Skip to contentSkip to conversation

FAILED: patch "[PATCH] drm/i915: Hold RPM wakelock while initializing OA buffer" failed to apply to 4.13-stable tree

PATCH [email protected] 2 participants 2 messages Sep 20, 2017

Subsystems (3)
lore
Share & tools
lore

[email protected]

author maintainer
#0 lore raw

The patch below does not apply to the 4.13-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

Quoted 1 line

From 987f8c444aa2c33d98e7030d0c5f0a5325cc84ea Mon Sep 17 00:00:00 2001

From: "[email protected]" <[email protected]>
Date: Tue, 27 Jun 2017 23:09:41 +0530
Subject: [PATCH] drm/i915: Hold RPM wakelock while initializing OA buffer

OA buffer initialization involves access to HW registers to set the OA base, head and tail. Ensure device is awake while setting these. With this, all oa.ops are covered under RPM and forcewake wakelock.

Cc Lionel Landwerlin <[email protected]>
Fixes d79651522e89c ("drm/i915: Enable i915 perf stream for Haswell OA unit")
Cc <[email protected]> # v4.11+
drivers/gpu/drm/i915/i915_perf.c +6 -6
index afd8260cd096..d9f77a4d85db 100644
@@ -2067,10 +2067,6 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 			return ret;
 	}
 
-	ret = alloc_oa_buffer(dev_priv);
-	if (ret)
-		goto err_oa_buf_alloc;
-
 	/* PRM - observability performance counters:
 	 *
 	 *   OACONTROL, performance counter enable, note:
@@ -2086,6 +2082,10 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 	intel_runtime_pm_get(dev_priv);
 	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
 
+	ret = alloc_oa_buffer(dev_priv);
+	if (ret)
+		goto err_oa_buf_alloc;
+
 	ret = dev_priv->perf.oa.ops.enable_metric_set(dev_priv);
 	if (ret)
 		goto err_enable;
@@ -2097,11 +2097,11 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 	return 0;
 
 err_enable:
-	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
-	intel_runtime_pm_put(dev_priv);
 	free_oa_buffer(dev_priv);
 
 err_oa_buf_alloc:
+	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
+	intel_runtime_pm_put(dev_priv);
 	if (stream->ctx)
 		oa_put_render_ctx_id(stream);
 
#1 lore raw

Hi Greg,

Something must have gone wrong with this patch.
Surprisingly it seem to be in 4.13 :

commit 04941829b0049d2446c7042ab9686dd057d809a6
Author: [email protected] <[email protected]> Date:�� Tue Jun 27 23:09:41 2017 +0530

��� drm/i915: Hold RPM wakelock while initializing OA buffer

Cheers,

-
Lionel

Quoted 70 lines · On 20/09/17 10:16, [email protected] wrote:

The patch below does not apply to the 4.13-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 987f8c444aa2c33d98e7030d0c5f0a5325cc84ea Mon Sep 17 00:00:00 2001
From: "[email protected]" <[email protected]>
Date: Tue, 27 Jun 2017 23:09:41 +0530
Subject: [PATCH] drm/i915: Hold RPM wakelock while initializing OA buffer

OA buffer initialization involves access to HW registers to set the OA base, head and tail. Ensure device is awake while setting these. With this, all oa.ops are covered under RPM and forcewake wakelock.

Cc Lionel Landwerlin <[email protected]>
Fixes d79651522e89c ("drm/i915: Enable i915 perf stream for Haswell OA unit")
Cc <[email protected]> # v4.11+
drivers/gpu/drm/i915/i915_perf.c +6 -6
index afd8260cd096..d9f77a4d85db 100644
@@ -2067,10 +2067,6 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
  			return ret;
  	}
 
-	ret = alloc_oa_buffer(dev_priv);
-	if (ret)
-		goto err_oa_buf_alloc;
-
  	/* PRM - observability performance counters:
  	 *
  	 *   OACONTROL, performance counter enable, note:
@@ -2086,6 +2082,10 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
  	intel_runtime_pm_get(dev_priv);
  	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
 
+	ret = alloc_oa_buffer(dev_priv);
+	if (ret)
+		goto err_oa_buf_alloc;
+
  	ret = dev_priv->perf.oa.ops.enable_metric_set(dev_priv);
  	if (ret)
  		goto err_enable;
@@ -2097,11 +2097,11 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
  	return 0;
 
  err_enable:
-	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
-	intel_runtime_pm_put(dev_priv);
  	free_oa_buffer(dev_priv);
 
  err_oa_buf_alloc:
+	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
+	intel_runtime_pm_put(dev_priv);
  	if (stream->ctx)
  		oa_put_render_ctx_id(stream);