Page 1 of 3 123 LastLast
Results 1 to 10 of 42

Thread: Intel Is Trying To Support The x32 ABI For LLVM/Clang

Hybrid View

  1. #1
    Join Date
    Jan 2007
    Posts
    15,438

    Default Intel Is Trying To Support The x32 ABI For LLVM/Clang

    Phoronix: Intel Is Trying To Support The x32 ABI For LLVM/Clang

    While adoption of the Linux x32 ABI hasn't really taken off with most developers and end-users doing just fine with x86_64-compiled software, Intel is trying to get things back on track for supporting x32 by LLVM and Clang...

    http://www.phoronix.com/vr.php?view=MTczMjg

  2. #2
    Join Date
    Dec 2010
    Location
    MA, USA
    Posts
    1,447

    Default

    Why? Seriously intel, just drop your 32 bit platforms already. This is getting old.

  3. #3
    Join Date
    Feb 2009
    Posts
    165

    Default

    Quote Originally Posted by schmidtbag View Post
    Why? Seriously intel, just drop your 32 bit platforms already. This is getting old.
    Please, if you don't know what you're talking about, just don't talk about it... If you had even read the article you would know it's not about 32 bit platforms...

  4. #4
    Join Date
    Dec 2010
    Location
    MA, USA
    Posts
    1,447

    Default

    Quote Originally Posted by mdias View Post
    Please, if you don't know what you're talking about, just don't talk about it... If you had even read the article you would know it's not about 32 bit platforms...
    "The Linux x32 ABI about allowing programs to take advantage of x86_64 features (and being dependent upon x86 64-bit CPUs) while using 32-bit memory pointers."

    To me, this sounds like trying to use 64 bit features on 32 bit architectures. While it doesn't explicitly say this is for 32 bit architectures, I feel that is the intended use in the long run. While I rarely mess with languages like C/C++, I don't see why (if you're on a 64 bit system) to use 32-bit pointers expecting them to use x86-64 features when you could just use 64-bit pointers. If what I'm saying is wrong then sure, it might not have been great to presume the worst, but it still doesn't change the fact that intel has been dragging 32 bit products even into 2013 and I'm getting tired of it.

  5. #5
    Join Date
    Sep 2012
    Posts
    296

    Default

    Quote Originally Posted by schmidtbag View Post
    "The Linux x32 ABI about allowing programs to take advantage of x86_64 features (and being dependent upon x86 64-bit CPUs) while using 32-bit memory pointers."

    To me, this sounds like trying to use 64 bit features on 32 bit architectures. While it doesn't explicitly say this is for 32 bit architectures, I feel that is the intended use in the long run. While I rarely mess with languages like C/C++, I don't see why (if you're on a 64 bit system) to use 32-bit pointers expecting them to use x86-64 features when you could just use 64-bit pointers. If what I'm saying is wrong then sure, it might not have been great to presume the worst, but it still doesn't change the fact that intel has been dragging 32 bit products even into 2013 and I'm getting tired of it.
    Exactly. I don't see any reason to keep using 32 bit architectures either. Not to mention that if it was not for AMD we would still be using 32-bit CPUs on the desktop and 64-bit only on servers ($$$$).

  6. #6
    Join Date
    Jul 2013
    Location
    Bordeaux, France
    Posts
    320

    Default

    Quote Originally Posted by wargames View Post
    Exactly. I don't see any reason to keep using 32 bit architectures either. .
    But x32 REQUIRES A 64 BIT ARCHITECTURE.

    x32 != x86

  7. #7
    Join Date
    Jul 2013
    Location
    Bordeaux, France
    Posts
    320

    Default

    Quote Originally Posted by schmidtbag View Post
    I don't see why (if you're on a 64 bit system) to use 32-bit pointers expecting them to use x86-64 features when you could just use 64-bit pointers.
    Since you have 64 bit of bandwith you can transfer two 32 bitspointers at a time which might make stuff faster.

  8. #8
    Join Date
    Dec 2010
    Location
    MA, USA
    Posts
    1,447

    Default

    Quote Originally Posted by doom_Oo7 View Post
    Since you have 64 bit of bandwith you can transfer two 32 bitspointers at a time which might make stuff faster.
    Well in that case, this is awesome.

  9. #9

    Default

    Quote Originally Posted by doom_Oo7 View Post
    Since you have 64 bit of bandwith you can transfer two 32 bitspointers at a time which might make stuff faster.
    In what real wold secenarios wouls be of bennifit over straight 64 bit? If I'm doing something that is that taxing to the point it makes a noticeable difference it probably gains more from full 64 bit anyways.

  10. #10
    Join Date
    Dec 2012
    Posts
    459

    Default

    Quote Originally Posted by schmidtbag View Post
    "The Linux x32 ABI about allowing programs to take advantage of x86_64 features (and being dependent upon x86 64-bit CPUs) while using 32-bit memory pointers."
    Quote Originally Posted by schmidtbag View Post
    To me, this sounds like trying to use 64 bit features on 32 bit architectures. While it doesn't explicitly say this is for 32 bit architectures, I feel that is the intended use in the long run. While I rarely mess with languages like C/C++, I don't see why (if you're on a 64 bit system) to use 32-bit pointers expecting them to use x86-64 features when you could just use 64-bit pointers.
    For some cases, the 32bit vs 64bit has some drawbacks.

    Quote Originally Posted by schmidtbag View Post
    If what I'm saying is wrong then sure, it might not have been great to presume the worst, but it still doesn't change the fact that intel has been dragging 32 bit products even into 2013 and I'm getting tired of it.
    x32 is not 32bit. You just mentioned this yourself.

Posting Permissions

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