[llvm-branch-commits] [llvm] release/23.x: [AArch64][SVE] Support lowering masked loads/stores of <4 x bf16> and <8 x bf16> (#208744) (PR #210497)

[Tobias Hieta via llvm-branch-commits]

https://github.com/tru approved this pull request.

Approving — this fixes a genuine crash (masked store of <8 x bf16> lowering, upstream #201149, confirmed reproducible ICE), not just a codegen improvement as first framed. Narrow, single-target (AArch64/SVE) change with two new regression tests directly exercising the fix. Clears the rc3+ bug-fix bar.

https://github.com/llvm/llvm-project/pull/210497