LinAlg Validation: MatVecMulAdd

[V-FEXrt]

Per the spec:

Validation will enforce that:

• The input vector length matches the K matrix dimension
• The bias vector length matches the M matrix dimension
• The matrix A is an A matrix of Thread scope
• The input and bias interpretation type must be one of the valid linalg component types specified in the list in the LinAlg Component Types section.
• The sign bit for output types should always be true if the output type is a vector of native floating point types.