Slashdot Mirror

← Back to Stories (view on slashdot.org)

GCC 3.0.2 Is Out

Posted by timothy on from the gee-see-sea dept.

anotherworld writes: "GCC 3.0.2 is out. That's good news for me, since I am having trouble compiling the 2.4.13 kernel with 3.0.1 (internal error) AND 2.96 (under investigation)... so I can try a new one :) But where is that good old 2.95.x? I just can't find it at the moment! Really, well done guys!" The site says this is mostly a bugfix release, but if you need those particular bugfixes, please use one of the mirrors. And remember, they do take appreciation is cold, hard cash, too :)

22 comments

  1. Let me get this straight... by Anonymous Coward · · Score: 0

    They repeatedly put out compilers that are breaking on your code, and you *thank* them? Are you being sarcastic? My sensors must be broken.

    1. Re:Let me get this straight... by doconnor · · Score: 1

      Seeing as they are generally not getting paid to put out these compilers, I think thanks are still in order dispite the imperfections.

    2. Re:Let me get this straight... by Anonymous Coward · · Score: 1, Insightful

      9 times out of 10, if gcc 3.x won't compile your code, its not gcc3's fault, especally if its C++ code.

      gcc3 is far better about standards than any previous release.

    3. Re:Let me get this straight... by Anonymous Coward · · Score: 0

      Exactly - a hell of a lot of C and C++ code on Linux is non-standards compliant.

    4. Re:Let me get this straight... by Anonymous Coward · · Score: 0

      Hmm. The article said:
      >That's good news for me, since I am having trouble compiling the 2.4.13 kernel with 3.0.1

      So what I'm hearing you say is that it's not
      a _gcc_ problem, it's the Linux kernel source that's dodgy. I'll drink to that...

    5. Re:Let me get this straight... by Arandir · · Score: 2

      Will it compile ISO Standard C++? That's all I care about. If it detects nonstandard code in my stuff, then I can fix my stuff.

      I can't wait for the day when a configure script says "g++ x.y found, no further compiler checks needed."

      --
      A Government Is a Body of People, Usually Notably Ungoverned
    6. Re:Let me get this straight... by Anonymous Coward · · Score: 0

      Actually, previous compiler releases would expose bugs in the Linux kernel (egcs and 2.0.x kernels, for instance), and every moron and their dog complained that egcs was buggy - but the problem was with Linux.

    7. Re:Let me get this straight... by Anonymous Coward · · Score: 0

      How does one code "non-standard" code?! Where are the books that teach it? In my view one adapts one's coding to the compiler.

  2. Using it already.... by Anonymous Coward · · Score: 0


    Compiled for Solaris 8. Rulez.

  3. 2.95.x by Ridge2001 · · Score: 3, Informative
    But where is that good old 2.95.x? I just can't find it at the moment!

    Huh?

    ftp://gcc.gnu.org/pub/gcc/releases/gcc-2.95.3

  4. KDE by Torulf · · Score: 1

    Will it compile KDE by now? Could anybody enlighten me on this?

  5. :-) by Anonymous Coward · · Score: 0

    I'm really glad that they made those bugfixes. 3.0.1 and 3.0 were horrible at compiling inline assembly (which I need in a certain application I am making), and so this release is the first serious gcc-3.0 I will use. I haven't had any problems with this compiler so far and I don't think there will be any more major problems with this.

  6. minor bugfixes by vlad_petric · · Score: 1

    gcc 3.0 needs "major bugfixes", not minor ones, especially in the c++ compiler.

    --

    The Raven

  7. Compiler that shipped with Redhat 7.0? by Fastball · · Score: 0, Flamebait

    I just did a full upgrade on my home workstation from Redhat 7.0 to 7.2. One of the main reasons was because I couldn't build a kernel from source with the compiler that shipped with 7.0 (gcc 2.96). I learned a valuable lesson: never run a .0 version of Redhat. Hopefully, the next time I attempt to rebuild my kernel, I won't run into the same problem.

    1. Re:Compiler that shipped with Redhat 7.0? by Anonymous Coward · · Score: 0

      7.2 still has 2.96, albeit a fixed version. It was fixed in 7.1 also.

  8. binary size by rv101 · · Score: 1

    From the same source code, gcc3 creates a binary three times as large as the one created with gcc2.95:

    $ make
    g++-2.95 -O3 -otexte.gcc.2.95 texte.C -L/usr/X11R6/lib -lGL -lX11
    g++-3.0 -O3 -otexte.gcc.3 texte.C -L/usr/X11R6/lib -lGL -lX11

    $ ll
    203 Makefile
    12760 texte.C
    12614 texte.gcc.2.95
    29899 texte.gcc.3

    1. Re:binary size by ChadN · · Score: 2

      Can you show us the sizes of the "stripped" binaries? Also, I just compiled my C++ program w/ 3.0.2, and it was the same size as 2.95.4 (prerelease), to within 0.1%. However, I'm not making heavy use of templates. Without seeing the source, we can only speculate as to your specific results...

      --
      "It's overkill, of course. But you can never have too much overkill." - Anonymous Slashdot Coward
    2. Re:binary size by Anonymous Coward · · Score: 0

      If you care about binary size, why are you doing O3 instead of Os? The big difference is probably more inlining and loop unrolling than the previous version. Just spitting out a size in bytes is really meaningless. You told it to make the program run fast and not really care what size the output is.

    3. Re:binary size by sir99 · · Score: 1

      I have noticed on some small programs that -Os actually made the binary bigger than -O3! Also, stripping the binaries produced with 2.95 and 3.0 doesn't reduce the size disparity.

      --
      The ocean parts and the meteors come down
      Laid out in amber, baby.
  9. C++ by akula1 · · Score: 1

    Anybody used it for any c++ code. does it work any better then the other 3.x's?

  10. GCPEE is oit by Anonymous Coward · · Score: 0

    Is it still a defeatured deballed bastardization?

    Stupid fucks took out some great tools.