$) and $( should be IOKp but not IOK

[FGasper]

> perl -MDevel::Peek -e'Dump( my $v = $) )'
SV = PVMG(0x7fe9b782e380) at 0x7fe9b7806ad0
  REFCNT = 1
  FLAGS = (IOK,POK,IsCOW,pIOK,pPOK)
  IV = 20
  NV = 0
  PV = 0x7fe9b6505760 "20 20 505 504 12 61 79 80 81 98 33 100 204 250 395 398 399"\0
  CUR = 58
  LEN = 64
  COW_REFCNT = 1

The fact that this is marked IOK tells XS modules (e.g., Sereal) that the IV value is legitimate. That’s sort of true, but it’s inconsistent with Perl’s typical behaviour of invalidating IOK whenever the IV mismatches the PV, which is the case here.

Note that leaving the private flag (IOKp) in place will prevent not-a-number warnings.

[FGasper]

This causes https://github.com/Sereal/Sereal/issues/263

[atoomic]

note that it's coming from the void)SvIOK_on(sv); /* what a wonderful hack! */ setting the IOK in all flags view https://github.com/Perl/perl5/blob/c683824536c9806b1b970459bf81690e006b59ae/mg.c#L1227

My understanding is the issue is when serializing the data. We probably should balance pros/cons of having the IOK flag set there

[toddr]

Bisecting, you can look at:

LD_LIBRARY_PATH=/root/projects/perl ./perl -Ilib -MDevel::Peek -e'Dump( my $v = $) )' 2>&1|egrep '\bIOK\b'

If it gives back IOK, then the bug is still there.

At which point you can bisect it to:

4bac9ae47b5ad7845a24e26b0e95609805de688a is the first bad commit
commit 4bac9ae47b5ad7845a24e26b0e95609805de688a
Author: Chip Salzenberg <chip@pobox.com>
Date:   Fri Jun 22 15:18:18 2012 -0700
    Magic flags harmonization.
    In restore_magic(), which is called after any magic processing, all of
    the public OK flags have been shifted into the private OK flags.  Thus
    the lack of an appropriate public OK flags was used to trigger both get
    magic and required conversions.  This scheme did not cover ROK, however,
    so all properly written code had to make sure mg_get was called the right
    number of times anyway.  Meanwhile the private OK flags gained a second
    purpose of marking converted but non-authoritative values (e.g. the IV
    conversion of an NV), and the inadequate flag shift mechanic broke this
    in some cases.
    This patch removes the shift mechanic for magic flags, thus exposing (and
    fixing) some improper usage of magic SVs in which mg_get() was not called
    correctly.  It also has the side effect of making magic get functions
    specifically set their SVs to undef if that is desired, as the new behavior
    of empty get functions is to leave the value unchanged.  This is a feature,
    as now get magic that does not modify its value, e.g. tainting, does not
    have to be special cased.
    The changes to cpan/ here are only temporary, for development only, to
    keep blead working until upstream applies them (or something like them).
    Thanks to Rik and Father C for review input.
 av.c                             |   2 +-
 cpan/Compress-Raw-Bzip2/Bzip2.xs |  68 ++---
 cpan/Compress-Raw-Zlib/Zlib.xs   |  82 +++---
 doio.c                           |   1 +
 doop.c                           |  10 +-
 embed.fnc                        |   7 +-
 embed.h                          |   3 +-
 ext/Devel-Peek/t/Peek.t          |   2 +-
 gv.c                             |   2 +
 mg.c                             |  51 ++--
 pp.c                             |   2 +-
 pp_ctl.c                         |   8 +-
 pp_hot.c                         |  13 +-
 pp_sys.c                         |  19 +-
 proto.h                          |  20 +-
 sv.c                             | 558 ++++++++++++++++++---------------------
 sv.h                             | 169 ++++++------
 t/op/eval.t                      |   2 +-
 t/op/tie.t                       |   5 +-
 19 files changed, 486 insertions(+), 538 deletions(-)

[nwc10]

Fix is in blead.