Freescale Semiconductor,inc.

Marvell Israel (misl) Ltd.

Yakov Tokar

Itay S Peled

Ziv Zamsky

Moshe Anschel

Lior Moheban

Doron Schupper

Jacob Efrat

Alex Pinskiy

Ilan Mayer-Wolf

Amit Gur

Kostantin Godin

Alon Eldar

Gregory Kovishaner

Itay Keidar

Ronen Goldberg

Moshe Bachar

Roman Landa

Itay Peled

A network device includes a memory configured to store a plurality of entries in respective locations in the memory, the plurality of entries corresponding to a trie data structure for performing a longest prefix match search. The network device also includes: a memory access engine configured to retrieve from a location in the memory, in a single memory lookup operation, i) longest prefix match information for a node corresponding to a network address in a header of a packet, and ii) pointer information that indicates a child node in the trie data structure. The network device also includes: a child node address calculator configured to use i) the longest prefix match information, and ii) the pointer information, to calculate a memory address of another location in the memory corresponding to the child node.

Method and apparatus for longest prefix match search

A first-in/first-out (FIFO) buffer includes at least one latch-based FIFO storage line, an input flip-flop stage upstream of the at least one latch-based storage line, an output flip-flop stage downstream of the at least one latch-based storage line. The output flip-flop stage functions as an additional storage line. Clock-gating circuitry separate from the device clock controls timing of the at least one latch-based FIFO storage line, the input flip-flop stage, and the output flip-flop stage. The input flip-flop stage functions as a second additional storage line, or as an input sampling stage. Optional bypass circuitry between the input flip-flop stage and the output flip-flop stage passes data for a storage line directly to the output flip-flop stage, without passing through the at least one latch-based storage line, when the buffer is empty.

Hybrid FIFO buffer

Embodiments described herein provide a dual-line FIFO structure without the use of any multiplexer. Instead, the dual-line FIFO described herein uses a selectively transparent latch and a flip-flop serially connected to the latch, such that the combination of the serially connected latch and the flip-flop can temporarily store up to two data units at two clock cycles.

Systems and methods for transferring data with a dual-line first-in-first-out (FIFO) memory array

A circuitry for a computing system comprising a first load/store unit, LSU, and a second LSU as well as a memory arrangement. The first LSU is connected to the memory arrangement via a first bus arrangement comprising a first write bus and a first read bus. The second LSU is connected to the memory arrangement via a second bus arrangement comprising a second write bus and a second read bus. The computing system is arranged to carry out a multiple load instruction to read data via the first read bus and the second read bus and/or to carry out a multiple store instruction to write data via the first write bus and the second write bus.

Circuitry for a computing system, LSU arrangement and memory arrangement as well as computing system

A write back allocate system that includes: (i) a store request circuit; (ii) a processor, adapted to generate a store request that comprises an information unit and an information unit address; and (iii) a cache module, connected to the store request circuit and to a high level memory unit. A single cache module line includes multiple segments, each segment is adapted to store a single information unit. A content of a cache module line is retrieved from the high level memory unit by generating a fetch burst that includes multiple segment fetch operations. The store request circuit includes a snooper and a controller. The snooper detects a portion of an address of a cache segment of a cache line that is being fetched during a fetch burst. The controller is adapted to request from the cache module to receive the information unit before a completion of the fetch burst if the portion of the address of the cache segment matches a corresponding portion of the information unit address.

System and method for fetching information to a cache module using a write back allocate algorithm

A virtual address cache and a method for sharing data. The virtual address cache includes: a memory, adapted to store virtual addresses, task identifiers and data associated with the virtual addresses and the task identifiers; and a comparator, connected to the memory, adapted to determine that data associated with a received virtual address and a received task identifier is stored in the memory if at least a portion of the received virtual address equals at least a corresponding portion of a certain stored virtual address and a stored task identifier associated with the certain stored virtual address indicates that the data is shared between multiple tasks. The method includes: receiving a virtual address and a task identifier for addressing data; determining as to whether the data is stored in a memory based upon a comparison of at least a portion of the received virtual address with at least a portion of an address associated with data stored in memory and based upon a fulfillment of the following criterion: a stored task identifier associated with the certain stored virtual address indicates that the data is shared between multiple tasks.

Virtual address cache and method for sharing data using a unique task identifier

A processing system has a processor, a cache, and a fetch unit. If there is a miss in the cache, the fetch unit generates a fetch address for the miss in the cache for the purpose of retrieving the requested data from external memory and providing the data to the processor and loading the data in a location in a line in the cache. The fetch unit also generates additional prefetch addresses for addresses consecutive with the fetch address. The prefetch addresses continue to be generated for all of the locations in the line in the cache that are consecutive with the fetch address. The generation of prefetch addresses will be stopped if another requests arrives that is not part of the already generated prefetched addresses. Further, the outstanding prefetches will be terminated if the external memory can handle such termination.

Growing community of inventors

Rishon Lezion, Israel

Yakov Tokar