Non-blocking hierarchical cache throttle

A data cache unit associated with a processor, the data cache unit comprising:

• a first non-blocking cache;
• a second non-blocking cache coupled to the first non-blocking cache to service misses in the first non-blocking cache;
• a memory scheduling window including a plurality of entries, each entry holding a memory access, and each entry including an address identifying a memory location having data that is a target of the memory access;
• a picker associated with the memory scheduling window for picking an entry from the memory scheduling window, and for applying an address within the picked entry to the first non-blocking cache;
• a resource monitor within the second non-blocking cache for monitoring resources in the second non-blocking cache, and for generating a control signal in response to a depletion of resources in the second non-blocking cache;
• and
• means coupling the control signal to the picker to stall the picker upon a predetermined depletion of resources in the second non-blocking cache.

A computer system comprising:

• a processor;
• memory;
• and
• a cache system, the cache system comprising:
• a first non-blocking cache;
• a second non-blocking cache coupled to the first non-blocking cache to service misses in the first non-blocking cache;
• a resource monitor within the second non-blocking cache for monitoring resources in the second non-blocking cache and for generating a control signal in accordance with resources available in the second non-blocking cache;
• a memory scheduling window having a plurality of entries holding memory accesses, each memory access including an address identifying a memory location having data that is a target of the memory access;
• a picker associated with the memory scheduling window for a picking memory access from the memory scheduling window and for applying an address within the picked memory access to the first non-blocking cache;
• and
• means connecting the control signal to the picker to stall the picker upon a depletion of resources available in the second non-blocking cache to service misses in the first non-blocking cache.