Announcement

Collapse
No announcement yet.

Apple Silicon / M1 Port Planned For GCC 12

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

  • perpetually high
    replied
    Originally posted by Sonadow View Post

    Indeed, inflated ego.

    You either produce code or you use someone else's code.

    Those who can only use code without producing code are lusers.
    Says the guy that uses the word "lusers"

    If I could block you, I would do it immediately. Dislike everything about you. That's how strongly I feel toward you.

    Don't reply anymore to me. Or at least, don't tag me. Don't want to see your notification.

    Leave a comment:


  • Sonadow
    replied
    Originally posted by perpetually high View Post

    I completely disagree. Power users are us.
    Indeed, inflated ego.

    You either produce code or you use someone else's code.

    Those who can only use code without producing code are lusers.

    Leave a comment:


  • perpetually high
    replied
    Originally posted by Sonadow View Post


    "Power users" are lusers with an inflated ego.

    Computing only has lusers and developers. Nothing exists in between.
    I completely disagree. Power users are us. Regular users are our parents and our friends who know jack shit about computers or even Linux. Post less.

    Let me know if you're still confused.

    Leave a comment:


  • Sonadow
    replied
    Originally posted by perpetually high View Post
    Power to the power users.

    "Power users" are lusers with an inflated ego.

    Computing only has lusers and developers. Nothing exists in between.

    Leave a comment:


  • Steffo
    replied
    Originally posted by kpedersen View Post

    I'm still not so sure. Think of all those i.e, hospital thin clients. Many still on Windows XP. Then in Russia, China and other countries where Apple is not quite so desirable.

    I will never get the stats but I am fairly sure there are more old Windows XP installs running on that era hardware than Mac installs.
    Well i386 and Windows XP have no future. Any further questions?

    Leave a comment:


  • perpetually high
    replied
    Gonna drop this here: https://openbenchmarking.org/result/...TJ-OSBENCHSE20

    This morning I ran the phoronix pts/osbench benchmark across five GCC versions (7, 8, 9, 10, 11) on my Ubuntu 20.04 LTS on the same exact kernel (5.4.147) and system settings, etc. Basically compiled, benchmarked, compiled, benchmarked, etc.

    Winner for my i5 Haswell.......drumroll......more...drumroll....... gcc 9.3.0 on Ubuntu 20.04

    Code:
    Geometric Mean > Higher Is Better
    5.4.147 gcc-7.5.0-6ubuntu2 ....... 4.615 |================================================= ================================================== ==========================
    5.4.147 gcc-11.1.0-1ubuntu120.04 . 4.632 |================================================= ================================================== ===========================
    5.4.147 gcc-8.4.0-3ubuntu2 ....... 4.653 |================================================= ================================================== ===========================
    5.4.147 gcc-10.3.0-1ubuntu120.04 . 4.676 |================================================= ================================================== ============================
    5.4.147 gcc-9.3.0-17ubuntu120.04 . 4.715 |================================================= ================================================== =============================
    I had a hunch going in, but I'm glad I finally ran my own benchmarks on my *own* system (this is important. Michael can run his, but means jack if your system doesn't exhibit the same results).

    Having said that, these are just 5 benchmarks. But for me, important indicators of overall OS performance (create processes, create threads). Sure, we're talking microperformances, but that's what we're here for baby. We're power usings. Power to the power users.

    Anyways, don't want to derail the thread. But wanted to throw this out there. Cheers.

    Leave a comment:


  • Markopolo
    replied
    Originally posted by discordian View Post

    In case someone figured "WTF", the changes are to adhere to function call conventions/ABI for interfacing with the system libraries. Has zero to do with the hardware implementation.
    That was definitely me, appreciate the clarification. Here I thought they had added a whole bunch of instructions beyond Arm v8.x or whatever they’re on.

    Leave a comment:


  • discordian
    replied
    functions and other fundamental differences, largely in the name of Apple M1 security enhancements requiring additional compiler alterations to yield working/supported code.
    In case someone figured "WTF", the changes are to adhere to function call conventions/ABI for interfacing with the system libraries. Has zero to do with the hardware implementation.

    Leave a comment:


  • stiiixy
    replied
    See the M1X news

    Leave a comment:


  • jabl
    replied
    Originally posted by kpedersen View Post

    I'm still not so sure. Think of all those i.e, hospital thin clients. Many still on Windows XP. Then in Russia, China and other countries where Apple is not quite so desirable.

    I will never get the stats but I am fairly sure there are more old Windows XP installs running on that era hardware than Mac installs.
    I'm sure there's lots of hospital thin clients, POS systems, etc. running 32-bit XP. But quite hard say they are relevant targets for GCC.

    You can see the current list of primary and secondary platforms along with criteria for them at https://gcc.gnu.org/gcc-12/criteria.html . I think there's quite a lot of historical inertia in that list. E.g. i686-apple-darwin is a secondary platform but x86_64-apple-darwin isn't, although 32-bit x86 was a rather short-lived Apple platform before they switched to 64-bit x86.

    Leave a comment:

Working...
X