Thread (5 messages) 5 messages, 2 authors, 2012-05-21
STALE5084d

[PATCH RT 2/2] fix printk flush of messages

From: Frank Rowand <hidden>
Date: 2012-05-17 01:10:06
Also in: lkml

Updates console-make-rt-friendly.patch

#ifdef CONFIG_PREEMPT_RT_FULL, printk() output is never flushed by
printk() because:

   # some liberties taken in this pseudo-code to make it easier to follow
   printk()
      vprintk()
         raw_spin_lock(&logbuf_lock)
            # increment preempt_count():
            preempt_disable()
      result = console_trylock_for_printk()
         retval = 0
         # lock will always be false, because preempt_count() will be >= 1
         lock = ... && !preempt_count()
         if (lock)
            retval = 1
         return retval
      # result will always be false since lock will always be false
      if (result)
         console_unlock()
            # this is where the printk() output would be flushed


On system boot some printk() output is flushed because register_console()
and tty_open() call console_unlock().


This change also fixes the problem that was previously fixed by
preempt-rt-allow-immediate-magic-sysrq-output-for-preempt_rt_full.patch

Signed-off-by: Frank Rowand <redacted>

---
 kernel/printk.c |    2 	1 +	1 -	0 !
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: b/kernel/printk.c
===================================================================
--- a/kernel/printk.c
+++ b/kernel/printk.c
@@ -847,7 +847,7 @@ static int console_trylock_for_printk(un
 	int retval = 0, wake = 0;
 #ifdef CONFIG_PREEMPT_RT_FULL
 	int lock = !early_boot_irqs_disabled && !irqs_disabled_flags(flags) &&
-		!preempt_count();
+		(preempt_count() <= 1);
 #else
 	int lock = 1;
 #endif
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help