Assigning memory to on-chip coherence domains
View Patent ↗A mechanism for assigning memory to on-chip cache coherence domains assigns caches within a processing unit to coherence domains. The mechanism assigns chunks of memory to the coherence domains. The mechanism monitors applications running on cores within the processing unit to identify needs of the applications. The mechanism may then reassign memory chunks to the cache coherence domains based on the needs of the applications running in the coherence domains. When a memory controller receives the cache miss, the memory controller may look up the address in a lookup table that maps memory chunks to cache coherence domains. Snoop requests are sent to caches within the coherence domain. If a cache line is found in a cache within the coherence domain, the cache line is returned to the originating cache by the cache containing the cache line either directly or through the memory controller.
1. A method, in a data processing system, for assigning memory to on-chip coherence domains, the method comprising:
assigning caches on a processing unit chip to a plurality of coherence domains;
assigning portions of memory to the plurality of coherence domains;
generating a mapping table, wherein the mapping table associates each portion of memory to an identifier of a coherence domain to which it is assigned;
responsive to a cache generating a cache miss, looking up an address of the cache miss in the mapping table to identify a coherence domain; and
snooping caches within the identified coherence domain.
2. The method of claim 1 , further comprising:
monitoring applications running on processing cores in the processing unit chip to determine memory usage by the applications; and
reassigning the portions of memory to the plurality of coherence domains based on the determined memory usage by the applications.
3. The method of claim 2 , further comprising:
updating the mapping table to reflect the reassignment portions of memory to the plurality of coherence domains.
4. The method of claim 1 , wherein the portions of memory are fixed sized ranges of memory.
5. The method of claim 1 , further comprising:
responsive to a snoop request, a cache in the identified coherence domain returning a cache line to the cache that generated the cache miss.
6. The method of claim 1 , further comprising:
accessing a portion of memory corresponding to the cache miss to receive a cache line; and
returning the cache line to the cache that generated the cache miss.