perf tools: Fix swap for samples with raw data

Luckily the traceevent library handles the endianity by itself (thank you Steven!), so we can pass the RAW data directly in the other endianity.

When we detect different endianity we swap event before processing. It's tricky for samples because we have no idea what's inside. We treat it as an array of u64s, swap them and later on we swap back parts which are different.

We mangle this way also the tracepoint raw data, which ends up in report showing wrong data:

1.95% comm=Q^B pid=29285 prio=16777216 target_cpu=000 1.67% comm=l^B pid=0 prio=16777216 target_cpu=000

Luckily the traceevent library handles the endianity by itself (thank you Steven!), so we can pass the RAW data directly in the other endianity.

2.51% comm=beah-rhts-task pid=1175 prio=120 target_cpu=002 2.23% comm=kworker/0:0 pid=11566 prio=120 target_cpu=000

The fix is basically to swap back the raw data if different endianity is detected.

Cc David Ahern <[email protected]>

Cc Steven Rostedt (VMware) <[email protected]>

Link http://lkml.kernel.org/n/[email protected] open here

Signed-off-by Jiri Olsa <[email protected]>

---
 tools/perf/util/evsel.c | 20 +++++++++++++++++---
 1 file changed, 17 insertions(+), 3 deletions(-)

tools/perf/util/evsel.c +17 -3

index 95853c51c0ca..09c68cf2d9d2 100644
@@ -36,6 +36,7 @@
 #include "debug.h"
 #include "trace-event.h"
 #include "stat.h"
+#include "memswap.h"
 #include "util/parse-branch-options.h"
 
 #include "sane_ctype.h"

@@ -2131,14 +2132,27 @@ int perf_evsel__parse_sample(struct perf_evsel *evsel, union perf_event *event,
 	if (type & PERF_SAMPLE_RAW) {
 		OVERFLOW_CHECK_u64(array);
 		u.val64 = *array;
-		if (WARN_ONCE(swapped,
-			      "Endianness of raw data not corrected!\n")) {
-			/* undo swap of u64, then swap on individual u32s */
+
+		/*
+		 * Undo swap of u64, then swap on individual u32s,
+		 * get the size of the raw area and undo all of the
+		 * swap. The pevent interface handles endianity by
+		 * itself.
+		 */
+		if (swapped) {
 			u.val64 = bswap_64(u.val64);
 			u.val32[0] = bswap_32(u.val32[0]);
 			u.val32[1] = bswap_32(u.val32[1]);
 		}
 		data->raw_size = u.val32[0];
+
+		/*
+		 * The raw data is aligned on 64bits including the
+		 * u32 size, so it's safe to use mem_bswap_64.
+		 */
+		if (swapped)
+			mem_bswap_64((void *) array, data->raw_size);
+
 		array = (void *)array + sizeof(u32);
 
 		OVERFLOW_CHECK(array, data->raw_size, max_size);

Signature · 1 line

2.13.6

The fix is basically to swap back the raw data if different endianity is detected.

+++ b/tools/perf/util/evsel.c

tools/perf/util/evsel.c +1 -0

@@ -36,6 +36,7 @@
 #include "trace-event.h"
 #include "stat.h"
+#include "memswap.h"
 

Em Wed, Nov 29, 2017 at 07:43:46PM +0100, Jiri Olsa escreveu:

Quoted 10 lines

The fix is basically to swap back the raw data if different endianity is detected.

+++ b/tools/perf/util/evsel.c

tools/perf/util/evsel.c +1 -0

@@ -36,6 +36,7 @@
 #include "trace-event.h"
 #include "stat.h"
+#include "memswap.h"
 

tools/perf/util/evsel.c int perf_evsel__parse_sample(struct perf_evsel *evsel, union perf_event *event, · line 2136

Quoted 5 lines

tools/perf/util/evsel.c +3 -0

@@ -2131,14 +2132,27 @@ int perf_evsel__parse_sample(struct perf_evsel *evsel, union perf_event *event,
 	if (type & PERF_SAMPLE_RAW) {
 
+		if (swapped)
+			mem_bswap_64((void *) array, data->raw_size);
+

you forgot to add util/memswap.c to tools/perf/util/python-ext-sources, breaking 'perf test python'.

Please run 'perf test' before sending pull requests :-)

Fixed it this time.