JP5606261B2 - Debug system and method of acquiring trace data of debug system - Google Patents

Description

  The present invention relates to, for example, a trace data acquisition device that acquires trace data of LSI (Large Scale Integration), a trace data acquisition method of the trace data acquisition device, and a debug system.

Due to the large scale of LSIs, LSIs mounted on a plurality of chips are now mounted on a single chip. If no special device is applied, information inside the LSI cannot be acquired. Therefore, a method for acquiring information inside the LSI is necessary.
As one of the methods, after tracing the signal inside the LSI and saving the trace data in the trace memory inside the LSI, the operation of the LSI is stopped and the trace data is extracted from the trace memory to an external device or the like. .
However, since the capacity of the trace memory is limited, it is not possible to obtain all trace data of the program under test without interruption. Therefore, the program under test is divided, and the trace data is acquired for each divided program under test.

  The prior art of Patent Document 1 monitors the free space in the trace memory in order to automatically perform a series of operations. When the trace memory is full, the LSI operation is stopped and the trace data is saved from the trace memory. Then, after all the trace data in the trace memory is saved, the operation of the LSI is resumed and the acquisition of the trace data is resumed.

  In addition, there is a method of compressing trace data by a compression circuit in order to utilize a limited capacity of the trace memory. As one of the techniques, the conventional technique of Patent Document 2 compresses trace data by a plurality of compression circuits, and selects data compressed by a compression circuit having the highest compression ratio among them.

JP-A-5-61720 JP 2009-087343 A

  The prior art of Patent Document 1 is inefficient because it is necessary to stop the operation of the LSI and save the trace data each time the trace memory runs out of space. Further, when the program under test is divided and executed, it is impossible to detect a problem that cannot be found unless the program under test is executed without being divided.

  Even if the trace data is efficiently compressed by the conventional method of Patent Document 2, the capacity of the trace memory for storing the trace data is limited, and thus the problem of Patent Document 1 cannot be solved.

  An object of the present invention is to make it possible to efficiently acquire necessary trace data, for example, without stopping execution of an LSI and without dividing a program under test.

The trace data acquisition apparatus of the present invention is
A trace data input unit for inputting each of a plurality of data of different types, each of which has a priority determined for each type, as trace data;
A trace data storage unit for storing trace data;
Based on the free capacity of the trace data storage unit, a trace data type selection unit that selects one or more types in descending order of priority as the type of trace data stored in the trace data storage unit;
When the trace data input unit inputs trace data, it is determined whether or not the input data input as the trace data to the trace data input unit is the selected type of trace data selected by the trace data type selection unit A trace data type determination unit for
The trace data storage unit stores the input data when the trace data type determination unit determines that the input data is the selected type of trace data.

  According to the present invention, for example, necessary trace data can be efficiently acquired without stopping execution of the LSI and without dividing the program under test.

1 is a configuration diagram of a debugging system 200 according to Embodiment 1. FIG. 1 is a functional configuration diagram of an LSI trace data acquisition apparatus 100 according to Embodiment 1. FIG. 3 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the first embodiment. 5 is a flowchart showing trace data input / output status analysis processing of a trace data input / output status analysis unit 140 according to the first embodiment. FIG. 6 is a schematic diagram showing an example of a trace data type selection process (S230) of the trace data input / output situation analysis unit 140 according to the first embodiment. 6 is a flowchart showing an example of a trace data type selection process (S230) of the trace data input / output situation analysis unit 140 according to the first embodiment. FIG. 4 is a functional configuration diagram of an LSI trace data acquisition apparatus 100 according to a second embodiment. 9 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the second embodiment. FIG. 10 is a functional configuration diagram of an LSI trace data acquisition apparatus 100 according to a third embodiment. 10 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the third embodiment. FIG. 10 is a functional configuration diagram of an LSI trace data acquisition apparatus 100 according to a fourth embodiment. 10 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the fourth embodiment. FIG. 10 is a functional configuration diagram of an LSI trace data acquisition apparatus 100 according to a fifth embodiment. 10 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the fifth embodiment. 10 is a flowchart showing trace data input / output status analysis processing of a trace data input / output status analysis unit 140 according to the fifth embodiment. 18 is a flowchart showing an example of a trace data type selection process (S230B) of the trace data input / output situation analysis unit 140 according to the fifth embodiment. FIG. 10 is a functional configuration diagram of an LSI trace data acquisition apparatus 100 according to a sixth embodiment. FIG. 10 is a configuration diagram of a debugging system 200 according to a seventh embodiment. 18 is a flowchart showing a trace data storage method of the debugging device 220 according to the seventh embodiment.

Embodiment 1 FIG.
A description will be given of a mode in which the type of trace data to be acquired is selected according to the free capacity of the buffer in the LSI (Large Scale Integration), and the selected type of trace data is acquired.
However, the LSI is an example of a target device that acquires trace data. For example, the trace data may be acquired inside or outside the computer for a computer such as a personal computer or a server device instead of the LSI.

FIG. 1 is a configuration diagram of a debugging system 200 according to the first embodiment.
The configuration of the debug system 200 will be described with reference to FIG.

  The debug system 200 includes an LSI 210 to be debugged and a debug device 220 for debugging a circuit of the LSI 210 or a program executed on the LSI 210 based on the trace data 101 of the LSI 210.

The LSI 210 includes a CPU 219 (Central Processing Unit), an internal bus 212, an external output interface 213, an internal memory 214, and the LSI trace data acquisition apparatus 100.
The LSI 210 also includes a plurality of trace signal generation sources 211. The trace signal generation source 211 is hardware (for example, a circuit) or software that generates (input / output) a trace signal representing data to be traced (trace data). The CPU 219, the internal bus 212, the external output interface 213, and the internal memory 214 may be used as the trace signal generation source 211.

The LSI trace data acquisition device 100 inputs a trace signal generated by each trace signal generation source 211 via a signal line, acquires the trace data 101 from the input trace signal, and transmits the acquired trace data 101 to the internal bus 212. The data is output to the debug device 220 via the external output interface 213.
However, if the trace data 101 cannot be output to the debug device 220 due to a failure of the external output interface 213, the LSI trace data acquisition device 100 outputs the trace data 101 to the internal memory 214 of the LSI 210.

  Hereinafter, the trace signal and the trace data are collectively referred to as “trace data 101”.

The trace data 101 includes a plurality of types such as a command signal indicating processing contents (instructions), an address signal indicating an address of a storage area in which processing target data and the like are stored, and a data signal indicating processing target data. .
In the first embodiment, the user sets a priority (importance) for each type of trace data 101 in advance, and sets information related to the determined priority (selection method information 103A described later) in the LSI trace data acquisition apparatus 100. Shall be kept.

FIG. 2 is a functional configuration diagram of the LSI trace data acquisition apparatus 100 according to the first embodiment.
A functional configuration of the LSI trace data acquisition apparatus 100 according to the first embodiment will be described with reference to FIG.

  The LSI trace data acquisition apparatus 100 (an example of a trace data acquisition apparatus) includes a trace data selection / selection unit 110, a trace data buffer 120, and a trace data input / output situation analysis unit 140.

The trace data input / output status analysis unit 140 (an example of the trace data type selection unit) sets the priority of the type of the trace data 101 stored in the trace data buffer 120 based on the free capacity 102 of the trace data buffer 120 as the selection type. Select one or more in descending order.
Hereinafter, the information indicating the selection type is referred to as “selection type information 103B”.

The trace data selection unit 110 (an example of a trace data input unit and a trace data type determination unit) inputs a plurality of pieces of data having different types and priorities determined for each type as the trace data 101. To do.
When the trace data selection unit 110 inputs the trace data 101, the trace data selection unit 110 determines whether the input trace data (input data) is the selected type of trace data 101 selected by the trace data input / output status analysis unit 140. To do.

  The trace data buffer 120 (an example of the trace data storage unit) stores the input data when the trace data sorting / selection unit 110 determines that the input data is the selected type of trace data 101.

For example, the trace data input / output status analysis unit 140 selects a selection type as follows.
The trace data input / output status analysis unit 140 selects more types of trace data 101 as selection types as the free capacity 102 of the trace data buffer 120 is larger. In addition, the trace data input / output situation analysis unit 140 selects the type of trace data 101 that is smaller as the free space 102 of the trace data buffer 120 is smaller as the selection type.
The trace data input / output status analysis unit 140 compares the free capacity 102 of the trace data buffer 120 with a predetermined capacity threshold. When the free capacity 102 of the trace data buffer 120 is larger than the capacity threshold value, the trace data input / output status analysis unit 140 selects the selection type of the trace data 101 by a predetermined first selection type number. When the free capacity 102 of the trace data buffer 120 is smaller than the capacity threshold, the trace data input / output status analysis unit 140 sets the selection type of the trace data 101 to a predetermined second selection type number less than the first selection type number. Just choose.

  The LSI trace data acquisition apparatus 100 further includes a trace data sequential output unit 130 and a trace data control register 141.

  The trace data sequential output unit 130 sequentially outputs the trace data 101 stored in the trace data buffer 120 to the debug device 220 or the internal memory 214.

The trace data control register 141 is a storage unit in which selection method information 103A (or an address of a storage area that stores the selection method information 103A) is set.
The selection method information 103A is information indicating a selection condition for causing the trace data input / output status analysis unit 140 to select the type of the trace data 101.

The configuration described as “˜unit” of the LSI trace data acquisition apparatus 100 may be “˜circuit”, “˜device”, “˜device”, and “˜step”, “˜procedure”, “˜processing”. It may be. In other words, the configuration described as “˜unit” may be implemented by any of firmware, software, hardware, or a combination thereof.
For example, the LSI trace data acquisition apparatus 100 includes a CPU and a memory. A program that executes a function described as “˜unit” is stored in a memory, and is read and executed by the CPU. That is, the program causes the computer to function as “to part”, and causes the computer to execute the procedures and methods of “to part”.

FIG. 3 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the first embodiment.
A processing flow of the trace data acquisition method according to the first embodiment will be described with reference to FIG.

In S110, the trace data sorting / selection unit 110 receives the trace data 101 output from the specific trace signal generation source 211.
It progresses to S120 after S110.

In S120, it is assumed that the trace data sorting selection unit 110 acquires the selection type information 103B from the trace data input / output status analysis unit 140 periodically or at a predetermined timing.
The selection type information 103B is information indicating one or more types of the trace data 101.
Hereinafter, the type of the trace data 101 indicated in the selection type information 103B is referred to as “selection type”.

The trace data sorting / selection unit 110 refers to the selection type information 103 </ b> B acquired last time from the trace data input / output status analysis unit 140.
Then, the trace data sorting selection unit 110 determines whether or not the trace data 101 input in S110 is the same type of trace data 101 as the selection type indicated in the selection type information 103B.

For example, the data type of the trace data 101 is set at a predetermined data position of the trace data 101.
In this case, the trace data sorting / selection unit 110 compares the data type set in the trace data 101 with the selection type indicated in the selection type information 103B to make a determination.

When the trace data 101 input in S110 is the same type of trace data 101 as the selection type indicated in the selection type information 103B (YES), the process proceeds to S130.
When the trace data 101 input in S110 is not the same type of trace data 101 as the selection type indicated in the selection type information 103B (NO), the trace data selection unit 110 discards the trace data 101 input in S110, and the process is Return to S110.

In S <b> 130, the trace data selection unit 110 stores the trace data 101 input in S <b> 110 in the trace data buffer 120.
It progresses to S140 after S130.

In S <b> 140, the trace data sequential output unit 130 sequentially acquires the trace data 101 previously stored from the trace data buffer 120.
Then, the trace data sequential output unit 130 outputs the acquired trace data 101 to the debug device 220 via the internal bus 212 and the external output interface 213 of the LSI 210. However, when the trace data 101 cannot be output to the debug device 220 due to a failure of the external output interface 213, the trace data sequential output unit 130 saves the trace data 101 in the internal memory 214 of the LSI 210.
After S140, the process returns to S110.

  The trace data sequential output unit 130 sequentially outputs the trace data 101 in S140, so that the trace data 101 can be output to the debug device 220 without stopping the operation of the LSI 210.

  However, since the output of the trace data 101 from the trace data sequential output unit 130 to the debug device 220 is performed via the external output interface 213, the output speed of the trace data 101 output from the trace data sequential output unit 130 is not fast.

Therefore, the output amount per unit time of the trace data 101 that can be output from the trace data sequential output unit 130 is usually larger than the input amount (generation amount) of the trace data 101 per unit time that is input to the trace data selection unit 110. There are few.
For this reason, if all the trace data 101 is stored in the trace data buffer 120, the free space of the trace data buffer 120 is lost, and important trace data 101 cannot be stored.

Therefore, the trace data input / output status analysis unit 140 selects the type of the trace data 101 to be stored in the trace data buffer 120 based on the free capacity 102 of the trace data buffer 120.
Then, the trace data selection / selection unit 110 stores only the selected type of trace data 101 selected by the trace data input / output situation analysis unit 140 in the trace data buffer 120 (S120, S130).

  Next, the processing of the trace data input / output status analysis unit 140 will be described.

FIG. 4 is a flowchart showing trace data input / output status analysis processing of the trace data input / output status analysis unit 140 according to the first embodiment.
The trace data input / output status analysis processing of the trace data input / output status analysis unit 140 in the first embodiment will be described with reference to FIG.

In S210, the trace data input / output status analysis unit 140 acquires the selection method information 103A from the trace data control register 141 (or the storage area specified by the address indicated in the trace data control register 141). The trace data input / output situation analysis unit 140 may acquire the selection method information 103A periodically or at a predetermined timing.
The selection method information 103A is information indicating a selection condition for selecting the type of the trace data 101. It is assumed that the selection method information 103A is preset in the memory of the trace data control register 141 or the LSI trace data acquisition apparatus 100 by the user.

For example, the selection method information 103A indicates a capacity threshold value to be compared with the free capacity 102 of the trace data buffer 120. Furthermore, the selection method information 103A indicates the type of trace data 101 to be selected in association with the magnitude relationship between the free capacity 102 of the trace data buffer 120 and the capacity threshold.
Further, for example, the selection method information 103 </ b> A indicates a priority for each type of trace data 101. Further, the selection method information 103A indicates the number of types of trace data 101 to be selected in association with the magnitude relationship between the free capacity 102 of the trace data buffer 120 and the capacity threshold.

  It progresses to S220 after S210.

In S <b> 220, the trace data input / output status analysis unit 140 acquires information indicating the free capacity of the trace data buffer 120 (free capacity 102). Information indicating the free capacity of the trace data buffer 120 can be obtained from the buffer management function of the trace data buffer 120.
The information indicating the free space means the free space size, the free space ratio, and the like. In the first embodiment, the information indicating the free capacity is referred to as “free capacity 102”.
It progresses to S230 after S220.

In S230, the trace data input / output status analysis unit 140 selects the type of the trace data 101 based on the free capacity 102 of the trace data buffer 120 and the selection method information 103A.
By selecting the type of the trace data 101 based on the free capacity 102 of the trace data buffer 120, the important trace data 101 can be saved even if the free capacity 102 of the trace data buffer 120 decreases.
Details of S230 will be described later.
It progresses to S240 after S230.

In S240, the trace data input / output situation analysis unit 140 generates selection type information 103B indicating the type of the selected trace data 101, and inputs the generated selection type information 103B to the trace data sorting and selection unit 110.
After S240, the process returns to S220.

  The trace data input / output status analysis unit 140 executes the processing from S220 to S240 periodically or at a predetermined timing.

FIG. 5 is a schematic diagram illustrating an example of the trace data type selection process (S230) of the trace data input / output situation analysis unit 140 according to the first embodiment.
An example of the trace data type selection process (S230) executed in the trace data input / output situation analysis process (see FIG. 4) will be described with reference to FIG.

For example, the type of the trace data 101 (trace signal) includes a command signal 101A indicating the processing content (instruction), an address signal 101B indicating the address of the storage area where the processing target data and the like are stored, and the processing target It is assumed that there are three types of data signals 101C indicating data.
Further, it is assumed that the priority (importance) is higher in the order of the command signal 101A, the address signal 101B, and the data signal 101C.

The trace data sorting and selection unit 110 stores the trace data 101 of the same type as the selection type indicated in the selection type information 103B among the command signal 101A, the address signal 101B, and the data signal 101C in the trace data buffer 120.
The trace data sequential output unit 130 acquires the trace data 101 from the trace data buffer 120 in the storage order, and outputs the acquired trace data 101.

Of the storage areas of the trace data buffer 120, an area where the trace data 101 is stored is referred to as a “use area”, and an area where the trace data 101 is not stored is referred to as an “empty area”.
The free capacity 102 of the trace data buffer 120 means the size or ratio of the free area of the trace data buffer 120.

The trace data input / output status analysis unit 140 compares the free capacity 102 of the trace data buffer 120 with a capacity threshold (large) and a capacity threshold (small).
When the free capacity 102 is larger than the capacity threshold (large), the trace data input / output status analysis unit 140 selects all types of the trace data 101.
When the free capacity 102 is smaller than the capacity threshold value (large) and larger than the capacity threshold value (small), the trace data input / output status analysis unit 140 selects two types (command signal 101A and address signal 101B) in descending order of priority. select.
When the free capacity 102 is smaller than the capacity threshold (small), the trace data input / output status analysis unit 140 selects the type (command signal 101A) having the highest priority.

  The trace data input / output status analysis unit 140 inputs selection type information 103B indicating the type of the selected trace data 101 to the trace data sorting selection unit 110.

FIG. 6 is a flowchart illustrating an example of the trace data type selection process (S230) of the trace data input / output situation analysis unit 140 according to the first embodiment.
A flowchart of the trace data type selection process (S230) described with reference to FIG. 5 will be described with reference to FIG.

In S231, the trace data input / output status analysis unit 140 compares the free capacity 102 of the trace data buffer 120 with the capacity threshold (large) indicated in the selection method information 103A.
If the free capacity 102 is greater than or equal to the capacity threshold (large) (YES), the process proceeds to S232.
When the free capacity 102 is less than the capacity threshold (large) (NO), the process proceeds to S233.

In S232, the trace data input / output status analysis unit 140 selects the command signal 101A, the address signal 101B, and the data signal 101C based on the selection method information 103A.
By S232, the trace data selection process (S230) ends.

In S233, the trace data input / output status analysis unit 140 compares the free capacity 102 of the trace data buffer 120 with the capacity threshold (small) indicated in the selection method information 103A.
When the free capacity 102 is equal to or larger than the capacity threshold (small) (YES), the process proceeds to S234.
When the free capacity 102 is less than the capacity threshold (small) (NO), the process proceeds to S235.

In S234, the trace data input / output status analysis unit 140 selects the command signal 101A and the address signal 101B based on the selection method information 103A.
The trace data selection process (S230) is terminated by S234.

In S235, the trace data input / output status analysis unit 140 selects the command signal 101A based on the selection method information 103A.
By S235, the trace data selection process (S230) ends.

The trace data type selection process (S230) described with reference to FIGS. 5 and 6 is an example of the embodiment.
For example, the number of capacity thresholds to be compared with the free capacity 102 of the trace data buffer 120 is not two, but may be one or three or more.
Further, the types of trace data 101 may be two types or four or more types instead of three types.

As shown in FIG. 5, the free capacity 102 corresponds to the capacity of the used area in a complementary manner. Therefore, instead of comparing the free capacity 102 and the capacity threshold, the capacity of the used area and the capacity threshold may be compared.
In the first embodiment, comparing the free capacity 102 and the capacity threshold is equivalent to comparing the capacity of the use area and the capacity threshold (however, the magnitude relationship is reversed).

  In the first embodiment, for example, the following LSI trace data acquisition apparatus 100 has been described.

The LSI trace data acquisition apparatus 100 includes a trace data sorting / selection unit 110, a trace data buffer 120, and a trace data input / output status analysis unit 140.
The trace data selection unit 110 selects the trace data 101.
The trace data buffer 120 temporarily stores the trace data 101.
The trace data input / output status analysis unit 140 analyzes the input / output status (free capacity 102) of the trace data 101. Then, the trace data input / output status analysis unit 140 dynamically controls the trace data selection unit 110 according to the remaining capacity (free capacity 102) of the trace data buffer 120.

The LSI trace data acquisition apparatus 100 further includes a trace data sequential output unit 130.
The trace data sequential output unit 130 sequentially outputs the trace data 101 stored in the trace data buffer 120 from the trace data buffer 120 to the outside.

  The LSI trace data acquisition apparatus 100 further includes a trace data control register 141 for controlling the operation of the trace data input / output status analysis unit 140.

In the debugging operation of the LSI 210, information on all the trace signal generation sources 211 is not necessary, and each trace signal generation source 211 has a necessary priority. Further, the amount of data for which the trace data 101 can be sequentially output varies depending on the time.
In the first embodiment, the trace data 101 is appropriately selected by analyzing the input / output status (free capacity 102) of the trace data buffer 120.
That is, many trace signals (trace data 101) are acquired when the amount of data that can sequentially output the trace data 101 is large. In addition, when the amount of data that can sequentially output the trace data 101 is small, only a limited trace signal having a high priority is selected.
Thus, the trace data 101 can be acquired without stopping the operation of the LSI 210. Furthermore, the storage capacity of the trace data buffer 120 can be minimized by setting an appropriate capacity threshold. That is, the storage capacity of the trace data buffer 120 can be reduced.

There is a method of measuring the performance of the LSI 210 using a counter called “performance monitor” embedded in the LSI 210. The performance monitor counts a predetermined number of events. For example, the performance monitor counts events such as arithmetic instructions, bus assertions, and pipeline installations.
To read the performance monitor count value, it is necessary to access the performance monitor. However, when accessing the performance monitor using the CPU of the LSI 210, it is necessary to add a code for accessing the performance monitor and reading the count value from the performance monitor to the program source of the LSI 210. When accessing the performance monitor for debugging the LSI 210 or measuring the performance of the LSI 210, the CPU of the LSI 210 may access the performance monitor, which may affect the usage environment of the LSI 210.
In the first embodiment, if the performance monitor is connected to the LSI trace data acquisition apparatus 100 as the trace signal generation source 211, the count value (an example of the trace data 101) of the performance monitor can be acquired by the LSI trace data acquisition apparatus 100. it can. That is, it is not necessary to add code to the program source of the LSI 210, and the usage environment of the LSI 210 is not affected.

Embodiment 2. FIG.
In order to store many types of trace data 101 in the trace data buffer 120, a form in which the trace data 101 is compressed and stored in the trace data buffer 120 will be described.
Hereinafter, items different from the first embodiment will be mainly described. Matters whose description is omitted are the same as those in the first embodiment.

FIG. 7 is a functional configuration diagram of the LSI trace data acquisition apparatus 100 according to the second embodiment.
A functional configuration of the LSI trace data acquisition apparatus 100 according to the second embodiment will be described with reference to FIG.

  The LSI trace data acquisition apparatus 100 includes a compression circuit 150 in addition to the configuration of the first embodiment (see FIG. 2).

  The compression circuit 150 (an example of the trace data compression unit) compresses the input data when the trace data sorting / selection unit 110 determines that the input data is the selected type of trace data.

  The trace data buffer 120 stores the input data compressed by the compression circuit 150.

FIG. 8 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the second embodiment.
A processing flow of the trace data acquisition method according to the second embodiment will be described with reference to FIG.

  The same processing number (Sxxx) as in the first embodiment is assigned to the same processing as in the first embodiment (see FIG. 3).

In S110, the trace data sorting / selection unit 110 inputs the trace data 101.
It progresses to S120 after S110.

In S120, the trace data sorting / selection unit 110 determines whether or not the input trace data 101 is the selected type of trace data 101.
When the input trace data 101 is the selected type of trace data 101 (YES), the process proceeds to S131.
If the input trace data 101 is not the selected type of trace data 101 (NO), the process returns to S110.

In S131, the compression circuit 150 inputs the trace data 101 from the trace data sorting / selection unit 110, and compresses the input trace data 101 by a predetermined compression method.
By compressing the trace data 101, the data size of the trace data 101 can be reduced, and the free capacity 102 of the trace data buffer 120 can be increased.
When the free space 102 of the trace data buffer 120 increases, as described in the first embodiment (see FIGS. 4 to 6), the types of the trace data 101 selected by the trace data input / output status analysis unit 140 increase. . That is, more types of trace data 101 can be stored in the trace data buffer 120.

For example, the compression circuit 150 compresses the trace data 101 by the run length encoding method.
In the run-length encoding method, when the same data as the previous time is input, the data is not output and the number of times the same data is input is counted. When data different from the previous time is input, the input data and the number of counts up to the previous time are output together. Thereby, the amount of output data can be reduced.
This run-length encoding method is particularly effective when the change in the trace data 101 is small.

  After S131, the process proceeds to S132.

In S <b> 132, the compression circuit 150 stores the compressed trace data 101 (compressed data 104) in the trace data buffer 120.
It progresses to S141 after S132.

In S <b> 141, the trace data sequential output unit 130 acquires the compressed data 104 from the trace data buffer 120 and outputs the acquired compressed data 104.
After S141, the process returns to S110.

In the second embodiment, the LSI trace data acquisition apparatus 100 including the compression circuit 150 has been described.
The compression circuit 150 compresses the trace data 101 generated from the trace data sorting / selection unit 110.
By compressing the trace signal (trace data 101) selected by the trace data selection / selection unit 110 by the compression circuit 150, the amount of trace data of the trace signal can be reduced, and more trace signals can be acquired.

Embodiment 3 FIG.
A description will be given of a mode in which more types of trace data 101 are stored in the trace data buffer 120 by providing a plurality of compression circuits.
Hereinafter, items different from the first and second embodiments will be mainly described. Matters whose description is omitted are the same as in the first and second embodiments.

FIG. 9 is a functional configuration diagram of the LSI trace data acquisition apparatus 100 according to the third embodiment.
A functional configuration of the LSI trace data acquisition apparatus 100 according to the third embodiment will be described with reference to FIG.

  The LSI trace data acquisition apparatus 100 includes a plurality of compression circuits 150, a compression circuit selection unit 151, a compression circuit analysis unit 152, and a compression circuit control register 153 in addition to the configuration of the first embodiment (see FIG. 2).

  The plurality of compression circuits 150 compress the input data using different compression methods when the input data is determined to be the selected type of trace data 101 by the trace data sorting and selection unit 110.

The compression circuit analysis unit 152 (an example of a compressed data selection unit) selects any input data from among a plurality of input data obtained by compression by the plurality of compression circuits 150.
Hereinafter, the compression circuit 150 that outputs the selected input data (compressed data 104) is referred to as a “selection circuit 105B”.

  The compression circuit selection unit 151 stores the input data selected by the compression circuit analysis unit 152 in the trace data buffer 120.

The compression circuit control register 153 is a storage unit in which compression condition information 105A (or an address of a storage area that stores the compression condition information 105A) is set.
The compression condition information 105A is information indicating a selection condition for causing the compression circuit analysis unit 152 to select the compressed data 104.

FIG. 10 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the third embodiment.
A processing flow of the trace data acquisition method according to the third embodiment will be described with reference to FIG.

  The same processing number (Sxxx) as in the first embodiment is assigned to the same processing as in the first embodiment (see FIG. 3).

In S110, the trace data sorting / selection unit 110 inputs the trace data 101.
It progresses to S120 after S110.

In S120, the trace data sorting / selection unit 110 determines whether or not the input trace data 101 is the selected type of trace data 101.
When the input trace data 101 is the selected type of trace data 101 (YES), the process proceeds to S133.
If the input trace data 101 is not the selected type of trace data 101 (NO), the process returns to S110.

  In S133, the plurality of compression circuits 150 input the trace data 101 from the trace data sorting / selection unit 110, and compress the input trace data 101 by different compression methods.

The compression rate of the trace data 101 varies depending on the compression method, and even with the same compression method, it varies depending on the data configuration of the trace data 101.
Therefore, a compression method that compresses a certain trace data 101 at a higher compression rate than other compression methods cannot always compress another trace data 101 at a higher compression rate than other compression methods.
Therefore, the trace data 101 can be compressed at a higher compression rate by including a plurality of compression circuits 150 that compress the trace data 101 using different compression methods.

  It progresses to S134 after S133.

In S134, it is assumed that the compression circuit analysis unit 152 acquires the compression condition information 105A from the compression circuit control register 153 (or the storage area specified by the address indicated by the compression circuit control register 153) in advance. The compression circuit analysis unit 152 may reacquire the compression condition information 105A periodically or at a predetermined timing.
The compression circuit analysis unit 152 selects one of the plurality of compression circuits 150 based on the compression condition information 105A, and inputs information indicating the selected compression circuit 150 to the compression circuit selection unit 151.
In the third embodiment, the selected compression circuit 150 is referred to as “selection circuit 105B”.

  The compression condition information 105 </ b> A is information indicating a selection condition for selecting any one of the trace data 101 (compressed data 104) that has been compressed. The compression condition information 105A is previously set by the user in the compression circuit control register 153 or the memory of the LSI trace data acquisition apparatus 100.

For example, the compression condition information 105A specifies “compressed data 104 with the smallest data size (compression circuit 150 with the highest compression rate)” as a selection condition for the compressed data 104.
In this case, the compression circuit analysis unit 152 identifies the compressed data 104 having the smallest data size among the plurality of compressed data 104 generated by the plurality of compression circuits 150, and determines the compression circuit 150 that generated the identified compressed data 104. Select as the selection circuit 105B.

For example, the compression condition information 105 </ b> A designates one of the plurality of compression circuits 150.
In this case, the compression circuit analysis unit 152 selects the compression circuit 150 specified in the compression condition information 105A as the selection circuit 105B.

  After S134, the process proceeds to S135.

In S135, the compression circuit selection unit 151 receives the compressed data 104 obtained by compressing the trace data 101 from each of the plurality of compression circuits 150, and receives information indicating the selection circuit 105B from the compression circuit analysis unit 152.
Then, the compression circuit selection unit 151 stores, in the trace data buffer 120, the compressed data 104 input from the selection circuit 105B among the plurality of compressed data 104 and the identification information indicating the compression method of the selection circuit 105B.
It progresses to S141 after S135.

In S <b> 141, the trace data sequential output unit 130 outputs the compressed data 104 and the identification information of the compression method from the trace data buffer 120.
After S141, the process returns to S110.

  The selection of the compression circuit 150 by the compression circuit analysis unit 152 (S134) may be performed periodically or at a predetermined timing without being performed every time the trace data 101 is input to the trace data sorting selection unit 110.

  In the third embodiment, for example, the following LSI trace data acquisition apparatus 100 has been described.

The LSI trace data acquisition apparatus 100 includes a plurality of compression circuits 150, a compression circuit selection unit 151, and a compression circuit control register 153.
The plurality of compression circuits 150 compress the trace data 101 using different compression methods.
The compression circuit selection unit 151 selects one of the compression circuits 150 and stores the trace data 101 compressed by the selected compression circuit 150 in the trace data buffer 120.
The compression circuit control register 153 gives information used for selecting the compression circuit 150.
Thereby, the amount of trace data can be reduced and more trace data 101 can be acquired than when the trace data 101 is compressed by a single compression method.

Embodiment 4 FIG.
A mode in which the trace data 101 is stored only when a predetermined event occurs will be described.
Hereinafter, items different from the first to third embodiments will be mainly described. Matters whose description is omitted are the same as those in the first to third embodiments.

FIG. 11 is a functional configuration diagram of the LSI trace data acquisition apparatus 100 according to the fourth embodiment.
A functional configuration of the LSI trace data acquisition apparatus 100 according to the fourth embodiment will be described with reference to FIG.

  The LSI trace data acquisition apparatus 100 includes an event monitoring unit 160 and an event control register 161 in addition to the configuration of the first embodiment (see FIG. 2).

  The event monitoring unit 160 (an example of an event determination unit) determines whether or not a predetermined event has occurred.

  The trace data buffer 120 stores input data when it is determined that a predetermined event has occurred by the event monitoring unit 160 and the input data is determined to be the selected type of trace data 101 by the trace data selection unit 110. .

The event control register 161 is a storage unit in which event information 106 (or an address of a storage area storing the event information 106) is set.
The event information 106 is information indicating an event to be determined by the event monitoring unit 160.

FIG. 12 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the fourth embodiment.
A processing flow of the trace data acquisition method according to the fourth embodiment will be described with reference to FIG.

  The same processing number (Sxxx) as in the first embodiment is assigned to the same processing as in the first embodiment (see FIG. 3).

In S110, the trace data sorting / selection unit 110 inputs the trace data 101.
It progresses to S120 after S110.

In S120, the trace data sorting / selection unit 110 determines whether or not the input trace data 101 is the selected type of trace data 101.
When the input trace data 101 is the selected type of trace data 101 (YES), the process proceeds to S121.
If the input trace data 101 is not the selected type of trace data 101 (NO), the process returns to S110.

  In step S <b> 121, the event monitoring unit 160 inputs the trace data 101 from the trace data selection unit 110.

  In addition, it is assumed that the event monitoring unit 160 has previously acquired the event information 106 from the event control register 161 (or the storage area specified by the address indicated in the event control register 161). The event information 106 is information indicating an event for which it is determined whether or not it has occurred. In the fourth embodiment, the event indicated in the event information 106 is referred to as a “target event”. The event monitoring unit 160 may acquire the event information 106 periodically or at a predetermined timing.

  Then, the event monitoring unit 160 determines whether or not the target event indicated in the event information 106 has already occurred.

An event is a specific event that occurs in the LSI 210 such as generation of a command, data transfer, or error detection.
In the fourth embodiment, a signal indicating an event is referred to as an event generation stimulus signal. When an event occurs in the LSI 210, the circuit of the LSI 210 that has generated or detected the event outputs an event generation stimulus signal.

When an event occurs in the LSI 210, the event monitoring unit 160 inputs an event generation stimulus signal from the circuit of the LSI 210 that has generated or detected the event.
In addition, the event monitoring unit 160 determines whether or not the event indicated in the input event occurrence stimulus signal is a target event indicated in the event information 106.

If the trace data 101 input in S110 is the trace data 101 input first after the event occurrence stimulus signal of the target event is input, the event monitoring unit 160 determines that the target event has already occurred. .
However, the event monitoring unit 160 may determine according to other determination conditions. For example, when a predetermined elapsed time has not elapsed since the event occurrence stimulus signal of the target event has been input, the event monitoring unit 160 may determine that the target event has already occurred. Further, the event monitoring unit 160 may determine that the target event has already occurred from the input of the event occurrence stimulus signal of the target event to the input of another event occurrence stimulus signal.

When the target event indicated in the event information 106 has already occurred (YES), the process proceeds to S136.
When the target event indicated by the event information 106 has not occurred (NO), the event monitoring unit 160 discards the trace data 101, and the process returns to S110.

In S <b> 136, the event monitoring unit 160 stores the trace data 101 in the trace data buffer 120.
After S136, the process proceeds to S140.

In S140, the trace data sequential output unit 130 acquires the trace data 101 from the trace data buffer 120, and outputs the acquired trace data 101.
After S140, the process returns to S110.

  In the fourth embodiment, for example, the following LSI trace data acquisition apparatus 100 has been described.

The LSI trace data acquisition device 100 includes an event monitoring unit 160 and an event control register 161.
The event monitoring unit 160 stores the trace data 101 generated from the trace data sorting / selection unit 110 in the trace data buffer 120 only when a prescribed event occurs.
The event control register 161 controls an event for saving the trace data 101 in the trace data buffer 120.

In the fourth embodiment, the LSI trace data acquisition apparatus 100 may include the compression circuit 150 as in the second embodiment.
In this case, the compression circuit 150 inputs the compressed data 104 obtained by compressing the trace data 101 to the event monitoring unit 160, and the event monitoring unit 160 stores the compressed data 104 input from the compression circuit 150 in the trace data buffer 120. To do.

Further, the LSI trace data acquisition apparatus 100 may include a plurality of compression circuits 150, a compression circuit selection unit 151, a compression circuit analysis unit 152, and a compression circuit control register 153, as in the third embodiment.
In this case, the compression circuit selection unit 151 inputs the compressed data 104 output from the selection circuit 105B and identification information indicating the compression method of the selection circuit 105B to the event monitoring unit 160. Then, the event monitoring unit 160 stores the compressed data 104 and the compression method identification information input from the compression circuit selection unit 151 in the trace data buffer 120.

Embodiment 5 FIG.
The type of the trace data 101 to be saved in the trace data buffer 120 is selected based on the input amount of the trace data 101 input (saved) in the trace data buffer 120 and the output amount of the trace data 101 output from the trace data buffer 120 The form to perform is demonstrated.
Hereinafter, items different from the first to fourth embodiments will be mainly described. Matters whose description is omitted are the same as those in the first to fourth embodiments.

FIG. 13 is a functional configuration diagram of the LSI trace data acquisition apparatus 100 according to the fifth embodiment.
A functional configuration of the LSI trace data acquisition apparatus 100 according to the fifth embodiment will be described with reference to FIG.

  The LSI trace data acquisition apparatus 100 includes an input trace data monitoring unit 170 and an output trace data monitoring unit 171 in addition to the configuration of the first embodiment (see FIG. 2).

  The input trace data monitoring unit 170 calculates the storage amount (input amount) of the trace data 101 stored in the trace data buffer 120.

  The output trace data monitoring unit 171 calculates the output amount of the trace data 101 output from the trace data buffer 120.

  The trace data input / output status analysis unit 140 is configured to generate trace data based on the storage amount of the trace data 101 calculated by the input trace data monitoring unit 170 and the output amount of the trace data 101 calculated by the output trace data monitoring unit 171. The type of the trace data 101 stored in the buffer 120 is selected.

For example, the trace data input / output status analysis unit 140 calculates a value obtained by subtracting the output amount of the trace data 101 from the storage amount of the trace data 101 as the trace data input / output difference.
The trace data input / output status analysis unit 140 compares the calculated trace data input / output difference with a predetermined input / output difference threshold.
When the trace data input / output difference is larger than the input / output difference threshold value, the trace data input / output status analysis unit 140 selects the selection type of the trace data 101 by the predetermined first selection type number.
When the trace data input / output difference is smaller than the input / output difference threshold, the trace data input / output status analysis unit 140 selects a predetermined second selected type number of the trace data 101 that is larger than the first selected type number. .

FIG. 14 is a flowchart showing a trace data acquisition method of the LSI trace data acquisition apparatus 100 according to the fifth embodiment.
A processing flow of the trace data acquisition method according to the fifth embodiment will be described with reference to FIG.

  The same processing number (Sxxx) as in the first embodiment is assigned to the same processing as in the first embodiment (see FIG. 3).

In S110, the trace data sorting / selection unit 110 inputs the trace data 101.
It progresses to S120 after S110.

In S120, the trace data sorting / selection unit 110 determines whether or not the input trace data 101 is the selected type of trace data 101.
When the input trace data 101 is the selected type of trace data 101 (YES), the process proceeds to S137.
If the input trace data 101 is not the selected type of trace data 101 (NO), the process returns to S110.

In S <b> 137, the input trace data monitoring unit 170 inputs the trace data 101 from the trace data sorting and selection unit 110 and stores the input trace data 101 in the trace data buffer 120.
It progresses to S138 after S137.

  In S138, the input trace data monitoring unit 170 calculates the input amount of the trace data 101 stored in the trace data buffer 120.

For example, the input trace data monitoring unit 170 calculates the total input amount, the maximum input amount, the current input amount, the average input amount, and the like of the trace data 101 as the input amount of the trace data 101.
The total input amount of the trace data 101 is the total size of the trace data 101 stored in the trace data buffer 120 so far. “Until then” is, for example, from when the LSI 210 (or the LSI trace data acquisition apparatus 100) is activated to when the calculation is performed.
The maximum input amount of the trace data 101 is the data size of the largest trace data 101 among the trace data 101 stored in the trace data buffer 120 so far.
The current input amount of the trace data 101 is the data size of the trace data 101 stored this time in the trace data buffer 120 in S137.
The average input amount of the trace data 101 is the data size per unit time of the trace data 101 stored in the trace data buffer 120 so far.

  It progresses to S139 after S138.

In S139, the output trace data monitoring unit 171 acquires the trace data 101 from the trace data buffer 120, and outputs the acquired trace data 101 to the trace data sequential output unit 130. However, it is assumed that the data amount of the trace data 101 output from the trace data buffer 120 to the trace data sequential output unit 130 matches the output amount of the trace data 101 output from the trace data sequential output unit 130.
Then, the output trace data monitoring unit 171 calculates the output amount of the trace data 101 output from the trace data buffer 120 to the trace data sequential output unit 130.

For example, the output trace data monitoring unit 171 calculates the total output amount, maximum output amount, current output amount, average output amount, outputable amount, and the like of the trace data 101 as the output amount of the trace data 101.
The total output amount of the trace data 101 is the total size of the trace data 101 output from the trace data buffer 120 so far. “Until then” is, for example, from when the LSI 210 (or the LSI trace data acquisition apparatus 100) is activated to when the calculation is performed.
The maximum output amount of the trace data 101 is the data size of the largest trace data 101 among the trace data 101 output from the trace data buffer 120 so far.
The current output amount of the trace data 101 is the data size of the trace data 101 output this time from the trace data buffer 120.
The average output amount of the trace data 101 is the data size per unit time of the trace data 101 output from the trace data buffer 120 so far.
The amount of trace data 101 that can be output is the data size of the trace data 101 that the trace data sequential output unit 130 can output per unit time.

  It progresses to S142 after S139.

In S <b> 142, the trace data sequential output unit 130 outputs the trace data 101 output from the trace data buffer 120 by the output trace data monitoring unit 171.
After S142, the process returns to S110.

FIG. 15 is a flowchart showing trace data input / output status analysis processing of the trace data input / output status analysis unit 140 according to the fifth embodiment.
The trace data input / output status analysis processing of the trace data input / output status analysis unit 140 in the fifth embodiment will be described with reference to FIG.

  The same processing number (Sxxx) as in the first embodiment is assigned to the same processing as in the first embodiment (see FIG. 4).

In S210, the trace data input / output status analysis unit 140 acquires the selection method information 103A.
In the fifth embodiment, the selection method information 103A indicates the selection condition of the trace data 101 based on the input / output amount of the trace data 101.
It progresses to S220 after S210.

In S <b> 220, the trace data input / output status analysis unit 140 acquires information indicating the free capacity of the trace data buffer 120 (free capacity 102).
It progresses to S221 after S220.

In step S <b> 221, the trace data input / output situation analysis unit 140 acquires the input amount information 107 </ b> A from the input trace data monitoring unit 170 and acquires the output amount information 107 </ b> B from the output trace data monitoring unit 171.
The input amount information 107A is information indicating the input amount of the trace data 101 calculated by the input trace data monitoring unit 170, and the output amount information 107B is the output amount of the trace data 101 calculated by the output trace data monitoring unit 171. Information.
It progresses to S230B after S221.

In S230B, the trace data input / output status analysis unit 140 selects the type of the trace data 101 based on the input amount information 107A, the output amount information 107B, the free space 102, and the selection method information 103A.
A specific example of S230B will be described later.
It progresses to S240 after S230B.

In S240, the trace data input / output status analysis unit 140 inputs selection type information 103B indicating the type of the selected trace data 101 to the trace data selection unit 110.
After S240, the process returns to S220.

FIG. 16 is a flowchart illustrating an example of the trace data type selection process (S230B) of the trace data input / output situation analysis unit 140 according to the fifth embodiment.
An example of the trace data type selection process (S230B) of the trace data input / output situation analysis unit 140 in the fifth embodiment will be described with reference to FIG.

In S231B, the trace data input / output status analysis unit 140 compares the free capacity 102 of the trace data buffer 120 with the capacity threshold indicated in the selection method information 103A.
If the free capacity 102 is equal to or greater than the capacity threshold (YES), the process proceeds to S232B.
When the free capacity 102 is less than the capacity threshold (NO), the process proceeds to S235B.

In S232B, the trace data input / output status analysis unit 140 calculates a value obtained by subtracting the average output amount of the trace data 101 from the average input amount of the trace data 101 as the trace data input / output difference. The average input amount of the trace data 101 is a value indicated by the input amount information 107A, and the average output amount of the trace data 101 is a value indicated by the output amount information 107B.
The trace data input / output status analysis unit 140 compares the calculated trace data input / output difference with the input / output difference threshold (small) indicated in the selection method information 103A.
When the trace data input / output difference is equal to or smaller than the input / output difference threshold (small) (YES), that is, when the difference between the average input amount of the trace data 101 and the average output amount of the trace data 101 is small, the process proceeds to S233B.
When the trace data input / output difference is larger than the input / output difference threshold (small) (NO), that is, when the difference between the average input amount of the trace data 101 and the average output amount of the trace data 101 is not small, the process proceeds to S234B.

In S233B, the trace data input / output status analysis unit 140 increases the selection type by one (or two or more predetermined numbers) in descending order of priority. However, when the currently selected types are all types, the trace data input / output status analysis unit 140 does not increase the selected types.
For example, when the current selection type is only the command signal with the highest priority, the trace data input / output status analysis unit 140 additionally selects the address signal with the second highest priority. In this case, there are two types of selection: command signal and address signal.
The trace data type selection process (S230B) is terminated by S233B.

In S234B, the trace data input / output status analysis unit 140 compares the trace data input / output difference with the input / output difference threshold (large) indicated in the selection method information 103A.
When the trace data input / output difference is less than or equal to the input / output difference threshold (large) (YES), that is, when the difference between the average input amount of the trace data 101 and the average output amount of the trace data 101 is not small, large, or medium The trace data input / output status analysis unit 140 does not change the selection type, and the trace data type selection process (S230B) ends.
When the trace data input / output difference is larger than the input / output difference threshold (large) (NO), that is, when the difference between the average input amount of the trace data 101 and the average output amount of the trace data 101 is large, the process proceeds to S235B.

In S235B, the trace data input / output status analysis unit 140 reduces the selection type by one (or a predetermined number of two or more) in ascending order of priority. However, when the current selection type is only one type, the trace data input / output status analysis unit 140 does not reduce the selection type.
For example, when the current selection type is all types (command signal, address signal, data signal), the trace data input / output status analysis unit 140 excludes the data signal having the lowest priority from the selection type. In this case, there are two types of selection: command signal and address signal.
The trace data type selection process (S230B) is terminated by S235B.

The trace data input / output status analysis unit 140 may select the type of the trace data 101 based on the input / output amount of the trace data 101 other than the average input / output amount.
Further, the trace data input / output status analysis unit 140 may select the type of the trace data 101 based on the input amount of the trace data 101 without using the output amount of the trace data 101.
Further, the trace data input / output status analysis unit 140 may select the type of the trace data 101 based on the output amount of the trace data 101 without using the input amount of the trace data 101.

  In the fifth embodiment, for example, the following LSI trace data acquisition apparatus 100 has been described.

The LSI trace data acquisition apparatus 100 includes an input trace data monitoring unit 170 and an output trace data monitoring unit 171.
The input trace data monitoring unit 170 monitors the trace data 101 input from the trace data sorting / selecting unit 110 to the trace data buffer 120.
The output trace data monitoring unit 171 monitors the trace data 101 output from the trace data buffer 120.

  The input / output amount of the trace data 101 varies depending on the execution status of the LSI 210. For this reason, the selection type of the trace data 101 is frequently switched only by the information on the remaining capacity (free capacity 102) of the trace data buffer 120, and conversely, the selection type of the trace data 101 is sufficiently switched. I will not be. By using the input / output amount of the trace data 101, the selection type of the trace data 101 can be switched more appropriately.

In the fifth embodiment, the LSI trace data acquisition apparatus 100 may include the compression circuit 150 as in the second embodiment.
In this case, the compression circuit 150 inputs the compressed data 104 obtained by compressing the trace data 101 to the input trace data monitoring unit 170, and the input trace data monitoring unit 170 converts the compressed data 104 input from the compression circuit 150 into the trace data. Save to buffer 120.
The input trace data monitoring unit 170 calculates the input amount of the compressed data 104, and the output trace data monitoring unit 171 calculates the output amount of the compressed data 104.

Further, the LSI trace data acquisition apparatus 100 may include a plurality of compression circuits 150, a compression circuit selection unit 151, a compression circuit analysis unit 152, and a compression circuit control register 153, as in the third embodiment.
In this case, the compression circuit selection unit 151 inputs the compressed data 104 output from the selection circuit 105B and identification information indicating the compression method of the selection circuit 105B to the input trace data monitoring unit 170. The input trace data monitoring unit 170 stores the compressed data 104 and the compression method identification information input from the compression circuit selection unit 151 in the trace data buffer 120.
The input trace data monitoring unit 170 calculates the input amount of the compressed data 104 (and the compression method identification information), and the output trace data monitoring unit 171 calculates the output amount of the compressed data 104 (and the compression method identification information).

Further, the LSI trace data acquisition apparatus 100 may include an event monitoring unit 160 and an event control register 161 as in the fourth embodiment.
In this case, the input trace data monitoring unit 170 inputs the trace data 101 (or compressed data 104) from the event monitoring unit 160 and stores the input trace data 101 (or compressed data 104) in the trace data buffer 120.

Embodiment 6 FIG.
A mode in which the trace data 101 is output to a predetermined output destination will be described.
Hereinafter, items different from the first to fifth embodiments will be mainly described. Matters whose description is omitted are the same as in the first to fifth embodiments.

FIG. 17 is a functional configuration diagram of the LSI trace data acquisition apparatus 100 according to the sixth embodiment.
A functional configuration of the LSI trace data acquisition apparatus 100 according to the sixth embodiment will be described with reference to FIG.

  The LSI trace data acquisition apparatus 100 includes a sequential output control register 131 in addition to the configuration of the first embodiment (see FIG. 2).

The sequential output control register 131 is a storage unit in which the output destination information 108 (or the address of the storage area storing the output destination information 108) is set.
The output destination information 108 is information indicating an output destination to which the trace data 101 is output.

  The trace data sequential output unit 130 acquires the output destination information 108 periodically or at a predetermined timing, and outputs the trace data 101 to the output destination indicated by the acquired output destination information 108.

The user can change the output destination of the trace data 101 by changing the output destination information 108.
For example, when a failure occurs in the external output interface 213, the user designates the internal memory 214 in the output destination information 108. Thereby, the trace data 101 can be saved from the trace data buffer 120 to the internal memory 214.

  It is conceivable to use a general-purpose interface such as PCI Express used in the LSI 210 as the external output interface 213. In that case, the external pins used exclusively for debugging and testing in the conventional method can be reduced.

In recent years, the LSI 210 is often connected to a network, and may operate in an environment integrated by the network. By transmitting the trace data 101 of the LSI 210 to such a network, it is possible to deal with a problem of low reproducibility. Therefore, an error signal can be output from the trace data sequential output unit 130 to the outside.
If a non-volatile memory is used as the internal memory 214, data is not lost even when the power is turned off, and an error can be dealt with.
Further, when security information such as a password is not set in the sequential output control register 131, the trace data 101 may be restricted from being output to the outside of the LSI 210. Further, the trace data sequential output unit 130 may encrypt and output the trace data 101. As a result, the trace data 101 can be protected from an attack that attempts to illegally acquire the trace data 101 of the LSI 210, and the processing information inside the LSI 210 can be concealed.

In the sixth embodiment, the LSI trace data acquisition apparatus 100 may compress the trace data 101 as in the second or third embodiment.
Further, the LSI trace data acquisition apparatus 100 may acquire the trace data 101 only when a specific event occurs as in the fourth embodiment.
Further, the LSI trace data acquisition apparatus 100 may determine the selection type of the trace data 101 based on the input / output amount of the trace data 101 as in the fifth embodiment.

Embodiment 7 FIG.
The debug device 220 that acquires the trace data 101 from the LSI trace data acquisition device 100 in the LSI 210 and debugs the LSI 210 based on the acquired trace data 101 will be described.

FIG. 18 is a configuration diagram of the debugging system 200 according to the seventh embodiment.
A functional configuration of the debugging device 220 according to the seventh embodiment will be described with reference to FIG.

  The debug system 200 according to the seventh embodiment includes any one of the LSI 210 and the LSI trace data acquisition apparatus 100 described in the first to sixth embodiments.

  The debugging device 220 includes a trace data restoring unit 221, an input data freezing unit 222, a data storage 223, and a debugging unit 224.

  The trace data restoration unit 221 (an example of a debug input unit) inputs the trace data 101 stored in the trace data buffer 120 from the LSI trace data acquisition device 100.

  The input data freezing unit 222 (an example of a save command input unit) inputs an input data freeze notification 229 (an example of a trace data save command) that designates saving of the trace data 101 stored in the data storage 223.

The data storage 223 (an example of a debug storage unit) has a storage area for storing the trace data 101.
When the storage area for storing the new trace data 101 is insufficient, the data storage 223 overwrites the new trace data 101 in the storage area in which the past trace data 101 is stored.
When the input data freezing notification 229 is input to the input data freezing unit 222, the data storage 223 does not store the new trace data 101 input to the trace data restoring unit 221.

The debug unit 224 displays the trace data 101 stored in the data storage 223 on the display device.
The debugging unit 224 inputs a user instruction based on the trace data 101 from an input device such as a keyboard or a mouse, and performs a debugging process according to the user instruction.

The debug device 220 includes hardware (not shown) such as a CPU, a memory, a magnetic disk device, a display device, an input device, and a communication device.
The configuration described as the “˜unit” of the debug device 220 may be “˜circuit”, “˜device”, “˜equipment”, and “˜step”, “˜procedure”, and “˜processing”. May be. In other words, the configuration described as “˜unit” may be implemented by any of firmware, software, hardware, or a combination thereof. A program that executes a function described as “unit” is stored in a memory, and is read and executed by the CPU. That is, the program causes the computer to function as “to part”, and causes the computer to execute the procedures and methods of “to part”.

FIG. 19 is a flowchart illustrating a trace data storage method of the debugging device 220 according to the seventh embodiment.
A flow of processing of the trace data storage method executed by the debug device 220 will be described with reference to FIG.

In S310, the LSI trace data acquisition apparatus 100 acquires the trace data 101 of the LSI 210 and outputs the acquired trace data 101 to the debug apparatus 220 as described in the first to sixth embodiments.
The trace data 101 may be compressed compressed data or encrypted encrypted data. However, if the compression method or encryption method is not determined, the LSI trace data acquisition device 100 outputs identification information indicating the compression method or encryption method to the debug device 220 together with the trace data 101. For example, the LSI trace data acquisition apparatus 100 according to the third embodiment outputs identification information indicating the compression method to the debug apparatus 220 together with the trace data 101.
It progresses to S320 after S310.

In S320, the trace data restoration unit 221 receives the trace data 101 (and the compression method or encryption method identification information) output from the LSI trace data acquisition apparatus 100.
It progresses to S330 after S320.

In S330, the trace data restoration unit 221 restores the input trace data 101.
For example, when the trace data 101 is compressed, the trace data restoration unit 221 determines the compression method of the trace data 101 based on the identification information of the compression method, and uses the decompression method corresponding to the determined compression method. Unzip. When the trace data 101 is encrypted, the trace data restoration unit 221 determines the encryption method of the trace data 101 based on the identification information of the encryption method, and the decryption method corresponding to the determined encryption method. The trace data 101 is decoded.
However, when the trace data 101 is neither compressed nor encrypted, the restoration process of S330 is not necessary.
It progresses to S340 after S330.

In S340, the input data freezing unit 222 inputs the trace data 101 decoded from the trace data restoration unit 221.
Also, the input data freezing unit 222 determines whether or not the input data freezing notification 229 has been input.
The input data freeze notification 229 is a command that prohibits overwriting the storage area of the data storage 223 in which the trace data 101 is stored with new trace data 101. The input data freezing notification 229 is input to the debugging device 220 by the user or notified (input) from the LSI 210.
When the input data freezing notification 229 is input (YES), the input data freezing unit 222 discards the trace data 101, and the processing of the trace data storage method ends.
When the input data freeze notification 229 is not input (NO), the process proceeds to S350.

In S350, the input data freezing unit 222 stores the trace data 101 in the data storage 223.
When the storage area for storing the new trace data 101 is insufficient, the data storage 223 overwrites the new trace data 101 in the storage area in which the past trace data 101 is stored. A storage device that stores data cyclically is called a ring buffer.
By S350, the process of the trace data storage method ends.

  In the seventh embodiment, the following debugging apparatus 220 has been described.

The debugging device 220 includes a trace data restoration unit 221, an input data freezing unit 222, and a data storage 223.
The trace data restoration unit 221 restores the trace data 101 to information that can be used for debugging or the like.
The data storage 223 stores the restored data.
The input data freezing unit 222 stops saving data in the data storage 223.

  For example, when an event signal indicating an error is notified from the LSI 221 to the debug device 220 as an input data freeze notification 229, overwriting of the new trace data 101 is stopped, and the trace data stored before the error occurs in the LSI 221 101 can be secured. The user determines the cause of the error using the trace data 101 and debugs the LSI 210.

  100 LSI trace data acquisition device, 101 trace data, 101A command signal, 101B address signal, 101C data signal, 102 free capacity, 103A selection method information, 103B selection type information, 104 compressed data, 105A compression condition information, 105B selection circuit, 106 Event information, 107A Input amount information, 107B Output amount information, 108 Output destination information, 110 Trace data selection unit, 120 Trace data buffer, 130 Trace data sequential output unit, 131 Sequential output control register, 140 Trace data input / output status Analysis unit, 141 Trace data control register, 150 compression circuit, 151 Compression circuit selection unit, 152 Compression circuit analysis unit, 153 Compression circuit control register, 160 Event monitoring unit, 161 Control register, 170 input trace data monitoring unit, 171 output trace data monitoring unit, 200 debug system, 210 LSI, 211 trace signal generation source, 212 internal bus, 213 external output interface, 214 internal memory, 219 CPU, 220 debug device 221 Trace data restoring unit, 222 Input data freezing unit, 223 Data storage, 224 Debugging unit, 229 Input data freezing notification.

Claims (11)

An LSI (Large Scale Integration) provided with a trace data acquisition device, and a debugging device connected to the LSI;
The trace data acquisition device includes:
A trace data input unit for inputting each of a plurality of data of different types, each of which has a priority determined for each type, as trace data;
A trace data storage unit for storing trace data;
Based on the free capacity of the trace data storage unit, a trace data type selection unit that selects one or more types in descending order of priority as the type of trace data stored in the trace data storage unit;
When the trace data input unit inputs trace data, it is determined whether or not the input data input as the trace data to the trace data input unit is the selected type of trace data selected by the trace data type selection unit A trace data type determination unit for
The trace data storage unit stores the input data when the trace data type determination unit determines that the input data is the selected type of trace data ,
The debugging device includes:
A debug input unit for inputting the trace data stored in the trace data storage unit from the trace data acquisition device;
A debug storage unit having a storage area for storing trace data, and when the storage area for storing new trace data is insufficient, the debug is performed by overwriting the new trace data in the storage area for storing past trace data. A storage unit;
A save command input unit for inputting a trace data save command for designating saving of the trace data stored in the debug storage unit,
Debug system said debug storage unit, when the stored command to the store command input unit is inputted, characterized in that does not store new trace data input to the debug input section of the debugging device. The trace data type selection unit selects a larger number of types of trace data as the free space of the trace data storage unit is larger, and selects a smaller type of trace data as the free space of the trace data storage unit is smaller. 2. The debugging system according to claim 1, wherein the debugging system is selected as a type. The trace data type selection unit compares the free capacity of the trace data storage unit with a predetermined capacity threshold value. If the free capacity of the trace data storage unit is larger than the capacity threshold value, the trace data type selection unit selects a predetermined type of trace data. When the first selection type number is selected and the free capacity of the trace data storage unit is smaller than the capacity threshold, the trace data selection type is a predetermined second selection type number less than the first selection type number. The debugging system according to claim 1, wherein the debugging system is selected . The trace data acquisition device further includes:
A trace data compression unit that compresses the input data when the trace data type determination unit determines that the input data is the selected type of trace data;
The debug system according to any one of claims 1 to 3, wherein the trace data storage unit stores the input data compressed by the trace data compression unit. The trace data acquisition device includes a plurality of trace data compression units that compress the input data using different compression methods when the trace data type determination unit determines that the input data is the selected type of trace data. ,
The trace data acquisition device further includes:
A compressed data selection unit that selects any one of a plurality of input data obtained by compression by a plurality of trace data compression units,
5. The debugging system according to claim 4, wherein the trace data storage unit stores the input data selected by the compressed data selection unit. The trace data acquisition device further includes:
An event determination unit for determining whether or not a predetermined event has occurred;
The trace data storage unit determines that a predetermined event has occurred by the event determination unit, and the input data when the trace data type determination unit determines that the input data is the selected type of trace data. 6. The debugging system according to claim 1, wherein the debugging system is stored. The trace data acquisition device further includes:
7. The debugging system according to claim 1, further comprising a trace data output unit that outputs the trace data stored in the trace data storage unit to a predetermined output destination. An LSI (Large Scale Integration) provided with a trace data acquisition device, and a debugging device connected to the LSI;
The trace data acquisition device includes:
A trace data input unit for inputting each of a plurality of data of different types, each of which has a priority determined for each type, as trace data;
A trace data storage unit for storing trace data;
A trace data type selection unit that selects one or more types in descending order of priority as the type of trace data stored in the trace data storage unit based on the amount of trace data stored in the trace data storage unit; ,
When the trace data input unit inputs trace data, it is determined whether or not the input data input as the trace data to the trace data input unit is the selected type of trace data selected by the trace data type selection unit A trace data type determination unit for
The trace data storage unit stores the input data when the trace data type determination unit determines that the input data is the selected type of trace data ,
The debugging device includes:
A debug input unit for inputting the trace data stored in the trace data storage unit from the trace data acquisition device;
A debug storage unit having a storage area for storing trace data, and when the storage area for storing new trace data is insufficient, the debug is performed by overwriting the new trace data in the storage area for storing past trace data. A storage unit;
A save command input unit for inputting a trace data save command for designating saving of the trace data stored in the debug storage unit,
Debug system said debug storage unit, when the stored command to the store command input unit is inputted, characterized in that does not store new trace data input to the debug input section of the debugging device. The trace data acquisition device further includes:
A trace data output unit for outputting the trace data stored in the trace data storage unit from the trace data storage unit;
The trace data type selection unit stores the trace data in the trace data storage unit based on the storage amount of the trace data stored in the trace data storage unit and the output amount of the trace data output from the trace data storage unit. 9. The debugging system according to claim 8, wherein one or more types of trace data are selected in descending order of priority as selection types. The trace data type selection unit calculates a value obtained by subtracting the output amount of the trace data from the storage amount of the trace data as a trace data input / output difference, and calculates the calculated trace data input / output difference and a predetermined input / output difference threshold value. When the trace data input / output difference is larger than the input / output difference threshold, the selected number of trace data is selected by the predetermined first selection type number, and the trace data input / output difference is smaller than the input / output difference threshold 10. The debugging system according to claim 9, wherein the selection type of trace data is selected by a predetermined second selection type number that is larger than the first selection type number. A trace data acquisition method for a debug system comprising an LSI (Large Scale Integration) provided with a trace data acquisition device and a debug device connected to the LSI,
The trace data input unit of the trace data acquisition device is a plurality of different types of data, and inputs each of a plurality of data whose priority is determined for each type as trace data,
The trace data type selection unit of the trace data acquisition device selects one or more types of trace data stored in the trace data storage unit as a selection type based on the free space of the trace data storage unit in descending order of priority. ,
When the trace data type determination unit of the trace data acquisition device inputs the trace data, the trace data type selection unit selects the input data input as the trace data to the trace data input unit. Whether the selected type of trace data is
The trace data storage unit of the trace data acquisition device stores the input data when the trace data type determination unit determines that the input data is the selected type of trace data ,
Debug input unit of the debugging device inputs the trace data stored in the trace data storage unit from the trace data acquisition device,
Debugging storage unit of the debugging device, have a storage area for storing trace data, wherein said debug input unit when the trace data storage instruction that specifies the storage of trace data storage command input unit of the debugging apparatus has been input If the storage area for storing new trace data is not stored in the storage command input unit without storing the trace data storage command , the past trace data is stored. trace data acquisition method of debugging system comprising a benzalkonium overwrite a new trace data in which storage area.

Images