linux-kernel - [RFC PATCH v1 0/2] kselftest/resctrl: CAT functional tests

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>] [thread-next>] [day] [month] [year] [list]

Message-Id: <20250616082453.3725-1-ilpo.jarvinen@linux.intel.com>
Date: Mon, 16 Jun 2025 11:24:51 +0300
From: Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>
To: linux-kselftest@...r.kernel.org,
	Reinette Chatre <reinette.chatre@...el.com>,
	Shuah Khan <shuah@...nel.org>,
	Tony Luck <tony.luck@...el.com>,
	Dave Martin <Dave.Martin@....com>,
	James Morse <james.morse@....com>,
	Shaopeng Tan <tan.shaopeng@...fujitsu.com>
Cc: Fenghua Yu <fenghua.yu@...el.com>,
	linux-kernel@...r.kernel.org,
	Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>
Subject: [RFC PATCH v1 0/2] kselftest/resctrl: CAT functional tests

Hi all,

In the last Fall Reinette mentioned functional tests of resctrl would
be preferred over selftests that are based on performance measurement.
This series tries to address that shortcoming by adding some functional
tests for resctrl FS interface and another that checks MSRs match to
what is written through resctrl FS. The MSR test is only available for
Intel CPUs at the moment.

Why RFC?

The new functional selftest itself works, AFAIK. However, calling
ksft_test_result_skip() in cat.c if MSR reading is found to be
unavailable is problematic because of how kselftest harness is
architected. The kselftest.h header itself defines some variables, so
including it into different .c files results in duplicating the test
framework related variables (duplication of ksft_count matters in this
case).

The duplication problem could be worked around by creating a resctrl
selftest specific wrapper for ksft_test_result_skip() into
resctrl_tests.c so the accounting would occur in the "correct" .c file,
but perhaps that is considered hacky and the selftest framework/build
systems should be reworked to avoid duplicating variables?

Ilpo Järvinen (2):
  kselftest/resctrl: CAT L3 resctrl FS function tests
  kselftest/resctrl: Add CAT L3 CBM functional test for Intel RDT

 tools/testing/selftests/resctrl/cat_test.c    | 210 ++++++++++++++++++
 tools/testing/selftests/resctrl/msr.c         |  55 +++++
 tools/testing/selftests/resctrl/resctrl.h     |   6 +
 .../testing/selftests/resctrl/resctrl_tests.c |   2 +
 tools/testing/selftests/resctrl/resctrlfs.c   |  48 ++++
 5 files changed, 321 insertions(+)
 create mode 100644 tools/testing/selftests/resctrl/msr.c

base-commit: c1d7e19c70cbb8a19f63c190cf53e71b5f970514
-- 
2.39.5