Jump to content
  • Advertisement
Sign in to follow this  
Climax777

Floating point precision

This topic is 3889 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Hi I know how to change the precision part of the control word of the FPU( there's more than one way) in x86 architecture. Either using __asm fldcw or __control87. But I read in MSDN's library and observed that it cannot be set in x64 bit architecture. Is there a way to set the precision in a 64-bit application? Thanx!

Share this post


Link to post
Share on other sites
Advertisement
Even when ASM programmers could old x86 architecture programmers should use SSE2 registers for floating point computations. AMD even specifically said an emulation of old architecture could disappear without further notice. So don't use it.

IIRC either larebee, or little further CPU will reroute FPU calls into SSE2 registers even in 32 bit mode. You are talking not only about outdated technology, you are talking about technology that would be eradicated.

Not only there are no (non SSE2) FPU registers in 64 bit mode, they wanna get rid of that outdated (and wrong) methodology.

Why do you ask?

Share this post


Link to post
Share on other sites
Quote:
Original post by Raghar
Even when ASM programmers could old x86 architecture programmers should use SSE2 registers for floating point computations. AMD even specifically said an emulation of old architecture could disappear without further notice. So don't use it.

IIRC either larebee, or little further CPU will reroute FPU calls into SSE2 registers even in 32 bit mode. You are talking not only about outdated technology, you are talking about technology that would be eradicated.

Not only there are no (non SSE2) FPU registers in 64 bit mode, they wanna get rid of that outdated (and wrong) methodology.

Why do you ask?


The reason I ask: I want to do intense and precise floating point operations (extended double precision mode) while building in 64-bit mode. And currently I am forced to use the standard double precision (53 bit floating point), unless I build in X86 mode and force the FPU to do otherwise (__control87 or __asm fldcw).

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

We are the game development community.

Whether you are an indie, hobbyist, AAA developer, or just trying to learn, GameDev.net is the place for you to learn, share, and connect with the games industry. Learn more About Us or sign up!

Sign me up!