natty cherry-pick request: drm/i915: Fix gen6 (SNB) missed BLT ring interrupts.

[Robert Hooker]

Please consider pulling the following change into natty to fix LP #761065.

SRU Justification:
 Fixes a constant stream of hangcheck errors flooding dmesg, and
removes the visible stuttering that was caused by it when using 3D
applications.
Impact:
 Fixes missed interrupts on sandybridge GPU's. It doesn't affect any
other GPU generation.
Fix:
 Upstream commit 498e720b96379d8ee9c294950a01534a73defcf3.
Testcase:
 1) Install mesa-utils on a system using sandybridge graphics on 11.04
 2) run vblank_mode=0 glxgears and let it run for 30 seconds or so
 3) kill it then check dmesg
 4) Without fix: hangcheck messages every ~5 seconds, massive
stuttering of the whole desktop observed. With fix: no hangcheck
messages, able to continue using the desktop.

This was sent to stable, but 2.6.38.y stable is dead so it will need
to be manually cherry-picked. It has been tested extensively locally
as well as by users on the bug whom I provided test kernels for. It
applies cleanly to ubuntu-natty.git.

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=498e720b96379d8ee9c294950a01534a73defcf3



drm/i915: Fix gen6 (SNB) missed BLT ring interrupts.

The failure appeared in dmesg as:

[drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer elapsed... blt
ring idle [waiting on 35064155, at 35064155], missed IRQ?

This works around that problem on by making the blitter command
streamer write interrupt state to the Hardware Status Page when a
MI_USER_INTERRUPT command is decoded, which appears to force the seqno
out to memory before the interrupt happens.

v1->v2: Moved to prior interrupt handler installation and RMW flags as
per feedback.
v2->v3: Removed RMW of flags (by anholt)

Cc: stable at kernel.org
Signed-off-by: Daniel J Blueman <daniel.blueman at gmail.com>
Signed-off-by: Eric Anholt <eric at anholt.net>
Tested-by: Chris Wilson <chris at chris-wilson.co.uk> [v1]
Tested-by: Eric Anholt <eric at anholt.net> [v1,v3]
   (incidence of the bug with a testcase went from avg 2/1000 to
   0/12651 in the latest test run (plus more for v1))
Tested-by: Kenneth Graunke <kenneth at whitecape.org> [v1]
Tested-by: Robert Hooker <robert.hooker at canonical.com> [v1]
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=33394
Signed-off-by: Dave Airlie <airlied at redhat.com>