Systems, methods and devices for determining work placement on processor cores
Apparatuses, methods and storage medium for computing including determination of work placement on processor cores are disclosed herein. In embodiments, an apparatus may include one or more processors, devices, and/or circuitry to identify a favored core of the processor cores. The one or more processors, devices, and/or circuitry may be configured to determine whether to migrate a thread to or from the favored core. In some embodiments, the determination may be by a process executed by a driver and/or by an algorithm executed by a power control unit of the processor.
1. An apparatus, comprising:
a processor including:
a plurality of cores; and
a favored core module operable by the processor to:
generate a ranking of the plurality of cores based on differing physical characteristics the plurality of cores;
identify one or more cores of the plurality of cores as favored with respect to one or more other cores of the plurality of cores based on the ranking, and include the one or more cores identified as favored in a favored core list;
for each thread of a plurality of threads:
determine whether a demand of the thread is greater than a threshold; and
in response to a determination that the demand of the thread is greater than the threshold, select the thread for affinitization;
determine whether a first count of selected thread(s) is less than or equal to a second count of the core(s) of the favored core list; and
in response to a determination that the first count is less than or equal to the second count, affinitize all selected thread(s) to a respective one of the core(s) of the favored core list, including:
perform a first affinitization of an initial thread of the selected thread(s) to a core of an initial entry in the favored core list; and
perform one or more second affinitizations of any one or more next threads, respectively, of the selected thread(s) to one or more cores of the one or more next entries in the favored core list, respectively; and
in response to a determination that the first count is greater than the second count, unaffinitize at least one thread-core affinitization.
2. The apparatus of claim 1 , wherein the one or more cores identified as favored have one or more first operating voltages, respectively, that are lower than one or more second operating voltages, respectively, of the one or more other cores.
3. The apparatus of claim 2 , wherein the at least one core of the one or more cores identified as favored is associated with a same operating frequency as at least one core of the one or more other cores.
4. The apparatus of claim 2 , wherein the at least one core of the one or more cores identified as favored is associated with a different operating frequency than at least one core of the one or more other cores.
5. The apparatus of claim 1 , wherein at least one of the one or more cores identified as favored corresponds to a user-selection input via a user interface.
6. The apparatus of claim 1 , wherein at least one of the one or more cores identified as favored corresponds to data of a fuse table.
7. The apparatus of claim 1 , wherein unaffinitize at least one thread-core affinitization comprises unaffinitize all thread-core affinitizations corresponding to at least one core of the core(s) of the favored core list.
8. The apparatus of claim 7 , wherein the favored core module is further to:
prior to affinitize all selected thread(s), sort the selected thread(s) based on at least one of an inclusion list or user request received by a user interface.
9. The apparatus of claim 1 , wherein the favored core module is further to, for each thread of the plurality of threads, in response to a determination that the demand of the thread is greater than the threshold:
add the thread to a tracker list and sort the threads of the tracker list based on an inclusion list or a user request received via a user interface; and
select the sorted thread for affinitization.
10. A non-transitory computer-readable medium having instructions stored thereon that, in response to execution by a processing device, cause the processing device to perform operations, to:
generate a ranking of cores of a processor based on differing physical characteristics of the cores;
identify one or more of the cores of the processor as favored with respect to one or more other cores of the cores of the processor based on the ranking, and include the one or more cores identified as favored in a favored core list;
for each thread of a plurality of threads:
determine whether a continuously or periodically monitored demand of the thread is greater than a threshold; and
in response to a determination that the continuously or periodically monitored demand of the thread is greater than the threshold, select the thread for affinitization;
determine whether a first count of selected thread(s) is less than or equal to a second count of the core(s) of the favored core list; and
in response to a determination that the first count is less than or equal to the second count, affinitize all selected thread(s) to a respective one of the core(s) of the favored core list:
perform a first affinitization of an initial thread of the selected thread(s) to a core of an initial entry in the favored core list; and
perform one or more second affinitizations of any one or more next threads, respectively, of the selected thread(s) to one or more cores of the one or more next entries in the favored core list, respectively; and
in response to a determination that the first count is greater than the second count, unaffinitize at least one thread-core affinitization.
11. The non-transitory computer-readable medium of claim 10 , wherein unaffinitize at least one thread-core affinitization comprises unaffinitize all thread-core affinitizations corresponding to at least one core of the core(s) of the favored core list.
12. The non-transitory computer-readable medium of claim 11 , wherein operations are further operable to:
prior to affinitize all selected thread(s), sort the selected thread(s) based on at least one of an inclusion list or user request received by a user interface.
13. The non-transitory computer-readable medium of claim 10 , wherein the operations include, for each thread of the plurality of threads:
add the thread to a tracker list and sort the threads of the tracker list based on an inclusion list or a user request received via a user interface; and
select the sorted thread for affinitization.
14. The non-transitory computer-readable medium of claim 10 , wherein the one or more cores identified as favored have one or more first operating voltages, respectively, that are lower than one or more second operating voltages, respectively, of the one or more other cores.
15. The non-transitory computer-readable medium of claim 14 , wherein the at least one core of the one or more cores identified as favored is associated with a same operating frequency as at least one core of the one or more other cores.
16. The non-transitory computer-readable medium of claim 14 , wherein the at least one core of the one or more cores identified as favored is associated with a different operating frequency than at least one core of the one or more other cores.
17. The non-transitory computer-readable medium of claim 10 , wherein at least one of the one or more cores identified as favored corresponds to a user-selection input via a user interface.
18. The non-transitory computer-readable medium of claim 10 , wherein at least one of the one or more cores identified as favored corresponds to data of a fuse table.
19. The non-transitory computer-readable medium of claim 10 , wherein the continuously or periodically monitored demand of the thread is based on current core utilization by the thread.
20. The apparatus of claim 1 , wherein the demand of the thread is based on current core utilization by the thread.