[llvm] AMDGPU: Fix packed 16-bit inline constants (PR #76522)

[Stanislav Mekhanoshin via llvm-commits]

Nicolai =?utf-8?q?Hähnle?= <nicolai.haehnle at amd.com>
Message-ID:
In-Reply-To: <llvm.org/llvm/llvm-project/pull/76522 at github.com>


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

LGTM. One thing left behind is BF16 constants. These are FP32 with cleared low 16 bits and can mostly use I16 with this FP immediates. However, using I16 does not allow FP modifiers on the operands. We will probably need to create new operand types for that eventually. But that is a pre-existing problem.

https://github.com/llvm/llvm-project/pull/76522