Invalid constant in WmmaOpsToNvvm lowering

[adam-smnk]

gpu.subgroup_mma_elementwise maxf lowering to nvvm dialect creates invalid constant:
%0 = llvm.mlir.constant(0x7E00 : f16) : vector<2xf16>
which prevents GPU binary generation with:
error: FloatAttr does not match expected type of the constant
The constant value has to be dense and match the result type as:
%0 = llvm.mlir.constant(dense<0x7E00> : vector<2xf16>) : vector<2xf16>

To be fixed in the upstream WmmaOpsToNvvm.cpp function createMinMaxF.