesting data point is that the *total* energy consumed for a latency bound test that passes a piece of work from one engine to the next reduced by 30%, with a speed increase of nearly 100% (on glk). -Chris _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
