From @eserte

Created by @eserte

One test failure on a FreeBSD 10.0 system​:

t/op/sprintf2 ................................................. # Failed test 1419 - '%a' '2.71828182845905' -> '0x1.5bf0a8b14576ap+1' cf '0x1.5bf0a8b145769p+1' at op/sprintf2.t line 577
# got "0x1.5bf0a8b14576ap+1"
# expected "0x1.5bf0a8b145769p+1"
FAILED at test 1419

Flags:
    category=core
    severity=low

Site configuration information for perl 5.21.3:

Configured by eserte at Thu Aug 21 07:37:37 CEST 2014.

Summary of my perl5 (revision 5 version 21 subversion 3) configuration:
   
  Platform:
    osname=freebsd, osvers=10.0-release-p1, archname=amd64-freebsd
    uname='freebsd cvrsnica-freebsd-10-64bit.herceg.de 10.0-release-p1 freebsd 10.0-release-p1 #0: tue apr 8 06:45:06 utc 2014 root@amd64-builder.daemonology.net:usrobjusrsrcsysgeneric amd64 '
    config_args='-ds -e -Dprefix=/usr/perl5.21.3 -Dusedevel -Dusemallocwrap=no'
    hint=recommended, useposix=true, d_sigaction=define
    useithreads=undef, usemultiplicity=undef
    use64bitint=define, use64bitall=define, uselongdouble=undef
    usemymalloc=n, bincompat5005=undef
  Compiler:
    cc='cc', ccflags ='-DHAS_FPSETMASK -DHAS_FLOATINGPOINT_H -fno-strict-aliasing -pipe -fstack-protector -I/usr/local/include -D_FORTIFY_SOURCE=2',
    optimize='-O',
    cppflags='-DHAS_FPSETMASK -DHAS_FLOATINGPOINT_H -fno-strict-aliasing -pipe -fstack-protector -I/usr/local/include'
    ccversion='', gccversion='4.2.1 Compatible FreeBSD Clang 3.3 (tags/RELEASE_33/final 183502)', gccosandvers=''
    intsize=4, longsize=8, ptrsize=8, doublesize=8, byteorder=12345678
    d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=16, longdblkind=3
    ivtype='long', ivsize=8, nvtype='double', nvsize=8, Off_t='off_t', lseeksize=8
    alignbytes=8, prototype=define
  Linker and Libraries:
    ld='cc', ldflags ='-Wl,-E  -fstack-protector -L/usr/local/lib'
    libpth=/usr/lib /usr/local/lib /usr/include/clang/3.3 /usr/lib
    libs=-lgdbm -lm -lcrypt -lutil -lc
    perllibs=-lm -lcrypt -lutil -lc
    libc=, so=so, useshrplib=false, libperl=libperl.a
    gnulibc_version=''
  Dynamic Linking:
    dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags=' '
    cccdlflags='-DPIC -fPIC', lddlflags='-shared  -L/usr/local/lib -fstack-protector'



@INC for perl 5.21.3:
    /usr/perl5.21.3/lib/site_perl/5.21.3/amd64-freebsd
    /usr/perl5.21.3/lib/site_perl/5.21.3
    /usr/perl5.21.3/lib/5.21.3/amd64-freebsd
    /usr/perl5.21.3/lib/5.21.3
    .


Environment for perl 5.21.3:
    HOME=/home/eserte
    LANG (unset)
    LANGUAGE (unset)
    LD_LIBRARY_PATH (unset)
    LOGDIR (unset)
    PATH=/usr/local/bin:/usr/bin:/bin:/usr/local/sbin:/usr/sbin:/sbin:/home/eserte/bin/freebsd10.0:/home/eserte/bin/sh:/home/eserte/bin:/usr/games:/home/eserte/devel
    PERLDOC=-MPod::Perldoc::ToTextOverstrike
    PERL_BADLANG (unset)
    PERL_HTML_DISPLAY_CLASS=HTML::Display::Mozilla
    SHELL=/usr/local/bin/zsh

From @jhi

Yeah, I managed to break that just before the release :-(

The HEAD from http​://perl5.git.perl.org/perl.git should work fine.

From [Unknown Contact. See original ticket]

Yeah, I managed to break that just before the release :-(

The HEAD from http​://perl5.git.perl.org/perl.git should work fine.

From @jhi

On closer look, this

# got "0x1.5bf0a8b14576ap+1"
# expected "0x1.5bf0a8b145769p+1"

is a narrower failure (and not a wider breakage which I suspect), which is caused by too rigid test (as you can see just the last digits differ).

and this tweak

http​://perl5.git.perl.org/perl.git/commitdiff/61b5c6185cc0aebfbd823e601ee6c8023c6aea0d

should allow the test to pass.

From [Unknown Contact. See original ticket]

On closer look, this

# got "0x1.5bf0a8b14576ap+1"
# expected "0x1.5bf0a8b145769p+1"

is a narrower failure (and not a wider breakage which I suspect), which is caused by too rigid test (as you can see just the last digits differ).

and this tweak

http​://perl5.git.perl.org/perl.git/commitdiff/61b5c6185cc0aebfbd823e601ee6c8023c6aea0d

should allow the test to pass.

From @jkeenan

On Thu Aug 21 13​:10​:14 2014, jhi wrote​:

On closer look, this

# got "0x1.5bf0a8b14576ap+1"
# expected "0x1.5bf0a8b145769p+1"

is a narrower failure (and not a wider breakage which I suspect),
which is caused by too rigid test (as you can see just the last digits
differ).

and this tweak

http​://perl5.git.perl.org/perl.git/commitdiff/61b5c6185cc0aebfbd823e601ee6c8023c6aea0d

should allow the test to pass.

Slaven (or anyone)​:

Can we get confirmation that Jarkko's fix resolved the problem on the OS in question?

Thank you very much.

--
James E Keenan (jkeenan@​cpan.org)

The RT System itself - Status changed from 'new' to 'open'

From @sisyphus

-----Original Message-----
From​: James E Keenan via RT
Sent​: Tuesday, September 23, 2014 8​:49 AM
To​: OtherRecipients of perl Ticket #122591​:
Cc​: perl5-porters@​perl.org
Subject​: [perl #122591] t/op/sprintf2 fails (hex floats?) (FreeBSD 10.0)

On Thu Aug 21 13​:10​:14 2014, jhi wrote​:

On closer look, this

# got "0x1.5bf0a8b14576ap+1"
# expected "0x1.5bf0a8b145769p+1"

Can we get confirmation that Jarkko's fix resolved the problem on the OS
in question?

Jarkko's fix (judging by the comments in the code) seems aimed at hiding
off-by-one discrepancies.

Are the failures on FreeBSD 10.0 off by one ? .... or off by more than one ?
On Linux and/or Windows I've seen discrepancies of 2 or more ULP's, though
the "problem" values vary from system to system.

Cheers,
Rob

From @eserte

Created by @eserte

One test failure on a FreeBSD 10.0 system​:

t/op/sprintf2 ................................................. # Failed test 1419 - '%a' '2.71828182845905' -> '0x1.5bf0a8b14576ap+1' cf '0x1.5bf0a8b145769p+1' at op/sprintf2.t line 577 # got "0x1.5bf0a8b14576ap+1" # expected "0x1.5bf0a8b145769p+1" FAILED at test 1419
Perl Info

This failure was reported on a version of FreeBSD which is no longer supported. According to our smoke-test aggregator, t/op/sprintf2.t is only failing on HP-UX. That should be the subject of a separate ticket. Closing this one.