On Wed, Mar 11, 2015 at 09:18:19PM +0000, Chris Wilson wrote: > Arguably busy-spinning on an idle system isn't totally evil, but it > certainly is likely to come at a power cost. On the other hand, spinning > is relatively rare outside of benchmarks. Rare enough to be useful? As a counterpoint, I plugged in the qm_pos trick we use for lowlatency dma interrupts (gmbus, dp-aux) and found no improvement at all over a plain interrupt driven __i915_wait_request. -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
