Method and apparatus for performing energy-efficient network packet processing in a multi processor core system

Abstract

A method and apparatus for managing core affinity for network packet processing is provided. Low-power idle state of a plurality of processing units in a system including the plurality of processing units is monitored. Network packet processing is dynamically reassigned to processing units that are in a non-low power idle state to increase the low-power idle state residency for processing units that are in a low-power idle state resulting in reduced energy consumption.

1 . A method, comprising:
 determining at least one respective low-power state associated with at least one of multiple processor cores; an   based on the determined at least one lower power state associated with the at least one of multiple processor cores, causing a change to data that maps network packet flows to respective multiple processor cores based on respective hashes of network packet header data of respective network packets belonging to the network packet flows, wherein the change to the data remaps network packet flows from the at least one of multiple processor cores in a low-power state to at least one processor core not in a low-power state.   
     
     
         2 . The method of  claim 1 ,
 wherein the low-power state comprises one of a subset of power states available to a processor core.   
     
     
         3 . The method of  claim 2 , wherein the low-power state comprises C1, C3, or C6. 
     
     
         4 . The method of  claim 1 , wherein the hashes of network packet header data comprise at least one hash of data in a transport header and in an internet header of a respective network packet. 
     
     
         5 . The method of  claim 1 , wherein the data comprises data stored by a network interface controller. 
     
     
         6 . The method of  claim 5 , wherein the causing the change to data comprises sending change data to the network interface controller. 
     
     
         7 . The method of  claim 1 , wherein the data comprises an indirection table indexed by the hashes to, at least in part, determine a receiving processor core. 
     
     
         8 . The method of  claim 1 , wherein the data that maps network packet flows to respective multiple processor cores comprises data that maps the network packet flows to respective receive queues associated with the respective multiple processor cores. 
     
     
         9 . An article including a machine-accessible medium having associated information, wherein the information when accessed, results in a machine to:
 determine at least one respective low-power state associated with at least one of multiple processor cores; and   based on the at least one respective low-power state associated with the at least one of multiple processor cores, cause a change to data that maps network packet flows to respective multiple processor cores based on respective hashes of network packet header data of respective network packets belonging to the network packet flows, wherein the change to the data remaps network packet flows from the at least one processor cores in a low-power state to at least one processor core not in a low-power state.   
     
     
         10 . The article of  claim 9 ,
 wherein the low-power state comprises one of a subset of power states available to a processor core.   
     
     
         11 . The article of  claim 10 , wherein the low-power state comprises C1, C3, or C6. 
     
     
         12 . The article of  claim 9 , wherein the hashes of network packet header data comprise at least one hash of data in a transport header and in an internet header of a respective network packet. 
     
     
         13 . The article of  claim 9 , wherein the data comprises data stored by a network interface controller. 
     
     
         14 . The article of  claim 13 , wherein the information when accessed, results in a machine to send data to the network interface controller to change the data. 
     
     
         15 . The article of  claim 9 , wherein the data comprises an indirection table indexed by the hashes to, at least in part, determine a receiving processor core. 
     
     
         16 . The article of  claim 9 , wherein the data that maps network packet flows to respective multiple processor cores comprises data that maps the network packet flows to respective receive queues associated with the respective multiple processor cores. 
     
     
         17 . A system, comprising:
 multiple processor cores; and a   a network interface controller;   wherein the system comprises logic to:
 determine at least one respective low-power state associated with at least one of multiple processor cores; 
 based on the at least one respective low-power state associated with at least one of multiple processor cores, cause a change to data that maps network packet flows to respective multiple processor cores based on respective hashes of network packet header data of respective network packets belonging to the network packet flows, wherein the change to the data remaps network packet flows from the at least one processor cores in a low-power state to at least one processor core not in a low-power state. 
   
     
     
         18 . The system of  claim 17 ,
 wherein the low-power state comprises one of a subset of power states available to a processor core.   
     
     
         19 . The system of  claim 17 , wherein the hashes of network packet header data comprise at least one hash of data in a transport header and in an internet header of a respective network packets. 
     
     
         20 . The system of  claim 17 , wherein the logic to cause the change to data comprises logic to send change data to the network interface controller. 
     
     
         21 . The article of  claim 17 , wherein the data that maps network packet flows to respective multiple processor cores comprises data that maps the network packet flows to respective receive queues associated with the respective multiple processor cores.