linux-kernel - [PATCH] sched_ext: Provide a sysfs enable

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [day] [month] [year] [list]
Message-ID: <20240921193921.75594-1-andrea.righi@linux.dev>
Date: Sat, 21 Sep 2024 21:39:21 +0200
From: andrea.righi@...ux.dev
To: Tejun Heo <tj@...nel.org>,
	David Vernet <void@...ifault.com>
Cc: Giovanni Gherdovich <giovanni.gherdovich@...e.com>,
	Kleber Sacilotto de Souza <kleber.souza@...onical.com>,
	Marcelo Henrique Cerri <marcelo.cerri@...onical.com>,
	Phil Auld <pauld@...hat.com>,
	linux-kernel@...r.kernel.org,
	Andrea Righi <andrea.righi@...ux.dev>
Subject: [PATCH] sched_ext: Provide a sysfs enable_seq counter

From: Andrea Righi <andrea.righi@...ux.dev>

As discussed during the distro-centric session within the sched_ext
Microconference at LPC 2024, introduce a sequence counter that is
incremented every time a BPF scheduler is loaded.

This feature can help distributions in diagnosing potential performance
regressions by identifying systems where users are running (or have ran)
custom BPF schedulers.

Example:

 arighi@...tme-ng~> cat /sys/kernel/sched_ext/enable_seq
 0
 arighi@...tme-ng~> sudo scx_simple
 local=1 global=0
 ^CEXIT: unregistered from user space
 arighi@...tme-ng~> cat /sys/kernel/sched_ext/enable_seq
 1

In this way user-space tools (such as Ubuntu's apport and similar) are
able to gather and include this information in bug reports.

Cc: Giovanni Gherdovich <giovanni.gherdovich@...e.com>
Cc: Kleber Sacilotto de Souza <kleber.souza@...onical.com>
Cc: Marcelo Henrique Cerri <marcelo.cerri@...onical.com>
Cc: Phil Auld <pauld@...hat.com>
Signed-off-by: Andrea Righi <andrea.righi@...ux.dev>
---
 Documentation/scheduler/sched-ext.rst | 10 ++++++++++
 kernel/sched/ext.c                    | 17 +++++++++++++++++
 tools/sched_ext/scx_show_state.py     |  1 +
 3 files changed, 28 insertions(+)

diff --git a/Documentation/scheduler/sched-ext.rst b/Documentation/scheduler/sched-ext.rst
index a707d2181a77..6c0d70e2e27d 100644
--- a/Documentation/scheduler/sched-ext.rst
+++ b/Documentation/scheduler/sched-ext.rst
@@ -83,6 +83,15 @@ The current status of the BPF scheduler can be determined as follows:
     # cat /sys/kernel/sched_ext/root/ops
     simple
 
+You can check if any BPF scheduler has ever been loaded since boot by examining
+this monotonically incrementing counter (a value of zero indicates that no BPF
+scheduler has been loaded):
+
+.. code-block:: none
+
+    # cat /sys/kernel/sched_ext/enable_seq
+    1
+
 ``tools/sched_ext/scx_show_state.py`` is a drgn script which shows more
 detailed information:
 
@@ -96,6 +105,7 @@ detailed information:
     enable_state  : enabled (2)
     bypass_depth  : 0
     nr_rejected   : 0
+    enable_seq    : 1
 
 If ``CONFIG_SCHED_DEBUG`` is set, whether a given task is on sched_ext can
 be determined as follows:
diff --git a/kernel/sched/ext.c b/kernel/sched/ext.c
index 9ee5a9a261cc..8057ab4c76da 100644
--- a/kernel/sched/ext.c
+++ b/kernel/sched/ext.c
@@ -874,6 +874,13 @@ static struct scx_exit_info *scx_exit_info;
 static atomic_long_t scx_nr_rejected = ATOMIC_LONG_INIT(0);
 static atomic_long_t scx_hotplug_seq = ATOMIC_LONG_INIT(0);
 
+/*
+ * A monotically increasing sequence number that is incremented every time a
+ * scheduler is enabled. This can be used by to check if any custom sched_ext
+ * scheduler has ever been used in the system.
+ */
+static atomic_long_t scx_enable_seq = ATOMIC_LONG_INIT(0);
+
 /*
  * The maximum amount of time in jiffies that a task may be runnable without
  * being scheduled on a CPU. If this timeout is exceeded, it will trigger
@@ -4154,11 +4161,19 @@ static ssize_t scx_attr_hotplug_seq_show(struct kobject *kobj,
 }
 SCX_ATTR(hotplug_seq);
 
+static ssize_t scx_attr_enable_seq_show(struct kobject *kobj,
+					struct kobj_attribute *ka, char *buf)
+{
+	return sysfs_emit(buf, "%ld\n", atomic_long_read(&scx_enable_seq));
+}
+SCX_ATTR(enable_seq);
+
 static struct attribute *scx_global_attrs[] = {
 	&scx_attr_state.attr,
 	&scx_attr_switch_all.attr,
 	&scx_attr_nr_rejected.attr,
 	&scx_attr_hotplug_seq.attr,
+	&scx_attr_enable_seq.attr,
 	NULL,
 };
 
@@ -5176,6 +5191,8 @@ static int scx_ops_enable(struct sched_ext_ops *ops, struct bpf_link *link)
 	kobject_uevent(scx_root_kobj, KOBJ_ADD);
 	mutex_unlock(&scx_ops_enable_mutex);
 
+	atomic_long_inc(&scx_enable_seq);
+
 	return 0;
 
 err_del:
diff --git a/tools/sched_ext/scx_show_state.py b/tools/sched_ext/scx_show_state.py
index d457d2a74e1e..8bc626ede1c4 100644
--- a/tools/sched_ext/scx_show_state.py
+++ b/tools/sched_ext/scx_show_state.py
@@ -37,3 +37,4 @@ print(f'switched_all  : {read_static_key("__scx_switched_all")}')
 print(f'enable_state  : {ops_state_str(enable_state)} ({enable_state})')
 print(f'bypass_depth  : {read_atomic("scx_ops_bypass_depth")}')
 print(f'nr_rejected   : {read_atomic("scx_nr_rejected")}')
+print(f'enable_seq    : {read_atomic("scx_enable_seq")}')
-- 
2.46.0