Hi!
I test with below two cpu to test: - Intel(R) Xeon(R) CPU E7-8890 v4 @ 2.20GHz - Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz
With below action: make -C cpuidle cpuidle_killer make -C cpufreq run_tests
And only cpufreq_02 and cpufreq_03 can pass. Log as below(ignore pass part):
make: Entering directory '/lkp/benchmarks/pm-qa/cpufreq' ### ### cpufreq_01: ### test the cpufreq framework is available for frequency ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_01.0/cpu0: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu0/cpufreq'... Err cpufreq_01.0/cpu1: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu1/cpufreq'... Err cpufreq_01.0/cpu10: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu10/cpufreq'... Err
### ### cpufreq_04: ### test the change of the frequency is effective in 'userspace' mode ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_04.0: userspace not supported... skip
cpufreq_04: skip
### ### cpufreq_05: ### test 'ondemand' and 'conservative' trigger correctly the configuration directory ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_05.0: ondemand not supported... skip cpufreq_05.1: conservative not supported... skip cpufreq_05.2: userspace not supported... skip cpufreq_05.3: checking 'ondemand' directory exists... Err cpufreq_05.4: checking 'conservative' directory exists... Err
cpufreq_05: fail
### ### cpufreq_06: ### test the change of the frequencies affect the performances of a test program ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_06.0: userspace not supported... skip ### ### cpufreq_07: ### test the load of the cpu affects the frequency with 'ondemand' ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_07.0: ondemand not supported... skip ### ### cpufreq_08: ### test the load of the cpu does not affect the frequency with 'userspace' ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_08.0: userspace not supported... skip ### ### cpufreq_09: ### test the load of the cpu does not affect the frequency with 'powersave' ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_09.0/cpu0: checking 'powersave' sets frequency to 1200000 KHz... Err cpufreq_09.1/cpu0: checking 'powersave' frequency 1200000 KHz is fixed... Err cpufreq_09.0/cpu1: checking 'powersave' sets frequency to 1200000 KHz... Err cpufreq_09.1/cpu1: checking 'powersave' frequency 1200000 KHz is fixed... Err cpufreq_09.0/cpu10: checking 'powersave' sets frequency to 1200000 KHz... Err
I check cpus and make sure two cpus can support cpufreq refer to https://www.kernel.org/doc/Documentation/cpu-freq/user-guide.txt.
Do i miss something important or this subtest can not all pass with Intel cpu?
Thanks Xiang
On 18 August 2017 at 08:33, Dai Xiang xiangx.dai@intel.com wrote:
Hi!
I test with below two cpu to test:
- Intel(R) Xeon(R) CPU E7-8890 v4 @ 2.20GHz
- Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz
With below action: make -C cpuidle cpuidle_killer make -C cpufreq run_tests
And only cpufreq_02 and cpufreq_03 can pass. Log as below(ignore pass part):
make: Entering directory '/lkp/benchmarks/pm-qa/cpufreq' ### ### cpufreq_01: ### test the cpufreq framework is available for frequency ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_01.0/cpu0: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu0/cpufreq'... Err cpufreq_01.0/cpu1: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu1/cpufreq'... Err cpufreq_01.0/cpu10: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu10/cpufreq'... Err
### ### cpufreq_04: ### test the change of the frequency is effective in 'userspace' mode ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_04.0: userspace not supported... skip
cpufreq_04: skip
### ### cpufreq_05: ### test 'ondemand' and 'conservative' trigger correctly the configuration directory ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_05.0: ondemand not supported... skip cpufreq_05.1: conservative not supported... skip cpufreq_05.2: userspace not supported... skip cpufreq_05.3: checking 'ondemand' directory exists... Err cpufreq_05.4: checking 'conservative' directory exists... Err
Your platform must be using acpi-cpufreq driver, check with this:
$ cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver
And that driver doesn't allow/support other cpufreq governors and so these tests aren't supposed to be run on Intel platform.
See if you can use intel_pstate driver instead in "passive" mode.
-- viresh
On Fri, Aug 18, 2017 at 09:42:19AM +0530, Viresh Kumar wrote:
On 18 August 2017 at 08:33, Dai Xiang xiangx.dai@intel.com wrote:
Hi!
I test with below two cpu to test:
- Intel(R) Xeon(R) CPU E7-8890 v4 @ 2.20GHz
- Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz
With below action: make -C cpuidle cpuidle_killer make -C cpufreq run_tests
And only cpufreq_02 and cpufreq_03 can pass. Log as below(ignore pass part):
make: Entering directory '/lkp/benchmarks/pm-qa/cpufreq' ### ### cpufreq_01: ### test the cpufreq framework is available for frequency ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_01.0/cpu0: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu0/cpufreq'... Err cpufreq_01.0/cpu1: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu1/cpufreq'... Err cpufreq_01.0/cpu10: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu10/cpufreq'... Err
### ### cpufreq_04: ### test the change of the frequency is effective in 'userspace' mode ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_04.0: userspace not supported... skip
cpufreq_04: skip
### ### cpufreq_05: ### test 'ondemand' and 'conservative' trigger correctly the configuration directory ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_05.0: ondemand not supported... skip cpufreq_05.1: conservative not supported... skip cpufreq_05.2: userspace not supported... skip cpufreq_05.3: checking 'ondemand' directory exists... Err cpufreq_05.4: checking 'conservative' directory exists... Err
Your platform must be using acpi-cpufreq driver, check with this:
$ cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver
And that driver doesn't allow/support other cpufreq governors and so these tests aren't supposed to be run on Intel platform.
See if you can use intel_pstate driver instead in "passive" mode.
In fact, i has used intel_pstate driver: $ cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver intel_pstate
And i try to echo into /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor like ondemand, but do not help.
Need i do some setting to enable those tests?
Thanks Xiang
-- viresh
On 18 August 2017 at 11:11, Dai Xiang xiangx.dai@intel.com wrote:
In fact, i has used intel_pstate driver: $ cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver intel_pstate
It may be running in *active* mode and you need to put it in *passive* mode.
Try passing following with cmdline:
intel_pstate=passive
On 18/08/2017 07:43, Viresh Kumar wrote:
On 18 August 2017 at 11:11, Dai Xiang xiangx.dai@intel.com wrote:
In fact, i has used intel_pstate driver: $ cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver intel_pstate
It may be running in *active* mode and you need to put it in *passive* mode.
Try passing following with cmdline:
intel_pstate=passive
I don't know the passive / active mode but the path changed.
/sys/devices/system/cpu/cpufreq/policy[0-9]/scaling_driver
So there is probably something to do in PM-QA.
On 18 August 2017 at 11:48, Daniel Lezcano daniel.lezcano@linaro.org wrote:
I don't know the passive / active mode but the path changed.
/sys/devices/system/cpu/cpufreq/policy[0-9]/scaling_driver
Naah, we wouldn't have broken userspace like that :)
$ ls -l /sys/devices/system/cpu/cpu0/cpufreq lrwxrwxrwx 1 root root 0 Aug 16 08:41 /sys/devices/system/cpu/cpu0/cpufreq -> ../cpufreq/policy0/
On Fri, Aug 18, 2017 at 11:13:35AM +0530, Viresh Kumar wrote:
On 18 August 2017 at 11:11, Dai Xiang xiangx.dai@intel.com wrote:
In fact, i has used intel_pstate driver: $ cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_driver intel_pstate
It may be running in *active* mode and you need to put it in *passive* mode.
Try passing following with cmdline:
intel_pstate=passive
Cool, after use this kernel_cmdline, 4\5\6\8 can pass now, but 1\7\9 still can not: cpufreq_01.0/cpu0: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu0/cpufreq'... Err cpufreq_01.0/cpu1: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu1/cpufreq'... Err cpufreq_01.0/cpu2: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu2/cpufreq'... Err cpufreq_01.0/cpu3: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu3/cpufreq'... Err cpufreq_01.0/cpu4: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu4/cpufreq'... Err cpufreq_01.0/cpu5: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu5/cpufreq'... Err cpufreq_01.0/cpu6: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu6/cpufreq'... Err cpufreq_01.0/cpu7: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu7/cpufreq'... Err
cpufreq_01: fail
cpufreq_07.0/cpu0: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu1: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu2: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu3: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu4: can not reach a quescient point for 'ondemand'... skip IPMI BMC is not supported on this machine, skip bmc-watchdog setup! cpufreq_07.0/cpu5: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu6: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu7: can not reach a quescient point for 'ondemand'... skip
cpufreq_07: skip
cpufreq_09.0/cpu0: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu0: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu1: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu1: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu2: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu2: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu3: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu3: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu4: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu4: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu5: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu5: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu6: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu6: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu7: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu7: checking 'powersave' frequency 1600000 KHz is fixed... Err
On 18/08/2017 08:20, Viresh Kumar wrote:
On 18 August 2017 at 11:48, Daniel Lezcano daniel.lezcano@linaro.org wrote:
I don't know the passive / active mode but the path changed.
/sys/devices/system/cpu/cpufreq/policy[0-9]/scaling_driver
Naah, we wouldn't have broken userspace like that :)
$ ls -l /sys/devices/system/cpu/cpu0/cpufreq lrwxrwxrwx 1 root root 0 Aug 16 08:41 /sys/devices/system/cpu/cpu0/cpufreq -> ../cpufreq/policy0/
Ah, yes. I'm not fully awake this morning :)
On 18 August 2017 at 11:52, Dai Xiang xiangx.dai@intel.com wrote:
Cool, after use this kernel_cmdline, 4\5\6\8 can pass now, but 1\7\9 still can not:
Why do you say that 4/5/6/8 have passed ?
cpufreq_01.0/cpu0: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu0/cpufreq'... Err cpufreq_01.0/cpu1: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu1/cpufreq'... Err cpufreq_01.0/cpu2: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu2/cpufreq'... Err cpufreq_01.0/cpu3: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu3/cpufreq'... Err cpufreq_01.0/cpu4: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu4/cpufreq'... Err cpufreq_01.0/cpu5: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu5/cpufreq'... Err cpufreq_01.0/cpu6: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu6/cpufreq'... Err cpufreq_01.0/cpu7: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu7/cpufreq'... Err
cpufreq_01: fail
cpufreq_07.0/cpu0: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu1: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu2: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu3: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu4: can not reach a quescient point for 'ondemand'... skip IPMI BMC is not supported on this machine, skip bmc-watchdog setup! cpufreq_07.0/cpu5: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu6: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu7: can not reach a quescient point for 'ondemand'... skip
cpufreq_07: skip
cpufreq_09.0/cpu0: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu0: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu1: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu1: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu2: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu2: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu3: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu3: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu4: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu4: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu5: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu5: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu6: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu6: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu7: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu7: checking 'powersave' frequency 1600000 KHz is fixed... Err
I can see all of them fail and the reason is quite simple to see. Even intel-pstate doesn't provide a frequency table and so whatsoever depends on a frequency table will fail.
These tests aren't really meant for x86 stuff, sorry about that :(
-- viresh
On Fri, Aug 18, 2017 at 11:59:47AM +0530, Viresh Kumar wrote:
On 18 August 2017 at 11:52, Dai Xiang xiangx.dai@intel.com wrote:
Cool, after use this kernel_cmdline, 4\5\6\8 can pass now, but 1\7\9 still can not:
Why do you say that 4/5/6/8 have passed ?
### ### cpufreq_04: ### test the change of the frequency is effective in 'userspace' mode ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ###
cpufreq_04: pass
### ### cpufreq_05: ### test 'ondemand' and 'conservative' trigger correctly the configuration directory ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ### cpufreq_05.0: checking 'ondemand' directory exists... Ok cpufreq_05.1: checking 'conservative' directory exists... Ok cpufreq_05.2: checking 'ondemand' directory is not there... Ok cpufreq_05.3: checking 'conservative' directory is not there... Ok cpufreq_05.4: checking 'ondemand' directory exists... Ok cpufreq_05.5: checking 'conservative' directory exists... Ok
cpufreq_05: pass
### ### cpufreq_06: ### test the change of the frequencies affect the performances of a test program ### https://wiki.linaro.org/WorkingGroups/PowerManagement/Resources/TestSuite/Pm... ###
cpufreq_06: pass
cpufreq_08: pass
cpufreq_01.0/cpu0: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu0/cpufreq'... Err cpufreq_01.0/cpu1: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu1/cpufreq'... Err cpufreq_01.0/cpu2: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu2/cpufreq'... Err cpufreq_01.0/cpu3: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu3/cpufreq'... Err cpufreq_01.0/cpu4: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu4/cpufreq'... Err cpufreq_01.0/cpu5: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu5/cpufreq'... Err cpufreq_01.0/cpu6: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu6/cpufreq'... Err cpufreq_01.0/cpu7: checking 'scaling_available_frequencies' exists in '/sys/devices/system/cpu/cpu7/cpufreq'... Err
cpufreq_01: fail
cpufreq_07.0/cpu0: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu1: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu2: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu3: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu4: can not reach a quescient point for 'ondemand'... skip IPMI BMC is not supported on this machine, skip bmc-watchdog setup! cpufreq_07.0/cpu5: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu6: can not reach a quescient point for 'ondemand'... skip cpufreq_07.0/cpu7: can not reach a quescient point for 'ondemand'... skip
cpufreq_07: skip
cpufreq_09.0/cpu0: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu0: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu1: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu1: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu2: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu2: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu3: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu3: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu4: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu4: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu5: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu5: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu6: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu6: checking 'powersave' frequency 1600000 KHz is fixed... Err cpufreq_09.0/cpu7: checking 'powersave' sets frequency to 1600000 KHz... Err cpufreq_09.1/cpu7: checking 'powersave' frequency 1600000 KHz is fixed... Err
I can see all of them fail and the reason is quite simple to see. Even intel-pstate doesn't provide a frequency table and so whatsoever depends on a frequency table will fail.
These tests aren't really meant for x86 stuff, sorry about that :(
So with intel cpu, the 1/7/9 subtests always fail, right? They are not for x86, so which type of cpu or arch can enable them?
-- viresh
On 18 August 2017 at 12:03, Dai Xiang xiangx.dai@intel.com wrote:
On Fri, Aug 18, 2017 at 11:59:47AM +0530, Viresh Kumar wrote:
On 18 August 2017 at 11:52, Dai Xiang xiangx.dai@intel.com wrote:
Cool, after use this kernel_cmdline, 4\5\6\8 can pass now, but 1\7\9 still can not:
Why do you say that 4/5/6/8 have passed ?
Ahh you haven't pasted the "pass" results earlier and so I got confused.
So with intel cpu, the 1/7/9 subtests always fail, right?
Yeah, as a frequency table is required for them.
They are not for x86, so which type of cpu or arch can enable them?
Any driver that provides a target_index() will pass.
Actually I was wrong about acpi-cpufreq driver earlier, and it does provide a target_index() callback. Try using that driver and all your tests may pass.
-- viresh
On 18/08/2017 08:29, Viresh Kumar wrote:
On 18 August 2017 at 11:52, Dai Xiang xiangx.dai@intel.com wrote:
[ ... ]
I can see all of them fail and the reason is quite simple to see. Even intel-pstate doesn't provide a frequency table and so whatsoever depends on a frequency table will fail.
These tests aren't really meant for x86 stuff, sorry about that :(
Actually they were meant for x86 but they were implemented before the intel_pstate driver and did not evolve to take it into account.
-
Dai Xiang -
Daniel Lezcano -
Viresh Kumar