Skip to content

Improve counter reporting on heterogeneous Arm systems #10

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
kieranhejmadi01 wants to merge 1 commit into
base: main
Choose a base branch
from
+10 −1
Open

Improve counter reporting on heterogeneous Arm systems #10

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 10 additions & 1 deletion src/sysreport.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1083,7 +1083,16 @@ def show(s):
if perf_inst:
print(" perf installed at: %s" % (perf_binary()))
print(" perf with OpenCSD: %s" % (colorize(perf_binary_has_opencsd())))
print(" perf counters: %s" % (s.perf_max_counters()))
counters = s.perf_max_counters()
# On heterogeneous systems with multiple CPU types (e.g. Arm big.LITTLE), perf
# redistributes events across PMUs rather than scheduling them as a single group.
# This means the group saturation point - which is how we detect the counter limit -
# is never reached, so detection returns None even though counters are available.
if counters is None and s.system.is_heterogeneous():
counter_str = "None (heterogeneous system - counter detection unreliable)"
else:
counter_str = str(counters)
print(" perf counters: %s" % (counter_str))
print(" perf sampling: %s" % (colorize_greenred(perf_noninvasive_sampling(s))))
print(" perf HW trace: %s" % (colorize_greenred(perf_hardware_trace(s))))
print(" perf paranoid: %s" % (perf_event_paranoid())) # 0 is not bad, it's good
Expand Down