I know this question might seem overly familiar to the community, but I swear I've never been able to reproduce the issue related to this question even once throughout my programming journey.
I understand what the strictfp modifier does and how it ensures full compliance with the IEEE754 standard. However, I've never encountered a situation in practice where the set of values with an extended exponent is used, as described in the official specification.
I've tried using options like -XX:+UseFPUForSpilling to stimulate the use of the FPU block for calculations on my relatively modern processor, but it had no effect.
I even went as far as installing Windows 98 SE on a virtual machine and emulating an Intel Pentium II processor through Bochs, which does not support the SSE instruction set, hoping that the use of the FPU block in this case would be virtually the only option. However, even such an experiment yielded no results.
The essence of the experiment was to take the maximum possible value of the double type and multiply it by 2 to take the intermediate result beyond the permissible range of the double type. Then, I divided the obtained value by 4, and the final result was saved back into a double variable. In theory, I should have gotten some more meaningful result, but in all situations, I ended up with Infinity. In general, I haven't found a single reproducible example on the entire internet (even as of 2024!) that would show different results with and without the use of strictfp. Is it really possible that in almost 30 years of the language's history, there isn't a single example on this topic that clearly demonstrates the difference?
P.S. I'm well aware of Java 17+. All experiments were conducted on earlier versions, where the difference should, in theory, be observable. I installed Java SE 1.3 on the virtual machine.