JPH0561720A - Trace data collector - Google Patents

Abstract

PURPOSE:To reduce the capacity of a trace memory and to improve work efficiency concerning the device for collecting trace data used for the evaluation test or analysis of a program. CONSTITUTION:This device is equipped with a collection part 3 to collect the trace data by simulating a program 2 to be tested, compression part 4 to compress the trace data from the collection part 3 so as to be restored, restorage part 5 to restore the compressed data to a state before the compression and to transfer the data to an external device 8, and memory area monitor part 7 to record the compressed data from the compression part 4 in a trace memory 6 while monitoring the remaining memory capacity of the trace memory 6 when there is a space in the memory, to interrupt the processing of the collection part 3 when there is no space and to restart the processing of the collection part 3 after restoring/transferring all the data in the trace memory 6 by activating the restorage part 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a trace data collecting device. In the development of computer programs,
Trace data is collected in order to improve the reliability and quality of the program by sufficiently performing evaluation tests and analysis of the program.
The trace data is data that lists the setting contents of registers and the addresses of the accessed memory at each operation step during program execution, and is sequentially stored in a predetermined area on the main memory (this is referred to as trace memory in this specification). Is stored.

However, in recent years, the amount of trace data to be collected has become enormous as the program becomes larger and more complex due to the diversification, high performance, and high quality of the program. The capacity of the trace memory may become insufficient, causing various problems. Therefore, this point must be improved.

[0003]

2. Description of the Related Art Trace data collected by simulating execution of a program under test is sequentially recorded and stored in a preset area (trace memory) on a main memory. The above method employs a method of recording the trace data in the trace memory as it is. Therefore, by increasing the capacity of the trace memory, an increase in the amount of trace data due to an increase in the scale of the program under test is dealt with.

Further, when recording the trace data in the trace memory, the following data compression is performed,
The amount of data recorded in the trace memory was reduced. That is, various data including the execution address, the access memory, the contents of the register, etc. are collected for the execution of one operation step, and if these series of data are continuously the same in one operation step, By counting the continuous number and recording the trace data and its number for one operation step in the trace memory,
The amount of data recorded in the trace memory was reduced.

However, there is a limit to the expansion of the capacity of the trace memory, while the above-mentioned data compression is effective only when the same trace data is frequently consecutive in each operation step, and it becomes complicated today. In such a program, there is almost no case where the trace data is continuous in each operation step, and the capacity of the trace memory may be insufficient.

Therefore, conventionally, the program under test is appropriately divided and processed, the trace data is referred to each time, or the trace data is collected several times while saving the data from the trace memory to an external storage device or the like. I was doing the work.

[0007]

However, dividing the program under test is complicated because the work of obtaining the address of the program under test and the work of transferring the contents of the trace memory to an external storage device are complicated. There was a problem such as missing data due to mistakes.

The present invention has been made in view of the above circumstances, and an object thereof is to reduce the capacity of a trace memory and to collect trace data efficiently and accurately. ..

[0009]

In order to solve the above problems, a trace data collecting apparatus as shown in FIG. 1 is provided.

That is, a collecting section 3 for executing the program under test 2 in a simulated manner and collecting trace data, and the collecting section 3
Compressor 4 that compresses trace data from
And a restoration unit 5 that restores the compressed data to the state before compression and transfers it to the external device 8, and the memory remaining amount of the trace memory 6 is monitored, and if there is a space, the compression from the compression unit 4 is performed. The traced data is recorded in the trace memory 6, and when there is no free space, the processing by the collection unit 3 is interrupted and the restoration unit 5 is activated to start the trace memory 6
A memory area monitoring unit 7 is provided for resuming the processing by the collecting unit 3 after restoring / transferring all the data therein.

When the trace data from the collecting unit 3 is continuous in 1-byte units, the compressing unit 4 counts the number of continuous trace data, and determines the number of continuous trace data corresponding to 1 byte. By outputting
It is configured to compress the trace data.

The external device 8 is a device for displaying or storing data, such as a display device, an external storage device, or a printer device.

[0013]

First, the collection unit 3 executes the program under test 2 in a simulated manner, and sequentially generates and collects trace data including the execution address of the program under test 2, the setting contents of the registers, the addresses of the accessed memory, etc. Pass to the compression unit 4.

In the compression unit 4, the trace data received from the collection unit 3 are sequentially compared byte by byte, and if the same 1-byte data (hereinafter referred to as byte data) continues, different byte data The number of continuous byte data is counted until the occurrence of, and the byte data and the number are passed to the memory area monitoring unit 7. When the same byte data is not continuous, the byte data is passed to the memory area monitoring unit 7 as it is.

The memory area monitoring unit 7 sequentially records and stores the byte data and the number data (compressed data) received from the compression unit 4 in the trace memory 6 until the remaining memory capacity is exhausted. The memory area monitoring unit 7 then collects data from the collection unit 3 when the remaining memory in the trace memory 6 is exhausted.
To the suspension unit 5 and to instruct the restoration unit 5 to start.

In response to this, the restoration unit 5 causes the trace memory 6
The compressed data is sequentially taken out of the compressed data, and the reverse processing of the compression unit 4 is performed to restore the same trace data as before the compression and output to the external device 8. When all the data in the trace memory 6 has been restored and output by the restoration unit 5, the memory area monitoring unit 7 instructs the collection unit 3 to restart the collection of trace data.

As described above, in the trace data collecting apparatus according to the present invention, since the trace data is compared and compressed in the compression unit 4 in units of 1 byte, the comparison and compression is performed in units of trace data for one operation step of the program as in the conventional case. It is more efficient because it can increase the data compression rate more than what it does.

Further, when the trace memory 6 becomes full, the collection of trace data is immediately interrupted, the data in the trace memory 6 is saved in the external device 8, and then the collection is continued. As described above, it is possible to omit the work or the like that has been performed when the trace memory is full, or when the trace memory is full, and it is possible to collect accurate trace data without data loss. Further, the capacity of the trace memory can be reduced.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIGS. 2 to 5 are flow charts showing an example of concrete processing of each section in FIG. The meaning of each symbol in these flowcharts is as follows.

TS_N: A buffer for storing the trace data collected by the collection unit TS_BUFF1: A buffer for storing the compressed trace data TS_BUFF2: A buffer for storing the counter data TS_BUFF: Both the compressed trace data and the counter data Named TS_N [MAX]: Number of bytes of TS_N n: Counter for reading TS_N byte by byte a: Counter for relative number of bytes c: Counter for counting the same data (bytes) af: a (number of relative bytes) Flag for determining whether or not is stored in the buffer TS_LST: The buffer that stores the previous 1 byte for comparison TS_By: The number of bytes in TS_BUFF TS_ALL: The number of bytes of the trace data currently stored in the trace memory T _MAX: maximum number of bytes of trace memory TS_MEM1: area for storing compressed data of trace memory TS_MEM2: area for storing counter data of trace memory In the following description, "additionally set" means that data is set before. In this case, the data is set in the form of being added to this, and when simply “set”, it means that the previous data is discarded and new data is set.

2 and 3 are flow charts showing the processing by the collection unit 3 and the compression unit 4 in FIG. First, refer to FIG. When the trace data collection device 1 is activated, the collection unit 3 first determines whether or not the collection of trace data is prohibited (step “abbreviated as ST” 1 below) 1 and simulates the program under test 2. Execute, collect trace data, and set TS_N sequentially (ST
2).

Next, the counters n, a, c and the flag af
Is cleared to 0 (ST3), TS_LST and TS_N
[N] is compared (ST4), and if they are equal, it is determined whether the counter af is "0" (S).
T5), if it is "0", the value of the counter a is TS
_BUFF2 is additionally set (ST6), the counter a is set to "0" (ST7), and the flag af is set to "1" (ST
8) Go to ST9. When the flag af is not "0" in ST5, the process directly proceeds to ST9.

In ST9, 1 is added to the counter c, then the value of the counter c is additionally set in TS_BUFF2 (ST10), and it is determined whether the counter c is equal to "0xFF" (ST11).

At ST4, TS_LST is TS_N.
If it is not equal to [n], or if the counter c is equal to “0xFF” in ST11, ST1 of FIG.
Go to 2.

In ST12, TS_N [n] is set to TS
_BUFF1 is additionally set, then 1 is added to the counter a (ST13), and the counter c and the flag af are cleared to 0 (ST14).

Next, it is determined whether or not the counter a is "0xFF" (ST15), and if they are equal, the value of the counter a is additionally set in TS_BUFF2 (ST1.
6), then “1 (counter value)” is TS_BUFF
2 is additionally set (ST17), and the value of the counter a is set to "1".
(ST18), and the process proceeds to ST19. In ST15, when the counter a is not equal to “0xFF”,
Go to ST19.

If the counter c is not equal to "0xFF" in ST11 of FIG. 2, the process proceeds to ST19 of FIG. In ST19, TS_N [n] is changed to TS_LS
Set to T, then add 1 to the counter n (ST
20), TS_N [MAX] is compared with the value of the counter n (ST21).

In ST21, if TS_N [MAX] is larger than the value of the counter n, the process returns to ST4 in FIG. 2, and if equal, TS_BUFF is sent to the memory area monitoring section 7 (ST22), and the program under test is tested. It is determined whether or not the collection of all trace data for 2 is completed (ST23), and if not completed, S of FIG.
Returning to T1, if completed, the processing by the collection unit 3 and the compression unit 4 is completed.

FIG. 4 is a flow chart showing the processing by the memory area monitoring unit 7 in FIG. Compressor 4 to T
When S_BUFF is passed, the memory area monitoring unit 7
First, the total number of bytes of TS_BUFF is set to TS_By (ST24), TS_By is added to TS_ALL (ST25), and TS_ALL and TS_MAX are compared (ST26).

In ST26, TS_ALL is TS_
TS_B if less than or equal to MAX
Add all the contents of UFF1 to TS_MEM1 (S
T27), then TS_BUFF2 contents TS_M
All the settings are additionally set in EM2 (ST28), and this processing is ended.

At ST26, TS_ALL becomes TS_
If it is larger than MAX, the collection unit 3 is notified that trace data collection is prohibited (ST29), and TS_BU
After saving the contents of FF to the temporary data save memory (ST
30), start the restoration unit 5, and wait until the processing (described later) by the restoration unit 5 ends (ST31).

When the processing by the restoration unit 5 is completed, the data saved in the data temporary save memory is stored in TS_ME.
Additional setting to M1 and TS_MEM2 (ST32, ST
33) Then, the collection unit 3 is notified of the release of the prohibition of the collection of trace data (ST34), and the processing by the memory area monitoring unit 7 ends.

FIG. 5 shows the processing of the restoration section 5 in FIG.
3 is a flowchart showing the processing of ST31 in FIG. When the restoration unit 5 is activated, it is confirmed whether the output destination of the trace data is the display device or the floppy disk device (FPD) (ST3).
5). For this confirmation, for example, the output destination of the trace data can be input in advance by the operator, or can be input each time as necessary.

A judgment is made based on this confirmation result (ST3
6) If the display device is designated, it is processed to output to the display device (ST37), and if the floppy disk device is designated, it is processed to output to the floppy disk device (ST38).

Then, the processes of ST40 to ST43 are repeated until there is no data set in TS_MEM1 (ST39). That is, 1-byte data is read from TS_MEM2 (ST40). This data is a relative number of bytes. Next, TS for this number of relative bytes
The data in _MEM1 is output (ST41), and TS_M
Read the next 1-byte data from EM2 (ST4
2). This data is a counter value. Next, TS_M
1 byte data at the local point of EM1 is converted to ST42
Output as many as the counter value (ST43), ST3
Return to 9.

In ST39, when the data of TS_MEM1 is lost (when all data is output), the processing by the restoration unit 5 is terminated. A specific example of the processing of each unit in FIG. 1 will be described in more detail with reference to the trace data compression transition diagram shown in FIG.

A1 in FIG. 6 shows a display example of trace data required for analysis of the program under test 2 and the like, where d1 is a clock, d2 is an execution address of the program under test, d3 is an assemble code, and d4 is. Is the content of the flag register, and d5 is the address of the accessed memory or the setting content of the register.

A2 is a list of the contents of A1 expressed in byte units, and shows the contents of data to be transferred from the restoration unit 5 to the external device 8 in FIG. A2 is data corresponding to A1 and 1: 1.

B1 indicates the contents of the data collected by the collection unit 3 in FIG. That is, the collection unit 3 collects data regarding the execution address (d2) of the program under test, the content of the flag register (d4), and the address of the accessed memory or the setting content of the register (d5), and passes the data to the compression unit 4. .. Here, the data is not collected for the clock (d1) and the assemble code (d3) because the clock (d1) is the assemble code (d3).
And the contents of the access memory or the register (d5) can be calculated backward, and the assemble code (d3) can be detected from the execution address of the program under test. Data is not passed to the compression unit 4. still,
The collecting unit 3 transfers the assemble code (d3) to the restoring unit 5 in response to a request from the restoring unit 5 (indicated by a dotted arrow in FIG. 1).

These data (B1) are compressed by the compression unit 4 into compressed data as shown by B2 and B3, and the memory area monitoring unit 7 causes the trace memory 6
(TS_MEM1, TS_MEM2) are sequentially set.

The decompression section 5 sequentially takes out the compressed data (B2, B3) from the trace memory 6 and starts from ST40 of FIG.
By repeating ST43, the data as shown in B1 is restored and the assemble code (d
3) is received from the collection unit 3, and the assemble code (d
The clock (d1) is calculated from 3) and the content (d5) of the access memory or the register to generate data as shown in A2, and the data is transferred to the external device 8.

The external device 8 converts the data (A2) transferred from the restoration unit 5 into data as shown in A1 and outputs it to the display device or the floppy disk device based on the instruction of the operator.

According to the present embodiment, the trace data is compared in byte units and compressed, and then the trace memory (T
S_MEM1 and TS_MEM2) are sequentially set. And the maximum number of bytes of trace memory (TS_MAX)
Trace memory (TS_MEM1, TS_
All the data in MEM2) is restored to the state before compression and automatically transferred to the display device or floppy disk device, after which the remaining trace data is collected, compressed, and set in the trace memory. ing.

Therefore, even when the amount of trace data to be collected exceeds the capacity of the trace memory, it is not necessary for the operator to perform special processing as in the conventional case, and trace data can be collected efficiently. Will be able to.

Although the restored data is transferred to the display device or the floppy disk device in the above embodiment, it may be transferred to the printer device.

[0046]

As described above in detail, the present invention compresses trace data in byte units and sequentially records it in the trace memory. When the trace memory becomes full, all the data in the trace memory is restored to the state before compression. Since it is restored and transferred to an external device, and then recording to the trace memory is continued, it is possible to collect a huge amount of trace data without increasing the capacity of the trace memory.
Since these processes are automatically performed, the number of man-hours required for collecting the trace data is reduced, and the trace data can be collected efficiently and accurately.

[Brief description of drawings]

FIG. 1 is a block diagram showing a principle configuration of the present invention.

FIG. 2 is a flowchart (1) showing processing of a collecting unit and a compressing unit in the embodiment of the present invention.

FIG. 3 is a flowchart (2) showing the processing of the collection unit and the compression unit in the embodiment of the present invention.

FIG. 4 is a flowchart showing processing of a memory area monitoring unit in the embodiment of the present invention.

FIG. 5 is a flowchart showing the processing of the restoration unit in the embodiment of the present invention.

FIG. 6 is a trace data compression transition diagram according to the embodiment of the present invention.

[Explanation of symbols]

 1 Trace Data Collection Device 2 Program Under Test 3 Collection Section 4 Compression Section 5 Restoration Section 6 Trace Memory 7 Memory Area Monitoring Section 8 External Device

Claims (2)

[Claims] 1. A collection unit (3) for simulating execution of a program under test (2) and collecting trace data, and a compression unit (4) for compressing trace data from the collection unit (3) in a recoverable manner. ), And restore the compressed data to the state before
The amount of memory remaining in the decompression unit (5) and the trace memory (6) to be transferred to (8) is monitored, and if there is free space, the compressed data from the compression unit (4) is transferred to the trace memory (6). ), The processing by the collection unit (3) is interrupted when there is no free space, and the restoration unit (5) is activated to restore / transfer all data in the trace memory (6). And a memory area monitoring unit (7) for restarting the processing by the collecting unit (3). 2. The trace data collecting device according to claim 1, wherein when the trace data from the collecting unit (3) continues in the compression unit (4) in 1-byte units, the continuous trace data is collected. The trace data collecting device is characterized in that the trace data is compressed by counting the number of the trace data and outputting the number continuous with the corresponding 1-byte trace data.