examples/perf/ipc: Use generic hardware events

The use of hardcoded raw events breaks compatibility as these are tied
to specific processors. On other processors, it could also be the case
that the raw event codes represent something completely different.

Hence, use the generic events for cycles and instructions as these are
available on most platforms with a core PMU.

Fixes: 59a753d ("Add perf event data collection example for an userspace application")
Signed-off-by: Sandipan Das <[email protected]>

examples/perf/ipc.py

@@ -158,20 +158,22 @@ def print_data(cpu, data, size):
 
 b["output"].open_perf_buffer(print_data)
 
-# Perf Event for Unhalted Cycles, The hex value is
-# combination of event, umask and cmask. Read Intel
-# Doc to find the event and cmask. Or use 
-# perf list --details to get event, umask and cmask
+# Perf Events for Unhalted Cycles and Retired Instructions are supported on
+# most platforms with a PMU and the kernel will attempt to translate these
+# into an architecture-specific event code. For architectures that do not have
+# these mappings, see perf list --details to find event details.
 # NOTE: Events can be multiplexed by kernel in case the
 # number of counters is greater than supported by CPU
 # performance monitoring unit, which can result in inaccurate
 # results. Counter values need to be normalized for a more
 # accurate value.
-PERF_TYPE_RAW = 4
+PERF_TYPE_HARDWARE = 0
+PERF_COUNT_HW_CPU_CYCLES = 0
+PERF_COUNT_HW_INSTRUCTIONS = 1
 # Unhalted Clock Cycles
-b["clk"].open_perf_event(PERF_TYPE_RAW, 0x0000003C)
+b["clk"].open_perf_event(PERF_TYPE_HARDWARE, PERF_COUNT_HW_CPU_CYCLES)
 # Instruction Retired
-b["inst"].open_perf_event(PERF_TYPE_RAW, 0x000000C0)
+b["inst"].open_perf_event(PERF_TYPE_HARDWARE, PERF_COUNT_HW_INSTRUCTIONS)
 
 while True:
  try: