Stata margins example with weights

I'm interested in assessing the effect of weights (importance, not sampling) when computing marginal effects for a continuous predictor in model. As part of that, I would like to see the results Stata gets from the following statements, which I think/hope should run successfully. If you would be willing to run it and reply with the results (from both glm and margins), that would be appreciated.

glm price mpg weight length foreign, family(gamma) link(log) [iweight=gear_ratio]
margins, dydx(mpg)