[PATCH V2 1/2] tools/perf/tests: perf all metrics test fails when perf_event access is restricted

Athira Rajeev atrajeev at linux.vnet.ibm.com
Fri Aug 4 15:00:46 AEST 2023


Perf all metrics test fails as below when perf_event access
is restricted.

   ./perf test -v "perf all metrics test"
   Metric 'Memory_RD_BW_Chip' not printed in:
   Error:
   Access to performance monitoring and observability operations is limited.
   Enforced MAC policy settings (SELinux) can limit access to performance
   —
   access to performance monitoring and observability operations for processes
   without CAP_PERFMON, CAP_SYS_PTRACE or CAP_SYS_ADMIN Linux capability.
   —
   test child finished with -1
   ---- end ----
   perf all metrics test: FAILED!

The perf all metrics test picks the input events from
"perf list --raw-dump metrics" and runs "perf stat -M "$m""
for each of the metrics in the list. It fails here for some
of the metrics which needs access, since it collects system
wide resource details/statistics. Fix the testcase to skip
those metric events. The check is added at two places:
- When metric is run for workload monitoring
- when metric is run for system wide.

It could happen that some of the metric events are not valid
in per-thread mode. Example from an x86 system:

    $ ./perf stat -M smi_cycles true
    Error:
    Invalid event (msr/smi/u) in per-thread mode, enable system wide with '-a'.

The test fallbacks to system wide if first stage fails.
But when run with system wide for this metric, it hits the
issue with access restriction.

    $ ./perf stat -M smi_cycles -a sleep 0.1
    Error:
    Access to performance monitoring and observability operations is limited.

Similar some events report not supported while running with system wide.

Example from an x86 system:

   $ ./perf stat -M tma_info_system_socket_clks -a sleep 0.1

   Performance counter stats for 'system wide':

    <not supported>      cbox_0/event=0x0/u

       0.102633747 seconds time elapsed

Hence the checks for unsupported events and access restrictions is
added for both cases.

Signed-off-by: Athira Rajeev <atrajeev at linux.vnet.ibm.com>
---
Changelog:
v1 -> v2:
 Added the check for access restriction in workload as
 well as system wide check.

 tools/perf/tests/shell/stat_all_metrics.sh | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/tools/perf/tests/shell/stat_all_metrics.sh b/tools/perf/tests/shell/stat_all_metrics.sh
index 54774525e18a..b26420885560 100755
--- a/tools/perf/tests/shell/stat_all_metrics.sh
+++ b/tools/perf/tests/shell/stat_all_metrics.sh
@@ -1,18 +1,18 @@
 #!/bin/bash
 # perf all metrics test
 # SPDX-License-Identifier: GPL-2.0
-
 err=0
+//list="tma_info_system_socket_clks"
 for m in $(perf list --raw-dump metrics); do
   echo "Testing $m"
   result=$(perf stat -M "$m" true 2>&1)
-  if [[ "$result" =~ ${m:0:50} ]] || [[ "$result" =~ "<not supported>" ]]
+  if [[ "$result" =~ ${m:0:50} ]] || [[ "$result" =~ "<not supported>" ]] || [[ "$result" =~ "Access to performance monitoring and observability operations is limited" ]]
   then
     continue
   fi
   # Failed so try system wide.
   result=$(perf stat -M "$m" -a sleep 0.01 2>&1)
-  if [[ "$result" =~ ${m:0:50} ]]
+  if [[ "$result" =~ ${m:0:50} ]] || [[ "$result" =~ "<not supported>" ]] ||  [[ "$result" =~ "Access to performance monitoring and observability operations is limited" ]]
   then
     continue
   fi
-- 
2.31.1



More information about the Linuxppc-dev mailing list