Chip testing method, device and equipment based on hardware accelerator and medium
Through the hardware accelerator-based method, the performance of the NVME controller in the chip is automatically tested, which solves the problem of difficulty in verifying the NVME controller in the prior art, and achieves efficient and accurate chip testing.
Patent Information
- Application Number
- CN202411157839.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-12-03
AI Technical Summary
The prior art is difficult to effectively verify the NVME controller inside the chip, affecting system performance and stability.
Through a hardware accelerator-based method, a transmission control protocol or Internet protocol link between servers is established, the operating status of the NVME controller is obtained, the target use cases are determined based on the status, and the use cases are executed through commands to automatically test the chip performance and generate an analysis report.
It realizes automated testing of chip performance, improves test efficiency and reliability of results, and can quickly locate the causes of errors, ensuring the accuracy and comprehensiveness of the test.
Smart Images

Figure CN119065912A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of chip testing technology, and in particular to a chip testing method, device, equipment and medium based on hardware accelerator. Background Art
[0002] With the development of semiconductor technology, the number of transistors has reached the tens of billions level, and the complexity of the System on Chip (SoC) has shown an exponential growth. In order to speed up verification, increase code coverage, and shorten product launch time, more and more companies have increased the application of hardware simulation accelerators in chip verification.
[0003] The Non-Volatile Memory Express (NVME) controller inside the chip is a key component connecting the host and the solid state drive (SSD). It plays a hub role. It not only implements the requirements of the NVME protocol specification, but also directly affects the performance of the SSD. During data transmission, the NVME controller is responsible for effectively managing data flow and ensuring efficient read and write operations. Its stability and efficiency directly determine the overall performance of the system. Therefore, in order to ensure stable operation of the system and to achieve optimal performance, the NVME controller needs to be accurately verified.
[0004] It can be seen that how to verify the NVME controller inside the chip based on the hardware accelerator is a technical problem that people in this field urgently need to solve. Summary of the invention
[0005] The purpose of the present invention is to provide a chip testing method, device, equipment and medium based on a hardware accelerator to solve the technical problem of verifying the NVME controller inside the chip based on the hardware accelerator.
[0006] In order to solve the above technical problems, the present invention provides a chip testing method based on a hardware accelerator, which is applied to a first server, and the method includes:
[0007] Pre-setting the universal asynchronous receiver / transmitter communication activator in the second server to a transmission control protocol service mode, and establishing a transmission control protocol link or an Internet protocol link with the second server;
[0008] Establishing a transmission control protocol link or an Internet protocol link with a third server; wherein the first server is a platform for chip testing, the second server is a chip side, and the third server is a host side;
[0009] Acquire the running state of the non-volatile memory fast channel controller inside the chip, and perform use case scheduling according to the running state to determine the target use case to be executed by the second server;
[0010] Sending a command representing a target use case to be executed by the second server to the third server, so that the second server executes the target use case according to the command;
[0011] Obtain the current running result of the second server executing the target use case, and when it is detected that the current running result is the same as the preset running result of the pre-set target use case, determine that the corresponding performance of the chip when executing the target use case is normal; otherwise, it is abnormal; wherein the first server pre-stores the running results of the second server executing multiple use cases.
[0012] On the one hand, performing use case scheduling according to the running state to determine a target use case to be executed by the second server includes:
[0013] Sending to the second server an instruction for characterizing reading the operating status of the non-volatile memory fast channel controller;
[0014] receiving the operating status sent by the non-volatile memory fast channel controller according to an instruction;
[0015] Determining whether the operating state of the non-volatile memory fast channel controller is a normal state;
[0016] If so, then when it is detected that the running state of the non-volatile memory fast channel controller is the initial state, determining that the target use case to be executed by the second server is the first use case; wherein the first use case is a basic use case related to a management command or an input / output operation command;
[0017] When it is detected that the running state of the non-volatile memory fast channel controller is a state of completing preliminary configuration and implementing execution of management commands or input / output operation commands, determining that the target use case to be executed by the second server is a second use case; wherein the second use case at least includes a use case for stress testing the chip;
[0018] After determining that the target use case to be executed by the second server is the second use case, the method further includes:
[0019] The use case for performing performance statistics on the chip is allocated to the second server to determine that the target use case to be executed by the second server is the use case for performing performance statistics on the chip; wherein the use case for performing performance statistics on the chip includes at least the number of input / output operations per second and the bandwidth of the chip.
[0020] On the other hand, in the case where it is determined that the performance of the chip corresponding to the target use case is abnormal, the method further includes:
[0021] Sending to the second server an instruction for querying data of a target register in the chip; wherein the target register is a register used by the chip to store data when executing the target use case;
[0022] receiving the data of the target register sent by the second server according to the instruction;
[0023] Comparing the data of the target register with the data in a preset error scenario model;
[0024] Output analysis report based on the comparison results.
[0025] On the other hand, use cases also include interrupt use cases and reset use cases;
[0026] Establishing the preset error scenario model includes:
[0027] The preset error scenario model is established based on the fact that the tail value and the head value in the queue corresponding to the execution use case are not equal and / or the preset error scenario model is established based on whether the register used to store data when executing the use case is set.
[0028] On the other hand, establishing the preset error scenario model according to the fact that the tail value and the head value in the queue corresponding to the execution use case are not equal includes:
[0029] Determine the management command path fault model according to the fact that the tail value and the head value in the queue corresponding to the execution management command are not equal;
[0030] According to the fact that the tail value and the head value in the queue corresponding to the execution of the input / output operation command are not equal, if the data of the counting register of the submission queue of the input / output operation command is not equal to the data actually sent, then the input / output operation command submission queue processing logic fault model is determined; if the data of the counting register of the submission queue of the input / output operation command is equal to the data actually sent, and is not equal to the data of the counting register of the input / output operation command completion queue, then the input / output operation command completion queue processing logic fault model is determined;
[0031] The establishing of the preset error scenario model according to whether the register used to store data when executing the use case is set includes:
[0032] When it is detected that the reset register used to store data is set when executing the reset case, and the value of the register to be reset is not cleared, the reset logic fault model is determined;
[0033] When it is detected that the interrupt status register used to store data is not set when executing the interrupt use case, and the interrupt generation condition is met, the interrupt logic fault model is determined.
[0034] On the other hand, after establishing the transmission control protocol link or the Internet protocol link with the second server and before establishing the transmission control protocol link or the Internet protocol link with the third server, the method further includes:
[0035] Sending a first initialization command frame to the second server, so that the second server completes the startup of the relevant clock and reset inside the chip according to the first initialization command frame, and establishes a fast peripheral component interconnection link with the virtual host in the third server via the fast peripheral component interconnection communication activator; wherein the first initialization command frame at least includes the initialization of the non-volatile memory fast channel controller;
[0036] After establishing the transmission control protocol link or the Internet protocol link with the third server and before acquiring the operating status of the non-volatile memory fast channel controller in the chip, the method further includes:
[0037] After detecting that the second server completes the initialization operation, controlling the second server to establish a transmission control protocol link or an Internet protocol link with the virtual host of the third server;
[0038] A second initialization command frame is sent to the third server so that the third server completes loading of a fast peripheral component interconnect device driver.
[0039] On the other hand, the first server, the second server and the third server transmit data according to a preset data format;
[0040] Wherein, the preset data transmission format includes a command frame and a status frame;
[0041] The command frame at least includes a command frame type, a use case type and a use case number; the command frame type at least includes a command frame for distinguishing environment initialization, a command frame for verifying use case operation and a command frame for querying chip status; the use case type at least includes a use case type for input and output operations, a use case corresponding to a management command, a reset use case, an interrupt use case and a statistical use case;
[0042] The status frame includes at least a status frame type, a use case type, a use case number and the number of status information packets; wherein the status frame type includes at least a data frame type used to distinguish normal use case outputs, a data frame type used to distinguish abnormal operation use case outputs and a data frame type used to distinguish chip error state outputs.
[0043] In order to solve the above technical problems, the present invention further provides a chip testing device based on a hardware accelerator, which is applied to a first server, and the device includes:
[0044] A setting and establishing module, used for pre-setting the universal asynchronous receiver / transmitter communication activator in the second server to a transmission control protocol service mode, and establishing a transmission control protocol link or an Internet protocol link with the second server;
[0045] An establishing module, used to establish a transmission control protocol link or an Internet protocol link with a third server; wherein the first server is a platform for chip testing, the second server is a chip side, and the third server is a host side;
[0046] An acquisition and scheduling module, used for acquiring the operating status of the non-volatile memory fast channel controller inside the chip, and performing use case scheduling according to the operating status to determine a target use case to be executed by the second server;
[0047] Sending a command, used to send a command representing a target use case to be executed by the second server to the third server, so that the second server executes the target use case according to the command;
[0048] A determination module is used to obtain the current running result of the second server executing the target use case, and when it is detected that the current running result is the same as the preset running result of the target use case, determine that the corresponding performance of the chip when executing the target use case is normal; otherwise, it is abnormal; wherein the first server pre-stores the running results of the second server executing multiple use cases.
[0049] In order to solve the above technical problems, the present invention also provides a chip testing device based on a hardware accelerator, comprising:
[0050] Memory for storing computer programs;
[0051] The processor is used to implement the steps of the above-mentioned chip testing method based on hardware accelerator when executing the computer program.
[0052] In order to solve the above technical problem, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned chip testing method based on hardware accelerator are implemented.
[0053] The chip testing method based on hardware accelerator provided by the present invention is applied to the first server. In the method, first, a transmission control protocol link or an Internet protocol link is established between the second server (i.e., the chip side) and the third server (i.e., the host side), respectively, to create a test environment for the first server to test the second server (i.e., the chip side), and then the target use case to be executed by the second server is determined. Finally, a command for the target use case to be executed by the second server is sent through the third server, so that the second server executes the target use case according to the command, thereby realizing the test of the chip performance according to the running result of the target use case.
[0054] The beneficial effects of the present invention are as follows: firstly, in the method, the first server sends a command for characterizing the target use case to be executed by the second server to the third server, so that the second server executes the target use case according to the command, thereby realizing the test of chip performance according to the running result of the target use case by the second server. In the process, there is no need to manually input the test case for comparison, thereby realizing the automatic test of chip performance and improving the test efficiency of the chip; secondly, before the second server executes the target use case, the universal asynchronous receiver / transmitter communication stimulator in the second server is first set to the transmission control protocol service mode, and a transmission control protocol link or an Internet protocol link is established with the second server and the third server respectively, thereby creating a test environment for chip testing and ensuring that the test can proceed normally; thirdly, the running results of the second server executing multiple use cases are pre-stored in the first server, thereby realizing the test of chip performance when the chip executes multiple use cases, thereby simulating a more realistic chip running scenario and improving the reliability of the test results; and, the target use case executed by the second server is determined while considering the running state of the non-volatile memory fast channel controller inside the chip, thereby controlling the second server to execute the target use case to test the chip performance, thereby making the selected target use case more appropriate and the chip performance test more reasonable and accurate.
[0055] In addition, when determining the target use case, the operating status sent by the non-volatile memory fast channel controller is first judged. When the operating status is normal, if the operating status of the non-volatile memory fast channel controller is in the initial state, the target use case to be executed by the second server is determined to be the basic use case; then, when the preliminary configuration is completed and the state of executing management commands or input and output operation commands is achieved, the target use case to be executed by the second server is determined to be the second use case (such as a stress test use case), and then the use case for performance statistics is allocated, thereby realizing the testing of multiple use cases and a more comprehensive test of the chip.
[0056] When determining the performance anomaly corresponding to the chip executing the target use case, the data of the target register is compared with the data in the preset error scenario model, and a comparison analysis is output based on the comparison result, so that the user can understand the reason for the test anomaly based on the analysis report. In addition, compared with manually reading the relevant registers on the chip side and querying the detailed meaning of the register based on the document to locate the error, since the number and type of registers involved in different verification scenarios are not completely consistent, the manual error locating method is inefficient in locating the error point. In the method provided by the present invention, the corresponding registers are automatically acquired through the first server, thereby realizing rapid location of the error.
[0057] When establishing a preset error scenario model, the preset error scenario model is established based on the fact that the tail value and the head value in the queue corresponding to the execution use case are not equal and / or the preset error scenario model is established based on whether the register used to store data when executing the use case is set, and the established error scenario model specifically includes a management command path fault model, an input and output operation command submission queue processing logic fault model, an input and output operation command completion queue processing logic fault model, a reset logic fault model and an interrupt logic fault model, which covers various errors that may occur in chip testing and realizes accurate testing of chip performance.
[0058] The first server, the second server and the third server transmit data according to a preset data format, which ensures the standardization of information interaction between the first server, the second server and the third server and improves the efficiency of data composition and management.
[0059] In addition, the present invention also provides a chip testing device based on a hardware accelerator, a chip testing equipment based on a hardware accelerator, and a computer-readable storage medium, which have the same or corresponding technical features as the above-mentioned chip testing method based on a hardware accelerator, and have the same effects as above. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] In order to more clearly illustrate the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0061] Figure 1 A flowchart of a chip testing method based on a hardware accelerator provided by an embodiment of the present invention;
[0062] Figure 2 A schematic diagram of a chip test through a virtual host provided by an embodiment of the present invention;
[0063] Figure 3 A schematic diagram of a chip testing system based on a hardware accelerator provided by an embodiment of the present invention;
[0064] Figure 4 A flowchart of a verification case automatic scheduling method provided by an embodiment of the present invention;
[0065] Figure 5 A schematic diagram of the operation of an automatic verification platform provided by an embodiment of the present invention;
[0066] Figure 6 A flowchart of an NVME Controller automated verification operation method provided by an embodiment of the present invention;
[0067] Figure 7 A structural diagram of a chip testing device based on a hardware accelerator provided in another embodiment of the present invention. DETAILED DESCRIPTION
[0068] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0069] The core of the present invention is to provide a chip testing method, device, equipment and medium based on a hardware accelerator to solve the technical problem of verifying the NVME controller inside the chip based on the hardware accelerator.
[0070] In order to enable those skilled in the art to better understand the solution of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Figure 1 A flowchart of a chip testing method based on a hardware accelerator provided in an embodiment of the present invention. The method is applied to a first server, and the method includes:
[0071] S10: pre-setting the universal asynchronous receiver / transmitter communication activator in the second server to a transmission control protocol service mode, and establishing a transmission control protocol link or an Internet protocol link with the second server;
[0072] S11: Establishing a transmission control protocol link or an Internet protocol link with a third server; wherein the first server is a platform for chip testing, the second server is a chip side, and the third server is a host side;
[0073] S12: Acquire the running status of the non-volatile memory fast channel controller inside the chip, and perform use case scheduling according to the running status to determine the target use case to be executed by the second server;
[0074] S13: Sending a command representing a target use case to be executed by the second server to the third server, so that the second server executes the target use case according to the command;
[0075] S14: Obtain the current running result of the second server executing the target use case. When it is detected that the current running result is the same as the preset running result of the preset target use case, determine that the corresponding performance of the chip when executing the target use case is normal; otherwise, it is abnormal.
[0076] The first server pre-stores the operation results of the second server executing various use cases.
[0077] Figure 2 A schematic diagram of a chip test through a virtual host provided by an embodiment of the present invention, such as Figure 2 As shown, the entire test environment includes the chip side and the host side. On the host side, manually input the verification case to the virtual host, and then interact with the non-volatile memory fast channel controller on the chip side after passing through the fast peripheral component interconnection communication stimulator. On the chip side, it is also necessary to manually input information to the universal asynchronous receiver / transmitter terminal, and then enter the non-volatile memory fast channel controller after passing through the universal asynchronous receiver / transmitter communication stimulator. When using this method for testing, two verification environments are generated on the host side and the chip side, and real-time information interaction cannot be performed between the environments, thus forming only a unilateral stimulation of the host to the chip. The method of testing the chip by the virtual host is limited by the independence of the two verification environments. During verification, the corresponding verification case needs to be manually run and the running results need to be manually compared. When the chip contains multiple NVME Controllers, this verification method will generate a huge workload, and frequent manual comparisons may further affect the accuracy of the comparison results.
[0078] Therefore, the embodiment of the present invention sets a first server (a platform for chip testing) to test the chip. The second server in the embodiment of the present invention is the chip side, and the third server corresponds to the host side. Figure 3 A schematic diagram of a chip testing system based on a hardware accelerator provided by an embodiment of the present invention is shown in FIG. Figure 3As shown, when the chip test system is working, the automatic verification platform (first server), the chip side (second server) and the host side (third server) exchange information. The host side includes a virtual host (Vhost) and a fast peripheral component interconnection communication activator, and the chip side includes a transmission control protocol server, a universal asynchronous receiver / transmitter communication activator and a non-volatile memory fast channel controller.
[0079] In order to test the chip through the first server, the test environment needs to be built first. That is, the Universal Asynchronous Receiver / Transmitter (UART) communication stimulator (Transactor) is set to the Transmission Control Protocol (TCP) service (Server) mode on the chip side. In this mode, the chip side will provide a TCP / Internet Protocol (IP)-oriented interface and wait for the automatic verification platform to connect; at the same time, a specified listening interface is created in the virtual host on the host side based on the TCP Socket programming specification, and the automatic verification platform is also waiting for the connection. At this point, the chip and the host have been connected to the Ethernet of the automatic verification platform and the basic environment required for the operation of the automatic verification platform has been built.
[0080] After the basic test environment is created, after the transmission control protocol link or the Internet protocol link is established with the second server, and before the transmission control protocol link or the Internet protocol link is established with the third server, the method further includes:
[0081] A first initialization command frame is sent to the second server so that the second server can complete the startup of the relevant clock and reset inside the chip according to the first initialization command frame, and establish a fast peripheral component interconnection link with the virtual host in the third server via the fast peripheral component interconnection communication stimulator; wherein the first initialization command frame includes at least the initialization of the non-volatile memory fast channel controller.
[0082] The above-mentioned process of initializing according to the first initialization command frame is chip-side initialization, which is the first stage of automated verification.
[0083] After the chip side is initialized, the second stage of automated verification, namely, host side initialization, is entered. Specifically, after establishing a transmission control protocol link or an Internet protocol link with the third server and before obtaining the operating status of the non-volatile memory fast channel controller inside the chip, it also includes:
[0084] After detecting that the second server completes the initialization operation, controlling the second server to establish a transmission control protocol link or an Internet protocol link with the virtual host of the third server;
[0085] A second initialization command frame is sent to the third server so that the third server completes loading of the fast peripheral component interconnect device driver.
[0086] After the host side is initialized, the third stage of automated verification is entered, namely, verification scenario identification and use case allocation. In the embodiment of the present invention, the operating status of the non-volatile memory fast channel controller inside the chip is obtained, and use case scheduling is performed according to the operating status to determine the target use case to be executed by the second server.
[0087] In order to enable the second server to execute the target use case, the first server needs to send a command to the third server for representing the target use case to be executed by the second server, so that the second server can execute the target use case according to the command. After that, the current running result of the second server executing the target use case is obtained. If it is detected that the current running result is the same as the preset running result of the preset target use case, it is determined that the corresponding performance of the chip when executing the target use case is normal; otherwise, it is abnormal.
[0088] In order to simulate a more realistic chip operation scenario and improve the reliability of the test results, the first server in the test system pre-stores the operation results of the second server executing multiple use cases.
[0089] It is worth noting that when determining the performance anomaly corresponding to the chip executing the target use case, it is also possible to enter the fourth stage of automated verification, that is, to perform error scenario analysis.
[0090] In the method provided by the embodiment of the present invention, first, the first server sends a command for characterizing the target use case to be executed by the second server to the third server, so that the second server executes the target use case according to the command, thereby realizing the test of chip performance according to the running result of the target use case by the second server. In this process, there is no need to manually input the test case for comparison, thereby realizing the automatic test of chip performance and improving the test efficiency of the chip; secondly, before the second server executes the target use case, the universal asynchronous receiver / transmitter communication activator in the second server is first set to the transmission control protocol service mode, and a transmission control protocol link or an Internet protocol link is established with the second server and the third server respectively, thereby creating a test environment for chip testing and ensuring that the test can proceed normally; thirdly, the running results of the second server executing multiple use cases are pre-stored in the first server, thereby realizing the test of chip performance when the chip executes multiple use cases, thereby simulating a more realistic chip running scenario and improving the reliability of the test results; and, the target use case executed by the second server is determined while considering the running state of the non-volatile memory fast channel controller inside the chip, thereby controlling the second server to execute the target use case to test the chip performance, thereby making the selected target use case more appropriate and the chip performance test more reasonable and accurate.
[0091] In order to determine the target use case to be executed by the second server, in some embodiments, performing use case scheduling according to the running state to determine the target use case to be executed by the second server includes:
[0092] Sending to the second server an instruction for characterizing the operation status of reading the non-volatile memory fast channel controller;
[0093] receiving an operation status sent by the non-volatile memory fast channel controller according to an instruction;
[0094] Determining whether the operation status of the non-volatile memory fast channel controller is normal;
[0095] If so, then when it is detected that the running state of the non-volatile memory fast channel controller is the initial state, determining that the target use case to be executed by the second server is the first use case; wherein the first use case is a basic use case related to an management (Admin) command or an input / output (IO) operation command;
[0096] When it is detected that the running state of the non-volatile memory fast channel controller is a state of completing preliminary configuration and implementing execution of management commands or input / output operation commands, determining that the target use case to be executed by the second server is a second use case; wherein the second use case at least includes a use case for stress testing the chip;
[0097] After determining that the target use case to be executed by the second server is the second use case, the method further includes:
[0098] The use case for chip performance statistics is allocated to the second server to determine that the target use case to be executed by the second server is the use case for chip performance statistics; wherein the use case for chip performance statistics includes at least the number of chip input / output operations per second and bandwidth.
[0099] In addition, when the running state is abnormal, in order to continue to implement the test of the use case, when it is determined that the running state of the non-volatile memory fast channel controller is abnormal, the target use case to be executed by the second server is determined to be a reset use case at each level;
[0100] After detecting that the running state of the non-volatile memory fast channel controller is normal, before determining that the target use case to be executed by the second server is the first use case, or after determining that the target use case to be executed by the second server is a reset use case at each level, it also includes:
[0101] Determine that the target use case to be executed by the second server is an interruption-related use case, and complete interruption enabling and registration;
[0102] In the case where it is detected that the first use case runs abnormally, a global reset use case is allocated to reset the non-volatile memory fast channel controller as a whole, and the process returns to the step of determining that the target use case to be executed by the second server is the first use case;
[0103] When it is detected that the second use case runs abnormally, a global reset use case is allocated to reset the non-volatile memory fast channel controller as a whole, and the process returns to the step of determining that the target use case to be executed by the second server is the second use case.
[0104] In this method, when the second server operates abnormally, a reset case is assigned to the second server and the second server is controlled to execute the reset case, so that the second server can perform subsequent case tests, that is, more chip operation scenarios can be simulated, thereby improving the comprehensiveness and reliability of the test.
[0105] In order to make those skilled in the art better understand the method for automatic scheduling of verification cases of the present invention, Figure 4 The entire process of use case scheduling of the present invention is described. Figure 4 A flowchart of a verification case automatic scheduling method provided by an embodiment of the present invention, such as Figure 4 As shown, the method includes:
[0106] S15: Send a command frame to read the operating status of the non-volatile memory fast channel controller;
[0107] S16: Determine whether the operating status is normal; if not, proceed to step S17; if yes, proceed to step S18;
[0108] S17: Analyze the abnormal state of the non-volatile memory fast channel controller and allocate reset use cases at each level;
[0109] S18: Allocate interrupt-related use cases and complete interrupt enable and registration;
[0110] S19: basic use cases related to allocating management commands or input and output operation commands;
[0111] S20: Determine whether the basic operation is normal; if not, proceed to step S21; if yes, proceed to step S22;
[0112] S21: Allocate a global reset use case to reset the module as a whole;
[0113] S22: Assign management commands / input and output operation commands related random and edge cases;
[0114] S23: Assigns statistics use cases such as the number of input / output operations per second / bandwidth, and enables data statistics;
[0115] S24: Determine whether the pressure operation is normal; if not, proceed to step S25; if yes, end;
[0116] S25: Allocate a global reset use case to reset the module as a whole.
[0117] It can be seen that the above method realizes the allocation and scheduling of use cases according to the operating status of the non-volatile memory fast channel controller.
[0118] When the performance anomaly corresponding to the chip executing the target use case is determined, the fourth stage of automated verification is entered, that is, error scenario analysis is performed. Specifically, when the performance anomaly corresponding to the chip executing the target use case is determined, the method further includes:
[0119] Sending to the second server an instruction for representing data of a target register in the query chip; wherein the target register is a register used by the chip to store data when executing a target use case;
[0120] Receiving data of the target register sent by the second server according to the instruction;
[0121] Compare the data of the target register with the data in the preset error scenario model;
[0122] Output analysis report based on the comparison results.
[0123] Use cases also include the interrupt use cases and reset use cases described above; establishing a preset error scenario model includes:
[0124] A preset error scenario model is established based on the fact that the tail value and the head value in the queue corresponding to the execution use case are not equal and / or based on whether the register used to store data when executing the use case is set.
[0125] More specifically, establishing a preset error scenario model according to the fact that the tail value and the head value in the queue corresponding to the execution use case are not equal includes:
[0126] The management command path fault model is determined based on the inequality between the tail value and the head value in the queue corresponding to the execution of the management command. The queues corresponding to the management command include the Admin Submission Queue (ASQ) and the Admin Completion Queue (ACQ). If the Tail and Head values related to the ASQ / ACQ are unequal, it indicates that there are still commands in the queue that have not been processed, and the ASQ or ACQ channel processing logic can be located.
[0127] According to the fact that the tail value and the head value in the queue corresponding to the execution of the input / output operation command are not equal, if the data of the counting register of the submission queue (I / O Submission Queue, IOSQ) of the input / output operation command are not equal to the data actually sent, then the input / output operation command submission queue processing logic fault model is determined; if the data of the counting register of the submission queue of the input / output operation command is equal to the data actually sent, and is not equal to the data of the counting register of the input / output operation command completion queue (I / O Completion Queue, IOCQ), then the input / output operation command completion queue processing logic fault model is determined;
[0128] The preset error scenario models are established based on whether the registers used to store data are set when executing the use case, including:
[0129] When it is detected that the reset register used to store data is set when executing the reset case, and the value of the register to be reset is not cleared, the reset logic fault model is determined;
[0130] When it is detected that the interrupt status register used to store data is not set when executing the interrupt use case, and the interrupt generation condition is met, the interrupt logic fault model is determined.
[0131] In the method provided in this embodiment, when determining the performance anomaly corresponding to the chip executing the target use case, the data of the target register is compared with the data in the preset error scenario model, and a comparison analysis is output based on the comparison result, so that the user can understand the cause of the test anomaly based on the analysis report. In addition, compared with manually reading the relevant registers on the chip side, querying the detailed meaning of the register based on the document, and then locating the error, since the number and type of registers involved in different verification scenarios are not completely consistent, the manual error locating method is inefficient in locating the error point. In the method provided by the present invention, the corresponding registers are automatically acquired through the first server, thereby achieving rapid location of the error.
[0132] It is worth noting that in the process of interaction between the first server, the second server and the third server described above, in order to improve the standardization of information interaction, the first server, the second server and the third server transmit data according to a preset data format;
[0133] Wherein, the preset data transmission format includes a command frame and a status frame;
[0134] The command frame at least includes the command frame type, the use case type and the use case number; the command frame type at least includes the command frame used to distinguish the environment initialization, the command frame used to verify the use case operation and the command frame used to query the chip status; the use case type at least includes the use case type of input and output operations, the use case corresponding to the management command, the reset use case, the interrupt use case and the statistical use case;
[0135] The status frame includes at least a status frame type, a use case type, a use case number and the number of status information packets; wherein the status frame type includes at least a data frame type used to distinguish normal use case outputs, a data frame type used to distinguish abnormal operation use case outputs and a data frame type used to distinguish chip error state outputs.
[0136] In order to enable those skilled in the art to better understand the format of the command frame and the format of the status frame of the present invention, the command frame and the status frame are further described below.
[0137] 1. Command frame: This frame consists of 32 bits and contains 4 fields: reserved bit, use case type, use case number and command frame type. Table 1 shows the command frame format.
[0138] Table 1
[0139]
[0140] For command frame type: used to distinguish three command frames: environment initialization, verification case running, and chip status query. When the sent frame is an environment initialization command, the remaining fields are set to 1. When it is a verification case running or chip status query command, the remaining fields are effective.
[0141] For use case types: it includes 5 main use case types: IO command, Admin command, reset, interrupt, and statistics.
[0142] For the use case number: it contains the specific verification points involved in the five main use cases, and the chip side, host side and platform identify the specified use case based on the number.
[0143] 2. Status frame: This frame consists of 32 bits and contains five fields: reserved bit, number of status information packets, use case type, scenario type and status frame type. Table 2 shows the status frame format.
[0144] Table 2
[0145]
[0146] For status frame type: used to distinguish between normal operation use case output, abnormal operation use case output and chip error status output, a total of three data frame types.
[0147] For use case types: it includes 5 main use case types: IO command, Admin command, reset, interrupt, and statistics.
[0148] For the use case number: it contains the specific verification points involved in the five main use case types. The chip side, host side and platform identify the specified use case based on the number.
[0149] For the number of status information packets: it indicates how many bytes the status information to be sent subsequently consists of.
[0150] In the method provided in this embodiment, data is transmitted among the first server, the second server and the third server according to a preset data format, thereby ensuring the standardization of information interaction among the first server, the second server and the third server and improving the efficiency of data composition and management.
[0151] In order to enable those skilled in the art to better understand the solution of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Figure 5 A schematic diagram of an automatic verification platform provided by an embodiment of the present invention is shown in FIG. Figure 5 As shown, the automatic verification platform (i.e., the first server) will parse the data frame, and then perform verification case identification and scheduling and data frame framing, and will also perform chip fault identification and fault report output.
[0152] The verification platform combines two independent verification environments into a unified whole through information interaction based on custom data frames, and acts as a node for information interaction to promote the realization of NVME Controller verification automation.
[0153] Figure 6 A flowchart of an NVME Controller automated verification operation method provided by an embodiment of the present invention, such as Figure 6 As shown, the method includes:
[0154] S26: Establishing a transmission control protocol / internet protocol link between the platform and the chip side;
[0155] S27: Sending chip startup and fast peripheral component interconnect bus link establishment related command frames;
[0156] S28: Determine whether the initialization is successful; if not, proceed to step S29; if yes, proceed to step S30;
[0157] S29: Exit the platform and report the error;
[0158] S30: Establishing a transmission control protocol / internet protocol link between the platform and the host side;
[0159] S31: Send a host side initialization command frame to complete the host side initialization;
[0160] S32: Match the verification point execution use case and compare the running results;
[0161] S33: Determine whether the use case is successfully executed; if so, return to step S32; if not, proceed to step S34;
[0162] S34: Send a command frame to query the relevant registers on the chip side;
[0163] S35: Import the error scenario model and output the analysis report.
[0164] like Figure 6 As shown, automated verification can be divided into four stages:
[0165] Phase 1: Chip side initialization: The automatic verification platform will establish a TCP / IP link with the TCP Server on the chip side, and send command frames such as chip basic function initialization and NVME Controller initialization to complete the startup of the relevant clock and reset inside the chip and establish a PCIE link with the Vhost via the PCIE Transactor. At this time, if the platform queries the relevant initialization command successfully, it will enter the second phase, otherwise it will automatically exit the current process and print the error status.
[0166] Phase 2: Host-side initialization: After the platform control chip completes the initialization operation, it will establish a TCP / IP link with the Vhost on the host side. The platform will then send a host-side initialization command frame to complete the loading of the PCIE device driver to generate a device node for the user layer, providing an interface for the subsequent verification use case.
[0167] Phase 3: Verification scenario identification and use case allocation: After completing the initialization work of the first and second phases, the platform will identify verification scenarios and allocate use cases and realize automatic scheduling of all verification use cases, such as Figure 7 As shown, the operation status use case scheduling mechanism based on NVMEController divides the verification scenarios into three types: module basic operation scenario, module abnormal operation scenario, and module stress operation scenario. The three scenarios intersect with each other and the use cases under the scenarios are randomly assigned to further improve the randomness of the use case operation and simulate a more realistic chip operation scenario.
[0168] Abnormal module operation scenario: Since the switching between individual verification use cases is in two mutually exclusive configurations, the NVME Controller will not be able to operate normally. Therefore, the scheduling mechanism will allocate reset use cases at all levels to globally initialize the logic and configuration of the module.
[0169] Basic module operation scenario: At this time, the NVME Controller is in the initial state. To ensure the normal operation of subsequent complex use cases, it is necessary to verify the basic functions and paths. Therefore, the scheduling mechanism will allocate relevant basic use cases.
[0170] Module stress operation scenario: At this time, the NVME Controller has completed the preliminary configuration and can implement the complete Admin / IO command process, so the scheduling mechanism will allocate random, stress use cases and related use cases such as performance statistics at the same time.
[0171] If an error occurs when the verification test case is run, the fourth stage will be entered. Otherwise, this stage will be repeated until all test cases pass the run.
[0172] Phase 4: Error scenario analysis: When the operation fails, the platform will read the relevant registers in the verification scenario, and import the obtained chip operation parameters into the custom fault model for identification and matching, so as to determine the approximate cause of the error, and finally combine the matching results, verification scenario and chip operation parameters into an analysis report and output it to the terminal as the basis for error location. According to the common fault scenarios of NVME Controller, the present invention defines four fault models as follows:
[0173] Admin command path fault model: The inequality between the Tail and Head values of ASQ / ACQ indicates that there are still unprocessed commands in the queue. This can be located in the ASQ or ACQ channel processing logic.
[0174] Reset logic fault model: The reset status register is set, but the register to be reset is not cleared, which can be located in the related reset logic.
[0175] Interrupt logic fault model: The interrupt status register is not set, the interrupt generation condition is met, and the relevant interrupt logic can be located.
[0176] IO command path fault model: The inequality between the IOSQ / IOCQ related Tail and Head values indicates that there are still commands in the queue that have not been processed. If the I / O Completion Queue Entry (IOSQE) count register is not equal to the actual transmission, the IOSQ processing logic can be located. If the IOSQE counter is equal to the actual transmission but the IOCQE counter is not equal to it, the IOCQ processing logic can be located.
[0177] Combined with the above description Figure 3 The automatic verification scheme provided by the present invention uses the automatic verification platform as a node for information exchange between the chip side and the host side. By switching the Uart Transactor to the TCP server mode and using the default virtual network of the host side Vhost, an information exchange channel based on the TCP / IP protocol is constructed, wherein the verification platform, the chip side and the host side will rely on the custom data frame to realize the three-party automatic interaction. The data frame can realize the comprehensive control of the entire verification process by characterizing the verification scenario, status and results. As the core of the automatic verification scheme, the automatic verification platform will automatically identify the current verification scenario and allocate verification cases to the chip side and the host side, and at the same time monitor the running of the case in real time and analyze the results. When an error in the chip behavior is detected, the platform will read and count the changes in the relevant register values in the scenario, and identify the causes of the abnormality by comparing with a variety of fault models and provide relevant configuration and scenario information to the verification personnel, thereby realizing rapid error location.
[0178] In the method provided by the embodiment of the present invention, the chip is automatically verified by the first server, so that the regression verification work of the chip at different stages can be completed more efficiently and more accurate comparison results can be obtained; the real-time operating parameters of the chip can be provided and the approximate cause of the error can be obtained when the use case fails to run, so as to quickly locate the error; through the customized data frame format, the data interaction between the automatic verification platform and the chip side and the host side has higher flexibility and readability, thereby realizing more efficient data processing and management; through the automatic scheduling mechanism of verification use cases, the autonomous identification of verification scenarios and the automatic scheduling of verification use cases are realized, which effectively improves the automation level and efficiency of the regression verification process.
[0179] In summary, the chip testing method provided by the present invention can bring the following effects:
[0180] 1) The two originally independent verification environments in NVME Controller verification are combined together through information interaction of custom data frames, and the automation of use case allocation, operation, and result comparison can be realized through the automated verification platform, thereby greatly improving the efficiency of verification.
[0181] 2) When an error occurs during the verification test case, the verification platform can read the value of the scenario-related register and import it into the fault model for automatic comparison to determine the approximate cause of the error. The verification platform will then present the analysis report to the terminal to facilitate the verification personnel to quickly locate the error.
[0182] 3) The verification platform can implement relevant print statements used in the original use case to display the chip operation status, thereby further reducing the amount of code in the verification use case.
[0183] In the above embodiments, the chip testing method based on the hardware accelerator is described in detail, and the present invention also provides a chip testing device based on the hardware accelerator and a chip testing equipment based on the hardware accelerator. It should be noted that the present invention describes the embodiments of the device part from two perspectives, one is based on the functional module perspective, and the other is based on the hardware perspective.
[0184] A chip testing device based on a hardware accelerator provided by an embodiment of the present invention includes, from the perspective of functional modules:
[0185] A setting and establishing module, used for pre-setting the universal asynchronous receiver / transmitter communication activator in the second server to a transmission control protocol service mode, and establishing a transmission control protocol link or an Internet protocol link with the second server;
[0186] An establishment module is used to establish a transmission control protocol link or an Internet protocol link with a third server; wherein the first server is a platform for chip testing, the second server is a chip side, and the third server is a host side;
[0187] An acquisition and scheduling module, used to acquire the operating status of the non-volatile memory fast channel controller inside the chip, and perform use case scheduling according to the operating status to determine the target use case to be executed by the second server;
[0188] Sending a command, used to send a command representing a target use case to be executed by the second server to the third server, so that the second server executes the target use case according to the command;
[0189] A determination module is used to obtain the current running result of the second server executing the target use case. When it is detected that the current running result is the same as the preset running result of the preset target use case, it is determined that the corresponding performance of the chip when executing the target use case is normal; otherwise, it is abnormal; wherein the first server pre-stores the running results of the second server executing multiple use cases.
[0190] In some embodiments, the chip testing device based on the hardware accelerator further includes:
[0191] A first sending module is used to send an instruction for representing data of a target register in the query chip to the second server; wherein the target register is a register used to store data when the chip executes a target use case;
[0192] A receiving module, used for receiving data of a target register sent by the second server according to an instruction;
[0193] A comparison module, used for comparing the data of the target register with the data in the preset error scenario model;
[0194] The output module is used to output an analysis report based on the comparison results.
[0195] In some embodiments, the chip testing device based on the hardware accelerator further includes:
[0196] The second sending module is used to send a first initialization command frame to the second server, so that the second server completes the startup of the relevant clock and reset inside the chip according to the first initialization command frame, and establishes a fast peripheral component interconnection link with the virtual host in the third server via the fast peripheral component interconnection communication activator; wherein the first initialization command frame at least includes the initialization of the non-volatile memory fast channel controller;
[0197] The chip test device based on hardware accelerator also includes:
[0198] A control module, for controlling the second server to establish a transmission control protocol link or an Internet protocol link with a virtual host of a third server after detecting that the second server has completed the initialization operation;
[0199] The third sending module is used to send a second initialization command frame to the third server so that the third server completes the loading of the fast peripheral component interconnect device driver.
[0200] Since the embodiments of the apparatus part correspond to the embodiments of the method part, please refer to the description of the embodiments of the method part for the embodiments of the apparatus part, which will not be repeated here.
[0201] Figure 7 This is a structural diagram of a chip testing device based on a hardware accelerator provided by another embodiment of the present invention. This embodiment is based on the hardware perspective, such as Figure 7 As shown, the chip testing equipment based on hardware accelerator includes:
[0202] A memory 20, for storing computer programs;
[0203] The processor 21 is used to implement the steps of the method for chip testing based on hardware accelerator mentioned in the above embodiment when executing the computer program.
[0204] Among them, the processor 21 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 21 can be implemented in at least one hardware form of a digital signal processor (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 21 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a central processing unit (CPU); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 21 may be integrated with a graphics processing unit (GPU), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 21 may also include an artificial intelligence (AI) processor, which is used to process computing operations related to machine learning.
[0205] The memory 20 may include one or more computer-readable storage media, which may be non-transitory. The memory 20 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In this embodiment, the memory 20 is at least used to store the following computer program 201, wherein, after the computer program is loaded and executed by the processor 21, it can implement the relevant steps of the chip testing method based on the hardware accelerator disclosed in any of the aforementioned embodiments. In addition, the resources stored in the memory 20 may also include an operating system 202 and data 203, etc., and the storage method may be temporary storage or permanent storage. Among them, the operating system 202 may include Windows, Unix, Linux, etc. Data 203 may include but is not limited to the data involved in the chip testing method based on the hardware accelerator mentioned above.
[0206] In some embodiments, the chip testing device based on the hardware accelerator may further include a display screen 22 , an input and output interface 23 , a communication interface 24 , a power supply 25 , and a communication bus 26 .
[0207] Those skilled in the art will understand that Figure 7 The structure shown in the figure does not constitute a limitation on the chip testing equipment based on the hardware accelerator, and may include more or fewer components than those shown in the figure.
[0208] The chip testing device based on hardware accelerator provided by the embodiment of the present invention includes a memory and a processor. When the processor executes the program stored in the memory, it can implement the following method: the chip testing method based on hardware accelerator has the same effect as above.
[0209] An embodiment of the present invention further provides a computer program product, including a computer program / instruction, which implements the steps of the above-mentioned chip testing method based on hardware accelerator when executed by a processor.
[0210] Finally, the present invention also provides an embodiment corresponding to a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps recorded in the above method embodiment are implemented.
[0211] It is understandable that if the method in the above embodiment is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc. Various media that can store program codes.
[0212] The computer-readable storage medium provided by the present invention includes the above-mentioned chip testing method based on hardware accelerator, and the effect is the same as above.
[0213] The above is a detailed introduction to a chip testing method, device, equipment and medium based on a hardware accelerator provided by the present invention. The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description. It should be pointed out that for ordinary technicians in this technical field, without departing from the principle of the present invention, the present invention can also be improved and modified in a number of ways, and these improvements and modifications also fall within the scope of protection of the present invention.
[0214] It should also be noted that, in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.
Claims
1. A chip testing method based on hardware accelerator, characterized in that: Applied to a first server, the method comprises: Pre-setting the universal asynchronous receiver / transmitter communication activator in the second server to a transmission control protocol service mode, and establishing a transmission control protocol link or an Internet protocol link with the second server; Establishing a transmission control protocol link or an Internet protocol link with a third server; wherein the first server is a platform for chip testing, the second server is a chip side, and the third server is a host side; Acquire the running state of the non-volatile memory fast channel controller inside the chip, and perform use case scheduling according to the running state to determine the target use case to be executed by the second server; Sending a command representing a target use case to be executed by the second server to the third server, so that the second server executes the target use case according to the command; Obtain the current running result of the second server executing the target use case, and when it is detected that the current running result is the same as the preset running result of the pre-set target use case, determine that the corresponding performance of the chip when executing the target use case is normal; otherwise, it is abnormal; wherein the first server pre-stores the running results of the second server executing multiple use cases.
2. The chip testing method based on hardware accelerator according to claim 1, characterized in that: The performing use case scheduling according to the running state to determine the target use case to be executed by the second server comprises: Sending to the second server an instruction for characterizing reading the operating status of the non-volatile memory fast channel controller; receiving the operating status sent by the non-volatile memory fast channel controller according to an instruction; Determining whether the operating state of the non-volatile memory fast channel controller is a normal state; If so, then when it is detected that the running state of the non-volatile memory fast channel controller is the initial state, determining that the target use case to be executed by the second server is the first use case; wherein the first use case is a basic use case related to a management command or an input / output operation command; When it is detected that the running state of the non-volatile memory fast channel controller is a state of completing preliminary configuration and implementing execution of management commands or input / output operation commands, determining that the target use case to be executed by the second server is a second use case; wherein the second use case at least includes a use case for stress testing the chip; After determining that the target use case to be executed by the second server is the second use case, the method further includes: The use case for performing performance statistics on the chip is allocated to the second server to determine that the target use case to be executed by the second server is the use case for performing performance statistics on the chip; wherein the use case for performing performance statistics on the chip includes at least the number of input / output operations per second and the bandwidth of the chip.
3. The chip testing method based on hardware accelerator according to claim 2, characterized in that: In the case where it is determined that the performance of the chip corresponding to the target use case is abnormal, the method further includes: Sending to the second server an instruction for querying data of a target register in the chip; wherein the target register is a register used by the chip to store data when executing the target use case; receiving the data of the target register sent by the second server according to the instruction; Comparing the data of the target register with the data in a preset error scenario model; Output analysis report based on the comparison results.
4. The chip testing method based on hardware accelerator according to claim 3, characterized in that: Use cases also include interrupt use cases and reset use cases; Establishing the preset error scenario model includes: The preset error scenario model is established based on the fact that the tail value and the head value in the queue corresponding to the execution use case are not equal and / or the preset error scenario model is established based on whether the register used to store data when executing the use case is set.
5. The chip testing method based on hardware accelerator according to claim 4, characterized in that: The establishing of the preset error scenario model according to the fact that the tail value and the head value in the queue corresponding to the execution use case are not equal includes: Determine the management command path fault model according to the fact that the tail value and the head value in the queue corresponding to the execution management command are not equal; According to the fact that the tail value and the head value in the queue corresponding to the execution of the input / output operation command are not equal, if the data of the counting register of the submission queue of the input / output operation command is not equal to the data actually sent, then the input / output operation command submission queue processing logic fault model is determined; if the data of the counting register of the submission queue of the input / output operation command is equal to the data actually sent, and is not equal to the data of the counting register of the input / output operation command completion queue, then the input / output operation command completion queue processing logic fault model is determined; The establishing of the preset error scenario model according to whether the register used to store data when executing the use case is set includes: When it is detected that the reset register used to store data is set when executing the reset case, and the value of the register to be reset is not cleared, the reset logic fault model is determined; When it is detected that the interrupt status register used to store data is not set when executing the interrupt use case, and the interrupt generation condition is met, the interrupt logic fault model is determined.
6. The chip testing method based on hardware accelerator according to any one of claims 1 to 5, characterized in that: After establishing the transmission control protocol link or the Internet protocol link with the second server and before establishing the transmission control protocol link or the Internet protocol link with the third server, the method further includes: Sending a first initialization command frame to the second server, so that the second server completes the startup of the relevant clock and reset inside the chip according to the first initialization command frame, and establishes a fast peripheral component interconnection link with the virtual host in the third server via the fast peripheral component interconnection communication activator; wherein the first initialization command frame at least includes the initialization of the non-volatile memory fast channel controller; After establishing the transmission control protocol link or the Internet protocol link with the third server and before acquiring the operating status of the non-volatile memory fast channel controller in the chip, the method further includes: After detecting that the second server completes the initialization operation, controlling the second server to establish a transmission control protocol link or an Internet protocol link with the virtual host of the third server; A second initialization command frame is sent to the third server so that the third server completes loading of a fast peripheral component interconnect device driver.
7. The chip testing method based on hardware accelerator according to claim 6, characterized in that: The first server, the second server and the third server transmit data according to a preset data format; Wherein, the preset data transmission format includes a command frame and a status frame; The command frame at least includes a command frame type, a use case type and a use case number; the command frame type at least includes a command frame for distinguishing environment initialization, a command frame for verifying use case operation and a command frame for querying chip status; the use case type at least includes a use case type for input and output operations, a use case corresponding to a management command, a reset use case, an interrupt use case and a statistical use case; The status frame includes at least a status frame type, a use case type, a use case number and the number of status information packets; wherein the status frame type includes at least a data frame type used to distinguish normal use case outputs, a data frame type used to distinguish abnormal operation use case outputs and a data frame type used to distinguish chip error state outputs.
8. A chip testing device based on hardware accelerator, characterized in that: Applied to a first server, the device comprises: A setting and establishing module, used for pre-setting the universal asynchronous receiver / transmitter communication activator in the second server to a transmission control protocol service mode, and establishing a transmission control protocol link or an Internet protocol link with the second server; An establishing module, used to establish a transmission control protocol link or an Internet protocol link with a third server; wherein the first server is a platform for chip testing, the second server is a chip side, and the third server is a host side; An acquisition and scheduling module, used for acquiring the operating status of the non-volatile memory fast channel controller inside the chip, and performing use case scheduling according to the operating status to determine a target use case to be executed by the second server; Sending a command, used to send a command representing a target use case to be executed by the second server to the third server, so that the second server executes the target use case according to the command; A determination module is used to obtain the current running result of the second server executing the target use case, and when it is detected that the current running result is the same as the preset running result of the target use case, determine that the corresponding performance of the chip when executing the target use case is normal; otherwise, it is abnormal; wherein the first server pre-stores the running results of the second server executing multiple use cases.
9. A chip testing device based on a hardware accelerator, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the steps of the chip testing method based on hardware accelerator as claimed in any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the chip testing method based on the hardware accelerator according to any one of claims 1 to 7 are implemented.
Citation Information
Cited By
Hardware detection method, device and equipment for distributed power supply access unit and medium
CN119316500A
PCIe function test method and device, medium and product
CN119862077A
Hard disk test method and device, storage medium and program product
CN120560918A
