Skip to contentSkip to conversation

tracing, memcg, vmscan: Hide trace events when not in use

PATCH In review Steven Rostedt 2 participants 3 messages Oct 12, 2017 – Oct 13, 2017 Tracing

lore
Share & tools
lore

Steven Rostedt

author maintainer
#0 lore raw

PATCH 1/1 tracing, memcg, vmscan: Hide trace events when not in use

From: Steven Rostedt (VMware) <[email protected]>

When trace events are defined but not used they still create data structures and functions for their use, even though nothing may be using them.

The trace events mm_vmscan_memcg_reclaim_begin, mm_vmscan_memcg_softlimit_reclaim_begin, mm_vmscan_memcg_reclaim_end, and mm_vmscan_memcg_softlimit_reclaim_end are not used if CONFIG_MEMCG is not defined. Do not create these trace events unless CONFIG_MEMCG is defined.

---

include/trace/events/vmscan.h +4 -0
Index: linux-trace.git/include/trace/events/vmscan.h
===================================================================
@@ -133,6 +133,7 @@ DEFINE_EVENT(mm_vmscan_direct_reclaim_be
 	TP_ARGS(order, may_writepage, gfp_flags, classzone_idx)
 );
 
+#ifdef CONFIG_MEMCG
 DEFINE_EVENT(mm_vmscan_direct_reclaim_begin_template, mm_vmscan_memcg_reclaim_begin,
 
 	TP_PROTO(int order, int may_writepage, gfp_t gfp_flags, int classzone_idx),
@@ -146,6 +147,7 @@ DEFINE_EVENT(mm_vmscan_direct_reclaim_be
 
 	TP_ARGS(order, may_writepage, gfp_flags, classzone_idx)
 );
+#endif /* CONFIG_MEMCG */
 
 DECLARE_EVENT_CLASS(mm_vmscan_direct_reclaim_end_template,
 
@@ -171,6 +173,7 @@ DEFINE_EVENT(mm_vmscan_direct_reclaim_en
 	TP_ARGS(nr_reclaimed)
 );
 
+#ifdef CONFIG_MEMCG
 DEFINE_EVENT(mm_vmscan_direct_reclaim_end_template, mm_vmscan_memcg_reclaim_end,
 
 	TP_PROTO(unsigned long nr_reclaimed),
@@ -184,6 +187,7 @@ DEFINE_EVENT(mm_vmscan_direct_reclaim_en
 
 	TP_ARGS(nr_reclaimed)
 );
+#endif /* CONFIG_MEMCG */
 
 TRACE_EVENT(mm_shrink_slab_start,
 	TP_PROTO(struct shrinker *shr, struct shrink_control *sc,
Acked-by on v1 1/1
#1 lore raw
Quoted 11 lines · Steven Rostedt

From: Steven Rostedt (VMware) <[email protected]>

When trace events are defined but not used they still create data structures and functions for their use, even though nothing may be using them.

The trace events mm_vmscan_memcg_reclaim_begin, mm_vmscan_memcg_softlimit_reclaim_begin, mm_vmscan_memcg_reclaim_end, and mm_vmscan_memcg_softlimit_reclaim_end are not used if CONFIG_MEMCG is not defined. Do not create these trace events unless CONFIG_MEMCG is defined.

I do not understand the ftrace machinery and I have to admit I was under impression that nothing really gets defined if the code using them doesn't exist but if this is not the case then this patch makes sense to me

Quoted 1 line
Quoted 37 lines

---
Index: linux-trace.git/include/trace/events/vmscan.h
===================================================================

include/trace/events/vmscan.h +4 -0
@@ -133,6 +133,7 @@ DEFINE_EVENT(mm_vmscan_direct_reclaim_be
 	TP_ARGS(order, may_writepage, gfp_flags, classzone_idx)
 );
 
+#ifdef CONFIG_MEMCG
 DEFINE_EVENT(mm_vmscan_direct_reclaim_begin_template, mm_vmscan_memcg_reclaim_begin,
 
 	TP_PROTO(int order, int may_writepage, gfp_t gfp_flags, int classzone_idx),
@@ -146,6 +147,7 @@ DEFINE_EVENT(mm_vmscan_direct_reclaim_be
 
 	TP_ARGS(order, may_writepage, gfp_flags, classzone_idx)
 );
+#endif /* CONFIG_MEMCG */
 
 DECLARE_EVENT_CLASS(mm_vmscan_direct_reclaim_end_template,
 
@@ -171,6 +173,7 @@ DEFINE_EVENT(mm_vmscan_direct_reclaim_en
 	TP_ARGS(nr_reclaimed)
 );
 
+#ifdef CONFIG_MEMCG
 DEFINE_EVENT(mm_vmscan_direct_reclaim_end_template, mm_vmscan_memcg_reclaim_end,
 
 	TP_PROTO(unsigned long nr_reclaimed),
@@ -184,6 +187,7 @@ DEFINE_EVENT(mm_vmscan_direct_reclaim_en
 
 	TP_ARGS(nr_reclaimed)
 );
+#endif /* CONFIG_MEMCG */
 
 TRACE_EVENT(mm_shrink_slab_start,
 	TP_PROTO(struct shrinker *shr, struct shrink_control *sc,
Signature · 2 lines
Michal Hocko
SUSE Labs
author maintainer
#2 lore raw
Quoted 1 line · On Fri, 13 Oct 2017 08:38:26 +0200 Michal Hocko <[email protected]> wrote:

I do not understand the ftrace machinery

Not many do :-/

Quoted 4 lines

and I have to admit I was under
impression that nothing really gets defined if the code using them doesn't exist but if this is not the case then this patch makes sense to me

Sometimes events are created by kernel proper that are used by modules, or created by modules to be used by other modules (kvm does this with kvm_intel and kvm_amd). Unfortunately, for these cases they need to stay. Or later, I plan on moving them to where they are used, because they have the same issue (wasting memory).

Quoted 4 lines
Quoted 2 lines

Thanks!

-- Steve