Is there a known DynamicC 9.62 floating point bug?

I’m working on a program that contains somewhat intensive floating point math calculations. In a nutshell, I can compile and run the program using DynamicC 7.33TSE and all is well. When I compile it with 9.62 the program either locks up during the calculations or calculates wrong numbers. Exactly the same source file works consistently in 7.33. Unfortunately I need to include SNMP support in the program which 7.33 does not provide. Any ideas?

Any known bugs are patched using the accumulated patched libraries at http://www.digi.com/support/productdetail?pid=5053&type=documentation under Development Software and Sample Code.