[PATCH RT] Remove double warning print with faulty preempt disable nesting

[PATCH RT] Remove double warning print with faulty preempt disable nesting

[Steven Rostedt]

On a faulty preempt disable bug we print out a warning but then we go
and call preempt_enable_no_resched, which also can print out a warning.
This is redundant.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

Index: linux-2.6.21-rt9/lib/smp_processor_id.c
===================================================================
--- linux-2.6.21-rt9.orig/lib/smp_processor_id.c
+++ linux-2.6.21-rt9/lib/smp_processor_id.c
@@ -47,7 +47,11 @@ unsigned int notrace debug_smp_processor
 	dump_stack();
 
 out_enable:
-	preempt_enable_no_resched();
+	/*
+	 * Don't print message if we enable to zero.
+	 * We arleady printed a warning up above.
+	 */
+	__preempt_enable_no_resched();
 out:
 	return this_cpu;
 }

Should I apply the forcedeth-rt-tweak?

[Free Ekanayaka]

Hi all,

I've noticed that since patch-2.6.21.4-rt13 (included) the
forcedeth-rt-tweak patch

http://people.redhat.com/mingo/realtime-preempt/testing/forcedeth-rt-tweak.patch

has been drop from the main rt patch. Does this mean that it should be
no longer necessary?

This little patch is supposed to prevent nVidia ethernet cards to
permanently suck about 3% of the CPU, even though it didn't work for
me. So far I've tried with

2.6.21 + 2.6.21.3 patch + rt9 patch (which includes the nforce patch)

2.6.21 + 2.6.21.4 patch + rt13 patch + nforce patch

Thanks,

Free