[llvm] [AMDGPU] Add missing intrinsic declaration to intrinsics.ll. NFC. (PR #138954)

[Matt Arsenault via llvm-commits]

================
@@ -329,6 +329,7 @@ bb:
 
 declare <4 x float> @llvm.amdgcn.mfma.f32.16x16x32.f16(<8 x half>, <8 x half>, <4 x float>, i32 immarg, i32 immarg, i32 immarg)
 declare <16 x float> @llvm.amdgcn.mfma.f32.32x32x16.f16(<8 x half>, <8 x half>, <16 x float>, i32 immarg, i32 immarg, i32 immarg)
+declare <16 x float> @llvm.amdgcn.mfma.f32.32x32x16.bf16(<8 x bfloat>, <8 x bfloat>, <16 x float>, i32 immarg, i32 immarg, i32 immarg)
----------------
arsenm wrote:

I don't like omitting them. It breaks autoupgrade on future intrinsic changes, and it's a handy reference for what the arguments actually are 

https://github.com/llvm/llvm-project/pull/138954