mirror of
https://github.com/intel/llvm.git
synced 2026-02-05 04:46:27 +08:00
Roll forward of commit 5684a12.
When outlining GPU kernels, put the kernel function inside a nested module. Then use a nested pipeline to generate the cubins, independently per kernel. In a final pass, move the cubins back to the parent module.
PiperOrigin-RevId: 270639748