This mostly works. Unfortunately it doesn't round the number up if the next decimal place is over 5.

e.g. 1.234 to 2dp is 1.23, whereas 1.236 should be 1.24, but this gives 1.23 still.

Code:

float round(float val, int dp)
{
  return int(val*pow(10,dp))/pow(10,dp);
} 

it should be noted that even when you 'round' a floating point number, it may not stay rounded.. floating point means that the numbers aren't stored per their decimal places, so a rounded 12.34566 to two places might come out as 12.3500071 or something like that.

the only way to truly fix the decimal places is to store things as integers, and divide them at the last minute by their decimal place.

and if you're just trying to do this for formatting text (the only way you'll get properly fixed decimal places) use the nf() function.