[PATCH 2/3] mm/cma: add new tracepoint, test_pages_isolated

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



cma allocation should be guranteeded to succeed, but, sometimes,
it could be failed in current implementation. To track down
the problem, we need to know which page is problematic and
this new tracepoint will report it.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
---
 include/trace/events/cma.h | 26 ++++++++++++++++++++++++++
 mm/page_isolation.c        |  5 +++++
 2 files changed, 31 insertions(+)

diff --git a/include/trace/events/cma.h b/include/trace/events/cma.h
index d7cd961..82281b0 100644
--- a/include/trace/events/cma.h
+++ b/include/trace/events/cma.h
@@ -60,6 +60,32 @@ TRACE_EVENT(cma_release,
 		  __entry->count)
 );
 
+TRACE_EVENT(test_pages_isolated,
+
+	TP_PROTO(
+		unsigned long start_pfn,
+		unsigned long end_pfn,
+		unsigned long fin_pfn),
+
+	TP_ARGS(start_pfn, end_pfn, fin_pfn),
+
+	TP_STRUCT__entry(
+		__field(unsigned long, start_pfn)
+		__field(unsigned long, end_pfn)
+		__field(unsigned long, fin_pfn)
+	),
+
+	TP_fast_assign(
+		__entry->start_pfn = start_pfn;
+		__entry->end_pfn = end_pfn;
+		__entry->fin_pfn = fin_pfn;
+	),
+
+	TP_printk("start_pfn=0x%lx end_pfn=0x%lx fin_pfn=0x%lx ret=%s",
+		__entry->start_pfn, __entry->end_pfn, __entry->fin_pfn,
+		__entry->end_pfn == __entry->fin_pfn ? "success" : "fail")
+);
+
 #endif /* _TRACE_CMA_H */
 
 /* This part must be outside protection */
diff --git a/mm/page_isolation.c b/mm/page_isolation.c
index 6f5ae96..bda0fea 100644
--- a/mm/page_isolation.c
+++ b/mm/page_isolation.c
@@ -7,6 +7,8 @@
 #include <linux/pageblock-flags.h>
 #include <linux/memory.h>
 #include <linux/hugetlb.h>
+#include <trace/events/cma.h>
+
 #include "internal.h"
 
 static int set_migratetype_isolate(struct page *page,
@@ -268,6 +270,9 @@ int test_pages_isolated(unsigned long start_pfn, unsigned long end_pfn,
 						skip_hwpoisoned_pages);
 	spin_unlock_irqrestore(&zone->lock, flags);
 
+#ifdef CONFIG_CMA
+	trace_test_pages_isolated(start_pfn, end_pfn, pfn);
+#endif
 	return (pfn < end_pfn) ? -EBUSY : 0;
 }
 
-- 
1.9.1

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>



[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux]     [Linux OMAP]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]