[PATCH 0/2] cpufreq: Specify the default governor on command line
2235d
Revision v1 of 2 in this series.
15 messages,
3 authors,
2020-06-23 · open the first message on its own page
This series enables users of prebuilt kernels (e.g. distro kernels) to
specify their CPUfreq governor of choice using the kernel command line,
instead of having to wait for the system to fully boot to userspace to
switch using the sysfs interface. This is helpful for 2 reasons:
1. users get to choose the governor that runs during the actual boot;
2. it simplifies the userspace boot procedure a bit (one less thing to
worry about).
To enable this, the first patch moves all governor init calls to
core_initcall, to make sure they are registered by the time the drivers
probe. This should be relatively low impact as registering a governor
is a simple procedure (it gets added to a llist), and all governors
already load at core_initcall anyway when they're set as the default
in Kconfig. This also allows to clean-up the governors' init/exit code,
and reduces boilerplate.
The second patch introduces the new command line parameter, inspired by
its cpuidle counterpart. More details can be found in the respective
patch headers.
Feedback is very much welcome :-)
Thanks,
Quentin
Quentin Perret (2):
cpufreq: Register governors at core_initcall
cpufreq: Specify default governor on command line
.../admin-guide/kernel-parameters.txt | 5 +++
Documentation/admin-guide/pm/cpufreq.rst | 6 ++--
.../platforms/cell/cpufreq_spudemand.c | 26 ++------------
drivers/cpufreq/cpufreq.c | 34 ++++++++++++++++---
drivers/cpufreq/cpufreq_conservative.c | 22 +++---------
drivers/cpufreq/cpufreq_ondemand.c | 24 ++++---------
drivers/cpufreq/cpufreq_performance.c | 14 ++------
drivers/cpufreq/cpufreq_powersave.c | 18 ++--------
drivers/cpufreq/cpufreq_userspace.c | 18 ++--------
include/linux/cpufreq.h | 14 ++++++++
kernel/sched/cpufreq_schedutil.c | 6 +---
11 files changed, 73 insertions(+), 114 deletions(-)
--
2.27.0.290.gba653c62da-goog
Currently, most CPUFreq governors are registered at core_initcall time
when used as default, and module_init otherwise. In preparation for
letting users specify the default governor on the kernel command line,
change all of them to use core_initcall unconditionally, as is already
the case for schedutil and performance. This will enable us to assume
builtin governors have been registered before the builtin CPUFreq
drivers probe.
And since all governors now have similar init/exit patterns, introduce
two new macros cpufreq_governor_{init,exit}() to factorize the code.
Signed-off-by: Quentin Perret <redacted>
---
Note: I couldn't boot-test the change to spudemand, by lack of hardware.
But I can confirm cell_defconfig compiles just fine.
---
.../platforms/cell/cpufreq_spudemand.c | 26 ++-----------------
drivers/cpufreq/cpufreq_conservative.c | 22 ++++------------
drivers/cpufreq/cpufreq_ondemand.c | 24 +++++------------
drivers/cpufreq/cpufreq_performance.c | 14 ++--------
drivers/cpufreq/cpufreq_powersave.c | 18 +++----------
drivers/cpufreq/cpufreq_userspace.c | 18 +++----------
include/linux/cpufreq.h | 14 ++++++++++
kernel/sched/cpufreq_schedutil.c | 6 +----
8 files changed, 36 insertions(+), 106 deletions(-)
diff --git a/arch/powerpc/platforms/cell/cpufreq_spudemand.c b/arch/powerpc/platforms/cell/cpufreq_spudemand.c
index 55b31eadb3c8..ca7849e113d7 100644
--- a/arch/powerpc/platforms/cell/cpufreq_spudemand.c
+++ b/arch/powerpc/platforms/cell/cpufreq_spudemand.c @@ -126,30 +126,8 @@ static struct cpufreq_governor spu_governor = {
. stop = spu_gov_stop ,
. owner = THIS_MODULE ,
};
- - /* - * module init and destoy - */ - - static int __init spu_gov_init ( void ) - { - int ret ; - - ret = cpufreq_register_governor ( & spu_governor ); - if ( ret ) - printk ( KERN_ERR "registration of governor failed \n " ); - return ret ; - } - - static void __exit spu_gov_exit ( void ) - { - cpufreq_unregister_governor ( & spu_governor ); - } - - - module_init ( spu_gov_init ); - module_exit ( spu_gov_exit ); + cpufreq_governor_init ( spu_governor ); + cpufreq_governor_exit ( spu_governor );
MODULE_LICENSE ( "GPL" );
MODULE_AUTHOR ( "Christian Krafft <krafft@de.ibm.com>" );
- diff --git a/drivers/cpufreq/cpufreq_conservative.c b/drivers/cpufreq/cpufreq_conservative.c
index 737ff3b9c2c0..aa39ff31ec9f 100644
--- a/drivers/cpufreq/cpufreq_conservative.c
+++ b/drivers/cpufreq/cpufreq_conservative.c @@ -322,17 +322,7 @@ static struct dbs_governor cs_governor = {
. start = cs_start ,
};
- #define CPU_FREQ_GOV_CONSERVATIVE (&cs_governor.gov) - - static int __init cpufreq_gov_dbs_init ( void ) - { - return cpufreq_register_governor ( CPU_FREQ_GOV_CONSERVATIVE ); - } - - static void __exit cpufreq_gov_dbs_exit ( void ) - { - cpufreq_unregister_governor ( CPU_FREQ_GOV_CONSERVATIVE ); - } + #define CPU_FREQ_GOV_CONSERVATIVE (cs_governor.gov)
MODULE_AUTHOR ( "Alexander Clouter <alex@digriz.org.uk>" );
MODULE_DESCRIPTION ( "'cpufreq_conservative' - A dynamic cpufreq governor for " @@ -343,11 +333,9 @@ MODULE_LICENSE("GPL");
#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_CONSERVATIVE
struct cpufreq_governor * cpufreq_default_governor ( void )
{
- return CPU_FREQ_GOV_CONSERVATIVE ; + return & CPU_FREQ_GOV_CONSERVATIVE ; }
- - core_initcall ( cpufreq_gov_dbs_init ); - #else - module_init ( cpufreq_gov_dbs_init ); #endif
- module_exit ( cpufreq_gov_dbs_exit ); + + cpufreq_governor_init ( CPU_FREQ_GOV_CONSERVATIVE ); + cpufreq_governor_exit ( CPU_FREQ_GOV_CONSERVATIVE ); diff --git a/drivers/cpufreq/cpufreq_ondemand.c b/drivers/cpufreq/cpufreq_ondemand.c
index 82a4d37ddecb..ac361a8b1d3b 100644
--- a/drivers/cpufreq/cpufreq_ondemand.c
+++ b/drivers/cpufreq/cpufreq_ondemand.c @@ -408,7 +408,7 @@ static struct dbs_governor od_dbs_gov = {
. start = od_start ,
};
- #define CPU_FREQ_GOV_ONDEMAND (&od_dbs_gov.gov) + #define CPU_FREQ_GOV_ONDEMAND (od_dbs_gov.gov)
static void od_set_powersave_bias ( unsigned int powersave_bias )
{ @@ -429,7 +429,7 @@ static void od_set_powersave_bias(unsigned int powersave_bias)
continue ;
policy = cpufreq_cpu_get_raw ( cpu );
- if ( ! policy || policy -> governor != CPU_FREQ_GOV_ONDEMAND ) + if ( ! policy || policy -> governor != & CPU_FREQ_GOV_ONDEMAND ) continue ;
policy_dbs = policy -> governor_data ; @@ -461,16 +461,6 @@ void od_unregister_powersave_bias_handler(void)
}
EXPORT_SYMBOL_GPL ( od_unregister_powersave_bias_handler );
- static int __init cpufreq_gov_dbs_init ( void ) - { - return cpufreq_register_governor ( CPU_FREQ_GOV_ONDEMAND ); - } - - static void __exit cpufreq_gov_dbs_exit ( void ) - { - cpufreq_unregister_governor ( CPU_FREQ_GOV_ONDEMAND ); - } - MODULE_AUTHOR ( "Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>" );
MODULE_AUTHOR ( "Alexey Starikovskiy <alexey.y.starikovskiy@intel.com>" );
MODULE_DESCRIPTION ( "'cpufreq_ondemand' - A dynamic cpufreq governor for " @@ -480,11 +470,9 @@ MODULE_LICENSE("GPL");
#ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND
struct cpufreq_governor * cpufreq_default_governor ( void )
{
- return CPU_FREQ_GOV_ONDEMAND ; + return & CPU_FREQ_GOV_ONDEMAND ; }
- - core_initcall ( cpufreq_gov_dbs_init ); - #else - module_init ( cpufreq_gov_dbs_init ); #endif
- module_exit ( cpufreq_gov_dbs_exit ); + + cpufreq_governor_init ( CPU_FREQ_GOV_ONDEMAND ); + cpufreq_governor_exit ( CPU_FREQ_GOV_ONDEMAND ); diff --git a/drivers/cpufreq/cpufreq_performance.c b/drivers/cpufreq/cpufreq_performance.c
index def9afe0f5b8..71c1d9aba772 100644
--- a/drivers/cpufreq/cpufreq_performance.c
+++ b/drivers/cpufreq/cpufreq_performance.c @@ -23,16 +23,6 @@ static struct cpufreq_governor cpufreq_gov_performance = {
. limits = cpufreq_gov_performance_limits ,
};
- static int __init cpufreq_gov_performance_init ( void ) - { - return cpufreq_register_governor ( & cpufreq_gov_performance ); - } - - static void __exit cpufreq_gov_performance_exit ( void ) - { - cpufreq_unregister_governor ( & cpufreq_gov_performance ); - } - #ifdef CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE
struct cpufreq_governor * cpufreq_default_governor ( void )
{ @@ -50,5 +40,5 @@ MODULE_AUTHOR("Dominik Brodowski <linux@brodo.de>");
MODULE_DESCRIPTION ( "CPUfreq policy governor 'performance'" );
MODULE_LICENSE ( "GPL" );
- core_initcall ( cpufreq_gov_performance_init ); - module_exit ( cpufreq_gov_performance_exit ); + cpufreq_governor_init ( cpufreq_gov_performance ); + cpufreq_governor_exit ( cpufreq_gov_performance ); diff --git a/drivers/cpufreq/cpufreq_powersave.c b/drivers/cpufreq/cpufreq_powersave.c
index 1ae66019eb83..7749522355b5 100644
--- a/drivers/cpufreq/cpufreq_powersave.c
+++ b/drivers/cpufreq/cpufreq_powersave.c @@ -23,16 +23,6 @@ static struct cpufreq_governor cpufreq_gov_powersave = {
. owner = THIS_MODULE ,
};
- static int __init cpufreq_gov_powersave_init ( void ) - { - return cpufreq_register_governor ( & cpufreq_gov_powersave ); - } - - static void __exit cpufreq_gov_powersave_exit ( void ) - { - cpufreq_unregister_governor ( & cpufreq_gov_powersave ); - } - MODULE_AUTHOR ( "Dominik Brodowski <linux@brodo.de>" );
MODULE_DESCRIPTION ( "CPUfreq policy governor 'powersave'" );
MODULE_LICENSE ( "GPL" ); @@ -42,9 +32,7 @@ struct cpufreq_governor *cpufreq_default_governor(void)
{
return & cpufreq_gov_powersave ;
}
- - core_initcall ( cpufreq_gov_powersave_init ); - #else - module_init ( cpufreq_gov_powersave_init ); #endif
- module_exit ( cpufreq_gov_powersave_exit ); + + cpufreq_governor_init ( cpufreq_gov_powersave ); + cpufreq_governor_exit ( cpufreq_gov_powersave ); diff --git a/drivers/cpufreq/cpufreq_userspace.c b/drivers/cpufreq/cpufreq_userspace.c
index b43e7cd502c5..50a4d7846580 100644
--- a/drivers/cpufreq/cpufreq_userspace.c
+++ b/drivers/cpufreq/cpufreq_userspace.c @@ -126,16 +126,6 @@ static struct cpufreq_governor cpufreq_gov_userspace = {
. owner = THIS_MODULE ,
};
- static int __init cpufreq_gov_userspace_init ( void ) - { - return cpufreq_register_governor ( & cpufreq_gov_userspace ); - } - - static void __exit cpufreq_gov_userspace_exit ( void ) - { - cpufreq_unregister_governor ( & cpufreq_gov_userspace ); - } - MODULE_AUTHOR ( "Dominik Brodowski <linux@brodo.de>, "
"Russell King <rmk@arm.linux.org.uk>" );
MODULE_DESCRIPTION ( "CPUfreq policy governor 'userspace'" ); @@ -146,9 +136,7 @@ struct cpufreq_governor *cpufreq_default_governor(void)
{
return & cpufreq_gov_userspace ;
}
- - core_initcall ( cpufreq_gov_userspace_init ); - #else - module_init ( cpufreq_gov_userspace_init ); #endif
- module_exit ( cpufreq_gov_userspace_exit ); + + cpufreq_governor_init ( cpufreq_gov_userspace ); + cpufreq_governor_exit ( cpufreq_gov_userspace ); diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h
index 3494f6763597..e62b022cb07e 100644
--- a/include/linux/cpufreq.h
+++ b/include/linux/cpufreq.h @@ -577,6 +577,20 @@ unsigned int cpufreq_policy_transition_delay_us(struct cpufreq_policy *policy);
int cpufreq_register_governor ( struct cpufreq_governor * governor );
void cpufreq_unregister_governor ( struct cpufreq_governor * governor );
+ #define cpufreq_governor_init(__governor) \ + static int __init __governor ## _init ( void ) \+ { \+ return cpufreq_register_governor ( & __governor ); \+ } \+ core_initcall ( __governor ## _init ) + + #define cpufreq_governor_exit(__governor) \ + static void __exit __governor ## _exit ( void ) \+ { \+ return cpufreq_unregister_governor ( & __governor ); \+ } \+ module_exit ( __governor ## _exit ) + struct cpufreq_governor * cpufreq_default_governor ( void );
struct cpufreq_governor * cpufreq_fallback_governor ( void );
diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c
index 7fbaee24c824..402a09af9f43 100644
--- a/kernel/sched/cpufreq_schedutil.c
+++ b/kernel/sched/cpufreq_schedutil.c @@ -909,11 +909,7 @@ struct cpufreq_governor *cpufreq_default_governor(void)
}
#endif
- static int __init sugov_register ( void ) - { - return cpufreq_register_governor ( & schedutil_gov ); - } - core_initcall ( sugov_register ); + cpufreq_governor_init ( schedutil_gov );
#ifdef CONFIG_ENERGY_MODEL
extern bool sched_energy_update ; --
2.27.0.290.gba653c62da-goog
Currently, the only way to specify the default CPUfreq governor is via
Kconfig options, which suits users who can build the kernel themselves
perfectly.
However, for those who use a distro-like kernel (such as Android, with
the Generic Kernel Image project), the only way to use a different
default is to boot to userspace, and to then switch using the sysfs
interface. Being able to specify the default governor on the command
line, like is the case for cpuidle, would enable those users to specify
their governor of choice earlier on, and to simplify slighlty the
userspace boot procedure.
To support this use-case, add a kernel command line parameter enabling
to specify a default governor for CPUfreq, which takes precedence over
the builtin default.
This implementation has one notable limitation: the default governor
must be registered before the driver. This is solved for builtin
governors and drivers using appropriate *_initcall() functions. And in
the modular case, this must be reflected as a constraint on the module
loading order.
Signed-off-by: Quentin Perret <redacted>
---
.../admin-guide/kernel-parameters.txt | 5 +++
Documentation/admin-guide/pm/cpufreq.rst | 6 ++--
drivers/cpufreq/cpufreq.c | 34 ++++++++++++++++---
3 files changed, 37 insertions(+), 8 deletions(-)
diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt
index fb95fad81c79..5fd3c9f187eb 100644
--- a/Documentation/admin-guide/kernel-parameters.txt
+++ b/Documentation/admin-guide/kernel-parameters.txt @@ -703,6 +703,11 @@
cpufreq.off=1 [CPU_FREQ]
disable the cpufreq sub-system
+ cpufreq.default_governor=+ [CPU_FREQ] Name of the default cpufreq governor to use.+ This governor must be registered in the kernel before+ the cpufreq driver probes.+ cpu_init_udelay=N
[X86] Delay for N microsec between assert and de-assert
of APIC INIT to start processors. This delay occurs diff --git a/Documentation/admin-guide/pm/cpufreq.rst b/Documentation/admin-guide/pm/cpufreq.rst
index 0c74a7784964..368e612145d2 100644
--- a/Documentation/admin-guide/pm/cpufreq.rst
+++ b/Documentation/admin-guide/pm/cpufreq.rst @@ -147,9 +147,9 @@ CPUs in it.
The next major initialization step for a new policy object is to attach a
scaling governor to it (to begin with, that is the default scaling governor
- determined by the kernel configuration, but it may be changed later- via ``sysfs`` ). First, a pointer to the new policy object is passed to the- governor's ``->init()`` callback which is expected to initialize all of the+ determined by the kernel command line or configuration, but it may be changed+ later via ``sysfs`` ). First, a pointer to the new policy object is passed to+ the governor's ``->init()`` callback which is expected to initialize all of the data structures necessary to handle the given policy and, possibly, to add
a governor ``sysfs`` interface to it. Next, the governor is started by
invoking its ``->start()`` callback. diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index 0128de3603df..0f05caedc320 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c @@ -50,6 +50,9 @@ static LIST_HEAD(cpufreq_governor_list);
#define for_each_governor(__governor) \
list_for_each_entry ( __governor , & cpufreq_governor_list , governor_list )
+ static char cpufreq_param_governor [ CPUFREQ_NAME_LEN ]; + static struct cpufreq_governor * default_governor ; + /**
* The "cpufreq driver" - the arch - or hardware - dependent low
* level driver of CPUFreq support , and its spinlock . This lock @@ -1055,7 +1058,6 @@ __weak struct cpufreq_governor *cpufreq_default_governor(void)
static int cpufreq_init_policy ( struct cpufreq_policy * policy )
{
- struct cpufreq_governor * def_gov = cpufreq_default_governor (); struct cpufreq_governor * gov = NULL ;
unsigned int pol = CPUFREQ_POLICY_UNKNOWN ;
@@ -1065,8 +1067,8 @@ static int cpufreq_init_policy(struct cpufreq_policy *policy)
if ( gov ) {
pr_debug ( "Restoring governor %s for cpu %d \n " ,
policy -> governor -> name , policy -> cpu );
- } else if ( def_gov ) { - gov = def_gov ; + } else if ( default_governor ) { + gov = default_governor ; } else {
return - ENODATA ;
} @@ -1074,8 +1076,8 @@ static int cpufreq_init_policy(struct cpufreq_policy *policy)
/* Use the default policy if there is no last_policy. */
if ( policy -> last_policy ) {
pol = policy -> last_policy ;
- } else if ( def_gov ) { - pol = cpufreq_parse_policy ( def_gov -> name ); + } else if ( default_governor ) { + pol = cpufreq_parse_policy ( default_governor -> name ); /*
* In case the default governor is neiter "performance"
* nor "powersave" , fall back to the initial policy @@ -2196,6 +2198,24 @@ __weak struct cpufreq_governor *cpufreq_fallback_governor(void)
return NULL ;
}
+ static void cpufreq_get_default_governor ( void ) + { + default_governor = cpufreq_parse_governor ( cpufreq_param_governor ); + if ( ! default_governor ) { + if ( * cpufreq_param_governor ) + pr_warn ( "Failed to find %s \n " , cpufreq_param_governor ); + default_governor = cpufreq_default_governor (); + } + } + + static void cpufreq_put_default_governor ( void ) + { + if ( ! default_governor ) + return ; + module_put ( default_governor -> owner ); + default_governor = NULL ; + } + static int cpufreq_init_governor ( struct cpufreq_policy * policy )
{
int ret ; @@ -2701,6 +2721,8 @@ int cpufreq_register_driver(struct cpufreq_driver *driver_data)
if ( driver_data -> setpolicy )
driver_data -> flags |= CPUFREQ_CONST_LOOPS ;
+ else + cpufreq_get_default_governor ();
if ( cpufreq_boost_supported ()) {
ret = create_boost_sysfs_file (); @@ -2769,6 +2791,7 @@ int cpufreq_unregister_driver(struct cpufreq_driver *driver)
subsys_interface_unregister ( & cpufreq_interface );
remove_boost_sysfs_file ();
cpuhp_remove_state_nocalls_cpuslocked ( hp_online );
+ cpufreq_put_default_governor ();
write_lock_irqsave ( & cpufreq_driver_lock , flags );
@@ -2792,4 +2815,5 @@ static int __init cpufreq_core_init(void)
return 0 ;
}
module_param ( off , int , 0444 );
+ module_param_string ( default_governor , cpufreq_param_governor , CPUFREQ_NAME_LEN , 0444 ); core_initcall ( cpufreq_core_init ); --
2.27.0.290.gba653c62da-goog
On Monday 15 Jun 2020 at 17:55:54 (+0100), Quentin Perret wrote: quoted hunk static int cpufreq_init_governor(struct cpufreq_policy *policy)
{
int ret; @@ -2701,6 +2721,8 @@ int cpufreq_register_driver(struct cpufreq_driver *driver_data)
if (driver_data->setpolicy)
driver_data->flags |= CPUFREQ_CONST_LOOPS;
+ else+ cpufreq_get_default_governor();
Looking at this again, it appears that the comment above
cpufreq_parse_governor() confused me a bit -- this needs doing
unconditionally I think.
I'll fix it in v2.
Thanks,
Quentin
On 15-06-20, 17:55, Quentin Perret wrote: Currently, most CPUFreq governors are registered at core_initcall time
when used as default, and module_init otherwise. In preparation for
letting users specify the default governor on the kernel command line,
change all of them to use core_initcall unconditionally, as is already
the case for schedutil and performance. This will enable us to assume
builtin governors have been registered before the builtin CPUFreq
drivers probe.
And since all governors now have similar init/exit patterns, introduce
two new macros cpufreq_governor_{init,exit}() to factorize the code.
Signed-off-by: Quentin Perret <redacted>
---
Note: I couldn't boot-test the change to spudemand, by lack of hardware.
But I can confirm cell_defconfig compiles just fine.
---
.../platforms/cell/cpufreq_spudemand.c | 26 ++-----------------
drivers/cpufreq/cpufreq_conservative.c | 22 ++++------------
drivers/cpufreq/cpufreq_ondemand.c | 24 +++++------------
drivers/cpufreq/cpufreq_performance.c | 14 ++--------
drivers/cpufreq/cpufreq_powersave.c | 18 +++----------
drivers/cpufreq/cpufreq_userspace.c | 18 +++----------
include/linux/cpufreq.h | 14 ++++++++++
kernel/sched/cpufreq_schedutil.c | 6 +----
8 files changed, 36 insertions(+), 106 deletions(-)
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
--
viresh
On 15-06-20, 17:55, Quentin Perret wrote: +static void cpufreq_get_default_governor(void)
+{
+ default_governor = cpufreq_parse_governor(cpufreq_param_governor);
+ if (!default_governor) {
+ if (*cpufreq_param_governor)
+ pr_warn("Failed to find %s\n", cpufreq_param_governor);
+ default_governor = cpufreq_default_governor();
A module_get() never happened for this case and so maybe a
module_put() should never get called.
quoted hunk + }
+}
+
+static void cpufreq_put_default_governor(void)
+{
+ if (!default_governor)
+ return;
+ module_put(default_governor->owner);
+ default_governor = NULL;
+}
+
static int cpufreq_init_governor(struct cpufreq_policy *policy)
{
int ret; @@ -2701,6 +2721,8 @@ int cpufreq_register_driver(struct cpufreq_driver *driver_data)
if (driver_data->setpolicy)
driver_data->flags |= CPUFREQ_CONST_LOOPS;
+ else+ cpufreq_get_default_governor();
if (cpufreq_boost_supported()) {
ret = create_boost_sysfs_file(); @@ -2769,6 +2791,7 @@ int cpufreq_unregister_driver(struct cpufreq_driver *driver)
subsys_interface_unregister(&cpufreq_interface);
remove_boost_sysfs_file();
cpuhp_remove_state_nocalls_cpuslocked(hp_online);
+ cpufreq_put_default_governor();
write_lock_irqsave(&cpufreq_driver_lock, flags);
@@ -2792,4 +2815,5 @@ static int __init cpufreq_core_init(void)
return 0;
}
And since this is a per boot thing, there is perhaps no need of doing
these at driver register/unregister, I would rather do it at:
cpufreq_core_init() time itself and so we will never need to run
cpufreq_put_default_governor() and so can be removed.
And another thing I am not able to understand (despite you commenting
about that in the commit log) is what happens if the default governor
chosen is built as a module ?
--
viresh
Hey Viresh,
On Tuesday 16 Jun 2020 at 10:01:43 (+0530), Viresh Kumar wrote: On 15-06-20, 17:55, Quentin Perret wrote: quoted +static void cpufreq_get_default_governor(void)
+{
+ default_governor = cpufreq_parse_governor(cpufreq_param_governor);
+ if (!default_governor) {
+ if (*cpufreq_param_governor)
+ pr_warn("Failed to find %s\n", cpufreq_param_governor);
+ default_governor = cpufreq_default_governor();
A module_get() never happened for this case and so maybe a
module_put() should never get called.
Correct, however cpufreq_default_governor() being a weak function, we're
basically guaranteed the governor we get from there is builtin, so
gov->owner is NULL. That is, module_put() is not actively useful, but it
doesn't harm. So I figured that should be fine. That could definitely
use a comment, though :)
quoted + }
+}
+
+static void cpufreq_put_default_governor(void)
+{
+ if (!default_governor)
+ return;
+ module_put(default_governor->owner);
+ default_governor = NULL;
+}
+
static int cpufreq_init_governor(struct cpufreq_policy *policy)
{
int ret; @@ -2701,6 +2721,8 @@ int cpufreq_register_driver(struct cpufreq_driver *driver_data)
if (driver_data->setpolicy)
driver_data->flags |= CPUFREQ_CONST_LOOPS;
+ else+ cpufreq_get_default_governor();
if (cpufreq_boost_supported()) {
ret = create_boost_sysfs_file(); @@ -2769,6 +2791,7 @@ int cpufreq_unregister_driver(struct cpufreq_driver *driver)
subsys_interface_unregister(&cpufreq_interface);
remove_boost_sysfs_file();
cpuhp_remove_state_nocalls_cpuslocked(hp_online);
+ cpufreq_put_default_governor();
write_lock_irqsave(&cpufreq_driver_lock, flags);
@@ -2792,4 +2815,5 @@ static int __init cpufreq_core_init(void)
return 0;
}
And since this is a per boot thing, there is perhaps no need of doing
these at driver register/unregister, I would rather do it at:
cpufreq_core_init() time itself and so we will never need to run
cpufreq_put_default_governor() and so can be removed.
Right, so the reason I avoided cpufreq_core_init() was because it is
called at core_initcall() time, which means I can't really assume the
governors have been loaded by that time. By waiting for the driver to
probe before detecting the default gov, we get that nice ordering. But
yes, it feels odd to have it here :/
Thinking about it more, the natural fit for this would rather be the
register/unregister path for governors directly. If that sounds good to
you (?) I'll try to move it there in v2.
And another thing I am not able to understand (despite you commenting
about that in the commit log) is what happens if the default governor
chosen is built as a module ?
So the answer is 'it depends'. If the driver is built as a module too,
then you should load the governor module first, and then the driver
module, and everything will work just fine.
But in the case where the governor is loaded _after_ the driver (either
because we got the module ordering wrong, or because the driver is
builtin), then the policies will be initialized with the builtin
default, and nothing special will happen when the governor module is
loaded.
That behaviour very much is open for discussion, though. A possible
alternative would be to automatically switch all policies to the default
governor upon loading. That would have the nice benefit or removing the
ordering dependency, but that is more involved and I didn't have a
use-case for it, so I went for the simpler option ('the-default
governor-needs-to-be-registered-before-the-policies-are-created').
Thoughts?
Thanks,
Quentin
On Tuesday 16 Jun 2020 at 09:58:31 (+0530), Viresh Kumar wrote: On 15-06-20, 17:55, Quentin Perret wrote: quoted Currently, most CPUFreq governors are registered at core_initcall time
when used as default, and module_init otherwise. In preparation for
letting users specify the default governor on the kernel command line,
change all of them to use core_initcall unconditionally, as is already
the case for schedutil and performance. This will enable us to assume
builtin governors have been registered before the builtin CPUFreq
drivers probe.
And since all governors now have similar init/exit patterns, introduce
two new macros cpufreq_governor_{init,exit}() to factorize the code.
Signed-off-by: Quentin Perret <redacted>
---
Note: I couldn't boot-test the change to spudemand, by lack of hardware.
But I can confirm cell_defconfig compiles just fine.
---
.../platforms/cell/cpufreq_spudemand.c | 26 ++-----------------
drivers/cpufreq/cpufreq_conservative.c | 22 ++++------------
drivers/cpufreq/cpufreq_ondemand.c | 24 +++++------------
drivers/cpufreq/cpufreq_performance.c | 14 ++--------
drivers/cpufreq/cpufreq_powersave.c | 18 +++----------
drivers/cpufreq/cpufreq_userspace.c | 18 +++----------
include/linux/cpufreq.h | 14 ++++++++++
kernel/sched/cpufreq_schedutil.c | 6 +----
8 files changed, 36 insertions(+), 106 deletions(-)
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Thanks!
Quentin
On 16-06-20, 09:31, Quentin Perret wrote: Right, so the reason I avoided cpufreq_core_init() was because it is
called at core_initcall() time, which means I can't really assume the
governors have been loaded by that time. By waiting for the driver to
probe before detecting the default gov, we get that nice ordering. But
yes, it feels odd to have it here :/
Thinking about it more, the natural fit for this would rather be the
register/unregister path for governors directly. If that sounds good to
you (?) I'll try to move it there in v2.
There is another problem here which we need to look at. Any governor
which is built as a module and isn't currently used, should be allowed
to unload. And this needs to be tested by you as well, should be easy
enough.
With the current implementation, you take a reference to the default
governor when the driver is registered and drop it only when the
driver goes away. Which means we won't be able to unload the module of
the governor even if it isn't used. Which is wrong. The solution I
proposed had the same issue as well.
You need to figure out a way where we don't need to keep holding the
module hostage even when it isn't used. I see two ways at least for
the same:
- Do that from the existing place: cpufreq_init_policy().
- And I think this can be done from governor-register/unregister as
well.
Second one sounds good, if it is feasible to do that.
--
viresh
On Tuesday 16 Jun 2020 at 14:57:59 (+0530), Viresh Kumar wrote: There is another problem here which we need to look at. Any governor
which is built as a module and isn't currently used, should be allowed
to unload. And this needs to be tested by you as well, should be easy
enough.
With the current implementation, you take a reference to the default
governor when the driver is registered and drop it only when the
driver goes away. Which means we won't be able to unload the module of
the governor even if it isn't used. Which is wrong. The solution I
proposed had the same issue as well.
You need to figure out a way where we don't need to keep holding the
module hostage even when it isn't used. I see two ways at least for
the same:
- Do that from the existing place: cpufreq_init_policy().
- And I think this can be done from governor-register/unregister as
well.
Second one sounds good, if it is feasible to do that.
Good point.
I'm thinking something along the lines of:
---8<--- diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index 0f05caedc320..a9219404e07f 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c @@ -2340,6 +2340,11 @@ int cpufreq_register_governor(struct cpufreq_governor *governor)
list_add ( & governor -> governor_list , & cpufreq_governor_list );
}
+ if ( ! strncasecmp ( cpufreq_param_governor , governor -> name , CPUFREQ_NAME_LEN )) + default_governor = governor ; + else if ( ! default_governor && cpufreq_default_governor () == governor ) + default_governor = cpufreq_default_governor (); + mutex_unlock ( & cpufreq_governor_mutex );
return err ;
} @@ -2368,6 +2373,8 @@ void cpufreq_unregister_governor(struct cpufreq_governor *governor)
mutex_lock ( & cpufreq_governor_mutex );
list_del ( & governor -> governor_list );
+ if ( governor == default_governor ) + default_governor = cpufreq_default_governor (); mutex_unlock ( & cpufreq_governor_mutex );
}
EXPORT_SYMBOL_GPL ( cpufreq_unregister_governor );
- --> 8 --- should do the trick. That removes the unnecessary reference count, and
feels like a good place to hook things -- that is how cpuidle does it
too IIRC.
I'll double check the locking/synchronization, but that shouldn't be too
bad (famous last words).
Cheers,
Quentin
On 16-06-20, 10:48, Quentin Perret wrote: quoted hunk ---8<--- diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index 0f05caedc320..a9219404e07f 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c @@ -2340,6 +2340,11 @@ int cpufreq_register_governor(struct cpufreq_governor *governor)
list_add ( & governor -> governor_list , & cpufreq_governor_list );
}
+ if ( ! strncasecmp ( cpufreq_param_governor , governor -> name , CPUFREQ_NAME_LEN )) + default_governor = governor ; + else if ( ! default_governor && cpufreq_default_governor () == governor ) + default_governor = cpufreq_default_governor ();
Instead of the else part here, maybe just do this from
cpufreq_core_init() only once, and so we will always have
default_governor set.
quoted hunk +
mutex_unlock(&cpufreq_governor_mutex);
return err;
} @@ -2368,6 +2373,8 @@ void cpufreq_unregister_governor(struct cpufreq_governor *governor)
mutex_lock(&cpufreq_governor_mutex);
list_del(&governor->governor_list);
+ if (governor == default_governor)+ default_governor = cpufreq_default_governor(); mutex_unlock(&cpufreq_governor_mutex);
}
EXPORT_SYMBOL_GPL(cpufreq_unregister_governor);
- -->8---should do the trick. That removes the unnecessary reference count, and
feels like a good place to hook things -- that is how cpuidle does it
too IIRC.
I'll double check the locking/synchronization, but that shouldn't be too
bad (famous last words).
Cheers,
Quentin
--
viresh
On Tuesday 16 Jun 2020 at 15:24:38 (+0530), Viresh Kumar wrote: On 16-06-20, 10:48, Quentin Perret wrote: quoted ---8<--- diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index 0f05caedc320..a9219404e07f 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c @@ -2340,6 +2340,11 @@ int cpufreq_register_governor(struct cpufreq_governor *governor)
list_add ( & governor -> governor_list , & cpufreq_governor_list );
}
+ if ( ! strncasecmp ( cpufreq_param_governor , governor -> name , CPUFREQ_NAME_LEN )) + default_governor = governor ; + else if ( ! default_governor && cpufreq_default_governor () == governor ) + default_governor = cpufreq_default_governor ();
Instead of the else part here, maybe just do this from
cpufreq_core_init() only once, and so we will always have
default_governor set.
Sounds good.
Thanks!
Quentin
Greeting,
FYI, we noticed a -30.2% regression of aim7.jobs-per-min due to commit:
commit: d83f959b5e7a6378a4afbff23de2a2d064d95749 ("[PATCH 2/2] cpufreq: Specify default governor on command line")
url: https://github.com/0day-ci/linux/commits/Quentin-Perret/cpufreq-Specify-the-default-governor-on-command-line/20200616-005920
base: https://git.kernel.org/cgit/linux/kernel/git/rafael/linux-pm.git linux-next
in testcase: aim7
on test machine: 72 threads Intel(R) Xeon(R) Gold 6139 CPU @ 2.30GHz with 128G memory
with following parameters:
disk: 1BRD_48G
fs: xfs
test: disk_src
load: 3000
cpufreq_governor: performance
ucode: 0x2006906
test-description: AIM7 is a traditional UNIX system level benchmark suite which is used to test and measure the performance of multiuser system.
test-url: https://sourceforge.net/projects/aimbench/files/aim-suite7/
In addition to that, the commit also has significant impact on the following tests:
If you fix the issue, kindly add following tag
Reported-by: kernel test robot <redacted>
Details are as below:
-------------------------------------------------------------------------------------------------->
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
bin/lkp install job.yaml # job file is attached in this email
bin/lkp run job.yaml
=========================================================================================
compiler/cpufreq_governor/disk/fs/kconfig/load/rootfs/tbox_group/test/testcase/ucode:
gcc-9/performance/1BRD_48G/xfs/x86_64-rhel-7.6/3000/debian-x86_64-20191114.cgz/lkp-skl-2sp7/disk_src/aim7/0x2006906
commit:
21bf3e69e5 ("cpufreq: Register governors at core_initcall")
d83f959b5e ("cpufreq: Specify default governor on command line")
21bf3e69e51e91cd d83f959b5e7a6378a4afbff23de
---------------- ---------------------------
fail:runs %reproduction fail:runs
| | |
:4 100% 4:4 kmsg.cpufreq:cpufreq_online:Failed_to_initialize_policy_for_cpu:#(-#)
%stddev %change %stddev
\ | \
37531 -30.2% 26214 aim7.jobs-per-min
479.85 +43.2% 686.95 aim7.time.elapsed_time
479.85 +43.2% 686.95 aim7.time.elapsed_time.max
1489 ± 2% +38.4% 2061 aim7.time.involuntary_context_switches
1126 +44.4% 1626 aim7.time.system_time
49.29 +49.9% 73.87 aim7.time.user_time
3.01 +1.7% 3.06 iostat.cpu.system
4535 -32.3% 3070 ± 2% vmstat.io.bo
133979 -29.0% 95145 vmstat.system.cs
24.32 +1.6% 24.71 boot-time.boot
15.32 +3.8% 15.90 boot-time.dhcp
1446 +1.6% 1468 boot-time.idle
5.532e+08 ± 5% -36.1% 3.538e+08 ± 2% cpuidle.C1.time
5739465 ± 2% -31.5% 3933307 cpuidle.C1.usage
78066984 ± 13% +24.5% 97217135 ± 9% cpuidle.C1E.usage
7739904 ± 5% +15.4% 8929792 ± 4% meminfo.DirectMap2M
590700 ± 5% -23.9% 449388 ± 7% meminfo.DirectMap4k
5963 -30.2% 4160 meminfo.max_used_kB
113.75 ± 57% +295.4% 449.75 ± 79% numa-meminfo.node0.Inactive(file)
1040851 ± 6% +17.5% 1222596 ± 14% numa-meminfo.node0.MemUsed
1316 ± 50% +2193.8% 30192 ±143% numa-meminfo.node0.PageTables
129108 ± 6% +33.7% 172589 ± 23% numa-meminfo.node0.Slab
100458 ± 37% +73.4% 174147 ± 21% numa-meminfo.node1.AnonHugePages
81000 ± 5% -16.3% 67824 ± 10% numa-meminfo.node1.KReclaimable
81000 ± 5% -16.3% 67824 ± 10% numa-meminfo.node1.SReclaimable
13.50 ±133% +596.3% 94.00 ±102% numa-vmstat.node0.nr_dirtied
28.25 ± 57% +296.5% 112.00 ± 79% numa-vmstat.node0.nr_inactive_file
328.50 ± 50% +2200.5% 7557 ±143% numa-vmstat.node0.nr_page_table_pages
13.50 ±133% +592.6% 93.50 ±103% numa-vmstat.node0.nr_written
28.25 ± 57% +296.5% 112.00 ± 79% numa-vmstat.node0.nr_zone_inactive_file
48.50 ± 37% +74.2% 84.50 ± 21% numa-vmstat.node1.nr_anon_transparent_hugepages
20249 ± 5% -16.3% 16952 ± 10% numa-vmstat.node1.nr_slab_reclaimable
4779 ± 6% +36.1% 6506 ± 2% slabinfo.Acpi-State.active_objs
4779 ± 6% +36.1% 6506 ± 2% slabinfo.Acpi-State.num_objs
536.25 ± 9% +13.8% 610.25 ± 7% slabinfo.UDP.active_objs
536.25 ± 9% +13.8% 610.25 ± 7% slabinfo.UDP.num_objs
2894 ± 10% +18.2% 3419 ± 6% slabinfo.eventpoll_pwq.active_objs
2894 ± 10% +18.2% 3419 ± 6% slabinfo.eventpoll_pwq.num_objs
648.00 ± 21% +34.6% 872.00 ± 4% slabinfo.kmalloc-rcl-128.active_objs
648.00 ± 21% +34.6% 872.00 ± 4% slabinfo.kmalloc-rcl-128.num_objs
268.25 ± 21% +57.9% 423.50 ± 7% slabinfo.nfs_read_data.active_objs
268.25 ± 21% +57.9% 423.50 ± 7% slabinfo.nfs_read_data.num_objs
685.00 ± 10% +16.8% 800.25 ± 5% slabinfo.skbuff_fclone_cache.active_objs
685.00 ± 10% +16.8% 800.25 ± 5% slabinfo.skbuff_fclone_cache.num_objs
94.50 +6.3% 100.50 proc-vmstat.nr_anon_transparent_hugepages
322.50 +2.8% 331.50 proc-vmstat.nr_dirtied
4433 +2.1% 4526 proc-vmstat.nr_inactive_anon
337.25 +2.6% 346.00 proc-vmstat.nr_inactive_file
322.00 +2.7% 330.75 proc-vmstat.nr_written
4433 +2.1% 4526 proc-vmstat.nr_zone_inactive_anon
337.25 +2.6% 346.00 proc-vmstat.nr_zone_inactive_file
1279746 +32.0% 1688708 proc-vmstat.numa_hit
1256250 +32.6% 1665258 proc-vmstat.numa_local
1568 ± 4% -25.6% 1166 ± 2% proc-vmstat.pgactivate
1474183 +28.3% 1890851 proc-vmstat.pgalloc_normal
1400035 +37.3% 1922052 proc-vmstat.pgfault
1425068 +29.5% 1845014 proc-vmstat.pgfree
857.00 ± 9% +42.1% 1218 ± 5% proc-vmstat.pgpgin
2188373 -3.2% 2118251 ± 2% proc-vmstat.pgpgout
8038 ± 6% +44.2% 11589 sched_debug.cfs_rq:/.exec_clock.avg
11048 ± 2% +38.4% 15291 ± 2% sched_debug.cfs_rq:/.exec_clock.max
6322 ± 7% +42.8% 9029 ± 9% sched_debug.cfs_rq:/.exec_clock.min
750.65 ± 6% +22.8% 922.15 ± 13% sched_debug.cfs_rq:/.exec_clock.stddev
18.09 ± 10% -21.3% 14.24 sched_debug.cfs_rq:/.load_avg.avg
394.62 ± 2% +9.3% 431.44 ± 3% sched_debug.cfs_rq:/.load_avg.max
0.32 ± 13% +121.3% 0.71 ± 11% sched_debug.cfs_rq:/.nr_spread_over.avg
3.22 ± 58% -54.1% 1.48 ± 36% sched_debug.cfs_rq:/.removed.load_avg.avg
116.86 ± 5% -28.2% 83.94 sched_debug.cfs_rq:/.removed.load_avg.max
18.23 ± 26% -40.3% 10.88 ± 17% sched_debug.cfs_rq:/.removed.load_avg.stddev
3.22 ± 58% -54.1% 1.48 ± 36% sched_debug.cfs_rq:/.removed.runnable_avg.avg
116.86 ± 5% -28.2% 83.94 sched_debug.cfs_rq:/.removed.runnable_avg.max
18.23 ± 26% -40.3% 10.88 ± 17% sched_debug.cfs_rq:/.removed.runnable_avg.stddev
149.95 ± 4% -22.3% 116.58 ± 4% sched_debug.cfs_rq:/.runnable_avg.avg
1407 ± 6% -8.3% 1291 ± 3% sched_debug.cfs_rq:/.runnable_avg.max
2.05 ± 20% -70.5% 0.60 ± 35% sched_debug.cfs_rq:/.runnable_avg.min
239.05 ± 8% -17.2% 197.94 sched_debug.cfs_rq:/.runnable_avg.stddev
102746 ± 14% -22.7% 79373 ± 22% sched_debug.cfs_rq:/.spread0.max
74.08 ± 2% -20.9% 58.61 ± 2% sched_debug.cfs_rq:/.util_avg.avg
1.82 ± 25% -73.7% 0.48 ± 46% sched_debug.cfs_rq:/.util_avg.min
123.17 ± 4% -12.9% 107.28 ± 3% sched_debug.cfs_rq:/.util_avg.stddev
60901 ± 7% +15.6% 70419 ± 6% sched_debug.cpu.avg_idle.min
260021 ± 5% +37.8% 358434 sched_debug.cpu.clock.avg
260042 ± 5% +37.8% 358462 sched_debug.cpu.clock.max
260000 ± 5% +37.8% 358404 sched_debug.cpu.clock.min
12.18 ± 2% +36.2% 16.59 ± 10% sched_debug.cpu.clock.stddev
260021 ± 5% +37.8% 358434 sched_debug.cpu.clock_task.avg
260042 ± 5% +37.8% 358462 sched_debug.cpu.clock_task.max
260000 ± 5% +37.8% 358404 sched_debug.cpu.clock_task.min
12.18 ± 2% +36.2% 16.59 ± 10% sched_debug.cpu.clock_task.stddev
10341 ± 3% +24.0% 12822 sched_debug.cpu.curr->pid.max
1421 ± 5% +17.8% 1673 ± 2% sched_debug.cpu.curr->pid.stddev
0.00 +11.2% 0.00 ± 4% sched_debug.cpu.next_balance.stddev
2036 ± 5% +26.6% 2578 sched_debug.cpu.ttwu_local.avg
4220 ± 5% +7.6% 4539 ± 3% sched_debug.cpu.ttwu_local.max
1074 ± 8% +39.8% 1502 ± 4% sched_debug.cpu.ttwu_local.min
260000 ± 5% +37.8% 358404 sched_debug.cpu_clk
259499 ± 5% +37.9% 357901 sched_debug.ktime
260390 ± 5% +37.8% 358806 sched_debug.sched_clk
1.659e+09 -27.6% 1.201e+09 perf-stat.i.branch-instructions
23511802 ± 2% -24.7% 17704718 perf-stat.i.branch-misses
16325833 ± 2% -29.2% 11553674 perf-stat.i.cache-misses
51488691 ± 2% -28.8% 36674264 perf-stat.i.cache-references
134382 -29.1% 95334 perf-stat.i.context-switches
1.14 ± 2% -6.6% 1.06 perf-stat.i.cpi
9.011e+09 -31.8% 6.145e+09 perf-stat.i.cpu-cycles
1591 -37.6% 993.54 perf-stat.i.cpu-migrations
597.52 -6.5% 558.71 perf-stat.i.cycles-between-cache-misses
2117105 ± 2% -29.2% 1499831 perf-stat.i.dTLB-load-misses
1.925e+09 -27.5% 1.396e+09 perf-stat.i.dTLB-loads
93028 ± 7% -28.7% 66336 ± 2% perf-stat.i.dTLB-store-misses
8.944e+08 -26.9% 6.541e+08 perf-stat.i.dTLB-stores
15.06 ± 2% -1.3 13.73 perf-stat.i.iTLB-load-miss-rate%
1042616 -30.9% 720885 perf-stat.i.iTLB-load-misses
5925503 -23.2% 4551782 perf-stat.i.iTLB-loads
8.199e+09 -27.7% 5.929e+09 perf-stat.i.instructions
10027 ± 3% -8.7% 9159 ± 3% perf-stat.i.instructions-per-iTLB-miss
0.91 +6.4% 0.96 perf-stat.i.ipc
0.13 -31.8% 0.09 perf-stat.i.metric.GHz
0.12 +709.6% 0.96 perf-stat.i.metric.K/sec
63.08 -27.4% 45.79 perf-stat.i.metric.M/sec
2820 -3.9% 2710 perf-stat.i.minor-faults
5163419 -29.7% 3630179 perf-stat.i.node-load-misses
845228 ± 3% -29.9% 592653 ± 2% perf-stat.i.node-loads
1288470 -30.3% 897482 perf-stat.i.node-store-misses
125883 -28.0% 90575 perf-stat.i.node-stores
2820 -3.9% 2710 perf-stat.i.page-faults
1.42 ± 2% +0.1 1.47 perf-stat.overall.branch-miss-rate%
1.10 -5.7% 1.04 perf-stat.overall.cpi
14.96 -1.3 13.68 perf-stat.overall.iTLB-load-miss-rate%
7864 +4.6% 8224 perf-stat.overall.instructions-per-iTLB-miss
0.91 +6.0% 0.96 perf-stat.overall.ipc
1.655e+09 -27.5% 1.2e+09 perf-stat.ps.branch-instructions
23464114 ± 2% -24.6% 17681251 perf-stat.ps.branch-misses
16290823 ± 2% -29.2% 11536638 perf-stat.ps.cache-misses
51379893 ± 2% -28.7% 36622024 perf-stat.ps.cache-references
134092 -29.0% 95192 perf-stat.ps.context-switches
8.992e+09 -31.8% 6.136e+09 perf-stat.ps.cpu-cycles
1587 -37.5% 992.51 perf-stat.ps.cpu-migrations
2112603 ± 2% -29.1% 1497800 perf-stat.ps.dTLB-load-misses
1.921e+09 -27.4% 1.394e+09 perf-stat.ps.dTLB-loads
92837 ± 7% -28.6% 66254 ± 2% perf-stat.ps.dTLB-store-misses
8.925e+08 -26.8% 6.532e+08 perf-stat.ps.dTLB-stores
1040393 -30.8% 719911 perf-stat.ps.iTLB-load-misses
5912777 -23.1% 4544925 perf-stat.ps.iTLB-loads
8.181e+09 -27.6% 5.921e+09 perf-stat.ps.instructions
2816 -3.8% 2708 perf-stat.ps.minor-faults
5152279 -29.6% 3624732 perf-stat.ps.node-load-misses
843418 ± 3% -29.8% 591785 ± 2% perf-stat.ps.node-loads
1285691 -30.3% 896142 perf-stat.ps.node-store-misses
125624 -28.0% 90468 perf-stat.ps.node-stores
2816 -3.8% 2708 perf-stat.ps.page-faults
3.934e+12 +3.5% 4.072e+12 perf-stat.total.instructions
148225 ± 3% +34.9% 199978 ± 2% softirqs.CPU0.RCU
68852 +36.5% 94003 softirqs.CPU0.SCHED
161692 ± 5% +56.1% 252411 ± 3% softirqs.CPU0.TIMER
149090 ± 3% +32.7% 197803 ± 3% softirqs.CPU1.RCU
64130 +38.9% 89072 ± 2% softirqs.CPU1.SCHED
156371 ± 6% +59.5% 249452 ± 3% softirqs.CPU1.TIMER
149696 ± 3% +35.0% 202059 ± 2% softirqs.CPU10.RCU
62512 ± 2% +44.0% 89989 softirqs.CPU10.SCHED
156527 ± 6% +60.0% 250425 ± 3% softirqs.CPU10.TIMER
147497 ± 2% +35.1% 199264 softirqs.CPU11.RCU
62553 +41.7% 88638 ± 2% softirqs.CPU11.SCHED
155703 ± 4% +60.2% 249398 ± 4% softirqs.CPU11.TIMER
148757 ± 3% +35.2% 201173 ± 2% softirqs.CPU12.RCU
63210 +42.0% 89782 softirqs.CPU12.SCHED
156322 ± 6% +59.8% 249797 ± 4% softirqs.CPU12.TIMER
149867 ± 3% +31.6% 197237 ± 5% softirqs.CPU13.RCU
63165 +44.3% 91168 ± 2% softirqs.CPU13.SCHED
157012 ± 5% +77.0% 277943 ± 13% softirqs.CPU13.TIMER
145801 ± 3% +36.7% 199320 softirqs.CPU14.RCU
62885 +42.0% 89294 softirqs.CPU14.SCHED
154607 ± 6% +61.1% 249128 ± 4% softirqs.CPU14.TIMER
125830 +47.3% 185328 softirqs.CPU15.RCU
63213 +39.1% 87898 ± 2% softirqs.CPU15.SCHED
155314 ± 5% +60.2% 248821 ± 3% softirqs.CPU15.TIMER
127243 +47.1% 187237 softirqs.CPU16.RCU
62817 ± 2% +42.8% 89726 softirqs.CPU16.SCHED
155933 ± 5% +60.1% 249636 ± 3% softirqs.CPU16.TIMER
128980 ± 2% +45.0% 187015 softirqs.CPU17.RCU
63146 +38.4% 87373 ± 2% softirqs.CPU17.SCHED
159461 ± 5% +56.9% 250262 ± 3% softirqs.CPU17.TIMER
149393 +35.6% 202597 softirqs.CPU18.RCU
66228 ± 2% +36.9% 90683 softirqs.CPU18.SCHED
159695 +57.9% 252201 softirqs.CPU18.TIMER
145542 +38.5% 201516 softirqs.CPU19.RCU
64394 ± 2% +39.9% 90093 softirqs.CPU19.SCHED
148693 ± 5% +64.3% 244278 softirqs.CPU19.TIMER
146496 ± 2% +36.2% 199533 softirqs.CPU2.RCU
62601 +42.5% 89193 ± 2% softirqs.CPU2.SCHED
156672 ± 4% +60.6% 251583 ± 3% softirqs.CPU2.TIMER
145548 ± 7% +41.3% 205694 softirqs.CPU20.RCU
64297 ± 2% +38.9% 89319 softirqs.CPU20.SCHED
160604 ± 8% +54.4% 248021 softirqs.CPU20.TIMER
150370 +36.5% 205196 softirqs.CPU21.RCU
64605 +37.7% 88956 ± 2% softirqs.CPU21.SCHED
154841 ± 2% +59.3% 246625 ± 2% softirqs.CPU21.TIMER
149245 +36.9% 204303 softirqs.CPU22.RCU
63598 +39.5% 88735 ± 2% softirqs.CPU22.SCHED
152785 ± 2% +60.4% 245140 softirqs.CPU22.TIMER
145943 +37.6% 200890 softirqs.CPU23.RCU
63605 ± 2% +41.2% 89835 softirqs.CPU23.SCHED
147754 ± 3% +64.9% 243701 softirqs.CPU23.TIMER
148007 +36.6% 202142 softirqs.CPU24.RCU
64937 ± 2% +38.3% 89833 softirqs.CPU24.SCHED
152782 ± 3% +59.2% 243256 softirqs.CPU24.TIMER
149238 +34.4% 200565 ± 2% softirqs.CPU25.RCU
64300 ± 2% +42.3% 91524 ± 3% softirqs.CPU25.SCHED
153715 ± 2% +77.9% 273383 ± 17% softirqs.CPU25.TIMER
149265 +36.5% 203791 softirqs.CPU26.RCU
64902 ± 2% +39.6% 90608 softirqs.CPU26.SCHED
155441 +57.7% 245113 softirqs.CPU26.TIMER
146233 +37.8% 201483 softirqs.CPU27.RCU
63266 ± 2% +39.5% 88228 ± 3% softirqs.CPU27.SCHED
147013 ± 4% +65.5% 243357 softirqs.CPU27.TIMER
145540 ± 3% +38.8% 201950 softirqs.CPU28.RCU
65407 ± 3% +36.7% 89434 ± 3% softirqs.CPU28.SCHED
148271 +36.3% 202125 softirqs.CPU29.RCU
64378 +38.8% 89361 softirqs.CPU29.SCHED
149114 ± 4% +62.7% 242571 softirqs.CPU29.TIMER
149246 ± 3% +34.5% 200795 softirqs.CPU3.RCU
63878 +37.7% 87943 ± 3% softirqs.CPU3.SCHED
157953 ± 6% +58.6% 250442 ± 3% softirqs.CPU3.TIMER
138500 +35.4% 187554 ± 5% softirqs.CPU30.RCU
63410 ± 5% +42.1% 90114 ± 2% softirqs.CPU30.SCHED
153064 ± 3% +61.1% 246521 ± 3% softirqs.CPU30.TIMER
138816 +39.5% 193704 ± 2% softirqs.CPU31.RCU
64501 ± 2% +40.2% 90406 ± 4% softirqs.CPU31.SCHED
152639 ± 4% +78.3% 272192 ± 16% softirqs.CPU31.TIMER
136214 +42.4% 193926 softirqs.CPU32.RCU
65755 ± 4% +37.3% 90313 softirqs.CPU32.SCHED
136873 +42.2% 194572 softirqs.CPU33.RCU
64588 ± 2% +37.1% 88570 ± 2% softirqs.CPU33.SCHED
149425 ± 3% +63.1% 243776 softirqs.CPU33.TIMER
137813 +41.7% 195263 softirqs.CPU34.RCU
63090 ± 2% +43.5% 90558 softirqs.CPU34.SCHED
150682 ± 3% +62.5% 244916 softirqs.CPU34.TIMER
138930 +40.8% 195547 softirqs.CPU35.RCU
64733 ± 2% +32.7% 85915 ± 6% softirqs.CPU35.SCHED
154054 ± 2% +58.8% 244669 softirqs.CPU35.TIMER
146677 ± 2% +38.8% 203648 softirqs.CPU36.RCU
61997 +42.5% 88335 softirqs.CPU36.SCHED
151515 ± 4% +63.8% 248136 ± 3% softirqs.CPU36.TIMER
148944 ± 3% +35.6% 201924 ± 2% softirqs.CPU37.RCU
62089 +40.3% 87128 ± 3% softirqs.CPU37.SCHED
152938 ± 5% +62.0% 247793 ± 3% softirqs.CPU37.TIMER
147059 +39.2% 204753 softirqs.CPU38.RCU
62171 +42.7% 88691 softirqs.CPU38.SCHED
152184 ± 5% +62.7% 247603 ± 3% softirqs.CPU38.TIMER
149805 ± 4% +37.0% 205264 softirqs.CPU39.RCU
61107 ± 2% +45.3% 88786 softirqs.CPU39.SCHED
152465 ± 4% +62.1% 247221 ± 3% softirqs.CPU39.TIMER
150072 ± 3% +34.4% 201733 softirqs.CPU4.RCU
63819 +40.5% 89697 softirqs.CPU4.SCHED
158029 ± 6% +58.6% 250631 ± 3% softirqs.CPU4.TIMER
152095 ± 3% +36.8% 208069 softirqs.CPU40.RCU
62807 +41.3% 88778 softirqs.CPU40.SCHED
152878 ± 5% +62.5% 248464 ± 3% softirqs.CPU40.TIMER
147032 ± 2% +39.3% 204813 softirqs.CPU41.RCU
60510 ± 5% +47.6% 89284 softirqs.CPU41.SCHED
152687 ± 3% +62.1% 247555 ± 3% softirqs.CPU41.TIMER
147261 ± 2% +38.9% 204553 softirqs.CPU42.RCU
62877 +40.8% 88551 softirqs.CPU42.SCHED
153002 ± 4% +62.1% 248053 ± 3% softirqs.CPU42.TIMER
149469 ± 2% +37.7% 205807 softirqs.CPU43.RCU
61857 ± 2% +43.1% 88502 softirqs.CPU43.SCHED
152893 ± 4% +62.0% 247754 ± 3% softirqs.CPU43.TIMER
144650 ± 9% +42.0% 205431 softirqs.CPU44.RCU
63153 +36.4% 86145 ± 5% softirqs.CPU44.SCHED
152723 ± 5% +62.2% 247666 ± 3% softirqs.CPU44.TIMER
142302 ± 2% +42.7% 203007 softirqs.CPU45.RCU
61870 +42.5% 88138 softirqs.CPU45.SCHED
151680 ± 4% +62.5% 246506 ± 3% softirqs.CPU45.TIMER
150475 ± 2% +37.1% 206324 softirqs.CPU46.RCU
61258 ± 2% +45.8% 89321 softirqs.CPU46.SCHED
152918 ± 5% +62.0% 247671 ± 3% softirqs.CPU46.TIMER
149712 ± 2% +37.7% 206114 softirqs.CPU47.RCU
61526 ± 4% +44.4% 88873 softirqs.CPU47.SCHED
152360 ± 4% +62.2% 247203 ± 3% softirqs.CPU47.TIMER
149904 ± 3% +38.1% 206972 softirqs.CPU48.RCU
62334 +42.1% 88572 softirqs.CPU48.SCHED
152221 ± 4% +61.9% 246446 ± 4% softirqs.CPU48.TIMER
152162 ± 2% +32.4% 201405 ± 6% softirqs.CPU49.RCU
62446 +40.8% 87927 ± 2% softirqs.CPU49.SCHED
151903 ± 4% +56.9% 238395 ± 10% softirqs.CPU49.TIMER
148700 ± 3% +35.2% 200969 softirqs.CPU5.RCU
63266 +40.3% 88735 softirqs.CPU5.SCHED
155744 ± 5% +60.9% 250589 ± 3% softirqs.CPU5.TIMER
144528 ± 5% +41.8% 204869 softirqs.CPU50.RCU
60457 ± 3% +45.1% 87749 ± 2% softirqs.CPU50.SCHED
152440 ± 3% +61.2% 245785 ± 4% softirqs.CPU50.TIMER
148237 ± 2% +37.9% 204469 softirqs.CPU51.RCU
60863 ± 4% +44.6% 88006 softirqs.CPU51.SCHED
152274 ± 4% +61.9% 246595 ± 3% softirqs.CPU51.TIMER
149127 ± 2% +38.4% 206376 softirqs.CPU52.RCU
62360 +43.4% 89443 softirqs.CPU52.SCHED
152512 ± 5% +62.2% 247332 ± 3% softirqs.CPU52.TIMER
151816 ± 3% +36.1% 206605 softirqs.CPU53.RCU
61809 ± 4% +44.6% 89378 softirqs.CPU53.SCHED
153584 ± 4% +61.4% 247836 ± 3% softirqs.CPU53.TIMER
138286 +41.3% 195411 softirqs.CPU54.RCU
63203 ± 3% +43.8% 90857 softirqs.CPU54.SCHED
148682 ± 3% +67.6% 249252 softirqs.CPU54.TIMER
136334 +43.5% 195592 softirqs.CPU55.RCU
63586 ± 2% +40.2% 89146 softirqs.CPU55.SCHED
145454 ± 5% +66.1% 241627 softirqs.CPU55.TIMER
137781 ± 2% +42.9% 196874 softirqs.CPU56.RCU
64602 ± 2% +34.8% 87089 ± 4% softirqs.CPU56.SCHED
139229 +41.5% 197025 softirqs.CPU57.RCU
59929 ± 8% +49.5% 89584 softirqs.CPU57.SCHED
145306 ± 4% +66.2% 241553 softirqs.CPU57.TIMER
138590 +42.8% 197852 softirqs.CPU58.RCU
63888 ± 2% +33.2% 85085 ± 8% softirqs.CPU58.SCHED
144762 ± 5% +67.1% 241869 softirqs.CPU58.TIMER
134272 ± 2% +43.9% 193254 softirqs.CPU59.RCU
62096 ± 2% +42.7% 88626 ± 2% softirqs.CPU59.SCHED
142782 ± 5% +68.7% 240865 softirqs.CPU59.TIMER
147604 ± 3% +35.4% 199910 softirqs.CPU6.RCU
63325 +41.3% 89494 softirqs.CPU6.SCHED
156352 ± 4% +59.9% 250043 ± 3% softirqs.CPU6.TIMER
137781 +37.2% 189074 softirqs.CPU60.RCU
61228 ± 6% +46.7% 89834 softirqs.CPU60.SCHED
147828 ± 4% +62.9% 240826 ± 2% softirqs.CPU60.TIMER
139277 +32.6% 184693 ± 5% softirqs.CPU61.RCU
60631 ± 3% +44.8% 87809 ± 3% softirqs.CPU61.SCHED
147103 ± 4% +57.3% 231356 ± 8% softirqs.CPU61.TIMER
139039 +37.0% 190516 softirqs.CPU62.RCU
63559 +41.4% 89882 softirqs.CPU62.SCHED
145849 ± 4% +65.1% 240854 ± 2% softirqs.CPU62.TIMER
134528 ± 2% +40.6% 189101 softirqs.CPU63.RCU
62093 +41.9% 88085 ± 2% softirqs.CPU63.SCHED
144435 ± 5% +66.8% 240912 ± 2% softirqs.CPU63.TIMER
131970 ± 7% +43.4% 189308 softirqs.CPU64.RCU
61853 ± 4% +43.5% 88780 softirqs.CPU64.SCHED
145033 ± 5% +66.0% 240731 ± 2% softirqs.CPU64.TIMER
139350 +36.7% 190526 softirqs.CPU65.RCU
61564 ± 5% +45.0% 89289 ± 2% softirqs.CPU65.SCHED
145387 ± 4% +65.2% 240188 ± 2% softirqs.CPU65.TIMER
138915 +35.3% 187968 ± 2% softirqs.CPU66.RCU
62678 ± 2% +43.3% 89802 ± 5% softirqs.CPU66.SCHED
145202 ± 5% +86.0% 270051 ± 18% softirqs.CPU66.TIMER
138859 ± 2% +33.2% 184905 ± 5% softirqs.CPU67.RCU
63267 +40.4% 88853 softirqs.CPU67.SCHED
145227 ± 5% +58.7% 230424 ± 8% softirqs.CPU67.TIMER
129176 ± 6% +45.1% 187491 softirqs.CPU68.RCU
60993 ± 3% +46.1% 89128 softirqs.CPU68.SCHED
144409 ± 6% +65.7% 239313 ± 2% softirqs.CPU68.TIMER
135472 ± 2% +39.2% 188515 softirqs.CPU69.RCU
62141 ± 2% +39.7% 86793 ± 4% softirqs.CPU69.SCHED
144417 ± 5% +67.4% 241757 ± 2% softirqs.CPU69.TIMER
147138 ± 3% +35.1% 198748 softirqs.CPU7.RCU
63426 +40.9% 89340 softirqs.CPU7.SCHED
156336 ± 5% +60.3% 250589 ± 3% softirqs.CPU7.TIMER
137717 +37.8% 189723 softirqs.CPU70.RCU
63375 ± 2% +41.5% 89708 softirqs.CPU70.SCHED
145921 ± 4% +66.0% 242291 softirqs.CPU70.TIMER
139796 +36.5% 190794 softirqs.CPU71.RCU
62089 ± 4% +44.3% 89621 softirqs.CPU71.SCHED
145373 ± 4% +66.5% 242008 softirqs.CPU71.TIMER
145025 ± 8% +38.2% 200408 softirqs.CPU8.RCU
64759 ± 2% +37.0% 88724 softirqs.CPU8.SCHED
185013 ± 19% +35.6% 250815 ± 3% softirqs.CPU8.TIMER
145962 ± 2% +36.7% 199584 softirqs.CPU9.RCU
62967 +41.8% 89283 softirqs.CPU9.SCHED
153151 ± 5% +62.8% 249403 ± 3% softirqs.CPU9.TIMER
10338644 +38.1% 14279978 softirqs.RCU
4540831 +41.2% 6411461 softirqs.SCHED
11032630 +61.3% 17798417 softirqs.TIMER
410.75 ± 3% +44.1% 591.75 ± 3% interrupts.35:PCI-MSI.31981568-edge.i40e-0000:3d:00.0:misc
1682 ±172% -100.0% 0.75 ±173% interrupts.87:PCI-MSI.31981620-edge.i40e-eth0-TxRx-51
962.50 +43.1% 1377 interrupts.9:IO-APIC.9-fasteoi.acpi
960488 +43.2% 1375718 interrupts.CPU0.LOC:Local_timer_interrupts
962.50 +43.1% 1377 interrupts.CPU1.9:IO-APIC.9-fasteoi.acpi
960657 +43.2% 1375739 interrupts.CPU1.LOC:Local_timer_interrupts
581.50 ± 5% -36.6% 368.50 ± 2% interrupts.CPU1.NMI:Non-maskable_interrupts
581.50 ± 5% -36.6% 368.50 ± 2% interrupts.CPU1.PMI:Performance_monitoring_interrupts
960352 +43.3% 1375770 interrupts.CPU10.LOC:Local_timer_interrupts
545.75 ± 12% -42.8% 312.25 ± 8% interrupts.CPU10.NMI:Non-maskable_interrupts
545.75 ± 12% -42.8% 312.25 ± 8% interrupts.CPU10.PMI:Performance_monitoring_interrupts
3029 ± 4% +52.2% 4609 ± 4% interrupts.CPU10.RES:Rescheduling_interrupts
960563 +43.2% 1375749 interrupts.CPU11.LOC:Local_timer_interrupts
580.75 ± 12% -38.9% 355.00 ± 7% interrupts.CPU11.NMI:Non-maskable_interrupts
580.75 ± 12% -38.9% 355.00 ± 7% interrupts.CPU11.PMI:Performance_monitoring_interrupts
3049 ± 5% +58.8% 4841 ± 3% interrupts.CPU11.RES:Rescheduling_interrupts
960595 +43.2% 1375735 interrupts.CPU12.LOC:Local_timer_interrupts
568.50 ± 7% -33.6% 377.50 ± 2% interrupts.CPU12.NMI:Non-maskable_interrupts
568.50 ± 7% -33.6% 377.50 ± 2% interrupts.CPU12.PMI:Performance_monitoring_interrupts
3001 ± 4% +53.8% 4615 ± 4% interrupts.CPU12.RES:Rescheduling_interrupts
960551 +43.2% 1375659 interrupts.CPU13.LOC:Local_timer_interrupts
571.50 ± 5% -30.6% 396.50 ± 7% interrupts.CPU13.NMI:Non-maskable_interrupts
571.50 ± 5% -30.6% 396.50 ± 7% interrupts.CPU13.PMI:Performance_monitoring_interrupts
3054 ± 3% +53.2% 4678 ± 6% interrupts.CPU13.RES:Rescheduling_interrupts
960642 +43.2% 1375651 interrupts.CPU14.LOC:Local_timer_interrupts
582.75 ± 9% -42.6% 334.25 ± 4% interrupts.CPU14.NMI:Non-maskable_interrupts
582.75 ± 9% -42.6% 334.25 ± 4% interrupts.CPU14.PMI:Performance_monitoring_interrupts
2990 ± 3% +50.8% 4511 ± 8% interrupts.CPU14.RES:Rescheduling_interrupts
960608 +43.2% 1375557 interrupts.CPU15.LOC:Local_timer_interrupts
560.75 ± 4% -30.5% 389.75 ± 8% interrupts.CPU15.NMI:Non-maskable_interrupts
560.75 ± 4% -30.5% 389.75 ± 8% interrupts.CPU15.PMI:Performance_monitoring_interrupts
3037 ± 2% +49.8% 4551 ± 7% interrupts.CPU15.RES:Rescheduling_interrupts
960636 +43.2% 1375721 interrupts.CPU16.LOC:Local_timer_interrupts
547.00 ± 8% -29.3% 387.00 ± 13% interrupts.CPU16.NMI:Non-maskable_interrupts
547.00 ± 8% -29.3% 387.00 ± 13% interrupts.CPU16.PMI:Performance_monitoring_interrupts
2956 ± 3% +52.3% 4502 ± 2% interrupts.CPU16.RES:Rescheduling_interrupts
960558 +43.2% 1375685 interrupts.CPU17.LOC:Local_timer_interrupts
2954 ± 6% +47.1% 4345 ± 3% interrupts.CPU17.RES:Rescheduling_interrupts
960329 +43.2% 1375542 interrupts.CPU18.LOC:Local_timer_interrupts
598.25 ± 10% -35.2% 387.50 ± 3% interrupts.CPU18.NMI:Non-maskable_interrupts
598.25 ± 10% -35.2% 387.50 ± 3% interrupts.CPU18.PMI:Performance_monitoring_interrupts
3027 ± 5% +56.4% 4735 ± 2% interrupts.CPU18.RES:Rescheduling_interrupts
960405 +43.3% 1375801 interrupts.CPU19.LOC:Local_timer_interrupts
533.75 ± 11% -41.1% 314.25 ± 16% interrupts.CPU19.NMI:Non-maskable_interrupts
533.75 ± 11% -41.1% 314.25 ± 16% interrupts.CPU19.PMI:Performance_monitoring_interrupts
2878 ± 3% +54.6% 4449 ± 2% interrupts.CPU19.RES:Rescheduling_interrupts
960642 +43.2% 1375585 interrupts.CPU2.LOC:Local_timer_interrupts
574.25 ± 17% -35.4% 371.25 ± 6% interrupts.CPU2.NMI:Non-maskable_interrupts
574.25 ± 17% -35.4% 371.25 ± 6% interrupts.CPU2.PMI:Performance_monitoring_interrupts
3107 ± 4% +58.3% 4921 ± 12% interrupts.CPU2.RES:Rescheduling_interrupts
960247 +43.3% 1375760 interrupts.CPU20.LOC:Local_timer_interrupts
547.50 ± 8% -48.4% 282.25 ± 24% interrupts.CPU20.NMI:Non-maskable_interrupts
547.50 ± 8% -48.4% 282.25 ± 24% interrupts.CPU20.PMI:Performance_monitoring_interrupts
2940 ± 6% +53.1% 4501 ± 5% interrupts.CPU20.RES:Rescheduling_interrupts
960353 +43.3% 1375737 interrupts.CPU21.LOC:Local_timer_interrupts
2926 ± 4% +54.6% 4523 interrupts.CPU21.RES:Rescheduling_interrupts
960365 +43.3% 1375751 interrupts.CPU22.LOC:Local_timer_interrupts
503.50 ± 13% -32.5% 339.75 ± 8% interrupts.CPU22.NMI:Non-maskable_interrupts
503.50 ± 13% -32.5% 339.75 ± 8% interrupts.CPU22.PMI:Performance_monitoring_interrupts
2936 ± 6% +53.3% 4500 interrupts.CPU22.RES:Rescheduling_interrupts
960336 +43.3% 1375691 interrupts.CPU23.LOC:Local_timer_interrupts
479.75 ± 12% -23.6% 366.75 ± 6% interrupts.CPU23.NMI:Non-maskable_interrupts
479.75 ± 12% -23.6% 366.75 ± 6% interrupts.CPU23.PMI:Performance_monitoring_interrupts
2958 ± 6% +56.2% 4619 ± 4% interrupts.CPU23.RES:Rescheduling_interrupts
960345 +43.2% 1375689 interrupts.CPU24.LOC:Local_timer_interrupts
549.00 ± 8% -33.3% 366.00 ± 8% interrupts.CPU24.NMI:Non-maskable_interrupts
549.00 ± 8% -33.3% 366.00 ± 8% interrupts.CPU24.PMI:Performance_monitoring_interrupts
2976 ± 4% +59.1% 4736 ± 4% interrupts.CPU24.RES:Rescheduling_interrupts
960335 +43.3% 1375732 interrupts.CPU25.LOC:Local_timer_interrupts
541.50 ± 11% -46.2% 291.50 ± 22% interrupts.CPU25.NMI:Non-maskable_interrupts
541.50 ± 11% -46.2% 291.50 ± 22% interrupts.CPU25.PMI:Performance_monitoring_interrupts
2986 ± 4% +59.7% 4770 interrupts.CPU25.RES:Rescheduling_interrupts
960059 +43.3% 1375626 interrupts.CPU26.LOC:Local_timer_interrupts
529.00 ± 9% -42.8% 302.50 ± 11% interrupts.CPU26.NMI:Non-maskable_interrupts
529.00 ± 9% -42.8% 302.50 ± 11% interrupts.CPU26.PMI:Performance_monitoring_interrupts
3041 +52.3% 4631 interrupts.CPU26.RES:Rescheduling_interrupts
960367 +43.2% 1375528 interrupts.CPU27.LOC:Local_timer_interrupts
2905 ± 4% +70.9% 4963 ± 6% interrupts.CPU27.RES:Rescheduling_interrupts
960325 +43.2% 1375585 interrupts.CPU28.LOC:Local_timer_interrupts
3009 ± 11% +57.6% 4742 ± 2% interrupts.CPU28.RES:Rescheduling_interrupts
960290 +43.3% 1375648 interrupts.CPU29.LOC:Local_timer_interrupts
2936 ± 14% +58.6% 4658 ± 2% interrupts.CPU29.RES:Rescheduling_interrupts
960600 +43.2% 1375689 interrupts.CPU3.LOC:Local_timer_interrupts
546.25 ± 11% -34.9% 355.50 ± 12% interrupts.CPU3.NMI:Non-maskable_interrupts
546.25 ± 11% -34.9% 355.50 ± 12% interrupts.CPU3.PMI:Performance_monitoring_interrupts
960286 +43.3% 1375748 interrupts.CPU30.LOC:Local_timer_interrupts
2853 ± 9% +61.4% 4604 ± 2% interrupts.CPU30.RES:Rescheduling_interrupts
960310 +43.3% 1375744 interrupts.CPU31.LOC:Local_timer_interrupts
2919 ± 6% +51.1% 4410 ± 7% interrupts.CPU31.RES:Rescheduling_interrupts
960330 +43.2% 1375667 interrupts.CPU32.LOC:Local_timer_interrupts
2996 ± 2% +46.1% 4375 ± 12% interrupts.CPU32.RES:Rescheduling_interrupts
960234 +43.2% 1375505 interrupts.CPU33.LOC:Local_timer_interrupts
547.50 ± 13% -30.2% 382.25 ± 6% interrupts.CPU33.NMI:Non-maskable_interrupts
547.50 ± 13% -30.2% 382.25 ± 6% interrupts.CPU33.PMI:Performance_monitoring_interrupts
3075 ± 6% +45.8% 4482 ± 10% interrupts.CPU33.RES:Rescheduling_interrupts
960291 +43.3% 1375718 interrupts.CPU34.LOC:Local_timer_interrupts
2883 ± 6% +56.4% 4511 ± 8% interrupts.CPU34.RES:Rescheduling_interrupts
960236 +43.3% 1375694 interrupts.CPU35.LOC:Local_timer_interrupts
2759 ± 6% +71.9% 4745 ± 5% interrupts.CPU35.RES:Rescheduling_interrupts
960636 +43.2% 1375756 interrupts.CPU36.LOC:Local_timer_interrupts
3057 ± 7% +46.1% 4466 ± 6% interrupts.CPU36.RES:Rescheduling_interrupts
960637 +43.2% 1375794 interrupts.CPU37.LOC:Local_timer_interrupts
584.75 ± 6% -39.5% 353.75 ± 4% interrupts.CPU37.NMI:Non-maskable_interrupts
584.75 ± 6% -39.5% 353.75 ± 4% interrupts.CPU37.PMI:Performance_monitoring_interrupts
3144 ± 6% +38.6% 4357 ± 13% interrupts.CPU37.RES:Rescheduling_interrupts
960594 +43.2% 1375658 interrupts.CPU38.LOC:Local_timer_interrupts
582.25 ± 16% -37.9% 361.50 ± 6% interrupts.CPU38.NMI:Non-maskable_interrupts
582.25 ± 16% -37.9% 361.50 ± 6% interrupts.CPU38.PMI:Performance_monitoring_interrupts
3104 ± 2% +40.8% 4373 ± 9% interrupts.CPU38.RES:Rescheduling_interrupts
960595 +43.2% 1375668 interrupts.CPU39.LOC:Local_timer_interrupts
575.25 ± 14% -39.5% 347.75 ± 12% interrupts.CPU39.NMI:Non-maskable_interrupts
575.25 ± 14% -39.5% 347.75 ± 12% interrupts.CPU39.PMI:Performance_monitoring_interrupts
3109 ± 5% +44.9% 4505 ± 5% interrupts.CPU39.RES:Rescheduling_interrupts
960626 +43.2% 1375725 interrupts.CPU4.LOC:Local_timer_interrupts
578.50 ± 8% -40.5% 344.25 ± 10% interrupts.CPU4.NMI:Non-maskable_interrupts
578.50 ± 8% -40.5% 344.25 ± 10% interrupts.CPU4.PMI:Performance_monitoring_interrupts
3200 ± 4% +43.1% 4579 ± 5% interrupts.CPU4.RES:Rescheduling_interrupts
960614 +43.2% 1375706 interrupts.CPU40.LOC:Local_timer_interrupts
581.00 ± 14% -40.1% 348.25 ± 9% interrupts.CPU40.NMI:Non-maskable_interrupts
581.00 ± 14% -40.1% 348.25 ± 9% interrupts.CPU40.PMI:Performance_monitoring_interrupts
3067 ± 6% +55.2% 4762 ± 3% interrupts.CPU40.RES:Rescheduling_interrupts
960561 +43.2% 1375676 interrupts.CPU41.LOC:Local_timer_interrupts
504.50 ± 21% -35.0% 327.75 ± 21% interrupts.CPU41.NMI:Non-maskable_interrupts
504.50 ± 21% -35.0% 327.75 ± 21% interrupts.CPU41.PMI:Performance_monitoring_interrupts
3106 ± 4% +46.0% 4536 ± 4% interrupts.CPU41.RES:Rescheduling_interrupts
960651 +43.2% 1375751 interrupts.CPU42.LOC:Local_timer_interrupts
503.25 ± 23% -42.7% 288.50 ± 10% interrupts.CPU42.NMI:Non-maskable_interrupts
503.25 ± 23% -42.7% 288.50 ± 10% interrupts.CPU42.PMI:Performance_monitoring_interrupts
3104 ± 2% +46.1% 4536 ± 3% interrupts.CPU42.RES:Rescheduling_interrupts
960627 +43.2% 1375681 interrupts.CPU43.LOC:Local_timer_interrupts
552.00 ± 11% -40.7% 327.50 ± 15% interrupts.CPU43.NMI:Non-maskable_interrupts
552.00 ± 11% -40.7% 327.50 ± 15% interrupts.CPU43.PMI:Performance_monitoring_interrupts
3032 ± 2% +55.4% 4711 ± 2% interrupts.CPU43.RES:Rescheduling_interrupts
960513 +43.2% 1375747 interrupts.CPU44.LOC:Local_timer_interrupts
588.25 ± 7% -45.4% 321.00 ± 11% interrupts.CPU44.NMI:Non-maskable_interrupts
588.25 ± 7% -45.4% 321.00 ± 11% interrupts.CPU44.PMI:Performance_monitoring_interrupts
3015 ± 3% +57.3% 4744 ± 2% interrupts.CPU44.RES:Rescheduling_interrupts
960612 +43.2% 1375625 interrupts.CPU45.LOC:Local_timer_interrupts
3155 ± 7% +49.0% 4701 ± 3% interrupts.CPU45.RES:Rescheduling_interrupts
960536 +43.2% 1375689 interrupts.CPU46.LOC:Local_timer_interrupts
2994 ± 8% +51.7% 4542 interrupts.CPU46.RES:Rescheduling_interrupts
960600 +43.2% 1375692 interrupts.CPU47.LOC:Local_timer_interrupts
3063 ± 8% +53.6% 4704 ± 5% interrupts.CPU47.RES:Rescheduling_interrupts
960789 +43.2% 1375688 interrupts.CPU48.LOC:Local_timer_interrupts
2935 ± 2% +56.0% 4578 ± 5% interrupts.CPU48.RES:Rescheduling_interrupts
960531 +43.2% 1375777 interrupts.CPU49.LOC:Local_timer_interrupts
3038 ± 6% +50.8% 4581 ± 3% interrupts.CPU49.RES:Rescheduling_interrupts
960587 +43.2% 1375671 interrupts.CPU5.LOC:Local_timer_interrupts
3227 +63.2% 5268 ± 27% interrupts.CPU5.RES:Rescheduling_interrupts
960627 +43.2% 1375680 interrupts.CPU50.LOC:Local_timer_interrupts
3031 ± 6% +56.4% 4742 ± 6% interrupts.CPU50.RES:Rescheduling_interrupts
1681 ±172% -100.0% 0.50 ±173% interrupts.CPU51.87:PCI-MSI.31981620-edge.i40e-eth0-TxRx-51
960555 +43.2% 1375716 interrupts.CPU51.LOC:Local_timer_interrupts
3062 ± 7% +50.7% 4616 ± 6% interrupts.CPU51.RES:Rescheduling_interrupts
960450 +43.2% 1375720 interrupts.CPU52.LOC:Local_timer_interrupts
551.25 ± 8% -30.5% 383.00 ± 13% interrupts.CPU52.NMI:Non-maskable_interrupts
551.25 ± 8% -30.5% 383.00 ± 13% interrupts.CPU52.PMI:Performance_monitoring_interrupts
3074 ± 5% +51.0% 4641 ± 2% interrupts.CPU52.RES:Rescheduling_interrupts
960557 +43.2% 1375674 interrupts.CPU53.LOC:Local_timer_interrupts
546.25 ± 8% -32.3% 370.00 ± 29% interrupts.CPU53.NMI:Non-maskable_interrupts
546.25 ± 8% -32.3% 370.00 ± 29% interrupts.CPU53.PMI:Performance_monitoring_interrupts
3061 +53.4% 4694 ± 4% interrupts.CPU53.RES:Rescheduling_interrupts
959955 +43.3% 1375697 interrupts.CPU54.LOC:Local_timer_interrupts
585.75 ± 6% -37.3% 367.25 ± 6% interrupts.CPU54.NMI:Non-maskable_interrupts
585.75 ± 6% -37.3% 367.25 ± 6% interrupts.CPU54.PMI:Performance_monitoring_interrupts
2918 ± 4% +61.9% 4726 ± 2% interrupts.CPU54.RES:Rescheduling_interrupts
960295 +43.3% 1375697 interrupts.CPU55.LOC:Local_timer_interrupts
535.50 ± 18% -38.7% 328.25 ± 15% interrupts.CPU55.NMI:Non-maskable_interrupts
535.50 ± 18% -38.7% 328.25 ± 15% interrupts.CPU55.PMI:Performance_monitoring_interrupts
2970 ± 3% +61.2% 4788 ± 3% interrupts.CPU55.RES:Rescheduling_interrupts
960252 +43.3% 1375697 interrupts.CPU56.LOC:Local_timer_interrupts
538.75 ± 7% -44.9% 296.75 ± 18% interrupts.CPU56.NMI:Non-maskable_interrupts
538.75 ± 7% -44.9% 296.75 ± 18% interrupts.CPU56.PMI:Performance_monitoring_interrupts
3066 ± 6% +54.6% 4741 ± 2% interrupts.CPU56.RES:Rescheduling_interrupts
960316 +43.3% 1375727 interrupts.CPU57.LOC:Local_timer_interrupts
573.00 ± 15% -43.8% 322.00 ± 19% interrupts.CPU57.NMI:Non-maskable_interrupts
573.00 ± 15% -43.8% 322.00 ± 19% interrupts.CPU57.PMI:Performance_monitoring_interrupts
2914 ± 4% +57.3% 4582 ± 7% interrupts.CPU57.RES:Rescheduling_interrupts
960271 +43.3% 1375634 interrupts.CPU58.LOC:Local_timer_interrupts
525.00 ± 10% -33.6% 348.50 ± 9% interrupts.CPU58.NMI:Non-maskable_interrupts
525.00 ± 10% -33.6% 348.50 ± 9% interrupts.CPU58.PMI:Performance_monitoring_interrupts
2919 ± 3% +51.3% 4415 ± 6% interrupts.CPU58.RES:Rescheduling_interrupts
960331 +43.2% 1375652 interrupts.CPU59.LOC:Local_timer_interrupts
488.00 ± 13% -26.1% 360.50 ± 3% interrupts.CPU59.NMI:Non-maskable_interrupts
488.00 ± 13% -26.1% 360.50 ± 3% interrupts.CPU59.PMI:Performance_monitoring_interrupts
2942 ± 5% +56.6% 4607 ± 6% interrupts.CPU59.RES:Rescheduling_interrupts
960564 +43.2% 1375751 interrupts.CPU6.LOC:Local_timer_interrupts
490.50 ± 21% -38.6% 301.00 ± 5% interrupts.CPU6.NMI:Non-maskable_interrupts
490.50 ± 21% -38.6% 301.00 ± 5% interrupts.CPU6.PMI:Performance_monitoring_interrupts
3214 ± 3% +44.9% 4658 ± 3% interrupts.CPU6.RES:Rescheduling_interrupts
960338 +43.2% 1375563 interrupts.CPU60.LOC:Local_timer_interrupts
550.50 ± 10% -33.3% 367.25 ± 11% interrupts.CPU60.NMI:Non-maskable_interrupts
550.50 ± 10% -33.3% 367.25 ± 11% interrupts.CPU60.PMI:Performance_monitoring_interrupts
2956 ± 4% +55.1% 4584 ± 6% interrupts.CPU60.RES:Rescheduling_interrupts
960364 +43.2% 1375687 interrupts.CPU61.LOC:Local_timer_interrupts
523.75 ± 14% -46.9% 278.00 ± 26% interrupts.CPU61.NMI:Non-maskable_interrupts
523.75 ± 14% -46.9% 278.00 ± 26% interrupts.CPU61.PMI:Performance_monitoring_interrupts
3118 ± 4% +44.4% 4502 ± 5% interrupts.CPU61.RES:Rescheduling_interrupts
960292 +43.3% 1375635 interrupts.CPU62.LOC:Local_timer_interrupts
3097 ± 4% +42.9% 4425 ± 4% interrupts.CPU62.RES:Rescheduling_interrupts
960341 +43.2% 1375680 interrupts.CPU63.LOC:Local_timer_interrupts
510.50 ± 14% -35.1% 331.25 ± 2% interrupts.CPU63.NMI:Non-maskable_interrupts
510.50 ± 14% -35.1% 331.25 ± 2% interrupts.CPU63.PMI:Performance_monitoring_interrupts
3174 ± 3% +42.0% 4508 ± 9% interrupts.CPU63.RES:Rescheduling_interrupts
960310 +43.3% 1375721 interrupts.CPU64.LOC:Local_timer_interrupts
529.50 ± 14% -32.3% 358.50 ± 8% interrupts.CPU64.NMI:Non-maskable_interrupts
529.50 ± 14% -32.3% 358.50 ± 8% interrupts.CPU64.PMI:Performance_monitoring_interrupts
3028 ± 7% +45.7% 4410 ± 3% interrupts.CPU64.RES:Rescheduling_interrupts
960268 +43.3% 1375675 interrupts.CPU65.LOC:Local_timer_interrupts
534.00 ± 12% -30.1% 373.25 ± 7% interrupts.CPU65.NMI:Non-maskable_interrupts
534.00 ± 12% -30.1% 373.25 ± 7% interrupts.CPU65.PMI:Performance_monitoring_interrupts
3059 ± 6% +49.7% 4578 ± 5% interrupts.CPU65.RES:Rescheduling_interrupts
960252 +43.3% 1375680 interrupts.CPU66.LOC:Local_timer_interrupts
547.25 ± 9% -33.9% 362.00 ± 8% interrupts.CPU66.NMI:Non-maskable_interrupts
547.25 ± 9% -33.9% 362.00 ± 8% interrupts.CPU66.PMI:Performance_monitoring_interrupts
3050 ± 6% +50.5% 4589 ± 5% interrupts.CPU66.RES:Rescheduling_interrupts
960281 +43.3% 1375628 interrupts.CPU67.LOC:Local_timer_interrupts
491.00 ± 6% -21.9% 383.50 ± 7% interrupts.CPU67.NMI:Non-maskable_interrupts
491.00 ± 6% -21.9% 383.50 ± 7% interrupts.CPU67.PMI:Performance_monitoring_interrupts
3090 ± 4% +43.6% 4439 interrupts.CPU67.RES:Rescheduling_interrupts
960327 +43.2% 1375585 interrupts.CPU68.LOC:Local_timer_interrupts
556.75 ± 13% -30.0% 390.00 ± 6% interrupts.CPU68.NMI:Non-maskable_interrupts
556.75 ± 13% -30.0% 390.00 ± 6% interrupts.CPU68.PMI:Performance_monitoring_interrupts
3148 +43.8% 4526 interrupts.CPU68.RES:Rescheduling_interrupts
960260 +43.3% 1375666 interrupts.CPU69.LOC:Local_timer_interrupts
550.00 ± 12% -31.4% 377.25 ± 7% interrupts.CPU69.NMI:Non-maskable_interrupts
550.00 ± 12% -31.4% 377.25 ± 7% interrupts.CPU69.PMI:Performance_monitoring_interrupts
3114 ± 2% +46.6% 4565 ± 2% interrupts.CPU69.RES:Rescheduling_interrupts
410.75 ± 3% +44.1% 591.75 ± 3% interrupts.CPU7.35:PCI-MSI.31981568-edge.i40e-0000:3d:00.0:misc
960569 +43.2% 1375605 interrupts.CPU7.LOC:Local_timer_interrupts
546.75 ± 9% -43.7% 307.75 ± 14% interrupts.CPU7.NMI:Non-maskable_interrupts
546.75 ± 9% -43.7% 307.75 ± 14% interrupts.CPU7.PMI:Performance_monitoring_interrupts
2889 ± 4% +53.5% 4435 ± 2% interrupts.CPU7.RES:Rescheduling_interrupts
960484 +43.2% 1375854 interrupts.CPU70.LOC:Local_timer_interrupts
3239 ± 5% +42.2% 4605 interrupts.CPU70.RES:Rescheduling_interrupts
960766 +43.3% 1376330 interrupts.CPU71.LOC:Local_timer_interrupts
568.25 ± 8% -32.4% 384.00 ± 2% interrupts.CPU71.NMI:Non-maskable_interrupts
568.25 ± 8% -32.4% 384.00 ± 2% interrupts.CPU71.PMI:Performance_monitoring_interrupts
3017 ± 6% +56.4% 4719 interrupts.CPU71.RES:Rescheduling_interrupts
960486 +43.2% 1375677 interrupts.CPU8.LOC:Local_timer_interrupts
578.25 ± 8% -44.5% 320.75 ± 17% interrupts.CPU8.NMI:Non-maskable_interrupts
578.25 ± 8% -44.5% 320.75 ± 17% interrupts.CPU8.PMI:Performance_monitoring_interrupts
3050 ± 7% +56.8% 4783 ± 11% interrupts.CPU8.RES:Rescheduling_interrupts
960636 +43.2% 1375675 interrupts.CPU9.LOC:Local_timer_interrupts
561.75 ± 7% -41.3% 329.75 ± 14% interrupts.CPU9.NMI:Non-maskable_interrupts
561.75 ± 7% -41.3% 329.75 ± 14% interrupts.CPU9.PMI:Performance_monitoring_interrupts
3299 ± 19% +40.4% 4631 ± 5% interrupts.CPU9.RES:Rescheduling_interrupts
69152219 +43.2% 99050063 interrupts.LOC:Local_timer_interrupts
72.00 +100.0% 144.00 interrupts.MCP:Machine_check_polls
37793 ± 4% -33.4% 25156 interrupts.NMI:Non-maskable_interrupts
37793 ± 4% -33.4% 25156 interrupts.PMI:Performance_monitoring_interrupts
222056 +50.1% 333203 interrupts.RES:Rescheduling_interrupts
20.17 -1.6 18.57 perf-profile.calltrace.cycles-pp.unlink
19.91 -1.6 18.34 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.unlink
19.91 -1.6 18.34 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.unlink
19.65 -1.5 18.13 perf-profile.calltrace.cycles-pp.do_unlinkat.do_syscall_64.entry_SYSCALL_64_after_hwframe.unlink
12.92 -1.1 11.87 perf-profile.calltrace.cycles-pp.xfs_create.xfs_generic_create.path_openat.do_filp_open.do_sys_openat2
13.06 -1.0 12.02 perf-profile.calltrace.cycles-pp.xfs_generic_create.path_openat.do_filp_open.do_sys_openat2.do_sys_open
17.17 ± 2% -1.0 16.18 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.creat
17.18 ± 2% -1.0 16.18 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.creat
17.38 ± 2% -1.0 16.39 perf-profile.calltrace.cycles-pp.creat
17.03 ± 2% -1.0 16.05 perf-profile.calltrace.cycles-pp.do_sys_open.do_syscall_64.entry_SYSCALL_64_after_hwframe.creat
17.02 ± 2% -1.0 16.05 perf-profile.calltrace.cycles-pp.do_sys_openat2.do_sys_open.do_syscall_64.entry_SYSCALL_64_after_hwframe.creat
16.78 ± 2% -0.9 15.85 perf-profile.calltrace.cycles-pp.do_filp_open.do_sys_openat2.do_sys_open.do_syscall_64.entry_SYSCALL_64_after_hwframe
16.76 ± 2% -0.9 15.84 perf-profile.calltrace.cycles-pp.path_openat.do_filp_open.do_sys_openat2.do_sys_open.do_syscall_64
7.43 ± 3% -0.9 6.57 perf-profile.calltrace.cycles-pp.destroy_inode.do_unlinkat.do_syscall_64.entry_SYSCALL_64_after_hwframe.unlink
7.22 ± 3% -0.8 6.38 perf-profile.calltrace.cycles-pp.xfs_fs_destroy_inode.destroy_inode.do_unlinkat.do_syscall_64.entry_SYSCALL_64_after_hwframe
7.18 ± 3% -0.8 6.34 perf-profile.calltrace.cycles-pp.xfs_inactive.xfs_fs_destroy_inode.destroy_inode.do_unlinkat.do_syscall_64
7.16 ± 3% -0.8 6.31 perf-profile.calltrace.cycles-pp.xfs_inactive_ifree.xfs_inactive.xfs_fs_destroy_inode.destroy_inode.do_unlinkat
9.69 -0.6 9.12 perf-profile.calltrace.cycles-pp.xfs_remove.xfs_vn_unlink.vfs_unlink.do_unlinkat.do_syscall_64
9.71 -0.6 9.15 perf-profile.calltrace.cycles-pp.xfs_vn_unlink.vfs_unlink.do_unlinkat.do_syscall_64.entry_SYSCALL_64_after_hwframe
9.84 -0.5 9.29 perf-profile.calltrace.cycles-pp.vfs_unlink.do_unlinkat.do_syscall_64.entry_SYSCALL_64_after_hwframe.unlink
4.00 -0.5 3.46 ± 4% perf-profile.calltrace.cycles-pp.__xfs_trans_commit.xfs_create.xfs_generic_create.path_openat.do_filp_open
3.93 -0.5 3.41 ± 4% perf-profile.calltrace.cycles-pp.xfs_log_commit_cil.__xfs_trans_commit.xfs_create.xfs_generic_create.path_openat
2.57 ± 3% -0.4 2.13 ± 3% perf-profile.calltrace.cycles-pp.__xfs_trans_commit.xfs_inactive_ifree.xfs_inactive.xfs_fs_destroy_inode.destroy_inode
2.51 ± 3% -0.4 2.08 ± 3% perf-profile.calltrace.cycles-pp.xfs_log_commit_cil.__xfs_trans_commit.xfs_inactive_ifree.xfs_inactive.xfs_fs_destroy_inode
4.33 ± 3% -0.4 3.97 perf-profile.calltrace.cycles-pp.xfs_ifree.xfs_inactive_ifree.xfs_inactive.xfs_fs_destroy_inode.destroy_inode
2.52 ± 2% -0.3 2.19 ± 4% perf-profile.calltrace.cycles-pp.xfs_log_commit_cil.__xfs_trans_commit.xfs_remove.xfs_vn_unlink.vfs_unlink
2.60 ± 2% -0.3 2.27 ± 4% perf-profile.calltrace.cycles-pp.__xfs_trans_commit.xfs_remove.xfs_vn_unlink.vfs_unlink.do_unlinkat
1.55 -0.3 1.24 ± 9% perf-profile.calltrace.cycles-pp.xlog_cil_insert_items.xfs_log_commit_cil.__xfs_trans_commit.xfs_create.xfs_generic_create
5.00 ± 2% -0.3 4.72 perf-profile.calltrace.cycles-pp.xfs_ialloc.xfs_dir_ialloc.xfs_create.xfs_generic_create.path_openat
5.00 ± 2% -0.3 4.73 perf-profile.calltrace.cycles-pp.xfs_dir_ialloc.xfs_create.xfs_generic_create.path_openat.do_filp_open
3.09 ± 4% -0.2 2.89 ± 2% perf-profile.calltrace.cycles-pp.xfs_difree.xfs_ifree.xfs_inactive_ifree.xfs_inactive.xfs_fs_destroy_inode
2.31 -0.2 2.10 ± 5% perf-profile.calltrace.cycles-pp.up.xfs_buf_unlock.xfs_buf_item_release.xfs_log_commit_cil.__xfs_trans_commit
4.41 ± 2% -0.2 4.22 perf-profile.calltrace.cycles-pp.xfs_dialloc.xfs_ialloc.xfs_dir_ialloc.xfs_create.xfs_generic_create
0.84 ± 5% -0.2 0.68 ± 8% perf-profile.calltrace.cycles-pp.ktime_get.clockevents_program_event.hrtimer_interrupt.smp_apic_timer_interrupt.__irqentry_text_start
0.94 ± 2% -0.1 0.79 ± 6% perf-profile.calltrace.cycles-pp.xlog_cil_insert_items.xfs_log_commit_cil.__xfs_trans_commit.xfs_remove.xfs_vn_unlink
0.84 ± 2% -0.1 0.70 ± 7% perf-profile.calltrace.cycles-pp.xfs_buf_item_format.xlog_cil_insert_items.xfs_log_commit_cil.__xfs_trans_commit.xfs_create
1.86 ± 3% -0.1 1.73 ± 2% perf-profile.calltrace.cycles-pp.xfs_check_agi_freecount.xfs_difree_inobt.xfs_difree.xfs_ifree.xfs_inactive_ifree
0.89 ± 4% -0.1 0.76 ± 9% perf-profile.calltrace.cycles-pp.xfs_buf_item_release.xfs_log_commit_cil.__xfs_trans_commit.xfs_inactive_ifree.xfs_inactive
0.84 ± 4% -0.1 0.71 ± 7% perf-profile.calltrace.cycles-pp.xfs_buf_unlock.xfs_buf_item_release.xfs_log_commit_cil.__xfs_trans_commit.xfs_inactive_ifree
1.16 ± 4% -0.1 1.04 ± 3% perf-profile.calltrace.cycles-pp.xfs_iunlink_remove.xfs_ifree.xfs_inactive_ifree.xfs_inactive.xfs_fs_destroy_inode
0.87 ± 4% -0.1 0.75 ± 8% perf-profile.calltrace.cycles-pp.xfs_iunlink.xfs_remove.xfs_vn_unlink.vfs_unlink.do_unlinkat
1.89 ± 2% -0.1 1.77 ± 2% perf-profile.calltrace.cycles-pp.sched_ttwu_pending.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
0.70 ± 4% -0.1 0.59 ± 7% perf-profile.calltrace.cycles-pp.xfs_trans_read_buf_map.xfs_read_agi.xfs_iunlink.xfs_remove.xfs_vn_unlink
0.73 ± 4% -0.1 0.63 ± 6% perf-profile.calltrace.cycles-pp.xfs_read_agi.xfs_iunlink.xfs_remove.xfs_vn_unlink.vfs_unlink
0.67 ± 8% -0.1 0.57 ± 5% perf-profile.calltrace.cycles-pp.xfs_btree_lookup_get_block.xfs_btree_lookup.xfs_check_agi_freecount.xfs_dialloc_ag.xfs_dialloc
0.88 ± 6% -0.1 0.79 ± 6% perf-profile.calltrace.cycles-pp.xfs_btree_lookup.xfs_check_agi_freecount.xfs_dialloc_ag.xfs_dialloc.xfs_ialloc
1.02 ± 4% -0.1 0.94 ± 5% perf-profile.calltrace.cycles-pp.xfs_dir2_node_removename.xfs_dir_removename.xfs_remove.xfs_vn_unlink.vfs_unlink
0.77 ± 3% -0.1 0.69 ± 5% perf-profile.calltrace.cycles-pp.xfs_buf_unlock.xfs_buf_item_release.xfs_log_commit_cil.__xfs_trans_commit.xfs_remove
1.16 ± 6% +0.2 1.37 ± 5% perf-profile.calltrace.cycles-pp.__softirqentry_text_start.irq_exit.smp_apic_timer_interrupt.__irqentry_text_start.cpuidle_enter_state
1.66 ± 7% +0.3 1.99 ± 3% perf-profile.calltrace.cycles-pp.irq_exit.smp_apic_timer_interrupt.__irqentry_text_start.cpuidle_enter_state.cpuidle_enter
1.74 ± 4% +0.3 2.08 ± 3% perf-profile.calltrace.cycles-pp.__hrtimer_run_queues.hrtimer_interrupt.smp_apic_timer_interrupt.__irqentry_text_start.cpuidle_enter_state
3.27 ± 2% +0.5 3.80 ± 4% perf-profile.calltrace.cycles-pp.hrtimer_interrupt.smp_apic_timer_interrupt.__irqentry_text_start.cpuidle_enter_state.cpuidle_enter
0.00 +0.6 0.61 ± 5% perf-profile.calltrace.cycles-pp.tick_nohz_next_event.tick_nohz_get_sleep_length.menu_select.do_idle.cpu_startup_entry
0.13 ±173% +0.6 0.74 ± 4% perf-profile.calltrace.cycles-pp.tick_nohz_get_sleep_length.menu_select.do_idle.cpu_startup_entry.start_secondary
1.41 ± 2% +0.6 2.05 ± 2% perf-profile.calltrace.cycles-pp.menu_select.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
5.38 ± 2% +1.1 6.43 ± 3% perf-profile.calltrace.cycles-pp.smp_apic_timer_interrupt.__irqentry_text_start.cpuidle_enter_state.cpuidle_enter.do_idle
53.08 +1.3 54.35 perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.do_idle.cpu_startup_entry.start_secondary
53.58 +1.5 55.05 perf-profile.calltrace.cycles-pp.cpuidle_enter.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
5.86 +1.9 7.78 ± 3% perf-profile.calltrace.cycles-pp.__irqentry_text_start.cpuidle_enter_state.cpuidle_enter.do_idle.cpu_startup_entry
58.56 +2.1 60.70 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
58.58 +2.2 60.73 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.secondary_startup_64
58.58 +2.2 60.73 perf-profile.calltrace.cycles-pp.start_secondary.secondary_startup_64
59.33 +2.2 61.52 perf-profile.calltrace.cycles-pp.secondary_startup_64
38.81 -2.4 36.36 perf-profile.children.cycles-pp.do_syscall_64
38.83 -2.4 36.39 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
20.21 -1.6 18.61 perf-profile.children.cycles-pp.unlink
19.65 -1.5 18.14 perf-profile.children.cycles-pp.do_unlinkat
9.19 -1.3 7.89 ± 2% perf-profile.children.cycles-pp.__xfs_trans_commit
8.99 -1.3 7.70 ± 2% perf-profile.children.cycles-pp.xfs_log_commit_cil
12.92 -1.1 11.87 perf-profile.children.cycles-pp.xfs_create
13.06 -1.0 12.02 perf-profile.children.cycles-pp.xfs_generic_create
17.40 ± 2% -1.0 16.41 perf-profile.children.cycles-pp.creat
17.07 ± 2% -1.0 16.09 perf-profile.children.cycles-pp.do_sys_openat2
17.08 ± 2% -1.0 16.11 perf-profile.children.cycles-pp.do_sys_open
16.82 ± 2% -0.9 15.89 perf-profile.children.cycles-pp.do_filp_open
16.81 ± 2% -0.9 15.88 perf-profile.children.cycles-pp.path_openat
7.43 ± 3% -0.9 6.57 perf-profile.children.cycles-pp.destroy_inode
7.22 ± 3% -0.8 6.38 perf-profile.children.cycles-pp.xfs_fs_destroy_inode
7.16 ± 3% -0.8 6.31 perf-profile.children.cycles-pp.xfs_inactive_ifree
7.18 ± 3% -0.8 6.34 perf-profile.children.cycles-pp.xfs_inactive
9.70 -0.6 9.13 perf-profile.children.cycles-pp.xfs_remove
3.01 -0.6 2.45 ± 5% perf-profile.children.cycles-pp.xlog_cil_insert_items
9.71 -0.6 9.15 perf-profile.children.cycles-pp.xfs_vn_unlink
9.84 -0.5 9.29 perf-profile.children.cycles-pp.vfs_unlink
4.00 -0.5 3.50 perf-profile.children.cycles-pp.xfs_trans_read_buf_map
4.33 ± 3% -0.4 3.97 perf-profile.children.cycles-pp.xfs_ifree
2.92 ± 2% -0.3 2.58 ± 3% perf-profile.children.cycles-pp.xfs_buf_read_map
2.76 ± 2% -0.3 2.44 ± 3% perf-profile.children.cycles-pp.xfs_buf_get_map
5.00 ± 2% -0.3 4.72 perf-profile.children.cycles-pp.xfs_ialloc
5.00 ± 2% -0.3 4.73 perf-profile.children.cycles-pp.xfs_dir_ialloc
2.90 -0.2 2.65 ± 3% perf-profile.children.cycles-pp.try_to_wake_up
2.48 ± 3% -0.2 2.23 ± 3% perf-profile.children.cycles-pp.xfs_buf_find
2.59 -0.2 2.37 ± 6% perf-profile.children.cycles-pp.xfs_buf_item_release
2.44 -0.2 2.22 ± 5% perf-profile.children.cycles-pp.xfs_buf_unlock
3.10 ± 4% -0.2 2.89 ± 2% perf-profile.children.cycles-pp.xfs_difree
1.16 ± 4% -0.2 0.95 ± 4% perf-profile.children.cycles-pp.xfs_trans_unreserve_and_mod_sb
2.32 -0.2 2.12 ± 5% perf-profile.children.cycles-pp.up
2.15 ± 2% -0.2 1.95 ± 4% perf-profile.children.cycles-pp.xfs_read_agi
1.28 ± 3% -0.2 1.09 ± 3% perf-profile.children.cycles-pp.xfs_btree_lookup_get_block
1.52 ± 3% -0.2 1.32 ± 9% perf-profile.children.cycles-pp.xfs_buf_item_format_segment
1.58 ± 3% -0.2 1.39 ± 8% perf-profile.children.cycles-pp.xfs_buf_item_format
4.41 ± 2% -0.2 4.22 perf-profile.children.cycles-pp.xfs_dialloc
1.04 ± 5% -0.2 0.85 ± 5% perf-profile.children.cycles-pp.__percpu_counter_compare
1.02 ± 5% -0.2 0.83 ± 6% perf-profile.children.cycles-pp.__percpu_counter_sum
0.58 ± 4% -0.2 0.40 ± 11% perf-profile.children.cycles-pp.xfs_inode_item_format
1.06 ± 6% -0.2 0.88 ± 5% perf-profile.children.cycles-pp.xfs_mod_ifree
1.92 ± 2% -0.2 1.75 perf-profile.children.cycles-pp.xfs_btree_lookup
0.99 ± 3% -0.2 0.84 ± 6% perf-profile.children.cycles-pp.xfs_btree_read_buf_block
0.92 ± 6% -0.1 0.77 ± 3% perf-profile.children.cycles-pp.rwsem_wake
1.02 ± 5% -0.1 0.88 ± 5% perf-profile.children.cycles-pp.xfs_da_read_buf
1.17 ± 4% -0.1 1.04 ± 3% perf-profile.children.cycles-pp.xfs_iunlink_remove
0.87 ± 4% -0.1 0.75 ± 8% perf-profile.children.cycles-pp.xfs_iunlink
0.85 ± 2% -0.1 0.73 ± 6% perf-profile.children.cycles-pp._xfs_trans_bjoin
0.25 ± 13% -0.1 0.14 ± 18% perf-profile.children.cycles-pp.xfs_inode_item_format_data_fork
0.79 ± 6% -0.1 0.69 ± 2% perf-profile.children.cycles-pp.wake_up_q
2.30 -0.1 2.20 perf-profile.children.cycles-pp.xfs_inobt_get_rec
0.22 ± 14% -0.1 0.12 ± 16% perf-profile.children.cycles-pp.xfs_iextents_copy
1.93 ± 2% -0.1 1.83 ± 2% perf-profile.children.cycles-pp.sched_ttwu_pending
0.56 ± 4% -0.1 0.47 ± 9% perf-profile.children.cycles-pp.xfs_verify_agino
0.30 ± 7% -0.1 0.21 ± 5% perf-profile.children.cycles-pp.xfs_inobt_btrec_to_irec
0.58 ± 4% -0.1 0.50 ± 5% perf-profile.children.cycles-pp.xfs_dir3_block_read
1.02 ± 4% -0.1 0.94 ± 5% perf-profile.children.cycles-pp.xfs_dir2_node_removename
0.21 ± 9% -0.1 0.14 ± 19% perf-profile.children.cycles-pp.calc_global_load_tick
0.47 ± 7% -0.1 0.40 ± 4% perf-profile.children.cycles-pp.select_task_rq_fair
0.34 ± 5% -0.1 0.27 ± 2% perf-profile.children.cycles-pp.__list_del_entry_valid
0.40 ± 2% -0.1 0.34 ± 11% perf-profile.children.cycles-pp.xfs_iget
0.42 ± 3% -0.1 0.36 ± 9% perf-profile.children.cycles-pp.xfs_iunlink_update_inode
0.48 ± 3% -0.1 0.42 ± 9% perf-profile.children.cycles-pp.xfs_trans_reserve
0.25 ± 5% -0.1 0.19 ± 12% perf-profile.children.cycles-pp._xfs_buf_obj_cmp
0.27 ± 12% -0.1 0.21 ± 6% perf-profile.children.cycles-pp.update_cfs_rq_h_load
0.14 ± 15% -0.1 0.08 ± 8% perf-profile.children.cycles-pp.xfs_bmap_validate_extent
0.23 ± 11% -0.0 0.18 ± 6% perf-profile.children.cycles-pp.down_write
0.13 ± 16% -0.0 0.08 ± 19% perf-profile.children.cycles-pp.complete_walk
0.07 ± 5% -0.0 0.03 ±100% perf-profile.children.cycles-pp.xfs_droplink
0.16 ± 10% -0.0 0.11 ± 14% perf-profile.children.cycles-pp.nr_iowait_cpu
0.25 ± 7% -0.0 0.21 ± 7% perf-profile.children.cycles-pp.xfs_log_ticket_ungrant
0.16 ± 12% -0.0 0.12 ± 6% perf-profile.children.cycles-pp.tick_nohz_idle_exit
0.08 ± 8% -0.0 0.05 ± 58% perf-profile.children.cycles-pp.xlog_cil_push_work
0.19 ± 2% -0.0 0.17 ± 6% perf-profile.children.cycles-pp.inode_permission
0.12 ± 7% -0.0 0.10 ± 5% perf-profile.children.cycles-pp.percpu_counter_add_batch
0.12 ± 4% -0.0 0.10 ± 10% perf-profile.children.cycles-pp.__list_add_valid
0.06 ± 7% +0.0 0.08 ± 6% perf-profile.children.cycles-pp.tick_nohz_idle_enter
0.07 ± 15% +0.0 0.10 ± 8% perf-profile.children.cycles-pp.__might_sleep
0.04 ± 58% +0.0 0.07 ± 12% perf-profile.children.cycles-pp.xfs_btree_delrec
0.07 ± 20% +0.0 0.10 ± 15% perf-profile.children.cycles-pp.__hrtimer_next_event_base
0.06 ± 11% +0.0 0.09 ± 13% perf-profile.children.cycles-pp.rcu_nmi_enter
0.23 ± 6% +0.0 0.26 ± 5% perf-profile.children.cycles-pp.process_one_work
0.08 ± 5% +0.0 0.11 ± 18% perf-profile.children.cycles-pp.menu_reflect
0.08 ± 6% +0.0 0.11 ± 4% perf-profile.children.cycles-pp.interrupt_entry
0.04 ± 58% +0.0 0.07 ± 11% perf-profile.children.cycles-pp.xfs_btree_delete
0.05 ± 9% +0.0 0.09 ± 14% perf-profile.children.cycles-pp.enqueue_hrtimer
0.24 ± 5% +0.0 0.27 ± 3% perf-profile.children.cycles-pp.worker_thread
0.04 ± 57% +0.0 0.07 ± 15% perf-profile.children.cycles-pp.timerqueue_add
0.05 ± 62% +0.0 0.09 ± 14% perf-profile.children.cycles-pp.rcu_eqs_enter
0.21 ± 9% +0.0 0.25 ± 5% perf-profile.children.cycles-pp.native_sched_clock
0.30 ± 6% +0.0 0.34 ± 7% perf-profile.children.cycles-pp.ret_from_fork
0.29 ± 7% +0.0 0.34 ± 8% perf-profile.children.cycles-pp.kthread
0.22 ± 6% +0.0 0.26 ± 6% perf-profile.children.cycles-pp.sched_clock
0.08 ± 15% +0.0 0.12 ± 13% perf-profile.children.cycles-pp.vfs_read
0.08 ± 15% +0.0 0.13 ± 10% perf-profile.children.cycles-pp.ksys_read
0.10 ± 18% +0.0 0.14 ± 12% perf-profile.children.cycles-pp.cpuidle_governor_latency_req
0.01 ±173% +0.0 0.06 perf-profile.children.cycles-pp.tsc_verify_tsc_adjust
0.06 ± 6% +0.1 0.11 ± 23% perf-profile.children.cycles-pp.run_timer_softirq
0.05 ± 67% +0.1 0.10 ± 15% perf-profile.children.cycles-pp.timekeeping_max_deferment
0.01 ±173% +0.1 0.07 ± 7% perf-profile.children.cycles-pp.new_sync_read
0.23 ± 8% +0.1 0.29 ± 3% perf-profile.children.cycles-pp.sched_clock_cpu
0.24 ± 14% +0.1 0.29 ± 7% perf-profile.children.cycles-pp.load_balance
0.09 ± 8% +0.1 0.15 ± 14% perf-profile.children.cycles-pp.read
0.08 ± 5% +0.1 0.14 ± 10% perf-profile.children.cycles-pp.__intel_pmu_enable_all
0.01 ±173% +0.1 0.07 ± 12% perf-profile.children.cycles-pp.arch_cpu_idle_enter
0.22 ± 13% +0.1 0.27 ± 3% perf-profile.children.cycles-pp.update_blocked_averages
0.12 ± 21% +0.1 0.18 ± 11% perf-profile.children.cycles-pp.arch_scale_freq_tick
0.07 ± 5% +0.1 0.13 ± 11% perf-profile.children.cycles-pp.__run_timers
0.23 ± 16% +0.1 0.29 ± 5% perf-profile.children.cycles-pp.run_rebalance_domains
0.18 ± 13% +0.1 0.24 ± 7% perf-profile.children.cycles-pp.ktime_get_update_offsets_now
0.04 ± 58% +0.1 0.11 ± 17% perf-profile.children.cycles-pp.vfprintf
0.00 +0.1 0.07 ± 17% perf-profile.children.cycles-pp.fprintf
0.01 ±173% +0.1 0.08 ± 5% perf-profile.children.cycles-pp.hrtimer_next_event_without
0.05 ± 8% +0.1 0.13 ± 3% perf-profile.children.cycles-pp.drm_fb_helper_dirty_work
1.04 ± 5% +0.1 1.12 ± 3% perf-profile.children.cycles-pp.update_process_times
0.04 ± 58% +0.1 0.13 ± 10% perf-profile.children.cycles-pp.__handle_mm_fault
1.07 ± 4% +0.1 1.16 ± 3% perf-profile.children.cycles-pp.tick_sched_handle
0.57 ± 6% +0.1 0.66 ± 4% perf-profile.children.cycles-pp.__orc_find
0.06 ± 13% +0.1 0.15 ± 13% perf-profile.children.cycles-pp.do_user_addr_fault
0.02 ±173% +0.1 0.11 ± 10% perf-profile.children.cycles-pp.delay_tsc
0.04 ± 57% +0.1 0.14 ± 11% perf-profile.children.cycles-pp.handle_mm_fault
0.12 ± 14% +0.1 0.22 ± 17% perf-profile.children.cycles-pp.__next_timer_interrupt
0.07 ± 16% +0.1 0.16 ± 12% perf-profile.children.cycles-pp.page_fault
0.70 ± 8% +0.1 0.80 ± 6% perf-profile.children.cycles-pp.orc_find
0.24 ± 13% +0.1 0.37 ± 12% perf-profile.children.cycles-pp.tick_irq_enter
0.35 ± 8% +0.1 0.48 ± 4% perf-profile.children.cycles-pp.read_tsc
0.29 ± 7% +0.1 0.43 ± 9% perf-profile.children.cycles-pp.perf_mux_hrtimer_handler
0.17 ± 12% +0.1 0.31 ± 15% perf-profile.children.cycles-pp.get_next_timer_interrupt
0.19 ± 15% +0.1 0.33 ± 13% perf-profile.children.cycles-pp.io_serial_in
0.25 ± 4% +0.2 0.41 ± 11% perf-profile.children.cycles-pp.native_irq_return_iret
0.28 ± 8% +0.2 0.45 ± 6% perf-profile.children.cycles-pp.lapic_next_deadline
0.32 ± 11% +0.2 0.48 ± 7% perf-profile.children.cycles-pp.irq_enter
0.38 ± 5% +0.2 0.56 ± 5% perf-profile.children.cycles-pp.native_write_msr
0.23 ± 5% +0.2 0.42 ± 7% perf-profile.children.cycles-pp.serial8250_console_putchar
0.23 ± 4% +0.2 0.42 ± 6% perf-profile.children.cycles-pp.uart_console_write
0.23 ± 4% +0.2 0.43 ± 7% perf-profile.children.cycles-pp.wait_for_xmitr
0.23 ± 4% +0.2 0.44 ± 6% perf-profile.children.cycles-pp.serial8250_console_write
0.25 ± 4% +0.2 0.46 ± 5% perf-profile.children.cycles-pp.irq_work_run
0.25 ± 4% +0.2 0.46 ± 5% perf-profile.children.cycles-pp.printk
0.25 ± 4% +0.2 0.46 ± 5% perf-profile.children.cycles-pp.vprintk_emit
0.25 ± 4% +0.2 0.46 ± 5% perf-profile.children.cycles-pp.console_unlock
0.26 ± 3% +0.2 0.47 ± 6% perf-profile.children.cycles-pp.irq_work_run_list
0.25 ± 4% +0.2 0.46 ± 6% perf-profile.children.cycles-pp.irq_work_interrupt
0.25 ± 4% +0.2 0.46 ± 6% perf-profile.children.cycles-pp.smp_irq_work_interrupt
1.22 ± 6% +0.2 1.45 ± 6% perf-profile.children.cycles-pp.__softirqentry_text_start
0.38 ± 8% +0.2 0.62 ± 5% perf-profile.children.cycles-pp.tick_nohz_next_event
0.00 +0.2 0.24 ± 14% perf-profile.children.cycles-pp.xfs_dir2_block_lookup
2.30 ± 2% +0.3 2.58 ± 3% perf-profile.children.cycles-pp.xfs_dir2_block_lookup_int
0.47 ± 8% +0.3 0.76 ± 4% perf-profile.children.cycles-pp.tick_nohz_get_sleep_length
0.00 +0.3 0.29 ± 9% perf-profile.children.cycles-pp.xfs_dir_lookup
0.00 +0.3 0.30 ± 9% perf-profile.children.cycles-pp.xfs_lookup
0.00 +0.3 0.30 ± 9% perf-profile.children.cycles-pp.xfs_vn_lookup
1.74 ± 7% +0.3 2.09 ± 4% perf-profile.children.cycles-pp.irq_exit
1.88 ± 4% +0.4 2.26 ± 2% perf-profile.children.cycles-pp.__hrtimer_run_queues
3.47 +0.6 4.08 ± 3% perf-profile.children.cycles-pp.hrtimer_interrupt
1.44 ± 2% +0.7 2.10 ± 3% perf-profile.children.cycles-pp.menu_select
5.66 +1.2 6.83 ± 3% perf-profile.children.cycles-pp.smp_apic_timer_interrupt
6.18 +1.4 7.55 ± 3% perf-profile.children.cycles-pp.__irqentry_text_start
54.26 +1.5 55.73 perf-profile.children.cycles-pp.cpuidle_enter_state
54.27 +1.5 55.75 perf-profile.children.cycles-pp.cpuidle_enter
58.58 +2.2 60.73 perf-profile.children.cycles-pp.start_secondary
59.33 +2.2 61.52 perf-profile.children.cycles-pp.secondary_startup_64
59.33 +2.2 61.52 perf-profile.children.cycles-pp.cpu_startup_entry
59.34 +2.2 61.55 perf-profile.children.cycles-pp.do_idle
0.58 ± 3% -0.1 0.45 ± 9% perf-profile.self.cycles-pp.xfs_log_commit_cil
0.95 ± 2% -0.1 0.83 ± 7% perf-profile.self.cycles-pp._raw_spin_lock
0.63 ± 2% -0.1 0.53 ± 3% perf-profile.self.cycles-pp.unwind_next_frame
1.06 ± 4% -0.1 0.96 ± 3% perf-profile.self.cycles-pp.memcpy_erms
0.46 ± 3% -0.1 0.36 ± 7% perf-profile.self.cycles-pp.xlog_cil_insert_items
0.58 ± 7% -0.1 0.48 ± 10% perf-profile.self.cycles-pp.__percpu_counter_sum
0.96 ± 5% -0.1 0.88 ± 7% perf-profile.self.cycles-pp.xfs_agino_range
0.21 ± 9% -0.1 0.13 ± 19% perf-profile.self.cycles-pp.calc_global_load_tick
0.34 ± 4% -0.1 0.27 ± 4% perf-profile.self.cycles-pp.__list_del_entry_valid
0.30 ± 9% -0.1 0.23 ± 10% perf-profile.self.cycles-pp.xfs_inode_item_format
0.28 ± 6% -0.1 0.21 ± 5% perf-profile.self.cycles-pp.xfs_inobt_btrec_to_irec
0.38 ± 6% -0.1 0.31 ± 7% perf-profile.self.cycles-pp.xfs_buf_item_init
0.27 ± 9% -0.1 0.20 ± 8% perf-profile.self.cycles-pp.xfs_buf_get_map
0.37 ± 2% -0.1 0.31 ± 7% perf-profile.self.cycles-pp._xfs_trans_bjoin
0.27 ± 12% -0.1 0.21 ± 6% perf-profile.self.cycles-pp.update_cfs_rq_h_load
0.24 ± 6% -0.1 0.19 ± 12% perf-profile.self.cycles-pp._xfs_buf_obj_cmp
0.24 ± 4% -0.0 0.20 ± 6% perf-profile.self.cycles-pp.xfs_log_ticket_ungrant
0.45 -0.0 0.41 ± 5% perf-profile.self.cycles-pp.do_syscall_64
0.15 ± 8% -0.0 0.11 ± 14% perf-profile.self.cycles-pp.nr_iowait_cpu
0.12 ± 12% -0.0 0.09 ± 13% perf-profile.self.cycles-pp.note_gp_changes
0.16 ± 4% -0.0 0.12 ± 8% perf-profile.self.cycles-pp.inode_permission
0.11 ± 10% -0.0 0.07 ± 26% perf-profile.self.cycles-pp.xfs_iget
0.14 ± 3% -0.0 0.10 ± 18% perf-profile.self.cycles-pp.rwsem_down_write_slowpath
0.18 ± 8% -0.0 0.15 ± 10% perf-profile.self.cycles-pp.down_write
0.17 ± 8% -0.0 0.14 ± 10% perf-profile.self.cycles-pp.select_task_rq_fair
0.15 ± 6% -0.0 0.12 ± 12% perf-profile.self.cycles-pp.xfs_verify_agbno
0.18 ± 6% -0.0 0.15 ± 8% perf-profile.self.cycles-pp.__radix_tree_lookup
0.14 ± 15% -0.0 0.11 ± 3% perf-profile.self.cycles-pp.___might_sleep
0.08 ± 15% -0.0 0.06 ± 7% perf-profile.self.cycles-pp.poll_idle
0.10 ± 10% -0.0 0.08 ± 13% perf-profile.self.cycles-pp.xfs_create
0.07 -0.0 0.05 ± 9% perf-profile.self.cycles-pp.xlog_space_left
0.23 ± 3% +0.0 0.26 ± 6% perf-profile.self.cycles-pp.xfs_btree_get_rec
0.08 ± 6% +0.0 0.11 ± 4% perf-profile.self.cycles-pp.interrupt_entry
0.12 ± 7% +0.0 0.15 ± 10% perf-profile.self.cycles-pp.ktime_get_update_offsets_now
0.10 ± 24% +0.0 0.13 ± 9% perf-profile.self.cycles-pp.native_read_msr
0.05 ± 58% +0.0 0.09 ± 7% perf-profile.self.cycles-pp.lapic_next_deadline
0.11 ± 10% +0.0 0.15 ± 7% perf-profile.self.cycles-pp.tick_nohz_next_event
0.32 ± 8% +0.1 0.37 ± 8% perf-profile.self.cycles-pp._raw_spin_unlock_irqrestore
0.09 ± 11% +0.1 0.14 ± 15% perf-profile.self.cycles-pp.hrtimer_interrupt
0.00 +0.1 0.06 ± 7% perf-profile.self.cycles-pp.tsc_verify_tsc_adjust
0.04 ±107% +0.1 0.10 ± 15% perf-profile.self.cycles-pp.timekeeping_max_deferment
0.56 ± 6% +0.1 0.65 ± 3% perf-profile.self.cycles-pp.__orc_find
0.02 ±173% +0.1 0.11 ± 10% perf-profile.self.cycles-pp.delay_tsc
0.32 ± 10% +0.1 0.45 ± 5% perf-profile.self.cycles-pp.read_tsc
0.19 ± 15% +0.1 0.33 ± 13% perf-profile.self.cycles-pp.io_serial_in
0.25 ± 4% +0.2 0.41 ± 11% perf-profile.self.cycles-pp.native_irq_return_iret
0.38 ± 5% +0.2 0.56 ± 5% perf-profile.self.cycles-pp.native_write_msr
0.60 ± 2% +0.2 0.80 ± 3% perf-profile.self.cycles-pp.cpuidle_enter_state
0.83 ± 4% +0.3 1.16 ± 4% perf-profile.self.cycles-pp.menu_select
aim7.jobs-per-min
45000 +-------------------------------------------------------------------+
| .+. .+. |
40000 |.+. +.+.+ ++.+ ++.+.+.+.+ .+.+.+.+ .+.+. .++.+.+.++ +.+.++.+.|
35000 |-+ + : : + + + |
| : : |
30000 |-O O O : :O |
25000 |-+ OO O OO O:O:O O O O OO O O O OO |
| : : |
20000 |-+ : : |
15000 |-+ : : |
| : : |
10000 |-+ :: |
5000 |-+ : |
| : |
0 +-------------------------------------------------------------------+
aim7.time.user_time
80 +----------------------------------------------------------------------+
| O O OO O O O O O O OO O O O O O O OO O O |
70 |-+ |
60 |-+ |
| |
50 |.+.+.++.+.+.+.+.+ ++.+.+.+.+.+.+.++.+.+.+.+.+.++.+.+.+. .+.++.+.+.|
| : : +.+ |
40 |-+ : : |
| : : |
30 |-+ : : |
20 |-+ : : |
| : : |
10 |-+ : |
| : |
0 +----------------------------------------------------------------------+
aim7.time.system_time
1800 +--------------------------------------------------------------------+
| O O O O O O |
1600 |-O OO O O O O OO O O O OO O O |
1400 |-+ |
| |
1200 |.+.++.+. .+.+.+.+.++.+.+.+.++.+.+.+.++. .+.+.++.+.|
1000 |-+ +.+.+.++ +.+.++ +.+ |
| : : |
800 |-+ : : |
600 |-+ : : |
| : : |
400 |-+ : : |
200 |-+ :: |
| : |
0 +--------------------------------------------------------------------+
aim7.time.elapsed_time
700 +---------------------------------------------------------------------+
| O O OO O O O O OO O O |
600 |-+ |
| |
500 |.+.+.++. .+ .+.+.+.+. .++.+.+.+.+.++.+. .+.++.+.+.|
| + +.+.+.+.+ +.+.+.+ + + +.+ |
400 |-+ : : |
| : : |
300 |-+ : : |
| : : |
200 |-+ :: |
| :: |
100 |-+ : |
| : |
0 +---------------------------------------------------------------------+
aim7.time.elapsed_time.max
700 +---------------------------------------------------------------------+
| O O OO O O O O OO O O |
600 |-+ |
| |
500 |.+.+.++. .+ .+.+.+.+. .++.+.+.+.+.++.+. .+.++.+.+.|
| + +.+.+.+.+ +.+.+.+ + + +.+ |
400 |-+ : : |
| : : |
300 |-+ : : |
| : : |
200 |-+ :: |
| :: |
100 |-+ : |
| : |
0 +---------------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
***************************************************************************************************
lkp-kblr-hp1: 8 threads Kaby Lake-R with 8G memory
=========================================================================================
compiler/iterations/kconfig/mode/rootfs/tbox_group/testcase:
gcc-9/10/x86_64-rhel-7.6/freeze/debian-x86_64-20180403.cgz/lkp-kblr-hp1/suspend-stress
commit:
21bf3e69e5 ("cpufreq: Register governors at core_initcall")
d83f959b5e ("cpufreq: Specify default governor on command line")
21bf3e69e51e91cd d83f959b5e7a6378a4afbff23de
---------------- ---------------------------
fail:runs %reproduction fail:runs
| | |
:4 100% 4:4 kmsg.cpufreq:cpufreq_online:Failed_to_initialize_policy_for_cpu:#(-#)
%stddev %change %stddev
\ | \
2514 ± 2% +7.9% 2713 suspend-stress.time.involuntary_context_switches
Disclaimer:
Results have been estimated based on internal Intel analysis and are provided
for informational purposes only. Any difference in system hardware or software
design or configuration may affect actual performance.
Thanks,
Rong Chen
Greeting,
FYI, we noticed the following commit (built with gcc-9):
commit: d83f959b5e7a6378a4afbff23de2a2d064d95749 ("[PATCH 2/2] cpufreq: Specify default governor on command line")
url: https://github.com/0day-ci/linux/commits/Quentin-Perret/cpufreq-Specify-the-default-governor-on-command-line/20200616-005920
base: https://git.kernel.org/cgit/linux/kernel/git/rafael/linux-pm.git linux-next
in testcase: kernel-selftests
with following parameters:
group: kselftests-x86
ucode: 0xdc
test-description: The kernel contains a set of "self tests" under the tools/testing/selftests/ directory. These are intended to be small unit tests to exercise individual code paths in the kernel.
test-url: https://www.kernel.org/doc/Documentation/kselftest.txt
on test machine: 8 threads Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz with 16G memory
caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
If you fix the issue, kindly add following tag
Reported-by: kernel test robot <redacted>
[ 8.715369] intel_pstate: Intel P-state driver initializing
[ 8.721146] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 0 (-61)
[ 8.728900] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 1 (-61)
[ 8.736615] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 2 (-61)
[ 8.744400] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 3 (-61)
[ 8.752222] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 4 (-61)
[ 8.760010] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 5 (-61)
[ 8.768077] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 6 (-61)
[ 8.775891] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 7 (-61)
[ 8.783861] hid: raw HID events driver (C) Jiri Kosina
[ 8.789211] usbcore: registered new interface driver usbhid
[ 8.794908] usbhid: USB HID core driver
[ 8.798902] drop_monitor: Initializing network drop monitor service
[ 8.805365] Initializing XFRM netlink socket
[ 8.809817] NET: Registered protocol family 10
[ 8.814586] Segment Routing with IPv6
[ 8.818389] NET: Registered protocol family 17
[ 8.822969] 9pnet: Installing 9P2000 support
[ 8.827367] mpls_gso: MPLS GSO support
[ 8.832204] microcode: sig=0x506e3, pf=0x2, revision=0xdc
[ 8.837953] microcode: Microcode Update Driver: v2.2.
[ 8.837955] IPI shorthand broadcast: enabled
[ 8.847612] ... APIC ID: 00000000 (0)
[ 8.848610] ... APIC VERSION: 01060015
[ 8.848610] 0000000000000000000000000000000000000000000000000000000000000000
[ 8.862620] 0000000000000000000000000000000000000000000000000000000000000000
[ 8.865617] usb 1-4: new low-speed USB device number 2 using xhci_hcd
[ 8.862620] 0000000000000000000000000000000000000000000000000000000000000000
[ 8.870017] number of MP IRQ sources: 15.
[ 8.870017] number of IO-APIC #2 registers: 120.
[ 8.870018] testing the IO APIC.......................
[ 8.897877] IO APIC #2......
[ 8.900891] .... register #00: 02000000
[ 8.904855] ....... : physical APIC id: 02
[ 8.909336] ....... : Delivery Type: 0
[ 8.913472] ....... : LTS : 0
[ 8.917609] .... register #01: 00770020
[ 8.921571] ....... : max redirection entries: 77
[ 8.926768] ....... : PRQ implemented: 0
[ 8.931161] ....... : IO APIC version: 20
[ 8.935641] .... register #02: 00000000
[ 8.939603] ....... : arbitration: 00
[ 8.943738] .... IRQ redirection table:
[ 8.947702] IOAPIC 0:
[ 8.950123] pin00, disabled, edge , high, V(00), IRR(0), S(0), physical, D(00), M(0)
[ 8.958130] pin01, disabled, edge , high, V(00), IRR(0), S(0), physical, D(00), M(0)
[ 8.966138] pin02, enabled , edge , high, V(02), IRR(0), S(0), remapped, I(0001), Z(0)
[ 8.974404] pin03, disabled, edge , high, V(00), IRR(0), S(0), physical, D(00), M(0)
[ 8.982411] pin04, disabled, edge , high, V(00), IRR(0), S(0), physical, D(00), M(0)
[ 8.990419] pin05, disabled, edge , high, V(00), IRR(0), S(0), physical, D(00), M(0)
[ 8.998425] pin06, disabled, edge , high, V(00), IRR(0), S(0), physical, D(00), M(0)
[ 9.006463] pin07, disabled, edge , high, V(00), IRR(0), S(0), physical, D(00), M(0)
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
bin/lkp install job.yaml # job file is attached in this email
bin/lkp run job.yaml
Thanks,
Rong Chen
Hi,
Thanks for the report.
On Monday 22 Jun 2020 at 08:54:57 (+0800), kernel test robot wrote: Greeting,
FYI, we noticed the following commit (built with gcc-9):
commit: d83f959b5e7a6378a4afbff23de2a2d064d95749 ("[PATCH 2/2] cpufreq: Specify default governor on command line")
url: https://github.com/0day-ci/linux/commits/Quentin-Perret/cpufreq-Specify-the-default-governor-on-command-line/20200616-005920
base: https://git.kernel.org/cgit/linux/kernel/git/rafael/linux-pm.git linux-next
in testcase: kernel-selftests
with following parameters:
group: kselftests-x86
ucode: 0xdc
test-description: The kernel contains a set of "self tests" under the tools/testing/selftests/ directory. These are intended to be small unit tests to exercise individual code paths in the kernel.
test-url: https://www.kernel.org/doc/Documentation/kselftest.txt
on test machine: 8 threads Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz with 16G memory
caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
If you fix the issue, kindly add following tag
Reported-by: kernel test robot <redacted>
[ 8.715369] intel_pstate: Intel P-state driver initializing
[ 8.721146] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 0 (-61)
[ 8.728900] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 1 (-61)
[ 8.736615] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 2 (-61)
[ 8.744400] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 3 (-61)
[ 8.752222] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 4 (-61)
[ 8.760010] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 5 (-61)
[ 8.768077] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 6 (-61)
[ 8.775891] cpufreq: cpufreq_online: Failed to initialize policy for cpu: 7 (-61)
That, I think, is because of the issue I reported here:
https://lore.kernel.org/lkml/20200615174141.GA235811@google.com/
The v2 (to be posted shortly) will address this.
Thanks,
Quentin