DPDK patches and discussions
 help / color / mirror / Atom feed
From: Wenzhuo Lu <wenzhuo.lu@intel.com>
To: dev@dpdk.org
Cc: Wenzhuo Lu <wenzhuo.lu@intel.com>
Subject: [RFC] usertools: enhance CPU layout
Date: Fri, 10 Feb 2023 15:41:34 +0800	[thread overview]
Message-ID: <1676014894-83654-1-git-send-email-wenzhuo.lu@intel.com> (raw)

CPU is becoming more and more complex.
Some CPUs are made up of several dies.
The cores in different dies may be different.
The user tool can be updated to show more
about CPU components.

This patch addes below informantion,
1, Group the cores based on the die.
2, A core is either a performance core or an
   efficency core.
   A performance core is shown as 'Core-P'.
   An efficency core is shown as 'Core-E'.

The known limitation/issue.
1, To tell a core is P-core or E-core is based on if
   this core shares L2 cache with others.
   Not sure if there's any better criteria.
2, OS shows there's only 1 die in a CPU although there're
   actually several.
   Hope the accurate information can be proved by the coming
   OSs.

There's also other information, like accelerators and even
memory are embedded into CPU. In the future the user tool
may be updated to show more.

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
---
 usertools/cpu_layout.py | 57 +++++++++++++++++++++++++++++++----------
 1 file changed, 43 insertions(+), 14 deletions(-)

diff --git a/usertools/cpu_layout.py b/usertools/cpu_layout.py
index 891b9238fa..9812df0503 100755
--- a/usertools/cpu_layout.py
+++ b/usertools/cpu_layout.py
@@ -1,11 +1,15 @@
 #!/usr/bin/env python3
 # SPDX-License-Identifier: BSD-3-Clause
-# Copyright(c) 2010-2014 Intel Corporation
+# Copyright(c) 2010-2023 Intel Corporation
 # Copyright(c) 2017 Cavium, Inc. All rights reserved.
 
 sockets = []
+dies = []
 cores = []
 core_map = {}
+core_p_e = {}
+title_len = 47
+die_len = 6
 base_path = "/sys/devices/system/cpu"
 fd = open("{}/kernel_max".format(base_path))
 max_cpus = int(fd.read())
@@ -20,19 +24,38 @@
     fd = open("{}/cpu{}/topology/physical_package_id".format(base_path, cpu))
     socket = int(fd.read())
     fd.close()
+    fd = open("{}/cpu{}/topology/die_id".format(base_path, cpu))
+    die = int(fd.read())
+    fd.close()
+    fd = open("{}/cpu{}/topology/thread_siblings_list".format(base_path, cpu))
+    threads_share = str(fd.read())
+    fd.close()
+    fd = open("{}/cpu{}/cache/index2/shared_cpu_list".format(base_path, cpu))
+    l2_cache_share = str(fd.read())
+    fd.close()
+    if (threads_share == l2_cache_share):
+        p_e = '-P'
+    else:
+        p_e = '-E'
     if core not in cores:
         cores.append(core)
+    if die not in dies:
+        dies.append(die)
     if socket not in sockets:
         sockets.append(socket)
-    key = (socket, core)
+    key = (socket, die, core)
+    key_p_e = (die, core)
     if key not in core_map:
         core_map[key] = []
+    if key_p_e not in core_p_e:
+        core_p_e[key_p_e] = p_e
     core_map[key].append(cpu)
 
-print(format("=" * (47 + len(base_path))))
+print(format("=" * (title_len + len(base_path))))
 print("Core and Socket Information (as reported by '{}')".format(base_path))
-print("{}\n".format("=" * (47 + len(base_path))))
+print("{}\n".format("=" * (title_len + len(base_path))))
 print("cores = ", cores)
+print("dies = ", dies)
 print("sockets = ", sockets)
 print("")
 
@@ -43,22 +66,28 @@
                       + len('[]') + len('Socket ')
 max_core_id_len = len(str(max(cores)))
 
-output = " ".ljust(max_core_id_len + len('Core '))
+socket_space_len = max_core_id_len + len('Core ') + die_len + len('-P')
+output = " ".ljust(socket_space_len)
 for s in sockets:
     output += " Socket %s" % str(s).ljust(max_core_map_len - len('Socket '))
 print(output)
 
-output = " ".ljust(max_core_id_len + len('Core '))
+output = " ".ljust(socket_space_len)
 for s in sockets:
     output += " --------".ljust(max_core_map_len)
     output += " "
 print(output)
 
-for c in cores:
-    output = "Core %s" % str(c).ljust(max_core_id_len)
-    for s in sockets:
-        if (s, c) in core_map:
-            output += " " + str(core_map[(s, c)]).ljust(max_core_map_len)
-        else:
-            output += " " * (max_core_map_len + 1)
-    print(output)
+for d in dies:
+    print("Die", die)
+    for c in cores:
+        output = " ".ljust(die_len)
+        output += "Core"
+        output += core_p_e[(d, c)]
+        output += " %s" % str(c).ljust(max_core_id_len)
+        for s in sockets:
+            if (s, d, c) in core_map:
+                output += " " + str(core_map[(s, d, c)]).ljust(max_core_map_len)
+            else:
+                output += " " * (max_core_map_len + 1)
+        print(output)
-- 
2.25.1


                 reply	other threads:[~2023-02-10  8:17 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=1676014894-83654-1-git-send-email-wenzhuo.lu@intel.com \
    --to=wenzhuo.lu@intel.com \
    --cc=dev@dpdk.org \
    /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;
as well as URLs for NNTP newsgroup(s).