mirror of
https://github.com/AdaCore/cuda.git
synced 2026-02-12 13:05:54 -08:00
ec6f9f15d9
The example now makes use of the new device-cuda runtime. The project file has also been split into seperate device and host projects because the host code depends on the opengl project which is incompatible with the device-cuda runtime, and I don't believe there's an easy way to use scenario variables to work around that. Beside, the device code is conceptionally different code anyway; ideally we could find a way to to make the device project a subproject of the host project. Source code has also been moved around to give the example a bit more structure. TN: U114-058
6 lines
219 B
Bash
Executable File
6 lines
219 B
Bash
Executable File
#!/bin/sh -e
|
|
|
|
Execution_Side=Device gprbuild -P ../../api/cuda_device.gpr -j0
|
|
Execution_Side=Device gprbuild -P marching_device_code.gpr -j0
|
|
Execution_Side=Host gprbuild -P marching_cubes -j0 -largs $PWD/obj/*.fatbin.o
|