Abstract: Methods and system for memory control in a computer system having a store-in cache. In response to main memory read or write requests from a secondary processor, data is transferred into a buffer during a snoop cycle to the store-in cache. The data in the buffer is merged with write-back data from the store-in cache in a write operation. Data is provided directly from the buffer to the secondary processor and to main memory in a read operation. The buffer can be placed on a memory controller of the computer system. A second store-in cache can also be used for main memory transfers.

Abstract: Methods and system for memory control in a computer system having a store-in cache. In response to main memory read or write requests from a secondary processor, data is transferred into a buffer during a snoop cycle to the store-in cache. The data in the buffer is merged with write-back data from the store-in cache in a write operation. Data is provided directly from the buffer to the secondary processor and to main memory in a read operation. The buffer can be placed on a memory controller of the computer system. A second store-in cache can also be used for main memory transfers.

Abstract: Incremental timing analyzer for selectively performing timing analysis on a revised electronic circuit design resulting from one or more modifications to an initial electronic circuit design having input nodes, output nodes, and active elements electrically connected therebetween in a set of signal paths interconnected by a plurality of nodes. Each signal path has a timing delay associated therewith. Data is recorded representative of the modification's affect on relative timing values for a set of signals propagated through the circuit design. The recorded data includes a leftmost frontier of change in relative timing values and a rightmost frontier of change in relative timing values. Upon presentation of a specific timing analysis request, incremental timing analysis on the selected portion of the modified electronic circuit design is conducted employing the recorded frontiers of change to limit the timing value analysis.