MIPSfpga Linux soft-float support?

This topic contains 3 replies, has 2 voices, and was last updated by  ZubairLK 1 year, 8 months ago.

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
    Posts
  • #52824

    Chun-Kai
    Member

    Hi all,

    Recently I’m working on running some benchmarking programs on MIPSfpga Linux Soc package. As the datasheet of the package stated the MIPSfpga does not have a Floating-point unit. Also I noticed that only hard-float libraries are provided in the mips-mti-linux-gnu toolchains. What should I do for compiling the code which contains floating point calculations in -msoft-float mode in order to run them in soft float mode?

    Thanks in advance.

    #52829

    ZubairLK
    Member

    Hi Kyle,

    I understand that you are trying to cross compile programs to run on Linux on MIPSfpga.

    And these programs require soft-float libraries.

    You could try using the mentor toolchain for this purpose.

    https://sourcery.mentor.com/GNUToolchain/release3133?@template=datasheet

    Regards,
    ZubairLK

    #52853

    Chun-Kai
    Member

    Hi ZubairLK

    I have tried the mentor toolchain and successfully compiled my program. However I was wondering if it was possible to add soft-float library into codescape sdk toolchain manually or not? Do I need to use Linux PC to do this?

    Regards,
    Kyle

    #52855

    ZubairLK
    Member

    Hi,

    Unfortunately it isn’t a simple matter of enabling a flag.

    Future versions of the codescape tool-chain will support soft-float libraries. But I can’t give any specific time line.

    Regards,
    ZubairLK

Viewing 4 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic.