From: John Kacur <jkacur@redhat.com>
To: linux-rt-users <linux-rt-users@vger.kernel.org>
Cc: Clark Williams <williams@redhat.com>, Tomas Glozar <tglozar@redhat.com>
Subject: [PATCH 1/7] rteval: Add CoreSiblings class for CPU core topology queries
Date: Tue, 21 Apr 2026 14:26:12 -0400 [thread overview]
Message-ID: <20260421182618.261347-2-jkacur@redhat.com> (raw)
In-Reply-To: <20260421182618.261347-1-jkacur@redhat.com>
Add a new CoreSiblings class in rteval/sysinfo/coresiblings.py that
provides a simple interface for querying which CPUs share physical cores.
The class reads /sys/devices/system/cpu/cpu*/topology/thread_siblings_list
to build a mapping of CPU core siblings and provides three methods:
- share_core(cpu1, cpu2): Check if two CPUs share a physical core
- get_siblings(cpu): Get all CPUs that share a core with a given CPU
- get_core_groups(): Get all unique core sibling groups
The implementation is generic and works with any SMT configuration:
- No SMT (1 CPU per core)
- 2-way SMT (typical Intel hyperthreading)
- 4-way SMT (some IBM POWER processors)
- 8-way SMT (IBM POWER8/9)
This provides a foundation for cpuset management features where we need
to understand CPU core topology for proper isolation.
Assisted-by: Claude Sonnet 4.5 <noreply@anthropic.com>
Signed-off-by: John Kacur <jkacur@redhat.com>
---
rteval/sysinfo/coresiblings.py | 110 +++++++++++++++++++++++++++++++++
1 file changed, 110 insertions(+)
create mode 100644 rteval/sysinfo/coresiblings.py
diff --git a/rteval/sysinfo/coresiblings.py b/rteval/sysinfo/coresiblings.py
new file mode 100644
index 000000000000..3c97439d0a39
--- /dev/null
+++ b/rteval/sysinfo/coresiblings.py
@@ -0,0 +1,110 @@
+# -*- coding: utf-8 -*-
+# SPDX-License-Identifier: GPL-2.0-or-later
+#
+# Copyright 2026 John Kacur <jkacur@redhat.com>
+#
+
+import os
+from rteval.cpulist_utils import expand_cpulist
+
+class CoreSiblings:
+ """Query CPU core topology to determine which CPUs share physical cores"""
+
+ def __init__(self, root="/"):
+ self.sysdir = os.path.join(root, 'sys', 'devices', 'system', 'cpu')
+ self.core_map = {} # Maps cpu -> set of sibling cpus
+ self._parse()
+
+ def _parse(self):
+ """Parse thread_siblings_list for each CPU"""
+ for dirname in os.listdir(self.sysdir):
+ # Only parse cpu<integer> directories
+ if dirname.startswith('cpu') and os.path.isdir(os.path.join(self.sysdir, dirname)):
+ try:
+ cpu_id = int(dirname[3:])
+ except ValueError:
+ continue
+
+ siblings_file = os.path.join(self.sysdir, dirname, 'topology', 'thread_siblings_list')
+ if os.path.exists(siblings_file):
+ with open(siblings_file, 'r') as f:
+ siblings_str = f.read().strip()
+ # expand_cpulist returns a list of cpu numbers
+ siblings = set(expand_cpulist(siblings_str))
+ self.core_map[cpu_id] = siblings
+
+ def share_core(self, cpu1, cpu2):
+ """
+ Check if two CPUs share the same physical core.
+
+ Args:
+ cpu1: First CPU ID (int)
+ cpu2: Second CPU ID (int)
+
+ Returns:
+ True if CPUs share a core, False otherwise
+ """
+ if cpu1 not in self.core_map:
+ return False
+ return cpu2 in self.core_map[cpu1]
+
+ def get_siblings(self, cpu):
+ """
+ Get all CPUs that share a core with the given CPU.
+
+ Args:
+ cpu: CPU ID (int)
+
+ Returns:
+ Set of CPU IDs that share a core with cpu (includes cpu itself)
+ """
+ return self.core_map.get(cpu, set())
+
+ def get_core_groups(self):
+ """
+ Get all unique core sibling groups.
+
+ Returns:
+ List of sets, where each set contains CPUs that share a core
+ """
+ seen = set()
+ groups = []
+
+ for cpu, siblings in self.core_map.items():
+ # Use frozenset as a hashable representation
+ group_key = frozenset(siblings)
+ if group_key not in seen:
+ seen.add(group_key)
+ groups.append(siblings)
+
+ return groups
+
+
+def unit_test():
+ """Simple unit test"""
+ try:
+ cs = CoreSiblings()
+
+ print("Core Sibling Groups:")
+ for i, group in enumerate(cs.get_core_groups()):
+ print(f" Core {i}: {sorted(group)}")
+
+ # Test share_core with first two CPUs if they exist
+ if len(cs.core_map) >= 2:
+ cpus = sorted(cs.core_map.keys())
+ cpu1, cpu2 = cpus[0], cpus[1]
+ print(f"\nDo CPU {cpu1} and CPU {cpu2} share a core? {cs.share_core(cpu1, cpu2)}")
+ print(f"CPU {cpu1} siblings: {sorted(cs.get_siblings(cpu1))}")
+ print(f"CPU {cpu2} siblings: {sorted(cs.get_siblings(cpu2))}")
+
+ return 0
+ except Exception as e:
+ print(f"** EXCEPTION: {e}")
+ import traceback
+ traceback.print_exc()
+ return 1
+
+
+if __name__ == '__main__':
+ import sys
+ sys.exit(unit_test())
--
2.53.0
next prev parent reply other threads:[~2026-04-21 18:26 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-04-21 18:26 [PATCH 0/7] Core sharing validation for CPU isolation John Kacur
2026-04-21 18:26 ` John Kacur [this message]
2026-04-21 18:26 ` [PATCH 2/7] rteval: Add core " John Kacur
2026-04-21 18:26 ` [PATCH 3/7] rteval: Include core sharing warnings in XML report John Kacur
2026-04-21 18:26 ` [PATCH 4/7] rteval: Add temporary test for core sharing validation with mocked isolated CPUs John Kacur
2026-04-21 18:26 ` [PATCH 5/7] rteval: Display core sharing warnings in text report John Kacur
2026-04-21 18:26 ` [PATCH 6/7] rteval: Add --warn-non-isolated-core-sharing option for measurement vs load warnings John Kacur
2026-04-21 18:26 ` [PATCH 7/7] rteval: Include --warn-non-isolated-core-sharing warnings in XML report John Kacur
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260421182618.261347-2-jkacur@redhat.com \
--to=jkacur@redhat.com \
--cc=linux-rt-users@vger.kernel.org \
--cc=tglozar@redhat.com \
--cc=williams@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox