spatial LM test on 2SLS residuals

Hello everyone,
I was wondering if it is correct (and possible) to carry an LM test on the residuals of a 2SLS estimated model for spatial dependence.
Thanks.
girogioc