Internal validation with small sample in R

Question

I am doing a prediction task with a small sample size (300 points with 18 predictors) using R.
Following Frank Harrel's suggestion, I would like to go with bootstrapping.
Below is my understanding of the procedure.

It uses bootstrap only once to generate the training data.

1. Create a bootstrapped sample that has the same size as the original sample.
2. Use the bootstrapped sample as a training set and validate with the original sample.
3. Average performance over +1000 repetitions.

So I am using a simple function, sample(data, replicate = TRUE), to accomplish the first step. Is there anything I am missing in the procedure, for ex., another bootstrapping needs to be done or the kind of bootstrap?

Thanks.