irq: Justify the various softirq stack choices
authorFrederic Weisbecker <fweisbec@gmail.com>
Thu, 5 Sep 2013 14:14:00 +0000 (16:14 +0200)
committerFrederic Weisbecker <fweisbec@gmail.com>
Tue, 1 Oct 2013 10:53:27 +0000 (12:53 +0200)
For clarity, comment the various stack choices for softirqs
processing, whether we execute them from ksoftirqd or
local_irq_enable() calls.

Their use on irq_exit() is already commented.

Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@au1.ibm.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul Mackerras <paulus@au1.ibm.com>
Cc: James Hogan <james.hogan@imgtec.com>
Cc: James E.J. Bottomley <jejb@parisc-linux.org>
Cc: Helge Deller <deller@gmx.de>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
kernel/softirq.c

index 9f8092b82a948bc9227779ddd6dc3b03219c60dd..2b4328ea769fb07ef0a97e9ce6059b8573ffd7d7 100644 (file)
@@ -170,8 +170,13 @@ static inline void _local_bh_enable_ip(unsigned long ip)
         */
        sub_preempt_count(SOFTIRQ_DISABLE_OFFSET - 1);
 
-       if (unlikely(!in_interrupt() && local_softirq_pending()))
+       if (unlikely(!in_interrupt() && local_softirq_pending())) {
+               /*
+                * Run softirq if any pending. And do it in its own stack
+                * as we may be calling this deep in a task call stack already.
+                */
                do_softirq();
+       }
 
        dec_preempt_count();
 #ifdef CONFIG_TRACE_IRQFLAGS
@@ -769,6 +774,10 @@ static void run_ksoftirqd(unsigned int cpu)
 {
        local_irq_disable();
        if (local_softirq_pending()) {
+               /*
+                * We can safely run softirq on inline stack, as we are not deep
+                * in the task stack here.
+                */
                __do_softirq();
                rcu_note_context_switch(cpu);
                local_irq_enable();