You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Expected outcome:
After PTX module have been compiled once, it gets cached on disk in an SQLite database. Before kernel gets compiled we look in cache to check if the kernel has been previously compiled
Comments:
Key should be hash (BLAKE3) of module text + compiler version + ZLUDA version + device (gfxXXXX) + flags (debug/release, windows/linux, compiler switches)
We will want to eventually support mechanism similar to CUDA_CACHE_MAXSIZE. This does not have to be implemented yet, but db should at least contain necessary information for evicting cache: time of last use for each entry and total size of all the kernels
You can get compile version from comgr by running a preprocessor on a file containing preprocessor directive that resolves to full clang version. It contains version and ROCm LLVM hash
ZLUDA version should be current git commit hash. There are several crates for it, last time I used vergen
The text was updated successfully, but these errors were encountered:
Expected outcome:
After PTX module have been compiled once, it gets cached on disk in an SQLite database. Before kernel gets compiled we look in cache to check if the kernel has been previously compiled
Comments:
CUDA_CACHE_MAXSIZE
. This does not have to be implemented yet, but db should at least contain necessary information for evicting cache: time of last use for each entry and total size of all the kernelsThe text was updated successfully, but these errors were encountered: