perf_counter: Introduce new (non-)paranoia level to allow raw tracepoint access

I want to sample inherited tracepoint workloads as a normal
user and the CAP_SYS_ADMIN check prevents me from doing that
right now.

Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c
index d988dfb4bbab96d4366cfb219bc0f4bd3fa5e358..0aa609f69103415f31cc42604793eee63928dff9 100644 (file)
--- a/kernel/perf_counter.c
+++ b/kernel/perf_counter.c
@@ -46,12 +46,18 @@ static atomic_t nr_task_counters __read_mostly;
 
 /*
  * perf counter paranoia level:
- *  0 - not paranoid
- *  1 - disallow cpu counters to unpriv
- *  2 - disallow kernel profiling to unpriv
+ *  -1 - not paranoid at all
+ *   0 - disallow raw tracepoint access for unpriv
+ *   1 - disallow cpu counters for unpriv
+ *   2 - disallow kernel profiling for unpriv
  */
 int sysctl_perf_counter_paranoid __read_mostly = 1;
 
+static inline bool perf_paranoid_tracepoint_raw(void)
+{
+       return sysctl_perf_counter_paranoid > -1;
+}
+
 static inline bool perf_paranoid_cpu(void)
 {
        return sysctl_perf_counter_paranoid > 0;
@@ -3971,6 +3977,7 @@ static const struct pmu *tp_perf_counter_init(struct perf_counter *counter)
         * have these.
         */
        if ((counter->attr.sample_type & PERF_SAMPLE_RAW) &&
+                       perf_paranoid_tracepoint_raw() &&
                        !capable(CAP_SYS_ADMIN))
                return ERR_PTR(-EPERM);