Multi-channel DMA with shared FIFO

• US 7,373,437 B2
• Filed: 03/15/2005
• Issued: 05/13/2008
• Est. Priority Date: 10/11/2004
• Status: Active Grant

• Current Assignee: Texas Instruments, Inc.
• Original Assignee: Texas Instruments, Inc.
• Inventors: Khalifa, Nabil, Seigneret, Franck, Kolli, Praveen, Ayinala, Sivayya
• Primary Examiner(s): PATEL, NIKETA I
• Application Number:
  US11/080,277
• Publication Number:
  US 20060080477A1
• Time in Patent Office: 1,155 Days
• Field of Search: 710/22, 710/26, 709/212
• US Class Current: 710/22
• CPC Class Codes: G06F 13/28 using burst mode transfer, e.g. direct memory access DMA, cycle steal G06F13/32 takes precedence

Abstract

A direct memory access (DMA) circuit (200) includes a read port (202) and a write port (204). The DMA circuit (200) is a multithreaded initiator with “m” threads on the read port (202) and “n” threads on the write port (204). The DMA circuit (200) includes a data FIFO (210) which is shared by all of the logical channels and the FIFO depth can be allocated dynamically allowing for the maximum number of channels to be scheduled and concurrently active. The FIFO (210) can also be allocated to a single channel if there is only one logical channel active. The FIFO (210) increases the DMA'"'"'s transfer performance, pre-fetch capacity and buffering, while maximizing pipelining.

Claims

1. A data storage area for use in a direct memory access (DMA) circuit comprising:

a data memory which is shared by a plurality of channels supported by the DMA circuit;

a next read address storage area coupled to the data memory, the next read address storage area stores a next read address indicative of where to read a next data associated with a DMA write over a first currently granted channel; and

a next write address storage area coupled to the data memory, the next write address storage area allocates a plurality of next write addresses, each of the plurality of next write addresses indicative of where to write the next data associated with a DMA read over a second currently granted channel.

2. The data storage area as defined in claim 1, wherein the data memory comprises a static random access memory (SRAM) or flip-flop based memory.

3. The data storage area as defined in claim 1, wherein the data memory is allocated dynamically in order to maximize the number of channels from among the plurality of channels to be scheduled and concurrently active.

4. The data storage area as defined in claim 1, a means coupled to the data memory for controlling the amount of memory in the data memory each channel that is active amongst the plurality of channels can use.

5. The data storage area as defined in claim 1, wherein the data memory stores data written into the data memory as part of the DMA write.

6. The data storage area as defined in claim 5, the data memory stores the next read address, indicative of where to read the next data associated with the DMA write over the first currently granted channel.

7. The data storage area as defined in claim 6, wherein if a requesting channel gets granted for a DMA write, a new next write address is retrieved from the next write storage area if the channel that has been granted has been scheduled for the DMA write for the first time.

8. The data storage area as defined in claim 6, wherein the next read address storage area is integral to the data memory.

9. The data storage area as defined in claim 6, wherein the next write address storage area comprises a first-in-first out (FIFO) circuit.

10. The data storage area as defined in claim 9, wherein the next write address output value from the next write address FIFO circuit is written into the next read address storage area.

11. A direct memory access (DMA) circuit, comprising:

a read port;

a write port; and

a unified shared FIFO circuit coupled to the read and write ports, the unified shared FIFO circuit includes;

a data memory which is shared amongst a plurality of channels supported by the DMA circuit; and

a next write address storage area coupled to the data memory, the next write address storage area allocates a plurality of next write addresses, each of the plurality of next write addresses indicative of where to write a next data of a channel currently being serviced on the read port.

12. The DMA circuit as defined in claim 11, wherein the data memory includes a next read address storage area, the next read address storage area stores a next read address indicative of where to read the next data from a channel currently granted on the write port.

13. The DMA circuit as defined in claim 12, wherein the data memory is allocated dynamically in order to maximize the number of channels from among the plurality of channels to be scheduled and concurrently active.

14. The DMA circuit as defined in claim 13, wherein the next write storage area comprises a first-in-first-out (FIFO) circuit.

15. The DMA circuit as defined in claim 13, further comprising:

an endianness conversion circuit coupled to an input of the data memory; and

an address alignment control circuit coupled to an output of the data memory.

16. The DMA circuit as defined in claim 15, wherein the endianness conversion circuit performs endianness conversion prior to the data being stored in the data memory.

17. The DMA circuit as defined in claim 13, further comprising:

a write port scheduler coupled to the write port, the write port scheduler selects the next channel to be serviced and for allocating a thread identifier to a thread associated with a DMA service.

18. The DMA circuit as defined in claim 17, further comprising:

a read port scheduler coupled to the read port for scheduling the next channel to be serviced by the read port.

19. The DMA circuit as defined in claim 18, wherein the read and write ports both include a high priority and a low priority queue for arbitrating which channel from the plurality of channels supported by the DMA will be serviced.

20. The DMA circuit as defined in claim 18, wherein the read and write ports can work on different data transfers at the same time.

21. The DMA circuit as defined in claim 18, further comprising:

a means for clearing channel data from the unified shared FIFO circuit in case of one or more channels from amongst the plurality of channels becoming disabled or aborted.