ftrace: Synchronize the initialization of ftrace_opsgrafted

There's some internal state that ftrace_ops needs to have set, but since
it can be declared outside of the ftrace.c code, it calls
ftrace_ops_init() on the ops in every global function. The issue is that
if two tasks call it on the same ops at the same time it is possible to
have the initialization of one corrupt the initialization of the other
call.

Create a ops_mutex to use to synchronize every initialization of the
ftrace_ops. The mutex is taken within checking the ftrace_ops flag that
states it was initializied but the flag is checked again after the mutex
has been taken. Checking first outside the mutex allows it to shortcut
having to take the mutex. But then the check needs to be done again after
the mute is taken in case of races.

Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260902095501.6b59af20@gandalf.local.home
Fixes: f04f24fb7e48d ("ftrace, kprobes: Fix a deadlock on ftrace_regex_lock")
Reported-by: sashiko-bot@kernel.org
Close: https://lore.kernel.org/all/20260829025528.49A831F000E9@smtp.kernel.org/
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

2 files changed, 94 insertions, 0 deletions

diff --git a/tools/tracing/rtla/example/timerlat_bpf_action.c b/tools/tracing/rtla/example/timerlat_bpf_action.c
new file mode 100644
index 000000000..ac1be049a
--- /dev/null
+++ b/tools/tracing/rtla/example/timerlat_bpf_action.c
@@ -0,0 +1,16 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <linux/bpf.h>
+#include <bpf/bpf_tracing.h>
+
+char LICENSE[] SEC("license") = "GPL";
+
+struct trace_event_raw_timerlat_sample {
+	unsigned long long timer_latency;
+} __attribute__((preserve_access_index));
+
+SEC("tp/timerlat_action")
+int action_handler(struct trace_event_raw_timerlat_sample *tp_args)
+{
+	bpf_printk("Latency: %lld\n", tp_args->timer_latency);
+	return 0;
+}
diff --git a/tools/tracing/rtla/example/timerlat_load.py b/tools/tracing/rtla/example/timerlat_load.py
new file mode 100644
index 000000000..a819c3588
--- /dev/null
+++ b/tools/tracing/rtla/example/timerlat_load.py
@@ -0,0 +1,78 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Copyright (C) 2024 Red Hat, Inc. Daniel Bristot de Oliveira <bristot@kernel.org>
+#
+# This is a sample code about how to use timerlat's timer by any workload
+# so rtla can measure and provide auto-analysis for the overall latency (IOW
+# the response time) for a task.
+#
+# Before running it, you need to dispatch timerlat with -U option in a terminal.
+# Then # run this script pinned to a CPU on another terminal. For example:
+#
+# timerlat_load.py 1 -p 95
+#
+# The "Timerlat IRQ" is the IRQ latency, The thread latency is the latency
+# for the python process to get the CPU. The Ret from user Timer Latency is
+# the overall latency. In other words, it is the response time for that
+# activation.
+#
+# This is just an example, the load is reading 20MB of data from /dev/full
+# It is in python because it is easy to read :-)
+
+import argparse
+import sys
+import os
+
+parser = argparse.ArgumentParser(description='user-space timerlat thread in Python')
+parser.add_argument("cpu", type=int, help='CPU to run timerlat thread')
+parser.add_argument("-p", "--prio", type=int, help='FIFO priority')
+args = parser.parse_args()
+
+try:
+    affinity_mask = {args.cpu}
+    os.sched_setaffinity(0, affinity_mask)
+except Exception as e:
+    print(f"Error setting affinity: {e}")
+    sys.exit(1)
+
+if args.prio:
+    try:
+        param = os.sched_param(args.prio)
+        os.sched_setscheduler(0, os.SCHED_FIFO, param)
+    except Exception as e:
+        print(f"Error setting priority: {e}")
+        sys.exit(1)
+
+try:
+    timerlat_path = f"/sys/kernel/tracing/osnoise/per_cpu/cpu{args.cpu}/timerlat_fd"
+    timerlat_fd = open(timerlat_path, 'r')
+except PermissionError:
+    print("Permission denied. Please check your access rights.")
+    sys.exit(1)
+except OSError:
+    print("Error opening timerlat fd, did you run timerlat -U?")
+    sys.exit(1)
+
+try:
+    data_fd = open("/dev/full", 'r')
+except Exception as e:
+    print(f"Error opening data fd: {e}")
+    sys.exit(1)
+
+while True:
+    try:
+        timerlat_fd.read(1)
+        data_fd.read(20 * 1024 * 1024)
+    except KeyboardInterrupt:
+        print("Leaving")
+        break
+    except IOError as e:
+        print(f"I/O error occurred: {e}")
+        break
+    except Exception as e:
+        print(f"Unexpected error: {e}")
+        break
+
+timerlat_fd.close()
+data_fd.close()