KR20120064553A - Method for measuring real-time performance in linux-based embedded system and the system - Google Patents

Method for measuring real-time performance in linux-based embedded system and the system Download PDF

Info

Publication number
KR20120064553A
KR20120064553A KR1020100125832A KR20100125832A KR20120064553A KR 20120064553 A KR20120064553 A KR 20120064553A KR 1020100125832 A KR1020100125832 A KR 1020100125832A KR 20100125832 A KR20100125832 A KR 20100125832A KR 20120064553 A KR20120064553 A KR 20120064553A
Authority
KR
South Korea
Prior art keywords
time
interrupt
real
task
value
Prior art date
Application number
KR1020100125832A
Other languages
Korean (ko)
Inventor
정영준
Original Assignee
한국전자통신연구원
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 한국전자통신연구원 filed Critical 한국전자통신연구원
Priority to KR1020100125832A priority Critical patent/KR20120064553A/en
Publication of KR20120064553A publication Critical patent/KR20120064553A/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3409—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment

Landscapes

  • Engineering & Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Quality & Reliability (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Debugging And Monitoring (AREA)

Abstract

PURPOSE: A method for measuring real-time performance in Linux-based embedded system is provided to accurately measure delay time which is performance measuring time in real time by using hardware system counter value. CONSTITUTION: A method for measuring real-time performance in Linux-based embedded system comprises: an RTC(Real-Time clock)(336) generating periodical interrupt; an interrupt handler(310) receiving interrupt from the RTC, recording system counter value as an interrupt generation timing value at interrupt receiving timing, and waking-up task corresponding to the interrupt; a scheduler(320) scheduling, in order to preferentially conduct a task with first priority after the task waked-up; and a performance measuring device in real time calculating preemption delay time.

Description

Method for measuring real-time performance in Linux-based embedded system and the system

The present invention relates to a real-time performance measurement method and a system in the Linux-based embedded system.

Recently, the demand for Linux-based smartphones such as Android is exploding, and the demand for Linux-based embedded systems is rapidly expanding in various fields such as IPTV set-top boxes, MIDs, and industrial devices. Among the services using the Linux-based embedded system, the real-time performance is required to be good in various services such as video playback, call function, and control through the Internet.

Due to the services that require good real-time performance of Linux-based embedded systems, technologies that support real-time performance for platforms for embedded systems are required. Therefore, in order to develop a technology supporting the real-time performance and apply it to an embedded system, a real-time performance measurement technique for a Linux-based embedded system is required.

In this regard, the real-time performance measurement technique for Linux-based embedded systems, which is mainly used, has not been widely used in development because the interval to measure performance is somewhat unclear, the measurement results are inaccurate and not intuitive.

Accordingly, the present invention is to solve the above-mentioned problems and to propose a new method, and an object of the present invention is to clarify a real-time performance measurement interval, and to accurately measure real-time performance in a Linux-based embedded system. It provides a real-time performance measurement method and system.

According to an aspect of the present invention for achieving the above object, a Linux-based embedded system is configured to operate in kernel mode and generates a real time clock (RTC) that generates periodic interrupts, and is configured to operate in kernel mode, the RTC An interrupt handler configured to receive an interrupt from the interrupt, write a system counter value at the time the interrupt is received as an interrupt occurrence value, and wake up a task corresponding to the interrupt; When a task wakes up, it is configured to operate in user mode, with a scheduler that performs scheduling so that the task with the highest priority executes preferentially and generates a task execution signal for executing the task, from the interrupt handler The interrupt foot Receives a starting point value, when receiving the task execution signal from the scheduler, records the system counter value at the received starting point as a task starting starting point value and subtracts an interrupt occurrence starting point value from the task starting starting point value It includes a real-time performance measurement device to calculate the.

According to another aspect of the present invention, a method for measuring real-time performance in a Linux-based embedded system includes generating a periodic interrupt by a Real Time Clock (RTC) configured to operate in kernel mode, and configured to operate in kernel mode. Receiving an interrupt from the RTC by an interrupt handler, recording a system counter value at the time the interrupt is received as an interrupt occurrence time value, and waking up the task corresponding to the interrupt by the interrupt handler; And scheduling by the scheduler configured to operate in kernel mode so that, when the task corresponding to the interrupt wakes up, the task having the highest priority is preferentially executed; A task to run it Generating a signal, receiving, by the real time performance measuring device configured to operate in a user mode, the interrupt occurrence time value from the interrupt handler, and receiving the task execution signal from the scheduler by the real time performance measuring device. Calculating a preemption delay time by subtracting an interrupt occurrence time point value from the task start time point value by the real time performance measuring apparatus by recording a system counter value at the received time point as a task start time point when receiving; It includes.

According to another aspect of the present invention, a method for measuring real-time performance in a real-time performance measuring device is set such that a real time clock (RTC) configured to operate in the kernel mode by the real-time performance measuring device periodically generates an interrupt. Receiving an interrupt occurrence time value from an interrupt handler configured to operate in kernel mode, and receiving a task execution signal for executing a task corresponding to the interrupt from a scheduler configured to operate in kernel mode; Recording a system counter value at the time point of receiving the task execution signal from the scheduler as a task start time point value, and calculating a preemption delay time by subtracting the interrupt occurrence time point value from the task start time point value; do.

According to the present invention, a delay time corresponding to a clear real-time performance measurement interval can be precisely measured using a hardware system counter than a real-time performance measurement of a conventional Linux-based embedded system, and is more user intuitive than the conventional methods. The advantage is that you get the result of time.

In addition, the apparatus and method for measuring real-time performance in the Linux-based embedded system according to the present invention can be utilized as a performance measurement tool for improving the real-time performance of Linux for embedded systems.

1 is a diagram schematically showing a Linux-based embedded system including a real-time performance measurement apparatus according to the present invention.
2 is a diagram showing a real-time performance measurement interval according to the present invention.
3 is a diagram illustrating a process of measuring a preemption delay time according to an embodiment of the present invention.
4 is a block diagram of a real-time performance measurement apparatus according to an embodiment of the present invention.
5 is a flowchart illustrating a real-time performance measurement method according to an embodiment of the present invention.

As the invention allows for various changes and numerous embodiments, particular embodiments will be illustrated in the drawings and described in detail in the written description. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

The terms first, second, A, B, etc. may be used to describe various elements, but the elements should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component. And / or < / RTI > includes any combination of a plurality of related listed items or any of a plurality of related listed items.

When a component is referred to as being "connected" or "connected" to another component, it may be directly connected to or connected to that other component, but it may be understood that other components may be present in between. Should be. On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting of the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, the terms "comprise" or "have" are intended to indicate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, and one or more other features. It is to be understood that the present invention does not exclude the possibility of the presence or the addition of numbers, steps, operations, components, components, or a combination thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art. Terms such as those defined in the commonly used dictionaries should be construed as having meanings consistent with the meanings in the context of the related art and shall not be construed in ideal or excessively formal meanings unless expressly defined in this application. Do not.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram schematically showing a Linux-based embedded system including a real-time performance measurement apparatus according to the present invention.

In the present invention, the meaning of the real-time performance measurement is to measure which event (RTC interrupt in this specification) can drive a task corresponding to the event within a certain delay time.

In general, task preemption delay is a task having a high priority by an interrupt when a interrupt occurs when a task having a higher priority than a task currently being performed due to a hardware interrupt occurs. Means the time until it starts to run.

However, a technique for measuring such task preemption delay is not currently explicitly defined. In addition, the task preemption delay time should be measured from the time when the hardware interrupt occurs, and the prior art does not define how to specify this time point.

The apparatus 100 for real-time performance measurement according to the present invention facilitates performance measurement by clearly defining a time point at which a task is performed from a time point at which an interrupt occurs.

To this end, the real-time performance measurement apparatus 100 according to the present invention is implemented in the user mode in the Linux-based embedded system, as shown in FIG.

The apparatus 100 for real-time performance measurement communicates with some of the components 310-330 in kernel mode via the system interface 200.

Kernel mode components include an interrupt handler 310, a scheduler 320, a device driver 330, and the like. The device driver 330 includes a network driver 332, a display driver 334, a real-time clock (RTC) 336, and the like.

The interrupt handler 310 refers to a routine or a function to be executed when a specific interrupt occurs. In other words, when an interrupt occurs, the interrupt handler 310 requests the scheduler 320 to execute a task corresponding to the generated interrupt. To this end, the interrupt handler 310 notifies the scheduler 320 of a task corresponding to the generated interrupt when an interrupt occurs. In detail, the interrupt handler 310 wakes up tasks having a high priority according to an interrupt. In this case, the interrupt handler 310 records the system counter value at the time when the interrupt is received as an interrupt generation time value and provides the interrupt generation time value to the real-time performance measuring apparatus 100 according to the present invention.

The scheduler 320 schedules tasks to be performed by the CPU when the task corresponding to the interrupt is woken up by the interrupt handler 310. That is, the scheduler 320 performs the scheduling so that the wake-up task, that is, the task having the highest priority, is executed first or priority. The scheduler 320 notifies the application performing the task in the user mode to perform the tasks as scheduled. That is, the scheduler 320 generates a task execution signal for executing the corresponding task and provides it to the application.

In accordance with the present invention, the RTC 336 is configured to generate an interrupt periodically by the real-time performance measurement device 100. The RTC 336 is mostly provided in a general embedded system.

According to the present invention, the apparatus 100 for real-time performance measurement defines a section between a time point at which the interrupt handler 310 is called and a time point at which the scheduler 320 notifies a corresponding application to perform a task according to an interrupt as a performance measurement section. do. The performance measurement interval according to the present invention will be described in detail with reference to FIG.

2 is a diagram showing a real-time performance measurement interval according to the present invention.

Referring to FIG. 2, the preemption delay time includes a hardware interrupt delay time 415, an interrupt handler duration 425, a scheduler delay time 435, and the like. Scheduling Latency 445 and Mode Switch Latency 455.

The hardware interrupt delay time 415 corresponds to the interval from the time point 410 at which the actual internal / external specific interrupt occurs to the time point 420 at which the interrupt handler 320 for the generated interrupt is called. The interrupt handler section 425 corresponds to a section for processing the generated interrupt in software. In other words, the interrupt handler section 425 corresponds to the time taken to perform a service routine for the generated interrupt. That is, the interrupt handler section 425 corresponds to the time from the time point 420 at which the interrupt handler 420 starts to the time point 430 at which the interrupt handler 430 ends. The scheduler delay time 435 corresponds to a delay time from the time point 430 when the interrupt handler 310 ends to the time point 440 when the scheduler 320 is executed. The scheduling delay time 445 corresponds to the time required for the actual scheduling in the scheduler 320. The mode switch delay time 455 corresponds to a time from a time point 450 when scheduling is completed to a time point 460 when a related application on the user mode is called to perform the highest priority task.

Originally, the preemption delay time corresponds to the time from the time point 410 to the time point 460. In other words, the part to be measured in time is the section from point 410 to point 460. However, the hardware interrupt latency 415 is difficult to measure unless additional equipment is provided, either realistically or technically.

However, this hardware interrupt delay time 415 is actually hardware time and takes up very little of the total preemption delay time. In addition, the interval of the hardware interrupt delay time 415 is very short so that the value can be ignored. Therefore, the present invention measures the time taken from point 420 to point 460 and uses the resulting measured value as the preemption delay time.

As described above, when the interrupt for the task is generated to measure the real-time performance according to the present invention, the apparatus 100 for measuring the real-time performance in the user mode starts from the time when the interrupt handler 310 for the generated interrupt is executed. Measure the time until the task starts executing.

To this end, the real-time performance measuring apparatus 100 receives the interrupt generation time point value from the interrupt handler 310. The apparatus 100 for real-time performance measurement records the system counter value at the time when the task execution signal is received from the scheduler 320 as the task start time value. Subsequently, the real-time performance measuring apparatus 100 calculates the preemption delay time by subtracting the interrupt occurrence time value A from the task start time value. Hereinafter, a process in which the real-time performance measuring apparatus 100 measures the preemption delay time will be described with reference to FIG. 3.

3 is a diagram illustrating a process of measuring a preemption delay time according to an embodiment of the present invention.

1 and 3, the apparatus 100 for real-time performance measurement sets the RTC 336 to periodically generate an interrupt in step 502 and drives the RTC 336. Accordingly, the RTC 336 generates a periodic interrupt at step 504. The RTC 336 may periodically generate an interrupt once set by the real-time performance measuring apparatus 100. The interrupts do not necessarily need to be generated periodically, but are preferably generated periodically for convenience of measurement. In this way, an interrupt or an event may be generated according to a preset value, that is, according to a predetermined period. This interrupt generation period can be easily adjusted. RTC driver 336 provides the interrupt generated to interrupt handler 301 in step 506.

The interrupt handler 310 records the interrupt occurrence time A in step 508 when receiving the interrupt. In detail, the interrupt handler 310 records the counter value at the time of receiving the interrupt as the interrupt generation time value A.

In this case, the high precision system counters available in both kernel mode and user mode are used. As is known in the art, embedded systems include a system clock. The system clock resides in the processor so that all the circuits and devices that operate within the computer system can work at the same time. It is a device that generates pulses at regular intervals to synchronize. The system counter can provide accurate time to the system by counting these clock pulses.

Therefore, the interrupt handler 310 reads the counter value at the time of receiving the interrupt and records it as the interrupt generation time value A. The interrupt handler 310 then wakes up a task corresponding to the interrupt among the tasks at the scheduler 320 at step 510. At the same time, or almost simultaneously, the interrupt handler 310 provides the interrupt occurrence time value A to the real-time performance measurement apparatus 100 in the user mode in step 512. In FIG. 3, the interrupt handler 310 is provided to provide an interrupt generation time point A, but in practice, the real-time performance measuring apparatus 100 reads the interrupt generation time value A recorded in the interrupt handler 310. Accordingly, the real-time performance measuring apparatus 100 may obtain the interrupt generation time point A.

Next, when the task corresponding to the interrupt wakes up by the interrupt handler 310, the scheduler 320 schedules tasks to be performed by the CPU in step 514. That is, the scheduler 320 performs the scheduling so that the wake-up task, that is, the task having the highest priority, is executed first. Next, the scheduler 320 provides the task execution signal to the real-time performance measurement apparatus 100 in step 516. In detail, the scheduler 320 notifies the real-time performance measurement apparatus 100 corresponding to the application performing the task in the user mode to perform the tasks as scheduled. In other words, the scheduler 320 generates a task execution signal for executing the task.

The apparatus 100 for real-time performance measurement records the system counter value at the time when the task execution signal is received by the scheduler 320 as the task start time value B. FIG. Subsequently, the real-time performance measuring apparatus 100 calculates the preemption delay time by subtracting the interrupt occurrence time value A from the task start time value B in step 520.

The process of Figure 3 is to measure one preemption delay time, according to an embodiment of the present invention to measure the average of a plurality of preemption delay time, and accordingly, the average of the preemption delay time is a real-time performance measurement value Used as The configuration of a real-time performance measuring apparatus operating as described above will be described with reference to FIG. 4.

4 is a block diagram of a real-time performance measurement apparatus according to an embodiment of the present invention. The real-time performance measuring device may be implemented in hardware or in software.

The real-time performance measuring apparatus 100 includes an input / output unit 110, a preprocessor 120, a counter reader 130, a delay time calculator 140, and a statistics processor 150.

The input / output unit 110 is for the real-time performance measuring apparatus 100 to input or output signals or data for kernel mode components.

The preprocessor 120 performs a memory lock. A memory lock prevents the task from being ejected or imported to memory so that tasks cannot be exported from memory and then fetched back to the hard disk. Specifically, since the amount of memory in the system is not infinite, the tasks are in memory and automatically exported to the hard disk if needed by the operating system 50, and then back to the memory if necessary. Get it and run it. The performance difference between a task that stays in memory and executes on and off the disk represents a huge performance difference. Thus, the preprocessor 120 performs the memory lock function to best measure real-time performance under the same conditions.

The preprocessor 120 also gives the highest priority to the task for performance measurement. Specifically, the task for real-time performance measurement has the highest priority so that the task for real-time performance measurement is performed before other tasks. This is because if there is a higher priority task in the system than the task for real-time performance measurement, the scheduler in the system may show lower performance since the task for real-time performance measurement is scheduled for later execution.

In addition, the preprocessor 120 performs a measurement time calibration. As described above, the interrupt occurrence time point value and the task start time point value are obtained from counter values provided by the system counter. Accordingly, the preprocessor 120 converts the measurement unit into a unit readable by the user, for example, μsec, so that the user can intuitively recognize the measurement result.

In addition, the preprocessor 120 opens, sets and runs the periodic RTC. As mentioned above, the RTC 336 is mostly embedded in a typical embedded system. The preprocessor 120 sets the RTC 336 to generate an interrupt periodically, that is, at predetermined time intervals, and executes the RTC 336. At this time, the interrupt does not necessarily need to be generated periodically, but periodic RTC is preferable for convenience of measurement.

The counter reading unit 130 reads the counter value from the system counter when the task start time is determined. In detail, the counter reading unit 130 reads the counter value of the system counter at the time when the task execution signal is received from the scheduler 320 and outputs the counter value to the delay time calculating unit 140 as the task start time value B. FIG.

The delay time calculator 140 receives the interrupt occurrence time point A from the interrupt handler 310, and calculates the preemption delay time when the task start time point B is provided from the counter reader 130. In other words, the delay time calculator 140 subtracts the interrupt occurrence time point A from the task start time point B and outputs the result as a preemption delay time.

The statistical processor 150 receives the preemption delay times from the delay time calculator 140 and performs statistical processing. The statistical processor 150 may average the plurality of delay times and output the average value as a real-time performance measurement value. The statistical processing unit 150 calculates the preemption delay time, sums up the at least one preemption delay time previously calculated, calculates an average value thereof, and outputs the average value as a real-time performance measurement value.

5 is a flowchart illustrating a real-time performance measurement method according to an embodiment of the present invention.

Referring to FIG. 5, the real-time performance measuring device performs a memory lock in step 602. As mentioned above, a memory lock prevents the task from being pulled out or pulled against the memory to prevent tasks from being exported from the memory and back to the hard disk.

Next, the apparatus for real-time performance measurement gives the highest priority to the task for performance measurement in step 604, and performs measurement time calibration in step 606, which is described above.

The real-time performance measuring device then sets up and executes the periodic RTC in step 608. Specifically, the real-time performance measuring apparatus sets the RTC to generate an interrupt, for example, at predetermined time intervals, and executes the RTC.

After executing the RTC, the real-time performance measuring apparatus determines whether an interrupt generation time point A is received from the interrupt handler in step 610. The interrupt occurrence time value A is a counter value from the system counter.

In operation 612, the apparatus for real-time performance measurement determines whether a task execution signal is received from the scheduler. The real-time performance measuring apparatus records the task start time point B in step 614 when the task execution signal is received from the scheduler. To this end, the real-time performance measuring apparatus reads the counter value of the system counter as the task start time point B at the time point when the task execution signal is received.

The real-time performance measuring apparatus then calculates the preemption delay time by subtracting the interrupt occurrence time value A from the task start time value B in step 616.

The real-time performance measuring device performs statistical processing on the currently calculated preemption delay time and at least one or more preemption delay times previously calculated. For example, the real-time performance measuring apparatus may average the plurality of delay times and output the average value as the real-time performance measurement.

In operation 620, the real-time performance measuring apparatus determines whether to receive an end signal for terminating the real-time performance measurement. Typically the exit signal is generated when the application terminates or can be sent by the user using the keyboard.

According to the present invention, it is possible to precisely measure the delay time corresponding to the clear real-time performance measurement interval using the hardware system counter than the real-time performance measurement of the existing Linux-based embedded system, and is much more user intuitive than the various methods. There is an advantage that the result of preemption delay time can be obtained.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the present invention as defined by the following claims It can be understood that

100: real-time performance measurement device 200: system interface
310: interrupt handler 320: scheduler
330: device driver 332: network driver
334: Display driver 336: Real-Time Clock (RTC)
415: hardware interrupt delay time 425: interrupt handler interval
435: Scheduler delay time 445: Scheduling delay time
455: mode switch delay time

Claims (17)

In a Linux-based embedded system,
Real Time Clock (RTC), which is configured to run in kernel mode and generates periodic interrupts,
An interrupt handler configured to operate in a kernel mode, receiving an interrupt from the RTC, recording a system counter value at the point of time at which the interrupt is received as an interrupt occurrence point value, and waking up a task corresponding to the interrupt;
A scheduler configured to operate in kernel mode, when a task corresponding to the interrupt wakes up, schedules the task having the highest priority to be executed first, and generates a task execution signal for executing the task;
Receive the interrupt occurrence time value from the interrupt handler, when receiving the task execution signal from the scheduler, record a system counter value at the received time point as a task start time value, and Linux-based embedded system including a real-time performance measurement device that calculates the preemption delay by subtracting the interrupt occurrence time value from the task start time value. The method of claim 1, wherein the real-time performance measuring device calculates the preemption delay time and adds the average of at least one preemption delay time and calculates the average value, and outputs the average value as a real-time performance measurement value. Featured Linux-based embedded system. The Linux-based embedded system of claim 1, wherein the apparatus for measuring real-time performance sets the RTC to periodically generate an interrupt. The Linux-based embedded system of claim 1, wherein the real-time performance measuring device gives the highest priority to a task for performance measurement. The Linux-based embedded system of claim 1, wherein the real-time performance measuring device implements a memory lock that prevents the task from being taken out of the memory or brought into the memory. The apparatus of claim 1, wherein the real-time performance measuring device
A preprocessor configured to periodically set the RTC to generate an interrupt;
A counter reading unit for reading a counter value of a system counter and outputting the counter value as a task starting time point at the time when the task execution signal is received from the scheduler;
And a delay time calculating unit configured to receive an interrupt occurrence time point value from the interrupt handler and calculate a preemption delay time by subtracting the interrupt occurrence time point value from the task start time point value when the task start time point value is provided from the counter reading unit. Featured Linux-based embedded system. The apparatus of claim 6, wherein the apparatus for measuring real-time performance receives the calculated preemption delay time from the delay time calculation unit, adds the previously calculated at least one preemption delay time, calculates an average thereof, and the average value. Linux-based embedded system further comprises a statistical processing unit for outputting a real-time performance measurement. In measuring real-time performance in Linux-based embedded system,
Generating a periodic interrupt by a Real Time Clock (RTC) configured to operate in kernel mode,
Receiving, by an interrupt handler configured to operate in kernel mode, an interrupt from the RTC, recording a system counter value at the time the interrupt was received as an interrupt occurrence time value,
Waking up the task corresponding to the interrupt by the interrupt handler;
Scheduling by the scheduler configured to operate in kernel mode so that when the task corresponding to the interrupt wakes up, the task having the fastest priority is executed first;
Generating a task execution signal for executing the task by the scheduler;
Receiving, by the real time performance measurement apparatus, configured to operate in a user mode, the interrupt occurrence time point value from the interrupt handler;
Recording the system counter value at the received time point as a task start time point value when the task execution signal is received from the scheduler by the real-time performance measuring device;
And calculating a preemption delay time by subtracting an interrupt occurrence time point value from the task start time point value by the real-time performance measurement device. The method of claim 8, further comprising: calculating the preemption delay time by the real-time performance measuring device and then adding the calculated preemption delays with at least one preemption delay time and calculating an average thereof;
And outputting the average value as a real-time performance measurement value by the real-time performance measurement device. The method of claim 8, further comprising setting the RTC to generate an interrupt periodically by the real-time performance measuring device before the interrupt generating step. 10. The method of claim 8, further comprising, prior to the interrupt generating step, assigning the earliest priority to a task for performance measurement by the real-time performance measurement device. 10. The method of claim 8, further comprising, prior to the interrupt generating step, assigning the earliest priority to a task for performance measurement by the real-time performance measurement device. 10. The method of claim 8, further comprising executing a memory lock prior to the interrupt generation step, the memory lock which prevents the real-time performance measuring device from fetching or bringing a task from the memory into the memory. Way. In the method for measuring the real-time performance in a real-time performance measurement device,
Setting a real time clock (RTC) configured to operate in the kernel mode by the real-time performance measuring device to generate an interrupt periodically;
Receiving an interrupt occurrence time value from an interrupt handler configured to operate in kernel mode,
Receiving a task execution signal for executing a task corresponding to the interrupt from a scheduler configured to operate in kernel mode;
Recording a system counter value at the time point of receiving the task execution signal from the scheduler as a task start time point value;
And calculating a preemption delay time by subtracting the interrupt occurrence time value from the task start time value. 15. The method of claim 14, further comprising: after calculating the preemption delay time, summing up at least one preemption delay time calculated and calculating an average value thereof;
And outputting said average value as a real-time performance measurement. 15. The method of claim 14, further comprising, prior to the setting step, assigning the earliest priority to the task for performance measurement. 10. The method of claim 8, further comprising executing a memory lock prior to the setting step to prevent a task from being taken from or brought into the memory.
KR1020100125832A 2010-12-09 2010-12-09 Method for measuring real-time performance in linux-based embedded system and the system KR20120064553A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020100125832A KR20120064553A (en) 2010-12-09 2010-12-09 Method for measuring real-time performance in linux-based embedded system and the system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020100125832A KR20120064553A (en) 2010-12-09 2010-12-09 Method for measuring real-time performance in linux-based embedded system and the system

Publications (1)

Publication Number Publication Date
KR20120064553A true KR20120064553A (en) 2012-06-19

Family

ID=46684640

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020100125832A KR20120064553A (en) 2010-12-09 2010-12-09 Method for measuring real-time performance in linux-based embedded system and the system

Country Status (1)

Country Link
KR (1) KR20120064553A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106155865A (en) * 2016-08-01 2016-11-23 浪潮(北京)电子信息产业有限公司 A kind of CPI acquisition methods, system and process performance judgement system in real time

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106155865A (en) * 2016-08-01 2016-11-23 浪潮(北京)电子信息产业有限公司 A kind of CPI acquisition methods, system and process performance judgement system in real time

Similar Documents

Publication Publication Date Title
US8402463B2 (en) Hardware threads processor core utilization
US20150082325A1 (en) Apparatuses and methods for generating and tracing event codes
US9400734B2 (en) Apparatuses and methods for generating event codes including event source
JP4840455B2 (en) Field control system
CN100338581C (en) Operating system-independent method and system of determining CPU utilization
Neukirchner et al. Monitoring arbitrary activation patterns in real-time systems
US11392475B2 (en) Job power predicting method and information processing apparatus
JP2007026095A (en) Parallel arithmetic operation device
US11249512B2 (en) Frequency execution monitoring in a real-time embedded system
US9600023B2 (en) Method and data processing unit for providing a timestamp
de Oliveira et al. Operating system noise in the linux kernel
KR20120064553A (en) Method for measuring real-time performance in linux-based embedded system and the system
US10819896B1 (en) Computer readable recording medium can perform image sensing system control method and image sensing system
Emde Long-term monitoring of apparent latency in PREEMPT RT Linux realtime systems
US20160299787A1 (en) System, method and managing device
US8160845B2 (en) Method for emulating operating system jitter
KR101674324B1 (en) Task scheduling device and method for real-time control applications
JP4996623B2 (en) Detection apparatus, program, and detection method
CN114071127A (en) Live video delay testing method and device, storage medium and electronic equipment
WO2009026361A2 (en) Method, system and apparatus for measuring an idle value of a central processing unit
CN112882857A (en) Performance problem positioning method and device, electronic equipment and storage medium
Lakhani et al. Applying design patterns to improve the reliability of embedded systems through a process of architecture migration
Fujimoto et al. PWU: Pre-Wakeup for CPU Idle to Reduce Latency and Power Consumption
CN115826731B (en) Sleep control method and device, storage medium and computing equipment
CN105630714B (en) Interface resource analytical equipment and its method

Legal Events

Date Code Title Description
WITN Withdrawal due to no request for examination