Two things seem to do the trick on my ivb machine here: - prevent the gt from powering down while waiting for seqno notification interrupts by grabbing the force_wake in get_irq (and dropping it in put_irq again). - ordering writes from the ring's CS by reading a CS register, ACTHD seems to work. Only the blt&bsd ring on ivb seem to be massively affected by this, but for paranoia do this dance also on the render ring and on snb (i.e. all gpus with forcewake). Tested with Eric's glCopyPixels loop which without this patch scores a missed irq every few seconds. This patch needs my forcewake rework to use a spinlock instead of dev->struct_mutex. Cc: stable at kernel.org Cc: Eric Anholt <eric at anholt.net> Cc: Kenneth Graunke <kenneth at whitecape.org> Cc: Eugeni Dodonov <eugeni.dodonov at intel.com> Signed-Off-by: Daniel Vetter <daniel.vetter at ffwll.ch> --- For easier testing grab my -next branch at http://cgit.freedesktop.org/~danvet/drm/log/?h=my-next Cheers, Daniel drivers/gpu/drm/i915/intel_ringbuffer.c | 21 +++++++++++++++++++-- 1 files changed, 19 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index cf6e159..dbbc565 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -631,6 +631,15 @@ render_ring_add_request(struct intel_ring_buffer *ring, } static u32 +gen6_ring_get_seqno(struct intel_ring_buffer *ring) +{ + /* Workaround for force correct ordering between irq and seqno writes on + * ivb (and maybe also on snb). */ + intel_ring_get_active_head(ring); + return intel_read_status_page(ring, I915_GEM_HWS_INDEX); +} + +static u32 ring_get_seqno(struct intel_ring_buffer *ring) { return intel_read_status_page(ring, I915_GEM_HWS_INDEX); @@ -795,6 +804,11 @@ gen6_ring_get_irq(struct intel_ring_buffer *ring, u32 gflag, u32 rflag) if (!dev->irq_enabled) return false; + /* It looks like we need to prevent the gt from suspending while waiting + * for an notifiy irq, otherwise irqs seem to get lost on at least the + * blt/bsd rings. */ + gen6_gt_force_wake_get(dev_priv); + spin_lock(&ring->irq_lock); if (ring->irq_refcount++ == 0) { ring->irq_mask &= ~rflag; @@ -819,6 +833,8 @@ gen6_ring_put_irq(struct intel_ring_buffer *ring, u32 gflag, u32 rflag) ironlake_disable_irq(dev_priv, gflag); } spin_unlock(&ring->irq_lock); + + gen6_gt_force_wake_put(dev_priv); } static bool @@ -1329,7 +1345,7 @@ static const struct intel_ring_buffer gen6_bsd_ring = { .write_tail = gen6_bsd_ring_write_tail, .flush = gen6_ring_flush, .add_request = gen6_add_request, - .get_seqno = ring_get_seqno, + .get_seqno = gen6_ring_get_seqno, .irq_get = gen6_bsd_ring_get_irq, .irq_put = gen6_bsd_ring_put_irq, .dispatch_execbuffer = gen6_ring_dispatch_execbuffer, @@ -1388,7 +1404,7 @@ static const struct intel_ring_buffer gen6_blt_ring = { .write_tail = ring_write_tail, .flush = blt_ring_flush, .add_request = gen6_add_request, - .get_seqno = ring_get_seqno, + .get_seqno = gen6_ring_get_seqno, .irq_get = blt_ring_get_irq, .irq_put = blt_ring_put_irq, .dispatch_execbuffer = gen6_ring_dispatch_execbuffer, @@ -1410,6 +1426,7 @@ int intel_init_render_ring_buffer(struct drm_device *dev) ring->flush = gen6_render_ring_flush; ring->irq_get = gen6_render_ring_get_irq; ring->irq_put = gen6_render_ring_put_irq; + ring->get_seqno = gen6_ring_get_seqno; } else if (IS_GEN5(dev)) { ring->add_request = pc_render_add_request; ring->get_seqno = pc_render_get_seqno; -- 1.7.7.3
