[PATCH][Lucid SRU] drm/nv50-nvc0: work around an evo channel hang that some people see

[Seth Forshee]

From: David Dillow <dave at thedillows.org>

BugLink: http://bugs.launchpad.net/bugs/583760

Signed-off-by: Ben Skeggs <bskeggs at redhat.com>
(backported from commit 59197c026257ab730c7fc8819a0e2b26e125a534 upstream)

Signed-off-by: Seth Forshee <seth.forshee at canonical.com>
---
 drivers/gpu/drm/nouveau/nv50_display.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index 90f0bf5..72e3048 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -329,6 +329,7 @@ nv50_display_init(struct drm_device *dev)
 	nv_wr32(dev, 0x610300, nv_rd32(dev, 0x610300) & ~1);
 
 	evo->dma.max = (4096/4) - 2;
+	evo->dma.max &= ~7;
 	evo->dma.put = 0;
 	evo->dma.cur = evo->dma.put;
 	evo->dma.free = evo->dma.max - evo->dma.cur;
-- 
1.7.4.1