tab with chi2 should be fine for looking at dichotomous predictor and outcome variables.
tabodds should be fine with nominal categories. however, it should only be used with case-control or cross-sectional data, so make sure your dataset is appropriate.