File _service:download_files:0003-revive-wakeup_gran... of Package linux-px

Overview Repositories Revisions Requests Users Attributes Meta

File _service:download_files:0003-revive-wakeup_granularity_ns.patch of Package linux-px (Revision 711b08596e189bb069b17e2052915d39)

Currently displaying revision 711b08596e189bb069b17e2052915d39 , Show latest

From 2366ba4f9edf504f7fc0bc859c699360954f2ac2 Mon Sep 17 00:00:00 2001
From: Masahito S <firelzrd@gmail.com>
Date: Wed, 3 May 2023 03:40:38 +0900
Subject: [PATCH] revive wakeup_granularity_ns

---
 kernel/sched/debug.c |  1 +
 kernel/sched/fair.c  | 84 +++++++++++++++++++++++++++++++++++++++-----
 kernel/sched/sched.h |  1 +
 3 files changed, 78 insertions(+), 8 deletions(-)

diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c
index 5f903d844..37016e049 100644
--- a/kernel/sched/debug.c
+++ b/kernel/sched/debug.c
@@ -309,6 +309,7 @@ static __init int sched_init_debug(void)
 #endif
 
 	debugfs_create_u32("base_slice_ns", 0644, debugfs_sched, &sysctl_sched_base_slice);
+	debugfs_create_u32("wakeup_granularity_ns", 0644, debugfs_sched, &sysctl_sched_wakeup_granularity);
 
 	debugfs_create_u32("latency_warn_ms", 0644, debugfs_sched, &sysctl_resched_latency_warn_ms);
 	debugfs_create_u32("latency_warn_once", 0644, debugfs_sched, &sysctl_resched_latency_warn_once);
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index a69648593..ed01d6f5b 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -69,17 +69,17 @@
  *   SCHED_TUNABLESCALING_LOG - scaled logarithmical, *1+ilog(ncpus)
  *   SCHED_TUNABLESCALING_LINEAR - scaled linear, *ncpus
  *
- * (default SCHED_TUNABLESCALING_LOG = *(1+ilog(ncpus))
+ * (default SCHED_TUNABLESCALING_NONE = *1)
  */
-unsigned int sysctl_sched_tunable_scaling = SCHED_TUNABLESCALING_LOG;
+unsigned int sysctl_sched_tunable_scaling = SCHED_TUNABLESCALING_NONE;
 
 /*
  * Minimal preemption granularity for CPU-bound tasks:
  *
- * (default: 0.75 msec * (1 + ilog(ncpus)), units: nanoseconds)
+ * (default: 3 msec * 1, units: nanoseconds)
  */
-unsigned int sysctl_sched_base_slice			= 750000ULL;
-static unsigned int normalized_sysctl_sched_base_slice	= 750000ULL;
+unsigned int sysctl_sched_base_slice			= 3000000ULL;
+static unsigned int normalized_sysctl_sched_base_slice	= 3000000ULL;
 
 /*
  * After fork, child runs first. If set to 0 (default) then
@@ -87,6 +87,18 @@ static unsigned int normalized_sysctl_sched_base_slice	= 750000ULL;
  */
 unsigned int sysctl_sched_child_runs_first __read_mostly;
 
+/*
+ * SCHED_OTHER wake-up granularity.
+ *
+ * This option delays the preemption effects of decoupled workloads
+ * and reduces their over-scheduling. Synchronous workloads will still
+ * have immediate wakeup/sleep latencies.
+ *
+ * (default: 9.6 msec * 1, units: nanoseconds)
+ */
+unsigned int sysctl_sched_wakeup_granularity			= 9600000UL;
+static unsigned int normalized_sysctl_sched_wakeup_granularity	= 9600000UL;
+
 const_debug unsigned int sysctl_sched_migration_cost	= 500000UL;
 
 #ifdef CONFIG_SCHED_BORE
@@ -329,6 +341,7 @@ static void update_sysctl(void)
 #define SET_SYSCTL(name) \
 	(sysctl_##name = (factor) * normalized_sysctl_##name)
 	SET_SYSCTL(sched_base_slice);
+	SET_SYSCTL(sched_wakeup_granularity);
 #undef SET_SYSCTL
 }
 
@@ -1051,6 +1064,7 @@ int sched_update_scaling(void)
 #define WRT_SYSCTL(name) \
 	(normalized_sysctl_##name = sysctl_##name / (factor))
 	WRT_SYSCTL(sched_base_slice);
+	WRT_SYSCTL(sched_wakeup_granularity);
 #undef WRT_SYSCTL
 
 	return 0;
@@ -5285,6 +5299,9 @@ set_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *se)
 	se->prev_sum_exec_runtime = se->sum_exec_runtime;
 }
 
+static int
+wakeup_preempt_entity(struct sched_entity *curr, struct sched_entity *se);
+
 /*
  * Pick the next process, keeping these things in mind, in this order:
  * 1) keep things fair between processes/task groups
@@ -5295,14 +5312,16 @@ set_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *se)
 static struct sched_entity *
 pick_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *curr)
 {
+	struct sched_entity *candidate = pick_eevdf(cfs_rq);
 	/*
 	 * Enabling NEXT_BUDDY will affect latency but not fairness.
 	 */
 	if (sched_feat(NEXT_BUDDY) &&
-	    cfs_rq->next && entity_eligible(cfs_rq, cfs_rq->next))
+	    cfs_rq->next && entity_eligible(cfs_rq, cfs_rq->next) && 
+			wakeup_preempt_entity(cfs_rq->next, candidate) < 1)
 		return cfs_rq->next;
 
-	return pick_eevdf(cfs_rq);
+	return candidate;
 }
 
 static bool check_cfs_rq_runtime(struct cfs_rq *cfs_rq);
@@ -6531,6 +6550,55 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags)
 	hrtick_update(rq);
 }
 
+static unsigned long wakeup_gran(struct sched_entity *se)
+{
+	unsigned long gran = sysctl_sched_wakeup_granularity;
+
+	/*
+	 * Since its curr running now, convert the gran from real-time
+	 * to virtual-time in his units.
+	 *
+	 * By using 'se' instead of 'curr' we penalize light tasks, so
+	 * they get preempted easier. That is, if 'se' < 'curr' then
+	 * the resulting gran will be larger, therefore penalizing the
+	 * lighter, if otoh 'se' > 'curr' then the resulting gran will
+	 * be smaller, again penalizing the lighter task.
+	 *
+	 * This is especially important for buddies when the leftmost
+	 * task is higher priority than the buddy.
+	 */
+	return calc_delta_fair(gran, se);
+}
+
+/*
+ * Should 'se' preempt 'curr'.
+ *
+ *             |s1
+ *        |s2
+ *   |s3
+ *         g
+ *      |<--->|c
+ *
+ *  w(c, s1) = -1
+ *  w(c, s2) =  0
+ *  w(c, s3) =  1
+ *
+ */
+static int
+wakeup_preempt_entity(struct sched_entity *curr, struct sched_entity *se)
+{
+	s64 gran, vdiff = curr->vruntime - se->vruntime;
+
+	if (vdiff <= 0)
+		return -1;
+
+	gran = wakeup_gran(se);
+	if (vdiff > gran)
+		return 1;
+
+	return 0;
+}
+
 static void set_next_buddy(struct sched_entity *se);
 
 /*
@@ -8005,7 +8073,7 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_
 	/*
 	 * XXX pick_eevdf(cfs_rq) != se ?
 	 */
-	if (pick_eevdf(cfs_rq) == pse)
+	if ((pick_eevdf(cfs_rq) == pse) && (wakeup_preempt_entity(se, pse) == 1))
 		goto preempt;
 
 	return;
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index bfce45b21..c61077166 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -2484,6 +2484,7 @@ extern const_debug unsigned int sysctl_sched_migration_cost;
 
 #ifdef CONFIG_SCHED_DEBUG
 extern unsigned int sysctl_sched_base_slice;
+extern unsigned int sysctl_sched_wakeup_granularity;
 extern int sysctl_resched_latency_warn_ms;
 extern int sysctl_resched_latency_warn_once;
 
-- 
2.25.1

Places

File _service:download_files:0003-revive-wakeup_granularity_ns.patch of Package linux-px (Revision 711b08596e189bb069b17e2052915d39)

Places