Blob Blame History Raw
From dc3421560a67361442f33ec962fc6dd48895a0df Mon Sep 17 00:00:00 2001
From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Date: Mon, 20 Mar 2023 15:14:23 +0000
Subject: [PATCH] drm/i915: Fix context runtime accounting
Git-commit: dc3421560a67361442f33ec962fc6dd48895a0df
Patch-mainline: v6.3-rc6
Alt-commit: b3e70051879c665acdd3a1ab50d0ed58d6a8001f
References: git-fixes

When considering whether to mark one context as stopped and another as
started we need to look at whether the previous and new _contexts_ are
different and not just requests. Otherwise the software tracked context
start time was incorrectly updated to the most recent lite-restore time-
stamp, which was in some cases resulting in active time going backward,
until the context switch (typically the heartbeat pulse) would synchronise
with the hardware tracked context runtime. Easiest use case to observe
this behaviour was with a full screen clients with close to 100% engine
load.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Fixes: bb6287cb1886 ("drm/i915: Track context current active time")
Cc: <stable@vger.kernel.org> # v5.19+
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230320151423.1708436-1-tvrtko.ursulin@linux.intel.com
[tursulin: Fix spelling in commit msg.]
(cherry picked from commit b3e70051879c665acdd3a1ab50d0ed58d6a8001f)

Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Acked-by: Takashi Iwai <tiwai@suse.de>

---
 drivers/gpu/drm/i915/gt/intel_execlists_submission.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_execlists_submission.c b/drivers/gpu/drm/i915/gt/intel_execlists_submission.c
index 1bbe6708d0a7..750326434677 100644
--- a/drivers/gpu/drm/i915/gt/intel_execlists_submission.c
+++ b/drivers/gpu/drm/i915/gt/intel_execlists_submission.c
@@ -2018,6 +2018,8 @@ process_csb(struct intel_engine_cs *engine, struct i915_request **inactive)
 	 * inspecting the queue to see if we need to resumbit.
 	 */
 	if (*prev != *execlists->active) { /* elide lite-restores */
+		struct intel_context *prev_ce = NULL, *active_ce = NULL;
+
 		/*
 		 * Note the inherent discrepancy between the HW runtime,
 		 * recorded as part of the context switch, and the CPU
@@ -2029,9 +2031,15 @@ process_csb(struct intel_engine_cs *engine, struct i915_request **inactive)
 		 * and correct overselves later when updating from HW.
 		 */
 		if (*prev)
-			lrc_runtime_stop((*prev)->context);
+			prev_ce = (*prev)->context;
 		if (*execlists->active)
-			lrc_runtime_start((*execlists->active)->context);
+			active_ce = (*execlists->active)->context;
+		if (prev_ce != active_ce) {
+			if (prev_ce)
+				lrc_runtime_stop(prev_ce);
+			if (active_ce)
+				lrc_runtime_start(active_ce);
+		}
 		new_timeslice(execlists);
 	}
 
-- 
2.35.3