I want to take advantage of the hardware FPU on the T20 with Windows CE 6. This article in the knowledge base seems to be outdated: Colibri Floating Point Unit | Toradex Developer Center
The article refers to disabling “floating point emulation” - but this compiler option doesn’t exit any more. Is it enough now to add the “QRfpe-” switch alone?
Thanks a lot and greetings from Franconia!
Yes, it should be fine. This is probably removed due to an issue in VS2008, the switch was never passed to the compiler, so it was probably removed by Microsoft and you now have to provide the switch manually. You details about this bug here: https://support.microsoft.com/en-us/kb/947894
Samuel, thanks for your quick reply to my question!
Using the referenced benchmark from here I was able to check the compiler settings for their effectiveness. In order to benefit from the FPU unit of the T20, you have to provde the “QRfpe-” switch to the compiler settings. Running the benchmark results in ca. 313 ms with the standard “Release” settings and ca. 79 ms when using the “QRfpe-” switch. That’s about the same numbers Toradex reports in their original knowledge base article here.