From @andk

git bisect

commit 8c12dc6
Author​: Jarkko Hietaniemi <jhi@​iki.fi>
Date​: Sat Aug 23 22​:49​:04 2014 -0400

  More robust inf/nan recognition and generation.

sample fail report

http​://www.cpantesters.org/cpan/report/fef572c0-2c93-11e4-b7c9-7fcbcc4edbc1

perl -V

Summary of my perl5 (revision 5 version 21 subversion 4) configuration​:
  Commit id​: 35c2140
  Platform​:
  osname=linux, osvers=3.14-2-amd64, archname=x86_64-linux-ld
  uname='linux k83 3.14-2-amd64 #1 smp debian 3.14.15-2 (2014-08-09) x86_64 gnulinux '
  config_args='-Dprefix=/home/sand/src/perl/repoperls/installed-perls/perl/v5.21.3-121-g35c2140/127e -Dmyhostname=k83 -Dinstallusrbinperl=n -Uversiononly -Dusedevel -des -Ui_db -Uuseithreads -Duselongdouble -DDEBUGGING=-g'
  hint=recommended, useposix=true, d_sigaction=define
  useithreads=undef, usemultiplicity=undef
  use64bitint=define, use64bitall=define, uselongdouble=define
  usemymalloc=n, bincompat5005=undef
  Compiler​:
  cc='cc', ccflags ='-fwrapv -fno-strict-aliasing -pipe -fstack-protector -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_FORTIFY_SOURCE=2',
  optimize='-O2 -g',
  cppflags='-fwrapv -fno-strict-aliasing -pipe -fstack-protector -I/usr/local/include'
  ccversion='', gccversion='4.9.1', 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='long double', nvsize=16, Off_t='off_t', lseeksize=8
  alignbytes=16, prototype=define
  Linker and Libraries​:
  ld='cc', ldflags =' -fstack-protector -L/usr/local/lib'
  libpth=/usr/local/lib /usr/lib/gcc/x86_64-linux-gnu/4.9/include-fixed /usr/include/x86_64-linux-gnu /usr/lib /lib/x86_64-linux-gnu /lib/../lib /usr/lib/x86_64-linux-gnu /usr/lib/../lib /lib
  libs=-lnsl -lgdbm -ldb -ldl -lm -lcrypt -lutil -lc -lgdbm_compat
  perllibs=-lnsl -ldl -lm -lcrypt -lutil -lc
  libc=libc-2.19.so, so=so, useshrplib=false, libperl=libperl.a
  gnulibc_version='2.19'
  Dynamic Linking​:
  dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-Wl,-E'
  cccdlflags='-fPIC', lddlflags='-shared -O2 -g -L/usr/local/lib -fstack-protector'

Characteristics of this binary (from libperl)​:
  Compile-time options​: HAS_TIMES PERLIO_LAYERS PERL_DONT_CREATE_GVSV
  PERL_HASH_FUNC_ONE_AT_A_TIME_HARD PERL_MALLOC_WRAP
  PERL_NEW_COPY_ON_WRITE PERL_PRESERVE_IVUV
  PERL_USE_DEVEL USE_64_BIT_ALL USE_64_BIT_INT
  USE_LARGE_FILES USE_LOCALE USE_LOCALE_COLLATE
  USE_LOCALE_CTYPE USE_LOCALE_NUMERIC USE_LOCALE_TIME
  USE_LONG_DOUBLE USE_PERLIO USE_PERL_ATOF
  Built under linux
  Compiled at Aug 25 2014 08​:06​:32
  %ENV​:
  PERL5LIB=""
  PERL5OPT=""
  PERL5_CPANPLUS_IS_RUNNING="8266"
  PERL5_CPAN_IS_RUNNING="8266"
  PERL_MM_USE_DEFAULT="1"
  @​INC​:
  /home/sand/src/perl/repoperls/installed-perls/perl/v5.21.3-121-g35c2140/127e/lib/site_perl/5.21.4/x86_64-linux-ld
  /home/sand/src/perl/repoperls/installed-perls/perl/v5.21.3-121-g35c2140/127e/lib/site_perl/5.21.4
  /home/sand/src/perl/repoperls/installed-perls/perl/v5.21.3-121-g35c2140/127e/lib/5.21.4/x86_64-linux-ld
  /home/sand/src/perl/repoperls/installed-perls/perl/v5.21.3-121-g35c2140/127e/lib/5.21.4
  .
--
andreas

From @schwern

Those are all tests for what should be invalid numbers. "near-misses (wrong spelling and wrong case) for the Win32 Inf and NaN. these should be invalid numbers on all platforms". I added some annotation to the tests to find out what exactly failed.

!isnan(1.#InD)
!isnan(-1.#InD)
!isinf(1.#InF)
!isinf(-1.#InF)

1.#InF and -1.#InF are considered infinity.
1.#InD and -1.#iND are considered NaN.

I'm no number wizard, but it looks like a regression in perl. I'll leave it to y'all to decide.

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

From @andk

Also affected​:
CMS/Mustache-Simple-v1.3.1.tar.gz
http​://www.cpantesters.org/cpan/report/76da0bc2-4c72-11e4-9aeb-3e61e0bfc7aa
--
andreas

From @arc

Michael G Schwern via RT <perlbug-followup@​perl.org> wrote​:

Those are all tests for what should be invalid numbers. "near-misses (wrong spelling and wrong case) for the Win32 Inf and NaN. these should be invalid numbers on all platforms". I added some annotation to the tests to find out what exactly failed.

!isnan(1.#InD)
!isnan(-1.#InD)
!isinf(1.#InF)
!isinf(-1.#InF)

1.#InF and -1.#InF are considered infinity.
1.#InD and -1.#iND are considered NaN.

I'm no number wizard, but it looks like a regression in perl. I'll leave it to y'all to decide.

This looks to me like the result of a deliberate change in Perl​:
grok_number() now accepts the Win32-style "1.#INF" and "1.#IND"
strings in any letter-case variants.

My view is that the Scalar​::Util​::Numeric test suite should change to
be happy for Perl to treat mixed-case "1.#InF" and "1.#InD" as
variants of the upper-case equivalents, and I've got a suitable patch
ready to submit to S​::U​::N. But I thought it would be worth checking
for dissenting opinions first; Jarkko, do you have a view on this?

--
Aaron Crane ** http​://aaroncrane.co.uk/

From @arc

Andreas Koenig <andreas.koenig.7os6VVqR@​franz.ak.mind.de> wrote​:

Also affected​:
CMS/Mustache-Simple-v1.3.1.tar.gz
http​://www.cpantesters.org/cpan/report/76da0bc2-4c72-11e4-9aeb-3e61e0bfc7aa

This looks like a genuine regression. This reduction prints "matched"
under 5.18 and 5.20, but not under blead​:

no warnings "experimental​::smartmatch";
my $x = "Ind";
$x + 0;
"Ind" ~~ $_ and print "matched\n" for $x;

--
Aaron Crane ** http​://aaroncrane.co.uk/

From @arc

Aaron Crane <arc@​cpan.org> wrote​:

My view is that the Scalar​::Util​::Numeric test suite should change to
be happy for Perl to treat mixed-case "1.#InF" and "1.#InD" as
variants of the upper-case equivalents, and I've got a suitable patch
ready to submit to S​::U​::N. But I thought it would be worth checking
for dissenting opinions first; Jarkko, do you have a view on this?

In private communication, Jarkko has concurred with my assessment.
I've submitted a patch for Scalar​::Util​::Numeric here​:

https://github.com/chocolateboy/Scalar-Util-Numeric/pull/2

--
Aaron Crane ** http​://aaroncrane.co.uk/

From @jhi

On Monday-201410-20, 18​:03, Aaron Crane wrote​:

no warnings "experimental​::smartmatch";
my $x = "Ind";
$x + 0;
"Ind" ~~ $_ and print "matched\n" for $x;

Huh? ... Oh, okay. Something is definitely broken here... mea culpa,
though I need to find out where the culpa exactly is.

("Ind" + 0 returns 'NaN'.)

It shouldn't trigger on its own, like "inf" and "nan" do,
only as part of the win32y "1.#ind".

From @jhi

On Monday-201410-20, 18​:13, Jarkko Hietaniemi wrote​:

On Monday-201410-20, 18​:03, Aaron Crane wrote​:

no warnings "experimental​::smartmatch";
my $x = "Ind";
$x + 0;
"Ind" ~~ $_ and print "matched\n" for $x;

Huh? ... Oh, okay. Something is definitely broken here... mea culpa,
though I need to find out where the culpa exactly is.

("Ind" + 0 returns 'NaN'.)

It shouldn't trigger on its own, like "inf" and "nan" do,
only as part of the win32y "1.#ind".

e855f54

@arc - Status changed from 'open' to 'resolved'