[LTP] [PATCH] high_freq_hwp_cap_cppc.c: new test

[Kubaj, Piotr]

2026-03-16 (月) の 13:33 +0100 に Petr Vorel さんは書きました:
> Hi Piotr,
> 
> > Verify for all online logical CPUs that their highest performance
> > value are
> > the same for HWP Capability MSR 0x771 and CPPC sysfs file.
> 
> You were supposed to add your RBT:
> Signed-off-by: Piotr Kubaj <piotr.kubaj@intel.com>
> 
> > ---
> >  testcases/kernel/power_management/.gitignore  |  1 +
> >  .../power_management/high_freq_hwp_cap_cppc.c | 57
> > +++++++++++++++++++
> 
> You did not add test into runtest/thermal. OTOH test will apply also
> after we
> finally merge thermal_interrupt_events.c.

It's actually not thermal-related, but there are going to be more
thermal tests anyway.

> 
> >  2 files changed, 58 insertions(+)
> >  create mode 100644
> > testcases/kernel/power_management/high_freq_hwp_cap_cppc.c
> 
> > diff --git a/testcases/kernel/power_management/.gitignore
> > b/testcases/kernel/power_management/.gitignore
> > index 0c2a3ed4b..c13bca1c4 100644
> > --- a/testcases/kernel/power_management/.gitignore
> > +++ b/testcases/kernel/power_management/.gitignore
> > @@ -1 +1,2 @@
> > +high_freq_hwp_cap_cppc
> >  pm_get_sched_values
> > diff --git
> > a/testcases/kernel/power_management/high_freq_hwp_cap_cppc.c
> > b/testcases/kernel/power_management/high_freq_hwp_cap_cppc.c
> > new file mode 100644
> > index 000000000..4cbb81f0b
> > --- /dev/null
> > +++ b/testcases/kernel/power_management/high_freq_hwp_cap_cppc.c
> > @@ -0,0 +1,57 @@
> > +// SPDX-License-Identifier: GPL-2.0-or-later
> > +
> > +/*
> > + * Copyright (C) 2025-2026 Intel - http://www.intel.com/
> > + */
> > +
> > +/*\
> > + * Verify for all online logical CPUs  that their highest
> > performance value are
> > + * the same for HWP Capability MSR 0x771 and CPPC sysfs file.
> > + */
> > +
> > +#include "tst_test.h"
> > +
> > +static int nproc;
> > +
> > +static void setup(void)
> > +{
> > +	nproc = tst_ncpus();
> > +}
> > +
> > +static void run(void)
> > +{
> > +	for (int i = 0; i < nproc; i++) {
> > +		char path[PATH_MAX];
> > +		unsigned long long msr_highest_perf = 0,
> > sysfs_highest_perf = 0;
> > +
> > +		snprintf(path, PATH_MAX,
> > "/sys/devices/system/cpu/cpu%d/acpi_cppc/highest_perf", i);
> > +		SAFE_FILE_SCANF(path, "%llu",
> > &sysfs_highest_perf);
> 
> I tried to run the test, but it fails due missing sysfs file:
> 
> $ grep CONFIG_X86_MSR /boot/config-6.19.0-rc1-1.g274aff5-default 
> CONFIG_X86_MSR=m
> 
> $ lsmod |grep msr
> intel_rapl_msr         20480  0
> intel_rapl_common      53248  1 intel_rapl_msr
> 
> # ./high_freq_hwp_cap_cppc 
> tst_test.c:1887: TINFO: Overall timeout per run is 0h 02m 00s
> ...
> high_freq_hwp_cap_cppc.c:28: TBROK: Failed to open FILE
> '/sys/devices/system/cpu/cpu0/acpi_cppc/highest_perf' for reading:
> ENOENT (2)
> 
> # modprobe msr
> 
> $ lsmod |grep msr
> msr                    12288  0
> intel_rapl_msr         20480  0
> intel_rapl_common      53248  1 intel_rapl_msr
> 
> # ./high_freq_hwp_cap_cppc
> ...
> high_freq_hwp_cap_cppc.c:28: TBROK: Failed to open FILE
> '/sys/devices/system/cpu/cpu0/acpi_cppc/highest_perf' for reading:
> ENOENT (2)
> 
> => obviously checking for CONFIG_X86_MSR is not enough.
Right, CONFIG_ACPI_CPPC_LIB is also necessary.

> 
> OT: it'd be nice to extend 'save_restore' library functionality to
> 1) support glob (/sys/devices/system/cpu/cpu*/acpi_cppc/highest_perf)
> 2) read value into the variable (in case of glob into array).
> 
> > +		tst_res(TDEBUG, "%s: %llu", path,
> > sysfs_highest_perf);
> > +
> > +		snprintf(path, PATH_MAX, "/dev/cpu/%d/msr", i);
> > +		int fd = SAFE_OPEN(path, O_RDONLY);
> > +
> > +		if (pread(fd, &msr_highest_perf,
> > sizeof(msr_highest_perf), 0x771) < 0) {
> > +			SAFE_CLOSE(fd);
> > +			tst_brk(TBROK | TERRNO, "MSR read error");
> > +		}
> > +		SAFE_CLOSE(fd);
> > +		msr_highest_perf &= (1ULL << 8) - 1;
> > +		tst_res(TDEBUG, "%s: %llu", path,
> > msr_highest_perf);
> > +
> > +		if (msr_highest_perf != sysfs_highest_perf)
> > +			tst_brk(TFAIL, "CPU %d: highest
> > performance values differ between sysfs and MSR", i);
> > +	}
> > +
> > +	tst_res(TPASS, "Test pass");
> 
> "Test pass" has zero information value.
> Maybe something like:
> "sysfs and MSR values are equal"
Done.

> 
> > +}
> > +
> > +static struct tst_test test = {
> > +	.needs_kconfigs = (const char *const []) {
> > +		"CONFIG_X86_MSR",
> > +		NULL
> > +	},
> For thermal_interrupt_events.c we also have use .supported_archs (for
> the docs
> purposes), why not to add it to this test as well?
I removed it for the same reason that was mentioned in the review for
thermal_interrupt_events.c. MSR is x86- and x86_64-only. Added in the
newer version.
> 
> +	.supported_archs = (const char *const []) {
> +		"x86",
> +		"x86_64",
> +		NULL
> +	},
> 
> Kind regards,
> Petr
> 
> > +	.needs_root = 1,
> > +	.setup = setup,
> > +	.test_all = run
> > +};
---------------------------------------------------------------------
Intel Technology Poland sp. z o.o.
ul. Slowackiego 173 | 80-298 Gdansk | Sad Rejonowy Gdansk Polnoc | VII Wydzial Gospodarczy Krajowego Rejestru Sadowego - KRS 101882 | NIP 957-07-52-316 | Kapital zakladowy 200.000 PLN.
Spolka oswiadcza, ze posiada status duzego przedsiebiorcy w rozumieniu ustawy z dnia 8 marca 2013 r. o przeciwdzialaniu nadmiernym opoznieniom w transakcjach handlowych.

Ta wiadomosc wraz z zalacznikami jest przeznaczona dla okreslonego adresata i moze zawierac informacje poufne. W razie przypadkowego otrzymania tej wiadomosci, prosimy o powiadomienie nadawcy oraz trwale jej usuniecie; jakiekolwiek przegladanie lub rozpowszechnianie jest zabronione.
This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). If you are not the intended recipient, please contact the sender and delete all copies; any review or distribution by others is strictly prohibited.