Data tracking method, system and device, computer equipment and storage medium
By splitting data processing tasks in high-performance storage control chips and setting tracking configurations, the data output timing is optimized, and the problem of packet loss and overlap under limited bandwidth is solved, efficient and reliable real-time tracking information output is achieved, reducing the Operation and maintenance costs.
Patent Information
- Application Number
- CN202411178350.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-11-29
AI Technical Summary
In high-performance storage control chips, the limited tracking data bandwidth leads to packet loss and overlap, making it difficult to achieve efficient capture and output of real-time tracking information, affecting system performance.
By obtaining the data processing tasks in the target chip, splitting them into subprocessing tasks based on the preset template, and setting up tracking configurations for each subtask, including export parameters, security parameters and level parameters, and determining the output timing based on the remaining bandwidth and data volume. Optimize data bit width and output timing to maximize bandwidth utilization and reduce packet loss and overlap.
It improves the reliability and output efficiency of data tracking information, reduces the workload of post-unpacking analysis, saves operation and maintenance costs, and realizes the capture and output of real-time tracking information without affecting system performance.
Smart Images

Figure CN119046097A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer technology, and in particular to a data tracking method, system, device, computer equipment and storage medium. Background Art
[0002] The chip trace function refers to tracking and recording the internal state and data information of the chip during operation, which is used for debugging, performance analysis, and problem diagnosis. With the continuous improvement of the performance indicators of high-performance storage control chips, especially the emergence of RAID (Redundant Array of Independent Disks) control chips for NVMe SSD (NVMe Solid State Drive), how to improve the efficiency and quality of trace data output under limited trace data bandwidth conditions is a challenging issue.
[0003] Specifically, high-performance storage control chips generate TB of internal data information per second during operation for tracking and output, but the on-chip tracking system can generally only provide a data throughput of less than 10GB / s. Figure 1 FIG. 1 is a schematic diagram of data tracking by a tracking hardware unit, wherein, due to the contradiction between the high data output demand and the low output bandwidth, when data from multiple data sources (i.e. Figure 1 When multiple tracking data packets of the tracking sources 1-3) in the system are output to external devices, data packet loss (partial or complete) and overlap between multiple data packets will inevitably occur. Then, it will be very difficult and inefficient to analyze and locate the problem through these massive (hundreds of GB) and fragmented data. In addition, for a storage control system that requires real-time response, how to achieve real-time tracking information capture and output without affecting system performance is also a technical difficulty. Summary of the invention
[0004] In view of this, the present invention provides a data tracking method, system, apparatus, computer device and storage medium to solve the problem of being unable to capture and output real-time tracking information without affecting chip performance.
[0005] In a first aspect, the present invention provides a data tracking method, the method comprising:
[0006] Obtaining a data processing task in a target chip, and splitting the data processing task based on a preset template corresponding to the data processing request to obtain sub-processing tasks;
[0007] Setting a tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task;
[0008] Execute the sub-processing task and output the tracing information according to the tracing configuration of the sub-processing task.
[0009] In an optional implementation, setting a tracking configuration for a sub-processing task includes:
[0010] When the tracking configuration is enabled for the sub-processing task, export parameters and security parameters are set for the sub-processing task, wherein the export parameters are used to indicate whether to export the tracking information corresponding to the sub-processing task when the sub-processing task fails to execute, and the security parameters are used to indicate whether to allow the tracking information to be lost;
[0011] Set the level parameters of the sub-processing tasks based on the data volume of the sub-processing tasks;
[0012] The data width of the trace configuration is set according to the export parameters, security parameters, and level parameters.
[0013] In the embodiment of the present invention, the timing of outputting the tracking information can be determined according to the remaining bandwidth in the target chip, thereby ensuring the reliability of the output tracking information and maximizing the use of the limited tracking bandwidth without increasing the operating frequency, increasing the data bit width, or increasing the internal data cache. By improving the utilization rate of the limited tracking bandwidth, packet loss and disordered overlap between multiple packets are reduced, the amount of tracking output data and the workload of subsequent unpacking analysis are greatly reduced, saving the labor cost of subsequent operation and maintenance.
[0014] In an optional implementation, the level parameter of the sub-processing task is set based on the data volume of the sub-processing task, including:
[0015] Obtaining a preset processing level, wherein different preset processing levels correspond to different amounts of tracking information data;
[0016] The data volume of the tracking information of the sub-processing task is estimated, and a target processing level matching the estimated result is determined in the preset processing levels, and a level parameter is generated according to the target processing level.
[0017] In the embodiment of the present invention, the output level can be set for the sub-processing task according to the estimated data volume of the tracking information of the sub-processing task, thereby providing a technical basis for determining the timing of outputting the tracking information of the sub-processing task according to the remaining bandwidth in the target chip.
[0018] In an optional implementation, outputting the tracking information according to the tracking configuration includes:
[0019] Monitor the remaining bandwidth in the target chip;
[0020] When the remaining bandwidth matches the trace configuration of the sub-processing task, the trace information is output.
[0021] In the embodiment of the present invention, the timing of outputting the tracking information can be determined according to the remaining bandwidth in the target chip, thereby ensuring the reliability of the output tracking information and maximizing the use of the limited tracking bandwidth without increasing the operating frequency, increasing the data bit width, or increasing the internal data cache. By improving the utilization rate of the limited tracking bandwidth, packet loss and disordered overlap between multiple packets are reduced, the amount of tracking output data and the workload of subsequent unpacking analysis are greatly reduced, saving the labor cost of subsequent operation and maintenance.
[0022] In an optional implementation, executing the sub-processing task includes:
[0023] Generate a task chain for the sub-processing task based on the execution steps corresponding to the data processing task;
[0024] Write the first address of the first sub-processing task in the task chain into the task queue;
[0025] Based on the first address, calculating the second addresses of the remaining sub-processing tasks in the task chain, and writing the second addresses into the task queue;
[0026] Based on the task queue, the sub-processing tasks in the task chain are executed sequentially.
[0027] In an embodiment of the present invention, the second address of each remaining sub-processing task in the control block chain can be calculated based on the data length and the first address of each sub-processing task in the control block chain, so as to write the control block chain into the task queue, thereby providing an implementation basis for organizing hardware units based on the work queue to schedule and execute sub-processing tasks in the control block chain.
[0028] In an optional implementation, calculating the second addresses of the remaining sub-processing tasks in the task chain based on the first address includes:
[0029] Get the data length of each sub-processing task in the first sub-processing task;
[0030] Based on the sum of the first address and the data length, the second address of each remaining sub-processing task is calculated respectively.
[0031] In an embodiment of the present invention, the second address of each remaining sub-processing task in the control block chain can be calculated based on the data length and the first address of each sub-processing task in the control block chain, so as to write the control block chain into the task queue, thereby providing an implementation basis for organizing hardware units based on the work queue to schedule and execute sub-processing tasks in the control block chain.
[0032] In an optional embodiment, the method further includes:
[0033] Execute the sub-processing tasks in the task queue in parallel and count the number of sub-processing tasks executed;
[0034] If the number of executions exceeds the number of sub-processing tasks in the task chain, an error message is generated.
[0035] In an embodiment of the present invention, in the work queue organization hardware unit, the sub-processing tasks in the control block chain can be assigned to different computing nodes for parallel execution. At the same time, when the number of control block chains is multiple, each control block chain can be assigned to different functional modules for parallel execution, thereby improving the execution efficiency of data processing tasks.
[0036] In an optional implementation, splitting the data processing task based on a preset template corresponding to the data processing request to obtain sub-processing tasks includes:
[0037] Determine a preset template that matches a task type of the data processing task;
[0038] Based on the preset template, the data processing request is split according to the execution steps of the data processing task to obtain sub-processing tasks.
[0039] In an embodiment of the present invention, the data processing task can be split according to a preset template to obtain sub-processing tasks corresponding to each execution step, so as to set tracking configuration for each sub-processing task respectively to refine the data tracking granularity and reduce the bandwidth pressure of the data transmission interface.
[0040] In an optional embodiment, the method further includes:
[0041] Based on the task type, a target control block is allocated for the task type in the control block of the target chip, and a preset template is determined according to the target control block;
[0042] Map and bind preset templates to task types.
[0043] In an embodiment of the present invention, the target control blocks corresponding to IO requests of different task types can be managed by the IO hardware management module to obtain a control block template matrix mapping, and a mapping relationship between IO requests of different task types and preset templates can be established based on the control block template matrix mapping, thereby providing a technical basis for splitting data processing tasks by selecting a corresponding mapping template based on the task type.
[0044] In an optional embodiment, the method further includes:
[0045] In response to the template adjustment instruction, the target control block matching the task type is recalculated, and the preset template is updated according to the calculation result.
[0046] In an embodiment of the present invention, the target control blocks corresponding to IO requests of different task types can be managed by the IO hardware management module to obtain a control block template matrix mapping, and a mapping relationship between IO requests of different task types and preset templates can be established based on the control block template matrix mapping, thereby providing a technical basis for splitting data processing tasks by selecting a corresponding mapping template based on the task type.
[0047] In an optional embodiment, the method further includes:
[0048] Generate a test chain based on the functional blocks corresponding to each functional module in the target chip;
[0049] Generate a test task through the test software in the target chip, wherein the test task is used to test the performance of the functional module in the target chip;
[0050] Execute the test task through the test chain, track the execution process of the test task, and obtain the test results.
[0051] After generating the test task according to the test chain, the tracking configuration corresponding to each control block can be set, and the configured test task can be scheduled to the work queue to organize the hardware unit for execution. It should be understood that when the target chip executes the test task, the test chain competes with the other IO request functions being executed for system resources. Therefore, the obtained test task can truthfully reflect the delay of the system at that time. At the same time, only the test chain in the entire system will generate tracking information at the level of tens of KB, which greatly reduces the size of the tracking cache and the occupation of the C / D port bandwidth, and the integrity of the tracking information is guaranteed. The tracking information will not overlap with different IO tracking data packets, and its decoding becomes very direct and efficient. Therefore, tracking data can be captured through a single test chain, and the corresponding impact on the host business can be minimized.
[0052] In an optional implementation, tracking the execution process of the test task to obtain the test result includes:
[0053] During the operation of the target chip, the execution process of each functional block in the test chain is tracked to obtain a tracking log;
[0054] The trace log is output through the data transmission interface in the target chip, and the test result is determined according to the trace log.
[0055] In the embodiment of the present invention, the system performance can be tested through the test task. Under the premise of not affecting the host IO business, the real-time status of the system operation can be captured in a targeted manner, which effectively saves the limited tracking bandwidth and storage space and greatly reduces the difficulty of decoding the tracking log, thereby significantly improving the iteration speed of problem solving and system performance tuning.
[0056] In an optional implementation, determining the test result according to the tracking log includes:
[0057] Based on the delay information in the tracking log, analyze the performance of the functional module corresponding to the tracking log to obtain the sub-test results;
[0058] The test results are determined based on the sub-test results corresponding to each functional module.
[0059] In the embodiment of the present invention, the system performance can be tested through the test task. Under the premise of not affecting the host IO business, the real-time status of the system operation can be captured in a targeted manner, which effectively saves the limited tracking bandwidth and storage space and greatly reduces the difficulty of decoding the tracking log, thereby significantly improving the iteration speed of problem solving and system performance tuning.
[0060] In a second aspect, the present invention provides a data tracking system, the system comprising: an IO management hardware module IOM, a work queue organization hardware unit WQS, a tracking hardware unit and a tracking cache;
[0061] The IOM is used to obtain the data processing task in the target chip, split the data processing task based on the preset template corresponding to the data processing request, obtain sub-processing tasks, and set the tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate the tracking mode corresponding to the sub-processing task;
[0062] WQS, used to allocate sub-processing tasks to corresponding functional modules in the target chip so that the functional modules perform the sub-processing tasks;
[0063] The tracing hardware unit is used to output tracing information to the tracing cache according to the tracing configuration of the sub-processing task during the execution of the sub-processing task by the functional module;
[0064] The trace cache is used to output the trace information to an external device through the data transmission interface of the target chip.
[0065] In an optional embodiment, the system further includes: a central processing unit CPU;
[0066] The CPU is used to generate a test chain based on the functional blocks corresponding to each functional module in the target chip, generate test tasks through the test software in the target chip, execute the test tasks through the test chain, and track the execution process of the test tasks to obtain test results, wherein the test tasks are used to test the performance of the functional modules in the target chip.
[0067] In an optional embodiment, the tracking cache is also used to: track the execution process of each functional block in the test chain during the operation of the target chip, obtain a tracking log, output the tracking log through the data transmission interface in the target chip, and determine the test result based on the tracking log.
[0068] In a third aspect, the present invention provides a data tracking device, the device comprising:
[0069] A splitting module is used to obtain the data processing task in the target chip and split the data processing task based on a preset template corresponding to the data processing request to obtain sub-processing tasks;
[0070] A setting module, used to set a tracking configuration for a sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task;
[0071] The execution module is used to execute the sub-processing task and output the tracking information according to the tracking configuration of the sub-processing task.
[0072] In a fourth aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the data tracking method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.
[0073] In a fifth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the data tracking method of the first aspect or any corresponding embodiment thereof.
[0074] In a sixth aspect, the present invention provides a computer program product, comprising computer instructions for causing a computer to execute the data tracking method of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS
[0075] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0076] Figure 1 is a schematic diagram of data tracking through a tracking hardware unit;
[0077] Figure 2 This is a block diagram of a typical system-on-chip data tracking implementation;
[0078] Figure 3 is a flow chart of a data tracking method according to an embodiment of the present invention;
[0079] Figure 4 It is a schematic diagram of the tracking configuration of the configuration CB;
[0080] Figure 5 is a flow chart of another data tracking method according to an embodiment of the present invention;
[0081] Figure 6 It is a schematic diagram of the control block template matrix mapping;
[0082] Figure 7 is a flow chart of another data tracking method according to an embodiment of the present invention;
[0083] Figure 8 is a flow chart of another data tracking method according to an embodiment of the present invention;
[0084] Fig. 9 It is a schematic diagram of allocating CB chains to WQS;
[0085] Fig.10 It is a schematic diagram of a chain of control blocks corresponding to executing multiple data processing tasks in parallel;
[0086] Fig.11 It is a timing diagram when the system performs the test task;
[0087] Fig.12 It is the architecture diagram of the data tracking system;
[0088] Fig.13 It is a flowchart of data tracking for a target chip;
[0089] Fig.14 is a structural block diagram of a data tracking device according to an embodiment of the present invention;
[0090] Fig.15 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION
[0091] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 those skilled in the art without creative work are within the scope of protection of the present invention.
[0092] In conjunction with the application scenarios on which the execution of the data tracking method depends, the application scenarios are described here.
[0093] The chip trace function refers to tracking and recording the internal state and data information of the chip during operation, which is used for debugging, performance analysis, and problem diagnosis. With the continuous improvement of the performance indicators of high-performance storage control chips, especially the emergence of RAID (Redundant Array of Independent Disks) control chips for NVMe SSD (NVMe Solid State Drive), how to improve the efficiency and quality of trace data output under limited trace data bandwidth conditions is a challenging issue.
[0094] Specifically, high-performance storage control chips generate TB of internal data information per second during operation for tracking and output, but the on-chip tracking system can generally only provide a data throughput of less than 10GB / s. Figure 1 FIG. 1 is a schematic diagram of data tracking by a tracking hardware unit, wherein, due to the contradiction between the high data output demand and the low output bandwidth, when data from multiple data sources (i.e. Figure 1 When multiple tracking data packets of the tracking sources 1-3) in the system are output to external devices, data packet loss (partial or complete) and overlap between multiple data packets will inevitably occur. Then, it will be very difficult and inefficient to analyze and locate the problem through these massive (hundreds of GB) and fragmented data. In addition, for a storage control system that requires real-time response, how to achieve real-time tracking information capture and output without affecting system performance is also a technical difficulty.
[0095] Specifically, Figure 2 The figure shows a typical structure diagram of data tracing implementation of a system on chip (SoC), wherein, when performing data tracing, the test software deployed in the central processing unit can output the tracing log to the tracing hardware unit through the bus bridge (A to Bt), and similarly, the hardware function modules 1 to n can also output the tracing information to the tracing hardware unit through the bus bridge (A1 / … / An to Bt). The tracing hardware unit is responsible for encapsulating the original tracing information and outputting it to the tracing buffer through the bus bridge (At to B1). The information is cyclically covered within the designated space range, and optionally, the tracing information can be output to an external tracing debugging device through the D interface. At the same time, the external debugging device can access the trace buffer (Trace Buffer) through the JTAG interface connected to the C interface through the bus bridge (Aj to B1).
[0096] Here, the data bandwidth of interfaces such as Bt, At, C and D can only provide very limited data throughput due to cost and power consumption considerations, which poses a great challenge to the integrity of tracking data. Of course, the data throughput of the tracking function can be effectively improved by increasing the operating frequency, increasing the data bit width or increasing the internal data cache, but this will also increase the area cost of the non-functional part of the chip and the corresponding power consumption overhead.
[0097] Based on this, an embodiment of the present invention provides a data tracking method, which obtains a data processing task in a target chip, and splits the data processing task based on a preset template corresponding to the data processing request to obtain a sub-processing task. Then, a tracking configuration can be set for the sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task. Next, when executing the sub-processing task, tracking information can be output according to the tracking configuration of the sub-processing task, thereby avoiding the solution in the related solution in which the CPU controls the start and stop of the tracking function in the chip as a whole to track the overall operation data of the chip. Here, the present invention refines the intensity of data tracking, realizes a more accurate tracking configuration at the sub-processing task granularity level, and thereby improves the utilization rate of the data interface in the chip.
[0098] According to an embodiment of the present invention, a data tracking method embodiment is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0099] In this embodiment, a data tracking method is provided, which can be used for the above-mentioned high-performance storage control chip. Figure 3 is a flow chart of a data tracking method according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:
[0100] Step S301, obtaining a data processing task in a target chip, and splitting the data processing task based on a preset template corresponding to the data processing request to obtain sub-processing tasks.
[0101] In the embodiment of the present invention, the data processing task may be a data processing task determined according to the IO request issued by the host, and each data processing task may be allocated to Figure 1 The corresponding hardware function module 1 is executed.
[0102] After obtaining the data processing task, the data processing task can be split according to the preset template to obtain multiple subtasks. Specifically, the preset template can be a split template pre-set for the IO request type of the data processing task, which can be used to split the data processing task of the request type into multiple sub-processing tasks according to the execution steps. For example, in the above data processing task, it is used to indicate that the target data is written into the cache and the content respectively, then, according to the writing order, the data processing task can be split into a first sub-processing task that writes the target data into the cache and a second sub-processing task that writes the target data into the content.
[0103] Step S302: Setting a tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task.
[0104] In an embodiment of the present invention, the sub-processing task may be assigned to a control block (hereinafter referred to as CB) corresponding to the hardware function module for execution. Here, a tracking configuration may be set for the CB, and the CB module may determine whether to start data tracking based on the tracking configuration. If so, data tracking is performed through a corresponding tracking mode. Specifically, the tracking mode may be used to indicate a specific time node for exporting tracking information, etc.
[0105] For example, the user can select the CB for which data tracing is to be enabled through the tracing configuration, and can also configure the time node for tracing the data in the CB and outputting the tracing information. For example, the time node can be the moment when the remaining bandwidth in the chip is detected to be sufficient to meet the transmission requirements of the tracing information, thereby avoiding packet loss during the output of the tracing information due to insufficient remaining bandwidth.
[0106] Here, if Figure 4 The figure shows a schematic diagram of the trace configuration of the configuration CB, wherein, in addition to the above-mentioned trace configuration fields, the CB may also include fields for dynamically calculating parameters according to IO requests and static template parameter fields. Specifically, the dynamically calculating parameters according to IO requests are used to indicate the process of calculating and adjusting the operation parameters in real time according to the specific IO requests during the execution of the sub-processing task, and the static template parameters are fixed parameters specified in the template declaration, which are mainly used for specific types of operations and specializations in class or function templates.
[0107] Step S303: execute the sub-processing task and output the tracking information according to the tracking configuration of the sub-processing task.
[0108] In an embodiment of the present invention, when executing the above-mentioned sub-processing tasks, the CBs corresponding to each sub-task can be connected front to back in the execution order to obtain a task chain, and the addresses of each CB in the task chain can be stored in the physical storage space of the chip in the execution order, so that the hardware function module can schedule and execute the sub-processing tasks according to the task chain.
[0109] In addition, by Figure 2 It can be seen that when the hardware function module outputs the above-mentioned tracing information, it can output the tracing information to the tracing hardware unit through the bus bridge. The hardware tracing unit is responsible for encapsulating the original tracing information and outputting it to the tracing buffer through the bus bridge (At interface to B1 interface). At the same time, the tracing information can be output to the external tracing debugging device through the D interface. In addition, the external debugging device can access the tracing buffer (Trace Buffer) through the JTAG interface connected to the C interface through the bus bridge (Aj to B1).
[0110] It can be known from the above description that in an embodiment of the present invention, the data processing task in the target chip can be obtained, and the data processing task can be split based on the preset template corresponding to the data processing request to obtain a sub-processing task, and then a tracking configuration can be set for the sub-processing task, wherein the tracking configuration is used to indicate the tracking mode corresponding to the sub-processing task. Next, when executing the sub-processing task, the tracking information can be output according to the tracking configuration of the sub-processing task, thereby avoiding the scheme in the related scheme that the CPU controls the start and stop of the tracking function in the chip as a whole to track the overall operation data of the chip. Here, the present invention refines the intensity of data tracking, realizes a more accurate tracking configuration at the sub-processing task granularity level, and thereby improves the utilization rate of the data interface in the chip.
[0111] In this embodiment, another data tracking method is provided, which can be used for the above-mentioned high-performance storage control chip. Figure 5 is a flow chart of another data tracking method according to an embodiment of the present invention. Figure 5 As shown, the process includes the following steps:
[0112] Step S501, obtaining a data processing task in a target chip, and splitting the data processing task based on a preset template corresponding to the data processing request to obtain sub-processing tasks.
[0113] Specifically, the above step S501 includes:
[0114] Step S5011, determining a preset template that matches the task type of the data processing task.
[0115] Step S5012: Based on the preset template, the data processing request is split according to the execution steps of the data processing task to obtain sub-processing tasks.
[0116] In the embodiment of the present invention, the task type can be used to indicate the IO request type corresponding to the data processing task. It should be understood that different IO request types have different execution steps and corresponding CBs. In the target chip, the functional modules can be managed by the IO management hardware module (hereinafter referred to as IOM), for example, a corresponding CB is allocated to each functional module, and a preset template corresponding to each task type is managed, wherein the preset template includes multiple CBs.
[0117] For example, if the task type corresponding to the data processing task is IO1, the target control block corresponding to IO1 may include CB1.x corresponding to function module 1, CB2.m corresponding to function module 2, ..., and CBn.w corresponding to function module n.
[0118] Next, the data processing request may be split according to the execution steps of the data processing task to obtain sub-processing task 1-sub-processing task n, and the sub-processing tasks may be respectively allocated to corresponding control blocks CB1.x-CBn.w for execution.
[0119] Step S502: Set a tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate the tracking mode corresponding to the sub-processing task. Figure 3 Step S302 of the illustrated embodiment will not be described in detail here.
[0120] Step S503: execute the sub-processing task and output the tracking information according to the tracking configuration of the sub-processing task. Figure 3 Step S303 of the illustrated embodiment will not be described in detail here.
[0121] In an embodiment of the present invention, the data processing task can be split according to a preset template to obtain sub-processing tasks corresponding to each execution step, so as to set tracking configuration for each sub-processing task respectively to refine the data tracking granularity and reduce the bandwidth pressure of the data transmission interface.
[0122] In some optional implementations, the above step S501 further includes:
[0123] Step a1, based on the task type, assigning a target control block to the task type in the control block of the target chip, and determining a preset template according to the target control block.
[0124] Step a2: Map and bind the preset template to the task type.
[0125] In the embodiment of the present invention, the target control blocks corresponding to the IO requests of different task types can be managed by the above-mentioned IOM to obtain the CB template matrix mapping. Here, Figure 6 Shown is a schematic diagram of CB template matrix mapping.
[0126] Depend on Figure 6 It can be seen that the CB template matrix is a matrix composed of multiple control blocks (hereinafter referred to as CB), for example, CB1.x-CB1.z, CB2.l-CB1.n, CBn.u-CBn.w.
[0127] Here, the mapping of preset templates corresponding to IO requests of different task types can be established in the CB template matrix. For example, the preset template 1 corresponding to IO1 includes target function blocks CB1.x, CB2.m, ..., CBn.w. Therefore, the target function blocks in the preset template 1 can be mapped and bound to IO1 in the CB template matrix. Similarly, the preset templates corresponding to IO requests IO 2-IO X of other task types can be determined respectively, and the preset templates can be mapped and bound to the corresponding IO requests in the CB template matrix. The specific mapping and binding process is similar to that of IO 1, which will not be repeated here.
[0128] It should be understood that, in the present invention, in response to the template adjustment instruction, the target control block matching the task type is recalculated, and the preset template is updated according to the calculation result.
[0129] In specific implementation, the template adjustment instruction can be used to adjust the target function block in the preset template corresponding to the task type, for example, to delete, add, or replace the target function block in the preset template. Here, the template adjustment instruction can be actively generated when the task type is detected to update the preset template, thereby reducing the reliance on manual work during the data tracking process.
[0130] In an embodiment of the present invention, the target control blocks corresponding to IO requests of different task types can be managed by IOM to obtain a CB template matrix mapping, and a mapping relationship between IO requests of different task types and preset templates can be established based on the CB template matrix mapping, thereby providing a technical basis for splitting data processing tasks by selecting the corresponding mapping template based on the task type.
[0131] In this embodiment, another data tracking method is provided, which can be used for the above-mentioned high-performance storage control chip. Figure 7 is a flow chart of another data tracking method according to an embodiment of the present invention. Figure 7 As shown, the process includes the following steps:
[0132] Step S701, obtain the data processing task in the target chip, and split the data processing task based on the preset template corresponding to the data processing request to obtain sub-processing tasks. Figure 3 Step S301 of the illustrated embodiment will not be described in detail here.
[0133] Step S702: Setting a tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task.
[0134] Specifically, the above step S702 includes:
[0135] Step S7021, when the tracking configuration is turned on for the sub-processing task, set export parameters and security parameters for the sub-processing task, where the export parameters are used to indicate whether to export the tracking information corresponding to the sub-processing task when the sub-processing task fails to execute, and the security parameters are used to indicate whether tracking information packet loss is allowed.
[0136] Step S7022: Setting the level parameter of the sub-processing task based on the data volume of the sub-processing task.
[0137] Step S7023, setting the data width of the tracking configuration according to the export parameters, security parameters and level parameters.
[0138] In an embodiment of the present invention, a tracing configuration may be set for the CB corresponding to the sub-processing task, wherein the tracing configuration of the CB may be represented as a data bit width of a preset number of bits, for example, a tracing configuration of a 5-bit data bit width, wherein configurations corresponding to different numbers of bits in the data bit width.
[0139] Specifically, if the data in the above 5-bit data width is all 0, that is, 00000b, it means that the data tracking function is turned off when executing the sub-processing task in this CB.
[0140] In addition, when the lower bit of the 5-bit data width is 1 and the remaining bits are X, that is, XXXX1b, the tracking configuration can be to use the "error export" function, which is used to indicate that when the sub-processing task fails to execute, the tracking information corresponding to the sub-processing task is exported for subsequent problem analysis, wherein the tracking information may include useful error site information.
[0141] In addition, when the high bit of the 5-bit data width is 1 and the remaining bits are X, that is, 1XXXXb, the tracking configuration can be to use the "safe tracking" function, and the "error export" function is used to indicate that the tracking information is not allowed to be lost in the safe mode. In this safe mode, after the tracking information is subjected to back pressure, the normal execution of the functional logic in the chip will be delayed so that the remaining bandwidth can ensure the safe output of the tracking information.
[0142] In addition, the present invention can also set the output level of the sub-processing task according to the data volume of the sub-processing task, so as to determine the output timing of the tracking information according to the level parameter. Here, the high 2 bits to the high 4 bits of the 5-bit data width can be determined as the output level, for example, X001Xb-X111Xb is used to indicate level 1-level 7. Here, the higher the output level, the larger the data volume of the sub-processing task, the larger the data volume of the tracking information, and the larger the bandwidth required for export.
[0143] In specific implementation, the tracking information output according to the tracking configuration includes:
[0144] Step b1, monitoring the remaining bandwidth in the target chip.
[0145] Step b2: when the remaining bandwidth matches the tracking configuration of the sub-processing task, output the tracking information.
[0146] In an embodiment of the present invention, the output level configured in the tracking configuration can be tracked, and the data volume of the tracking information matching the output level can be obtained. Then, whether the remaining bandwidth in the target chip matches the data volume can be monitored.
[0147] Specifically, if the remaining bandwidth is greater than or equal to the data volume of the tracking information, it can be considered that the remaining bandwidth matches the data volume, and the tracking information can be output. Here, the remaining bandwidth can be calculated based on the occupancy status of the data transmission interface in the chip used to transmit the tracking information. For example, when the tracking information is transmitted to an external tracking and debugging device outside the chip, the remaining bandwidth can be calculated based on the bandwidth occupancy status of the C / D port of the above-mentioned data transmission interface. The specific method of calculating the remaining bandwidth is mainly based on what can be implemented, and will not be repeated in the present invention.
[0148] Step S703: execute the sub-processing task and output the tracking information according to the tracking configuration of the sub-processing task. Figure 3 Step S303 of the illustrated embodiment will not be described in detail here.
[0149] In the embodiment of the present invention, the timing of outputting the tracking information can be determined according to the remaining bandwidth in the target chip, thereby ensuring the reliability of the output tracking information and maximizing the use of the limited tracking bandwidth without increasing the operating frequency, increasing the data bit width, or increasing the internal data cache. By improving the utilization rate of the limited tracking bandwidth, packet loss and disordered overlap between multiple packets are reduced, the amount of tracking output data and the workload of subsequent unpacking analysis are greatly reduced, saving the labor cost of subsequent operation and maintenance.
[0150] In some optional implementations, the above step S7022 includes:
[0151] Step c1, obtaining a preset processing level, wherein different preset processing levels correspond to different amounts of tracking information data.
[0152] Step c2, estimating the data volume of the tracking information of the sub-processing task, and determining a target processing level that matches the estimated result in the preset processing levels, and generating level parameters according to the target processing level.
[0153] In the embodiment of the present invention, the data volume range of the corresponding tracking information can be pre-set for the above processing level, wherein the preset processing level is the above output level. Here, when the output level includes level 1 to level 7, the higher the output level, the larger the data volume range of the corresponding tracking information.
[0154] When estimating the data volume of the tracking information of a sub-processing task, the estimate can be made based on the data volume corresponding to the IO request of the sub-processing task. For example, when the IO request is an IO read request, the data volume of the tracking information of the sub-processing task can be estimated based on the data volume of the target data requested to be read by the IO request.
[0155] After estimating the data volume of the tracking information of the sub-processing task, the preset processing level corresponding to the data volume range to which the data volume belongs can be determined as the target processing level. For example, when the data volume range described in the estimated data volume of the tracking information is range 1, and the preset processing level corresponding to range 1 is level 1, then the target processing level of the sub-processing task can be determined as level 1.
[0156] In the embodiment of the present invention, the output level can be set for the sub-processing task according to the estimated data volume of the tracking information of the sub-processing task, thereby providing a technical basis for determining the timing of outputting the tracking information of the sub-processing task according to the remaining bandwidth in the target chip.
[0157] In this embodiment, another data tracking method is provided, which can be used for the above-mentioned high-performance storage control chip. Figure 8 is a flow chart of another data tracking method according to an embodiment of the present invention. Figure 8 As shown, the process includes the following steps:
[0158] Step S801, obtain the data processing task in the target chip, and split the data processing task based on the preset template corresponding to the data processing request to obtain sub-processing tasks. Figure 3 Step S301 of the illustrated embodiment will not be described in detail here.
[0159] Step S802: Set a tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate the tracking mode corresponding to the sub-processing task. Figure 3 Step S302 of the illustrated embodiment will not be described in detail here.
[0160] Step S803: execute the sub-processing task, and output the tracking information according to the tracking configuration of the sub-processing task.
[0161] Specifically, the above step S803 includes:
[0162] Step S8031, generating a task chain for the sub-processing task based on the execution steps corresponding to the data processing task.
[0163] Step S8032, writing the first address of the first sub-processing task in the task chain into the task queue.
[0164] Step S8033: Based on the first address, calculate the second addresses of the remaining sub-processing tasks in the task chain, and write the second addresses into the task queue.
[0165] Step S8034: based on the task queue, execute the sub-processing tasks in the task chain in sequence.
[0166] In an embodiment of the present invention, after generating a task chain (hereinafter referred to as CB chain) of IO requests corresponding to a data processing task, the task chain can be assigned to a WQS (Work Queue Scheduler, a work queue organization hardware module) so that the WQS executes the sub-processing task based on the task chain. Figure 6 It can be seen that the CB dynamic parameters can be calculated and updated to output the CB address corresponding to the first sub-processing task in the CB chain (i.e. the first address mentioned above), and the address is handed over to the WQS so that the WQS schedules the CB chain and executes the sub-processing tasks of each CB therein.
[0167] Specifically, the WQS may write the first address into the task queue, and sequentially calculate the CB addresses of the remaining sub-processing tasks in the CB chain according to the first address to obtain the second address, and sequentially write the second address into the task queue.
[0168] Specifically, step S8033 calculates the second addresses of the remaining sub-processing tasks in the task chain based on the first address, including:
[0169] Step d1, obtaining the data length of each sub-processing task in the first sub-processing task.
[0170] Step d2, calculating the second address of each remaining sub-processing task based on the sum of the first address and the data length.
[0171] In the embodiment of the present invention, the acquired data length may include the data length of each sub-processing task in the CB chain, and the second addresses of the remaining sub-processing tasks except the first sub-processing task in the CB chain are calculated in sequence.
[0172] Specifically, Fig. 9 The figure shows a schematic diagram of allocating CB chains to WQS, wherein the data length of each sub-processing task in the CB chain can be obtained to obtain a data length collection {L1, L2, ..., Ln}, wherein L1 is the data length of the first sub-processing task CB1 in the CB chain, L2 is the data length of the first sub-processing task CB2 in the CB chain, and so on.
[0173] When the CB chain is allocated to the WQS, the first address addr_cb1 of the first sub-processing task CB1 can be written into the WQS, and the sum of the first address and the data length L1 is calculated to obtain the address addr_cb2 of the second sub-processing task CB2 in the CB chain, where addr_cb2=addr_cb1+L1. Then, the address addr_cb3 of CB3 can be calculated based on the sum of the address of CB2 and the data length L2 of CB2, where addr_cb3=addr_cb2+L2, and so on, and the addresses of CB3, CB4, ... until CBn are calculated in sequence. The specific calculation method is as described in the calculation method of CB1 and CB2, and will not be repeated here.
[0174] In an embodiment of the present invention, the second address of each remaining sub-processing task in the CB chain can be calculated based on the data length and the first address of each sub-processing task in the CB chain, so as to write the CB chain into the task queue, thereby providing an implementation basis for scheduling and executing sub-processing tasks in the CB chain based on WQS.
[0175] In some optional implementations, the above step S803 further includes:
[0176] Step e1, executing the sub-processing tasks in the task queue in parallel, and counting the number of sub-processing tasks executed.
[0177] Step e2: when the number of executions exceeds the number of sub-processing tasks in the task chain, an error message is generated.
[0178] In the embodiment of the present invention, in order to improve the execution efficiency of sub-processing tasks, the sub-processing tasks in the CB chain can be allocated to different computing nodes for parallel execution in the WQS. It should be understood that after receiving CB chains corresponding to multiple data processing tasks in the WQS at the same time, these CB chains can also be executed in parallel at the same time. Here, each CB chain can be allocated to different hardware function modules for execution. The specific allocation method is not limited in this disclosure. For example, CB chain 1 is allocated to Figure 2 It is executed in the hardware function module 1 in.
[0179] When implementing it, Fig.10The figure shows a schematic diagram of CB chains corresponding to multiple data processing tasks executed in parallel, where task queue 1 is used to execute CB chain 1, and task queue n is used to execute CB chain n. Here, CB chain 1 is taken as an example, and CB chain 1 contains CB templates such as CB1.x, CB1.y, CB1.z, ... When performing data tracking, you can flexibly select the sub-processing tasks you want to track by setting the configuration parameters corresponding to the CB. For example, if you do not need to track the sub-processing tasks in CB1, you can set the configuration parameters of CB1 to turn off the tracking function.
[0180] In an embodiment of the present invention, in QS, the sub-processing tasks in the CB chain can be assigned to different computing nodes for parallel execution. At the same time, when the number of CB chains is multiple, each CB chain can be assigned to different functional modules for parallel execution, thereby improving the execution efficiency of data processing tasks.
[0181] In some optional embodiments, the above Figure 3 The corresponding embodiments also include:
[0182] Step S11, generating a test chain based on the functional blocks corresponding to each functional module in the target chip.
[0183] Step S12, generating a test task through the test software in the target chip, wherein the test task is used to test the performance of the functional module in the target chip.
[0184] Step S13, executing the test task through the test chain, and tracking the execution process of the test task to obtain the test result.
[0185] In an embodiment of the present invention, a "pseudo IO" processing CB chain (i.e., the above-mentioned test chain) can be created through software. The test chain can be used to test the delay of each module in the system. Therefore, when executing the test chain, it is necessary to cover each module to be tested.
[0186] On this basis, when generating the above test chain, the test chain may include at least one CB corresponding to each functional module. For example, if the CB corresponding to the above hardware functional module 1 is CB1.x-CB1.z in the CB template matrix, the CB corresponding to the hardware functional module 2 is CB2.1-CB2.n in the above CB template matrix, ..., the CB corresponding to the hardware functional module 2 is CBn.u-CBn.w in the CB template matrix. Therefore, when generating the test chain, the test chain may include CB1.x in the hardware functional module 1, CB2.n in the hardware functional module 2, ..., and CBn.v in the hardware functional module n.
[0187] After generating the test task according to the test chain, the trace configuration corresponding to each CB can be set, and the configured test task can be scheduled to WQS for execution. It should be understood that when the target chip executes the test task, the test chain competes with the other IO request functions being executed for system resources. Therefore, the obtained test task can truthfully reflect the delay of the system at that time. At the same time, only the test chain in the entire system will generate trace information at the level of tens of KB, which greatly reduces the size of the trace cache and the occupation of the C / D port bandwidth, and the integrity of the trace information is guaranteed. The trace information will not overlap with different IO trace data packets, and its decoding becomes very direct and efficient. Therefore, trace data can be captured through a single test chain, and the corresponding impact on the host business can be minimized.
[0188] In specific implementation, the above step S13 executes the test task through the test chain, and tracks the execution process of the test task to obtain the test result, including:
[0189] Step f1, during the operation of the target chip, the execution process of each functional block in the test chain is tracked to obtain a tracking log.
[0190] Step f2, outputting the tracking log through the data transmission interface in the target chip, and determining the test result according to the tracking log.
[0191] In the embodiment of the present invention, Fig.11 The figure shows a timing diagram when the system executes the above test task, wherein the host is the CPU in the above target chip, and the CPU can initiate the test through the host IO request, generate the test chain CB1.x / CB2.n / … / CBn.v, and distribute the test chain to the IOM.
[0192] Then, the IOM can submit the address CB1.x corresponding to the first sub-processing task CB1 in the above test chain to the WQS, so that the WQS can allocate the first sub-processing task to the corresponding hardware function module 1 according to CB1.x, and the hardware function module 1 can read and execute CB1.x to output the tracking information. It should be understood that the tracking information corresponding to each functional module can be stored in the above tracking cache to obtain a tracking log.
[0193] After CB1 is executed, CB2-CBn in the test chain can be executed in sequence according to the execution steps of CB1, and after CBn.x is executed, the IOM is responded to so that the IOM responds to the host, thereby completing the test task.
[0194] After receiving the notification from the IOM, the host can read the tracking log from the above tracking cache and return the tracking log to the host management tool for further processing. It should be understood that the host can spontaneously and periodically enable the test process and initiate the test task, thereby providing a technical basis for the adaptive adjustment of the dynamic performance of the system.
[0195] In addition, the above step S13, executing the test task through the test chain and tracking the execution process of the test task to obtain the test result, also includes:
[0196] Step g1, based on the delay information in the tracking log, analyze the performance of the functional module corresponding to the tracking log to obtain the sub-test result.
[0197] Step g2, determining the test result according to the sub-test results corresponding to each functional module.
[0198] In an embodiment of the present invention, the delay information may include the delay information of the functional modules corresponding to each CB in the test chain during the execution of the sub-processing task. Therefore, the performance of the functional module can be analyzed based on the delay information of the functional module to obtain the sub-test result. The sub-test result can be used to indicate whether the functional module has timed out when executing the sub-processing task. The specific timeout judgment standard depends on the usage situation and is not limited in this disclosure.
[0199] Next, a coarser-grained test result can be generated based on the sub-test results. The test result can be used to summarize and analyze the delay information of each step when executing the test task, and obtain a performance adjustment plan for the system to improve the overall performance of the system.
[0200] In the embodiment of the present invention, the system performance can be tested through the test task. Under the premise of not affecting the host IO business, the real-time status of the system operation can be captured in a targeted manner, which effectively saves the limited tracking bandwidth and storage space and greatly reduces the difficulty of decoding the tracking log, thereby significantly improving the iteration speed of problem solving and system performance tuning.
[0201] In summary, in an embodiment of the present invention, a data processing task in a target chip can be obtained, and the data processing task can be split based on a preset template corresponding to a data processing request to obtain a sub-processing task. Then, a tracking configuration can be set for the sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task. Next, when executing a sub-processing task, tracking information can be output according to the tracking configuration of the sub-processing task, thereby avoiding the solution in the related solution in which the CPU controls the start and stop of the tracking function in the chip as a whole to track the overall operation data of the chip. Here, the present invention refines the intensity of data tracking, realizes a more accurate tracking configuration at the sub-processing task granularity level, and thereby improves the utilization rate of the data interface in the chip.
[0202] In this embodiment, a data tracking system is also provided. The system is used to implement the above embodiments and preferred implementation modes. The descriptions that have been made will not be repeated. Fig.12 The diagram shows an architecture of a data tracking system, wherein the system comprises: an IO management hardware module IOM, a work queue organization hardware unit WQS, a tracking hardware unit and a tracking cache;
[0203] The IOM is used to obtain the data processing task in the target chip, split the data processing task based on the preset template corresponding to the data processing request, obtain sub-processing tasks, and set the tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate the tracking mode corresponding to the sub-processing task;
[0204] WQS, used to allocate sub-processing tasks to corresponding functional modules in the target chip so that the functional modules perform the sub-processing tasks;
[0205] The tracing hardware unit is used to output tracing information to the tracing cache according to the tracing configuration of the sub-processing task during the execution of the sub-processing task by the functional module;
[0206] The trace cache is used to output the trace information to an external device through the data transmission interface of the target chip.
[0207] In the embodiment of the present disclosure, the external device is Fig.12 The external tracing and debugging device in the system is used to perform fault analysis and other operations based on the tracing information, so as to adjust the performance of the system to improve the overall performance of the system. The specific method of determining the tracing information is as described above. Figure 3 The corresponding embodiments are described in detail and will not be repeated here.
[0208] In some optional implementations, the above system further includes: a central processing unit CPU.
[0209] The CPU is used to generate a test chain based on the functional blocks corresponding to each functional module in the target chip, generate test tasks through the test software in the target chip, execute the test tasks through the test chain, and track the execution process of the test tasks to obtain test results, wherein the test tasks are used to test the performance of the functional modules in the target chip.
[0210] In addition, the above-mentioned tracking cache is also used to: track the execution process of each functional block in the test chain during the operation of the target chip, obtain a tracking log, output the tracking log through the data transmission interface in the target chip, and determine the test result based on the tracking log.
[0211] In the embodiment of the present disclosure, a tracking log is generated by a test task, and a method of determining a test result according to the tracking log is as described in the above steps S11 to S13 and the corresponding embodiments, which will not be repeated here.
[0212] In addition, in an embodiment of the present invention, Fig.13 The figure shows a flow chart of data tracking for a target chip, wherein the host IO stream is used to indicate the IO data stream when the target chip performs the above-mentioned data processing task, the pseudo IO is used to indicate the IO data stream when the test task is performed, the tracking data is used to indicate the data stream for tracking the above-mentioned data processing task and the test task respectively, the hardware modules 1-n are the above-mentioned hardware functional modules 1-n, and the software is the above-mentioned CPU including the test software.
[0213] Specifically, when executing the above-mentioned data processing tasks, IOM can be used to manage hardware modules. At the same time, the sub-processing tasks in the data processing tasks can be stored in WQS through IOM, so that WQS can assign the sub-processing tasks to the hardware module for execution. After completing the current sub-processing task, the hardware module writes back the status information to the "completion" register of WQS.
[0214] When executing the above test tasks, the software can generate test tasks and store the test tasks in WQS, so that WQS can assign the test tasks to the hardware module for execution. After completing the test tasks, the hardware module writes back the status information to the "Complete" register of WQS. In addition, the CPU can also control the tracking function through data interaction with the IOM.
[0215] When performing data tracking, the tracking log for the test task can be stored in the tracking cache. In addition, the tracking hardware unit can output the tracking log corresponding to each hardware module to the tracking cache. It should be understood that on the time axis, the acquisition time of the tracking log of the hardware modules 1-n increases step by step.
[0216] In this embodiment, a data tracking device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.
[0217] This embodiment provides a data tracking device, such as Fig.14 As shown, including:
[0218] The splitting module 1401 is used to obtain the data processing task in the target chip, and split the data processing task based on the preset template corresponding to the data processing request to obtain sub-processing tasks;
[0219] A setting module 1402, used to set a tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task;
[0220] The execution module 1403 is used to execute the sub-processing task and output the tracking information according to the tracking configuration of the sub-processing task.
[0221] In some optional implementations, the setting module 1402 is further configured to:
[0222] When the tracking configuration is enabled for the sub-processing task, export parameters and security parameters are set for the sub-processing task, wherein the export parameters are used to indicate whether to export the tracking information corresponding to the sub-processing task when the sub-processing task fails to execute, and the security parameters are used to indicate whether to allow the tracking information to be lost;
[0223] Set the level parameters of the sub-processing tasks based on the data volume of the sub-processing tasks;
[0224] The data width of the trace configuration is set according to the export parameters, security parameters, and level parameters.
[0225] In some optional implementations, the setting module 1402 is further configured to:
[0226] Obtaining a preset processing level, wherein different preset processing levels correspond to different amounts of tracking information data;
[0227] The data volume of the tracking information of the sub-processing task is estimated, and a target processing level matching the estimated result is determined in the preset processing levels, and a level parameter is generated according to the target processing level.
[0228] In some optional implementations, the execution module 1403 is further configured to:
[0229] Monitor the remaining bandwidth in the target chip;
[0230] When the remaining bandwidth matches the trace configuration of the sub-processing task, the trace information is output.
[0231] In some optional implementations, the execution module 1403 is further configured to:
[0232] Generate a task chain for the sub-processing task based on the execution steps corresponding to the data processing task;
[0233] Write the first address of the first sub-processing task in the task chain into the task queue;
[0234] Based on the first address, calculating the second addresses of the remaining sub-processing tasks in the task chain, and writing the second addresses into the task queue;
[0235] Based on the task queue, the sub-processing tasks in the task chain are executed sequentially.
[0236] In some optional implementations, the execution module 1403 is further configured to:
[0237] Get the data length of each sub-processing task in the first sub-processing task;
[0238] Based on the sum of the first address and the data length, the second address of each remaining sub-processing task is calculated respectively.
[0239] In some optional implementations, the execution module 1403 is further configured to:
[0240] Execute the sub-processing tasks in the task queue in parallel and count the number of sub-processing tasks executed;
[0241] If the number of executions exceeds the number of sub-processing tasks in the task chain, an error message is generated.
[0242] In some optional implementations, the splitting module 1401 is further configured to:
[0243] Determine a preset template that matches a task type of the data processing task;
[0244] Based on the preset template, the data processing request is split according to the execution steps of the data processing task to obtain sub-processing tasks.
[0245] In some optional implementations, the splitting module 1401 is further configured to:
[0246] Based on the task type, a target control block is allocated for the task type in the control block of the target chip, and a preset template is determined according to the target control block;
[0247] Map and bind preset templates to task types.
[0248] In some optional implementations, the splitting module 1401 is further configured to:
[0249] In response to the template adjustment instruction, the target control block matching the task type is recalculated, and the preset template is updated according to the calculation result.
[0250] In some optional embodiments, the device is also used for:
[0251] Generate a test chain based on the functional blocks corresponding to each functional module in the target chip;
[0252] Generate a test task through the test software in the target chip, wherein the test task is used to test the performance of the functional module in the target chip;
[0253] Execute the test task through the test chain, track the execution process of the test task, and obtain the test results.
[0254] In some optional embodiments, the device is also used for:
[0255] During the operation of the target chip, the execution process of each functional block in the test chain is tracked to obtain a tracking log;
[0256] The trace log is output through the data transmission interface in the target chip, and the test result is determined according to the trace log.
[0257] In some optional embodiments, the device is also used for:
[0258] Based on the delay information in the tracking log, analyze the performance of the functional module corresponding to the tracking log to obtain the sub-test results;
[0259] The test results are determined based on the sub-test results corresponding to each functional module.
[0260] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.
[0261] The data tracking device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.
[0262] The embodiment of the present invention also provides a computer device having the above Fig.14 The data tracking device is shown.
[0263] See also Fig.15 , Fig.15is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Fig.15 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Fig.15 A processor 10 is taken as an example.
[0264] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware integrated circuit. The hardware integrated circuit may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable logic gate array, a general purpose array logic or any combination thereof.
[0265] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.
[0266] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
[0267] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.
[0268] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.
[0269] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.
[0270] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.
[0271] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the accompanying description.
Claims
1. A data tracking method, characterized in that: The method comprises: Acquire a data processing task in a target chip, and split the data processing task based on a preset template corresponding to the data processing request to obtain sub-processing tasks; Setting a tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task; The sub-processing task is executed, and tracking information is output according to the tracking configuration of the sub-processing task.
2. The method according to claim 1, characterized in that The step of setting a tracking configuration for the sub-processing task includes: In the case where the sub-processing task has the tracing configuration turned on, setting export parameters and security parameters for the sub-processing task, wherein the export parameters are used to indicate whether to export the tracing information corresponding to the sub-processing task when the sub-processing task fails to execute, and the security parameters are used to indicate whether to allow the tracing information to be lost; Setting a level parameter of the sub-processing task based on the data volume of the sub-processing task; The data bit width of the tracking configuration is set according to the export parameter, the security parameter and the level parameter.
3. The method according to claim 2, characterized in that The step of setting the level parameter of the sub-processing task based on the data volume of the sub-processing task includes: Obtaining a preset processing level, wherein different preset processing levels correspond to different amounts of tracking information data; The data volume of the tracking information of the sub-processing task is estimated, and a target processing level matching the estimated result is determined in the preset processing levels, and a level parameter is generated according to the target processing level.
4. The method according to claim 1, characterized in that: Outputting the tracking information according to the tracking configuration includes: monitoring the remaining bandwidth in the target chip; When the remaining bandwidth matches the tracking configuration of the sub-processing task, the tracking information is output.
5. The method according to claim 1, characterized in that The executing the sub-processing task includes: generating a task chain for the sub-processing task based on the execution steps corresponding to the data processing task; Writing the first address of the first sub-processing task in the task chain into the task queue; Based on the first address, calculating the second addresses of the remaining sub-processing tasks in the task chain, and writing the second addresses into the task queue; Based on the task queue, the sub-processing tasks in the task chain are executed in sequence.
6. The method according to claim 5, characterized in that The calculating, based on the first address, second addresses of remaining sub-processing tasks in the task chain includes: Obtaining the data length of each sub-processing task in the first sub-processing task; Based on the sum of the first address and the data length, the second address of each of the remaining sub-processing tasks is calculated respectively.
7. The method according to claim 5, characterized in that The method further comprises: Executing the sub-processing tasks in the task queue in parallel, and counting the number of sub-processing tasks executed; When the execution quantity exceeds the number of sub-processing tasks in the task chain, an error message is generated.
8. The method according to claim 1, characterized in that The step of splitting the data processing task based on a preset template corresponding to the data processing request to obtain sub-processing tasks includes: Determining a preset template that matches the task type of the data processing task; Based on the preset template, the data processing request is split according to the execution steps of the data processing task to obtain sub-processing tasks.
9. The method according to claim 8, characterized in that The method further comprises: Based on the task type, assigning a target control block to the task type in a control block of a target chip, and determining the preset template according to the target control block; The preset template is mapped and bound to the task type.
10. The method according to claim 9, characterized in that The method further comprises: In response to the template adjustment instruction, the target control block matching the task type is recalculated, and the preset template is updated according to the calculation result.
11. The method according to claim 1, characterized in that: The method further comprises: Generate a test chain based on the functional blocks corresponding to each functional module in the target chip; Generate a test task through the test software in the target chip, wherein the test task is used to test the performance of the functional module in the target chip; The test task is executed through the test chain, and the execution process of the test task is tracked to obtain the test result.
12. The method according to claim 11, characterized in that The tracking of the execution process of the test task to obtain the test result includes: During the operation of the target chip, the execution process of each functional block in the test chain is tracked to obtain a tracking log; The tracking log is outputted through a data transmission interface in the target chip, and the test result is determined according to the tracking log.
13. The method according to claim 12, characterized in that Determining the test result according to the tracking log includes: Based on the delay information in the tracking log, analyzing the performance of the functional module corresponding to the tracking log to obtain a sub-test result; The test result is determined according to the sub-test results corresponding to each of the functional modules.
14. A data tracking system, characterized in that: The system comprises: an IO management hardware module IOM, a work queue organization hardware unit WQS, a tracking hardware unit and a tracking cache; The IOM is used to obtain a data processing task in a target chip, split the data processing task based on a preset template corresponding to the data processing request to obtain a sub-processing task, and set a tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task; The WQS is used to allocate the sub-processing task to the corresponding functional module in the target chip so that the functional module performs the sub-processing task; The tracing hardware unit is used to output tracing information to the tracing cache according to the tracing configuration of the sub-processing task during the process of the functional module executing the sub-processing task; The tracking cache is used to output the tracking information to an external device through the data transmission interface of the target chip.
15. The system according to claim 14, characterized in that The system also includes: a central processing unit CPU; The CPU is used to generate a test chain based on the functional blocks corresponding to each functional module in the target chip, and generate a test task through the test software in the target chip, and execute the test task through the test chain, and track the execution process of the test task to obtain a test result, wherein the test task is used to test the performance of the functional module in the target chip.
16. The system according to claim 14, characterized in that The tracking cache is also used to: track the execution process of each functional block in the test chain during the operation of the target chip, obtain a tracking log, output the tracking log through the data transmission interface in the target chip, and determine the test result based on the tracking log.
17. A data tracking device, characterized in that: The device comprises: A splitting module, used for acquiring a data processing task in a target chip, and splitting the data processing task based on a preset template corresponding to the data processing request to obtain sub-processing tasks; A setting module, used to set a tracking configuration for the sub-processing task, wherein the tracking configuration is used to indicate a tracking mode corresponding to the sub-processing task; The execution module is used to execute the sub-processing task and output the tracking information according to the tracking configuration of the sub-processing task.
18. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the data tracking method according to any one of claims 1 to 14 by executing the computer instructions.
19. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the data tracking method according to any one of claims 1 to 14.
20. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the data tracking method according to any one of claims 1 to 14.
