forked from OSchip/llvm-project
f57b2420b2
Depends On D104999 Automatic reference counting based on the liveness analysis can add a lot of reference counting overhead at runtime. If the IR is known to be constrained to few particular "shapes", it's much more efficient to provide a custom reference counting policy that will specify where it is required to update the async value reference count. Reviewed By: mehdi_amini Differential Revision: https://reviews.llvm.org/D105037 |
||
---|---|---|
.. | ||
async-parallel-for-async-dispatch.mlir | ||
async-parallel-for-canonicalize.mlir | ||
async-parallel-for-seq-dispatch.mlir | ||
async-runtime-policy-based-ref-counting.mlir | ||
async-runtime-ref-counting-opt.mlir | ||
async-runtime-ref-counting.mlir | ||
async-to-async-runtime.mlir | ||
coro.mlir | ||
ops.mlir | ||
runtime.mlir | ||
verify.mlir |