Method and device for distributing partitions on a multicore processor
地域：
Courbevoie
摘要
A method and a device for distributing partitions of a sequence of partitions on the cores of a multicore processor are provided. The method makes it possible to identify parameters characterizing the hardware architecture of a multicore processor, and parameters characterizing an initial ordering of the partitions of a sequence; and then to profile and classify each partition of the sequence in order to assign the execution of each partition to a core of the multicore processor while maintaining the initial sequential ordering of the partitions.
說明書
This application is a National Stage of International patent application PCT/EP2017/068937, filed on Jul. 26, 2017, which claims priority to foreign French patent application No. FR 1601198, filed on Aug. 4, 2016, the disclosures of which are incorporated by reference in their entirety.
The invention relates to the field of onboard systems, and more particularly to a method and a device for distributing partitions on a multicore processor of an avionics computer.
權利要求
The invention claimed is:1. A computer implemented method for distributing partitions of a sequence of partitions on cores of a multicore processor, the method comprising steps of:identifying first parameters characterizing a hardware architecture of a multicore processor; identifying second parameters characterizing an initial ordering of a plurality of partitions of a periodic sequence, said second parameters comprising at least a number of the partitions, a time interval allocated to each of the partitions, an activation date for each of the partitions, and a total time of executing each of the partitions; generating a profile for each of the partitions on the basis of the first parameters and of the initial ordering, wherein each of the generations takes into account one or more criteria for estimating spatial and temporal locality of memory accesses of the respective partition, a number of the memory accesses, a volume of data accessed, and processor load; classifying each of the partitions according to the respective profile; assigning the execution of each of the partitions according to the respective classification to a core of the multicore processor such that the execution of each of the partitions is performed on the same core, while maintaining the initial ordering; and executing the sequence of partitions. 2. The method as claimed in claim 1 , wherein the identification of the first parameters comprises at least:defining the hardware architecture in terms of a number of cores of the multicore processor, a hierarchy of shared and/or private memories, memory controllers, and an interconnection bus. 3. The method as claimed in claim 1 , wherein the generations comprise determining the partitions that have a performance gain, to be coupled to a dedicated memory. 4. The method as claimed in claim 1 , wherein the generations comprise determining the partitions the execution of which gives rise to a hot spot to be reduced. 5. The method as claimed in claim 1 , wherein each of the classifications comprises:calculating a value for the respective partition on the basis of the one or more criteria, and classifying the respective partition according to the value. 6. A device for distributing partitions of a periodic sequence on cores of a multicore processor, the device comprising means for implementing the steps of the method as claimed in claim 1 . 7. A non-transitory, computer-readable medium comprising instructions for carrying out the steps of the method according to claim 1 , when said instructions are executed on a computer. 8. The method as claimed in claim 1 , further comprising:assigning an execution of each of a plurality of other partitions of another periodic sequence according to a respective classification to another core of the multicore processor such that the execution of each of the other partitions is on the same other core, while maintaining an initial ordering of the other partitions of the other periodic sequence. 9. The method as claimed in claim 8 , further comprising:disabling the assigned cores during the time interval in which the partition allocated to the cores is not executed. 10. The method as claimed in claim 8 , further comprising:synchronizing the executions of the partitions between the respective cores.
意見反饋