On Mon, Apr 20, 2015 at 09:07:28AM -0700, Yu Dai wrote: > > > On 04/18/2015 06:48 AM, Chris Wilson wrote: > >On Fri, Apr 17, 2015 at 02:21:13PM -0700, yu.dai@xxxxxxxxx wrote: > >> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c > >> index de8c074..8f13e80 100644 > >> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c > >> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c > >> @@ -1993,7 +1993,7 @@ static int intel_init_ring_buffer(struct drm_device *dev, > >> INIT_LIST_HEAD(&ring->request_list); > >> INIT_LIST_HEAD(&ring->execlist_queue); > >> i915_gem_batch_pool_init(dev, &ring->batch_pool); > >> - ringbuf->size = 32 * PAGE_SIZE; > >> + ringbuf->size = 4 * PAGE_SIZE; > >> ringbuf->ring = ring; > >> memset(ring->semaphore.sync_seqno, 0, sizeof(ring->semaphore.sync_seqno)); > > > >NAK. > > > First of all, GuC firmware reserves limited bits for ring buffer. 4 > pages is max for now. Second, considering the ring buffer is > per-context now, there is no need to allocate 32 pages for it. Please look at which function you are changing and explain how this is not in the least broken. -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx