技术领域technical field
本发明涉及计算机应用领域，特别涉及一种基于MESI的缓存方法、装置和处理器。The invention relates to the field of computer applications, in particular to a MESI-based caching method, device and processor.
技术背景technical background
随着现代社会科技的发展，计算机已经广泛应用于各种领域，随之而来的是处理器技术的飞速发展。处理器作为计算机中的主要设备之一，是电脑中的核心配件。其功能主要是解释计算机指令以及处理计算机软件中的数据。随着人们对于计算速度和计算规模的需求不断提高，由于单处理器计算机系统的处理器运算性能受芯片速度极限和加工工艺极限的限制，不可能无限提高，这就使得多处理器的技术应运而生。With the development of science and technology in modern society, computers have been widely used in various fields, followed by the rapid development of processor technology. As one of the main devices in the computer, the processor is the core accessory in the computer. Its function is mainly to interpret computer instructions and process data in computer software. As people's demand for computing speed and computing scale continues to increase, since the processor computing performance of a single-processor computer system is limited by the chip speed limit and processing technology limit, it is impossible to increase infinitely, which makes the multi-processor technology come into play. And born.
多处理器中存在共享和私有数据的高速缓存。私有数据是被单个处理器使用的，而共享数据则是被多个处理器使用的，本质上是通过读写共享数据完成处理器之间的通信，这就使得多处理器的缓存一致性成为必须的技术。There are caches for shared and private data in a multiprocessor. Private data is used by a single processor, while shared data is used by multiple processors. In essence, the communication between processors is completed by reading and writing shared data, which makes multi-processor cache coherence a required technology.
MESI是广泛应用的一种缓存一致性协议。MESI是四种缓存段状态Invalid、Shared、Exclusive、Modified的首字母缩写，分别代表无效状态、共享状态、独占状态和修改状态，四种缓存段状态的转换可以如图1所示。多处理器系统中任何缓存段都处于这四种状态之一。MESI协议是一个合适的状态机，既能处理来自本地处理器的请求，也能把信息广播到总线上。MESI is a widely used cache coherence protocol. MESI is an acronym for the four cache segment states Invalid, Shared, Exclusive, and Modified, which respectively represent the invalid state, shared state, exclusive state, and modified state. The transitions of the four cache segment states can be shown in Figure 1. Any cache segment in a multiprocessor system is in one of these four states. The MESI protocol is a suitable state machine that handles requests from the local processor and broadcasts information onto the bus.
现有技术中，在一些对实时性较高的多处理器系统，处理器需要实时地迅速读取数值并进行分析，在这种场景下，保证最新的数据是非常重要的。但现有的MESI协议中却存在着数据过期的风险。例如，A处理器读取数据后B处理器可能迅速对相应的缓存进行写，这时A处理器读取的数据就成了过期值。操作已过期的数据可能会造成偏差，这在一些实时性高、对时效要求苛刻的系统中尤为严重。In the prior art, in some multi-processor systems with high real-time performance, the processor needs to quickly read and analyze values in real time. In this scenario, it is very important to ensure the latest data. However, there is a risk of data expiration in the existing MESI protocol. For example, after processor A reads data, processor B may quickly write to the corresponding cache, and at this time, the data read by processor A becomes an expired value. Operation of expired data may cause deviation, which is especially serious in some systems with high real-time performance and strict requirements on timeliness.
发明内容Contents of the invention
本发明实施例所要解决的技术问题是，提供一种基于MESI的缓存方法、装置和处理器，以解决本地读数据过期的问题，提高系统稳定性和用户体验。The technical problem to be solved by the embodiments of the present invention is to provide a MESI-based caching method, device and processor to solve the problem of local read data expiration and improve system stability and user experience.
为解决上述技术问题，本发明提供了一种基于MESI的缓存方法，该方法包括：In order to solve the above technical problems, the present invention provides a MESI-based caching method, the method comprising:
本地读操作后在缓存段添加过期风险标识；Add an expiration risk flag to the cache segment after the local read operation;
在所述缓存段进行下一操作后，对存在所述过期风险标识的改变的缓存段进行本地重读，删除所述过期风险标识。After the next operation is performed on the cache segment, local reread is performed on the cache segment with the change of the expiration risk identifier, and the expiration risk identifier is deleted.
可选的，在所述缓存段进行下一操作后，对存在所述过期风险标识的改变的缓存段进行本地重读，删除所述过期风险标识，包括：Optionally, after the cache segment performs the next operation, locally reread the cache segment with the change of the expiration risk identifier, and delete the expiration risk identifier, including:
判断下一操作是否为本地写或远端写操作；Determine whether the next operation is a local write or a remote write operation;
若是，在所述缓存段进行本地写或远端写操作后，对存在所述过期风险标识的改变的缓存段进行本地重读，删除所述过期风险标识；If so, after the local write or remote write operation is performed on the cache segment, locally reread the cache segment with the change of the expiration risk identifier, and delete the expiration risk identifier;
若否，在远端读操作前直接删除所述缓存段的所述过期风险标识。If not, directly delete the expiration risk identifier of the cache segment before the remote read operation.
可选的，所述过期风险标识具体为：R标识。Optionally, the expiration risk identifier is specifically: R identifier.
此外，本发明提供了一种基于MESI的缓存装置，该装置包括：In addition, the present invention provides a MESI-based caching device, which includes:
添加模块，用于本地读操作后在缓存段添加过期风险标识；Add a module for adding an expiration risk flag to the cache segment after a local read operation;
重读模块，用于在所述缓存段进行下一操作后，对改变的缓存段进行重读，删除所述过期风险标识。The rereading module is configured to reread the changed cache segment after the next operation is performed on the cache segment, and delete the expiration risk identifier.
可选的，所述重读模块，包括：Optionally, the rereading module includes:
判断单元，用于判断下一操作是否为本地写或远端写操作；若是，向重读单元发送重读信号；若否，向删除单元发送删除信号；The judging unit is used to judge whether the next operation is a local write or a remote write operation; if so, send a reread signal to the reread unit; if not, send a delete signal to the delete unit;
重读单元，用于接收所述重读信号，在本地写或远端写操作后进行重读，删除所述过期风险标识；A reread unit, configured to receive the reread signal, reread after the local write or remote write operation, and delete the expired risk identifier;
删除单元，用于接收所述删除信号，在远端读操作前直接删除所述标识。The deletion unit is configured to receive the deletion signal, and delete the identification directly before the remote read operation.
本发明还提供了一种处理器，包括：根据上述任一项所述的基于MESI的缓存装置。The present invention also provides a processor, including: the MESI-based cache device according to any one of the above items.
本发明所提供的一种基于MESI的缓存方法，包括本地读操作后在缓存段添加过期风险标识；在所述缓存段进行下一操作后，对存在所述过期风险标识的改变的缓存段进行本地重读，删除所述过期风险标识；A MESI-based caching method provided by the present invention includes adding an expiration risk identifier to a cache segment after a local read operation; Reread locally, delete the expired risk identification;
可见，本发明通过本地读操作后在缓存段添加过期风险标识，标记了存在过期风险的数据，再通过所述缓存段进行下一操作后，对存在所述过期风险标识的改变的缓存段进行本地重读，删除所述过期风险标识，可以根据本地重读的数据更新原来本地读的数据，消除数据过期的风险，从而提高系统的稳定性和用户体验。此外，本发明还提供了一种基于MESI的缓存装置和处理器，同样具有上述有益效果。It can be seen that the present invention adds an expiration risk identifier to the cache segment after a local read operation, and marks data with an expiration risk. Local rereading, deleting the expiration risk identifier, can update the original locally read data according to the local rereading data, eliminating the risk of data expiration, thereby improving system stability and user experience. In addition, the present invention also provides a MESI-based cache device and processor, which also have the above beneficial effects.
附图说明Description of drawings
为了更清楚地说明本申请实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those skilled in the art, other drawings can also be obtained based on these drawings without any creative effort.
图1为现有技术中MESI状态转变的示意图；FIG. 1 is a schematic diagram of MESI state transition in the prior art;
图2为本发明实施例提供的一种基于MESI的缓存方法的流程图；Fig. 2 is a flow chart of a MESI-based caching method provided by an embodiment of the present invention;
图3为本发明实施例提供的另一种基于MESI的缓存方法的关系图；FIG. 3 is a relational diagram of another MESI-based caching method provided by an embodiment of the present invention;
图4为本发明实施例提供的另一种基于MESI的缓存方法在本地读操作之后添加R标识的示意图；FIG. 4 is a schematic diagram of adding an R mark after a local read operation in another MESI-based caching method provided by an embodiment of the present invention;
图5为本发明实施例提供的另一种基于MESI的缓存方法在本地读操作之后下一操作的示意图；5 is a schematic diagram of another MESI-based caching method provided in an embodiment of the present invention for a next operation after a local read operation;
图6为本发明实施例提供的一种基于MESI的缓存装置的结构图。FIG. 6 is a structural diagram of a MESI-based caching device provided by an embodiment of the present invention.
具体实施方式detailed description
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。The following will clearly and completely describe the technical solutions in the embodiments of the application with reference to the drawings in the embodiments of the application. Apparently, the described embodiments are only some of the embodiments of the application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.
请参考图2，图2为本发明实施例提供的一种基于MESI的缓存方法的流程图。该方法可以包括：Please refer to FIG. 2 , which is a flow chart of a MESI-based caching method provided by an embodiment of the present invention. The method can include:
步骤101：本地读操作后在缓存段添加过期风险标识。Step 101: Add an expiration risk identifier to the cache segment after the local read operation.
其中，过期风险标识可以为缓存段在经过本地读操作后，在相应的缓存段额外添加的用于标记过期风险的标识，并不会对接下来的操作造成影响。Wherein, the expiration risk identifier may be an additional identifier for marking the expiration risk added to the corresponding cache segment after the cache segment undergoes a local read operation, and will not affect subsequent operations.
需要说明的是，过期风险标识可以为R标识，也可以为L标识等其他标识，本实施例对此不受任何限制。It should be noted that the expiration risk identifier may be an R identifier, or another identifier such as an L identifier, which is not limited in this embodiment.
步骤102：在所述缓存段进行下一操作后，对存在所述过期风险标识的改变的缓存段进行本地重读，删除所述过期风险标识。Step 102: After the next operation is performed on the cache segment, locally reread the cache segment with the change of the expiration risk identifier, and delete the expiration risk identifier.
其中，本地重读可以是本地读，只是在缓存段进行本地写或远端写操作后，对存在过期风险标识的改变的缓存段再次进行本地读操作，更新原来本地读操作缓存的数据。Among them, the local reread can be a local read, but after the local write or remote write operation is performed on the cache segment, the local read operation is performed again on the cache segment with the change of the expiration risk flag, and the data in the original local read operation cache is updated.
需要说明的是，下一操作可以是本地写、远端写或远端读，而主要可以针对下一操作为本地写、远端写时，可能会改变本地读时的数据，导致本地读的数据过期。而对于下一操作为远端读操作时，由于不会改变本地读的数据，可以不用进行本地重读操作，而是在本地读操作后直接删除缓存段的过期风险标识。本实施例对此不受任何限制。It should be noted that the next operation can be local write, remote write, or remote read, and it can mainly be aimed at that when the next operation is local write or remote write, the data in local read may be changed, resulting in local read Data expired. When the next operation is a remote read operation, since the locally read data will not be changed, it is not necessary to perform a local re-read operation, but directly delete the expiration risk flag of the cache segment after the local read operation. This embodiment is not limited in any way.
可以理解的是，对于下一操作为远端读，不进行本地重读操作的技术方案，可以加入判断环节，用于判断下一操作的类型。It can be understood that, for the technical solution in which the next operation is a remote read and no local re-read operation is performed, a judging link may be added to judge the type of the next operation.
本实施例中，提供的基于MESI的缓存方法通过本地读操作后在缓存段添加过期风险标识，标记了存在过期风险的数据，再通过所述缓存段进行下一操作后，对存在所述过期风险标识的改变的缓存段进行本地重读，删除所述过期风险标识，可以根据本地重读的数据更新原来本地读的数据，消除数据过期的风险，从而提高系统的稳定性和用户体验。In this embodiment, the provided MESI-based caching method adds an expiration risk identifier to the cache segment after a local read operation, and marks data with an expiration risk. After the next operation is performed through the cache segment, the expired The changed cache segment of the risk identifier is reread locally, and the expired risk identifier is deleted, and the original locally read data can be updated according to the locally reread data, eliminating the risk of data expiration, thereby improving system stability and user experience.
请参考图3、图4、图5，图3为本发明实施例提供的另一种基于MESI的缓存方法的关系图，图4为本发明实施例提供的另一种基于MESI的缓存方法在本地读操作之后添加R标识的示意图，图5为本发明实施例提供的另一种基于MESI的缓存方法在本地读操作之后下一操作的示意图。该方法可以包括：Please refer to FIG. 3, FIG. 4, and FIG. 5. FIG. 3 is a relationship diagram of another MESI-based caching method provided by an embodiment of the present invention, and FIG. 4 is another MESI-based caching method provided by an embodiment of the present invention. A schematic diagram of adding an R flag after a local read operation, and FIG. 5 is a schematic diagram of another MESI-based caching method provided by an embodiment of the present invention for a next operation after a local read operation. The method can include:
步骤201：本地读操作后在缓存段添加R标识。Step 201: After the local read operation, add an R mark to the cache segment.
其中，R标识可以为缓存段在进行本地读操作时直接添加的，也可以是在本地读操作后再添加的，本实施对此不受任何限制。Wherein, the R identifier may be added directly when the cache segment is performing a local read operation, or may be added after the local read operation, which is not subject to any restrictions in this implementation.
需要说明的是，添加R标识的步骤可以为如图4所示，在M、S、E状态进行本地读操作后在对应的缓存段添加R标识，如E状态经本地读后成为E+R状态，也就是E状态对应的缓存段中存在R标识。It should be noted that the step of adding the R mark can be as shown in Figure 4, after performing local read operations in the M, S, and E states, add the R mark in the corresponding cache segment, such as the E state becomes E+R after local reading state, that is, the R identifier exists in the cache segment corresponding to the E state.
具体实施中，可以为如表1展示的基于MESI的缓存方法的操作实例中的指令周期2和6所示：In the specific implementation, it can be shown in instruction cycles 2 and 6 in the operation example of the MESI-based caching method shown in Table 1:
表1Table 1
在本地读操作后，直接添加R标识。而图中的状态M+R表示的可以是在M状态的缓存段存在R标识。After the local read operation, add the R flag directly. The state M+R in the figure indicates that there may be an R flag in the cache segment of the M state.
步骤202：判断下一操作是否为本地写或远端写操作；若是，进入步骤203；若否，进入步骤204。Step 202: Determine whether the next operation is a local write operation or a remote write operation; if yes, go to step 203; if not, go to step 204.
可以理解的是，判断下一操作是否为本地写或远端写操作是因为本实施例的方法主要是针对本地写或远端写后数据可能改变，使得本地读的数据过期，所以对本地写或远端写操作进行判断。判断下一操作是否为远端读，或不进行判断而是识别下一操作的类型也可以达到本实施例的目的，本实施例对此不受任何限制。It can be understood that the reason for judging whether the next operation is a local write or a remote write operation is because the method of this embodiment is mainly aimed at data that may change after local write or remote write, making the locally read data expire, so the local write Or remote write operation to judge. Judging whether the next operation is a remote read, or not judging but identifying the type of the next operation can also achieve the purpose of this embodiment, which is not limited in this embodiment.
步骤203：在所述缓存段进行本地写或远端写操作后，对存在所述过期风险标识的改变的缓存段进行本地重读，删除所述R标识。Step 203: After the local write or remote write operation is performed on the cache segment, locally re-read the cache segment with the change of the expiration risk identifier, and delete the R identifier.
可以理解的是，在进行本地写或远端写操作后，缓存段中存在R标识的状态可以为如图5所示，如S状态在本地读后为S+R状态，其后的动作是本地写，则发生S->M的状态转变，由于携带着R，实际上是S+R->M+R。风险出现的情景是M、S、E状态因为本地写而成为M状态，或远端写而成为I状态。其中I状态的缓存在本地写后成为M状态，但因为在此之前I状态缓存不能进行读操作，因此不会有风险，不需要产生R状态。It can be understood that after the local write or remote write operation, the state of the R flag in the cache segment can be as shown in Figure 5, such as the S state after the local read is the S+R state, and the subsequent actions are When writing locally, the state transition of S->M occurs. Since it carries R, it is actually S+R->M+R. The scenario where the risk arises is that the M, S, and E states become the M state due to local writing, or become the I state due to remote writing. The cache in the I state becomes the M state after being written locally, but because the I state cache cannot be read before then, there is no risk, and there is no need to generate the R state.
具体实施中，本步骤可以为如表1中指令周期3和4所示，对缓存段存在R标识的10101进行远端写操作，新值为00111，这时本地缓存段成为I状态。由于R标识的存在，本地缓存请求读入新值，并成为S状态。此时仍携带R标识。为解决本地读数据过期问题，此时将进行重读，也就是本地重读。读取到新值00111后，删除缓存段的R标识，继续新的操作。In specific implementation, this step can be as shown in instruction cycles 3 and 4 in Table 1, perform a remote write operation on the 10101 with the R mark in the cache segment, and the new value is 00111, and the local cache segment becomes the I state at this time. Due to the presence of the R flag, the local cache requests to read new values and become the S state. At this time, it still carries the R logo. In order to solve the problem of local read data expiration, rereading will be performed at this time, that is, local rereading. After reading the new value 00111, delete the R mark of the cache segment and continue the new operation.
步骤204：在远端读操作前直接删除所述缓存段的所述R标识。Step 204: Delete the R identifier of the cache segment directly before the remote read operation.
其中，本步骤可以是远端读操作前直接删除缓存段的R标识，也可以是在远端读操作后删除缓存段的R标识，本实施例对此步骤任何限制。Wherein, this step may be directly deleting the R identifier of the cache segment before the remote read operation, or may be deleting the R identifier of the cache segment after the remote read operation, and there is no restriction on this step in this embodiment.
具体实施中，本步骤可以为如表1中指令周期7所示，在指令周期6进行本地读操作后，此时缓存段存在R标识。指令周期7进行的操作为远端读，不会产生数据过期风险，因此可以直接删除缓存段的R状态标志。In specific implementation, this step can be as shown in instruction cycle 7 in Table 1, after the local read operation is performed in instruction cycle 6, there is an R flag in the cache segment at this time. The operation performed in instruction cycle 7 is a remote read, and there is no risk of data expiration, so the R status flag of the cache segment can be directly deleted.
本实施中，本发明实施例通过加入判断环节，对于没有过期风险的远端读操作，在远端读操作前直接删除所述缓存段的所述R标识，减少了本地重读的次数，可以进一步提高系统的效率，提高用户体验。In this implementation, the embodiment of the present invention adds the judging link. For the remote read operation without the risk of expiration, the R identifier of the cache segment is directly deleted before the remote read operation, which reduces the number of local re-reads, and can further Improve system efficiency and improve user experience.
请参考图6，为本发明实施例提供的一种基于MESI的缓存装置的结构图。该装置可以包括：Please refer to FIG. 6 , which is a structural diagram of a MESI-based caching device provided by an embodiment of the present invention. The device can include:
添加模块100，用于本地读操作后在缓存段添加过期风险标识。The adding module 100 is used for adding an expiration risk identifier to the cache segment after the local read operation.
重读模块200，用于在所述缓存段进行下一操作后，对改变的缓存段进行重读，删除所述过期风险标识。The reread module 200 is configured to reread the changed cache segment after the next operation is performed on the cache segment, and delete the expiration risk identifier.
可选的，所述重读模块200，可以包括：Optionally, the rereading module 200 may include:
判断单元，用于判断下一操作是否为本地写或远端写操作；若是，向重读单元发送重读信号；若否，向删除单元发送删除信号。The judging unit is used to judge whether the next operation is a local write or a remote write operation; if yes, send a reread signal to the reread unit; if not, send a delete signal to the delete unit.
重读单元，用于接收所述重读信号，在本地写或远端写操作后进行重读，删除所述过期风险标识。The reread unit is configured to receive the reread signal, perform reread after the local write or remote write operation, and delete the expired risk identifier.
删除单元，用于接收所述删除信号，在远端读操作前直接删除所述标识。The deletion unit is configured to receive the deletion signal, and delete the identification directly before the remote read operation.
本实施例中，本发明实施例通过添加模块100在本地读操作后的缓存段添加过期风险标识，标记了存在过期风险的数据，再通过重读模块200在所述缓存段进行下一操作后，对存在所述过期风险标识的改变的缓存段进行本地重读，删除所述过期风险标识，可以根据本地重读的数据更新原来本地读的数据，消除数据过期的风险，从而提高系统的稳定性和用户体验。In this embodiment, the embodiment of the present invention adds an expiration risk identifier to the cache segment after the local read operation through the adding module 100, marking data with an expiration risk, and then performs the next operation on the cache segment through the reread module 200, Perform local re-reading on the cache segment with the change of the expiration risk identification, delete the expiration risk identification, update the original locally read data according to the local re-read data, and eliminate the risk of data expiration, thereby improving system stability and user experience.
本发明还提供了一种处理器，包括：如上一实施例所述的基于MESI的缓存装置。The present invention also provides a processor, including: the MESI-based cache device as described in the previous embodiment.
需要说明的是，本说明书中的实施例均采用递进的方式描述，每个实施例重点说明的都是与其他实施例的不同之处，各个实施例之间相同相似的部分互相参见即可。对于装置类和处理器实施例而言，由于其与方法实施例基本相似，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。It should be noted that the embodiments in this specification are described in a progressive manner, each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments can be referred to each other . For the device class and processor embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and for relevant parts, please refer to part of the description of the method embodiments.
以上对本申请所提供的基于MESI的缓存方法、装置和处理器进行了详细介绍，本文中应用了具体实施例对本申请的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本申请的方法及其核心思想；同时，对于本领域的一般技术人员，依据本申请的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本申请的限制。The MESI-based caching method, device, and processor provided by this application have been introduced in detail above. The principles and implementation methods of this application have been explained by using specific embodiments in this paper. The descriptions of the above embodiments are only used to help understand this application. The method of application and its core idea; at the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation and scope of application. In summary, the content of this specification should not be understood For the limitation of this application.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610979384.5ACN106502920B (en) | 2016-11-08 | 2016-11-08 | A kind of caching method based on MESI, device and processor |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610979384.5ACN106502920B (en) | 2016-11-08 | 2016-11-08 | A kind of caching method based on MESI, device and processor |
| Publication Number | Publication Date |
|---|---|
| CN106502920Atrue CN106502920A (en) | 2017-03-15 |
| CN106502920B CN106502920B (en) | 2019-09-24 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610979384.5AActiveCN106502920B (en) | 2016-11-08 | 2016-11-08 | A kind of caching method based on MESI, device and processor |
| Country | Link |
|---|---|
| CN (1) | CN106502920B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5522057A (en)* | 1993-10-25 | 1996-05-28 | Intel Corporation | Hybrid write back/write through cache having a streamlined four state cache coherency protocol for uniprocessor computer systems |
| CN1196530A (en)* | 1997-04-14 | 1998-10-21 | 国际商业机器公司 | Cache-coherency protocol with upstream undefined state |
| CN1849594A (en)* | 2003-09-12 | 2006-10-18 | 英特尔公司 | Method and apparatus for joint cache coherency states in multi-interface caches |
| CN102144218A (en)* | 2008-07-28 | 2011-08-03 | 超威半导体公司 | Virtualizable advanced synchronization facility |
| CN102270180A (en)* | 2011-08-09 | 2011-12-07 | 清华大学 | Multicore processor cache and management method thereof |
| CN102741826A (en)* | 2009-12-15 | 2012-10-17 | 英特尔公司 | Performing mode switching in an unbounded transactional memory (UTM) system |
| US20150143049A1 (en)* | 2013-11-20 | 2015-05-21 | Electronics And Telecommunications Research Institute | Cache control apparatus and method |
| CN105512053A (en)* | 2015-12-07 | 2016-04-20 | 中南大学 | Mirror caching method for mobile transparent computing system server terminal multi-user access |
| WO2016106057A1 (en)* | 2014-12-26 | 2016-06-30 | Intel Corporation | Implementation of data coherence among devices |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5522057A (en)* | 1993-10-25 | 1996-05-28 | Intel Corporation | Hybrid write back/write through cache having a streamlined four state cache coherency protocol for uniprocessor computer systems |
| CN1196530A (en)* | 1997-04-14 | 1998-10-21 | 国际商业机器公司 | Cache-coherency protocol with upstream undefined state |
| CN1849594A (en)* | 2003-09-12 | 2006-10-18 | 英特尔公司 | Method and apparatus for joint cache coherency states in multi-interface caches |
| CN102144218A (en)* | 2008-07-28 | 2011-08-03 | 超威半导体公司 | Virtualizable advanced synchronization facility |
| CN102741826A (en)* | 2009-12-15 | 2012-10-17 | 英特尔公司 | Performing mode switching in an unbounded transactional memory (UTM) system |
| CN102270180A (en)* | 2011-08-09 | 2011-12-07 | 清华大学 | Multicore processor cache and management method thereof |
| US20150143049A1 (en)* | 2013-11-20 | 2015-05-21 | Electronics And Telecommunications Research Institute | Cache control apparatus and method |
| WO2016106057A1 (en)* | 2014-12-26 | 2016-06-30 | Intel Corporation | Implementation of data coherence among devices |
| CN105512053A (en)* | 2015-12-07 | 2016-04-20 | 中南大学 | Mirror caching method for mobile transparent computing system server terminal multi-user access |
| Publication number | Publication date |
|---|---|
| CN106502920B (en) | 2019-09-24 |
| Publication | Publication Date | Title |
|---|---|---|
| US8321635B2 (en) | Synchronizing commands for preventing data corruption | |
| CN104252501B (en) | A kind of computing device and method executing database manipulation order | |
| CN103218176B (en) | Data processing method and device | |
| CN111737564B (en) | Information query method, device, equipment and medium | |
| CN103049392B (en) | The implementation method of CACHE DIRECTORY and device | |
| KR102287677B1 (en) | Data accessing method, apparatus, device, and storage medium | |
| CN110765206A (en) | Data synchronization system, method, device, equipment and storage medium | |
| WO2019010703A1 (en) | Method for reading and partially writing data, and relevant apparatus | |
| CN115104092B (en) | Data synchronization method and related device | |
| US20150227562A1 (en) | Database device | |
| CN102971698A (en) | Snapshot data-processing method and system, storage system and snapshot agency | |
| CN108108392B (en) | Commodity data management method, device, computer equipment and storage medium | |
| CN110764796A (en) | Method and device for updating cache | |
| CN108805590A (en) | A kind of agricultural product traceability system based on block chain | |
| CN107391402A (en) | A kind of data operating method, device and a kind of data operation card | |
| US20120059996A1 (en) | Avoiding Cross-Interrogates in a Streaming Data Optimized L1 Cache | |
| CN113722389B (en) | Data management method, device, electronic equipment and computer readable storage medium | |
| CN107003932B (en) | Cache directory processing method and directory controller of multi-core processor system | |
| CN117891754B (en) | Processor Cache structure caching method, computing subsystem and intelligent computing platform | |
| CN106502920B (en) | A kind of caching method based on MESI, device and processor | |
| CN111858665B (en) | Method, system, terminal and storage medium for improving soft copy reading performance | |
| CA3089270C (en) | Systems and methods for storing object state on hash chains | |
| US10417126B2 (en) | Non-coherent read in a strongly consistent cache system for frequently read but rarely updated data | |
| CN116756138A (en) | Data access method and device and electronic equipment | |
| CN115408010A (en) | Data processing method, device, equipment, readable storage medium and program product |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |