Having many triggers still hits a bottleneck in LLVM leading to long compile times. Instead of setting triggers bit-wise, set them as a whole 64-bit word when possible. This improves C++ compile times by ~4x on some large designs and has minor run-time performance benefit. |
||
|---|---|---|
| .. | ||
| t | ||
| .gdbinit | ||
| .gitignore | ||
| CMakeLists.txt | ||
| Makefile | ||
| Makefile_obj | ||
| driver.py | ||
| input.vc | ||
| input.xsim.vc | ||