Hewlett-packard Development Company, L.p.

Mosur Kumaraswamy Ravishankar

Kourosh Gharachorloo

Robert J Stets, Jr

Luiz Andre Barroso

Andreas Georg Nowatzyk

Daniel J Scales

Luiz Andr&eacute; Barroso

A protocol engine is for use in each node of a computer system having a plurality of nodes. Each node includes an interface to a local memory subsystem that stores memory lines of information, a directory, and a memory cache. The directory includes an entry associated with a memory line of information stored in the local memory subsystem. The directory entry includes an identification field for identifying sharer nodes that potentially cache the memory line of information. The identification field has a plurality of bits at associated positions within the identification field. Each respective bit of the identification field is associated with one or more nodes. The protocol engine furthermore sets each bit in the identification field for which the memory line is cached in at least one of the associated nodes. In response to a request for exclusive ownership of a memory line, the protocol engine sends an initial invalidation request to no more than a first predefined number of the nodes associated with set bits in the identification field of the directory entry associated with the memory line.

System and method for limited fanout daisy chaining of cache invalidation requests in a shared-memory multiprocessor system

A computer system has a plurality of processor nodes and a plurality of input/output nodes. Each processor node includes a multiplicity of processor cores, an interface to a local memory system and a protocol engine implementing a predefined cache coherence protocol. Each processor core has an associated memory cache for caching memory lines of information. Each input/output node includes no processor cores, an input/output interface for interfacing to an input/output bus or input/output device, a memory cache for caching memory lines of information and an interface to a local memory subsystem. The local memory subsystem of each processor node and input/output node stores a multiplicity of memory lines of information. The protocol engine of each processor node and input/output node implements the same predefined cache coherence protocol.

Multiprocessor cache coherence system and method in which processor nodes and input/output nodes are equal participants

L cache synonyms in a two-level cache system are detected and resolved by logic in the L cache. Duplicate copies of the L cache tags and state (&ldquo;Dtags&rdquo;) are maintained in the L cache. After a miss occurs in the L cache, the Dtags in the second-level cache that correspond to all possible synonym locations in the L cache are searched for synonyms. If a synonym is found, the L cache notifies the L cache where the requested cache line can be found in the L cache. The L cache then copies the cache line from the location where the synonym was found to the location where the miss occurred, and it invalidates the cache line at the original location. The Dtags in the second-level cache are updated to reflect the changes made in the L cache.

Method and system for detecting and resolving virtual address synonyms in a two-level cache hierarchy

The present invention relates generally to multiprocessor computer system, and particularly to a multiprocessor system designed to be highly scalable, using efficient cache coherence logic and methodologies. More specifically, the present invention is a system and method including a plurality of processor nodes configured to execute a cache coherence protocol that avoids the use of negative acknowledgment messages (NAKs) and ordering requirements on the underlying transaction-message interconnect/network and services most 3-hop transactions with only a single visit to the home node.

Scalable multiprocessor system and cache coherence method

A system including a plurality of processor nodes is configured to execute a cache coherence protocol that avoids the use of negative acknowledgments and ordering requirements on the underlying transaction-message interconnect/network, and implements store-conditional memory transactions. A store-conditional memory transaction succeeds if a directory tracking the state of a memory line of information unambiguously indicates that the requesting node is the exclusive owner of the memory line, if the directory ambiguously indicates that the requesting node is sharing the memory line and the requesting node is in fact sharing the memory line, or if the directory unambiguously indicates that the requesting node is sharing the memory line. The store-conditional memory transaction fails if the directory unambiguously indicates that the requesting node is not sharing the memory line, or if the directory ambiguously indicates that the requesting node may be sharing the memory line and the requesting node is in fact not sharing the memory line.

Scalable multiprocessor system and cache coherence method implementing store-conditional memory transactions while an associated directory entry is encoded as a coarse bit vector

An invalid-to-dirty request permits a transition from an invalid memory state to a dirty state without requiring an up-to-date copy of the memory. The present invention is a system for supporting invalid-to-dirty memory transactions in an aggressive cache coherence protocol that minimizes directory entry locking. The nodes of a multiprocessor system each include a protocol engine that is configured to implement a distinct invalidation request that corresponds to an invalid-to-dirty memory transaction. If node O receives this distinct invalidation request while waiting for a response to an outstanding request for exclusive ownership, the protocol engine of node O is configured to treat the distinct invalidation request as applying to the memory line of information that is the subject of the outstanding request for exclusive ownership. Furthermore, if the node O receives a normal invalidation request, while it has an outstanding exclusive request, the invalidation request applies to a previous copy of the memory line of information held by the node N and therefore the protocol engine of node O is configured to ignore the normal invalidation request in this circumstance.

Scalable multiprocessor system and cache coherence method incorporating invalid-to-dirty requests

In a chip multiprocessor system, the coherence protocol is split into two cooperating protocols implemented by different hardware modules. One protocol is responsible for cache coherence management within the chip, and is implemented by a second-level cache controller. The other protocol is responsible for cache coherence management across chip multiprocessor nodes, and is implemented by separate cache coherence protocol engines. The cache controller and the protocol engine within each node communicate and synchronize memory transactions involving multiple nodes to maintain cache coherence within and across the nodes. The present invention addresses race conditions that arise during this communication and synchronization.

System for handling coherence protocol races in a scalable shared memory system based on chip multiprocessing

The present invention relates generally to a protocol engine for use in a multiprocessor computer system. The protocol engine, which implements a cache coherence protocol, includes a clock signal generator for generating signals denoting interleaved even clock periods and odd clock periods, a memory transaction state array for storing entries, each denoting the state of a respective memory transaction, and processing logic. The memory transactions are divided into even and odd transactions whose states are stored in distinct sets of entries in the memory transaction state array. The processing logic has interleaving circuitry for processing during even clock periods the even memory transactions and for processing during odd clock periods the odd memory transactions. Moreover, the protocol engine is configured to transition from one memory transaction to another in a minimum number of clock cycles. This design improves efficiency for processing commercial workloads, such as on-line transaction processing (OLTP) by taking certain steps in parallel.

Cache coherence protocol engine and method for efficient processing of interleaved memory transactions in a multiprocessor system

The present invention relates generally to a protocol engine for use in a multiprocessor computer system. The protocol engine, which implements a cache coherence protocol, includes a clock signal generator for generating signals denoting interleaved even clock periods and odd clock periods, a memory transaction state array for storing entries, each denoting the state of a respective memory transaction, and processing logic. The memory transactions are divided into even and odd transactions whose states are stored in distinct sets of entries in the memory transaction state array. The processing logic has interleaving circuitry for processing during even clock periods the even memory transactions and for processing during odd clock periods the odd memory transactions.

Cache coherence protocol engine system and method for processing memory transaction in distinct address subsets during interleaved time periods in a multiprocessor system

Growing community of inventors

Mountain View, CA, United States of America

Mosur Kumaraswamy Ravishankar