Page table cache invalidation method, data processing device and readable medium

By extending TLBI operations in the CHI protocol and utilizing invalid operation information from non-ARM instruction sets, the compatibility issues of the CHI protocol on non-ARM instruction set CPU systems are resolved, achieving broader instruction set support.

CN119557243BActive Publication Date: 2026-11-17SANECHIPS TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202311085596.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-08-25
Publication Date
2026-11-17
Estimated Expiration
2043-08-25

AI Technical Summary

Technical Problem

The CHI protocol has compatibility issues in on-chip systems that include CPUs with non-ARM instruction sets, especially since it does not support ASID-related TLBI operations under EL3.

Method used

The first processing core sends invalid operation information to miscellaneous nodes, including operation information that is not defined in the ARM instruction set but is defined in non-ARM instruction sets such as RISC-V instruction sets, extending the CHI protocol to support TLBI operations for more instruction sets.

Benefits of technology

It decouples the CHI protocol from different instruction sets, solves the compatibility problem of non-ARM instruction set CPUs, and improves the scalability and applicability of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119557243B_ABST
    Figure CN119557243B_ABST
Patent Text Reader

Abstract

This disclosure provides a page table cache invalidation method. A first data processing device in a first processing core sends invalidation operation information to a miscellaneous node. The invalidation operation information includes at least invalidation operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set. The miscellaneous node then notifies a second data processing device in a second processing core to perform a page table cache invalidation operation based on the invalidation operation information. This disclosure extends the CHI protocol, supplementing the page table cache invalidation operations supported by the CHI protocol. It can support comprehensive TLBI operations, decoupling the CHI protocol from the instruction set used by the processing core, making the CHI protocol compatible with more instruction sets. This solves the compatibility problem of on-chip systems containing CPUs with non-ARM instruction sets, improving scalability and applicability. This disclosure also provides a data processing device and a readable medium.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and specifically to a method for invalidating page table cache, a data processing apparatus, and a readable medium. Background Technology

[0002] In a System-on-Chip (SoC), which consists of a single processor, processor cluster, graphics processor, memory controller, etc., the CHI (Coherent Hub Interface) bus is one of the most commonly used interconnect bus protocols. It is highly flexible, scalable, and can meet different performance, power consumption, and area requirements.

[0003] The CHI architecture provides a complete DVM (Distributed Virtual Memory) processing flow, enabling efficient management of virtual memory. However, the supported TLBI (Translation LookasideBuffer Invalidate) request types are limited by the ARM (Advanced RISC-V) instruction set. This means that various transactions defined by the CHI protocol are strongly correlated with the ARM instruction set, and the CHI protocol does not provide corresponding solutions for TLBI operations not supported by the ARM instruction set. For example, the CHI protocol does not support ASID (Address Space Identifier) ​​related TLBI operations at EL (Exception Level) 3, while some instruction sets, such as the RISC-V instruction set, support ASID-related TLBI operations in M-mode (EL3). This leads to compatibility issues when the CHI protocol is implemented in SoCs containing non-ARM instruction set CPUs. Summary of the Invention

[0004] This disclosure provides a method for invalidating page table cache, a data processing apparatus, and a readable medium.

[0005] In a first aspect, embodiments of this disclosure provide a method for invalidating a page table cache, the method being applied to a first data processing device in a first processing core, the method comprising:

[0006] Invalid operation information is sent to miscellaneous nodes so that the miscellaneous nodes can notify the second data processing device in the second processing core to perform an invalid page table cache operation based on the invalid operation information; wherein, the invalid operation information includes at least invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set.

[0007] In another aspect, embodiments of this disclosure also provide a data processing apparatus, including: one or more processors; a storage device storing one or more programs thereon; when the one or more programs are executed by the one or more processors, causing the one or more processors to implement the page table cache invalidation method as described above; and one or more I / O interfaces connected between the processors and the storage device, configured to enable information interaction between the processors and the storage device.

[0008] In another aspect, embodiments of this disclosure also provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed, implements the page table cache invalidation method as described above.

[0009] The page table cache invalidation method provided in this disclosure involves a first data processing device in a first processing core sending invalidation operation information to a miscellaneous node. This invalidation operation information includes at least invalidation operation information not defined in the ARM instruction set but defined in a non-ARM instruction set. The miscellaneous node then notifies a second data processing device in a second processing core to perform a page table cache invalidation operation based on the invalidation operation information. This disclosure extends the CHI protocol, supplementing the page table cache invalidation operations supported by the CHI protocol. It can support comprehensive TLBI operations, decoupling the CHI protocol from the instruction set used by the processing core. This allows the CHI protocol to be compatible with more instruction sets, solving the compatibility issues of on-chip systems containing CPUs with non-ARM instruction sets, and improving scalability and applicability. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of a page table cache invalidation process provided in an embodiment of the present disclosure, with the first data processing device as the execution entity. Figure 1 ;

[0011] Figure 2 This is a schematic diagram of the process for sending invalid operation information to a miscellaneous node according to an embodiment of this disclosure;

[0012] Figure 3 This is a schematic diagram of a page table cache invalidation process provided in an embodiment of the present disclosure, with the first data processing device as the execution entity. Figure 2 ;

[0013] Figure 4 A schematic diagram of the page table cache invalidation process provided for a specific example of this disclosure;

[0014] Figure 5 This embodiment of the present disclosure provides a schematic diagram of a page table cache invalidation process performed by a second data processing device. Figure 1 ;

[0015] Figure 6This embodiment of the present disclosure provides a schematic diagram of a page table cache invalidation process performed by a second data processing device. Figure 2 ;

[0016] Figure 7 This is a schematic diagram of the structure of a data processing apparatus provided in an embodiment of the present disclosure. Detailed Implementation

[0017] Exemplary embodiments will be described more fully below with reference to the accompanying drawings; however, these exemplary embodiments may be embodied in different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will enable those skilled in the art to fully understand the scope of this disclosure.

[0018] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0019] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. As used herein, the singular forms “a” and “the” are also intended to include the plural forms unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the said feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded.

[0020] The embodiments described herein can be described with reference to plan views and / or cross-sectional views using the ideal schematic diagrams of this disclosure. Therefore, the example illustrations can be modified according to manufacturing techniques and / or tolerances. Therefore, the embodiments are not limited to those shown in the drawings, but include modifications to configurations formed based on manufacturing processes. Therefore, the areas illustrated in the drawings are schematic in nature, and the shapes of the areas shown in the figures illustrate specific shapes of areas of an element, but are not intended to be limiting.

[0021] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and this disclosure, and will not be interpreted as having an idealized or overly formal meaning, unless expressly so defined herein.

[0022] To achieve efficient virtual memory management, the CHI protocol provides DVMOp (Distributed Virtual Memory operations), which includes the TLB maintenance instruction TLBI. The TLBI instruction can invalidate page tables at different granularities, thereby efficiently managing the virtual memory space of different processes, making physical memory invisible to applications, and achieving software and hardware isolation.

[0023] TLBI operations are performed only according to EL, VA (Virtual Address), and ASID; that is, invalid operation information can include EL, ASID, and VA information. The TLBI operations currently supported by the CHI protocol are shown in Tables 1 and 2 below:

[0024] Table 1

[0025] ASID valid VA valid Operation ARM Instruction 0b0 0b0 TLBI ALL EL1 TLBI ALLE1 0b0 0b1 TLBI by VA EL1 TLBI VAAE1 0b1 0b0 TLBI by ASID EL1 TLBI ASIDE1 0b1 0b1 TLBI by ASID and VA EL1 TLBI VAE1

[0026] Table 1 shows the TLBI operations corresponding to the ARM instructions related to invalid operation information under the first exception level (EL1). As shown in Table 1, when both ASID and VA are invalid, the "TLBI ALLE1" instruction in the ARM instruction set causes all RNs (Request Nodes) in EL1 to perform a TLBI operation; when ASID is invalid and VA is valid, the "TLBI VAAE1" instruction in the ARM instruction set causes the RN in EL1 that matches VA to perform a TLBI operation; when ASID is valid and VA is invalid, the "TLBI ASIDE1" instruction in the ARM instruction set causes the RN in EL1 that matches ASID to perform a TLBI operation; when both ASID and VA are valid, the "TLBI VAE1" instruction in the ARM instruction set causes the RN in EL1 that matches both ASID and VA to perform a TLBI operation.

[0027] Table 2

[0028] ASID valid VA valid Operation ARM Instruction 0b0 0b0 TLBI ALL EL3 TLBI ALLE3 0b0 0b1 TLBI by VA EL3 TLBI VAE3 0b1 0b0 Not Support NA 0b1 0b1 Not Support NA

[0029] Table 2 shows the TLBI operations corresponding to the ARM instructions related to invalid operation information under the third exception level (EL3). As shown in Table 2, when both ASID and VA are invalid, the "TLBI ALLE3" instruction in the ARM instruction set causes all RNs in EL3 to perform the TLBI operation; when ASID is invalid and VA is valid, the "TLBIVAE3" instruction in the ARM instruction set causes the RN in EL3 that matches VA to perform the TLBI operation; when ASID is valid and VA is invalid, and when both ASID and VA are valid, there is no corresponding operation instruction in the ARM instruction set, that is, the ARM instruction set does not define the corresponding operation, the ARM instruction set does not support the corresponding TLBI operation, and the CHI protocol does not provide a corresponding handling solution.

[0030] As can be seen from Tables 1 and 2, the CHI protocol does not support ASID-related TLBI operations in EL3. However, some non-RAM instruction sets, such as the RISC-V instruction set, support ASID-related TLBI operations in M-mode (EL3). This leads to compatibility issues when the CHI protocol is implemented in a SoC containing a non-ARM instruction set CPU.

[0031] To address the aforementioned problems, this disclosure provides a method for invalidating page table caches. The method is applied to a first data processing device within a first processing core. The first data processing device is a request node. It should be noted that different data processing devices reside within different processing cores of the same CPU. Figure 1 As shown, the page table cache invalidation method includes the following steps:

[0032] Step S11: Send invalid operation information to the miscellaneous node so that the miscellaneous node can notify the second data processing device in the second processing core to perform an invalid operation on the page table cache according to the invalid operation information; wherein, the invalid operation information includes at least invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set.

[0033] The first data processing device in the first processing core sends invalid operation information to the Miscellaneous Node (MN). The Miscellaneous Node then forwards this invalid operation information to other data processing devices in other processing cores within the same CPU, so that these other data processing devices can perform TLBI operations based on the invalid operation information. This invalid operation information is invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set. For example, this invalid operation information is defined in the RISC-V instruction set but not in the ARM instruction set, i.e., as shown in Table 2, for information where ASID is valid and VA is invalid, and for information where both ASID and VA are valid. This extends the CHI protocol, supplements and improves the ARM instruction set, and solves the compatibility issues of on-chip systems containing CPUs with non-ARM instruction sets.

[0034] The page table cache invalidation method provided in this disclosure involves a first data processing device in a first processing core sending invalidation operation information to a miscellaneous node. This invalidation operation information includes at least invalidation operation information not defined in the ARM instruction set but defined in a non-ARM instruction set. The miscellaneous node then notifies a second data processing device in a second processing core to perform a page table cache invalidation operation based on the invalidation operation information. This disclosure extends the CHI protocol, supplementing the page table cache invalidation operations supported by the CHI protocol. It can support comprehensive TLBI operations, decoupling the CHI protocol from the instruction set used by the processing core. This allows the CHI protocol to be compatible with more instruction sets, solving the compatibility issues of on-chip systems containing CPUs with non-ARM instruction sets, and improving scalability and applicability.

[0035] In some embodiments, the non-ARM instruction set may be the RISC-V instruction set.

[0036] In some embodiments, the invalid operation information not defined in the ARM instruction set but defined in a non-ARM instruction set includes first invalid operation information, or first invalid operation information and second invalid operation information. The first invalid operation information may include EL information and ASID information, and the second invalid operation information may include VA information, wherein the EL information may be third exception level information (EL3). That is, invalid operation information not defined in the ARM instruction set but defined in a non-ARM instruction set includes EL3 information and ASID information, or EL3 information, ASID information, and VA information.

[0037] Taking the RISC-V instruction set (a non-ARM instruction set) as an example, the RISC-V instruction set provides the Sfence.VAM and Hfence.VVAM instructions for TLB maintenance. The Sfence.VAM instruction supports operations in S-mode (corresponding to EL1 in ARM architecture) and M-mode (corresponding to EL3 in ARM architecture), and both instructions also have operations for ASID and VA. The mapping relationship between TLBI operations in the RISC-V and ARM instruction sets is shown in Table 3.

[0038] Table 3

[0039] RISC-V ARM Operation CHI Support Sfence VAM All S TLBI ALLE1 TLBI ALL Y Sfence VAM VA S TLBI VAAE1 TLBI by VA Y Sfence VAM ASID S TLBI ASIDE1 TLBI by ASID Y Sfence VAM VAA S TLBI VAE1 TLBI by ASID and VA Y Sfence VAM All M TLBI ALLE3 TLBI ALL Y Sfence VAM VA M TLBI VAE3 TLBI by VA Y Sfence VAM ASID M NA TLBI by ASID N Sfence VAM VAA M NA TLBI by ASID and VA N

[0040] As shown in Table 3, invalid operation information includes not only invalid operation information not defined in the ARM instruction set but defined in non-ARM instruction sets, but also invalid operation information defined in both the ARM and non-ARM instruction sets. That is, invalid operation information also includes third invalid operation information, or third invalid operation information and fourth invalid operation information. The third invalid operation information includes EL information, and the fourth invalid operation information includes VA information; wherein, the third invalid operation information is either first exception level information or third exception level information. That is, invalid operation information also includes EL information (EL1 information or EL3 information), or invalid operation information also includes EL information (EL1 information or EL3 information) and VA information. Furthermore, as shown in Table 3, the EL information in the first invalid operation information can also be EL1 information, meaning the first invalid operation information can also be EL1 information and ASID information.

[0041] As can be seen from Table 3, the CHI protocol does not support ASID-related TLBI operations under EL3. Through the extension of the CHI protocol in the embodiments of this disclosure, the Sfence VAM ASID M instruction and Sfence VAM VAA M instruction in the RISC-V instruction set can also be correctly processed in the CHI protocol.

[0042] In some embodiments, such as Figure 2 As shown, sending invalid operation information to the miscellaneous node (i.e., step S11) includes the following steps:

[0043] Step S111: Send a distributed virtual memory operation request message to the miscellaneous node. The distributed virtual memory operation request message carries the first invalid operation information.

[0044] The first data processing device in the first processing core carries first invalid operation information (i.e., EL information and ASID information) in the Distributed Virtual Memory Operation Request message (DVMOp) sent to the miscellaneous node, so that the miscellaneous node can inform other data processing devices in other processing cores of the CPU to which the first data processing device belongs of the first invalid operation information.

[0045] Step S112: If the invalid operation information includes second invalid operation information, after receiving the packet identifier response message returned by the miscellaneous node, send the target data and second invalid operation information to the miscellaneous node.

[0046] After receiving the DVMOp message, the miscellaneous node returns a packet identifier response message (DBIDResp) to the first data processing device. If the invalid operation information also includes second invalid operation information, the first data processing device sends the second invalid operation information along with the target data (NCBWrData) to the miscellaneous node so that the miscellaneous node can inform other data processing devices within other processing cores of the CPU to which the first data processing device belongs of the second invalid operation information. The target data is the data to be processed by the first data processing device.

[0047] It should be noted that miscellaneous nodes can broadcast invalid operation information (first invalid operation information, or first invalid operation information and second invalid operation information) in a distributed virtual memory operation snooping response message (SnpDVMOp) to the second data processing device in the second processing core.

[0048] In the case where the invalid operation information is the third invalid operation information, or both the third and fourth invalid operation information, the first data processing device in the first processing core carries the third invalid operation information (i.e., EL1 information or EL3 information) in the Distributed Virtual Memory Operation Request message (DVMOp) sent to the miscellaneous node. If the invalid operation information also includes the fourth invalid operation information, the first data processing device in the first processing core, after receiving the packet identifier response message (DBIDResp) returned by the miscellaneous node, sends the target data (NCBWrData) and the fourth invalid operation information (AV information) to the miscellaneous node.

[0049] In some embodiments, such as Figure 3 As shown, after sending the target data and the second invalid operation information to the miscellaneous node (i.e., step S112), the page table cache invalidation method may further include the following steps:

[0050] Step S113: Receive a completion response message sent by the miscellaneous node. The completion response message is sent by the miscellaneous node after the second data processing device completes the page table cache invalidation operation.

[0051] After receiving invalid operation information sent by the miscellaneous node, the second data processing device in the second processing core performs a TLBI operation based on the invalid operation information and returns a distributed virtual memory operation snooping response message (SnpResp_I) to the miscellaneous node. After receiving the distributed virtual memory operation snooping response message, the miscellaneous node returns a completion response message (Comp) to the first data processing device in the first processing core to inform the first data processing device that the second data processing device has completed the corresponding TLBI operation.

[0052] To clearly illustrate the solutions of the embodiments of this disclosure, the following is combined with... Figure 4 This section provides a detailed explanation of the page table cache invalidation process through a specific example. Figure 4 As shown, RN-F0 is the first data processing unit within processing core 0, and RN-F1 is the second data processing unit within processing core 1. The page table cache invalidation process includes the following steps:

[0053] Step 1: RN-F0 sends a DVMOp request message to MN on the bus. The DVMOp request message carries EL information and ASID information, which are carried in the address payload of the DVMOp request message.

[0054] Step 2: After receiving the DVMOp request message, MN returns a DBIDResp response message to RN-F0.

[0055] Step 3: After receiving the DBIDResp response message, RN-F0 sends the VA information in NCBWrData to MN.

[0056] Step 4: After receiving the NCBWrData carrying the VA information, MN broadcasts the SnpDVMOp message for the corresponding TLBI operation. The SnpDVMOp message is split into two parts: SnpDVMOp_P1 and SnpDVMOp_P2. Both SnpDVMOp_P1 and SnpDVMOp_P2 messages carry EL information, ASID information, and VA information.

[0057] Step 5: After receiving the SnpDVMOp_P1 and SnpDVMOp_P2 messages, RN-F1 returns a SnpResp_I response message to MN after completing the corresponding TLBI operation.

[0058] Step 6: After receiving the snpresp_I response message returned by RN-F1, MN returns a Comp response message to RN-F0, completing the entire transaction processing flow.

[0059] As can be seen from steps 1-6 above, it is only necessary to modify the information carried by the DVMOp message and the message carrying NCBWrData to realize the ASID-related TLBI operation under EL3. Its processing flow can directly reuse the current CHI protocol's processing flow for regular TLBI.

[0060] This disclosure extends the TLBI operation in distributed virtual memory operation transactions supported by the CHI protocol, making the CHI protocol compatible with more instruction set architectures. This disclosure can be applied to all SOC systems built on the CHI protocol that require virtual memory maintenance.

[0061] This disclosure also provides a method for invalidating page table caches, the method being applied to a second data processing device in a second processing core, such as... Figure 5 As shown, the page table cache invalidation method includes the following steps:

[0062] Step S21: Receive invalid operation information sent by the miscellaneous node; wherein the invalid operation message is sent to the miscellaneous node by the first data processing device in the first processing core, and the invalid operation information includes at least invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set.

[0063] In some embodiments, the non-ARM instruction set includes the RISC-V instruction set.

[0064] In some embodiments, receiving invalid operation information sent by the miscellaneous node includes: receiving a distributed virtual memory operation snooping message broadcast by the miscellaneous node, wherein the distributed virtual memory operation snooping message carries invalid operation information. That is, the miscellaneous node can broadcast the invalid operation information in a distributed virtual memory operation snooping response message (SnpDVMOp) and transmit it to the second data processing device in the second processing core.

[0065] Step S22: Perform an invalid operation on the page table cache based on the invalid operation information.

[0066] After receiving invalid operation information sent by the miscellaneous node, the second data processing device in the second processing core performs TLBI operation based on the invalid operation information.

[0067] The page table cache invalidation method provided in this disclosure involves a second data processing device in a second processing core performing a page table cache invalidation operation based on invalidation operation information sent by a miscellaneous node after receiving such information. The invalidation operation information includes at least invalidation operation information not defined in the ARM instruction set but defined in a non-ARM instruction set. This disclosure extends the CHI protocol and supplements the page table cache invalidation operations supported by the CHI protocol, enabling comprehensive TLBI operations. It decouples the CHI protocol from the instruction set used by the processing core, making the CHI protocol compatible with more instruction sets. This solves the compatibility problem of on-chip systems containing CPUs with non-ARM instruction sets, improving scalability and applicability.

[0068] In some embodiments, the invalid operation information not defined in the ARM instruction set but defined in a non-ARM instruction set includes first invalid operation information, or first invalid operation information and second invalid operation information. The first invalid operation information includes exception level (EL) information and ASID information, and the second invalid operation information includes VA information; wherein, the EL information is third exception level information or first exception level information. That is, invalid operation information not defined in the ARM instruction set but defined in a non-ARM instruction set includes EL3 information and ASID information, or EL3 information, ASID information, and VA information.

[0069] It should be noted that, as shown in Table 3, invalid operation information may also include third invalid operation information, or third invalid operation information and fourth invalid operation information. The third invalid operation information includes EL information, and the fourth invalid operation information includes VA information; wherein, the third invalid operation information is either first or third anomaly level information. That is, invalid operation information also includes EL information (EL1 or EL3 information), or invalid operation information also includes EL information (EL1 or EL3 information) and VA information. Furthermore, as shown in Table 3, the EL information in the first invalid operation information may also be EL1 information, meaning the first invalid operation information may also be EL1 information and ASID information.

[0070] In some embodiments, such as Figure 6 As shown, after performing a page table cache invalidation operation based on the invalidation operation information (i.e., step S22), the page table cache invalidation method may further include the following steps:

[0071] Step S23: Return a distributed virtual memory operation snooping response message to the miscellaneous node so that the miscellaneous node can send a completion response message to the first data processing device.

[0072] After performing a TLBI operation based on invalid operation information, the second data processing device in the second processing core returns a distributed virtual memory operation snooping response message (SnpResp_I) to the miscellaneous node. After receiving the distributed virtual memory operation snooping response message, the miscellaneous node returns a completion response message (Comp) to the first data processing device in the first processing core, informing the first data processing device that the second data processing device has completed the corresponding TLBI operation.

[0073] This disclosure also provides a data processing apparatus, such as... Figure 7 As shown, it includes:

[0074] At least one processor 701;

[0075] The memory 702 stores at least one program, which, when executed by the at least one processor, causes the at least one processor to implement the page table cache invalidation method provided in the foregoing embodiments;

[0076] At least one I / O interface 703 is connected between the processor and the memory and configured to enable information exchange between the processor and the memory.

[0077] The processor 701 is a device with data processing capabilities, including but not limited to a central processing unit (CPU); the memory 702 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), and flash memory (FLASH); the I / O interface (read / write interface) 703 is connected between the processor 701 and the memory 702, enabling information exchange between the processor 701 and the memory 702, including but not limited to a data bus (Bus).

[0078] In some embodiments, the processor 701, memory 702, and I / O interface 703 are interconnected via a bus, and thus connected to other components of the computing device.

[0079] This disclosure also provides a computer-readable medium having a computer program stored thereon, wherein the computer program, when executed, implements the page table cache invalidation method as provided in the foregoing embodiments.

[0080] It will be understood by those skilled in the art that all or some of the steps in the methods disclosed above, and the functional modules / units in the apparatus, can be implemented as software, firmware, hardware, and suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0081] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A method for invalidating page table cache, characterized in that, The method is applied to a first data processing device in a first processing core to extend the CHI protocol to support page table cache invalidation operations under non-ARM instruction sets. The method includes: Invalid operation information is sent to miscellaneous nodes so that the miscellaneous nodes can notify the second data processing device in the second processing core to perform a page table cache invalidation operation based on the invalid operation information; wherein, the invalid operation information includes at least invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set; the invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set includes first invalid operation information, which includes exception level (EL) information and address space identifier (ASID) information, wherein the EL information is third exception level information, and the first invalid operation information corresponds to a page table cache invalidation operation based on ASID that is not defined in the ARM instruction set but is performed in the RISC-V instruction set M-mode, wherein the page table cache invalidation operation has no corresponding instruction in the ARM instruction set.

2. The method as described in claim 1, characterized in that, The invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set also includes second invalid operation information, which includes virtual address (VA) information.

3. The method as described in claim 2, characterized in that, Sending invalid operation information to miscellaneous nodes includes: Send a distributed virtual memory operation request message to the miscellaneous node, the distributed virtual memory operation request message carrying the first invalid operation information; If the invalid operation information includes the second invalid operation information, after receiving the packet identifier response message returned by the miscellaneous node, the target data and the second invalid operation information are sent to the miscellaneous node.

4. The method as described in claim 3, characterized in that, After sending the target data and the second invalid operation information to the miscellaneous node, the method further includes: The system receives a completion response message sent by the miscellaneous node, which is sent by the miscellaneous node after the second data processing device has completed the page table cache invalidation operation.

5. A method for invalidating page table cache, characterized in that, The method is applied to a second data processing device in a second processing core to extend the CHI protocol to support page table cache invalidation operations under non-ARM instruction sets. The method includes: The system receives invalid operation information sent by a miscellaneous node. The invalid operation message is sent to the miscellaneous node by a first data processing device in the first processing core, and the invalid operation information includes at least invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set. The invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set includes first invalid operation information, which includes exception level (EL) information and address space identifier (ASID) information. The EL information is third exception level information. The first invalid operation information corresponds to a page table cache invalid operation based on ASID that is not defined in the ARM instruction set and is performed in the RISC-V instruction set M-mode. The page table cache invalid operation has no corresponding instruction in the ARM instruction set. Perform an invalid page table cache operation based on the invalid operation information.

6. The method as described in claim 5, characterized in that, The invalid operation information that is not defined in the ARM instruction set but is defined in a non-ARM instruction set also includes second invalid operation information, which includes virtual address (VA) information.

7. The method as described in claim 5, characterized in that, The invalid operation information sent by the receiving miscellaneous node includes: Receive a distributed virtual memory operation snooping message broadcast by a miscellaneous node, the distributed virtual memory operation snooping message carrying the invalid operation information.

8. The method as described in claim 5, characterized in that, After performing a page table cache invalidation operation based on the invalidation operation information, the method further includes: The miscellaneous node returns a distributed virtual memory operation snooping response message, which allows the miscellaneous node to send a completion response message to the first data processing device.

9. A data processing apparatus, comprising: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the page table cache invalidation method as described in any one of claims 1-4, or the page table cache invalidation method as described in any one of claims 5-8; One or more I / O interfaces are connected between the processor and the storage device and configured to enable information interaction between the processor and the storage device.

10. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed, it implements the page table cache invalidation method as described in any one of claims 1-4, or the page table cache invalidation method as described in any one of claims 5-8.

Citation Information

Patent Citations

  • Integrated circuit computing device and computing processing system

    CN111400986A

  • Apparatus and method in which control functions and synchronization events are performed

    GB202303353D0