An internal signal acquisition device and method for on-chip configurable logic blocks.

By integrating an internal signal acquisition device into the microcontroller, the problem of difficult observation of internal signals in the CLB in the prior art is solved, realizing flexible and accurate observation and diagnosis, supporting dynamic adjustment and closed-loop linkage, and suitable for on-site diagnosis.

CN123113840APending Publication Date: 2026-11-17青岛本原微电子有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611635515.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-22
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies struggle to efficiently and non-intrusively observe and diagnose node signals within on-chip configurable logic blocks (CLBs) in microcontrollers, particularly the timing behavior of LUT outputs, FSM status registers, and counter values. Furthermore, existing methods are resource-intensive or dependent on external toolchains, hindering flexible observation and diagnosis.

Method used

Design a signal acquisition device integrated within a configurable logic block, including observation point selection, trigger condition detection, sampling control, and storage units. The observation points and trigger conditions are configured by the CPU to achieve real-time acquisition and storage of internal signals without occupying configurable logic resources or relying on external toolchains.

Benefits of technology

It enables flexible, accurate, and non-intrusive observation of CLB internal signals without consuming CLB resources, supports dynamic adjustment of observation points and triggering conditions, shortens the debugging iteration cycle, provides closed-loop linkage diagnostic capabilities, and is suitable for on-site diagnostics after deployment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123113840A_ABST
    Figure CN123113840A_ABST
Patent Text Reader

Abstract

This invention discloses an internal signal acquisition device and method for an on-chip configurable logic block, belonging to the field of microcontroller debugging and diagnostic technology. The device is embedded within the on-chip configurable logic block. During operation, the observation point selection unit selects the observed signal from multiple internal observation nodes of the configurable logic array. A trigger condition unit performs condition detection on the observed signal and / or user logic trigger signals. The sampling control unit controls the acquisition window according to pre-trigger depth and post-trigger depth, selectively storing data based on sampling enable conditions. A readback interface allows the CPU to read the acquired data stored in the storage unit. Trigger events can be fed back to the configurable logic array to achieve closed-loop linkage between acquisition and the logic under test. This invention does not occupy configurable logic resources, the observation point can be dynamically adjusted during operation, the trigger and the signal under test are in the same domain, and no external debugging toolchain is required. It can cover observation needs from development debugging to field diagnostics and even during processor power domain power outages.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of microcontroller debugging and diagnostic technology, and in particular to an internal signal acquisition device and method for on-chip configurable logic blocks. Background Technology

[0002] A Configurable Logic Block (CLB) is a programmable logic unit integrated within a microcontroller (MCU) chip. It includes submodules such as lookup tables (LUTs), finite state machines (FSMs), and counters, allowing users to implement customized digital logic functions through software configuration. The logic functions implemented in the CLB typically handle high-speed real-time tasks, such as pulse protection, protocol interfaces, and signal filtering. The timing behavior of its internal node signals directly determines the correctness of the function.

[0003] In the development, debugging, and troubleshooting of CLBs, observing the signals of their internal nodes presents significant challenges. The first existing approach involves using external testing instruments: routing the signal to be observed to the chip pins and probing the pins with an oscilloscope or logic analyzer. This approach has several limitations: the number of chip pins is limited, and many are multiplexed as functional pins, resulting in a very small number of signals that can be observed simultaneously; pin driving and routing introduce additional delays, meaning the observation results cannot reflect the true timing of internal nodes; it cannot observe intermediate nodes in unexposed interconnect networks; and high-speed signals exiting via pins may cause signal integrity issues.

[0004] The second existing approach utilizes processor-side debugging units within the microcontroller, such as breakpoint units, data watchpoint units, and instruction tracing units. These debugging units observe access behavior on the processor bus, program execution flow, or variable access. Their triggering and acquisition mechanisms are designed around the processor architecture and cannot observe user logic node signals such as LUT outputs, FSM status registers, counter values, and interconnect buses within configurable logic blocks (CLBs), rendering them ineffective for debugging the internal logic of the CLBs.

[0005] The third existing approach comes from the field of programmable logic devices: integrating the logic analyzer core as a soft core along with the user design into the programmable logic resources, collecting data and storing it in the device's internal memory block, then reading it out through the debug port. This approach has fundamental drawbacks when applied to CLB scenarios: First, the analyzer core occupies valuable configurable logic and storage resources, competing with the user logic for the already limited CLB resources, which can severely hinder the complete implementation of the user logic. Second, since the analyzer core is compiled along with the user design, the set of observation signals is fixed during construction, making it impossible to dynamically adjust the observation points based on debugging progress at runtime. Third, this approach relies on external debugging toolchains and dedicated debug ports for configuration and readback, which cannot be independently completed by the microcontroller's own CPU in software, making it unsuitable for post-deployment field diagnostics.

[0006] Existing technologies also include schemes that store processor trace information in the chip's in-system memory, and schemes that integrate a logic analyzer as solidified hardware within the chip to observe processor bus transactions or inter-module interface signals. While the latter does not occupy configurable logic resources, its observation targets are bus interfaces and protocol signals that are fixed during the chip design phase. The acquired content is bus transaction records rather than the raw waveforms of user logic nodes, and therefore cannot be used to analyze the timing behavior of LUT outputs, FSM state transitions, or counter values ​​within the configurable logic. Summary of the Invention

[0007] The purpose of this invention is to propose an internal signal acquisition device and method for an on-chip configurable logic block. Without occupying the configurable logic resources of the on-chip configurable logic block or relying on an external debugging toolchain, it can perform real-time acquisition of node signals inside the configurable logic block in a runtime selectable manner. The triggering condition is jointly generated by the user logic within the configurable logic block and the triggering configuration of the acquisition device and is in the same domain as the signal being measured. The acquired data is stored in a fixed storage resource integrated inside the configurable logic block, and the CPU can access it through conventional registers or read it back directly via DMA.

[0008] The present invention is implemented using the following technical solutions: An internal signal acquisition device for an on-chip configurable logic block is proposed, integrated within the on-chip configurable logic block, wherein the configurable logic block includes a configurable logic array, comprising: The observation point selection unit has its input coupled to multiple internal observation nodes of a configurable logic array, and is configured to select several internal observation nodes as the observed signal output according to the observation point configuration information. The trigger condition unit is connected to the observation point selection unit and the trigger signal output terminal of the configurable logic array, and is configured to: perform condition detection on the observed signal and / or the user logic trigger signal output from the configurable logic array according to the trigger configuration information, and generate a trigger event when the condition is met; The sampling control unit, connected to the triggering condition unit, is configured to: write the observed signal into the storage unit under the control of the sampling clock, and control the position of the acquisition window relative to the triggering event according to the configuration of the pre-trigger depth and the post-trigger depth; The storage unit is configured to: cyclically write the observed signal until the subsequent depth counting is completed, then stop writing and set the acquisition completion flag; The readback interface is connected to the CPU bus interface and is configured to allow the CPU to read the acquired data in the storage unit and the configuration and status registers of the acquisition device.

[0009] In some embodiments of the present invention, the observation point selection unit is fixed hardware and does not occupy the configurable logic resources of the configurable logic array.

[0010] In some embodiments of the present invention, the storage unit is a ring buffer memory integrated within the configurable logic block.

[0011] In some embodiments of the present invention, the triggering condition unit includes a comparator group and a trigger sequence state machine; The comparator group is configured to perform equal to, greater than, less than, and edge-changing matching detection on the observed signal, and combine multiple comparison results according to a Boolean expression; The trigger sequence state machine is configured to detect the sequential fulfillment of multiple trigger conditions in chronological order to achieve sequence triggering.

[0012] In some embodiments of the present invention, the trigger source of the trigger condition unit includes a user logic trigger signal derived from the configurable logic array; the user logic trigger signal is generated by a user-configured logic function and is in the same clock domain as the observed signal.

[0013] In some embodiments of the present invention, the sampling control unit includes a sampling enable condition detector, configured to perform writing only when the observed signal meets the sampling enable condition, thereby achieving selective storage.

[0014] In some embodiments of the present invention, the apparatus further includes: The trigger feedback path is configured to feed the trigger event as an event input to the configurable logic array, so that the user logic can perform pause, snapshot or state switching actions when the trigger occurs, thereby realizing closed-loop linkage between data acquisition and the logic under test.

[0015] In some embodiments of the present invention, the sampling clock is the clock of the configurable logic block or its frequency division clock, and the frequency division ratio is configurable.

[0016] In some embodiments of the present invention, the device is set in a second power domain independent of the processor power domain along with the configurable logic block, and continues to sample during the power failure of the processor power domain, collecting data for readback after the processor power domain is restored.

[0017] Compared with the prior art, the advantages and positive effects of the present invention are: 1. Zero logical resource consumption.

[0018] The acquisition device of this invention is a solidified hardware integrated inside a configurable logic block. The selection of observation points, trigger detection, sampling control and data storage do not consume the resources of the configurable logic array such as LUT, FSM, Counter. The user logic implementation and debugging capabilities do not interfere with each other, which is different from the soft-core logic analyzer solution that occupies programmable logic resources.

[0019] 2. Observation points can be configured during runtime.

[0020] The observation point selection information is configuration data written to registers at runtime. During debugging, the set of observation signals can be changed at any time according to fault clues without modifying the configuration data of user logic or recompiling, which significantly shortens the debugging iteration cycle.

[0021] 3. The trigger and the logic being tested are in the same domain.

[0022] The trigger signal can come directly from the user logic in the configurable logic array. It is in the same clock domain and synchronized with the observed signal. There is no cross-domain sampling deviation between the trigger time and the signal transition, ensuring that the trigger positioning is accurate to the sampling clock edge.

[0023] 4. Non-invasive observation.

[0024] The internal observation node is brought out to the observation point selection unit during chip design. The observation action does not change the timing path and resource usage of the logic under test. The observation result reflects the actual behavior of the internal node, avoiding the additional delay and signal integrity issues of the pin-out method.

[0025] 5. Complete history before and after triggering.

[0026] The ring buffer, combined with the pre-trigger depth configuration, can save both the signal history before the triggering event and the behavior after the triggering event. With the sampling enable condition, selective storage can be achieved, covering the most critical observation window with a limited storage depth.

[0027] 6. Closed-loop linkage diagnosis.

[0028] The triggered event can be fed back to the configurable logic array as an event input. The user logic can perform actions such as state snapshot and fault freeze at the moment of triggering. The acquisition device and the logic under test form a closed loop, supporting a diagnostic mode that freezes the scene when the signal is abnormal.

[0029] 7. Available on site.

[0030] Configuration and readback are both completed through CPU regular register access or DMA, without relying on external debugging toolchains and dedicated debugging ports. The firmware can autonomously perform data acquisition and diagnosis in the deployed field system.

[0031] This invention also proposes an internal signal acquisition method for on-chip configurable logic blocks, applied to the aforementioned internal signal acquisition device for on-chip configurable logic blocks, comprising: The CPU configures the observation point selection information and selects several internal observation nodes from the configurable logic array as the observed signals. CPU configuration trigger condition information; the trigger conditions include matching conditions for the observed signal and / or user logic trigger signals derived from the configurable logic array; The CPU configures the pre-trigger depth and post-trigger depth and starts acquisition, using a sampling clock to cyclically write the observed signal into the storage unit; The triggering condition is detected, and when the condition is met, a triggering event is generated and depth counting is started. When the depth counting is completed, writing stops, the acquisition completion flag is set, and an acquisition completion interrupt is generated. In response to the interrupt, the CPU reads the acquired data from the storage unit. The pre-trigger data in the acquired data is located in the address range prior to the trigger record, reflecting the signal history before the trigger.

[0032] In some embodiments of the present invention, the method further includes: The CPU modifies the observation point selection information and / or trigger condition information according to the diagnostic needs, and repeatedly executes the acquisition to achieve dynamic adjustment of the observation points during runtime.

[0033] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In the method of the present invention, the observation point selection information and trigger condition information are configured by the CPU using register access, so that the observation point set and trigger conditions can be determined at runtime without recompiling the configuration data of the configurable logic array; the CPU further configures the pre-trigger depth and post-trigger depth and starts acquisition; the device writes the observed signal to the storage unit cyclically with the sampling clock; the trigger condition unit detects the trigger condition step by step; when the condition is met, a trigger event is generated and the post-trigger depth counting begins; after the counting is completed, writing stops, the acquisition completion flag is set, and an acquisition completion interrupt is generated; the CPU responds to the interrupt and reads the acquired data in the storage unit, wherein the pre-trigger data is located in the address range before the trigger record, reflecting the signal history before the trigger; the CPU can also modify the observation points and trigger conditions according to the current diagnostic results. Repeated acquisition results in the following technical advantages: First, the entire acquisition process is completed autonomously by the CPU through conventional bus access and interrupt mechanisms, without relying on external debugging toolchains and dedicated debugging ports. Second, the adjustment of observation points and trigger conditions only involves register writing, without modifying user logic configuration data or recompiling. Third, by configuring pre-trigger depth and post-trigger depth, a single acquisition can simultaneously obtain the signal history before triggering and the behavior record after triggering, avoiding the loss of key information before triggering. Fourth, the CPU can dynamically change the set of observation signals and trigger conditions after each acquisition based on newly discovered fault clues, forming a closed-loop iterative observation process of configuration → acquisition → analysis → reconfiguration. This balances on-site availability, observation flexibility, and diagnostic completeness in CLB internal logic debugging, significantly shortening the cycle from fault phenomenon to root cause location.

[0034] Other features and advantages of the present invention will become clearer after reading the detailed description of the embodiments of the present invention in conjunction with the accompanying drawings. Attached Figure Description

[0035] Figure 1 This invention presents the device architecture for an internal signal acquisition device with on-chip configurable logic blocks. Figure 2 These are the execution steps of the internal signal acquisition method for on-chip configurable logic blocks proposed in this invention; Figure 3 This is a timing diagram of internal signal acquisition and triggering in an embodiment of the present invention. Detailed Implementation

[0036] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0037] like Figure 1 As shown, this is the signal acquisition device integrated within the configurable logic block proposed in this embodiment. It is a fixed hardware macro consisting of five parts: The observation point selection unit has its input coupled to multiple internal observation nodes of the configurable logic array. The outgoing observation nodes include, but are not limited to: the outputs of each lookup table submodule, the state register outputs of each finite state machine submodule, the count value outputs of each counter submodule, the signal bits of the logic interconnect bus, and the outputs of the input signal processing paths (filters / synchronizers), totaling N channels (e.g., 64 channels). The observation point selection unit contains M (e.g., 8) multiplexers, each controlled by a corresponding observation point configuration register, selecting one path from the N observation nodes. The outputs of the M multiplexers constitute an M-bit observed signal bus. The observation point configuration register is a standard memory-mapped register, which can be rewritten by the CPU at any time during runtime, taking effect immediately in the next clock cycle, without requiring recompilation of the configurable logic array's configuration data.

[0038] The trigger condition unit comprises a comparator group and a trigger sequence state machine. The comparator group independently sets matching conditions for each observed signal channel: equal to a specified value, greater than a threshold, less than a threshold, rising edge detected, falling edge detected, or a change occurs. The comparison results of each channel are combined into trigger conditions using configurable Boolean combinational logic (AND / OR / NOT). The trigger sequence state machine supports up to K levels (e.g., 4 levels) of sequential triggering: each level holds independent conditions and timeout windows, and a final trigger event is generated only when the conditions at each level are met sequentially. Furthermore, the trigger condition unit also has an external trigger input terminal, receiving user logic trigger signals directly derived from the configurable logic array. These signals are generated by user-configured logic functions and reside in the same clock domain as the observed signal, such as a custom protection flag or a specific state decoded output of the state machine.

[0039] The sampling control unit operates under the control of a sampling clock. The sampling clock is a configurable logic block clock or a configurable division of it (1 / 1, 1 / 2, 1 / 4, etc.) to adapt to different observation needs. For example, full-speed sampling is used to capture single-cycle glitches, while reduced-frequency sampling is used for long-term trend observation. The sampling control unit contains a pre-trigger counter and a post-trigger counter: acquired data is continuously and cyclically written to the storage unit; a post-trigger counter is started upon the arrival of a trigger event, and writing stops when the count reaches the configured post-trigger depth. The sampling control unit also includes a sampling enable condition detector, which only performs writing when the observed signal meets the enable condition (such as a valid strobe signal), achieving selective storage and improving the effective information density of the storage window.

[0040] The storage unit is a ring-shaped buffer memory integrated within the configurable logic block, with a depth and width that can be designed, for example, 256×32 bits. The write pointer increments cyclically, and the read address is generated by the readback interface. The storage unit is a fixed SRAM macro and does not occupy configurable logic array resources.

[0041] The readback interface includes a configuration register group, a status register group, and a data read port. The CPU reads the acquired data word by word via the bus interface using register read mode, or configures the DMA channel to move the entire acquired data segment to system memory in batches. Events such as acquisition completion, trigger occurrence, and buffer overwrite are reported in the form of status bits and interrupts.

[0042] The trigger feedback path sends the trigger event as an event input to the event selection network of the configurable logic array. The user logic can configure it as a pause signal, snapshot signal, or state transition condition.

[0043] As attached Figure 2 As shown, the internal signal acquisition device for the on-chip configurable logic block described above performs the internal signal acquisition method according to the following steps: Step S201 (Configure Observation Points): The CPU configures the observation point selection information and selects several internal observation nodes from the configurable logic array as the observed signals.

[0044] The option is to perform this at runtime without recompiling the configuration data of the configurable logic array. For example, if a debugger suspects that a malfunction of a certain protection logic is caused by a filter output glitch, the CPU can configure channel 0 of the observation point selection unit as the output node of the filter, channel 1 as the state machine state register, and channel 2 as the counter count value. This configuration only writes to the observation point configuration register, and the configuration data of the user logic remains unchanged.

[0045] Step S202 (Configure Trigger Conditions): Configure CPU trigger condition information.

[0046] Configure the trigger condition as follows: the status value of channel 1 is equal to error code 0x5A. Optionally, further enable sequence triggering: first, a channel 2 count overflow occurs, followed by a status value equal to 0x5A within 100 sampling periods.

[0047] Step S203 (Configure Depth and Start): The CPU configures the pre-trigger depth and post-trigger depth and starts acquisition, so that the observed signal is cyclically written to the storage unit by the sampling clock.

[0048] The pre-trigger depth is set to 64, the post-trigger depth to 192, and the total window size is 256 sampling points. The start bit is written, and the acquisition device begins cyclically writing to the circular buffer at the sampling clock.

[0049] Step S204 (Trigger Detection): Detect the trigger condition. When the condition is met, generate a trigger event and start triggering depth counting.

[0050] As attached Figure 3As shown, the trigger condition unit detects the trigger condition frame by frame; when the observed signal SIG meets the condition, a trigger event TRIG is generated, the sampling control unit locks the trigger recording position, and trigger counting is started after activation. In the diagram, SCLK is the sampling clock, TRIG_COND is the trigger condition, WEN is the memory write enable, ADDR is the memory write address pointer, and DONE is the acquisition completion flag.

[0051] Step S205 (Stop and Notification): When the depth counting is completed, stop writing, set the acquisition completion flag, and generate an acquisition completion interrupt.

[0052] After the trigger count reaches 192, writing stops, the acquisition completion flag is set, and an interrupt is generated.

[0053] Step S206 (Readback Analysis): The CPU responds to the interrupt and reads the collected data from the storage unit.

[0054] The CPU responds to the interrupt and reads 256 sampling points via the readback interface: 64 points before the trigger record reflect the signal history before the trigger (e.g., whether a glitch occurred before the state transition), and the remaining 192 points reflect the behavior after the trigger. If DMA is configured, the entire data segment is moved into system memory for software analysis.

[0055] Step S207 (Dynamic Reconfiguration): The CPU modifies the observation point selection information and / or trigger condition information according to the diagnostic needs, and repeats the acquisition to achieve dynamic adjustment of the observation points during runtime.

[0056] Based on the observation results, the debugging team replaced the observation point with a new node, modified the triggering conditions, and repeated the data collection. The entire iterative process only involved register writing and did not require recompiling the user logic.

[0057] The following specific embodiment illustrates a typical use of the trigger feedback path in the internal signal acquisition device of the present invention, demonstrating the closed-loop linkage between the device and user logic.

[0058] Suppose the user has implemented motor protection logic in a configurable logic array and configured the trigger event of the acquisition device as one of the event inputs of the protection logic. The trigger condition for the acquisition device is configured as: overcurrent flag valid.

[0059] When a trigger event occurs, on the one hand, sampling stops and the fault window data is frozen in the ring buffer; on the other hand, the trigger feedback path sends the event to the protection logic in the configurable logic array. The protection logic then performs a state snapshot: latches all internal register values ​​into the snapshot register group and switches the output to the safe state.

[0060] After a fault occurs, the CPU simultaneously obtains two pieces of information: the signal waveform in the ring buffer (a dynamic process with millisecond-level time resolution) and the instantaneous state in the snapshot register (all static states at the moment of triggering). The two corroborate each other, significantly shortening the fault location time.

[0061] The following specific embodiment describes how the sampling enable condition is used to expand the effective observation window in the sampling control of the present invention.

[0062] Assuming the logic under test only produces meaningful behavior during the validity of the strobe signal SEL, the observation target is an occasional anomaly during long-term operation. However, the memory depth of the ring buffer is fixed. If it is written to every sampling clock, it is highly unlikely that an anomaly that occurs only after several hours of operation will be captured.

[0063] Under the design of this invention, the sampling enable condition is configured as "SEL active," and the circular buffer is written only during the active period of the logic under test, with storage during invalid periods being skipped. The buffer depth of 256, combined with the enable condition filtering, expands the equivalent observation window to the entire active period, allowing occasional anomalies that occur after running for several hours to be captured with only a single acquisition.

[0064] The following example demonstrates the application of the internal signal acquisition device designed in this invention to autonomous acquisition in a low-power scenario.

[0065] This embodiment combines the internal acquisition device with the independent power domain mechanism of the configurable logic block (independent of the CPU's power domain): when the processor's power domain is powered down and the configurable logic block runs independently, the acquisition device maintains power supply and sampling along with the configurable logic block, and the triggered events and acquired data can still be read back after the processor wakes up. This provides a direct observation method for the difficult problem of "faults only occurring during system hibernation": acquisition is started and trigger conditions are configured before the processor goes into hibernation, and the signal records frozen during hibernation are read after waking up.

[0066] It should be noted that the above description is not intended to limit the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. An internal signal acquisition device for an on-chip configurable logic block, integrated within the on-chip configurable logic block, wherein the configurable logic block includes a configurable logic array, characterized in that, include: The observation point selection unit has its input coupled to multiple internal observation nodes of a configurable logic array, and is configured to select several internal observation nodes as the observed signal output according to the observation point configuration information. The trigger condition unit is connected to the observation point selection unit and the trigger signal output terminal of the configurable logic array, and is configured to: perform condition detection on the observed signal and / or the user logic trigger signal output from the configurable logic array according to the trigger configuration information, and generate a trigger event when the condition is met; The sampling control unit, connected to the triggering condition unit, is configured to: write the observed signal into the storage unit under the control of the sampling clock, and control the position of the acquisition window relative to the triggering event according to the configuration of the pre-trigger depth and the post-trigger depth; The storage unit is configured to: cyclically write the observed signal until the subsequent trigger depth counting is completed, at which point writing stops and the acquisition completion flag is set; The readback interface is connected to the CPU bus interface and is configured to allow the CPU to read the acquired data in the storage unit and the configuration and status registers of the acquisition device.

2. The internal signal acquisition device with on-chip configurable logic blocks according to claim 1, characterized in that, The observation point selection unit is fixed hardware and does not occupy the configurable logic resources of the configurable logic array.

3. The internal signal acquisition device with on-chip configurable logic blocks according to claim 1, characterized in that, The triggering condition unit includes a comparator group and a trigger sequence state machine; The comparator group is configured to perform equal to, greater than, less than, and edge-changing matching detection on the observed signal, and combine multiple comparison results according to a Boolean expression; The trigger sequence state machine is configured to detect the sequential fulfillment of multiple trigger conditions in chronological order to achieve sequence triggering.

4. The internal signal acquisition device with on-chip configurable logic blocks according to claim 1, characterized in that, The trigger source of the trigger condition unit includes a user logic trigger signal derived from the configurable logic array; the user logic trigger signal is generated by the user-configured logic function and is in the same clock domain as the observed signal.

5. The internal signal acquisition device with on-chip configurable logic blocks according to claim 1, characterized in that, The sampling control unit includes a sampling enable condition detector, configured to perform writing only when the observed signal meets the sampling enable condition, thereby achieving selective storage.

6. The internal signal acquisition device with on-chip configurable logic blocks according to claim 1, characterized in that, The device further includes: The trigger feedback path is configured to feed the trigger event as an event input to the configurable logic array, so that the user logic can perform pause, snapshot, or state switching actions when the trigger occurs.

7. The internal signal acquisition device with on-chip configurable logic blocks according to claim 1, characterized in that, The sampling clock is the clock of the configurable logic block or its divided clock, and the division ratio is configurable.

8. The internal signal acquisition device with on-chip configurable logic blocks according to claim 1, characterized in that, The device is located in a second power domain, independent of the processor power domain, along with the configurable logic block. It continues to sample during the power outage of the processor power domain and collects data for readback after the processor power domain is powered on again.

9. A method for acquiring internal signals of an on-chip configurable logic block, applied in the on-chip configurable logic block internal signal acquisition device as described in any one of claims 1-8, characterized in that, include: The CPU configures the observation point selection information and selects several internal observation nodes from the configurable logic array as the observed signals. CPU configuration trigger condition information; the trigger conditions include matching conditions for the observed signal and / or user logic trigger signals derived from the configurable logic array; The CPU configures the pre-trigger depth and post-trigger depth and starts acquisition, using a sampling clock to cyclically write the observed signal into the storage unit; The triggering condition is detected, and when the condition is met, a triggering event is generated and depth counting is started. When the depth counting is completed, writing stops, the acquisition completion flag is set, and an acquisition completion interrupt is generated. In response to the interrupt, the CPU reads the acquired data from the storage unit.

10. The method for acquiring internal signals of an on-chip configurable logic block according to claim 9, characterized in that, The method further includes: The CPU modifies the observation point selection information and / or trigger condition information according to the diagnostic needs, and repeatedly executes the acquisition to achieve dynamic adjustment of the observation points during runtime.