Announcement

Collapse
No announcement yet.

The Project To Compile The Linux Kernel With LLVM Clang Is Still Alive

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • The Project To Compile The Linux Kernel With LLVM Clang Is Still Alive

    Phoronix: The Project To Compile The Linux Kernel With LLVM Clang Is Still Alive

    While for years there has been ongoing work to build the Linux kernel with Clang, in 2015 there wasn't much progress to report and the mainline LLVM Clang compiler still can't build the mainline Linux kernel tree successfully. What's going on?..

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Clang and LTO efforts would be greatly appreciated by embedded users.

    Comment


    • #3
      Is this the year of the <strike>Linux Desktop</strike> Clang kernel?

      Comment


      • #4
        I love to see competitions. I hope we'll have a LLVM Clang-compiled kernel soon. Then there'll be a post on phoronix comparing clang-compiled kernel and gcc-compiled kernel.

        Comment


        • #5
          Originally posted by zman0900 View Post
          Is this the year of the <strike>Linux Desktop</strike> Clang kernel?
          I'm not holding my breath, but I hope so. Then all that would be left to shut Stallman fans up about that "GNU/Linux" name (too impractical and inaccurate to ever happen for a laundry list of reasons) is seeing BusyBox+musl mature into something closed-source apps get along with more consistently.

          Comment


          • #6
            What is the work that needs to be done?
            Is it kernel code that needs changes? Is it llvm clang changes? Or something else?

            Comment


            • #7
              Originally posted by ArthurBorsboom View Post
              What is the work that needs to be done?
              Is it kernel code that needs changes? Is it llvm clang changes? Or something else?
              Don't expect to find answers here (at least from the article). Some forthcoming voluntary comments may or may not shed some light on this.

              Comment


              • #8
                Originally posted by ArthurBorsboom View Post
                What is the work that needs to be done?
                Is it kernel code that needs changes? Is it llvm clang changes? Or something else?
                The last status report i saw a year or two ago went something like this:

                Person A: The kernel won't build because of these 20 places that use GCC specific behaviour. We need to either fix the kernel, or add that into LLVM so it works.
                Person B: We don't want to add that GCC specific stuff into LLVM. You should port the kernel.
                Person C: We don't want to change the kernel, that stuff is useful. You should fix LLVM.
                Person A: Ok, we're just going to maintain these patches ourselves and forget about merging them upstream.

                Comment


                • #9
                  Originally posted by smitty3268 View Post

                  The last status report i saw a year or two ago went something like this:

                  Person A: The kernel won't build because of these 20 places that use GCC specific behaviour. We need to either fix the kernel, or add that into LLVM so it works.
                  Person B: We don't want to add that GCC specific stuff into LLVM. You should port the kernel.
                  Person C: We don't want to change the kernel, that stuff is useful. You should fix LLVM.
                  Person A: Ok, we're just going to maintain these patches ourselves and forget about merging them upstream.

                  Pretty much this.

                  Granted, no matter how useful, compiler specific behavior doesn't belong in the kernel. So I side with LLVM in this instance; the kernel should really be free of any platform limiting behavior, period.

                  Comment


                  • #10
                    Originally posted by smitty3268 View Post

                    The last status report i saw a year or two ago went something like this:

                    Person A: The kernel won't build because of these 20 places that use GCC specific behaviour. We need to either fix the kernel, or add that into LLVM so it works.
                    Person B: We don't want to add that GCC specific stuff into LLVM. You should port the kernel.
                    Person C: We don't want to change the kernel, that stuff is useful. You should fix LLVM.
                    Person A: Ok, we're just going to maintain these patches ourselves and forget about merging them upstream.
                    Could you be more specific about those "20 places" where Linux kernel is unavoidably gcc-dependant ?

                    Comment

                    Working...
                    X