Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1348214

[PATCH RT 05/17] net: provide a way to delegate processing a softirq to ksoftirqd

Show all headers | View raw

3.14.61-rt64-rc1 stable review patch.
If anyone has any objections, please let me know.

------------------

From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>

If the NET_RX uses up all of his budget it moves the following NAPI
invocations into the `ksoftirqd`. On -RT it does not do so. Instead it
rises the NET_RX softirq in its current context again.

In order to get closer to mainline's behaviour this patch provides
__raise_softirq_irqoff_ksoft() which raises the softirq in the ksoftird.

Cc: stable-rt@vger.kernel.org
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
 include/linux/interrupt.h |  8 ++++++++
 kernel/softirq.c          | 21 +++++++++++++++++++++
 net/core/dev.c            |  2 +-
 3 files changed, 30 insertions(+), 1 deletion(-)

diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
index 2d231b330f49..6116748bf43f 100644
--- a/include/linux/interrupt.h
+++ b/include/linux/interrupt.h
@@ -440,6 +440,14 @@ extern void thread_do_softirq(void);
 extern void open_softirq(int nr, void (*action)(struct softirq_action *));
 extern void softirq_init(void);
 extern void __raise_softirq_irqoff(unsigned int nr);
+#ifdef CONFIG_PREEMPT_RT_FULL
+extern void __raise_softirq_irqoff_ksoft(unsigned int nr);
+#else
+static inline void __raise_softirq_irqoff_ksoft(unsigned int nr)
+{
+	__raise_softirq_irqoff(nr);
+}
+#endif
 
 extern void raise_softirq_irqoff(unsigned int nr);
 extern void raise_softirq(unsigned int nr);
diff --git a/kernel/softirq.c b/kernel/softirq.c
index e88582a86803..8821db3f07e6 100644
--- a/kernel/softirq.c
+++ b/kernel/softirq.c
@@ -703,6 +703,27 @@ void __raise_softirq_irqoff(unsigned int nr)
 }
 
 /*
+ * Same as __raise_softirq_irqoff() but will process them in ksoftirqd
+ */
+void __raise_softirq_irqoff_ksoft(unsigned int nr)
+{
+	unsigned int mask;
+
+	if (WARN_ON_ONCE(!__this_cpu_read(ksoftirqd) ||
+			 !__this_cpu_read(ktimer_softirqd)))
+		return;
+	mask = 1UL << nr;
+
+	trace_softirq_raise(nr);
+	or_softirq_pending(mask);
+	if (mask & TIMER_SOFTIRQS)
+		__this_cpu_read(ktimer_softirqd)->softirqs_raised |= mask;
+	else
+		__this_cpu_read(ksoftirqd)->softirqs_raised |= mask;
+	wakeup_proper_softirq(nr);
+}
+
+/*
  * This function must run with irqs disabled!
  */
 void raise_softirq_irqoff(unsigned int nr)
diff --git a/net/core/dev.c b/net/core/dev.c
index 7bbcef273c74..4259edb6d29b 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -4525,7 +4525,7 @@ out:
 
 softnet_break:
 	sd->time_squeeze++;
-	__raise_softirq_irqoff(NET_RX_SOFTIRQ);
+	__raise_softirq_irqoff_ksoft(NET_RX_SOFTIRQ);
 	goto out;
 }
 
-- 
2.7.0

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH RT 00/17] Linux 3.14.61-rt64-rc1 Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 04/17] softirq: split timer softirqs out of ksoftirqd Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 08/17] kernel: migrate_disable() do fastpath in atomic & irqs-off Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 06/17] latencyhist: disable jump-labels Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 10/17] kernel/stop_machine: partly revert "stop_machine: Use raw spinlocks" Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 14/17] f2fs: Mutex cant be used by down_write_nest_lock() Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 01/17] ptrace: dont open IRQs in ptrace_freeze_traced() too early Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 15/17] rcu/torture: Comment out rcu_bh ops on PREEMPT_RT_FULL Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 09/17] kernel: softirq: unlock with irqs on Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 05/17] net: provide a way to delegate processing a softirq to ksoftirqd Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 16:50 +0100
  [PATCH RT 03/17] preempt-lazy: Add the lazy-preemption check to preempt_schedule() Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 17:00 +0100
  [PATCH RT 11/17] sched,rt: __always_inline preemptible_lazy() Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 17:10 +0100
  [PATCH RT 02/17] net: move xmit_recursion to per-task variable on -RT Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 17:10 +0100
  [PATCH RT 13/17] trace: Use rcuidle version for preemptoff_hist trace point Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 17:10 +0100
  [PATCH RT 12/17] drm,radeon,i915: Use preempt_disable/enable_rt() where recommended Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 17:10 +0100
  [PATCH RT 16/17] kernel: sched: Fix preempt_disable_ip recodring for preempt_disable() Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 17:10 +0100
  [PATCH RT 07/17] arm64: replace read_lock to rcu lock in call_step_hook Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 17:10 +0100
    Re: [PATCH RT 07/17] arm64: replace read_lock to rcu lock in  call_step_hook Pratyush Anand <panand@redhat.com> - 2016-03-03 05:30 +0100
      Re: [PATCH RT 07/17] arm64: replace read_lock to rcu lock in  call_step_hook Steven Rostedt <rostedt@goodmis.org> - 2016-03-03 15:20 +0100
        Re: [PATCH RT 07/17] arm64: replace read_lock to rcu lock in  call_step_hook Steven Rostedt <rostedt@goodmis.org> - 2016-03-03 15:30 +0100
  [PATCH RT 17/17] Linux 3.14.61-rt64-rc1 Steven Rostedt <rostedt@goodmis.org> - 2016-03-02 17:10 +0100

csiph-web