Skip to contentSkip to conversation

BLD-3.18 release.

ANNOUNCE Rakib Mullick 1 participant 1 message Dec 16, 2014 Scheduler

lore
Share & tools
lore
#0 lore raw

BLD patch for Linux-3.18. Changes since previous release:

* Contains changes to address issue at the time of wakeup
of rt tasks, it was reported by Peter Junos <[email protected]>.

BLD has some positive impact on vmlinux size too, the following shows the picture:

$ cat size.mainline 
   text	   data	    bss	    dec	    hex	filename
12769041	2056008	11722752	26547801	1951659	vmlinux
$ cat size.bld
   text	   data	    bss	    dec	    hex	filename
12755462	2056040	11722752	26534254	194e16e	vmlinux

and the config could be found here:

https://raw.githubusercontent.com/rmullick/bld-patches/master/config.benchmark-3.17

After previous release, Mike Galbraith shows that on systems BLD can reduce throughput significantly, due to L2 misses and where no L3 is available, that issue is yet to address, I lack those kind of systems, so it might take some time.

Thanks,
Rakib

---

init/Kconfig +9 -0
index 2081a4d..becfd85 100644
@@ -36,6 +36,15 @@ config BROKEN_ON_SMP
 	depends on BROKEN || !SMP
 	default y
 
+config BLD
+	bool "An alternate CPU load distribution technique for task scheduler"
+	depends on SMP
+	default y
+	help
+	  This is an alternate CPU load distribution technique based for task
+	  scheduler based on The Barbershop Load Distribution algorithm. Not
+	  suitable for NUMA, should work well on SMP.
+
 config INIT_ENV_ARG_LIMIT
 	int
 	default 32 if !UML
kernel/sched/bld.h new file +215 -0

215 lines not on this page · raw message

kernel/sched/core.c +51 -7

160 lines not on this page · raw message

kernel/sched/fair.c +42 -30

158 lines not on this page · raw message

kernel/sched/rt.c +4 -0
index 20bca39..03a0ed3 100644
@@ -1295,6 +1295,7 @@ static void yield_task_rt(struct rq *rq)
 #ifdef CONFIG_SMP
 static int find_lowest_rq(struct task_struct *task);
 
+#ifndef	CONFIG_BLD
 static int
 select_task_rq_rt(struct task_struct *p, int cpu, int sd_flag, int flags)
 {
@@ -1348,6 +1349,7 @@ select_task_rq_rt(struct task_struct *p, int cpu, int sd_flag, int flags)
 out:
 	return cpu;
 }
+#endif	/* CONFIG_BLD */
 
 static void check_preempt_equal_prio(struct rq *rq, struct task_struct *p)
 {
@@ -2111,7 +2113,9 @@ const struct sched_class rt_sched_class = {
 	.put_prev_task		= put_prev_task_rt,
 
 #ifdef CONFIG_SMP
+#ifndef	CONFIG_BLD
 	.select_task_rq		= select_task_rq_rt,
+#endif
 
 	.set_cpus_allowed       = set_cpus_allowed_rt,
 	.rq_online              = rq_online_rt,
kernel/sched/sched.h +11 -3
index 2df8ef0..79fce51 100644
@@ -366,9 +366,8 @@ struct cfs_rq {
 #endif /* CONFIG_FAIR_GROUP_SCHED */
 #endif /* CONFIG_SMP */
 
-#ifdef CONFIG_FAIR_GROUP_SCHED
 	struct rq *rq;	/* cpu runqueue to which this cfs_rq is attached */
-
+#ifdef CONFIG_FAIR_GROUP_SCHED
 	/*
 	 * leaf cfs_rqs are those that hold tasks (lowest schedulable entity in
 	 * a hierarchy). Non-leaf lrqs hold other higher schedulable entities
@@ -392,6 +391,11 @@ struct cfs_rq {
 	struct list_head throttled_list;
 #endif /* CONFIG_CFS_BANDWIDTH */
 #endif /* CONFIG_FAIR_GROUP_SCHED */
+
+#ifdef CONFIG_BLD
+	struct list_head bld_cfs_list;
+	char pos;
+#endif
 };
 
 static inline int rt_bandwidth_enabled(void)
@@ -425,12 +429,16 @@ struct rt_rq {
 	/* Nests inside the rq lock: */
 	raw_spinlock_t rt_runtime_lock;
 
+	struct rq *rq;
 #ifdef CONFIG_RT_GROUP_SCHED
 	unsigned long rt_nr_boosted;
 
-	struct rq *rq;
 	struct task_group *tg;
 #endif
+#ifdef CONFIG_BLD
+	struct list_head bld_rt_list;
+	int lowbit;
+#endif
 };
 
 /* Deadline class' related fields in a runqueue */