JP3838804B2 - Analysis device and recording medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an analysis apparatus and a recording medium, and in particular, an analysis apparatus that analyzes an operation state of an application program composed of a plurality of call units and a computer-readable recording program that causes a computer to execute such analysis processing. The present invention relates to a recording medium.
[0002]
[Prior art]
In recent years, with the advancement of hardware of information processing devices such as computers, software for operating hardware has become more complex and sophisticated, and the labor required for debugging and optimization has been increasing. .
[0003]
There are several load analysis tools that analyze the load on hardware and software components as a tool for supporting such software development. These can be broadly classified as follows: (1) Based on sampling There are two types: tools and (2) tools based on the calling sequence of calling units such as functions and objects.
[0004]
The former sampling-based tool generates an interrupt at a predetermined period during the execution of the application program to be analyzed, and acquires the address information of the application program being executed when the interrupt occurs. Then, by calculating the function being executed from the address information, it is displayed how often each function is executed.
[0005]
In the latter case, if the call unit is a function, for example, after the function is called, the process of that function is terminated and the function is returned to the other function that called it. By measuring the time until, the actual execution time of each function is displayed.
[0006]
[Problems to be solved by the invention]
However, in the former tool based on sampling, only information indicating the number of executions of each function can be obtained. For example, when the number of executions of a predetermined function is large (the load is high) There is a problem in that it is difficult to know whether it becomes high and it is difficult to perform optimization.
[0007]
In addition, the tool based on the latter function call sequence needs to generate an interrupt every time the function is called and execute processing for load analysis.-Since it becomes a head and causes an error, there is a problem that accurate measurement is difficult.
[0008]
The present invention has been made in view of these points, and provides an analysis apparatus that can smoothly perform optimization of application programs and the like by collecting necessary information and providing it to the user. For the purpose.
[0009]
[Means for Solving the Problems]
  In the present invention, in order to solve the above problems,In an analysis device for analyzing the operation status of an application program composed of a plurality of calling units, an interrupt setting means for setting a control unit so that an interrupt is generated at a predetermined cycle, and execution of the target application program is started And when the interrupt set by the interrupt setting means occurs during execution of the application program, all the calls that have passed through until the call unit being executed is called when the interrupt occurs Call information specifying a unit is acquired from a stack of a memory included in the control unit, and when there is the same call information as the already stored call information, the already stored call information is collected in the same pattern. If there is no new Call information acquisition means stored as a simple pattern, and when the operation of the application program ends, the cumulative number of calls for all the call units indicated by the call information stored by the call information acquisition means is accumulated. Aggregating as a count number, totaling means for totaling the number of calls of the call unit corresponding to the deepest nest shown in the call information as an actual count number, and the name of each call unit obtained by the aggregation means, Display means for displaying the cumulative count number and the actual count number in association with each other.The analysis apparatus characterized by this is provided.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing a configuration example of an embodiment of the present invention. As shown in this figure, the analysis apparatus 1 according to the present invention includes an interrupt setting means 1a, an application program execution means 1b, a recording device 1c, a call information acquisition means 1d, a counting means 1e, a display means 1g, an input means 1f, and The control unit 1h is formed. Also, a CRT (Cathode Ray Tube) monitor 2 is connected to the outside of the analysis apparatus 1.
[0012]
When executing the application program to be analyzed, the interrupt setting unit 1a sets the control unit 1h so that an interrupt is generated at a predetermined cycle.
The control unit 1h is configured by, for example, a CPU (Central Processing Unit), a memory, and the like, and actually executes an application program and controls each unit of the apparatus.
[0013]
The recording device 1c records an application program and an operating system (OS) to be analyzed, and reads and supplies predetermined information in response to a request from the application program execution unit 1b. Further, predetermined data or the like is recorded as necessary.
[0014]
The application program execution unit 1b reads an application program to be analyzed from the recording device 1c and starts its execution.
When the application program is being executed, the call information acquisition unit 1d is configured to execute the call of the control unit 1h when a call unit (for example, a function, an object, a procedure, or a subroutine) constituting the application program is called. Get call information stacked in internal memory.
[0015]
For example, when the call unit is a function, this call information includes the address information of all the functions passed from the main function up to the function being executed at the time of occurrence of the interrupt, and each function. It is constituted by a time stamp or the like indicating the time at the time when is called.
[0016]
The totaling unit 1e totals the information acquired by the call information acquisition unit 1d for each call pattern.
The input unit 1f receives information input from a keyboard or the like, for example. Information for specifying an application program to be analyzed and information for changing the display form of the CRT monitor 2 are input from the input means 1f.
[0017]
The display unit 1g causes the display device 2 to display the call information aggregated by the aggregation unit 1e in a predetermined display format based on the information input from the input unit 1f.
[0018]
Next, the operation of the above embodiment will be described.
Now, assuming that information for specifying an application program to be analyzed is input from the input unit 1f, the application program executing unit 1b acquires the corresponding application program from the recording device 1c.
[0019]
At this time, the interrupt setting unit 1a sets the control unit 1h so that an interrupt is generated at a predetermined cycle (for example, a cycle of 500 msec). As a result, when the execution of the application program is started, an interrupt is generated at a predetermined cycle.
[0020]
When an interrupt occurs, the call information acquisition unit 1d acquires call information stacked in a memory or the like inside the control unit 1h. Note that, as described above, the call information is address information of all functions that have passed from the main function until the function being executed is called when an interrupt occurs. For example, if the function “sub7” is currently being executed and the functions that have passed through the main function up to this function are “main”, “sub2”, “sub3”, and “sub5”, Addresses corresponding to all these functions are acquired as call information. At this time, a time stamp indicating the time when each function is called is also acquired.
[0021]
The counting means 1e classifies the acquired call information according to the pattern. That is, since the types of function call patterns are limited to some extent, the amount of information is reduced by classifying the call information for each pattern.
[0022]
For example, when the function currently being executed (the function in which the interrupt has occurred) is “sub5”, the functions that have passed through the main function up to this function are “main”, “sub2”, and “sub3”. In this case, “sub5” is not included as compared with the case of the call information described above, but an interrupt occurred before the processing of “sub5” (or after the processing of “sub5”). Therefore, it is determined that the pattern is the same as that described above.
[0023]
The above processing is repeatedly executed while the control unit 1h executes the application program.
When the execution of the application program by the control unit 1h is completed, the interrupt setting unit 1a stops the generation of the interrupt. Further, the counting unit 1e refers to the address information included in the call information classified for each pattern, and determines the cumulative count number that is the cumulative number of calls of each function and the last of each call information. And the actual count that is the number of times included in the.
[0024]
For example, it is assumed that functions included in certain call information are “main”, “sub2”, and “sub3”, and the cumulative counts of these functions are acm, ac2, and ac3, respectively, If the real count numbers of these functions are rcm, rc2, and rc3, respectively, the accumulated count number and the real count number for the above call information are acm = ac1 = ac2 = ac3 = 1, and rcm = rc1 = rc2 = 0 and rc3 = 1.
[0025]
The counting unit 1e calculates the cumulative count number and the actual count number of each function by repeating the above processing.
The display unit 1g obtains the cumulative count number and the actual count number for each function aggregated by the aggregation unit 1e, acquires the function name of the function from the recording device 1c, and associates them.
[0026]
When the association is completed, the display unit 1g displays each pattern on the CRT monitor 2 while indenting according to the depth of the nest. FIG. 2 is an example of a display screen displayed on the CRT monitor 2 as a result of such processing.
[0027]
In this display example, a window 20 titled “Analysis result display” is displayed, and the information aggregated by the aggregation means 1e is displayed in this window 20.
[0028]
The buttons 20a to 20c displayed on the upper right of the window 20 are operated when the window 20 is reduced, enlarged, or ended.
In the display area 20d, information aggregated by the aggregation unit 1e is displayed. For example, in the case of “<00> main [a0, b0]” displayed on the first line of this display example, the number “00” in <> indicates that the nesting depth of this function is “0”. ". The next “main” indicates a function name. Furthermore, variables a0 and b0 in [] indicate the cumulative count number and the actual count number, respectively.
[0029]
“<01> sub1 [a1, b1]” displayed on the second line has a nesting depth of “1”, a function name of “sub1”, and the cumulative count and actual count. The numbers are displayed as a1 and b1, respectively. Since the head of each line is indented according to the nesting depth, it is easy to grasp the correspondence between the functions when the application program is structured.
[0030]
Further, by appropriately operating the slider 20e or the buttons 20f and 20g displayed on the right end of the window 20, the content displayed in the display area 20d is scrolled, so that the user can freely set a desired range. It is possible to select and refer to.
[0031]
Next, an example of processing executed in the embodiment shown in FIG. 1 in order to realize the above processing will be described with reference to FIG. When the flowchart shown in FIG. 3 is started, the following processing is executed.
[S1] The application program name to be analyzed is input from the input means 1f.
[S2] When information instructing to change the interrupt cycle is input from the input unit 1f, the setting of the interrupt cycle is changed.
[0032]
If the interrupt cycle setting is not changed, the interrupt cycle is automatically set to a default value (for example, 500 ms).
[S3] It is determined whether or not information for instructing execution of the application program has been input from the input unit 1f. If input, the process proceeds to step S4. Otherwise, the process returns to step S3 to perform the same processing. repeat.
[S4] The application program execution unit 1b acquires the specified application program from the recording device 1c, supplies it to the control unit 1h, and starts its execution.
[S5] If an interrupt has occurred, the process proceeds to step S6. Otherwise, the process proceeds to step S7.
[S6] An interrupt process is executed. Details of this process will be described later with reference to FIG.
[S7] It is determined whether or not the execution of the application program is finished. If finished, the process proceeds to step S8. Otherwise, the process returns to step S5.
[S8] The counting unit 1e executes a counting process for counting the call information acquired by the interrupt process.
[0033]
Details of this processing will be described later with reference to FIG.
[S9] The display unit 1g supplies the information aggregated by the aggregation unit 1e to the CRT monitor 2 for display.
[0034]
Next, the details of the “interrupt processing” shown in FIG. 3 will be described with reference to FIG. When the flowchart shown in FIG. 4 is started, the following processing is executed.
[S20] The call information acquisition unit 1d acquires the call information stacked in the memory as the return destination information of the function when the control unit 1h executes the application program at that time (when the interrupt occurs). .
[S21] The call information acquisition unit 1d searches the call information already acquired for information having the same pattern as the call information acquired in step S20.
[S22] As a result of the search in step S21, if call information having the same pattern already exists, the process proceeds to step S23. Otherwise, the process proceeds to step S24.
[S23] The call information acquired in step S20 is stored together with other already stored same patterns. Then, the process returns to the original process.
[S24] The call information acquired in step S20 is stored as a new pattern. Then, the process returns to the original process.
[0035]
Next, details of the “aggregation process” shown in FIG. 3 will be described with reference to FIG. When the flowchart shown in FIG. 5 is started, the following processing is executed.
[S40] The tabulating unit 1e obtains the function name and its address constituting the application program from the control unit 1h.
[S41] For each function acquired in step S40, a variable ax for storing the cumulative count number and a variable bx for storing the actual count number are prepared.
[S42] Initially set the values of variables i and j for counting the number of processings to "1".
[S43] The i-th pattern is acquired. For example, when i = 1, all functions belonging to the first pattern are acquired.
[S44] Variables (variables ax and bx included in the i-th pattern) corresponding to each function acquired in step S43 are prepared.
[S45] The counting unit 1e acquires the j-th call information from the call information acquisition unit 1d.
[S46] All the functions included in the call information acquired in step S45 are acquired.
[S47] The values of the variables ax corresponding to all the functions acquired in step S46 are incremented by 1.
[S48] The function (the deepest nested function) included at the end of the call information acquired in step S45 is acquired.
[S49] The value of the variable bx corresponding to the function acquired in step S48 is incremented by one.
[S50] In the i-th pattern, it is determined whether or not there is unprocessed call information. If there is unprocessed call information, the process proceeds to step S51; otherwise, the process proceeds to step S52.
[S51] The value of the variable j is incremented by 1, and the process returns to step S45.
[S52] It is determined whether or not there is an unprocessed pattern. If there is an unprocessed pattern, the process proceeds to step S53. Otherwise, the process returns (returns) to the original process.
[S53] The value of the variable i is incremented by 1, and the process proceeds to step S54.
[S54] The value of variable j is set to "1", and the process returns to step S43.
[0036]
According to the above processing, when the application program to be analyzed is executed, the call information is collected by generating an interrupt at a predetermined cycle, and after the execution of the application program is finished, each function is obtained from the call information. Can be calculated and output to the CRT monitor 2.
[0037]
As described above, according to the present invention, while executing an application program to be analyzed, an interrupt is generated at a predetermined cycle, and when an interrupt occurs, the main function is followed by a function being executed at that time. Call information composed of function addresses is acquired, classified and stored for each pattern. When the execution of the application program is completed, the cumulative count number and the actual count number of each function are calculated and displayed in a predetermined display format, so that detailed information is available compared to conventional sampling-based tools. Can be obtained. That is, in the case of a conventional sampling-based tool, the information acquired when an interrupt occurs is only information related to the function being executed at that time (for example, the address of the function). Since not only the function being executed at the time but also information (stack information) about all functions from the main function to that function is acquired, it is possible to provide detailed information compared to the conventional case it can.
[0038]
In addition, in the conventional tool based on the function call sequence, it is necessary to generate an interrupt each time the function is called and execute processing for load analysis. However, in the case of the present invention, by appropriately setting the interrupt generation period, instead of sacrificing the detail level of information, the occurrence of errors can be suppressed or the detail level of information can be improved. It is possible to make settings according to the user's purpose, such as sacrificing the occurrence of errors.
[0039]
Furthermore, since the function call information is classified and displayed for each pattern, when the load of a certain function is high, it is possible to specify what kind of call pattern the load will be high. The application program can be optimized smoothly.
[0040]
Next, a display example when the display format for the CRT monitor 2 is changed will be described with reference to FIG.
For example, it is assumed that information indicating that the nest depth displayed on the CRT monitor 2 is limited to 0 to 2 is input from the input unit 1f. In this case, the display unit 1g does not output information about a function having a nesting depth of 3 or more when the information obtained by the totaling unit 1e is output to the CRT monitor 2. As a result, the screen shown in FIG. 6 is displayed on the CRT monitor 2.
[0041]
In the display example shown in FIG. 6, compared to the case of FIG. 2, information corresponding to three or more nestings is not displayed in the information displayed in the display area 20d. For example, the function “sub3” displayed in the fourth line of FIG. 2 is not displayed in FIG. Further, at the end of the previous line, a mark “*” is added, indicating that the subsequent function is omitted.
[0042]
In such a display screen, for example, if the mark “*” displayed at the end of the sixth line is designated by the cursor 25, a dialog box 30 is additionally displayed on the CRT monitor 2 as shown in FIG. It is possible to refer to information that is displayed and omitted. In this example, a dialog box 30 entitled “Detailed display” is displayed, and information to be displayed next to the mark “*” specified in the window 20 is displayed in the display area 30b. Yes.
[0043]
By adopting such a display format, when the number of functions included in the application program is large, only information with shallow nesting is displayed, and information that is omitted only when necessary is displayed as appropriate. Therefore, the work can be performed efficiently.
[0044]
It should be noted that the information to be displayed is not limited by the depth of nesting, and for example, only the functions whose cumulative count number or actual count number exceeds a predetermined value may be displayed. By doing so, it is possible to display on the screen only functions that are frequently called, so that only main information can be referred to.
[0045]
In the above display example, the analysis result is displayed on the screen as text information, but may be displayed in a chart format as shown in FIG.
In the display example shown in FIG. 8, a window 40 entitled “Time Series Display” is displayed. On the upper right of the window 40, buttons 40a to 40c that are operated when the window 40 is reduced, enlarged, or terminated are displayed.
[0046]
The horizontal axis of the display area 40d indicates the elapsed time since the execution of the application program is started, and the vertical axis indicates each function constituting the application program. Moreover, the arrow shown in the display area 40d shows the period during which the function was executed.
[0047]
The slider 40e and the buttons 40f and 40g provided on the right side of the display area 40d are operated when the content displayed in the display area 40d is scrolled up and down. The slider 40h and the buttons 40i and 40j provided below the display area 40d are operated when the content displayed in the display area 40d is scrolled in the left-right direction.
[0048]
By the way, when the display as shown in FIG. 8 is performed, information on the time when each function is called is also required. This is because the time stamp acquired by the call information acquisition unit 1d described above is used. Is possible.
[0049]
That is, since the time stamp is associated with and stored in the address information of each function, the acquired address information is rearranged in chronological order, and an arrow is displayed at a position corresponding to the address information. By doing so, the diagram shown in FIG. 8 can be obtained.
[0050]
On the screen shown in FIG. 8, when a predetermined arrow is designated by the cursor 25 as shown in FIG. 9, a dialog box in which detailed information corresponding to that part is described is displayed. In this example, as a result of the center portion of the first arrow of the function “sub1” being designated by the cursor 25, a dialog box 50 entitled “Detailed information (sub1)” is displayed, and a call to the function “sub1” is performed. A list of other functions that performed is displayed. That is, before the function “sub1” is executed, the function “subD” is first called, and then the functions “subC”, “subB”, and “subA” are sequentially called, and then the function “sub1”. Is called.
[0051]
As described above, according to the display forms shown in FIGS. 8 and 9, since the acquired call information is displayed in a chart format, it is acquired in a display format that makes it easy to understand time-series operations. Information can be displayed. In the embodiment of the present invention, since the call information is acquired every time an interrupt occurs, information may be lost when the interrupt cycle is long, but the same information is used in the program. Since calling by pattern is often repeated, the information of the missing part can often be supplemented by other parts, and the influence caused by the missing information is small.
[0052]
In the above embodiment, the case where a single application program is executed has been described as an example. However, the control unit 1h has a plurality of CPUs, and a plurality of application programs are executed in parallel. It is possible to apply the present invention to the case where it is done. Note that the present invention can also be applied to a case where one application program is shared by a plurality of processors and executed.
[0053]
When the control unit 1h has a plurality of CPUs, an interrupt is generated for each CPU, and the call information acquisition unit 1d acquires the call information from each CPU. The acquired call information is aggregated for each CPU by the aggregation means 1e and then supplied to the display means 1g.
[0054]
The display means 1g displays the information on the plurality of CPUs aggregated by the aggregation means 1e side by side on the CRT monitor 2. The screen displayed as a result is shown in FIG. In this display example, a window 60 titled “Analysis result display” is displayed. On the upper right of the window 60, buttons 60a to 60c that are operated when the window 60 is reduced, enlarged, or closed are displayed. In the display area 60d, information regarding the process 0 (in this example, the application program 0) executed by the first CPU is displayed. In the display area 60e, information related to the process 1 executed in the second CPU is displayed.
[0055]
In this display example, only the information related to the process 0 and the process 1 is displayed. However, the slider 60i or the buttons 60j and 60k displayed at the lower part of the window 60 are appropriately operated to scroll the screen in the horizontal direction. This makes it possible to refer to information regarding a desired process.
[0056]
In addition, by appropriately operating the slider 60f or the buttons 60g and 60h and scrolling the screen in the vertical direction, it is possible to refer to a predetermined range of displayed information.
[0057]
When a predetermined item is selected by the cursor 25 on such a display screen, a dialog box 80 as shown in FIG. 11 is displayed, and each CPU in the case where the designated item is being executed is displayed. Information about the load distribution will be displayed.
[0058]
In this display example, the CPU number is displayed in the display area 80b, and the load factor of each CPU is indicated in the display area 80c by a bar graph centering on the average value of 70%. . Further, a deviation based on an average value of 70% is displayed in the display area 80d.
[0059]
In the above display example, the load distribution is shown by a bar graph. However, as shown in FIG. 12, it may be displayed by a numerical value. In this display example, a dialog box 90 is displayed corresponding to the designation of the display item by the cursor 25. The display area 90b of the dialog box 90 displays the CPU number, and the display area 90c displays the load factor as a numerical value. Furthermore, the number of samples in each CPU is displayed in the display area 90d adjacent to the right.
[0060]
Next, with reference to FIG. 13, a display example in a case where call information acquired from each CPU is displayed in a chart format will be described.
In this display example, a window 100 entitled “Time Series Display” is displayed. In the upper right portion of the window 100, buttons 100a to 100c that are operated when the window is reduced, enlarged, or closed are displayed.
[0061]
In the display area 100d, information related to the process 0 executed by the first CPU is displayed. In addition, information regarding the process 1 executed by the second CPU is displayed in the display area 100e.
[0062]
In this display example, only information related to the processes executed by the first and second CPUs is displayed. However, by appropriately operating the button 100f or the button 100g to scroll the screen, information related to a predetermined process is referred to. It becomes possible to do. Further, by operating the slider 100h or the buttons 100i and 100j, the range of information displayed in the display area 100d can be selected. Similarly, the range of information displayed in the display area 100e can be selected by operating the slider 100k or the buttons 100l and 100m. Furthermore, by operating the slider 100n or the buttons 100o and 100p, it is possible to select a range of information displayed in the display area 100d and the display area 100e.
[0063]
In such a display example, when a predetermined arrow is designated, detailed information may be displayed as shown in FIG.
As described above, even when processes are executed in parallel in a plurality of CPUs, the present invention can be applied, and overhead can be reduced as compared with conventional tools. Therefore, it is possible to perform an analysis with a small error.
[0064]
The above processing functions can be realized by a computer. In this case, the processing contents of the functions that the analysis device should have are described in a program recorded on a computer-readable recording medium, and the above processing is realized by the computer by executing this program on the computer. The Examples of the computer-readable recording medium include a magnetic recording device and a semiconductor memory.
[0065]
When distributing to the market, store the program on a portable recording medium such as a CD-ROM (Compact Disk Read Only Memory) or floppy disk, or store it in a computer storage device connected via a network. In addition, it can be transferred to another computer through the network. When executed by a computer, the program may be stored in a hard disk device or the like in the computer, loaded into the main memory, and executed.
[0066]
【The invention's effect】
  As described above, in the present invention,Since the function call information is classified and displayed for each pattern, it is possible to specify the call pattern when the load of a certain function is high. Can be smoothly performed.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration example of an embodiment of the present invention.
FIG. 2 is a display example of an analysis result display screen displayed on the CRT monitor of the embodiment shown in FIG. 1;
FIG. 3 is a flowchart for explaining an example of processing executed in the embodiment shown in FIG. 1;
4 is a flowchart illustrating details of “interrupt processing” shown in FIG. 3;
FIG. 5 is a flowchart for explaining the details of “aggregation processing” shown in FIG. 3;
FIG. 6 is another display example of the analysis result display screen displayed on the CRT monitor of the embodiment shown in FIG. 1;
7 is a display example of a dialog box displayed when a mark “*” is selected on the display screen shown in FIG.
FIG. 8 is still another display example of the analysis result display screen displayed on the CRT monitor of the embodiment shown in FIG. 1;
FIG. 9 is a display example of a dialog box displayed when a predetermined arrow is designated on the display screen shown in FIG. 8;
FIG. 10 is a display example of an analysis result display screen when the analysis target is a parallel program.
11 is a display example of a dialog box displayed when a predetermined item is selected on the screen shown in FIG.
12 is another display example of the dialog box displayed when a predetermined item is selected on the screen shown in FIG.
FIG. 13 is another display example of the analysis result display screen when the analysis target is a parallel program.
[Explanation of symbols]
1 Analysis device
1a Interrupt setting means
1b Application program execution means
1c Recording device
1d call information acquisition means
1e Aggregation means
1f Input means
1g Display means
2 CRT monitor

Claims (6)

In an analysis device that analyzes the operation status of an application program composed of multiple call units,
Interrupt setting means for setting the control unit to generate an interrupt at a predetermined cycle;
And application program execution means to start execution of the application program in question,
When an interrupt set by the interrupt setting means occurs during the execution of the application program, call information that identifies all the call units that have passed through until the call unit being executed is called when the interrupt occurs. When the same information as the call information already stored is obtained from the memory stack provided in the control unit, the call information already stored is collectively stored in the same pattern, and does not exist Call information acquisition means for storing as a new pattern ,
When the operation of the application program is completed , the cumulative number of calls for all the call units indicated by the call information stored by the call information acquisition unit is totaled as a cumulative count number, and is indicated by the call information. A counting means for counting the number of calls of the call unit corresponding to the deepest nested as a real count;
Display means for associating and displaying the name of each call unit obtained by the counting means, the cumulative count number, and the actual count number;
Analyzer characterized in that it comprises a.   The display means performs indentation according to the nesting depth of each call unit, and displays the name of each call unit, the cumulative count number, and the actual count number in association with each other. The analysis device according to claim 1.   The analysis apparatus according to claim 2, wherein the display unit displays only information related to the calling unit shallower than the predetermined nest.   3. The analysis apparatus according to claim 2, wherein the display unit displays only information related to the calling unit in which the accumulated count number or the actual count number is greater than a predetermined value. It further has an input means for inputting predetermined information,
5. The display unit displays information corresponding to the designated part when there is an input from the input part to designate a part for which the display is omitted. The analysis device described. In a computer-readable recording medium in which a program for analyzing an operation state of an application program composed of a plurality of calling units is recorded,
Computer
Interrupt setting means for setting the control unit to generate an interrupt at a predetermined cycle;
When an interrupt set by the interrupt setting means occurs during execution of the target application program , when the interrupt occurs, all the call units that have passed through until the call unit being executed is called are specified. Call information is acquired from a stack of memory provided in the control unit, and if the same information as the call information that has already been stored, the call information that has already been stored is collectively stored in the same pattern, Call information acquisition means for storing as a new pattern if it does not exist,
When the operation of the application program is completed, the cumulative number of calls for all the call units indicated by the call information stored by the call information acquisition unit is totaled as a cumulative count number, and is indicated by the call information. A counting means for counting the number of calls of the call unit corresponding to the deepest nested as a real count number;
Display means for displaying the name of each call unit obtained by the counting means, the cumulative count number, and the actual count number in association with each other;
Computer readable recording medium recording a program to function as a.

Images