|
Hybrid write back/write through cache having a streamlined four state cache coherency protocol for uniprocessor computer systems
| Details |
Inventors: Lichy, Joe;
Assignee: Intel Corporation (Santa Clara, CA)
Primary Examiner: Kim; Matthew M.
Assistant Examiner:
Attorney, Agent or Firm: Blakely, Sokoloff, Taylor & Zafman
A cache controller of a hybrid write back/write through cache of a uniprocessor system is provided with state transition and complimentary logic that implements a streamlined (modified, exclusive, shared and invalid) MESI cache coherency protocol, (modified, exclusive, pseudo shared and invalid) ME.SIGMA.I, having a pseudo shared state ".SIGMA.". Under the streamlined ME.SIGMA.I, a cache line will enter the .SIGMA. state only through allocation. From the .SIGMA. state, the cache line will remain in the .SIGMA. state when it is read, written into, or snoop-inquired, and it will transition into the I state when it is snoop-invalidated. Additionally, if a cache line in the .SIGMA. state is written into, the cache controller will always cause the data to be written to memory, effectively treating the cache line as a dedicated write through cache line. |
|
DETAILED DESCRIPTION Under the present invention, the desired results are advantageously achieved by providing the cache controller of a hybrid write back/write through cache of an uniprocessor system with state transition and complementary logic that implements a streamlined (modified, exclusive, shared and invalid) MESI cache coherency protocol, (modified, exclusive, pseudo shared and invalid) ME.
SIGMA.
I, having a pseudo shared state ".
SIGMA.
".
Under the streamlined ME.
SIGMA.
I, a cache line will enter the .
SIGMA.
state only through allocation.
From the .
SIGMA.
state, the cache line will remain in the .
SIGMA.
state when it is read, written into, or snoop-inquired, and it will transition into the I state when it is snoop-invalidated.
Additionally, if a cache line in the .
SIGMA.
state is written into, the cache controller will always cause the data to be written to memory, effectively treating the cache line as a dedicated write through cache line.
As a result, cache coherency is maintained without the complexity and cost of the prior art MESI protocol, but without the disadvantages of the prior art MVI protocol.
|
| Related patents |
|
|
Method and apparatus to control cache memory in multiprocessor system utilizing a shared memory
It is an object of the present invention to provide a cache memory control method which is suitable for all kinds of software architectures. It is another object of the ...
|
|
|
Method and system for efficient maintenance of data coherency in a multiprocessor system utilizing cache synchronization
It is therefore one object of the present invention to provide enhanced data coherency in a data processing system. It is another object of the present invention to ...
|
|
|
Dynamic mechanism to upgrade o state memory-consistent cache lines
In view of the above and other shortcomings in the art recognized by the present invention, the present invention introduces an O cache consistency state that permits ...
|
|
|
Non-volatile semiconductor memory capable of readily erasing data
Therefore, one object of the present invention is to provide a non-volatile semiconductor memory in which the area occupied by one memory cell can be reduced. Another ...
|
|
|
Main memory system and checkpointing protocol for a fault-tolerant computer system using a read buffer
Embodiments of the present invention provide a main memory device and a process for maintaining, in a computer system, a consistent, periodically-updated, checkpoint ...
|
|
|
Interrupts between asynchronously operating CPUs in fault tolerant computer system
In accordance with one embodiment of the invention, a computer system employs three identical CPUs typically executing the same instruction stream, and has two identical,...
|
|
|
Memory processing system and method for accessing memory including reordering memory requests to reduce mode switching
In accordance with the present invention, memory accesses are reordered to improve efficiency. A memory controller is used to arbitrate memory access requests from a ...
|
|
|
Method and apparatus for snoop stretching using signals that convey snoop results
The present invention provides a protocol and related apparatus for snoop stretching in a computer system having at least one requesting agent for issuing bus ...
|
|
|
Processing Data 
