mirror of
https://github.com/llvm/llvm-project.git
synced 2025-04-26 10:16:07 +00:00
dc3258c617
This op is the inverse of all-gather. It is useful to have an explicit concise representation instead of having a blob of slicing logic. Add lowering for the op that slices from the tensor based on the in-group process index. Make resharding generate an all-slice instead of inserting the slicing logic directly.