On Tue, 4 Dec 2012 15:11:59 +0200, Mika Kuoppala <mika.kuoppala at linux.intel.com> wrote: > Hi, > > As Ben suggested, explicitly syncing across wrap boundary > revealed yet another set of issues that gem_stress didn't find. > > So here is a patchset which addresses the syncing problems > across wrap boundary. Chris helped me to carve out seqno handling > ugliness from my initial attempts to handle the ring wrapping. > > As I can't break this anymore, i added 6/6. > --Mika > > Mika Kuoppala (6): > drm/i915: Fix debugfs seqno info print to use uint > drm/i915: Don't emit semaphore wait if wrap happened > drm/i915: Split intel_ring_begin > drm/i915: Add intel_ring_handle_seqno wrap Reviewed-by: Chris Wilson <chris at chris-wilson.co.uk> > drm/i915: Add debugfs entry to read/write next_seqno Acked-by: Chris Wilson <chris at chris-wilson.co.uk> > drm/i915: Set initial seqno value close to wrap boundary *washes hands* -Chris -- Chris Wilson, Intel Open Source Technology Centre
