JP2015088068A - Information processing device, method for analyzing information processing device, and program for analyzing information processing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an information processing device or the like capable of identifying, with high accuracy, a function that constitutes the cause of idling.SOLUTION: The information processing device comprises: an arithmetic processing unit for executing a program; a function information collection unit for collecting, at prescribed intervals, function information which is the information indicating the function under execution, and storing, in a storage unit, the function information and a sampling time of day which is the time of day at which the function information was collected; a transition acquisition unit for acquiring, when the arithmetic unit transitions to an idle state, a transition time of day and storing the acquired time of day in the storage unit; an extraction unit for extracting a difference between the transition time of day stored in the storage unit and a sampling time of day, among the sampling times of day stored in the storage unit, that immediately precedes the transition time of day, and the function information corresponding to the sampling time of day; and an identification unit for identifying a function that constitutes the cause of idling on the basis of the distribution of differences corresponding to each function information extracted by the extraction unit.

Description

  The present invention relates to an information processing apparatus, an information processing apparatus analysis method, and an information processing apparatus analysis program.

  Collecting information indicating the processing contents being executed from the information processing apparatus that has executed the program, and confirming the operation of the executed program based on the aggregation result obtained by performing aggregation processing on the collected information. Conventionally, performance analysis of programs has been performed (see, for example, Patent Document 1).

JP 2011-164839 A

  However, the analysis by the conventional profiler can only grasp the phenomenon that the ratio of the idle time to the execution time is high. For this reason, in the conventional technique, since the program is not executed, there is a problem that the function that causes the idle that creates the idle state that affects the performance cannot be accurately specified.

  In one aspect, an object of the present invention is to provide an information processing apparatus or the like that can specify a function that causes idle more accurately.

  The information processing device disclosed in the present application collects function information, which is information indicating a function being executed, and an arithmetic processing unit that executes a program at predetermined intervals, and at the time when the function information and the function information are collected. A function information collection unit that stores a certain sampling time in a storage unit, a transition acquisition unit that acquires a transition time and stores it in the storage unit when the arithmetic processing unit shifts to an idle state, and a memory that is stored in the storage unit The difference between the transition time and the sampling time immediately before the transition time among the sampling times stored in the storage unit, and an extraction unit that extracts function information corresponding to the sampling time, and each function extracted by the extraction unit And a specifying unit that specifies a function that causes idle, based on a distribution of differences corresponding to the information.

  In one aspect, it is possible to specify a function that causes idle more accurately.

It is a block diagram which shows the hardware group of information processing apparatus. It is explanatory drawing which shows the image of a sampling process. It is explanatory drawing which shows the record layout of a sampling table. It is explanatory drawing which shows the record layout of a transition time table. It is explanatory drawing which shows the record layout of a difference table. It is explanatory drawing which shows the record layout of a statistics table. It is a flowchart which shows the procedure of the production | generation process of a sampling table. It is a flowchart which shows the acquisition processing procedure of a transition time. It is a flowchart which shows the acquisition process procedure of a difference and a function. It is a flowchart which shows the extraction process procedure of a function with high possibility of causing an idle. It is explanatory drawing which shows a sampling result. It is a graph which shows the change of the data memorized by the statistics table. 10 is a flowchart illustrating a function specifying process procedure according to the second embodiment. 10 is a flowchart illustrating a function specifying process procedure according to the second embodiment. FIG. 10 is a block diagram illustrating a hardware group of a computer according to a third embodiment. It is explanatory drawing which shows the image of a sampling process. It is explanatory drawing which shows the record layout of an end time table. It is explanatory drawing which shows the record layout of an end difference table. It is explanatory drawing which shows the record layout of a transfer statistics table. It is explanatory drawing which shows the record layout of an end statistics table. It is explanatory drawing which shows the record layout of the statistics table which concerns on Embodiment 2. FIG. It is a flowchart which shows the acquisition process procedure of end time. It is a flowchart which shows the difference and function extraction process procedure which concern on end time. It is a flowchart which shows the extraction process procedure of a function with high possibility of causing an idle. It is a flowchart which shows the procedure of a weighting process. It is a functional block diagram which shows operation | movement of the computer of the form mentioned above. FIG. 16 is a block diagram illustrating a hardware group of a computer according to a fifth embodiment.

Embodiment 1
Hereinafter, embodiments will be described with reference to the drawings. FIG. 1 is a block diagram illustrating a hardware group of the information processing apparatus 1. The information processing apparatus 1 is a personal computer, server computer, mobile phone, game machine, book reader, PDA (Personal Digital Assistant) or the like to be analyzed. Hereinafter, the information processing apparatus 1 will be described as being replaced with the computer 1. The computer 1 includes a central processing unit (CPU) 11 as a control unit, a random access memory (RAM) 12, an input unit 13, a display unit 14, a clock unit 18, a storage unit 15, a communication unit 16, and the like. The CPU 11 is connected to each part of the hardware via the bus 17. The CPU 11 controls each part of the hardware according to the control program 15P stored in the storage unit 15. The RAM 12 is, for example, SRAM (Static RAM), DRAM (Dynamic RAM), flash memory, or the like. The RAM 12 also functions as a storage unit, and temporarily stores various data generated when the CPU 11 executes various programs.

  The input unit 13 is an input device such as a mouse, a keyboard, or a touch panel, and outputs received operation information to the CPU 11. The display unit 14 is a liquid crystal display, an organic EL display, or the like, and displays various information according to instructions from the CPU 11. The communication unit 16 is a communication module and transmits / receives information to / from another computer (not shown) via a communication network.

  The clock unit 18 outputs time information to the CPU 11. The storage unit 15 is a hard disk or a large-capacity memory, and stores a control program 15P, an OS (Operating System) 151, a user program 153, a sampling table 154, a transition time table 155, a difference table 156, a statistics table 157, and the like. The CPU 11 loads the user program 153 into the RAM 12. The CPU 11 generates interrupts at predetermined intervals by the profiler 152 which is a function of the OS 151. The profiler 152 acquires identification information for identifying a process (hereinafter referred to as PID), an execution instruction address, and an interrupt time (hereinafter referred to as sampling time) by periodic sampling processing by interruption.

  The execution instruction address is an address on the RAM 12 where the sampling target function read from the OS 151 and the user program 153 is stored. In the present embodiment, the profiler 152 is described as a function of the OS 151, but may be an independent program different from the OS 151.

  FIG. 2 is an explanatory diagram showing an image of sampling processing. The horizontal axis represents time, and shows a state in which functions (hereinafter referred to as user functions) are successively called from the user program 153 by the CPU 11 as time passes, and are being executed. The OS 151 calls an idle function for setting the CPU 11 in an idle state. As shown in FIG. 2, the user function and the idle function are sequentially called by the CPU 11 and various processes are executed. Here, the greater the proportion of the idle state, the lower the usage rate of the CPU 11.

  A downward arrow in FIG. 2 indicates the timing of the sampling process by the profiler 152. The CPU 11 obtains the PID, the execution instruction address, and the sampling time at predetermined time intervals stored in the storage unit 15 by the interrupt handler of the profiler 152. Note that the sampling interval can be changed by the user from the input unit 13 or the like.

  FIG. 3 is an explanatory diagram showing a record layout of the sampling table 154. The sampling table 154 includes a PID field, an execution instruction address field, a sampling time field, a process name field, and a function name field. The PID field stores PIDs for specifying processes related to user functions and idle functions called from the OS 151 and the user program 153. In the execution instruction address field, an address on the RAM 12 where the sampling target function is stored is stored in association with the PID.

  In the sampling time field, the time when the profiler 152 performs sampling is stored in association with the execution instruction address. In the process name field, the name of the process executed at the time of sampling is stored in association with the PID or the like. In the function name field, the function name of the OS 151 or the user program 153 at the time of sampling is stored in association with the sampling time. The CPU 11 stores the PID, execution instruction address, and sampling time periodically acquired by the profiler 152 in the sampling table 154.

  The CPU 11 acquires a process name and a function name corresponding to the PID and the execution instruction address from the OS 151. The CPU 11 stores the acquired process name and function name in the sampling table 154 in association with the execution instruction address or the like. The CPU 11 acquires the process name and function name corresponding to the PID and execution instruction address of the user program 153 from the binary data of the user program 153. The CPU 11 stores the acquired process name and function name in the sampling table 154 in association with the execution instruction address or the like. In the example of FIG. 3, it can be understood that the user function “func_c” is executed during the first sampling, and the idle function indicated by “poll_idle” is executed during the second and third samplings. .

  Returning to FIG. The upward arrow in FIG. 2 indicates the transition timing to the idle state. When the CPU 11 uses the debug function or the event registration function of the OS 151 and detects the switching from the user function to the idle function, the CPU 11 calls the profiler 152 and obtains the time of transition to the idle state (hereinafter referred to as transition time). To do. The event registration function of the OS 151 may use, for example, Linux (registered trademark) Kprobes or Systemap. Specifically, the CPU 11 may register the transition from the user function to the idle function as an event, and acquire the transition time when the registered event is detected. As for the debugging function of the CPU 11, for example, a method described in “Intel (registered trademark) 64 and IA-32 Architectures Software Developer's Manual Volume 3B: System Programming Guide, Part 2” by Intel may be used. For example, the CPU 11 may store a specific execution instruction address of the user function or the idle function, and may determine that the state has shifted to the idle state when the stored execution instruction address is detected. Hereinafter, an example using the debug function will be described. The CPU 11 acquires the transition time associated with the transition to the idle state together with the sampling process.

  FIG. 4 is an explanatory diagram showing a record layout of the transition time table 155. When acquiring the transition time, the CPU 11 sequentially stores the transition time in the transition time table 155. FIG. 5 is an explanatory diagram showing a record layout of the difference table 156. The difference table 156 includes a PID field, a process name field, a function name field, a difference field, and the like. The CPU 11 generates a difference table 156 based on the stored contents of the sampling table 154 and the transition time table 155. The CPU 11 refers to the sampling table 154 and the transition time table 155 to extract the sampling time immediately before the transition time. The CPU 11 refers to the sampling table 154 and extracts the PID, process name, and function name corresponding to the extracted sampling time.

  The CPU 11 stores the extracted PID, process name, and function name in the difference table 156. The CPU 11 calculates the difference between the transition time and the immediately preceding sampling time and stores it in the difference field. In the difference field, the real time is indicated in parentheses in parentheses for reference.

  FIG. 6 is an explanatory diagram showing a record layout of the statistics table 157. The CPU 11 classifies the difference distribution of each function based on the difference time zone stored in advance in the storage unit 15. In the example of FIG. 6, the time zone is 5 μs. Specifically, it is classified every 5 μs, such as less than 5 μs, 5 μs or more, less than 10 μs, 10 μs or more, and less than 15 μs. The time zone is merely an example and is not limited to this. Further, the user can change the time zone from the input unit 13 or the like.

  The CPU 11 refers to the difference table 156 and counts the number of differences belonging to the time zone for each function based on the function name and the difference. In the example of FIG. 6, it can be understood that the number of differences less than 5 μs is 2015 in the function a indicated by “func_a”, and the number of differences between 5 μs and less than 10 μs is 2. In addition, it can be understood that the function b indicated by “func_b” has 186 differences of less than 5 μs, and 1861 has a difference of 5 μs or more and less than 10 μs. Similarly, in the function c indicated by “func_c”, it can be understood that the number of differences less than 5 μs is 144, and the number of differences between 5 μs and less than 10 μs is 1822.

  The CPU 11 specifies the function with the largest number classified into the time zone with the shortest time zone. In the example of FIG. 6, the function a having the largest number classified into 5 μs with the shortest time zone, which is 2015, is specified as the function causing the idle. In the above hardware group, various software processes will be described with reference to flowcharts.

  FIG. 7 is a flowchart showing a procedure for generating the sampling table 154. The CPU 11 refers to the RAM 12 and acquires the PID, the execution instruction address, and the sampling time that are being executed at predetermined time intervals by the profiler 152 (step S71). The CPU 11 stores the acquired PID, execution instruction address, and sampling time in the sampling table 154 (step S72). The CPU 11 acquires a process name and a function name corresponding to the PID and the execution instruction address from the OS 151 (step S73).

  The CPU 11 acquires a process name and a function name corresponding to the PID and the execution instruction address from the user program 153 (step S74). The CPU 11 stores the process name and function name acquired in steps S73 and S74 in the sampling table 154 in association with the PID and the execution instruction address (step S75). When the execution instruction address is the OS 151 area, the CPU 11 acquires the process name and function name from the OS 151. When the execution instruction address is the user program area, the CPU 11 and the OS 151 and user program information (a symbol created by the program binary or the OS 151 when the program is loaded) The process name and function name may be acquired from the information. Note that these two acquisition processes may be performed as one process. Further, the control program 15P operates integrally with the profiler 152. Here, when the profiler 152 is incorporated in the OS 151, the control program 15P is also incorporated in the OS 151 or provided in the form of an add-on module, and is loaded into the OS 151 and executed. When the profiler 152 is an independent program, the profiler 152 is provided including the control program 15P or provided as a program operating in cooperation.

  FIG. 8 is a flowchart showing a procedure for acquiring the transition time. The CPU 11 determines whether or not the transition to the idle state is detected by the debug function (step S81). If the CPU 11 determines that the transition to the idle state has not been detected (NO in step S81), the CPU 11 waits until the idle state is detected. If the CPU 11 detects the transition to the idle state (YES in step S81), the CPU 11 proceeds to step S82. The CPU 11 acquires the transition time by the profiler 152 (step S82). The CPU 11 stores the acquired transition time in the transition time table 155 (step S83).

  FIG. 9 is a flowchart showing a difference and function acquisition processing procedure. The CPU 11 refers to the transition time table 155 and extracts the transition time (step S91). The CPU 11 refers to the sampling table 154 and extracts the sampling time immediately before the transition time (step S92). The CPU 11 calculates the difference by subtracting the sampling time extracted in step S92 from the transition time extracted in step S91 (step S93). The CPU 11 reads the PID, process name, and function name corresponding to the sampling time from the sampling table 154 (step S94).

  The CPU 11 associates the difference calculated in step S93 with the PID, process name, and function name read in step S94, and stores them in the difference table 156 (step S95). The CPU 11 determines whether or not the above-described processing has been completed for all the transition times stored in the transition time table 155 (step S96). If the CPU 11 determines that the process has not ended (NO in step S96), the process proceeds to step S97. The CPU 11 refers to the transition time table 155 and reads the next transition time in time series (step S97). Thereafter, the CPU 11 returns the process to step S92. As a result, the function name and the difference are accumulated. CPU11 complete | finishes a process, when it is judged that the process was complete | finished about all the transition times (it is YES at step S96).

  FIG. 10 is a flowchart showing a procedure for extracting a function that is likely to cause idle. CPU11 reads a function from the difference table 156 (step S101). The CPU 11 reads the difference corresponding to the function from the difference table 156 (step S102). The CPU 11 reads a plurality of time zones continuous in a plurality of time series from the storage unit 15. The CPU 11 counts the number of differences belonging to each read time zone (step S103). The CPU 11 stores the counted number in the statistical table 157 in association with the function and the time zone (step S104). Thus, the number of differences belonging to each time zone can be grasped for one function.

  The CPU 11 determines whether the processing has been completed for all functions (step S105). If the CPU 11 determines that the process has not ended (NO in step S105), the CPU 11 reads an unprocessed function from the difference table 156 (step S106). Thereafter, the CPU 11 returns the process to step S102. If the CPU 11 determines that the processing has been completed for all functions (YES in step S105), the CPU 11 proceeds to step S107. By repeating the above processing, difference distributions of all functions can be acquired.

  The CPU 11 refers to the statistical table 157 and extracts the function having the largest number classified into the time zone with the shortest difference (step S107). The CPU 11 outputs the extracted function to the display unit 14 (step S108). Note that the output destination of the extracted function is not limited to the display unit 14. In addition to outputting sound from a microphone (not shown), the sound may be output to another computer (not shown) via a communication network such as the Internet.

  Subsequently, an experiment was conducted to confirm the effect. FIG. 11 is an explanatory diagram showing a sampling result. There are three user programs 153, function A, function B, and function C, and sampling processing was performed at a period of 50 μs. Note that the function A is a function that causes idling. FIG. 11A shows the ratio of the number of samplings with respect to each of the idle function and functions A to C. The ratio extracted as an idle function was 76.4%, the function A was 4.2%, the function B was 10.9%, and the function C was 8.6%. It can be seen that the proportion in the idle state is high and the proportion occupied by the function B is also large.

  Subsequently, the number of functions sampled immediately before the transition time was investigated. FIG. 11B shows the sampling number and ratio immediately before the transition time for the functions A to C. The sampling number of the function A was 4138, the ratio was 17.6%, the sampling number of the function B was 10867, the ratio was 46.2%, the sampling number of the function C was 8534, and the ratio was 36.3%. If the determination is made based only on the count value of the function obtained by sampling immediately before shifting to the idle state, the function B is erroneously recognized as a function related to the cause of idle.

  FIG. 12 is a graph showing changes in data stored in the statistical table 157. As shown in FIG. 12, in the method according to the present embodiment, it is noticeable that the function A is the function having the largest time zone with the shortest difference. On the other hand, the functions B and C are concentrated in a time zone in which the difference is relatively long. From the above, in the method according to the present embodiment, it is possible to accurately identify the function that causes idling as the function A.

Embodiment 2
The second embodiment relates to a mode in which a function that causes idle is determined based on the number of a plurality of time zones. In the first embodiment, the time zone having the shortest difference is specified, and in the example of FIG. 6, the function having the maximum number of less than 5 μs is specified. However, the present invention is not limited to this. A weight may be set for each time slot, and the number of each time slot may be corrected according to the weight. Further, when the number of the shortest time zones is not greatly different between the functions, the function may be specified based on the number of the next shortest time zones. Details will be described below.

  13 and 14 are flowcharts showing the function specifying process procedure according to the second embodiment. The CPU 11 refers to the statistical table 157 and extracts the function having the largest number classified in the shortest time zone (hereinafter referred to as the first function) (step S131). The CPU 11 extracts the function having the next largest number classified into the shortest time zone (hereinafter referred to as the second function) (step S132). The CPU 11 subtracts the number of second functions extracted in step S132 from the number of first functions extracted in step S131 (step S133). The CPU 11 reads a threshold value stored in advance in the storage unit 15 (step S134).

  The CPU 11 determines whether or not the subtraction value calculated in step S133 exceeds a threshold value (step S135). If the CPU 11 determines that the subtraction value exceeds the threshold value (YES in step S135), the process proceeds to step S139. The CPU 11 outputs the first function as a function estimated to be the cause of idle (step S139). When the CPU 11 determines that the subtraction value does not exceed the threshold value (NO in step S135), the CPU 11 shifts the processing to step S136. CPU11 extracts the number classified into the next short time slot about the 1st function (Step S136).

  CPU11 extracts the number classified into the next short time slot about the 2nd function (Step S137). The CPU 11 compares the number extracted in step S136 with the number extracted in step S137, and outputs a function with many extracted numbers as a function estimated to be the cause of idle (step S138). This makes it possible to specify a function that causes idle more accurately with respect to a function having characteristics that are difficult to discriminate.

  The second embodiment is as described above, and the other parts are the same as those of the first embodiment. Therefore, the corresponding parts are denoted by the same reference numerals, and detailed description thereof is omitted.

Embodiment 3
The third embodiment relates to a mode of using an end time when shifting from an idle function to a user function. FIG. 15 is a block diagram illustrating a hardware group of the computer 1 according to the third embodiment. The storage unit 15 further includes an end time table 158, an end difference table 159, a migration statistics table 1510, and an end statistics table 1511. FIG. 16 is an explanatory diagram showing an image of sampling processing. In addition to the processing of the first embodiment, the CPU 11 uses the debugging function or the event registration function of the OS 151 as in the first embodiment to call the profiler 152 when detecting the switching from the idle function to the user function, The time when the user function is returned from the idle state (hereinafter referred to as the end time) is acquired. In the second embodiment, an example using the debug function will be described. The end time in FIG. 16 is indicated by an upward dotted arrow. The CPU 11 calculates a difference from the sampling time immediately after the end time is detected. The CPU 11 specifies a function that causes idling in consideration of the difference calculated based on the end time.

  FIG. 17 is an explanatory diagram showing a record layout of the end time table 158. When the CPU 11 detects the switching from the idle function to the user function using the debug function, the CPU 11 calls the profiler 152 and obtains the time when the user function is returned from the idle state. Each time the CPU 11 acquires the end time, the CPU 11 stores the acquired end time in the end time table 158.

  FIG. 18 is an explanatory diagram showing a record layout of the end difference table 159. The end difference table 159 includes a PID field, a process name field, a function name field, a difference field, and the like. The CPU 11 generates an end difference table 159 based on the stored contents of the sampling table 154 and the end time table 158. The CPU 11 refers to the sampling table 154 and the end time table 158 to extract the sampling time immediately after the end time. The CPU 11 refers to the sampling table 154 and extracts the PID, process name, and function name corresponding to the extracted sampling time.

  The CPU 11 stores the extracted PID, process name, and function name in the end difference table 159. The CPU 11 obtains a difference between the end time and the immediately following sampling time and stores it in the difference field. In the difference field, the real time is indicated in parentheses in parentheses for reference.

  FIG. 19 is an explanatory diagram showing a record layout of the migration statistics table 1510. Since the stored contents of the migration statistics table 1510 are the same as the statistics table 157 described in the first embodiment, detailed description thereof is omitted. FIG. 20 is an explanatory diagram showing a record layout of the end statistics table 1511. The CPU 11 refers to the end difference table 159 shown in FIG. 18 and classifies the difference distribution of each function based on the difference time zone stored in advance in the storage unit 15. This time zone may be the same time zone as the migration statistics table 1510 shown in FIG.

  The CPU 11 refers to the end difference table 159 and counts the number of differences belonging to the time zone for each function based on the function name and the difference. In the example of FIG. 20, it can be understood that the function a indicated by “func_a” has 2013 the number of differences of less than 5 μs and 4 the number of 5 μs or more and less than 10 μs. In addition, in the function b indicated by “func_b”, it can be understood that the number of differences less than 5 μS is 184, and the number of differences between 5 μs and less than 10 μs is 1865. Similarly, in the function c indicated by “func_c”, it can be understood that the number of differences less than 5 μs is 139 and the number of differences between 5 μs and less than 10 μs is 1827.

  FIG. 21 is an explanatory diagram showing a record layout of the statistics table 157 according to the second embodiment. The statistics table 157 is a total of the migration statistics table 1510 and the end statistics table 1511 shown in FIG. The CPU 11 refers to the transition statistics table 1510 and the end statistics table 1511, and for each function, the total value of the number of differences calculated based on the transition time of each time zone and the number of differences calculated based on the end time Is calculated. The CPU 11 stores the total value in the statistics table 157 in association with the function and the time zone. In the example of FIG. 21, it can be understood that the number of differences less than 5 μs is 4028 and the number of functions a greater than 5 μs is less than 10 μs in the function a indicated by “func_a”. Further, it can be understood that the function b indicated by “func_b” has 370 the number of differences of less than 5 μs and 3726 the number of 5 μs or more and less than 10 μs. Similarly, in the function c indicated by “func_c”, it can be understood that the number of differences less than 5 μs is 283, and the number of differences between 5 μs and less than 10 μs is 3649.

  FIG. 22 is a flowchart showing an end time acquisition processing procedure. The CPU 11 determines whether or not a return from the idle state to the execution of the user program 153 is detected by the debug function (step S221). If the CPU 11 determines that no return has been detected (NO in step S221), the CPU 11 waits until a return is detected. When the CPU 11 detects return (YES in step S221), the CPU 11 proceeds to step S222. The CPU 11 obtains the end time using the profiler 152 (step S222). The CPU 11 stores the acquired end time in the end time table 158 (step S223).

  FIG. 23 is a flowchart showing a difference and function extraction process procedure related to the end time. The CPU 11 refers to the end time table 158 and extracts the end time (step S231). The CPU 11 refers to the sampling table 154 and extracts the sampling time immediately after the end time (step S232). The CPU 11 calculates the difference by subtracting the end time extracted in step S231 from the sampling time extracted in step S232 (step S233). The CPU 11 reads the PID, process name, and function name corresponding to the sampling time from the sampling table 154 (step S234).

  The CPU 11 associates the difference calculated in step S233 with the PID, process name, and function name read in step S234 and stores them in the end difference table 159 (step S235). The CPU 11 determines whether or not the above-described processing has been completed for all the end times stored in the end time table 158 (step S236). If the CPU 11 determines that the process has not ended (NO in step S236), the process proceeds to step S237. The CPU 11 reads the next end time in time series with reference to the end time table 158 (step S237). Thereafter, the CPU 11 returns the process to step S232. As a result, the function name and the difference are accumulated. If the CPU 11 determines that the process has been completed for all end times (YES in step S236), the process ends.

  FIG. 24 is a flowchart showing a procedure for extracting a function having a high possibility of causing idling. The CPU 11 reads the function from the end difference table 159 (step S241). The CPU 11 reads the difference corresponding to the function from the end difference table 159 (step S242). The CPU 11 reads a plurality of time zones continuous in a plurality of time series from the storage unit 15. The CPU 11 counts the number of differences belonging to each read time zone (step S243). The CPU 11 stores the counted number in the end statistics table 1511 in association with the function and the time zone (step S244).

  The CPU 11 determines whether or not processing has been completed for all functions (step S245). If the CPU 11 determines that the process has not ended (NO in step S245), the CPU 11 reads an unprocessed function from the end difference table 159 (step S246). Thereafter, the CPU 11 returns the process to step S242. If the CPU 11 determines that the processing has been completed for all functions (YES in step S245), the CPU 11 proceeds to step S247. By repeating the above processing, difference distributions of all functions can be acquired.

  The CPU 11 reads the statistical table 157 generated by the processing described in FIG. 10 as the migration statistical table 1510, and further reads the end statistical table 1511 generated by the processing of steps S241 to S246 (step S247). The CPU 11 refers to the migration statistics table 1510 and the end statistics table 1511, calculates the total value of the time zones of each function, and generates the statistics table 157 (step S248).

  The CPU 11 refers to the statistical table 157 and extracts the function having the largest number classified into the time zone with the shortest difference (step S249). The CPU 11 outputs the extracted function to the display unit 14 (step S2410). As described above, since a function having a large number of times of returning from the idle state to the execution of the program function is also taken into consideration, it becomes possible to specify the function causing the idle more accurately.

  The third embodiment is as described above, and the others are the same as in the first and second embodiments. Therefore, the corresponding parts are denoted by the same reference numerals, and detailed description thereof is omitted.

Embodiment 4
The fourth embodiment relates to a mode in which weighting processing is performed on the number of the migration statistics table 1510 or the end statistics table 1511. FIG. 25 is a flowchart showing the procedure of the weighting process. The CPU 11 refers to the migration statistics table 1510 and reads the number of each function classified in the shortest time zone (step S251). The CPU 11 refers to the end statistics table 1511 and reads the number of each function classified in the shortest time zone (step S252). CPU11 reads the coefficient as a weight previously memorize | stored in the memory | storage part 15 (step S253). The CPU 11 corrects the number of functions by multiplying the number of each function read in step S252 in the end statistics table 1511 by a coefficient (step S254).

  For example, the coefficient may be a number smaller than 1 (for example, 0.8) or a number larger than 1 (for example, 1.2). In this embodiment, the number of end statistics tables 1511 is multiplied by a coefficient. However, the number of migration statistics tables 1510 read in step S251 may be multiplied by a coefficient. Alternatively, both may be multiplied by different values. The CPU 11 adds the number of functions read for the transition statistics table 1510 in step S251 and the number of corrected functions corrected in step S254 (step S255). The CPU 11 outputs the function having the largest added value as a function that causes idle (step S256). In the present embodiment, the number of functions classified in the shortest time zone is weighted, but the present invention is not limited to this. A weighting process may be performed on the number of functions classified in other time zones. Then, the CPU 11 adds the numbers classified in all the time zones of the migration statistics table 1510 and the corrected end statistics table 1511. Thereafter, as described in the second embodiment, the CPU 11 may specify a function that causes an idle based on the number of the shortest time zones and the number of the next shortest time zones. As described above, by performing the weighting process according to the characteristics of the function, it is possible to specify the function that becomes the idle factor with higher accuracy.

  The fourth embodiment is as described above, and the others are the same as those of the first to third embodiments. Therefore, the corresponding parts are denoted by the same reference numerals and detailed description thereof is omitted.

Embodiment 5
FIG. 26 is a functional block diagram showing the operation of the computer 1 of the above-described form. When the CPU 11 executes the control program 15P, the computer 1 operates as follows. The function information collection unit 261 collects function information, which is information indicating a function being executed, at predetermined intervals, and stores a sampling time, which is a time when the function information and the function information are collected, in the storage unit 15. When the CPU 11 shifts to the idle state, the shift acquisition unit 262 acquires the shift time and stores it in the storage unit 15. The extraction unit 263 extracts the difference between the transition time stored in the storage unit 15 and the sampling time immediately before the transition time among the sampling times stored in the storage unit 15, and the function information corresponding to the sampling time To do. The specifying unit 264 specifies a function that causes an idle based on the distribution of differences corresponding to each function information extracted by the extracting unit 263. The classification unit 265 classifies the difference of each function according to a plurality of difference time zones. The function specifying unit 267 specifies the function that causes the idle based on the number classified into the time zone with a short difference.

  FIG. 27 is a block diagram illustrating a hardware group of the computer 1 according to the fifth embodiment. A program for operating the computer 1 reads a portable recording medium 1A such as a CD-ROM, a DVD (Digital Versatile Disc) disk, a memory card, or a USB (Universal Serial Bus) memory into a reading unit 10A such as a disk drive. It may be stored in the storage unit 15. Further, a semiconductor memory 1B such as a flash memory storing the program may be mounted in the computer 1. Further, the program can be downloaded from another server computer (not shown) connected via a communication network such as the Internet. The contents will be described below.

  The computer 1 shown in FIG. 27 reads a program for executing the above-described various software processes from the portable recording medium 1A or the semiconductor memory 1B or downloads it from another server computer (not shown) via a communication network. . The program is installed as the control program 15P, loaded into the RAM 12, and executed. Thereby, it functions as the computer 1 described above.

  The fifth embodiment is as described above, and the other parts are the same as those of the first embodiment. Therefore, corresponding parts are denoted by the same reference numerals, and detailed description thereof is omitted.

  With respect to the embodiments including the first to fifth embodiments, the following additional notes are disclosed.

(Appendix 1)
An arithmetic processing unit for executing a program;
A function information collection unit that collects function information that is information indicating a function being executed at a predetermined interval, and stores a sampling time that is a time at which the function information and the function information are collected;
When the arithmetic processing unit transitions to an idle state, a transition acquisition unit that acquires a transition time and stores it in the storage unit;
A difference between the transition time stored in the storage unit and the sampling time immediately before the transition time among the sampling times stored in the storage unit, and an extraction unit that extracts function information corresponding to the sampling time;
An information processing apparatus comprising: a specifying unit that specifies a function that causes an idle based on a distribution of differences corresponding to each function information extracted by the extraction unit.

(Appendix 2)
The specific part is:
A classification unit that classifies the difference of each function according to a plurality of difference time zones;
The information processing apparatus according to appendix 1, further comprising: a function specifying unit that specifies a function that causes the idle based on a number classified into a time zone in which the difference is short.

(Appendix 3)
The function specifying unit is
The information processing apparatus according to claim 2, wherein a function having the largest number classified into the shortest time zone is specified as the function causing the idle.

(Appendix 4)
An end acquisition unit that acquires an end time when the idle state ends;
A difference between the end time and the sampling time immediately after the end time, and an end extraction unit that extracts a function corresponding to the sampling time,
The specific part is:
The information processing apparatus according to claim 1, wherein a function that causes idle is identified based on a difference distribution of each function extracted by the extraction unit and the end extraction unit.

(Appendix 5)
The specific part is:
A plurality of classification units for classifying the difference of each function extracted by the extraction unit and the end extraction unit according to a plurality of difference time zones;
The information processing apparatus according to appendix 4, further comprising: a plurality of specifying units that specify a large number of functions classified into a time zone with a short difference.

(Appendix 6)
A weight part for weighting the number classified by time zone of each function extracted by the extraction unit, or the number classified by time zone of each function extracted by the end extraction unit,
The plurality of specific parts are:
The information processing apparatus according to claim 4, wherein after the weighting process is performed by the weight part, a large number of functions classified into a time zone having a short difference are specified.

(Appendix 7)
The plurality of specific parts are:
The information processing apparatus according to claim 6, wherein after the weighting process is performed by the weight part, the function having the largest number classified into the time zone having the shortest difference is identified.

(Appendix 8)
In an analysis method of an information processing apparatus having an arithmetic processing unit that executes a program and a storage unit,
The information processing apparatus is
Collecting function information that is information indicating a function being executed at a predetermined interval, and storing a sampling time that is a time when the function information and the function information are collected in the storage unit,
When the arithmetic processing unit shifts to the idle state, obtain the transition time and store in the storage unit,
A difference between the transition time stored in the storage unit and the sampling time immediately before the transition time among the sampling times stored in the storage unit, and an extraction unit that extracts function information corresponding to the sampling time;
An information processing apparatus analysis method for identifying a function that causes an idle based on a distribution of differences corresponding to each extracted function information.

(Appendix 9)
In an analysis program for an information processing apparatus having an arithmetic processing unit that executes a program and a storage unit,
In the information processing apparatus,
The function information which is information indicating the function being executed is collected at predetermined intervals, and the sampling time which is the time when the function information and the function information are collected is stored in the storage unit,
When the arithmetic processing unit shifts to an idle state, the transition time is acquired and stored in the storage unit,
The difference between the transition time stored in the storage unit and the sampling time immediately before the transition time among the sampling times stored in the storage unit, and the function information corresponding to the sampling time are extracted,
An analysis program for an information processing apparatus that identifies a function that causes an idle based on a distribution of differences corresponding to each extracted function information.

DESCRIPTION OF SYMBOLS 1 Information processing apparatus 1A Portable recording medium 1B Semiconductor memory 10A Reading part 11 CPU
12 RAM
13 Input unit 14 Display unit 15 Storage unit 15P Control program 16 Communication unit 18 Clock unit 151 OS
152 Profiler 153 User program 154 Sampling table 155 Transition time table 156 Difference table 157 Statistics table 158 End time table 159 End difference table 1510 Transition statistics table 1511 End statistics table 261 Function information collection unit 262 Migration acquisition unit 263 Extraction unit 264 Identification unit 265 Classification part 267 Function identification part

Claims (5)

An arithmetic processing unit for executing a program;
A function information collection unit that collects function information that is information indicating a function being executed at a predetermined interval, and stores a sampling time that is a time at which the function information and the function information are collected;
When the arithmetic processing unit transitions to an idle state, a transition acquisition unit that acquires a transition time and stores it in the storage unit;
A difference between the transition time stored in the storage unit and the sampling time immediately before the transition time among the sampling times stored in the storage unit, and an extraction unit that extracts function information corresponding to the sampling time;
An information processing apparatus comprising: a specifying unit that specifies a function that causes an idle based on a distribution of differences corresponding to each function information extracted by the extraction unit. The specific part is:
A classification unit that classifies the difference of each function according to a plurality of difference time zones;
The information processing apparatus according to claim 1, further comprising: a function specifying unit that specifies a function that causes the idle based on a number classified into a time zone in which the difference is short. The function specifying unit is
The information processing apparatus according to claim 2, wherein a function having the largest number classified into the shortest time zone is specified as the function causing the idle. In an analysis method of an information processing apparatus having an arithmetic processing unit that executes a program and a storage unit,
The information processing apparatus is
Collecting function information that is information indicating a function being executed at a predetermined interval, and storing a sampling time that is a time when the function information and the function information are collected in the storage unit,
When the arithmetic processing unit shifts to the idle state, obtain the transition time and store in the storage unit,
A difference between the transition time stored in the storage unit and the sampling time immediately before the transition time among the sampling times stored in the storage unit, and an extraction unit that extracts function information corresponding to the sampling time;
An information processing apparatus analysis method for identifying a function that causes an idle based on a distribution of differences corresponding to each extracted function information. In an analysis program for an information processing apparatus having an arithmetic processing unit that executes a program and a storage unit,
In the information processing apparatus,
The function information which is information indicating the function being executed is collected at predetermined intervals, and the sampling time which is the time when the function information and the function information are collected is stored in the storage unit,
When the arithmetic processing unit shifts to an idle state, the transition time is acquired and stored in the storage unit,
The difference between the transition time stored in the storage unit and the sampling time immediately before the transition time among the sampling times stored in the storage unit, and the function information corresponding to the sampling time are extracted,
An analysis program for an information processing apparatus that identifies a function that causes an idle based on a distribution of differences corresponding to each extracted function information.

Images