CodeGuru Home VC++ / MFC / C++ .NET / C# Visual Basic VB Forums Developer.com
Results 1 to 5 of 5
  1. #1
    Join Date
    Sep 2002
    Location
    Singapore
    Posts
    673

    [RESOLVED] When is 80bit long double coming back?

    80bit long double support is removed due to Windows NT is running on non-x86 platforms which does not have 80bit floating point.

    Since, now Windows is running mostly on x86 and x64, what about having 80bit long double back in Visual C++?

    Thanks in advance for answering my question!

  2. #2
    Join Date
    May 2006
    Posts
    17

    Re: When is 80bit long double coming back?

    Hi,

    There are no current plans to add 80 bit FP support.

    The major reason is that FP code generation has been switching to the use of SSE/SSE2/SSE3 instruction sets instead of the x87 FP stack since that is what both the AMD and Intel recent and future chip generations are focusing their performance efforts on. These instruction sets only support 32 and 64 bit FP formats.

    Ronald Laeremans
    Acting Product Unit Manager
    Visual C++ Team

  3. #3
    Join Date
    Sep 2002
    Location
    Singapore
    Posts
    673

    Re: When is 80bit long double coming back?

    If both long double and double are 64bit, is there a need to make them 2 incompatible types?

  4. #4
    Join Date
    May 2006
    Posts
    17

    Re: When is 80bit long double coming back?

    Yes, just like long and int are distinct types although they happen to map to the same machine type on both 32 and 64 bit Windows platforms.

    Doing it differently would hurt cross platform portability of C++ code now and would make it harder to cleanly support extended precision in the future if and when it becomes critical (say if HW architectures commonly support 128 bit FP types and algorithms that require that precision become relevant).

    Ronald

  5. #5
    Join Date
    Jun 2002
    Location
    Germany
    Posts
    1,557

    Re: [RESOLVED] When is 80bit long double coming back?

    I don't think legacy support for the exotic 80-bit floating point makes any sense.

    However, I would love to see...

    - Native compiler support for 128-bit floating point representations in C/C++.

    - As well as native compiler support INT128 / UINT128 in C / C++.

    The SSE2 instruction coding for the integer operations is rather straight forward.

    However the 128-bit floating point is difficult to specify and much more difficult to implement in <cmath>. However it's not as bad as you might think once you've got the FP representation and the basic add / sub / mul / div.

    Any plans here.

    Chris.
    You're gonna go blind staring into that box all day.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  





Click Here to Expand Forum to Full Width

Featured