chisq.test - Kolmogorov-Smirnov test for goodness of fit

How can I test with chisq.test the Null hypothesis that "the sample comes from an exp(1) distribution" against the alternative hypothesis "the sample do not comes from an exp(1) distribution? I wrote the following code in R, is this right?
> observed<-z # I generated m = 500 random values from exp(1) distribution
> prob.exp<-dexp(z)
> prob.exp<-dexp(z,rate=1)
> chisq.test(observed,p=prob.exp,rescale.p=TRUE)

Chi-squared test for given probabilities

data: observed
X-squared = 40909, df = 499, p-value < 2.2e-16

Warning message:
In chisq.test(observed, p = prob.exp, rescale.p = TRUE) :
Chi-squared approximation may be incorrect
How can I test the same hypothesis with Kolmogorov-Smirnov test?