# US 2019/0171466 A1 · Method and system for multiple embedded device links in a host executable

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

- Number: US 2019/0171466 A1
- Document ID: US20190171466A1
- Status: Published application
- Kind: A1
- Application number: 16/268,106
- Filed: 2019-02-05
- Published: 2019-06-06
- Earliest claimed: 2012-05-09
- Assignees: NVIDIA Corporation
- Applicants: Nvidia Corporation
- Inventors: Jaydeep Marathe; Michael Murphy; Sean Y. Lee
- Web page: https://patentgrep.com/patent/US20190171466A1

## CPC

- G06F 9/44521: {Dynamic linking or loading; Link editing at or after load time, e.g. Java class loading}
- G06F 8/41: Compilation
- G06F 8/54: Link editing before load time
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Embodiments of the present invention provide a novel solution to generate multiple linked device code portions within a final executable file. Embodiments of the present invention are operable to extract device code from their respective host object filesets and then link them together to form multiple linked device code portions. Also, using the identification process described by embodiments of the present invention, device code embedded within host objects may also be uniquely identified and linked in accordance with the protocols of conventional programming languages. Furthermore, these multiple linked device code portions may be then converted into distinct executable forms of code that may be encapsulated within a single executable file.

## Claims

### Claim 1

A computer-implemented method, comprising:
linking a plurality of compiled device code portions to produce a plurality of linked device code portions, each compiled device code portion of the plurality of compiled device code portions corresponding to a compiled host code portion of a plurality of compiled host code portions;
generating an executable file comprising an executable form of the plurality of compiled host code portions and the plurality of linked device code portions; and
accessing the executable file for execution using different processor types.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1, wherein the different processor types comprise a central processor type and a graphics processor type.

### Claim 3 (depends on 1)

The computer-implemented method of claim 1, further comprising assigning a unique identifier to each compiled device code portion of the plurality of compiled device code portions.

### Claim 4 (depends on 3)

The computer-implemented method of claim 3, wherein the assigning further comprises using the unique identifier to prevent each compiled device code portion of the plurality of compiled device code portions from being linked more than once.

### Claim 5 (depends on 1)

The computer-implemented method of claim 1, wherein the plurality of compiled host code portions comprise instructions to be executed by a central processing unit (CPU) and the plurality of compiled device code portions comprise instructions to be executed by a graphics processing unit (GPU).

### Claim 6 (depends on 1)

The computer-implemented method of claim 1, wherein the plurality of compiled device code portions are written in a version of a Compute Unified Device Architecture programming language (CUDA).

### Claim 7 (depends on 1)

The computer-implemented method of claim 1, wherein each compiled device code portion of the plurality of compiled device code portions and the corresponding compiled host code portion of the plurality of complied host code portions are within a host object fileset.

### Claim 8

A system for executing code, the system comprising:
a linking module to link a plurality of compiled device code portions to produce a plurality of linked device code portions, each compiled device code portion of the plurality of compiled device code portions corresponding to a compiled host code portion of a plurality of compiled host code portions;
an executable file generation module to generate an executable file comprising an executable form of the plurality of compiled host code portions and the plurality of linked device code portions; and
at least two different processor types to execute the executable file.

### Claim 9 (depends on 8)

The system of claim 8, wherein the at least two different processor types comprise a central processor type and a graphics processor type.

### Claim 10 (depends on 8)

The system of claim 8, further comprising an identification module to assign a unique identifier to each compiled device code portion of the plurality of compiled device code portions.

### Claim 11 (depends on 10)

The system of claim 10, wherein the linking module is further operable to use the unique identifier to prevent each compiled device code portion of the plurality of compiled device code portions from being linked more than once.

### Claim 12 (depends on 8)

The system of claim 8, wherein the plurality of compiled host code portions comprise instructions to be executed by a central processing unit (CPU) and the plurality of compiled device code portions comprise instructions to be executed by a graphics processing unit (GPU).

### Claim 13 (depends on 8)

The system of claim 8, wherein the plurality of compiled device code portions are written in a version of a Compute Unified Device Architecture programming language (CUDA).

### Claim 14 (depends on 8)

The system of claim 8, wherein each compiled device code portion of the plurality of compiled device code portions and the corresponding compiled host code portion of the plurality of compiled host code portions are within a host object fileset.

### Claim 15

A non-transitory computer-readable storage medium including instructions to execute code, the instructions when executed by at least one processor of a computing device causing the computing device to:
link a plurality of compiled device code portions to produce a plurality of linked device code portions, each compiled device code portion of the plurality of compiled device code portions corresponding to a compiled host code portion of a plurality of compiled host code portions;
generate an executable file comprising an executable form of the plurality of compiled host code portions and the plurality of linked device code portions; and
access the executable file for execution using different processor types.

### Claim 16 (depends on 15)

The non-transitory computer-readable storage medium of claim 15, wherein the different processor types comprise a central processor type and a graphics processor type.

### Claim 17 (depends on 15)

The non-transitory computer-readable storage medium of claim 15, wherein accessing further comprises assigning a unique identifier to each device code portion of the plurality of device code portions.

### Claim 18 (depends on 17)

The non-transitory computer-readable storage medium of claim 17, further comprising instructions when executed by the at least one processor of the computing device causing the computing device to assign a unique identifier to each compiled device code portion of the plurality of compiled device code portions.

### Claim 19 (depends on 15)

The non-transitory computer-readable storage medium of claim 15, wherein the plurality of compiled device code portions comprise instructions to be executed by a graphics processing unit (GPU).

### Claim 20 (depends on 15)

The non-transitory computer-readable storage medium of claim 15, wherein the plurality of compiled device code portions are written in a version of a Compute Unified Device Architecture programming language (CUDA).
