Fifo read interface protocol

Electrical computers and digital processing systems: memory – Storage accessing and control – Access timing

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

Reexamination Certificate

active

06629226

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to a method and/or architecture for implementing a multiqueue first-in-first-out (FIFO) memory read interface generally and, more particularly, to a method and/or architecture for implementing a multiqueue FIFO read interface protocol for eliminating synchronizing problems for configuration dependent latencies where the protocol may be capable of handling variable size packets.
The present invention may also relate to a method and/or architecture for variable stage pipeline system generally and, more particularly, to a method and/or architecture for implementing an event driven variable stage pipeline system for handling variable size blocks that may have a minimum block size less than total round-time delay.
BACKGROUND OF THE INVENTION
Referring to
FIG. 1
, a diagram illustrating a conventional circuit
10
for exchanging data between a first-in-first-out (FIFO) device
12
and a read device
14
is shown. A read signal RD is presented from the read device
14
to the FIFO
12
. After a fixed number of latency cycles, the signal DATA is presented to the read device
14
. The signals READ_CLOCK and RD control the timing of the presentation of the data signal DATA.
Referring to
FIG. 2
, a diagram illustrating a conventional system
30
for exchanging data between a multiqueue FIFO
32
and a read device
34
is shown. The signal ADDRESS is a queue address configured to determine a queue number of the multiqueue FIFO
32
. The signals READ_CLOCK and READ_EN control the timing of the presentation of the data signal DATA.
The read signal RD in
FIG. 1
is replaced by the read enable signal READ_EN in FIG.
2
. The signal READ_EN controls whether to continue or to stop a particular read. The queue address signal ADDRESS is an additional signal not present in FIG.
1
. Since there are multiple queues in the FIFO
32
, a read occurs from the particular queue that is addressed by the signal ADDRESS.
Because of particular architectures (e.g., the cited co-pending applications) and specifications of particular devices, the latency between enabling the queue address signal ADDRESS and presenting the signal DATA can differ depending on the particular configuration. The configuration information needs to be written into the external read device
34
. The only event reference available to the external read device
34
is an end of packet or a start of packet (EOP/SOP). In such an environment, the read device
34
is required to monitor this event to generate the queue address signal ADDRESS in a sufficient number of cycles ahead of the read.
The circuit
30
has the disadvantage of requiring a fixed packet size. The circuit
30
can be required to generate the queue address ADDRESS a certain number of cycles before the EOP occurs. The particular number of cycles is the same as the minimum latency requirement. For certain configurations, there is a specific latency between the queue address signal ADDRESS and presenting the signal DATA. If the packet size varies randomly, such as when the size of the packet is less than the number of cycles of latency, a read of one or more unwanted packets occurs. The circuit
30
additionally requires a pipeline memory (within the multiqueue
32
) to handle variable sized packets in an asynchronous configuration.
It may also be difficult for the read device
34
to synchronize the queue address signal ADDRESS with the data received from the FIFO
32
. Therefore, the read device
34
needs to be configured with enough logic to respond to the different latencies. Such a configuration requires extra overhead for the read device
34
.
Handling of slow read clock speeds is also difficult within the circuit
30
. The read operation occurs at one clock and the internal logic operates at another clock, such as the system clock (e.g., the cited co-pending applications). This requires the queue address signal ADDRESS and the read enable signal READ_EN to be synchronized before other processing can be executed. The data read also needs to be synchronized. As a result, there is an uncertainty of 1-2 clock cycles resulting in the latency. This makes the synchronization between the queue address signal ADDRESS and the data read extremely difficult.
SUMMARY OF THE INVENTION
One aspect of the present invention concerns an interface coupled to a multiqueue storage device and configured to interface the multiqueue storage device with one or more handshaking signals. The multiqueue storage device and the interface may be configured to transfer variable size data packets.
Another aspect of the present invention concerns an apparatus configured to interface a first clock speed of a multiqueue storage device and a second clock speed of an interface. The apparatus may be configured to control a flow of variable size data packets.
The objects, features and advantages of the present invention include providing a method and/or architecture for implementing a multiqueue FIFO read interface protocol that may (i) eliminate synchronizing problems with configuration dependent latencies; (ii) be capable of handling variable size packets; (iii) allow back-to-back reads of variable size packets; (iv) exchange address and data between an external read device and a multiqueue storage device; (v) generate an address request for an external device from the storage device; (vi) generate a valid queue address in response to the address request; (vii) provide data in response to the valid queue address; (viii) provide a single clock domain or a dual clock domain between a multiqueue storage device and a read interface device; (ix) vary a latency between an address request and an address validate; (x) provide an event driven variable stage pipeline system; and/or (xi) handle variable size packets with a minimum packet size less than a total round-time delay.


REFERENCES:
patent: 5673396 (1997-09-01), Smolansky et al.
patent: 6442646 (2002-08-01), Tsuruta
patent: 6466581 (2002-10-01), Yee et al.
patent: 6470415 (2002-10-01), Starr et al.
Tamir et al, “Dynamically-Allocated Multi-Queue Butters for VLSI Communication Switches”, IEEE, 1992, p 725-737.*
S. Babar Raza et al., “Architecture for Implementing Virtual Multiqueue Fifos”, Ser. No. 09/676,704, Filed: Sep. 29, 2000.
S. Babar Raza et al., “Method and Logic for Storing and Extracting In-Band Multicast Port Information Stored Along With the Data in a Single Memory Without Memory Read Cycle Overhead”, Ser. No. 09/676,171, Filed: Sep. 29, 2000.
S. Babar Raza et al., “Logic for Generating Multicast/Unicast Address (ES)”, Ser. No. 09/676,706, Filed: Sep. 29, 2000.
S. Babar Raza et al., “Logic for Initializing the Depth of the Queue Pointer Memory”, Ser. No. 09/676,705, Filed: Sep. 29, 2000.
S. Babar Raza et al., “Method and Logic for Initializing the Forward-Pointer Memory During Normal Operation of the Device as a Background Process”, Ser. No. 09/676,170, Filed: Sep. 29, 2000.
S. Babar Raza et al., “Method and/or Architecture for Implementing Queue Expansion in Multiqueue Devices”, Ser. No. 09/714,441, Filed: Nov. 16, 2000.
Somnath Paul et al., “Fifo Read Interface Protocol”, Ser. No. 09/732,686, Filed: Dec. 8, 2000.
Somnath Paul et al., “Out-of-Band Look-Ahead Arbitration Method and/or Architecture”, Ser. No. 09/732,687, Filed: Dec. 8, 2000.
S. Babar Raza et al., “Logic for Providing Arbitration for Synchronous Dual-Port Memory”, Ser. No. 09/676,169, Filed: Sep. 29, 2000.
Jiann-Cheng Chen et al., “Configurable Fast Clock Detection Logic With Programmable Resolution”, Ser. No. 09/775,372, Filed: Feb. 1, 2001.

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

Fifo read interface protocol does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Fifo read interface protocol, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Fifo read interface protocol will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-3109065

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.