Re: [PATCH 1/2] tracing/mm: don't trace kfree on offline CPUs

From: Steven Rostedt
Date: Mon Feb 15 2016 - 12:33:05 EST


On Sun, 14 Feb 2016 23:19:42 +0300
Denis Kirjanov <kda@xxxxxxxxxxxxxxxxx> wrote:


> >> TP_PROTO(unsigned long call_site, const void *ptr),
> >>
> >> - TP_ARGS(call_site, ptr)
> >> + TP_ARGS(call_site, ptr),
> >> +
> >> + /*
> >> + * This trace can be potentially called from an offlined cpu.
> >> + * Since trace points use RCU and RCU should not be used from
> >> + * offline cpus, filter such calls out.
> >> + * While this trace can be called from a preemptable section,
> >> + * it has no impact on the condition since tasks can migrate
> >> + * only from online cpus to other online cpus. Thus its safe
> >> + * to use raw_smp_processor_id.
> >> + */
> >> + TP_CONDITION(cpu_online(raw_smp_processor_id()))
> >
> > This is starting to become a common occurrence. Perhaps it is best to
> > just hardcode this into the tracepoint code itself?
>
> Can you take it as a fix for now. I'll post the follow-up patch then
> for rcu and offline cpus
> issue.

Actually, I prefer not to add more checks for cpu_online(), if its
going to become the default.

Does this work for you instead?

-- Steve

diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h
index acd522a91539..acfdbf353a0b 100644
--- a/include/linux/tracepoint.h
+++ b/include/linux/tracepoint.h
@@ -14,8 +14,10 @@
* See the file COPYING for more details.
*/

+#include <linux/smp.h>
#include <linux/errno.h>
#include <linux/types.h>
+#include <linux/cpumask.h>
#include <linux/rcupdate.h>
#include <linux/tracepoint-defs.h>

@@ -132,6 +134,9 @@ extern void syscall_unregfunc(void);
void *it_func; \
void *__data; \
\
+ if (!cpu_online(raw_smp_processor_id())) \
+ return; \
+ \
if (!(cond)) \
return; \
prercu; \