# US 2008/0307121 A1 · Direct Memory Access Transfer Completion Notification

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2008/0307121 A1
- Document ID: US20080307121A1
- Status: Published application
- Kind: A1
- Application number: 11/758,167
- Filed: 2007-06-05
- Published: 2008-12-11
- Earliest claimed: 2007-06-05
- Assignees: International Business Machines Corporation
- Inventors: Dong Chen; Mark E. Giampapa; Philip Heidelberger; Sameer Kumar; Jeffrey J. Parker; Burkhard D. Steinmacher-Burow; Pavlos Vranas
- Web page: https://patentgrep.com/patent/US20080307121A1

## CPC

- G06F 13/28: using burst mode transfer, e.g. direct memory access {DMA}, cycle steal (G06F13/32 takes precedence)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Methods, compute nodes, and computer program products are provided for direct memory access (‘DMA’) transfer completion notification. Embodiments include determining, by an origin DMA engine on an origin compute node, whether a data descriptor for an application message to be sent to a target compute node is currently in an injection first-in-first-out (‘FIFO’) buffer in dependence upon a sequence number previously associated with the data descriptor, the total number of descriptors currently in the injection FIFO buffer, and the current sequence number for the newest data descriptor stored in the injection FIFO buffer; and notifying a processor core on the origin DMA engine that the message has been sent if the data descriptor for the message is not currently in the injection FIFO buffer.

## Claims

### Claim 1

A computer-implemented method for direct memory access (‘DMA’) transfer completion notification, the method comprising:
determining, by an origin DMA engine on an origin compute node, whether a data descriptor for an application message to be sent to a target compute node is currently in an injection first-in-first-out (‘FIFO’) buffer in dependence upon a sequence number previously associated with the data descriptor, the total number of descriptors currently in the injection FIFO buffer, and the current sequence number for the newest data descriptor stored in the injection FIFO buffer; and
notifying a processor core on the origin compute node that the message has been sent if the data descriptor for the message is not currently in the injection FIFO buffer.

### Claim 2 (depends on 1)

The method of claim 1 wherein determining, by the origin DMA engine, whether a data descriptor for the message is currently in the injection FIFO buffer further comprises:
calculating the sequence number of the oldest data descriptor stored in the injection FIFO buffer; and
determining whether the sequence number for the data descriptor for the message is less than the sequence number of the oldest data descriptor stored in the injection FIFO buffer

### Claim 3 (depends on 1)

The method of claim 1 further comprising:
storing, by the DMA engine, the data descriptor for the message in the injection FIFO buffer; and
assigning, by the DMA engine, an injection FIFO buffer sequence number to the data descriptor.

### Claim 4 (depends on 3)

The method of claim 3 wherein assigning, by the DMA engine, an injection FIFO buffer sequence number to the data descriptor further comprises storing the injection FIFO sequence number for the data descriptor in application storage allocated to an application sending the message.

### Claim 5 (depends on 1)

The method of claim 1 wherein the data descriptor specifies a shared DMA counter; and
notifying a processor core on the origin DMA engine that the message has been sent further comprises disregarding a value for the shared DMA counter.

### Claim 6 (depends on 1)

The method of claim 1 the origin compute node and the target compute node are comprised in a parallel computer, the parallel computer comprising a plurality of compute nodes connected for data communications through a data communications network, the data communications network optimized for point to point data communications.

### Claim 7

An origin compute node capable of direct memory access (‘DMA’) transfer completion notification, the compute node comprising a direct memory access (‘DMA’) controller, an origin DMA engine installed upon the DMA controller, a computer processor, computer memory operatively coupled to the computer processor, the DMA controller, and the origin DMA engine, the computer memory having disposed within it computer program instructions capable of:
determining, by an origin DMA engine on an origin compute node, whether a data descriptor for an application message to be sent to a target compute node is currently in an injection first-in-first-out (‘FIFO’) buffer in dependence upon a sequence number previously associated with the data descriptor, the total number of descriptors currently in the injection FIFO buffer, and the current sequence number for the newest data descriptor stored in the injection FIFO buffer; and
notifying a processor core on the origin compute node that the message has been sent if the data descriptor for the message is not currently in the injection FIFO buffer.

### Claim 8 (depends on 7)

The compute node of claim 7 wherein computer program instructions capable of determining, by the origin DMA engine, whether a data descriptor for the message is currently in the injection FIFO buffer further comprise computer program instructions capable of:
calculating the sequence number of the oldest data descriptor stored in the injection FIFO buffer; and
determining whether the sequence number for the data descriptor for the message is less than the sequence number of the oldest data descriptor stored in the injection FIFO buffer.

### Claim 9 (depends on 7)

The compute node of claim 7 wherein the computer memory also has disposed within it computer program instructions capable of:
storing, by the DMA engine, the data descriptor for the message in the injection FIFO buffer; and
assigning, by the DMA engine, an injection FIFO buffer sequence number to the data descriptor.

### Claim 10 (depends on 9)

The compute node of claim 9 wherein computer program instructions capable of assigning, by the DMA engine, an injection FIFO buffer sequence number to the data descriptor further comprise computer program instructions capable of storing the injection FIFO sequence number for the data descriptor in application storage allocated to an application sending the message.

### Claim 11 (depends on 7)

The compute node of claim 7 wherein the data descriptor specifies a shared DMA counter; and
computer program instructions capable of notifying a processor core on the origin DMA engine that the message has been sent further comprise computer program instructions capable of disregarding a value for the shared DMA counter.

### Claim 12 (depends on 7)

The compute node of claim 7, wherein the origin compute node is comprised in a parallel computer, the parallel computer comprising a plurality of compute nodes connected for data communications through a data communications network, the data communications network optimized for point to point data communications.

### Claim 13

A computer program product for direct memory access (‘DMA’) transfer completion notification, the computer program product disposed upon a computer readable medium, the computer program product comprising computer program instructions capable of:
determining, by an origin DMA engine on an origin compute node, whether a data descriptor for an application message to be sent to a target compute node is currently in an injection first-in-first-out (‘FIFO’) buffer in dependence upon a sequence number previously associated with the data descriptor, the total number of descriptors currently in the injection FIFO buffer, and the current sequence number for the newest data descriptor stored in the injection FIFO buffer; and
notifying a processor core on the origin compute node that the message has been sent if the data descriptor for the message is not currently in the injection FIFO buffer.

### Claim 14 (depends on 13)

The computer program product of claim 13 wherein computer program instructions capable of determining, by the origin DMA engine, whether a data descriptor for the message is currently in the injection FIFO buffer further comprise computer program instructions capable of:
calculating the sequence number of the oldest data descriptor stored in the injection FIFO buffer; and
determining whether the sequence number for the data descriptor for the message is less than the sequence number of the oldest data descriptor stored in the injection FIFO buffer.

### Claim 15 (depends on 13)

The computer program product of claim 13 further comprising computer program instructions capable of:
storing, by the DMA engine, the data descriptor for the message in the injection FIFO buffer; and
assigning, by the DMA engine, an injection FIFO buffer sequence number to the data descriptor.

### Claim 16 (depends on 15)

The computer program product of claim 15 wherein computer program instructions capable of assigning, by the DMA engine, an injection FIFO buffer sequence number to the data descriptor further comprise computer program instructions capable of storing the injection FIFO sequence number for the data descriptor in application storage allocated to an application sending the message.

### Claim 17 (depends on 13)

The computer program product of claim 13 wherein the data descriptor specifies a shared DMA counter; and
computer program instructions capable of notifying a processor core on the origin DMA engine that the message has been sent further comprise computer program instructions capable of disregarding a value for the shared DMA counter.

### Claim 18 (depends on 13)

The computer program product of claim 13 wherein the computer readable medium comprises a recordable medium.

### Claim 19 (depends on 13)

The computer program product of claim 13 wherein the computer readable medium comprises a transmission medium.
