Professional Web Applications Themes

Math::GMP tests and Crypt::Random fail on Compaq (Math::Pari related ??) - PERL Modules

Hi, I've been compiling Math::GMP for different OS's in order to use Net::SFTP and I have successfully compiled and used it for Solaris, HP-UX.11 and Linux, but I'm running into trouble on Compaq (Tru5.1b). GMP library compiled without trouble and all tests were passed. I'm trying to make a dynamic and only OS dependant version of Math::GMP (and of Net::SFTP) and thus GMP was configured as follows: (hardware model from config.guess) ../configure --prefix=$(INSTALL_DIR)/GMP --disable-static --build=alphaev6-dec-osf5.1 --host=none-unknown-osf5.1 Now the next step was to compile Math::GMP and here are the failed test results: PERL_DL_NONLAZY=1 /vobs/prod/tools/perl/OSF1alpha.NG/bin/perl "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/gmppm....ok ...

  1. #1

    Default Math::GMP tests and Crypt::Random fail on Compaq (Math::Pari related ??)

    Hi,

    I've been compiling Math::GMP for different OS's in order to use
    Net::SFTP and I have successfully compiled and used it for Solaris,
    HP-UX.11 and Linux, but I'm running into trouble on Compaq (Tru5.1b).

    GMP library compiled without trouble and all tests were passed. I'm
    trying to make a dynamic and only OS dependant version of Math::GMP
    (and of Net::SFTP) and thus GMP was configured as follows: (hardware
    model from config.guess)

    ../configure --prefix=$(INSTALL_DIR)/GMP --disable-static
    --build=alphaev6-dec-osf5.1 --host=none-unknown-osf5.1

    Now the next step was to compile Math::GMP and here are the failed
    test results:

    PERL_DL_NONLAZY=1 /vobs/prod/tools/perl/OSF1alpha.NG/bin/perl
    "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib',
    'blib/arch')" t/*.t
    t/gmppm....ok 149/349# Test 277 got: '9999999999' (t/gmppm.t at line
    117 fail #277)
    # Expected: '1410065407'
    # Test 278 got: '99999999999' (t/gmppm.t at line 117 fail #278)
    # Expected: '1215752191'
    # Test 279 got: '999999999999' (t/gmppm.t at line 117 fail #279)
    # Expected: '3567587327'
    # Test 284 got: '9999999999' (t/gmppm.t at line 117 fail #284)
    # Expected: '1410065407'
    t/gmppm....FAILED tests 277-279, 284
    Failed 4/349 tests, 98.85% okay
    Failed Test Stat Wstat Total Fail Failed List of Failed
    -------------------------------------------------------------------------------
    t/gmppm.t 349 4 1.15% 277-279 284
    Failed 1/1 test scripts, 0.00% okay. 4/349 subtests failed, 98.85%
    okay.
    gmake[1]: *** [test_dynamic] Error 2

    I am out of ideas what to try next.

    I tried to ignore this and continued to compile my Net::SFTP but ran
    into problems with Crypt::Random. The Crypt::Random tests fail due to
    invalid arguments passed to divll.


    PERL_DL_NONLAZY=1 /vobs/prod/tools/perl/OSF1alpha.NG/bin/perl
    "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib',
    'blib/arch')" t/*.t
    t/generator..........ok
    t/makerandom_itv.....ok 4/5PARI: *** Invalid arguments to divll.
    at /vobs/prod/AMD/ftpcopy/build/OSF1/Crypt-Random-1.13/blib/lib/Crypt/Random.pm
    line 72.
    t/makerandom_itv.....dubious
    Test returned status 255 (wstat 65280, 0xff00)
    DIED. FAILED test 5
    Failed 1/5 tests, 80.00% okay
    t/octet_string.......ok
    t/pari_misfeature....PARI: *** Invalid arguments to divll. at
    lib/Crypt/Random.pm line 72.
    t/pari_misfeature....dubious
    Test returned status 255 (wstat 65280, 0xff00)
    DIED. FAILED test 1
    Failed 1/1 tests, 0.00% okay
    t/random.............ok
    Failed Test Stat Wstat Total Fail Failed List of Failed
    -------------------------------------------------------------------------------
    t/makerandom_itv.t 255 65280 5 1 20.00% 5
    t/pari_misfeature.t 255 65280 1 1 100.00% 1
    Failed 2/5 test scripts, 60.00% okay. 2/26 subtests failed, 92.31%
    okay.
    gmake[1]: *** [test_dynamic] Error 2
    gmake[1]: Leaving directory
    `/vobs/prod/AMD/ftpcopy/build/OSF1/Crypt-Random-1.13'

    Can anyone say whether this problem is related to the Math::Pari ie.
    is Crypt::Random using Math::Pari incorrectly or what? Also
    Crypt::Random tests take forever (1/2-2 hours) to execute. I haven't
    experienced this kind of slowness on any other platform.

    I'm using Perl 5.8.0 and it says:

    Summary of my perl5 (revision 5.0 version 8 subversion 0)
    configuration:
    Platform:
    osname=dec_osf, osvers=5.1, archname=OSF1alpha-dec_osf
    uname='osf1 sinika v5.1 732 alpha '
    config_args='-Dcc=cc -Dprefix=/vobs/prod/tools/perl/OSF1alpha'
    hint=recommended, useposix=true, d_sigaction=define
    usethreads=undef use5005threads=undef useithreads=undef
    usemultiplicity=undef
    useperlio=define d_sfio=undef uselargefiles=define usesocks=undef
    use64bitint=define use64bitall=define uselongdouble=undef
    usemymalloc=n, bincompat5005=undef
    Compiler:
    cc='cc', ccflags ='-std -D_INTRINSICS -fprm d -ieee
    -I/usr/local/include -DLANGUAGE_C',
    optimize='-O4',
    cppflags='-std -D_INTRINSICS -ieee -I/usr/local/include
    -DLANGUAGE_C'
    ccversion='V6.3-029', gccversion='', gccosandvers=''
    intsize=4, longsize=8, ptrsize=8, doublesize=8, byteorder=12345678
    d_longlong=define, longlongsize=8, d_longdbl=define,
    longdblsize=16
    ivtype='long', ivsize=8, nvtype='double', nvsize=8, Off_t='off_t',
    lseeksize=8
    alignbytes=8, prototype=define
    Linker and Libraries:
    ld='ld', ldflags =' -L/usr/local/lib'
    libpth=/usr/local/lib /usr/shlib /usr/ccs/lib /usr/lib/cmplrs/cc
    /usr/lib /var/shlib
    libs=-lpthread -ldbm -ldb -lm -lutil
    perllibs=-lpthread -lm -lutil
    libc=/usr/shlib/libc.so, so=so, useshrplib=true,
    libperl=libperl.so
    gnulibc_version=''
    Dynamic Linking:
    dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='
    -Wl,-rpath,/vobs/prod/tools/perl/OSF1alpha/lib/5.8.0/OSF1alpha-dec_osf/CORE'
    cccdlflags=' ', lddlflags='-shared -expect_unresolved "*" -O4
    -msym -std -s -L/usr/local/lib'


    Characteristics of this binary (from libperl):
    Compile-time options: USE_64_BIT_INT USE_64_BIT_ALL USE_LARGE_FILES
    Built under dec_osf
    Compiled at Oct 15 2002 10:03:55
    %ENV:
    PERL5LIB="/vobs/prod/AMD/ftpcopy/lib/OSF1/lib:/vobs/prod/AMD/ftpcopy/lib/OSF1/lib/site_perl:/vobs/prod/AMD/ftpcopy/lib/OSF1/lib/site_perl/5.8.0/OSF1"
    INC:
    /vobs/prod/AMD/ftpcopy/lib/OSF1/lib/5.8.0/OSF1alpha-dec_osf
    /vobs/prod/AMD/ftpcopy/lib/OSF1/lib/5.8.0
    /vobs/prod/AMD/ftpcopy/lib/OSF1/lib
    /vobs/prod/AMD/ftpcopy/lib/OSF1/lib/site_perl/5.8.0/OSF1alpha-dec_osf
    /vobs/prod/AMD/ftpcopy/lib/OSF1/lib/site_perl/5.8.0
    /vobs/prod/AMD/ftpcopy/lib/OSF1/lib/site_perl
    /vobs/prod/AMD/ftpcopy/lib/OSF1/lib/site_perl/5.8.0/OSF1
    /vobs/prod/tools/perl/OSF1alpha/lib/5.8.0/OSF1alpha-dec_osf
    /vobs/prod/tools/perl/OSF1alpha/lib/5.8.0
    /vobs/prod/tools/perl/OSF1alpha/lib/site_perl/5.8.0/OSF1alpha-dec_osf
    /vobs/prod/tools/perl/OSF1alpha/lib/site_perl/5.8.0
    /vobs/prod/tools/perl/OSF1alpha/lib/site_perl

    my platform is:

    OSF1 V5.1 2650 alpha

    best regards,

    /jUSSi
    Jussi Mononen Guest

  2. #2

    Default Re: Math::GMP tests and Crypt::Random fail on Compaq (Math::Parirelated ??)

    Jussi Mononen wrote:
    >
    > Now the next step was to compile Math::GMP and here are the failed
    > test results:
    >
    > PERL_DL_NONLAZY=1 /vobs/prod/tools/perl/OSF1alpha.NG/bin/perl
    > "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib',
    > 'blib/arch')" t/*.t
    > t/gmppm....ok 149/349# Test 277 got: '9999999999' (t/gmppm.t at line
    > 117 fail #277)
    > # Expected: '1410065407'
    > # Test 278 got: '99999999999' (t/gmppm.t at line 117 fail #278)
    > # Expected: '1215752191'
    > # Test 279 got: '999999999999' (t/gmppm.t at line 117 fail #279)
    > # Expected: '3567587327'
    > # Test 284 got: '9999999999' (t/gmppm.t at line 117 fail #284)
    > # Expected: '1410065407'
    > t/gmppm....FAILED tests 277-279, 284
    > Failed 4/349 tests, 98.85% okay
    > Failed Test Stat Wstat Total Fail Failed List of Failed
    > -------------------------------------------------------------------------------
    > t/gmppm.t 349 4 1.15% 277-279 284
    > Failed 1/1 test scripts, 0.00% okay. 4/349 subtests failed, 98.85%
    > okay.
    > gmake[1]: *** [test_dynamic] Error 2
    The GMP doentation suggests that those tests will fail if your
    'unsigned long' and 'signed long' are not 32 bits. Your 'perl -V' (which
    you thoughtfully provided) would suggest that your 'unsigned long' and
    'signed long' are 64 bits. That being the case, the tests have returned
    the values that are to be expected - ie they have passed, not failed.

    The functions that reported a failure are 'uintify_gmp' and
    'intify_gmp'. If you want to know what these functions do, go to:

    [url]http://swox.com/gmp/manual/Converting-Integers.html#Converting%20Integers[/url]

    and read the doentation on 'mpz_get_ui()' and 'mpz_get_si()'.
    ('uintify_gmp' is simply the perl interface to the former, and
    'intify_gmp' the perl interface to the latter.)

    As for the rest, I can't really help - though I'd be wondering how those
    other modules also cope with 64 bit ints.

    Cheers,
    Rob


    --
    To reply by email u have to take out the u in kalinaubears.

    Sisyphus Guest

  3. #3

    Default Re: Math::GMP tests and Crypt::Random fail on Compaq (Math::Parirelated ??)

    Sisyphus wrote:
    >
    >
    > The GMP doentation suggests that those tests will fail if your
    > 'unsigned long' and 'signed long' are not 32 bits.
    That's possibly a bit misleading. Just to make sure I'm being clear, the
    tests *haven't* failed. A failure is being reported because the
    "expected values" that the Math::GMP author provided are incorrect in
    your case. (He assumed the tests were going to be conducted on a 32 bit
    machine.)

    Cheers,
    Rob

    --
    To reply by email u have to take out the u in kalinaubears.

    Sisyphus Guest

  4. #4

    Default Re: Math::GMP tests and Crypt::Random fail on Compaq (Math::Pari related ??)

    > Sisyphus <kalinaubearsiinet.net.au> wrote in message news:
    > <3fdfe44f$0$1749$5a62ac22freenews.iinet.net.au>.. .
    >
    > > The GMP doentation suggests that those tests will fail if your
    > > 'unsigned long' and 'signed long' are not 32 bits. Your 'perl -V' (which
    > > you thoughtfully provided) would suggest that your 'unsigned long' and
    > > 'signed long' are 64 bits. That being the case, the tests have returned
    > > the values that are to be expected - ie they have passed, not failed.
    >
    > That's possibly a bit misleading. Just to make sure I'm being clear, the
    > tests *haven't* failed. A failure is being reported because the
    > "expected values" that the Math::GMP author provided are incorrect in
    > your case. (He assumed the tests were going to be conducted on a 32 bit
    > machine.)
    So basicly everything _should_ be okay? The results returned from
    Math::GMP are _correct_ and can not cause errors in future use?

    Can this be considered as a bug in the Math::GMP test suite? I'd hate
    to install this and later on find out that it does not work correctly.

    anyway, thanks for the information :-)

    /jUSSi

    ps. just for the record, this module is a prerequisite for Net::SSH
    and thus for Net::SFTP which is my actual target. I don't know the
    extent of GMP library usage so with a bit of luck this could be
    ignored.
    Jussi Mononen Guest

  5. #5

    Default Re: Math::GMP tests and Crypt::Random fail on Compaq (Math::Parirelated ??)

    Jussi Mononen wrote:
    >
    > Can this be considered as a bug in the Math::GMP test suite?
    Sure can.
    For these tests you really need to know whether it's 32 or 64 bit
    architecture before you can work out what the expected return value is.
    (The author has simply assumed that it's 32 bit architecture.)
    >
    > ps. just for the record, this module is a prerequisite for Net::SSH
    > and thus for Net::SFTP which is my actual target. I don't know the
    > extent of GMP library usage so with a bit of luck this could be
    > ignored.
    Wouldn't be hard to find out which Math::GMP functions are used. Just
    look for any imported functions when the module is loaded in these other
    modules - also search them for the string 'Math::GMP::'.
    However, at this stage I haven't seen anything which suggests to me that
    you're going to strike trouble with *any* of the Math::GMP functions.

    Cheers,
    Rob


    --
    To reply by email u have to take out the u in kalinaubears.

    Sisyphus Guest

Similar Threads

  1. Math::Pari 'factor' is wrong
    By gamo in forum PERL Modules
    Replies: 8
    Last Post: March 18th, 11:36 AM
  2. Segmentation violation in Math:Pari when used in connection with
    By Joern Wuestenfeld in forum PERL Modules
    Replies: 0
    Last Post: February 2nd, 09:18 AM
  3. Math::Pari test fail on Compaq OSF1 5.1b
    By Jussi Mononen in forum PERL Modules
    Replies: 7
    Last Post: January 15th, 06:38 PM
  4. Math::Pari and libsunmath.so.1 - is it really required?
    By Jussi Mononen in forum PERL Modules
    Replies: 4
    Last Post: December 17th, 06:33 AM
  5. Math::PARI tests fail on HP-UX.11 and Linux RH7.3
    By Jussi Mononen in forum PERL Modules
    Replies: 1
    Last Post: November 8th, 09:03 AM

Bookmarks

Posting Permissions

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

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139