Migrating sleeping and waking threads between wake-and-go mechanisms in a multiple processor data processing system

1. A method, in a data processing system, for performing a wake-and-go operation, the method comprising:

responsive to a first wake-and-go mechanism associated with a first processing unit within a plurality of processing units detecting a thread running on the first processing unit that is waiting for an event that modifies a data value associated with a target address, placing the thread in a sleep state;

creating a wake-and-go instance for the thread comprising the target address and a thread identifier associated with the thread;

assigning the wake-and-go instance to a second processing unit within the plurality of processing units;

storing, by a second wake-and-go mechanism associated with the second processing unit, the wake-and-go instance in a wake-and-go storage array associated with the second processing unit; and

responsive to the second wake-and-go mechanism detecting the event that modifies the data value associated with the target address, placing the thread in a non-sleep state,

wherein placing the thread in a non-sleep state comprises:

identifying a third processing unit within the plurality of processing units that has a lowest processor utilization within the plurality of processing units or a lowest priority thread running within the plurality of processing units; and

placing the thread in a run queue of the third processing unit.

2. The method of claim 1 , wherein the third processing unit has the lowest priority thread running within the plurality of processing units and wherein placing the thread in a non-sleep state on the third processing unit comprises placing the lowest priority thread in a sleep state on the third processing unit.

3. The method of claim 1 , wherein assigning the wake-and-go instance to the second processing unit comprises transferring the wake-and-go instance from the first wake-and-go mechanism to the second wake-and-go mechanism.

4. The method of claim 1 , wherein the wake-and-go storage array is a content addressable memory and an address on a bus is used to address the content addressable memory.

5. The method of claim 4 , wherein responsive to detecting the event that modifies the data value associated with the target address, the content addressable memory returns a storage address at which the target address is stored in the content addressable memory.

6. The method of claim 1 , wherein assigning the wake-and-go instance to the second processing unit comprises:

identifying the second processing unit from the plurality of processing units based on a number of wake-and-go instances in the wake-and-go storage array associated with the second processing unit.

7. The method of claim 1 , wherein placing the thread in a sleep state comprises storing thread state information of the thread in a thread state storage, wherein the wake-and-go instance comprises a thread state pointer referencing the thread state information in the thread state storage, and wherein placing the thread in a non-sleep state comprises restoring the thread state information using the thread state pointer and placing the thread in a run queue of the third processing unit.

8. A data processing system, comprising:

a plurality of processing units;

a first wake-and-go mechanism associated with a first processing unit within the plurality of processing units;

a second wake-and-go mechanism associated with a second processing unit within the plurality of processing units; and

a wake-and-go storage array associated with the second processing unit,

wherein the first wake-and-go mechanism is configured to:

responsive to detecting a thread running on the first processing unit that is waiting for an event that modifies a data value associated with a target address, place the thread in a sleep state;

create a wake-and-go instance for the thread comprising the target address and a thread identifier associated with the thread; and

assign the wake-and-go instance to the second processing unit;

wherein the second wake-and-go mechanism is configured to:

store the wake-and-go instance in the wake-and-go storage array; and

responsive to the second wake-and-go mechanism detecting the event that modifies the data value associated with the target address, place the thread in a non-sleep state;

wherein placing the thread in a non-sleep state comprises:

identifying a third processing unit within the plurality of processing units that has a lowest processor utilization within the plurality of processing units or a lowest priority thread running within the plurality of processing units; and

placing the thread in a run queue of the third processing unit.

9. The data processing system of claim 8 , wherein the third processing unit has the lowest priority thread running within the plurality of processing units and wherein placing the thread in a non-sleep state on the third processing unit comprises placing the lowest priority thread in a steep state on the third processing unit.

10. The data processing system of claim 8 , wherein assigning the wake-and-go instance to the second processing unit comprises transferring the wake-and-go instance from the first wake-and-go mechanism to the second wake-and-go mechanism.

11. The data processing system of claim 8 , wherein the wake-and-go storage array is a content addressable memory and an address on a bus is used to address the content addressable memory.

12. The data processing system of claim 11 , wherein responsive to detecting the event that modifies the data value associated with the target address, the content addressable memory returns a storage address at which the target address is stored in the content addressable memory.

13. The data processing system of claim 8 , wherein the first wake-and-go mechanism is a hardware component within the first processing unit and wherein the second wake-and-go mechanism is a hardware component within the second processing unit.

14. A computer program product comprising a non-transitory computer readable medium storing a computer readable program, wherein the computer readable program, when executed on a computing device, causes the computing device to:

responsive to a first wake-and-go mechanism associated with a first processing unit within a plurality of processing units detecting a thread running on the first processing unit that is waiting for an event that modifies a data value associated with a target address, place the thread in a sleep state;

create a wake-and-go instance for the thread comprising the target address and a thread identifier associated with the thread;

assign the wake-and-go instance to a second processing unit within the plurality of processing units;

store the wake-and-go instance in a wake-and-go storage array associated with the second processing unit; and

responsive to a second wake-and-go mechanism associated with the second processing unit detecting the event that modifies the data value associated with the target address, place the thread in a non-sleep state,

wherein placing the thread in a non-sleep state comprises:

identifying a third processing unit within the plurality of processing units that has a lowest processor utilization within the plurality of processing units or a lowest priority thread running within the plurality of processing units; and

placing the thread in a run queue of the third processing unit.

15. The computer program product of claim 14 , wherein assigning the wake-and-go instance to the second processing unit comprises:

identifying the second processing unit from the plurality of processing units based on a number of wake-and-go instances in the wake-and-go storage array associated with the second processing unit.

16. The computer program product of claim 14 , wherein placing the thread in a sleep state comprises storing thread state information of the thread in a thread state storage, wherein the wake-and-go instance comprises a thread state pointer referencing the thread state information in the thread state storage, and wherein placing the thread in a non-sleep state comprises restoring the thread state information using the thread state pointer and placing the thread in a run queue of the third processing unit.