Slashdot Mirror

← Back to Stories (view on slashdot.org)

MinGW and MSVCRT Conflict Causes Floating-Point Value Corruption

Posted by timothy on from the internal-conflict dept.

jones_supa writes: If you are working on a C++ program where you need very accurate floating point numbers, you might have decided to use long double data type for the extra precision. After a few calculations, you happen to print your number. To your shock, instead of the number being 123.456789, it is printed out as -6.518427 × 10^264 (or 2.745563 depending on your computer). This is actually a bug in some versions of MinGW g++ 4.8.1 (MinGW is a port of GNU programming tools for Windows). Microsoft's C++ runtime library reserves 80 bits for double and long double. When MinGW uses the Microsoft DLL to print out the value, the number is interpreted as using only 64 bits. This discrepancy causes garbage results to be output.

19 of 98 comments (clear)

  1. Bad title. by olsmeister · · Score: 2

    The title implies that the floating point value becomes corrupt. Without looking into it, it sounds like the value does not become corrupt but rather is just not output correctly. The underlying value is still intact.

    1. Re:Bad title. by Blaskowicz · · Score: 2

      If you wanted to rely on that output, say write to a csv file, the data in the csv file is then effectively corrupt.

  2. Re:Surprised? by DRJlaw · · Score: 2

    The sage of Wikipedia states

    The actual size and behavior of floating-point types also vary by implementation. The only guarantee is that long double is not smaller than double, which is not smaller than float. Usually, the 32-bit and 64-bit IEEE 754 binary floating-point formats are used, if supported by hardware.

    So the program implementation assumed a behavior that was not guaranteed, and was burned when it used an outside library which was specification compliant but not in the same way as that particular implementation.

    A poor workman blames his tools. And in this case I'm not referring to MinGW, which is admirably neutral in its reporting.

  3. the 80 bit issue is well known by 140Mandak262Jamuna · · Score: 4, Insightful
    Intel chips provide 80 bit floating point registers, but the storage is 64 bits. GCC would let you compute all calc in 80 bits all registers are loaded with 64 bit fetch and final result is stored in 64 bits. Intermediate results are 80 bit accurate. Some carefully written expressions can limit their truncation errors.

    This is well known. I had a bug in a tree class due to this. The key stored in the instance was 64 bit, but the compare class evaluated and compared it in 80 bits. One of the most difficult bugs I ever encountered. Highly recursive calls to the compare function failing once in about a billion calls... But that was almost 10 or 12 years ago.

    But one thing. GCC handled the truncations correctly. It allows the 80 bit evaluations turned off by compiler options. I don't mix GCC with msvcrt so I am not sure how old / new this is. My 80 bit adventure was in Linux on Intel chips.

    --
    sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
    1. Re:the 80 bit issue is well known by ledow · · Score: 2

      Why the hell were you storing / manipulating a 64-bit key in a tree class as floating-point?

    2. Re:the 80 bit issue is well known by sribe · · Score: 2

      Why the hell were you storing / manipulating a 64-bit key in a tree class as floating-point?

      One would guess that maybe just maybe the natural type of the key was floating point? So then what unnatural type would you suggest he should have used in its place?

      Yes, comparing floating-point numbers is tricky and you have to aware of the issues. No, it is not always wrong to compare them, nor to use them as keys. Would you really argue that it is inappropriate for a database to provide the ability to index a floating-point column???

    3. Re:the 80 bit issue is well known by gl4ss · · Score: 3, Funny

      "I don't mix GCC with msvcrt so I am not sure how old / new this is." Why anybody WOULD mix the two and ever expect anything to work, ever, is kind of beyond me. I guess if you have a week to go through everything line-by-line and your project is manageably small enough to do so.

      yeah why would you link a windows application against windows ui dll's? maybe to display something using them..

      --
      world was created 5 seconds before this post as it is.
  4. Re:Surprised? by TheRaven64 · · Score: 5, Informative

    It's an ABI mismatch, and the summary is nonsense, saying almost the exact opposite of TFA (which I actually read, because the summary is obvious nonsense). The issue is that the Windows ABI defines long double as being a 64-bit floating point value (which is fine, because the only requirement for long double is that it have no less precision than double. If you're using it and expecting some guaranteed precision for vaguely portable code then you're an idiot). For some reason, MinGW (which aims to be ABI-compatible with MS, at least for C libraries) uses 80-bit x87 values for long double, so you get truncation. I forget the exact calling conventions for Windows i386, but I believe that in some cases this will be silently hidden, as the value will be passed in x87 register and so be transparently extended to 80 bits in the caller and truncated in the callee anyway. It's only if it's passed on the stack (or indirectly via a pointer) that it's a problem.

    It's not obvious which definition of long double is better. On modern x86, you'll use SSE for 32- and 64-bit values, and may lose precision moving between x87 and SSE registers. You also get worse IEEE compliance out of the x87 unit, which may matter more than the extra 16 bits of precision. 80-bit floats are not available on any platform other than x86 (128-bit is more common, though PowerPC has its own special non-IEEE version of these and on some other platforms they're entirely done in software), so they're a bad choice if you want portable code that generates the same output on different platforms.

    --
    I am TheRaven on Soylent News
  5. Well, from Dice's perspective... by Anonymous Coward · · Score: 5, Informative

    ... this is better than posting stories about SourceForge getting caught highjacking the dev accounts of major OSS projects I guess.

    1. Re: Well, from Dice's perspective... by Anonymous Coward · · Score: 5, Informative

      The account was hijacked. They locked the gimp-win developer out of his account in order to add malware to an open source project. This is big news, but because it's SlashdotMedia doing the shenanigans, it won't appear on the front page. That's pretty stunning that a tech news site censors the most important tech news of the day, but that's what is going on.

    2. Re:Well, from Dice's perspective... by michelcolman · · Score: 2

      Why don't the GIMP maintainers just accept to take back control, then "update" the project by removing all the source code files and replacing them with funny cat pictures?

  6. 64 bits by simula · · Score: 2

    64 bits should be enough for anybody

    1. Re:64 bits by Carewolf · · Score: 2

      64 bits should be enough for anybody

      Yes, and 80bit floating point gives you exactly 64 significant bits ;)

  7. Re: Surprised? by jones_supa · · Score: 2

    Also earlier versions of MinGW do not have the problem. It's a regression.

  8. Re: Surprised? by jones_supa · · Score: 5, Informative

    Which means it's a 100% windows bug

    No. The MinGW version of GCC allowed to compile programs against the Microsoft C++ runtime library, but the compiler created code which did not follow the spec of the Microsoft library. There really isn't anything to blame about Windows here.

  9. Useful to know... by Anonymous Coward · · Score: 5, Insightful

    But once I've debugged my software and uploaded it to SourceForge can I be sure it won't have an advertising spyware package added to the installer by DICE?

  10. Re:Something seems to be badly garbled in the stor by wonkey_monkey · · Score: 2

    It's the other way around. long double is 80 bits long and most decidedly does fit into 16 bytes, which is does so presumably for alignment purposes.

    --
    systemd is Roko's Basilisk.
  11. exactly by rewindustry · · Score: 5, Informative

    SourceForge, the code repository site owned by Slashdot Media, has apparently seized control of the account hosting GIMP for Windows on the service, according to e-mails and discussions amongst members of the GIMP community—locking out GIMP's lead Windows developer. And now anyone downloading the Windows version of the open source image editing tool from SourceForge gets the software wrapped in an installer replete with advertisements.

    http://arstechnica.com/informa...

    The GIMP developers aren't happy at all about this. They say that Sourceforge impersonated the GIMP developers, and abused the trademarks owned by the GNOME foundation.

    https://mail.gnome.org/archive...

  12. Re:Surprised? by raxx7 · · Score: 2

    x87 can produce IEEE 754 compliant results if the compiler either sets the correct rounding mode before each operation OR if it stores and reloads the results of each operation into memory (forcing the correct rounding).
    However, both are expensive to do, performance wise, and no compiler does so by default.
    Instead, x87 is normally used a way which is not IEEE 754 compliant, although it's actually a bit more accurate: internally, everything is done with 80 bit precision.
    This results from the fact the x87 unit actually predates the final version of the IEEE 754 standard.

    The IEEE 754 standard only covers a few operations: add, subtract, multiply, divide, FMA,
    The transcendals (sin, cos, tag, exp, pow, etc) functions have never been part of the IEEE 754. Historically, most x87 FPUs have had errors larger than 1 ulp, at least for some part of the range.it
    If I am not mistaken, only the AMD K5 FPU actually provided errors of less than 1 ulp for the entire range of inputs. And please, take this with a large grain of salt.