drm/i915: Invalidate TLBs for the rings after a reset
authorChris Wilson <chris@chris-wilson.co.uk>
Tue, 6 Aug 2013 18:01:14 +0000 (19:01 +0100)
committerBen Hutchings <ben@decadent.org.uk>
Tue, 10 Sep 2013 00:57:25 +0000 (01:57 +0100)
commit132823e00683166597c74acc07ab410360875ac0
treeb9876534b07e0f3bee3d222521af727048bf3835
parent8ab959a23fae461cf05093cfd5b0e6ebdc50aa87
drm/i915: Invalidate TLBs for the rings after a reset

commit 884020bf3d2a3787a1cc6df902e98e0eec60330b upstream.

After any "soft gfx reset" we must manually invalidate the TLBs
associated with each ring. Empirically, it seems that a
suspend/resume or D3-D0 cycle count as a "soft reset". The symptom is
that the hardware would fail to note the new address for its status
page, and so it would continue to write the shadow registers and
breadcrumbs into the old physical address (now used by something
completely different, scary). Whereas the driver would read the new
status page and never see any progress, it would appear that the GPU
hung immediately upon resume.

Based on a patch by naresh kumar kachhi <naresh.kumar.kacchi@intel.com>

Reported-by: Thiago Macieira <thiago@kde.org>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=64725
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: Thiago Macieira <thiago@kde.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
[bwh: Backported to 3.2: add definition of RING_INSTPM() from
 commit c1cd90ed7957 'drm/i915: collect more per ring error state']
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
drivers/gpu/drm/i915/i915_reg.h
drivers/gpu/drm/i915/intel_ringbuffer.c