Patch "drm/i915/gen9: fix the WaWmMemoryReadLatency implementation" has been added to the 4.8-stable tree

This is a note to let you know that I've just added the patch titled

drm/i915/gen9: fix the WaWmMemoryReadLatency implementation

to the 4.8-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
drm-i915-gen9-fix-the-wawmmemoryreadlatency-implementation.patch
and it can be found in the queue-4.8 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree, please let <[email protected]> know about it.

From: Paulo Zanoni <[email protected]>
Date: Thu, 22 Sep 2016 18:00:30 -0300
Subject: drm/i915/gen9: fix the WaWmMemoryReadLatency implementation

From: Paulo Zanoni <[email protected]>

commit 4e4d3814a9bb4d71cd3ff0701d8d7041edefd8f0 upstream.

Bspec says:
"The mailbox response data may not account for memory read latency.
If the mailbox response data for level 0 is 0us, add 2 microseconds to the result for each valid level."

This means we should only do the +2 in case wm[0] == 0, not always.

So split the sanitizing implementation from the WA implementation and fix the WA implementation.

v2: Add Fixes tag (Maarten).

---
 drivers/gpu/drm/i915/intel_pm.c |   42 ++++++++++++++++++++--------------------
 1 file changed, 22 insertions(+), 20 deletions(-)

@@ -2119,32 +2119,34 @@ static void intel_read_wm_latency(struct
 				GEN9_MEM_LATENCY_LEVEL_MASK;
 
 		/*
+		 * If a level n (n > 1) has a 0us latency, all levels m (m >= n)
+		 * need to be disabled. We make sure to sanitize the values out
+		 * of the punit to satisfy this requirement.
+		 */
+		for (level = 1; level <= max_level; level++) {
+			if (wm[level] == 0) {
+				for (i = level + 1; i <= max_level; i++)
+					wm[i] = 0;
+				break;
+			}
+		}
+
+		/*
 		 * WaWmMemoryReadLatency:skl
 		 *
 		 * punit doesn't take into account the read latency so we need
-		 * to add 2us to the various latency levels we retrieve from
-		 * the punit.
-		 *   - W0 is a bit special in that it's the only level that
-		 *   can't be disabled if we want to have display working, so
-		 *   we always add 2us there.
-		 *   - For levels >=1, punit returns 0us latency when they are
-		 *   disabled, so we respect that and don't add 2us then
-		 *
-		 * Additionally, if a level n (n > 1) has a 0us latency, all
-		 * levels m (m >= n) need to be disabled. We make sure to
-		 * sanitize the values out of the punit to satisfy this
-		 * requirement.
+		 * to add 2us to the various latency levels we retrieve from the
+		 * punit when level 0 response data us 0us.
 		 */
-		wm[0] += 2;
-		for (level = 1; level <= max_level; level++)
-			if (wm[level] != 0)
+		if (wm[0] == 0) {
+			wm[0] += 2;
+			for (level = 1; level <= max_level; level++) {
+				if (wm[level] == 0)
+					break;
 				wm[level] += 2;
-			else {
-				for (i = level + 1; i <= max_level; i++)
-					wm[i] = 0;
-
-				break;
 			}
+		}
+
 	} else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		uint64_t sskpd = I915_READ64(MCH_SSKPD);
 

Patches currently in stable-queue which might be from [email protected] are

queue-4.8/drm-i915-gen9-fix-the-watermark-res_blocks-value.patch queue-4.8/drm-i915-gen9-only-add-the-planes-actually-affected-by-ddb-changes.patch queue-4.8/drm-i915-gen9-fix-the-wawmmemoryreadlatency-implementation.patch queue-4.8/drm-i915-introduce-intel_has_sagv.patch queue-4.8/drm-i915-gen9-fix-plane_blocks_per_line-on-watermarks-calculations.patch queue-4.8/drm-i915-kbl-kbl-also-needs-to-run-the-sagv-code.patch queue-4.8/drm-i915-gen9-minimum-scanlines-for-y-tile-is-not-always-4.patch queue-4.8/drm-i915-sagv-is-not-skl-only-so-rename-a-few-things.patch