[PATCH]: Fix some fortran internal error messages

[Jerry DeLisle]

On 09/24/2009 06:37 PM, Kaveh R. GHAZI wrote:
> While reading gfortran code, I noticed that some of the error messages
> were bogus.  Many of the math functions erroneously claim to be cos,
> likely as a result of overly aggressive cut-n-paste'ing.
>
> Tested on x86_64-unknown-linux-gnu, no regressions.  However I expect
> nothing in the testsuite actually triggers these as they should be
> impossible conditions.
>
> Okay for mainline?
>
Yes, OK, thanks for catching these.

Jerry