Power state synchronization in a multi-core processor

• US 8,782,451 B2
• Filed: 11/17/2011
• Issued: 07/15/2014
• Est. Priority Date: 12/22/2010
• Status: Active Grant

• Current Assignee: VIA Technologies Incorporated (VIA Technologies)
• Original Assignee: VIA Technologies Incorporated (VIA Technologies)
• Inventors: Henry, G. Glenn, Gaskins, Darius D.
• Primary Examiner(s): Patel, Nimesh G
• Application Number:
  US13/299,059
• Publication Number:
  US 20120166845A1
• Time in Patent Office: 971 Days
• Field of Search: 713/323, 713/320
• US Class Current: 713/320
• CPC Class Codes: G06F 1/26 Power supply means, e.g. regulation thereof for memories G11C
• CPC Class Codes: G06F 1/3206 Monitoring of events, devices or parameters that trigger a change in power modality
• CPC Class Codes: G06F 1/3237 by disabling clock generation or distribution
• CPC Class Codes: G06F 1/324 by lowering clock frequency
• CPC Class Codes: G06F 1/3287 by switching off individual functional units in the computer system
• CPC Class Codes: G06F 1/3296 by lowering the supply or operating voltage
• CPC Class Codes: G06F 9/5094 where the allocation takes into account power or heat criteria power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

A multi-core processor includes microcode distributed in each core enabling each core to participate in a de-centralized inter-core state discovery process. In a related microcode-implemented method, states of a multi-core processor are discovered by at least two cores participating in a de-centralized inter-core state discovery process. The inter-core state discovery process is carried out through a combination of microcode executing on each participating core and signals exchanged between the cores through sideband non-system-bus communication wires. The discovery process is unmediated by any centralized non-core logic. Applicable discoverable states include target and composite power states, whether and how many cores are enabled, the availability and distribution of various resources, and hierarchical structures and coordination systems for the cores. The inter-core state discovery process may be carried out in accordance with various hierarchical coordination systems involving chained inter-core communications.

Claims

1. A multi-core processor comprising:

a plurality of physical processing cores; and

inter-core state discovery microcode in each core enabling the core to participate in a de-centralized inter-core state discovery process through signals, unmediated by any centralized non-core logic, sent to or received from other cores;

wherein;

the inter-core state discover microcode includes synchronization logic, provided to each core, synchronized instances of which are operable to be invoked on multiple cores for purposes of an inter-core state discovery process; and

wherein each native instance is operable both to invoke new instances of the synchronization logic on other cores, and to respond to any prior instance of the synchronization logic on another core that invoked the native instance;

each core has a target operating state;

the processor includes a domain comprising at least two of the microprocessor'"'"'s cores;

the processor provides a resource to the domain, which resource is shared by the cores of the domain;

the synchronization logic is configured to discover whether the domain is prepared to implement a restricted power-conserving operating state for the resource which would limit the power, speed, or efficiency with which the cores sharing the resource are able to operate; and

wherein the domain is prepared to implement the restricted operating state if any only if each of the enabled cores in the domain sharing the resource has a target operating state at least as restrictive as the restricted operating state.

2. The multi-core processor of claim 1, wherein:

the inter-core state discovery microcode exchanges signals with other cores through sideband communication wires that are independent of a system bus connecting the multi-core processor to a chipset; and

the inter-core state discovery microcode determines an applicable state value that is a function, at least, of a state of another core, without the assistance of any centralized non-core logic.

3. The multi-core processor of claim 1, wherein:

the shared resource is a system bus connected to a chipset;

the domain includes all of the enabled cores of the multi-core processor; and

the restricted operating state is a C-state that disables a bus clock of the system bus.

4. The multi-core processor of claim 1, wherein:

the shared resource is a phased-lock-loop (PLL) on a multi-core die of the microprocessor;

the domain includes all of the enabled cores whose clock signals are supplied by the PLL; and

the restricted operating state is a less-than-maximum performance frequency ratio that can be used by the cores sharing the PLL.

5. The multi-core processor of claim 1, wherein:

the shared resource is a voltage source;

the domain includes all of, but is limited to, the enabled cores of the microprocessor sharing that voltage source; and

the restricted operating state is a less-than-maximum performance voltage level that can be used by the cores sharing the voltage source.

6. The multi-core processor of claim 1, wherein:

each instance of synchronization logic is configured, unless terminated earlier by a termination condition, to recursively invoke instances of synchronization logic on other cores until synchronized instances of the synchronization logic have been invoked on all of the cores of an applicable domain of the processor; and

wherein the synchronization logic is configured with a termination condition to halt invocations of instances of synchronization logic on other not-yet-synchronized cores if it discovers a core having a target operating state that is less restrictive than the restricted power-conserving operating state;

wherein the synchronization logic is configured to coordinate with a minimally sufficient number of other cores to discover whether the applicable domain is prepared to implement a restricted power-conserving operating state.

7. The multi-core processor of claim 1, wherein each instance of the synchronization logic is configured to invoke dependent instances of the synchronization logic on other cores of the multi-core processor in accordance with a hierarchical coordination system that structures inter-core coordinations in a hierarchical fashion.

8. The multi-core processor of claim 1, wherein the hierarchical coordination system groups cores into domains in accordance with resources shared by cores within those domains, and wherein for each domain, a single core is designated as a master of that domain for purposes of a coordinated configuration of those resources.

9. The multi-core processor of claim 1, wherein:

the hierarchical coordination system groups cores into a plurality of domain levels, including at least;

a primary-level domain of highest rank that includes all of the cores; and

two or more secondary-level domains of equal next rank, immediately below the highest rank, that are constituents of and nested within the primary-level domain, each secondary-level group comprising respectively exclusive subgroups of the cores;

for each multi-core domain level, a single core is designated as a master of that domain;

each multi-core domain other than the lowest-level multi-core domains defines a kinship group consisting of the master cores of constituent domains of the immediately next lower rank;

each lowest-level multi-core domain defines a kinship group consisting of all of its cores;

each core belongs to at least one kinship group; and

each native instance of the synchronization logic is restricted from invoking new instances of the synchronization logic to cores not belonging to a native core kinship group.

10. The multi-core processor of claim 1, wherein a single one of the multi-core processor'"'"'s cores is designated as a master for each multi-core domain of the hierarchical coordination system.

11. The multi-core processor of claim 1, wherein each core is configured to employ its de-centralized inter-core state discovery microcode to discover whether others of the multi-core processor'"'"'s cores are disabled.

12. The multi-core processor claim 1, wherein each core is configured to employ its de-centralized inter-core state discovery microcode to discover how many enabled cores the multi-core processor has.

13. The multi-core processor of claim 1, wherein each core is configured to employ its de-centralized inter-core state discovery microcode to discover a hierarchical coordination system of the multi-core processor.

14. A multi-core processor comprising:

a plurality of physical processing cores; and

inter-core state discovery microcode in each core enabling the core to participate in a de-centralized inter-core state discovery process through signals, unmediated by any centralized non-core logic, sent to or received from other cores;

wherein;

the inter-core state discover microcode includes synchronization logic, provided to each core, synchronized instances of which are operable to be invoked on multiple cores for purposes of an inter-core state discovery process; and

wherein each native instance is operable both to invoke new instances of the synchronization logic on other cores, and to respond to any prior instance of the synchronization logic on another core that invoked the native instance;

each core has a target operating state;

the processor includes a domain comprising at least two of the microprocessor'"'"'s cores;

the processor provides a resource to the domain, which resource is shared by the cores of the domain; and

the synchronization logic is configured to;

discover if one of the enabled cores of the domain sharing the resource has a target operating state less restrictive than a currently implemented power-conserving operating state;

enable the core, if it is authorized to coordinate with that resource, to repeal a power-conserving operating state for the resource if the synchronization logic has discovered that an enabled core of the domain has a target operating state less restrictive than the power-conserving operating state.

15. A decentralized, microcode-implemented method of discovering states of a multi-core processor comprising a plurality of physical processing cores, the method comprising:

at least two cores participating in a de-centralized inter-core state discovery process through signals, unmediated by any centralized non-core logic, exchanged by the cores;

wherein the method is applied to discover at least one of the following states;

a composite power state for the processor;

a composite power state for a domain of the processor, the domain comprising a group of cores that share a configurable resource operable to be configured, for purposes of power conservation, according to one of a plurality of configurations;

a target power state of another core;

a least restrictive target power state of any of a group of cores sharing a configurable resource;

a most restrictive operating state that can be implemented by a core without interfering with corresponding target operating states of other cores;

whether a core is enabled or disabled;

how many cores of the multi-core processor are enabled;

an identification of shared resources and domains of cores amongst which various configurable resources are shared;

a hierarchical coordination system of the cores used for managing shared resources;

an availability within the multi-core processor of sideband communication wires to coordinate cores, which sideband communication wires are independent of a system bus connecting the multi-core processor to a chipset; and

a hierarchical coordination system of the cores applied to inter-core communications over sideband communication wires that are independent of a system bus connecting the multi-core processor to a chipset.

16. The method of claim 15, wherein each participating core exchanges state-related signals with another participating core using sideband communication wires that are independent of a system bus connecting the multi-core processor to a chipset.

17. The method of claim 15, further comprising participating in the de-centralized inter-core discovery process to discover a target power state of another core.

18. The method of claim 15, further comprising participating in the de-centralized inter-core discovery process to discover a composite power state of group of cores.

19. The method of claim 15, further comprising participating in the de-centralized inter-core discovery process to restrict implementation of operating states for configuring a shared resource to an operating state that is no more restrictive, with respect to a limitation that its configuration of the resource would impose on the power, speed, or efficiency with which a core sharing the resource would be able to operate, than the least restrictive target operating state of any core sharing that resource.

20. The method of claim 15, further comprising:

each core receiving a target operating state;

each core, in response to receiving the target operating state, invoking a native instance of synchronization logic, embodied in the core'"'"'s microcode, to discover an applicable state;

wherein the applicable state is a most restrictive operating state, no greater than the core'"'"'s own target operating state, that can be implemented by the core without interfering with corresponding target operating states of other cores;

the native instance of synchronization logic invoking at least one new dependent instance of the synchronization logic on another core, and passing the native core'"'"'s target operating state to the other core; and

the dependent instance calculating a compound operating state as a function, at least, of a target operating state applicable to itself and the target operating state received from the native core, and returning the compound operating state to the native core.

21. The method of claim 20, further comprising:

each instance of synchronization logic, unless terminated earlier by a termination condition, recursively invoking instances of synchronization logic on other not-yet-synchronized cores until synchronized instances of the synchronization logic have been invoked on all of the cores of an applicable domain of the processor.

22. The method of claim 21, further comprising:

each instance of synchronization logic conditionally preventing invocations of further dependent instances of synchronization logic on other not-yet-synchronized cores if that instance discovers a core having a target operating state that is not more restrictive than the resource'"'"'s least restrictive operating state;

wherein the synchronization logic is configured to coordinate with a minimally sufficient number of other cores to discover whether a restrictive operating state may be applied to the shared resource.

23. A microcode routine encoded in a non-transitory computer readable storage medium of a physical core of a multi-core processor, the microcode routine comprising code for discovering an applicable state of the multi-core processor using a de-centralized inter-core state discovery process through signals, unmediated by any centralized non-core logic, exchanged by the cores;

wherein the applicable state is one of the following states;

a composite power state for the processor;

a composite power state for a domain of the processor, the domain comprising a group of cores that share a configurable resource operable to be configured, for purposes of power conservation, according to one of a plurality of configurations;

a target power state of another core;

a least restrictive target power state of any of a group of cores sharing a configurable resource;

a most restrictive operating state that can be implemented by a core without interfering with corresponding target operating states of other cores;

whether a core is enabled or disabled;

how many cores of the multi-core processor are enabled;

an identification of shared resources and domains of cores amongst which various configurable resources are shared;

a hierarchical coordination system of the cores used for managing shared resources;

an availability within the multi-core processor of sideband communication wires to coordinate cores, which sideband communication wires are independent of a system bus connecting the multi-core processor to a chipset; and

a hierarchical coordination system of the cores applied to inter-core communications over sideband communication wires that are independent of a system bus connecting the multi-core processor to a chipset.