Skip to contentSkip to conversation

scheduler fixes

GIT PULL Ingo Molnar 1 participant 1 message Dec 5, 2011 x86

lore
Share & tools
lore
#0 lore raw

Linus,

Please pull the latest sched-urgent-for-linus git tree from:

git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git sched-urgent-for-linus

Thanks,

Ingo

------------------>
Carsten Emde (1):
sched: Set the command name of the idle tasks in SMP kernels

Hui Kang (1):
sched_fair: Fix a typo in the comment describing update_sd_lb_stats

J. Bruce Fields (1):
sched: Document wait_for_completion_*() return values

Paul Turner (1):
sched: Fix buglet in return_cfs_rq_runtime()

Peter Zijlstra (3):
sched: Add a comment to effective_load() since it's a pain sched, rt: Provide means of disabling cross-cpu bandwidth sharing
sched: Avoid SMT siblings in select_idle_sibling() if possible

Salman Qazi (1):
sched, x86: Avoid unnecessary overflow in sched_clock

 arch/x86/include/asm/timer.h |   23 ++++++-
 include/linux/init_task.h    |    4 +-
 kernel/sched.c               |   17 +++++
 kernel/sched_fair.c          |  159 +++++++++++++++++++++++++++++++++---------
 kernel/sched_features.h      |    1 +
 kernel/sched_rt.c            |    3 +
 6 files changed, 171 insertions(+), 36 deletions(-)
arch/x86/include/asm/timer.h +22 -1
index fa7b917..431793e 100644
@@ -32,6 +32,22 @@ extern int no_timer_check;
  *  ([email protected])
  *
  *			[email protected] "math is hard, lets go shopping!"
+ *
+ * In:
+ *
+ * ns = cycles * cyc2ns_scale / SC
+ *
+ * Although we may still have enough bits to store the value of ns,
+ * in some cases, we may not have enough bits to store cycles * cyc2ns_scale,
+ * leading to an incorrect result.
+ *
+ * To avoid this, we can decompose 'cycles' into quotient and remainder
+ * of division by SC.  Then,
+ *
+ * ns = (quot * SC + rem) * cyc2ns_scale / SC
+ *    = quot * cyc2ns_scale + (rem * cyc2ns_scale) / SC
+ *
+ *			- [email protected]
  */
 
 DECLARE_PER_CPU(unsigned long, cyc2ns);
@@ -41,9 +57,14 @@ DECLARE_PER_CPU(unsigned long long, cyc2ns_offset);
 
 static inline unsigned long long __cycles_2_ns(unsigned long long cyc)
 {
+	unsigned long long quot;
+	unsigned long long rem;
 	int cpu = smp_processor_id();
 	unsigned long long ns = per_cpu(cyc2ns_offset, cpu);
-	ns += cyc * per_cpu(cyc2ns, cpu) >> CYC2NS_SCALE_FACTOR;
+	quot = (cyc >> CYC2NS_SCALE_FACTOR);
+	rem = cyc & ((1ULL << CYC2NS_SCALE_FACTOR) - 1);
+	ns += quot * per_cpu(cyc2ns, cpu) +
+		((rem * per_cpu(cyc2ns, cpu)) >> CYC2NS_SCALE_FACTOR);
 	return ns;
 }
 
include/linux/init_task.h +3 -1
index 08ffab0..b6e5b8b 100644
@@ -126,6 +126,8 @@ extern struct cred init_cred;
 # define INIT_PERF_EVENTS(tsk)
 #endif
 
+#define INIT_TASK_COMM "swapper"
+
 /*
  *  INIT_TASK is used to set up the first task table, touch at
  * your own risk!. Base=0, limit=0x1fffff (=2MB)
@@ -162,7 +164,7 @@ extern struct cred init_cred;
 	.group_leader	= &tsk,						\
 	RCU_INIT_POINTER(.real_cred, &init_cred),			\
 	RCU_INIT_POINTER(.cred, &init_cred),				\
-	.comm		= "swapper",					\
+	.comm		= INIT_TASK_COMM,				\
 	.thread		= INIT_THREAD,					\
 	.fs		= &init_fs,					\
 	.files		= &init_files,					\
kernel/sched.c +17 -0
index 0e9344a..d6b149c 100644
@@ -71,6 +71,7 @@
 #include <linux/ctype.h>
 #include <linux/ftrace.h>
 #include <linux/slab.h>
+#include <linux/init_task.h>
 
 #include <asm/tlb.h>
 #include <asm/irq_regs.h>
@@ -4810,6 +4811,9 @@ EXPORT_SYMBOL(wait_for_completion);
  * This waits for either a completion of a specific task to be signaled or for a
  * specified timeout to expire. The timeout is in jiffies. It is not
  * interruptible.
+ *
+ * The return value is 0 if timed out, and positive (at least 1, or number of
+ * jiffies left till timeout) if completed.
  */
 unsigned long __sched
 wait_for_completion_timeout(struct completion *x, unsigned long timeout)
@@ -4824,6 +4828,8 @@ EXPORT_SYMBOL(wait_for_completion_timeout);
  *
  * This waits for completion of a specific task to be signaled. It is
  * interruptible.
+ *
+ * The return value is -ERESTARTSYS if interrupted, 0 if completed.
  */
 int __sched wait_for_completion_interruptible(struct completion *x)
 {
@@ -4841,6 +4847,9 @@ EXPORT_SYMBOL(wait_for_completion_interruptible);
  *
  * This waits for either a completion of a specific task to be signaled or for a
  * specified timeout to expire. It is interruptible. The timeout is in jiffies.
+ *
+ * The return value is -ERESTARTSYS if interrupted, 0 if timed out,
+ * positive (at least 1, or number of jiffies left till timeout) if completed.
  */
 long __sched
 wait_for_completion_interruptible_timeout(struct completion *x,
@@ -4856,6 +4865,8 @@ EXPORT_SYMBOL(wait_for_completion_interruptible_timeout);
  *
  * This waits to be signaled for completion of a specific task. It can be
  * interrupted by a kill signal.
+ *
+ * The return value is -ERESTARTSYS if interrupted, 0 if completed.
  */
 int __sched wait_for_completion_killable(struct completion *x)
 {
@@ -4874,6 +4885,9 @@ EXPORT_SYMBOL(wait_for_completion_killable);
  * This waits for either a completion of a specific task to be
  * signaled or for a specified timeout to expire. It can be
  * interrupted by a kill signal. The timeout is in jiffies.
+ *
+ * The return value is -ERESTARTSYS if interrupted, 0 if timed out,
+ * positive (at least 1, or number of jiffies left till timeout) if completed.
  */
 long __sched
 wait_for_completion_killable_timeout(struct completion *x,
@@ -6099,6 +6113,9 @@ void __cpuinit init_idle(struct task_struct *idle, int cpu)
 	 */
 	idle->sched_class = &idle_sched_class;
 	ftrace_graph_init_idle_task(idle, cpu);
+#if defined(CONFIG_SMP)
+	sprintf(idle->comm, "%s/%d", INIT_TASK_COMM, cpu);
+#endif
 }
 
 /*
kernel/sched_fair.c +125 -34

226 lines not on this page · raw message

kernel/sched_features.h +1 -0
index efa0a7b..8480224 100644
@@ -67,3 +67,4 @@ SCHED_FEAT(NONTASK_POWER, 1)
 SCHED_FEAT(TTWU_QUEUE, 1)
 
 SCHED_FEAT(FORCE_SD_OVERLAP, 0)
+SCHED_FEAT(RT_RUNTIME_SHARE, 1)
kernel/sched_rt.c +3 -0
index 056cbd2..583a136 100644
@@ -560,6 +560,9 @@ static int balance_runtime(struct rt_rq *rt_rq)
 {
 	int more = 0;
 
+	if (!sched_feat(RT_RUNTIME_SHARE))
+		return more;
+
 	if (rt_rq->rt_time > rt_rq->rt_runtime) {
 		raw_spin_unlock(&rt_rq->rt_runtime_lock);
 		more = do_balance_runtime(rt_rq);