[PATCH AUTOSEL for 4.9 221/281] perf/core: Correct event creation with PERF_FORMAT_GROUP

From: Sasha Levin
Date: Mon Mar 19 2018 - 13:53:17 EST


From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>

[ Upstream commit ba5213ae6b88fb170c4771fef6553f759c7d8cdd ]

Andi was asking about PERF_FORMAT_GROUP vs inherited events, which led
to the discovery of a bug from commit:

3dab77fb1bf8 ("perf: Rework/fix the whole read vs group stuff")

- PERF_SAMPLE_GROUP = 1U << 4,
+ PERF_SAMPLE_READ = 1U << 4,

- if (attr->inherit && (attr->sample_type & PERF_SAMPLE_GROUP))
+ if (attr->inherit && (attr->read_format & PERF_FORMAT_GROUP))

is a clear fail :/

While this changes user visible behaviour; it was previously possible
to create an inherited event with PERF_SAMPLE_READ; this is deemed
acceptible because its results were always incorrect.

Reported-by: Andi Kleen <ak@xxxxxxxxxxxxxxx>
Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
Cc: Alexander Shishkin <alexander.shishkin@xxxxxxxxxxxxxxx>
Cc: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
Cc: Jiri Olsa <jolsa@xxxxxxxxxx>
Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Stephane Eranian <eranian@xxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Vince Weaver <vince@xxxxxxxxxx>
Fixes: 3dab77fb1bf8 ("perf: Rework/fix the whole read vs group stuff")
Link: http://lkml.kernel.org/r/20170530094512.dy2nljns2uq7qa3j@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
Signed-off-by: Sasha Levin <alexander.levin@xxxxxxxxxxxxx>
---
kernel/events/core.c | 15 ++++++++++-----
1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/kernel/events/core.c b/kernel/events/core.c
index d7914fa6bc09..c4100c38a467 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -5669,9 +5669,6 @@ static void perf_output_read_one(struct perf_output_handle *handle,
__output_copy(handle, values, n * sizeof(u64));
}

-/*
- * XXX PERF_FORMAT_GROUP vs inherited events seems difficult.
- */
static void perf_output_read_group(struct perf_output_handle *handle,
struct perf_event *event,
u64 enabled, u64 running)
@@ -5716,6 +5713,13 @@ static void perf_output_read_group(struct perf_output_handle *handle,
#define PERF_FORMAT_TOTAL_TIMES (PERF_FORMAT_TOTAL_TIME_ENABLED|\
PERF_FORMAT_TOTAL_TIME_RUNNING)

+/*
+ * XXX PERF_SAMPLE_READ vs inherited events seems difficult.
+ *
+ * The problem is that its both hard and excessively expensive to iterate the
+ * child list, not to mention that its impossible to IPI the children running
+ * on another CPU, from interrupt/NMI context.
+ */
static void perf_output_read(struct perf_output_handle *handle,
struct perf_event *event)
{
@@ -9259,9 +9263,10 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu,
local64_set(&hwc->period_left, hwc->sample_period);

/*
- * we currently do not support PERF_FORMAT_GROUP on inherited events
+ * We currently do not support PERF_SAMPLE_READ on inherited events.
+ * See perf_output_read().
*/
- if (attr->inherit && (attr->read_format & PERF_FORMAT_GROUP))
+ if (attr->inherit && (attr->sample_type & PERF_SAMPLE_READ))
goto err_ns;

if (!has_branch_stack(event))
--
2.14.1