Skip to contentSkip to conversation

high_freq_hwp_cap_cppc.c: new test

PATCH Superseded Piotr Kubaj 2 participants 4 messages Mar 23 – Apr 29

v6 of 14 Superseded Latest is v14 →

lore
Share & tools
lore
#0 lore raw

PATCH v6 1/1 high_freq_hwp_cap_cppc.c: new test

Verify for all online logical CPUs that their highest performance value are the same for HWP Capability MSR 0x771 and CPPC sysfs file.

---
Online check needs to be true by default for cpu0 to be checked.

 runtest/power_management_tests                |  1 +
 testcases/kernel/power_management/.gitignore  |  1 +
 .../power_management/high_freq_hwp_cap_cppc.c | 84 +++++++++++++++++++
 3 files changed, 86 insertions(+)
 create mode 100644 testcases/kernel/power_management/high_freq_hwp_cap_cppc.c
runtest/power_management_tests +1 -0
index 884e615cd..6d87dfb7f 100644
@@ -1,4 +1,5 @@
 #POWER_MANAGEMENT
+high_freq_hwp_cap_cppc high_freq_hwp_cap_cppc
 runpwtests01 runpwtests01.sh
 runpwtests02 runpwtests02.sh
 runpwtests03 runpwtests03.sh
testcases/kernel/power_management/.gitignore +1 -0
index 0c2a3ed4b..c13bca1c4 100644
@@ -1 +1,2 @@
+high_freq_hwp_cap_cppc
 pm_get_sched_values
testcases/kernel/power_management/high_freq_hwp_cap_cppc.c new file +84 -0

84 lines not on this page · raw message

Signature · 13 lines
2.47.3

---------------------------------------------------------------------
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.


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp
#1 lore raw

Hi Piotr,

Thanks for the new test. A few issues below.

Quoted 3 lines

+/*
+ * Copyright (C) 2026 Intel - http://www.intel.com/
+ */

The copyright line should include an author name and email, e.g.:

Copyright (c) 2026 Piotr Kubaj <[email protected]>

testcases/kernel/power_management/high_freq_hwp_cap_cppc.c line 32

Quoted 2 lines
testcases/kernel/power_management/high_freq_hwp_cap_cppc.c new file +2 -0
+		if (i)
+			SAFE_FILE_SCANF(path, "%d", &online);

SAFE_FILE_SCANF with "%d" expects a pointer to int, but online is bool (typically 1 byte). This is undefined behavior and will likely corrupt the stack on little-endian. Please use int online = 1; instead.

This should be spotted by any static analyzer or clangd..and even gcc shows this error.

testcases/kernel/power_management/high_freq_hwp_cap_cppc.c line 51

Quoted 4 lines
testcases/kernel/power_management/high_freq_hwp_cap_cppc.c new file +4 -0
+		snprintf(path, sizeof(path), "/dev/cpu/%d/msr", i);
+		int fd = SAFE_OPEN(path, O_RDONLY);
+
+		SAFE_PREAD(1, fd, &msr_highest_perf, sizeof(msr_highest_perf), 0x771);

SAFE_CLOSE(fd) right after it has been used. There's no need to keep it open after pread().

testcases/kernel/power_management/high_freq_hwp_cap_cppc.c line 52

Quoted 1 line
testcases/kernel/power_management/high_freq_hwp_cap_cppc.c new file +1 -0
+		msr_highest_perf &= (1ULL << 8) - 1;

0x771 and the 8-bit mask are magic numbers. Named defines would help readability, e.g.:

  #define MSR_HWP_CAPABILITIES  0x771
  #define HIGHEST_PERF_MASK     0xFF

The test fails because msr device is not available, so you need:

  .needs_drivers = (const char *const []) {                                                                                                        
      "msr",                                                                                                                                       
      NULL                                                                                                                                         
  },

And probably the CONFIG_X86_MSR=y check as well.

Regards,
--
Andrea Cervesato
SUSE QE Automation Engineer Linux
[email protected]

Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#2 lore raw

2026-03-25 (水) の 12:24 +0000 に Andrea Cervesato さんは書きました:

Quoted 20 lines

Hi Piotr,

Thanks for the new test. A few issues below.

Quoted 3 lines

+/*
+ * Copyright (C) 2026 Intel - http://www.intel.com/
+ */

The copyright line should include an author name and email, e.g.:

  Copyright (c) 2026 Piotr Kubaj <[email protected]>

testcases/kernel/power_management/high_freq_hwp_cap_cppc.c line 32

Quoted 2 lines
testcases/kernel/power_management/high_freq_hwp_cap_cppc.c new file +2 -0
+		if (i)
+			SAFE_FILE_SCANF(path, "%d", &online);

SAFE_FILE_SCANF with "%d" expects a pointer to int, but online is bool (typically 1 byte). This is undefined behavior and will likely corrupt the stack on little-endian. Please use int online = 1; instead.

Quoted 2 lines

This should be spotted by any static analyzer or clangd..and even gcc shows this error.

For the record:
GCC does NOT print a warning while compiling that code. At least with the default set of options when building on Debian 13 with stock GCC 14.2.0.

Quoted 35 lines
Quoted 7 lines
(quoted diff) +1 -0
+		snprintf(path, sizeof(path), "/dev/cpu/%d/msr",

i);

testcases/kernel/power_management/high_freq_hwp_cap_cppc.c new file +3 -0
+		int fd = SAFE_OPEN(path, O_RDONLY);
+
+		SAFE_PREAD(1, fd, &msr_highest_perf,

sizeof(msr_highest_perf), 0x771);

SAFE_CLOSE(fd) right after it has been used. There's no need to keep it
open after pread().

testcases/kernel/power_management/high_freq_hwp_cap_cppc.c line 52

Quoted 1 line
testcases/kernel/power_management/high_freq_hwp_cap_cppc.c new file +1 -0
+		msr_highest_perf &= (1ULL << 8) - 1;

0x771 and the 8-bit mask are magic numbers. Named defines would help readability, e.g.:

  #define MSR_HWP_CAPABILITIES  0x771
  #define HIGHEST_PERF_MASK     0xFF

The test fails because msr device is not available, so you need:

  .needs_drivers = (const char *const [])
{


"msr",



NULL


  },

Quoted 7 lines

And probably the CONFIG_X86_MSR=y check as well.

Regards,
--
Andrea Cervesato
SUSE QE Automation Engineer Linux
[email protected]

---------------------------------------------------------------------
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.

Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp
#3 lore raw
In reply to Piotr Kubaj

Hi Piotr,

this patch looks much better.

testcases/kernel/power_management/high_freq_hwp_cap_cppc.c line 27

Quoted 1 line
testcases/kernel/power_management/high_freq_hwp_cap_cppc.c new file +1 -0
+		char path[PATH_MAX];

Please move this out of the loop, snprintf() already terminate the string with \0 so we don't need to reset the buffer before using it.

Regards,
--
Andrea Cervesato
SUSE QE Automation Engineer Linux
[email protected]

Signature · 1 line
Mailing list info: https://lists.linux.it/listinfo/ltp