drm/i915/guc/slpc: Add waitboost functionality for SLPC

Add helper in RPS code for handling SLPC and non-SLPC paths.
When boost is requested in the SLPC path, we can ask GuC to ramp
up the frequency req by setting the minimum frequency to boost freq.
Reset freq back to the min softlimit when there are no more waiters.

v2: Schedule a worker thread which can boost freq from within
an interrupt context as well.

v3: No need to check against requested freq before scheduling boost
work (Ashutosh)

Cc: Ashutosh Dixit <ashutosh.dixit@intel.com>
Signed-off-by: Vinay Belgaumkar <vinay.belgaumkar@intel.com>
Reviewed-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211102012608.8609-3-vinay.belgaumkar@intel.com

drivers/gpu/drm/i915/gt/intel_rps.c

@@ -936,8 +936,23 @@ void intel_rps_park(struct intel_rps *rps)
 	GT_TRACE(rps_to_gt(rps), "park:%x\n", rps->cur_freq);
 }
 
+void intel_rps_dec_waiters(struct intel_rps *rps)
+{
+	struct intel_guc_slpc *slpc;
+
+	if (rps_uses_slpc(rps)) {
+		slpc = rps_to_slpc(rps);
+
+		intel_guc_slpc_dec_waiters(slpc);
+	} else {
+		atomic_dec(&rps->num_waiters);
+	}
+}
+
 void intel_rps_boost(struct i915_request *rq)
 {
+	struct intel_guc_slpc *slpc;
+
 	if (i915_request_signaled(rq) || i915_request_has_waitboost(rq))
 		return;
 
@@ -945,6 +960,16 @@ void intel_rps_boost(struct i915_request *rq)
 	if (!test_and_set_bit(I915_FENCE_FLAG_BOOST, &rq->fence.flags)) {
 		struct intel_rps *rps = &READ_ONCE(rq->engine)->gt->rps;
 
+		if (rps_uses_slpc(rps)) {
+			slpc = rps_to_slpc(rps);
+
+			/* Return if old value is non zero */
+			if (!atomic_fetch_inc(&slpc->num_waiters))
+				schedule_work(&slpc->boost_work);
+
+			return;
+		}
+
 		if (atomic_fetch_inc(&rps->num_waiters))
 			return;
 

drivers/gpu/drm/i915/gt/intel_rps.h

@@ -23,6 +23,7 @@ void intel_rps_disable(struct intel_rps *rps);
 void intel_rps_park(struct intel_rps *rps);
 void intel_rps_unpark(struct intel_rps *rps);
 void intel_rps_boost(struct i915_request *rq);
+void intel_rps_dec_waiters(struct intel_rps *rps);
 
 int intel_rps_set(struct intel_rps *rps, u8 val);
 void intel_rps_mark_interactive(struct intel_rps *rps, bool interactive);

drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.c

@@ -446,7 +446,11 @@ int intel_guc_slpc_set_min_freq(struct intel_guc_slpc *slpc, u32 val)
 	    val > slpc->max_freq_softlimit)
 		return -EINVAL;
 
+	/* Need a lock now since waitboost can be modifying min as well */
+	mutex_lock(&slpc->lock);
+
 	with_intel_runtime_pm(&i915->runtime_pm, wakeref) {
+
 		ret = slpc_set_param(slpc,
 				     SLPC_PARAM_GLOBAL_MIN_GT_UNSLICE_FREQ_MHZ,
 				     val);
@@ -459,6 +463,8 @@ int intel_guc_slpc_set_min_freq(struct intel_guc_slpc *slpc, u32 val)
 	if (!ret)
 		slpc->min_freq_softlimit = val;
 
+	mutex_unlock(&slpc->lock);
+
 	return ret;
 }
 
@@ -644,6 +650,20 @@ int intel_guc_slpc_enable(struct intel_guc_slpc *slpc)
 	return 0;
 }
 
+void intel_guc_slpc_dec_waiters(struct intel_guc_slpc *slpc)
+{
+	/*
+	 * Return min back to the softlimit.
+	 * This is called during request retire,
+	 * so we don't need to fail that if the
+	 * set_param fails.
+	 */
+	mutex_lock(&slpc->lock);
+	if (atomic_dec_and_test(&slpc->num_waiters))
+		slpc_force_min_freq(slpc, slpc->min_freq_softlimit);
+	mutex_unlock(&slpc->lock);
+}
+
 int intel_guc_slpc_print_info(struct intel_guc_slpc *slpc, struct drm_printer *p)
 {
 	struct drm_i915_private *i915 = slpc_to_i915(slpc);

drivers/gpu/drm/i915/gt/uc/intel_guc_slpc.h

@@ -39,5 +39,6 @@ int intel_guc_slpc_get_min_freq(struct intel_guc_slpc *slpc, u32 *val);
 int intel_guc_slpc_print_info(struct intel_guc_slpc *slpc, struct drm_printer *p);
 void intel_guc_pm_intrmsk_enable(struct intel_gt *gt);
 void intel_guc_slpc_boost(struct intel_guc_slpc *slpc);
+void intel_guc_slpc_dec_waiters(struct intel_guc_slpc *slpc);
 
 #endif

drivers/gpu/drm/i915/i915_request.c

@@ -339,7 +339,7 @@ bool i915_request_retire(struct i915_request *rq)
 	}
 
 	if (test_and_set_bit(I915_FENCE_FLAG_BOOST, &rq->fence.flags))
-		atomic_dec(&rq->engine->gt->rps.num_waiters);
+		intel_rps_dec_waiters(&rq->engine->gt->rps);
 
 	/*
 	 * We only loosely track inflight requests across preemption,