[PATCH 4.17 157/336] drm/nouveau/fifo/gk104-: poll for runlist update completion

From: Greg Kroah-Hartman
Date: Wed Aug 01 2018 - 14:35:59 EST


4.17-stable review patch. If anyone has any objections, please let me know.

------------------

From: Ben Skeggs <bskeggs@xxxxxxxxxx>

[ Upstream commit 4f2fc25c0f8bcc8db1b8a7b21e88c3d7f35c5acb ]

Newer HW doesn't appear to send this event, which will cause long delays
in runlist updates if they don't complete immediately.

RM doesn't use these events anywhere, and an NVGPU commit message notes
that polling is the preferred method even on HW that supports the event.

Signed-off-by: Ben Skeggs <bskeggs@xxxxxxxxxx>
Signed-off-by: Sasha Levin <alexander.levin@xxxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)

--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -155,10 +155,10 @@ gk104_fifo_runlist_commit(struct gk104_f
(target << 28));
nvkm_wr32(device, 0x002274, (runl << 20) | nr);

- if (wait_event_timeout(fifo->runlist[runl].wait,
- !(nvkm_rd32(device, 0x002284 + (runl * 0x08))
- & 0x00100000),
- msecs_to_jiffies(2000)) == 0)
+ if (nvkm_msec(device, 2000,
+ if (!(nvkm_rd32(device, 0x002284 + (runl * 0x08)) & 0x00100000))
+ break;
+ ) < 0)
nvkm_error(subdev, "runlist %d update timeout\n", runl);
unlock:
mutex_unlock(&subdev->mutex);