A performance monitoring method for integrated circuit simulation
By generating interface performance monitors and atomic probe instances, basic performance statistics and data aggregation are performed, solving the problems of intrusiveness and functional limitations in existing integrated circuit simulation performance monitoring, and realizing efficient monitoring and flexible expansion of multi-level performance data.
Patent Information
- Application Number
- CN202611427589.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-15
- Publication Date
- 2026-11-17
AI Technical Summary
Existing integrated circuit simulation performance monitoring methods are intrusive, coupled, and have functional limitations. They are difficult to achieve systematic correlation analysis of multi-level performance data, and have low code reuse rates, making it impossible to flexibly expand observation points.
By obtaining the logical path corresponding to the target protocol, an interface performance monitor instance and an atomic probe instance are generated. Basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization are performed. The logical path decomposition, multi-level instance generation, and three-level statistical process are adopted to achieve code reuse and expansion flexibility for performance monitoring.
It enables efficient monitoring of multi-level performance data, improves the scalability and code reusability of performance monitoring, and supports correlation analysis of multi-level performance data.
Smart Images

Figure CN123117010A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of integrated circuit technology, and in particular to a performance monitoring method for integrated circuit simulation. Background Technology
[0002] In integrated circuit design verification, performance verification is a key step to ensure that the chip functions correctly and meets performance indicators, while performance monitoring is a necessary prerequisite for performance verification.
[0003] Currently, existing performance monitoring methods for integrated circuit simulation typically involve capturing bus signals through software models in the verification platform and performing post-processing statistics, or instantiating dedicated hardware performance counter registers in the register transfer level design and reading the registers through firmware to obtain statistical data. However, software monitoring schemes in verification platforms suffer from intrusiveness and coupling issues. Dedicated performance statistical logic needs to be written for each bus protocol, requiring independent implementation for different protocols, resulting in low code reusability. Furthermore, the monitoring logic is tightly coupled with protocol details; when the interface or protocol is modified in the design, the performance monitoring code in the verification platform also requires extensive modification. Secondly, hardware performance counter schemes have functional limitations. On-chip hardware performance counters typically only provide simple event counting functions, and the number and type of performance counters are fixed in the design phase, making it difficult to flexibly expand the observation points. Finally, existing schemes usually only focus on single-level performance indicators, lacking a systematic correlation analysis method from micro-interface indicators to macro-system throughput, making it difficult to simultaneously acquire and correlate multi-level performance data in a single simulation. Summary of the Invention
[0004] This invention provides a performance monitoring method for integrated circuit simulation, which enables code reuse for performance monitoring, improves the flexibility of performance monitoring expansion, and enables efficient monitoring of multi-level performance data.
[0005] According to one aspect of the present invention, a performance monitoring method for integrated circuit simulation is provided, comprising: Obtain each logical path corresponding to the target protocol, and generate an interface performance monitor instance corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance; When a performance monitoring trigger for the target protocol is detected, the performance monitoring interval is obtained; Based on the performance monitoring interval, basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization are sequentially performed through each interface performance monitor instance and the corresponding atomic probe instance to obtain performance monitoring data at each level.
[0006] According to another aspect of the present invention, a performance monitoring device for integrated circuit simulation is provided, comprising: The instance generation module is used to obtain each logical path corresponding to the target protocol, and generate an interface performance monitor instance corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance. The interval acquisition module is used to acquire the performance monitoring interval when a performance monitoring trigger for the target protocol is detected. The data acquisition module is used to perform basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization sequentially through each of the interface performance monitor instances and the corresponding atomic probe instances according to the performance monitoring interval, so as to obtain performance monitoring data at all levels.
[0007] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the performance monitoring method for integrated circuit simulation as described in any embodiment of the present invention.
[0008] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program for causing a processor to execute and implement the performance monitoring method for integrated circuit simulation as described in any embodiment of the present invention.
[0009] According to another aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the performance monitoring method for integrated circuit simulation as described in any embodiment of the present invention.
[0010] The technical solution of this invention obtains each logical path corresponding to the target protocol and generates interface performance monitor instances corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance. When a performance monitoring trigger for the target protocol is detected, a performance monitoring interval is obtained. Based on the performance monitoring interval, basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization are sequentially executed through each interface performance monitor instance and its corresponding atomic probe instance to obtain performance monitoring data at all levels. By adopting logical path decomposition, multi-level instance generation, and a three-level statistical process, performance monitoring code reuse can be achieved, the scalability of performance monitoring can be improved, and efficient monitoring of multi-level performance data can be achieved.
[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description
[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0013] Figure 1 This is a flowchart of a performance monitoring method for integrated circuit simulation provided in Embodiment 1 of the present invention; Figure 2 This is a flowchart of a performance monitoring method for integrated circuit simulation provided in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the structure of a performance monitoring device for integrated circuit simulation provided in Embodiment 3 of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device that implements the performance monitoring method for integrated circuit simulation according to an embodiment of the present invention. Detailed Implementation
[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0015] It should be noted that the terms "first," "second," "target," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0016] Example 1 Figure 1 This is a flowchart of a performance monitoring method for integrated circuit simulation provided in Embodiment 1 of the present invention. This embodiment is applicable to the situation of performing layered composable performance monitoring of integrated circuit simulation. The method can be executed by a performance monitoring device for integrated circuit simulation, which can be implemented in hardware and / or software. Typically, the performance monitoring device for integrated circuit simulation can be configured in electronic devices, such as computer equipment, servers, etc. Figure 1 As shown, the method includes: S110. Obtain each logical path corresponding to the target protocol, and generate an interface performance monitor instance corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance.
[0017] In this invention, the performance monitoring system comprises three core layers: atomic performance probes (L0 layer), interface performance monitors (L1 layer), and protocol performance monitors (L2 layer). The atomic performance probe is the smallest functional unit for performance monitoring. Each atomic performance probe receives two event inputs (input event signal and output event signal) and one event weight input, deriving multiple statistical quantities from these three basic signals, including the total number of event occurrences, the weighted cumulative value of events and its maximum value, the initial event delay, the cumulative delay and maximum delay, and the average delay. The atomic performance probe is protected by conditional compilation, participating in functional behavior during simulation and automatically eliminating itself as a constant zero output during synthesis, achieving zero area and zero timing impact on the design.
[0018] The interface performance monitor instantiates multiple atomic probe instances on the same interface signal and employs different binding strategies for the input and output events of each atomic probe instance. This allows the same atomic performance probe to act as: an outstanding tracker, a stall counter, a bandwidth meter, a cycle counter, and a latency analyzer in different instantiations. The complete interface-level performance metrics are then obtained through the arithmetic combination of the outputs of the atomic probe instances. Here, the atomic performance probe can be considered a general template (module definition), and the atomic probe instance can be considered a specific running entity (instantiated object).
[0019] The protocol performance monitor decomposes the complex bus protocol into multiple independent logical paths (request-response channels) through semantic analysis, and instantiates an interface performance monitor instance for each logical path to achieve full-channel performance monitoring of the complex bus protocol.
[0020] Specifically, firstly, the target protocol is semantically decomposed using a protocol performance monitor to obtain multiple logical request-response paths, which serve as logical paths. Then, an interface performance monitor instance is instantiated for each logical path, and multiple atomic probe instances are instantiated for each interface performance monitor instance. Each interface performance monitor instance can employ different input and output event binding strategies, which can be pre-defined.
[0021] S120. When a performance monitoring trigger for the target protocol is detected, the performance monitoring interval is obtained.
[0022] Specifically, if a preset performance monitoring trigger condition is detected, such as a reset release or the receipt of an enable signal, performance monitoring of the target protocol is triggered. Then, based on preset interval parameters or real-time monitoring start and stop signals, the current performance monitoring interval is determined.
[0023] Optionally, the performance monitoring interval can include: Obtain the current time window mode, and based on the current time window mode, obtain the performance monitoring interval; The time window modes include manual control mode, first input event to last output event mode, first input event to last input event mode, and first output event to last output event mode.
[0024] In this embodiment, a configurable time window mechanism is supported to determine the performance monitoring interval. Typically, four time window modes are supported, selected via a two-bit selection signal. 00 represents the manual control mode, corresponding to a performance monitoring interval of the number of cycles from the external start signal to the external stop signal; 01 represents the mode from the first input event to the last output event, corresponding to a performance monitoring interval of the number of cycles from the first input event to the last output event; 10 represents the mode from the first input event to the last input event, corresponding to a performance monitoring interval of the number of cycles from the first input event to the last input event; and 11 represents the mode from the first output event to the last output event, corresponding to a performance monitoring interval of the number of cycles from the first output event to the last output event. Mode 00 is suitable for scenarios where precise control of the measurement window is required in the verification environment, while modes 01, 10, and 11 are suitable for scenarios where the measurement range is naturally defined by the events themselves.
[0025] S130. Based on the performance monitoring interval, perform basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization sequentially through each interface performance monitor instance and the corresponding atomic probe instance to obtain performance monitoring data at each level.
[0026] Specifically, within the performance monitoring interval, basic performance statistics are first performed on each atomic probe instance to obtain basic performance indicators. Then, each interface performance monitor instance aggregates the basic performance indicators of its corresponding atomic probe instances at the interface level to obtain interface-level performance indicators. Next, protocol-level data is summarized for each interface-level performance indicator to obtain protocol-level performance statistics. Finally, the basic performance indicators, interface-level performance indicators, and protocol-level performance statistics are combined as performance monitoring data. The performance indicators at each level, as well as the corresponding indicator aggregation and data summarization methods, can be pre-defined.
[0027] Optionally, based on the performance monitoring interval, basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization are sequentially performed through each of the interface performance monitor instances and the corresponding atomic probe instances to obtain performance monitoring data at each level, which may include: Within the performance monitoring interval, each atomic probe instance statistically obtains incomplete transaction data, response latency data, request activity cycle data, backpressure cycle data, bus bandwidth data, and effective byte bandwidth data based on pre-configured input events, output events, and weight data. By using each of the interface performance monitor instances, arithmetic combinations are performed based on the statistical results of the corresponding atomic probe instances to obtain interface-level performance indicators, and the interface-level performance indicators are summarized to obtain protocol-level performance statistics.
[0028] In this embodiment, for each interface performance monitor instance, five atomic probe instances can be instantiated, each employing a different input, output, and decrement event binding strategy. The first atomic probe instance is used for incomplete transaction tracking and latency measurement: it binds a valid request handshake (the logical AND of a valid request signal and a ready request signal) as an input event (weight 1), a valid response handshake as an output event, and a valid response handshake as a decrement event (weight 1). The weighted accumulator register of the atomic probe instance naturally reflects the real-time number of incomplete transactions, with the maximum weighted value representing the peak number of incomplete transactions. The pairing of input and output event queues provides response latency statistics. Thus, incomplete transaction data and response latency data can be statistically obtained.
[0029] The second atomic probe instance is used for request active cycle statistics: it binds a valid request signal (regardless of the ready signal state) as an input event, and sets the time window mode to the first input event to the last input event. The total number of events counted in this case is the total number of valid request cycles. Thus, request active cycle data can be obtained. The third atomic probe instance is used for backpressure (stall) cycle statistics: it binds the condition of a high valid request signal but a low ready signal as an input event. The total number of events counted in this case is the total number of backpressure cycles. Combining the active cycle count from the second atomic probe instance, the backpressure percentage can be calculated. Thus, backpressure cycle data can be obtained.
[0030] The fourth atomic probe instance is used for bus bandwidth statistics: a valid handshake is bound as an input event, with the weight being the number of bus bytes transmitted in each transfer. The weighted accumulator register then provides the total number of bytes transmitted, which, combined with the performance monitoring interval and clock frequency parameters, allows calculation of the actual bandwidth. This provides the bus bandwidth data. The fifth atomic probe instance is used for effective byte bandwidth statistics: similar to the fourth instance, but the weight is changed to the actual number of effective bytes (calculated using the byte enable mask). This is used to calculate the actual data throughput after deducting invalid bytes, thus obtaining the effective byte bandwidth data.
[0031] After obtaining incomplete transaction data, response latency data, request activity cycle data, backpressure cycle data, bus bandwidth data, and effective byte bandwidth data, these statistical results can be further combined arithmetically and summarized to obtain interface-level performance indicators and protocol-level performance statistics.
[0032] Optionally, the incomplete transaction data can be statistically obtained by each of the atomic probe instances based on pre-configured input events, output events, and weight data, which may include: Through each of the aforementioned atomic probe instances, the values of the event count register, weighted accumulation register, and maximum value register are updated according to the pre-configured input events, output events, and weight data to obtain the total number of transactions, the number of incomplete transactions, and the historical maximum number of incomplete transactions. Based on the total number of transactions, the number of incomplete transactions, and the historical maximum number of incomplete transactions, the incomplete transaction data is obtained.
[0033] In this embodiment, the atomic probe instance supports an event counting and weighted accumulation mechanism. Internally, the atomic probe instance maintains an event count register, a weighted accumulation register, and a maximum value register. Whenever an input event is valid, the event count register is incremented by one, and the weighted accumulation register accumulates the current event weight. The event weight is determined based on the weight data. Simultaneously, the atomic probe instance also supports a decrement event input; when the decrement event is valid, the corresponding decrement weight is subtracted from the weighted value. When the input event and the decrement event are valid simultaneously in the same period, a one-time addition / subtraction operation is performed. Furthermore, through the maximum value register, the atomic probe instance can also track the historical maximum value of the weighted value in real time. This addition / decrement mechanism allows a single atomic probe instance to track the number of incomplete transactions in real time, using request events as the increment input (weight 1) and response events as the decrement input (weight 1), the weighted accumulation register naturally reflects the current number of incomplete transactions.
[0034] Specifically, by reading the real-time values of the event count register, weighted accumulator register, and maximum value register, the total number of transactions, the number of incomplete transactions, and the historical maximum number of incomplete transactions are obtained respectively, and the obtained quantities are used together as the incomplete transaction data.
[0035] Optionally, by statistically analyzing the response latency data of each of the atomic probe instances based on pre-configured input events, output events, and weight data, the following may be included: The timestamps of the input and output events are recorded in the input event queue and the output event queue, respectively. When the current output event is triggered, the first timestamp of the current output event is obtained from the output event queue, and the second timestamp of the current input event corresponding to the current output event is obtained from the input event queue. The single event delay is calculated based on the first occurrence timestamp and the second occurrence timestamp, and the response delay data is obtained based on the single event delay.
[0036] In this embodiment, the atomic probe instance also supports a paired delay measurement mechanism. Internally, the probe uses two dynamic queues to record the timestamps (values of an internal tick counter) of input and output events, respectively. When an output event occurs, a timestamp is popped from both the output and input event queues, and the difference is calculated to obtain the processing delay for a single event. Next, the delays of all events are summed to obtain the total delay, and the maximum delay and the first event delay are tracked. The average delay is calculated by dividing the total delay by the total number of events. The delay measurement function can be independently controlled via an enable signal. Finally, the single event delay, total delay, maximum delay, first event delay, and average delay are combined to form the response delay data.
[0037] Optionally, the technical solution of this embodiment of the invention may further include: Each interface performance monitor instance establishes a counter corresponding to each transaction identifier; the counter is incremented by one when the request handshake of the corresponding transaction identifier is successful, and decremented by one when the response handshake of the corresponding transaction identifier is successful. Based on the count values of each counter, obtain the number of identifiers of concurrently existing incomplete transactions; Each of the interface performance monitor instances contains embedded protocol compliance assertions, including at least one of the following: no backpressure on the response channel, integrity of request and response pairing by transaction identifier, and read / write return response specifications.
[0038] In this embodiment, the interface performance monitor instance may also include incomplete transaction tracking logic based on transaction identifiers. Typically, an independent counter is maintained for each possible transaction identifier. When a request handshake for a transaction identifier is successful, its corresponding counter is incremented by one, and when the response handshake for that transaction identifier is successful, the counter is decremented by one. By reading the count values of each counter, the number of different transaction identifiers with concurrent incomplete transactions is counted to obtain the number of identifiers with concurrent incomplete transactions (the number of inflight IDs), and the historical maximum value of this number is tracked. Specifically, if the count value of the counter corresponding to a transaction identifier is not 0, it is determined that it has an incomplete transaction.
[0039] Secondly, the interface performance monitor instance also embeds protocol compliance assertions, including but not limited to: no backpressure on the response channel, integrity of request and response pairing by transaction identifier, and read / write return response specifications. Among these, the protocol compliance assertions are real-time protocol behavior verification logic implemented using a hardware description language in digital integrated circuit simulation verification. By setting protocol compliance assertions, performance monitoring and protocol compliance checks can be integrated.
[0040] The technical solution of this invention obtains each logical path corresponding to the target protocol and generates interface performance monitor instances corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance. When a performance monitoring trigger for the target protocol is detected, a performance monitoring interval is obtained. Based on the performance monitoring interval, basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization are executed sequentially through each interface performance monitor instance and its corresponding atomic probe instance to obtain performance monitoring data at all levels. By adopting logical path decomposition, multi-level instance generation, and a three-level statistical process, performance monitoring code reuse can be achieved, the scalability of performance monitoring can be improved, and efficient monitoring of multi-level performance data can be achieved.
[0041] Example 2 Figure 2This is a flowchart illustrating a performance monitoring method for integrated circuit simulation provided in Embodiment 2 of the present invention. This embodiment is a further refinement of the above technical solution, and the technical solution in this embodiment can be combined with one or more of the above implementation methods. Figure 2 As shown, the method includes: S210. Obtain each logical path corresponding to the target protocol, and generate an interface performance monitor instance corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance.
[0042] In a specific example, the target protocol is the Advanced eXtensible Interface (AXI) bus protocol, which defines five independent transmission channels: Write Address Channel (AW), Write Data Channel (W), Write Response Channel (B), Read Address Channel (AR), and Read Data Channel (R). By semantically decomposing this channel, four logical request-response paths are obtained, and an interface performance monitor instance is instantiated for each logical path to achieve full channel performance coverage of the AXI bus. Each interface performance monitor instance can further instantiate five corresponding atomic probe instances, thus allowing for the deployment of 20 atomic probe instances for each observation point.
[0043] The first interface performance monitor instance monitors the write address channel: it binds the valid or ready handshake of the AW channel as a request event (input event) and the valid or ready handshake of the B channel as a response event (output event), and enables transaction identifier tracking (ID_CHECK=1), thereby simultaneously obtaining full-dimensional indicators such as the number of write requests, backpressure statistics of the write address channel, the number of incomplete write transactions and tracking by transaction identifier, and write response latency (end-to-end latency from the AW handshake to the B handshake). The second interface performance monitor instance monitors the write data channel: it binds the valid or ready handshake of the W channel as a request event and the handshake of the B channel as a response event, disables latency measurement (because the latency of the W channel is statistically analyzed by the AW channel instance), but enables bandwidth statistics—it binds the byte enable mask (wstrb) width of the W channel to the bus byte weight, and binds the actual number of effective bytes (calculated by popcount(wstrb), where popcount represents a set count operation) to the effective byte weight, so as to obtain the bus bandwidth and effective data bandwidth respectively.
[0044] The third interface performance monitor instance monitors the read address channel: It binds a valid or ready handshake for the AR channel as a request event, and binds a valid handshake for the R channel with rlast (the R channel burst end marker signal) high as a response event (i.e., the last data beat of the entire burst transmission is used as a response completion marker for a transaction). It enables transaction identifier tracking to obtain the number of read requests, the number of incomplete read transactions, tracking by transaction identifier, and read response latency. The fourth interface performance monitor instance monitors the read data channel: It binds a valid or ready handshake for the R channel as a request event, disables latency measurement, and only counts read data bandwidth.
[0045] The four examples above share the same set of probe control signals, such as enable, frequency parameters, and end signals, and share a result port containing over 30 performance metrics. The entire functional module is wrapped by a conditional compilation guard; during synthesis, all output ports are assigned a constant value of zero, having no impact on chip area or timing. Furthermore, the AXI performance monitor also embeds protocol compliance assertions, such as checking whether the BUSER field (user-defined extended signal field of channel B) of the write response channel conforms to the expected specifications. Performance monitoring and protocol compliance checks are integrated into the same module, avoiding the extra workload of developing a separate protocol checker.
[0046] It should be noted that the AXI performance monitor, as a typical example of an L2 layer protocol performance monitor, can be extended to other bus protocols: simply decompose the channel structure of the target protocol into request-response paths and specify the event binding strategy of the L1 layer interface performance monitor for each path to quickly build a new protocol performance monitor.
[0047] S220. When a performance monitoring trigger for the target protocol is detected based on the design embedded mode or the verification platform control mode, the performance monitoring interval is obtained.
[0048] This embodiment provides a dual-mode observation architecture that simultaneously supports both design-embedded and verification platform control. In design-embedded mode, the atomic performance probe and its combined monitor are directly instantiated in the register-transfer level design source code, located inside the conditional compilation guard. In auto-enable mode, the atomic probe instance automatically begins statistics after reset and release, without any intervention from the verification platform. Design engineers can directly observe the output port signals of the atomic probe instance through a waveform viewer, such as the real-time number of incomplete transactions and cumulative latency, and perform online performance debugging. In verification platform control mode, the verification platform can actively send corresponding control signals to the atomic probe instance and the interface performance monitor instance to trigger performance monitoring.
[0049] Among them, the automatic enable mode and the controlled mode mentioned later are the enable mode dimensions of the atom probe instance. The enable mode can be configured in both observation modes.
[0050] S230. Based on the performance monitoring interval, perform basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization sequentially through each interface performance monitor instance and the corresponding atomic probe instance to obtain performance monitoring data at each level.
[0051] S240. In the embedded design mode, a simulation report is generated based on performance monitoring data at all levels. In the verification platform control mode, an interface log is generated based on performance monitoring data at all levels.
[0052] In this embodiment, a dual-output mechanism is supported, allowing statistical results from the same atom probe instance to be output simultaneously through two channels: a simulation report printed to the simulation log for direct review by design engineers, and an interface log appended for automated post-processing and performance regression comparison by verification engineers. The data from both channels originates from the same atom probe instance to ensure data consistency. For example, in the design-embedded mode, when the simulation ends (triggered by a termination signal), the atom probe instance can automatically output a formatted statistical summary to the simulation log via the simulation report mechanism.
[0053] Furthermore, for the verification platform control mode, a parameterizable macro definition can be used to establish a forced binding between the verification interface and the control and result ports of the internal instance of the design at the top level of the verification platform. This connection mechanism forces the control signals of the verification interface (such as enable, window start / stop, end, etc.) to the corresponding input ports, and simultaneously forces the result ports of the instance to the result signals of the verification interface. The verification interface is distributed to the corresponding performance monitoring application interface instances through a unified configuration database. Each performance monitoring application interface instance provides the following task interfaces: Enable control task—selects the time window mode and activates the atomic probe instance; Window control task—precisely controls the start and end times of the measurement in manual control mode; Result acquisition task—waits for the atomic probe instance to output a valid flag, reads all statistical result data, and writes it to the interface log.
[0054] Finally, the verification environment component automatically performs termination and result acquisition operations on all registered performance monitoring application interface instances during the final stage of the simulation, ensuring the complete output of performance monitoring data. The number of performance monitoring application interface instances in the verification environment component can be dynamically adjusted through configuration parameters, supporting simultaneous monitoring of performance data from multiple different observation points in a single simulation.
[0055] Optionally, the technical solution of this embodiment of the invention may further include: In the verification platform control mode, multiple performance monitoring application interface instances are established, and the verification interface associated with each performance monitoring application interface instance is obtained based on the configuration database. By forcibly binding macros, a forced connection is established between the verification interface and the control port and result port of each of the atomic probe instances. Through the forced connection, an enable signal is sent to the control port of each of the atomic probe instances to trigger performance monitoring.
[0056] In this embodiment, a set of performance monitoring application programming interface components and their automated data collection mechanism for verification platforms are provided. These components expose the instance control and result access rights embedded in the design to the verification environment, thereby enabling precise control of the performance measurement process and automated result collection.
[0057] The performance monitoring application programming interface (API) system of the verification platform can include a verification interface (perfmon_if) component, a forced binding macro (PERFMON_INTF_BIND) component, a performance monitoring API component (perfmon_api) component, and a performance monitoring environment component (perfmon_env). The verification interface component defines a set of signals corresponding one-to-one with instance control and result ports, including enable signals, latency measurement enable, time window mode selection, window start / end control, end signals, and output signals for all statistical results, such as time window length, initial latency, cumulative latency, total number of events, cumulative event value, and validity flags. The forced binding macro component defines a parameter that accepts the hierarchical path of an instance within the design and a verification interface. It uses SystemVerilog's `force` statement to force the control signals of the verification interface to the instance's input ports and simultaneously force the instance's result ports to the result signals of the verification interface. This mechanism enables the verification environment to have complete control over atomic probe instances and interface performance monitor instances without modifying the design code, and distributes the verification interface to the corresponding performance monitoring application interface instance through the configuration database of the Universal Verification Methodology (UVM) (such as uvm_config_db).
[0058] In this embodiment, the performance monitoring environment component can dynamically create N performance monitoring application interface instances (N can be configured via the API_NUM parameter) during the construction phase according to preset configuration parameters, to support simultaneous monitoring of performance data at N different observation points in a single simulation. Next, the association between the performance monitoring application interface instances and the verification interface can be determined based on a preset configuration database. Then, using SystemVerilog's force statement, a forced connection is established between the verification interface and the control port and result port of each corresponding atom probe instance. Subsequently, each performance monitoring application interface instance sends an enable signal to the control port of each corresponding atom probe instance based on the corresponding forced connection to trigger performance monitoring of the target protocol.
[0059] It should be noted that the Atom Probe instance supports dual-mode enable control. In automatic mode, the Atom Probe instance automatically begins statistics after reset release, requiring no external control signal; while in controlled mode, the Atom Probe instance is controlled by an external enable signal to start and stop. Automatic mode is suitable for designing embedded default observations, while controlled mode is suitable for verifying precise control of the environment. Secondly, the Atom Probe instance supports result output and reporting mechanisms. When the external end signal is valid, the Atom Probe instance latches all statistical results to the result port and sets the valid flag, while simultaneously printing a formatted statistical summary to the simulation log via the simulation report task. If any counter overflows, an error report is output.
[0060] Optionally, after establishing a forced connection between the verification interface and the control port and result port of each of the atomic probe instances through forced binding macros, the following may also be included: A time window mode selection signal is generated through the verification interface associated with each of the performance monitoring application interface instances, and the time window mode selection signal is sent to each of the atom probe instances through the forced connection corresponding to each of the verification interfaces. When the time window mode selection signal corresponds to the manual control mode, a window start control signal and a window end control signal are generated through the verification interface associated with each of the performance monitoring application interface instances, and the window start control signal and the window end control signal are sent to each of the atomic probe instances through the forced connection corresponding to each of the verification interfaces. The performance monitoring interval is obtained using each of the aforementioned atomic probe instances, based on the window start control signal and the window end control signal.
[0061] In this embodiment, the performance monitoring application interface instance can be configured with the following task interfaces: perfmon_en(mode): used to enable the atomic probe instance and select one of the four time window modes to activate delay measurement; perfmon_window_start(): used to issue a window start pulse in manual control mode to accurately mark the start time of the measurement; perfmon_window_stop(): used to issue a window stop pulse to accurately mark the end time of the measurement; perfmon_end(): used to trigger the end signal of the atomic probe instance to latch the statistical results; perfmon_result(): used to wait for the probe output valid flag to be set, read all statistical results, such as time window length, initial delay, cumulative delay, total number of events, cumulative event value, etc., and write them to the structured log file in an append-only manner. Each acquisition includes complete hierarchical path identifiers and all statistical indicators.
[0062] Specifically, firstly, the performance monitoring environment component concurrently calls the task interface `perfmon_en(mode)` of each performance monitoring application interface instance. This controls the performance monitoring application interface instance to send time window mode selection signals to the input ports of each corresponding atom probe instance through the corresponding verification interface and forced connection, in order to select the current time window mode. When the current time window mode is manual control mode, the performance monitoring environment component concurrently calls the task interfaces `perfmon_window_start()` and `perfmon_window_stop()` of each performance monitoring application interface instance. This controls the performance monitoring application interface instance to send window start control signals and window stop control signals to the input ports of each corresponding atom probe instance through the corresponding verification interface and forced connection, in order to mark the measurement start time and measurement end time, thereby determining the performance monitoring interval.
[0063] Optionally, after establishing a forced connection between the verification interface and the control port and result port of each of the atomic probe instances through forced binding macros, the following may also be included: Through the verification interface associated with each of the performance monitoring application interface instances, a monitoring end control signal and a result reading control signal are generated, and the monitoring end control signal and the result reading control signal are sent to each of the atomic probe instances through the forced connection corresponding to each of the verification interfaces; Using each of the aforementioned atomic probe instances, performance monitoring is terminated according to the monitoring end control signal, and the control signal is read according to the result. The statistical result data is then written to the interface log via the result port and the verification interface.
[0064] In this embodiment, during the final stage of the simulation, the performance monitoring environment component automatically performs the termination and result acquisition operations in parallel for all registered performance monitoring application interface instances. Specifically, for each performance monitoring application interface instance, `perfmon_end()` and `perfmon_result()` are called concurrently using a parallel process startup statement (fork-join_none). This controls the performance monitoring application interface instance to send monitoring termination control signals and result reading control signals to the input ports of its corresponding atom probe instances through the corresponding verification interface and forced connection, thereby achieving efficient batch acquisition. Then, the atom probe instances terminate performance monitoring according to the monitoring termination control signal, latch the statistical result data, and then automatically write the statistical result data to the interface log via the result port and verification interface according to the result reading control signal. The performance monitoring environment component uses UVM phase mechanism objection control to ensure that simulation control is released only after all acquisition tasks are completed, thus guaranteeing the integrity of the performance data.
[0065] In this embodiment, a force binding mechanism enables non-intrusive control of the internal probes by the verification environment. The verification environment does not need to understand the internal implementation details of the probes; it can control atomic probe instances at any location simply through a unified application programming interface. Simultaneously, the automated post-main-phase acquisition mechanism integrates performance data acquisition entirely into the UVM verification process, eliminating the need for test engineers to write additional acquisition code and reducing the complexity of the verification work.
[0066] The technical solution of this invention involves obtaining each logical path corresponding to the target protocol and generating an interface performance monitor instance corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance. When a performance monitoring trigger for the target protocol is detected based on either the design-embedded mode or the verification platform control mode, a performance monitoring interval is obtained. Based on the performance monitoring interval, basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization are sequentially executed through each interface performance monitor instance and the corresponding atomic probe instance to obtain performance monitoring data at each level. In the design-embedded mode, a simulation report is generated based on the performance monitoring data at each level, and in the verification platform control mode, an interface log is generated based on the performance monitoring data at each level. By adopting a dual-mode observation architecture that simultaneously supports design-embedded and verification platform control, and a dual-output mechanism for results, the fusion observation of performance data between the design and verification environments can be achieved, which can improve the efficiency of locating performance problems and reduce the difficulty of cross-team collaboration.
[0067] The technical effects of the present invention are as follows: (1) High reusability: Through the parameterized design of atomic performance probes and different event binding strategies, the same probe module can act as an event counter, delay measurer, incomplete transaction tracker, bandwidth meter and backpressure analyzer in different instantiation contexts, reducing the development cost of developing independent monitoring logic for each performance indicator. (2) Hierarchical composability: The three-layer architecture of L0, L1 and L2 enables the system to be gradually combined from the simplest event statistics to complex bus protocol analysis. Each layer only needs to focus on the semantic mapping logic of its own layer, reducing the complexity of development and maintenance. (3) Zero synthesis area impact: All monitoring logic is within the conditional compilation guard and is automatically eliminated during synthesis, without affecting the area, power consumption and timing of the final chip. (4) Design-verification fusion observation: The same probe infrastructure serves both design debugging (waveform + log) and verification automation (application programming interface + result file), eliminating the redundancy of the design team and verification team maintaining independent performance observation systems in the traditional solution. (5) Flexible measurement control: Four time window modes and dual-mode enable control enable the probe to automatically collect full-process performance data without external intervention, and to perform targeted performance analysis for specific test stages under the precise control of the verification environment.
[0068] Example 3 Figure 3 This is a schematic diagram of a performance monitoring device for integrated circuit simulation provided in Embodiment 3 of the present invention. Figure 3 As shown, the device includes: an instance generation module 310, an interval acquisition module 320, and a data acquisition module 330; wherein, The instance generation module 310 is used to obtain each logical path corresponding to the target protocol, and generate an interface performance monitor instance corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance. The interval acquisition module 320 is used to acquire the performance monitoring interval when a performance monitoring trigger for the target protocol is detected. The data acquisition module 330 is used to perform basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization sequentially through each of the interface performance monitor instances and the corresponding atomic probe instances according to the performance monitoring interval, so as to obtain performance monitoring data at all levels.
[0069] The technical solution of this invention obtains each logical path corresponding to the target protocol and generates interface performance monitor instances corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance. When a performance monitoring trigger for the target protocol is detected, a performance monitoring interval is obtained. Based on the performance monitoring interval, basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization are sequentially executed through each interface performance monitor instance and its corresponding atomic probe instance to obtain performance monitoring data at all levels. By adopting logical path decomposition, multi-level instance generation, and a three-level statistical process, performance monitoring code reuse can be achieved, the scalability of performance monitoring can be improved, and efficient monitoring of multi-level performance data can be achieved.
[0070] Optionally, the interval acquisition module 320 is specifically used to acquire the current time window mode and acquire the performance monitoring interval based on the current time window mode; The time window modes include manual control mode, first input event to last output event mode, first input event to last input event mode, and first output event to last output event mode.
[0071] Optionally, the data acquisition module 330 includes: The data statistics unit is used to, within the performance monitoring interval, statistically analyze, based on pre-configured input events, output events, and weight data, the incomplete transaction data, response latency data, request activity cycle data, backpressure cycle data, bus bandwidth data, and effective byte bandwidth data, respectively, through each of the atomic probe instances. The result acquisition unit is used to perform arithmetic combination of the statistical result data of the corresponding atomic probe instances through each of the interface performance monitor instances to obtain interface-level performance indicators, and to summarize the interface-level performance indicators to obtain protocol-level performance statistics.
[0072] Optionally, the data statistics unit is specifically used to update the values of the event count register, weighted accumulation register, and maximum value register according to the pre-configured input events, output events, and weight data through each of the atomic probe instances, so as to obtain the total number of transactions, the number of incomplete transactions, and the historical maximum number of incomplete transactions, and to obtain the incomplete transaction data based on the total number of transactions, the number of incomplete transactions, and the historical maximum number of incomplete transactions.
[0073] Optionally, the data statistics unit is also used to record the occurrence timestamps of input events and output events through the input event queue and the output event queue respectively, and when the current output event is triggered, to obtain the first occurrence timestamp of the current output event from the output event queue and the second occurrence timestamp of the current input event corresponding to the current output event from the input event queue; The single event delay is calculated based on the first occurrence timestamp and the second occurrence timestamp, and the response delay data is obtained based on the single event delay.
[0074] Optionally, the performance monitoring device for integrated circuit simulation also includes: The counter establishment module is used to establish a counter corresponding to each transaction identifier through each of the interface performance monitor instances; the counter is incremented by one when the request handshake of the corresponding transaction identifier is successful, and decremented by one when the response handshake of the corresponding transaction identifier is successful. The quantity acquisition module is used to acquire the number of identifiers of simultaneously incomplete transactions based on the count values of each of the counters; Each of the interface performance monitor instances contains embedded protocol compliance assertions, including at least one of the following: no backpressure on the response channel, integrity of request and response pairing by transaction identifier, and read / write return response specifications.
[0075] Optionally, the interval acquisition module 320 is also used to detect a performance monitoring trigger for the target protocol based on the design embedded mode or the verification platform control mode. The data acquisition module 330 is also used to generate simulation reports based on performance monitoring data at all levels in the design embedded mode, and to generate interface logs based on performance monitoring data at all levels in the verification platform control mode.
[0076] Optionally, the performance monitoring device for integrated circuit simulation also includes: The interface acquisition module is used to establish multiple performance monitoring application interface instances in the verification platform control mode, and to obtain the verification interface associated with each performance monitoring application interface instance based on the configuration database. The connection establishment module is used to establish a forced connection between the verification interface and the control port and result port of each of the atomic probe instances through a forced binding macro, and to send an enable signal to the control port of each of the atomic probe instances through the forced connection to trigger performance monitoring.
[0077] Optionally, the connection establishment module is also used to generate a time window mode selection signal through the verification interface associated with each of the performance monitoring application interface instances, and send the time window mode selection signal to each of the atomic probe instances through the forced connection corresponding to each of the verification interfaces. When the time window mode selection signal corresponds to the manual control mode, a window start control signal and a window end control signal are generated through the verification interface associated with each of the performance monitoring application interface instances, and the window start control signal and the window end control signal are sent to each of the atomic probe instances through the forced connection corresponding to each of the verification interfaces. The performance monitoring interval is obtained using each of the aforementioned atomic probe instances, based on the window start control signal and the window end control signal.
[0078] Optionally, the connection establishment module is also used to generate a monitoring end control signal and a result reading control signal through the verification interface associated with each of the performance monitoring application interface instances, and send the monitoring end control signal and the result reading control signal to each of the atomic probe instances through the forced connection corresponding to each of the verification interfaces; Using each of the aforementioned atomic probe instances, performance monitoring is terminated according to the monitoring end control signal, and the control signal is read according to the result. The statistical result data is then written to the interface log via the result port and the verification interface.
[0079] The performance monitoring device for integrated circuit simulation provided in this embodiment of the invention can execute the performance monitoring method for integrated circuit simulation provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.
[0080] Example 4 Figure 4 A schematic diagram of an electronic device 40 that can be used to implement embodiments of the present invention is shown. The electronic device 40 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.
[0081] like Figure 4 As shown, the electronic device 40 includes at least one processor 41 and a memory, such as a read-only memory (ROM) 42 or a random access memory (RAM) 43, communicatively connected to the at least one processor 41. The memory stores computer programs executable by the at least one processor. The processor 41 can perform various appropriate actions and processes based on the computer program stored in the read-only memory 42 or loaded from the storage unit 48 into the random access memory 43. The RAM 43 can also store various programs and data required for the operation of the electronic device 40. The processor 41, ROM 42, and RAM 43 are interconnected via a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.
[0082] Multiple components in electronic device 40 are connected to I / O interface 45, including: input unit 46, such as keyboard, mouse, etc.; output unit 47, such as various types of monitors, speakers, etc.; storage unit 48, such as disk, optical disk, etc.; and communication unit 49, such as network card, modem, wireless transceiver, etc. Communication unit 49 allows electronic device 40 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.
[0083] Processor 41 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 41 include, but are not limited to, central processing units, graphics processing units, various special-purpose artificial intelligence computing chips, various processors running machine learning model algorithms, digital signal processors, and any suitable processor, controller, microcontroller, etc. Processor 41 performs the various methods and processes described above, such as performance monitoring methods for integrated circuit simulation.
[0084] In some embodiments, the performance monitoring method for integrated circuit emulation can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 48. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 40 via ROM 42 and / or communication unit 49. When the computer program is loaded into RAM 43 and executed by processor 41, one or more steps of the performance monitoring method for integrated circuit emulation described above can be performed. Alternatively, in other embodiments, processor 41 can be configured to perform the performance monitoring method for integrated circuit emulation by any other suitable means (e.g., by means of firmware).
[0085] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays, application-specific integrated circuits (ASICs), application-specific standard products (ASICs), system-on-a-chip (SoCs), complex programmable logic devices, computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.
[0086] Computer programs used to implement the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.
[0087] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination thereof.
[0088] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device 40, which includes: a display device (e.g., a cathode ray tube or liquid crystal display) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device 40. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).
[0089] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.
[0090] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact via a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server.
[0091] This embodiment may also include a computer program product, which includes a computer program that, when executed by a processor, implements the performance monitoring method for integrated circuit simulation provided in any embodiment of the present invention.
[0092] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.
[0093] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.
Claims
1. A performance monitoring method for integrated circuit simulation, characterized by, include: Obtain each logical path corresponding to the target protocol, and generate an interface performance monitor instance corresponding to each logical path, as well as multiple atomic probe instances corresponding to each interface performance monitor instance; When a performance monitoring trigger for the target protocol is detected, the performance monitoring interval is obtained; Based on the performance monitoring interval, basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization are sequentially performed through each interface performance monitor instance and the corresponding atomic probe instance to obtain performance monitoring data at each level.
2. The method of claim 1, wherein, Obtain the performance monitoring interval, including: Obtain the current time window mode, and based on the current time window mode, obtain the performance monitoring interval; The time window modes include manual control mode, first input event to last output event mode, first input event to last input event mode, and first output event to last output event mode.
3. The method of claim 1, wherein, Based on the performance monitoring interval, basic performance statistics, interface-level indicator aggregation, and protocol-level data summarization are sequentially performed through each interface performance monitor instance and its corresponding atomic probe instance to obtain performance monitoring data at each level, including: Within the performance monitoring interval, each atomic probe instance statistically obtains incomplete transaction data, response latency data, request activity cycle data, backpressure cycle data, bus bandwidth data, and effective byte bandwidth data based on pre-configured input events, output events, and weight data. By using each of the interface performance monitor instances, arithmetic combinations are performed based on the statistical results of the corresponding atomic probe instances to obtain interface-level performance indicators, and the interface-level performance indicators are summarized to obtain protocol-level performance statistics.
4. The method according to claim 3, characterized in that, Based on pre-configured input events, output events, and weight data, each of the aforementioned atomic probe instances statistically obtains incomplete transaction data, including: Through each of the aforementioned atomic probe instances, the values of the event count register, weighted accumulation register, and maximum value register are updated according to the pre-configured input events, output events, and weight data to obtain the total number of transactions, the number of incomplete transactions, and the historical maximum number of incomplete transactions. Based on the total number of transactions, the number of incomplete transactions, and the historical maximum number of incomplete transactions, the incomplete transaction data is obtained.
5. The method according to claim 3, characterized in that, The response latency data is statistically obtained by analyzing the pre-configured input events, output events, and weight data of each atomic probe instance, including: The timestamps of the input and output events are recorded in the input event queue and the output event queue, respectively. When the current output event is triggered, the first timestamp of the current output event is obtained from the output event queue, and the second timestamp of the current input event corresponding to the current output event is obtained from the input event queue. The single event delay is calculated based on the first occurrence timestamp and the second occurrence timestamp, and the response delay data is obtained based on the single event delay.
6. The method according to claim 3, characterized in that, Also includes: Each interface performance monitor instance establishes a counter corresponding to each transaction identifier; the counter is incremented by one when the request handshake of the corresponding transaction identifier is successful, and decremented by one when the response handshake of the corresponding transaction identifier is successful. Based on the count values of each counter, obtain the number of identifiers of concurrently existing incomplete transactions; Each of the interface performance monitor instances contains embedded protocol compliance assertions, including at least one of the following: no backpressure on the response channel, integrity of request and response pairing by transaction identifier, and read / write return response specifications.
7. The method according to claim 1, characterized in that, Performance monitoring for the target protocol was detected, including: Based on the design embedded mode or the verification platform control mode, a performance monitoring trigger for the target protocol is detected. After acquiring performance monitoring data at all levels, the following is also included: In the embedded design mode, simulation reports are generated based on performance monitoring data at all levels. In the verification platform control mode, interface logs are generated based on performance monitoring data at all levels.
8. The method according to claim 7, characterized in that, Also includes: In the verification platform control mode, multiple performance monitoring application interface instances are established, and the verification interface associated with each performance monitoring application interface instance is obtained based on the configuration database. By forcibly binding macros, a forced connection is established between the verification interface and the control port and result port of each of the atomic probe instances. Through the forced connection, an enable signal is sent to the control port of each of the atomic probe instances to trigger performance monitoring.
9. The method according to claim 8, characterized in that, After establishing a forced connection between the verification interface and the control port and result port of each atomic probe instance through forced binding macros, the method further includes: A time window mode selection signal is generated through the verification interface associated with each of the performance monitoring application interface instances, and the time window mode selection signal is sent to each of the atom probe instances through the forced connection corresponding to each of the verification interfaces. When the time window mode selection signal corresponds to the manual control mode, a window start control signal and a window end control signal are generated through the verification interface associated with each of the performance monitoring application interface instances, and the window start control signal and the window end control signal are sent to each of the atomic probe instances through the forced connection corresponding to each of the verification interfaces. The performance monitoring interval is obtained using each of the aforementioned atomic probe instances, based on the window start control signal and the window end control signal.
10. The method according to claim 8, characterized in that, After establishing a forced connection between the verification interface and the control port and result port of each atomic probe instance through forced binding macros, the method further includes: Through the verification interface associated with each of the performance monitoring application interface instances, a monitoring end control signal and a result reading control signal are generated, and the monitoring end control signal and the result reading control signal are sent to each of the atomic probe instances through the forced connection corresponding to each of the verification interfaces; Using each of the aforementioned atomic probe instances, performance monitoring is terminated according to the monitoring end control signal, and the control signal is read according to the result. The statistical result data is then written to the interface log via the result port and the verification interface.
