Re: [PATCH 2/2] mm/page_ref: add tracepoint to track down page reference manipulation

From: Sergey Senozhatsky
Date: Thu Feb 18 2016 - 19:33:13 EST


On (02/18/16 09:29), Steven Rostedt wrote:
> > diff --git a/include/linux/page_ref.h b/include/linux/page_ref.h
> > index 534249c..fd6d9a5 100644
> > --- a/include/linux/page_ref.h
> > +++ b/include/linux/page_ref.h
> > @@ -1,6 +1,54 @@
> > #include <linux/atomic.h>
> > #include <linux/mm_types.h>
> > #include <linux/page-flags.h>
> > +#include <linux/tracepoint-defs.h>
> > +
> > +extern struct tracepoint __tracepoint_page_ref_set;
> > +extern struct tracepoint __tracepoint_page_ref_mod;
> > +extern struct tracepoint __tracepoint_page_ref_mod_and_test;
> > +extern struct tracepoint __tracepoint_page_ref_mod_and_return;
> > +extern struct tracepoint __tracepoint_page_ref_mod_unless;
> > +extern struct tracepoint __tracepoint_page_ref_freeze;
> > +extern struct tracepoint __tracepoint_page_ref_unfreeze;
> > +
> > +#ifdef CONFIG_DEBUG_PAGE_REF
>
> Please add a comment here. Something to the effect of:
>
> /*
> * Ideally we would want to use the trace_<tracepoint>_enabled() helper
> * functions. But due to include header file issues, that is not
> * feasible. Instead we have to open code the static key functions.
> *
> * See trace_##name##_enabled(void) in include/linux/tracepoint.h
> */
>

not sure if it's worth mentioning in the comment, but the other
concern here is the performance impact of an extra function call,
I believe. otherwise, Joonsoo would just do:

in include/linux/page_ref.h

static inline void set_page_count(struct page *page, int v)
{
atomic_set(&page->_count, v);
__page_ref_set(page, v);
}
...



and in mm/debug_page_ref.c

void __page_ref_set(struct page *page, int v)
{
if (trace_page_ref_set_enabled())
trace_page_ref_set(page, v);
}
EXPORT_SYMBOL(__page_ref_set);
EXPORT_TRACEPOINT_SYMBOL(page_ref_set);
...

-ss