On Thu, Jan 20, 2011 at 9:38 AM, Chris Wilson <chris at chris-wilson.co.uk> wrote: > >> because from looking at the code, I get the notion that >> "intel_read_status_page()" may not be exact. But what happens if that >> inexact value matches our cached ring->actual_head, so we never even >> try to read the exact case? Does it _stay_ inexact for arbitrarily >> long times? If so, we might wait for the ring to empty forever (well, >> until the timeout - the behavior I see), even though the ring really >> _is_ empty. No? > > Ah. Your analysis is spot on and this will cause a hang whilst polling if > we enter the loop with the last known head the same as the reported value.
So how about just doing this in the loop? It will mean that the _first_ read uses the fast cached one (the common case, hopefully), but then if we loop, we'll use the slow exact one. (cut-and-paste, so whitespace isn't good): diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 03e3370..11bbfb5 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -961,6 +961,8 @@ int intel_wait_ring_buffer(struct intel_ring_buffer *ring, int n) msleep(1); if (atomic_read(&dev_priv->mm.wedged)) return -EAGAIN; + /* Force a re-read. FIXME: what if read_status_page says 0 too */ + ring->actual_head = 0; } while (!time_after(jiffies, end)); trace_i915_ring_wait_end (dev); return -EBUSY; but to get rid of the FIXME you should probably get rid of "actual_head" entirely, and just use a local variable as a simple boolean flag instead. Hmm? Linus