37results about How to "Solve the deadlock problem" patented technology

The invention provides a control message sending method and device. The method comprises the steps that first network side edge equipment PE receives a first control message from a second PE, wherein the first control message is used for setting the state of the physical port of a first access circuit AC connected with the first PE as the preset state; and the first PE sends a second control message to the second PE according to the preset state and the actual state of the physical port of the first AC, wherein the second control message is used for controlling the state of the physical port of a second AC connected with the second PE. With application of the control message sending method and device, the problem of deaklock of end-to-end fault transmission in the relevant technologies can be solved so that the effect of avoiding the problem of deaklock of end-to-end fault transmission can be achieved.

The invention discloses a method for achieving hot pull and hot plug and electronic devices. The method for achieving hot pull and hot plug is applied to a first device which is connected with a second device through an inter-integrated circuit (I2C) bus. The first device is provided with M I2C components which are connected on the I2C bus based on the I2C protocol. The second device is provided with N I2C components which are connected on the I2C bus on the basis of the I2C protocol. The M and the N are integral numbers larger than zero. The method includes the steps of obtaining error information which carries out I2C communication with the second device, carrying out ergodic operations on the M I2C components and the N I2C components on the basis of the error information, obtaining a great amount of feedback information on the basis of the ergodic operations, and determining that the I2C bus between the first device and the second device is switched off when the condition that the great amount of feedback information all comes from the M I2C components is determined.

The invention relates to a monitoring answer processing method on the basis of double correlation chains. A missing buffer MB is set in a Cache and memory commands in a Cache pipeline, which cannot be completed in one step, are stored into the buffer MB; the memory commands comprise local memory commands and remote monitoring commands; the local memory commands are memory commands sent to the Cache by a core; the remote monitoring commands are monitoring commands sent by a directory controller; a local correlation chain is established for the local memory commands and the memory commands with the same address enter the local correlation chain and wait for the execution of the previous memory commands with the same address; and a monitoring chain is established for monitoring access entering the buffer MB and the monitoring commands with the same address enter the monitoring correlation chain and wait for the execution of the previous memory commands with the same address. The invention can solve the problem of the condition of deadlock or livelock so as to avoid command repetition, and can efficiently solve the problem of a monitoring response processing sequence.

The present invention provides a method and device for sending a control message, wherein the method includes: a first network-side edge device PE receiving a first control message from a second PE, wherein the first control message is used to communicate with the first The state of the physical port of the first access circuit AC connected to the PE is set to a predetermined state; the first PE sends a second control message to the second PE according to the predetermined state and the actual state of the physical port of the first AC, wherein the first The second control message is used to control the state of the physical interface of the second AC connected to the second PE. The present invention solves the deadlock problem of end-to-end fault transmission existing in the related art, and further achieves the effect of avoiding the deadlock problem of end-to-end fault transmission.

This application discloses a multi-robot path planning method and device, which belong to the technical field of robot path planning. The A* algorithm is used to plan an optimal path for a single robot in the intersection graph environment; according to the optimal path, plan all passages in the intersection graph environment as one-way streets; for the one-way Road, through the heuristic A* algorithm, plan the quadratic optimal path for a single robot in the intersection graph environment; for the quadratic optimal path, rebuild the grid environment to obtain the grid The optimal path in grid environment. Solve the shortcomings of traditional heuristic algorithms that cause deadlock due to robot conflicts in the channel.

The invention discloses a data processing method and device, and belongs to the technical field of computers. The method comprises the following steps of receiving data to be processed; according to the data to be processed, creating a distribution thread in a single thread pool, wherein at most one distribution thread under an operation state is in the single thread pool; through the distribution thread under the operation state, creating m1 pieces of processing threads in a maximum operation thread pool; and in the maximum operation thread pool, through the processing threads under the operation state, processing the data, wherein the at most m2 pieces of processing threads under the operation state are in the maximum operation thread pool. By use of the method, through a condition that at most one distribution thread under the operation state is in the single thread pool, a situation that a plurality of processing threads need to access the same resource is avoided, and the problem of thread deadlock is solved. Calculation resources are reasonably utilized, and a multi-processing-thread mechanism is stably used for quickly processing the data so as to achieve an effect on improving the calculation performance of a system.

The present invention relates to the field of robot technology, and provides a path scheduling method, device, background server and storage medium, the method includes: receiving and generating a first schedule from the current position of the first robot to the first robot according to the obstacle handling request Set the first navigation path of the mission point; obtain the current navigation path of the second robot; when there is a path conflict between the current navigation path and the first navigation path, generate a second preset mission point from the current position of the second robot to the second robot the second navigation path; when there is no path conflict between the second navigation path and the first navigation path, send the first navigation path and the second navigation path to the first robot and the second robot respectively. The invention solves the "deadlock" problem of group robots in the same map by reasonably dispatching the moving paths of robots encountering obstacles and robots not encountering obstacles.

This invention discloses a message transfer method including: determining the necessary transferred sequence-keeping information in distributed network devices, determining the synchronous controlled sequence-keeping information necessary for the applied layer to control and transfer synchronously in the information, at the applied layer, said sequence-keeping information is transmitted to the transmission layer of said network device orderly to be further sent out in terms of the asynchronous sending mode on the transmission layer, among which, the sending of said information is synchronously controlled at the applied layer. Besides, this invention also discloses a related information transfer device.