public inbox for ltp@lists.linux.it
 help / color / mirror / Atom feed
From: Andrea Cervesato via ltp <ltp@lists.linux.it>
To: "Piotr Kubaj" <piotr.kubaj@intel.com>
Cc: daniel.niestepski@intel.com, tomasz.ossowski@intel.com,
	helena.anna.dubel@intel.com, rafael.j.wysocki@intel.com,
	ltp@lists.linux.it
Subject: Re: [LTP] [PATCH v4] high_freq_hwp_cap_cppc.c: new test
Date: Mon, 23 Mar 2026 08:00:28 +0000	[thread overview]
Message-ID: <69c0f31f.050a0220.943db.4f2a@mx.google.com> (raw)
In-Reply-To: <20260321083813.138283-2-piotr.kubaj@intel.com>

Hi Piotr,

A few issues below.

> +// SPDX-License-Identifier: GPL-2.0-or-later
> +
> +/*
> + * Copyright (C) 2026 Intel - http://www.intel.com/

The blank line between the SPDX identifier and the copyright block is
non-standard. They should be adjacent (no blank line between them).

> +	nproc = tst_ncpus();

tst_ncpus() returns _SC_NPROCESSORS_ONLN — the *count* of online CPUs,
not the highest CPU number. On a system where CPUs 0, 1, 3 are online
and CPU 2 is offline, tst_ncpus() returns 3, and the loop below iterates
over CPUs 0, 1, 2, accessing the wrong set of sysfs paths.

Use tst_ncpus_conf() as the upper bound and skip offline CPUs
individually, or iterate over the actual online CPU numbers from
/sys/devices/system/cpu/online.

> +		snprintf(path, sizeof(path), "/sys/devices/system/cpu/cpu%d/acpi_cppc/highest_perf", i);
> +		SAFE_FILE_SCANF(path, "%llu", &sysfs_highest_perf);

CONFIG_ACPI_CPPC_LIB=y does not guarantee the acpi_cppc/ sysfs
directory exists at runtime — it only appears when the CPU firmware
exposes the CPPC _CPC ACPI method. If the path is absent, SAFE_FILE_SCANF
aborts with TBROK. Check for existence and return TCONF instead:

  if (access(path, F_OK)) {
      tst_res(TCONF, "CPPC sysfs not available, skipping");
      return;
  }

> +		int fd = SAFE_OPEN(path, O_RDONLY);
> +
> +		SAFE_PREAD(1, fd, &msr_highest_perf, sizeof(msr_highest_perf), 0x771);
> +		msr_highest_perf &= (1ULL << 8) - 1;
> +		tst_res(TDEBUG, "%s: %llu", path, msr_highest_perf);
> +
> +		if (msr_highest_perf != sysfs_highest_perf)
> +			status = false;

Two issues:

1. fd is never closed. Opening it once per CPU iteration leaks one
   file descriptor per CPU. Add SAFE_CLOSE(fd) at the end of the loop
   body.

2. When the values differ, status is set to false but there is no
   per-CPU message. The single TFAIL at the end gives no indication of
   which CPU triggered the failure. Add a TINFO or TFAIL per CPU:

     tst_res(TFAIL, "cpu%d: sysfs=%llu MSR=%llu",
             i, sysfs_highest_perf, msr_highest_perf);


Regards,
--
Andrea Cervesato
SUSE QE Automation Engineer Linux
andrea.cervesato@suse.com

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

      reply	other threads:[~2026-03-23  8:01 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-03-21  8:38 [LTP] [PATCH v4] high_freq_hwp_cap_cppc.c: new test Piotr Kubaj
2026-03-23  8:00 ` Andrea Cervesato via ltp [this message]

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=69c0f31f.050a0220.943db.4f2a@mx.google.com \
    --to=ltp@lists.linux.it \
    --cc=andrea.cervesato@suse.com \
    --cc=daniel.niestepski@intel.com \
    --cc=helena.anna.dubel@intel.com \
    --cc=piotr.kubaj@intel.com \
    --cc=rafael.j.wysocki@intel.com \
    --cc=tomasz.ossowski@intel.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