If the following question is irrelevant to the current forum, please direct me to the appropriate forum:

I'm using version 2.07 of NR in FORTRAN. I'm using the Newton's method for multidimensional root-finding. And for a particular system of equations (there's nothing really odd about this system), the code refuses to compile. The error reported by the compiler is F1001 : "Internal compiler error".

Please note that I get no error message for many other (and more complicated) systems of equations I solved.

The error does not seem to be related to NR, but it might be. I use Windows XP.

Any information about this particular error message, and why it possibly occurred, is welcome. The "Help" file in FORTRAN has not proved to be of much use.

Hi, sachin_iit!

"Internal compiler error" is really bad news. It probably has nothing to do with the particular set of equations, but with some compiler bug triggered by (e.g.) the length or structure of some particular expression, or some particular optimization opportunity that the compiler thinks it can handle and then bombs on.

It could also be a missing piece of the compiler (e.g., a DLL or other run-time component). You might try de- and then re-installing the whole compiler; but this might also be a wild goose chase.

Good luck!

Thanks. You seem to be correct in suggesting an un-install and re-install of the compiler. Because, I just tried to run the 'same' file on another computer. And it worked fine!!

I'm not sure if the following piece of information might help to track why this particular error occured, but the compiler appears to get stuck on:

(compiler file //..../p2/p2symtab.c, line 1702)

Again, I'm thankful for your suggestion and will try to un- and re-install the compiler. Let's see if it works.