In perl.git, the branch smoke-me/tonyc/5.21.1-blockers has been created
<http://perl5.git.perl.org/perl.git/commitdiff/1922d1674cfbae9d0808143b9bbe1bd6671e6016?hp=0000000000000000000000000000000000000000>
at 1922d1674cfbae9d0808143b9bbe1bd6671e6016 (commit)
- Log -----------------------------------------------------------------
commit 1922d1674cfbae9d0808143b9bbe1bd6671e6016
Author: Daniel Dragan <bul...@hotmail.com>
Date: Sat May 17 22:10:01 2014 -0400
refactor pp_list
-move PL_stack_sp and PL_stack_base reads into the branch in which they
are used, this also removes 1 var from being saved across the function
call in GIMME, which removes saving and restoring 1 non-vol register
-write SP to PL_stack_sp (PUTBACK) only if it was changed
-POPMARK is mutable, it must execute on all branches
this reduced pp_list's machine code size of the function from 0x58 to
0x53 bytes on VC 2003 -01 32 bits
M pp.c
commit b3af97dbd9c42635dc0cb91154f883944e701d04
Author: Tony Cook <t...@develop-help.com>
Date: Wed May 21 10:45:58 2014 +1000
remove comment that no longer applies
M x2p/a2py.c
commit e7b3022dec9bb51dae04a45df57a749f96be1f77
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Fri Apr 25 21:52:54 2014 -0400
Fix for Coverity perl5 CIDs 29203, 29207, 29211, 29214, 29217, 29218,
29222: Unused pointer value (UNUSED_VALUE) Pointer foo returned by bar() is
overwritten.
Pointers set but then (immediately or very shortly) overwritten.
M ext/XS-APItest/APItest.xs
M perlio.c
M toke.c
M x2p/a2py.c
M x2p/walk.c
commit 7d84cea9f88f757dc692069f376c9fa3ddd87772
Author: Tony Cook <t...@develop-help.com>
Date: Wed May 21 14:17:13 2014 +1000
bump $threads::VERSION to 1.94
M dist/threads/lib/threads.pm
commit 1e791dd17140775224b1d119ae25650dea034974
Author: Tony Cook <t...@develop-help.com>
Date: Wed May 21 14:16:57 2014 +1000
bump $PerlIO::mmap::VERSION to 0.012
M ext/PerlIO-mmap/mmap.pm
commit d762b860deef2337948797739a29341d9e27eadd
Author: Tony Cook <t...@develop-help.com>
Date: Wed May 21 14:16:42 2014 +1000
bump $IO::VERSION to 1.32
M dist/IO/IO.pm
commit 9da0e8bdcd15871ec62502a5cfddcabfcf44e223
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Fri May 2 22:12:24 2014 -0400
Check fileno/numgroups -1, check fcntl (+fgetc) failures.
(merged fix for perl #121743 and perl #121745)
M dist/IO/IO.xs
M dist/threads/threads.xs
M doio.c
M ext/PerlIO-mmap/mmap.xs
M mg.c
M perl.c
M perlio.c
M pod/perldiag.pod
M pp_sys.c
M util.c
commit 15e2d1a76908821f0869b875d3e8b77874b36724
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Wed May 7 09:25:04 2014 -0400
slen may be uninitialized.
Fix for Coverity perl5 CID 29081: Uninitialized scalar
variable (UNINIT) uninit_use_in_call: Using uninitialized value slen when
calling Perl_croak.
If all fails, slen hasn't been set, and croak will be called with that.
M utf8.c
commit 97703a550ad58f60ea7cfce6e607e4bdbb18e1ab
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Wed May 7 09:22:16 2014 -0400
Leaked string in failure path.
Fix for Coverity perl5 CID 29058: Resource leak
(RESOURCE_LEAK) leaked_storage: Variable codeset going out of scope leaks
the
storage it points to.
The savepv-ed codeset was not freed in failure path.
(The save_input_locale is freed just few lines later.)
M locale.c
commit bb9bcdb318ae60463a78fcf33078957eb280ea67
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Wed May 7 09:30:35 2014 -0400
fd closes for failure paths.
Fix for Coverity perl5 CIDs 29053, 29055, 29057: Resource leak
(RESOURCE_LEAK)
leaked_handle: Handle variable fd going out of scope leaks the handle.
M perlio.c
commit e76db56a3f6770003ab2eee3f4a00651d9f011b5
Author: Tony Cook <t...@develop-help.com>
Date: Tue Apr 15 03:57:57 2014 +0200
fix the I32 bug for index() and rindex()
M MANIFEST
M pp.c
A t/bigmem/index.t
commit 19b161710f609ccead374d2404ec0a740e4c3c7b
Author: Smylers <smyl...@stripey.com>
Date: Fri Mar 28 16:33:05 2014 +0000
Document -l on symlinkless file systems
We can rely on -l always being false on platforms that don't support
symlinks:
http://www.nntp.perl.org/group/perl.perl5.porters/2014/03/msg214012.html
M pod/perlfunc.pod
commit 27b2432b97e5b5d923bbf188dac43b7d632bb7c4
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Tue May 6 10:40:21 2014 -0400
UCHARAT unnecessary with isSPACE().
UCHARAT is not only unnecessary here (and no other spot in the core
seems to use UCHARAT with isFOO()), but it also narrows the type so
that some of the code in isSPACE() becomes unreachable.
Fix for Coverity perl5 CIDs 28937, 28938.
M regcomp.c
commit f307a898156d6cc48d255bcbfc91e44765bcb3fc
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Tue May 6 11:52:02 2014 -0400
Strengthen the ix zero min-clamping to panic.
Currently the ix cannot ever be less than zero (detected by Coverity).
Furthermore, it probably should never get less than zero, so panic if
that ever happens.
Fix for Coverity perl5 CID 28946.
M pp_ctl.c
commit 0bf3fdcf040a921d8d5d8f58c1d6bf4199d3f9ff
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Mon May 19 17:57:06 2014 -0400
The max size of the flop operator range.
Firstly, rename {max,j} -> {j,n} since j as the count is very confusing.
The j > SSize_t_MAX cannot (usually) fire.
(1) If the IV and SSize_t are the same (very likely) the count (an IV)
simply cannot be larger than the max of SSize_t.
(2) If IV is larger than SSize_t (e.g. long longs for IVs but only
32-bit pointers available) the count could conceivably be larger than
the max of SSize_t, but we still need to correctly dance around the
max: the signed maxima are tricky since the behavior of overflow is
undefined.
(3) The IV cannot be smaller than than SSize_t (e.g. I32 for IV but
64-bit pointers available): IV is supposed to be large enough to
contain pointers).
NOTE: this logic only protects against the wraparound, not against
an OOM: we will run out of memory long before we create (even)
SSize_t_MAX SVs. There is nothing magical about SSize_t as such,
except that it is a likely wraparound spot.
Fix for Coverity perl5 CID 28933.
M pp_ctl.c
commit d14f72661ca069ab32eb5936ed86e624f9048ec3
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Sun May 4 19:44:24 2014 -0400
Annotate intentional case fallthrough, or add breaks.
Fix suspicious (*) switch cases found by Coverity by
annotating with either "/* FALLTHROUGH */" or adding a break;
The FALLTHROUGH was much more common, but the break turned
out to be the right choice in three spots. All changes tested
to work.
(*) suspicious = case1 + code + case2, and neither flow control
(like break) nor fallthrough annotation between code and case2.
Fix for Coverity perl5 CIDs 28977..28989, 45353.
M dist/Data-Dumper/Dumper.xs
M gv.c
M op.c
M pp_ctl.c
M regcomp.c
M toke.c
commit f571ea34c7765bd82297c03a323b01d65c4aa8af
Author: Tony Cook <t...@develop-help.com>
Date: Tue May 20 15:05:47 2014 +1000
bump $Devel::Peek::VERSION to 1.17
M ext/Devel-Peek/Peek.pm
commit 40e6adad549bff487a0c6f0de5aca86634020145
Author: Tony Cook <t...@develop-help.com>
Date: Tue May 20 15:05:31 2014 +1000
bump $Data::Dumper::VERSION to 2.152
M dist/Data-Dumper/Dumper.pm
commit 1c34e776c9b5325fa0aaa30acfe2d9df42971d58
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Mon May 19 07:24:57 2014 -0400
Insert asserts to paths suspected by Coverity.
Coverity suspects paths like this:
(1)
p = foo();
if (!p) p = bar();
baz(p->q);
since it cannot prove that p is always non-NULL at the dereference.
(2) Or simply something like:
mg = mg_find(...);
foo(mg->blah);
Since mg_find() can fail returning NULL.
Adding assert() calls before the dereferences. Testing with -DDEBUGGING.
Hopefully there are regular smokes doing the same.
Fix for Coverity perl5 CIDs
28950,28952..28955,28964,28967,28970..28795,49921:
CID ...: Dereference after null check (FORWARD_NULL)
var_deref_op: Dereferencing null pointer p->q
(TODO: Coverity perl5 CIDs 28962,28968,28969: the same issue, but would
conflict with already in-flight changes, prepare catch-up patch later.)
M dist/Data-Dumper/Dumper.xs
M dump.c
M ext/Devel-Peek/Peek.xs
M ext/XS-APItest/APItest.xs
M mg.c
M mro.c
M op.c
M perlio.c
M pp_hot.c
M regcomp.c
M regexec.c
M universal.c
M util.c
commit 869d5331c09861cdcef241c5706b08abeba16361
Author: Tony Cook <t...@develop-help.com>
Date: Tue May 20 15:06:33 2014 +1000
bump $XS::APItest::VERSION to 0.61
M ext/XS-APItest/APItest.pm
commit eda3564dcac906554e0b594a12321c095f7bbc23
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Fri May 9 10:34:09 2014 -0400
Unused dTHX.
Fix for Coverity perl5 CID 49935.
M ext/XS-APItest/APItest.xs
commit 46b0b1454429c21dd997917c19fcc0094a084c3d
Author: Tony Cook <t...@develop-help.com>
Date: Tue May 20 15:05:59 2014 +1000
bump $NDBM_File::VERSION to 0.15
M ext/NDBM_File/NDBM_File.pm
commit 0511934e5f36d2f29de78df3578597d8cfd1fc5a
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Thu May 8 17:17:02 2014 -0400
Printfed as %s, cast appropriately.
Fix for Coverity perl5 CID 29050.
M ext/NDBM_File/NDBM_File.xs
commit 00626dc34bb3156e5bb72c8d71d2c75957cec756
Author: Daniel Dragan <bul...@hotmail.com>
Date: Wed May 14 04:08:13 2014 -0400
fix sv_usepvn_flags's docs
Newx != malloc, mixing Newx and malloc leads to heap corruption on some
builds like Win32, use the official Perl API for allocating memory.
M sv.c
commit 1fed408b91a4d56985882723d821857c19b17309
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Tue May 6 11:18:49 2014 -0400
Use system default locale only if there is one.
Fix for Coverity perl5 CID 28946.
M locale.c
commit b59c39d189183216540a54b7017ea82fdda1a536
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Fri May 9 11:05:30 2014 -0400
Do not invert a NULL cp_list.
Fix for Coverity perl5 CID 28966.
M regcomp.c
commit 88ecf9a3a954ac54d389f26de0250043372c390a
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Tue May 6 12:50:55 2014 -0400
Quad_t and Uquad_t cannot unpack as NVs.
If IVSIZE >= 8, a Quad_t is always >= IV_MIN, and <= IV_MAX, and an
Uquad_t is always (>= 0 aka UV_MIN and) <= UV_MAX; they cannot escape
their quadness and be NVs. (This logic may fail if Quad_t is not 8
bytes, but then other things would no doubt fail.)
Also tighten the logic by adding HAS_QUAD, also for the pack case.
Fix for Coverity perl5 CID 28942.
M pp_pack.c
commit e99e4fe71eac5895efd27aa074dcf3e0330044c0
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Tue May 13 08:18:05 2014 -0400
Cannot rotl u32 (hek_hash) by 64 bits.
Fix for Coverity perl5 CID 28935:
Operands don't affect result (CONSTANT_EXPRESSION_RESULT)
result_independent_of_operands: (unsigned long)entry->hent_hek->hek_hash >>
47 /* 64 - 17 */ is 0 regardless of the values of its operands. This occurs as
the bitwise second operand of '|'.
M hv.c
commit b3b3ee5125c038717c047007bace1c6d09a5c432
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Mon May 19 06:52:24 2014 -0400
Use the C_ARRAY_LENGTH.
Use the C_ARRAY_LENGTH instead of sizeof(c_array)/sizeof(c_array[0])
or sizeof(c_array)/sizeof(type_of_element_in_c_array), and C_ARRAY_END
for c_array + C_ARRAY_LENGTH(c_array).
While doing this found potential off-by-one error in sv.c:Perl_sv_magic:
how > C_ARRAY_LENGTH(PL_magic_data)
should probably have been
how >= C_ARRAY_LENGTH(PL_magic_data)
No tests fail, but this seems to be more of an internal sanity check.
M dump.c
M handy.h
M patchlevel.h
M perl.c
M sv.c
M universal.c
M util.c
commit 0b81a8a0335fbf90a61aefcade5e8b30ecf8afd6
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Wed May 7 12:04:37 2014 -0400
Accessing array before its start is dubious.
Fix by petermartini.
Fix for Coverity perl5 CID 28909:
Out-of-bounds access (ARRAY_VS_SINGLETON)
ptr_arith: Using &unsliced_keysv as an array.
This might corrupt or misinterpret adjacent memory locations.
M pp.c
commit 7655198e4408eb7b3c0fdad5e192d25101e0e3d6
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Fri May 9 18:49:25 2014 -0400
si_names access one past the end.
Fix for Coverity perl5 CID 45359.
M deb.c
commit 1babc4623baf958b9e07db453ff245372878b177
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Wed Apr 30 07:49:07 2014 -0400
For ptr masking use PTRSIZE-1, not nested conditional statement.
M perl.c
commit 7448b3afdeab4495a6dbc4cbeb9f01bc3790ed4b
Author: Daniel Dragan <bul...@hotmail.com>
Date: Tue Apr 29 12:27:42 2014 -0400
cleanup on commit "Pseudo-fork dups arg array on argless calls"
less branching, less cpu intructions, easier to read
M sv.c
commit f760caa89984f643d829b39a84d564bddc5abeff
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Sun Apr 27 20:52:54 2014 -0400
Add PERL_UNUSED_RESULT() and use that instead of the V_Gconvert().
M perl.h
M sv.c
commit 2dcfa4d64645e51b859378198c9f155c80a0931e
Author: Tony Cook <t...@develop-help.com>
Date: Tue May 20 15:06:16 2014 +1000
bump $PerlIO::via::VERSION to 0.15
M ext/PerlIO-via/via.pm
commit 40588c6344fbe227e8a6975bfaa4433cd005c33c
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Wed Apr 23 12:53:42 2014 -0400
Fix for Coverity perl5 CIDs 29225, 29226, 29227, 29228, 29229: Missing
varargs init or cleanup (VARARGS) missing va_end: va_end was not called for foo.
Use of va_args must be finished off with va_end (in other words,
use of va_start or va_copy must be bracketed off with va_end).
In most platforms va_end is a no-op, but in some platforms it is
required for proper cleanup (or face stack smash, or memory leak).
M ext/PerlIO-via/via.xs
M mathoms.c
M perlio.c
M util.c
commit 7f77710cb807f151c21229cc0c251d5c4fb79003
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Thu Apr 24 12:23:18 2014 -0400
Fix for Coverity perl5 CID 29068: Insecure temporary file (SECURE_TEMP)
secure_temp: Calling mkstemp() without securely setting umask first.
The umask used for mkstemp should be secure, but umask 0600 has been
the required umask only since POSIX.1-2008. In glibc 2.06 and earlier
the default was 0666, which is not secure. And no explicit knowledge
of how well non-glibc platforms implement mkstemp. Better err on the
side security, so set the umask temporarily to 0600, and then restore it.
M perl.c
M perlio.c
commit 812add2007d7b08b15dd04c8484b1ff66208a40d
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Tue May 6 08:35:02 2014 -0400
O_BINARY versus O_TEXT.
O_BINARY and O_TEXT can be either different (mostly in Windowsy platforms,
and in some hybrids, and then they can be either effective or no-ops);
or equal (in UNIXy platforms, no-ops). If they are no-ops, and especially
if one or both of them are zeros, one cannot even test for them (bit-and),
without introducing dead/non-sensical code.
M op.c
M perl.h
M perlio.c
commit b244fdcc82c4f2133f07da2506b952a0c10759e4
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Fri Apr 25 08:25:52 2014 -0400
Fix for Coverity perl5 CID 28936: Wrong operator used
(CONSTANT_EXPRESSION_RESULT)
operator_confusion: ret->flags | 0x10 is always 1/true regardless of
the values of its operand. This occurs as the logical first operand of '&&'.
Did you intend to use '&' rather than '|'?
(confirmed by Karl Williamson)
M regcomp.c
commit 7125b431de27dbb5e4f3a141a2753055220ee3e0
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Wed May 7 09:26:52 2014 -0400
Uninitialized tmbuf.
Fix for Coverity perl5 CID 29088: Uninitialized scalar variable (UNINIT)
uninit_use: Using uninitialized value tmbuf.tm_year.
There is a code path that can lead to accessing uninitialized tmbuf:
when the too-small or too-large time inputs to gmtime/localtime
happen.
- make it so that the tm_year is used only on successful code path:
pp_sys.c
- add the gmtime failed / localtime failed errors to perldiag:
pod/perldiag.pod
- test those errors: t/op/time.t
M pod/perldiag.pod
M pp_sys.c
M t/op/time.t
commit 19a8a18f191d537f49bd7880af0fdf3e6a4b28b6
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Wed May 7 09:19:00 2014 -0400
Off-by-one in PL_fold_locale use.
Fix for Coverity perl5 CID 29033: Out-of-bounds read
(OVERRUN) overrun-local: Overrunning array PL_fold_locale of 256 bytes at
byte offset 256 using index c1 (which evaluates to 256).
- the "c1 > 256" was off-by-one, it needed to be "c1 > 255",
it could have caused the PL_fold_locale to be accessed one past the end,
at offset 256, but we have dodged the bullet thanks to the regex engine
optimizing the bad case away before we hit it (analysis by Karl
Williamson):
regexec.c
- comment fixes (pointed out by Karl Williamson): regexec.c
- add tests to nail down the behaviour of fold matching
for the last of Latin-1 (0xFF, lowercase which curiously does not have
uppercase within Latin-1). and the first pure Unicode: t/re/pat.t
M regexec.c
M t/re/pat.t
commit 42be2a1b53290f6c385255fed1ee85d29ab2c1d3
Author: Tony Cook <t...@develop-help.com>
Date: Tue May 13 11:05:56 2014 +1000
Doug Bell is now a perl author
M AUTHORS
commit 4ceb35ffa9d7df8e220f2e8a5cb592cd28d1df62
Author: Doug Bell <madcity...@gmail.com>
Date: Tue May 13 10:50:41 2014 +1000
move given/when ~~ note below item introducing it
The note explaining to use $c ~~ $_ instead of $_ ~~ $c was put under the
item
introducing all binary operators, and not the one mentioning the explicit
smartmatch operator.
M pod/perlsyn.pod
commit d80712fb4e40aa5efb56331f3c432f6ae40ab9e6
Author: Jarkko Hietaniemi <j...@iki.fi>
Date: Sat Apr 19 15:38:40 2014 -0400
Fixes for running a clean minitest (at least in OS X). - detect POSIX
locale support differently: t/op/lc.t, t/uni/fold.t - detect PerlIO::scalar
differently: t/op/pwent.t - bail out earlier if no dscl: t/op/op/pwent.t -
avoid wide character warning: t/op/tr.t
M t/op/lc.t
M t/op/pwent.t
M t/op/tr.t
M t/uni/fold.t
commit 8fd50a82ea6c3be30bee46ddaff76d7effbca443
Author: pjacklam <pjack...@online.no>
Date: Tue Apr 15 23:03:18 2014 +0200
Fix RT #62052: Math::BigFloat -> bdiv() in list context
In list context, Math::BigFloat -> bdiv() returns the quotient and
remainder.
By convention, this means that the quotient and remainder, q and r, after
dividing x by y, satisfies x = q*y + r. The current behaviour is to return
x/y
and r, which do not satisfy this relation, and - more importantly - is not
what
people expect.
dist/Math-BigInt/lib/Math/BigFloat.pm: Patched to fix this bug.
t/bigfltpm.inc and t/upgrade.inc: Fixed test cases to match the new
behaviour.
M dist/Math-BigInt/lib/Math/BigFloat.pm
M dist/Math-BigInt/t/bigfltpm.inc
M dist/Math-BigInt/t/upgrade.inc
commit c614e824f70584d09d8770196348cce73ee02f8c
Author: pjacklam <pjack...@online.no>
Date: Tue Apr 15 17:57:02 2014 +0200
Fix RT #43692: Math::BigFloat -> blog() is sometimes incorrect
dist/Math-BigInt/lib/Math/BigFloat.pm: Patched as suggested in RT #43692.
blog(x) would sometimes return blog(2*x) when the accuracy was greater than
70 digits.
t/biglog.t: Fixed incorrect use of ok(). Edited a test case so it fails
without
this patch.
M dist/Math-BigInt/lib/Math/BigFloat.pm
M dist/Math-BigInt/t/biglog.t
commit 8e81fff25658bcbc0224bedfc5172656e84488bd
Author: Tony Cook <t...@develop-help.com>
Date: Tue May 13 10:21:57 2014 +1000
bump dist/Math-BigInt/ $VERSIONs
M dist/Math-BigInt/lib/Math/BigFloat.pm
M dist/Math-BigInt/lib/Math/BigInt.pm
M dist/Math-BigInt/lib/Math/BigInt/Calc.pm
M dist/Math-BigInt/lib/Math/BigInt/CalcEmu.pm
commit 0c9b7bbc87d0f2b9f56504923f4a26ef2a80d47b
Author: Peter John Acklam <pjack...@online.no>
Date: Wed Apr 9 11:46:01 2014 +0200
Sync Math-BigInt blead vs. CPAN
Some POD changes, and a version number increase for consistency between all
.pm files, were accidentally included in the CPAN release of Math-BigInt
before they were included in blead. This patch gets the modules in sync.
M dist/Math-BigInt/lib/Math/BigFloat.pm
M dist/Math-BigInt/lib/Math/BigInt.pm
M dist/Math-BigInt/lib/Math/BigInt/Calc.pm
M dist/Math-BigInt/lib/Math/BigInt/CalcEmu.pm
commit 9a4167f79bae7b8f82b564e82e599291885f118a
Author: Tony Cook <t...@develop-help.com>
Date: Tue May 13 10:09:34 2014 +1000
bump Math::BigRat $VERSION
M dist/Math-BigRat/lib/Math/BigRat.pm
commit c7e212a2ee3670dd95f955f159e711e672ffb5fb
Author: Peter John Acklam <pjack...@online.no>
Date: Wed Apr 9 19:30:40 2014 +0200
Sync Math-BigRat blead vs. CPAN
Some POD changes were accidentally included in the CPAN release of
Math-BigRat before they were included in blead. This patch gets the
module in sync.
M dist/Math-BigRat/lib/Math/BigRat.pm
commit b6665025023a01f1bcf7f59e3f3c1701916a55df
Author: Tony Cook <t...@develop-help.com>
Date: Mon May 12 15:52:04 2014 +1000
Pierre Bogossian is now a perl author
M AUTHORS
commit effc69c50ba9b5b91bb75c764629273161479afd
Author: Pierre Bogossian <bogoss...@mail.com>
Date: Fri Apr 4 20:07:56 2014 +0200
Remove redundant test
M sv.c
commit 5e6f59ea3d632b9b5453d78c1aa91b139f0bcd51
Author: Karen Etheridge <et...@cpan.org>
Date: Thu Apr 3 11:08:20 2014 -0700
Ignore mismatched versions between Carp and Carp::Heavy if the version
discrepancy does not actually lead to broken code
Carp::Heavy's guts were merged into Carp in 1.12. After this point, if a
different version of Carp::Heavy is loaded than Carp, no harm will be done
and
the error can be ignored.
M dist/Carp/lib/Carp.pm
M dist/Carp/lib/Carp/Heavy.pm
commit 8ed94a84a633c1f0ae58486fb301a97f2fa971fb
Author: Shlomi Fish <shlo...@shlomifish.org>
Date: Thu Mar 27 19:07:50 2014 +0200
Fix RT #121509 : perl -d handling chdir().
See: https://rt.perl.org/Ticket/Display.html?id=121509
[perl #121509] perl debugger doesn't save starting dir to restart from
Thanks to Linda Walsh for reporting the problem and RJBS for commenting
for initial approval. Fix and a regression test by Shlomi Fish.
M MANIFEST
M lib/perl5db.pl
M lib/perl5db.t
A lib/perl5db/t/rt-121509-restart-after-chdir
commit 0674d84f54970631bd37182da19580f896891468
Author: Niels Thykier <ni...@thykier.net>
Date: Sat Mar 22 15:54:41 2014 +0100
pp_sys.c: Optimize out a call to PerlIO_get_cnt
If PerlIO_getc returns EOF, PerlIO_get_cnt is not going to return a
positive value all of a sudden.
Signed-off-by: Niels Thykier <ni...@thykier.net>
M pp_sys.c
commit 2f8b09c2c0b036f297eddc7c9a8ddd820040a559
Author: Daniel Dragan <bul...@hotmail.com>
Date: Thu Mar 13 02:51:51 2014 -0400
refactor av_delete
Some bad code generation/bad optimization from VC caused me to clean this
up. Dont write AvARRAY(av)[key] = NULL twice. It appears in 2 branches
just as written VC asm wise. Don't call sv_2mortal on a NULL SV*. It
works but less efficient. On VC2008 x64 -O1 this func dropped from 0x208
to 0x202 bytes of machine code after this patch.
M av.c
commit 6207b9908cc45ddb228877091d573ba39aadad06
Author: Rob Hoelz <r...@hoelz.ro>
Date: Sun Jun 9 21:25:09 2013 +0200
Implement get_fork_TTY for tmux
M lib/perl5db.pl
commit 7df3792e3ae5b2bd5cee801e4f92f3ad2279b477
Author: Tony Cook <t...@develop-help.com>
Date: Mon May 12 15:51:48 2014 +1000
Norman Koch is now a perl author
M AUTHORS
commit cf15aa33c2c6a98e671aa010195704c7d105f1d3
Author: Norman Koch <kochnor...@rocketmail.com>
Date: Sat May 4 23:18:31 2013 +0200
Clearified "require <>"-Message
Changed "<> should be quotes" to
"<> at require-statement should be quotes" when someone
writes "require <Module>", because this way it is way
easier to find the specific command that caused this.
M pod/perldiag.pod
M t/op/require_errors.t
M toke.c
commit ecb768a0c0e0215f0e3095de5e57a376978f1a2d
Author: Tony Cook <t...@develop-help.com>
Date: Mon May 12 13:55:36 2014 +1000
[perl #121112] only warn if newline is the last non-NUL character
M doio.c
M embed.fnc
M embed.h
M inline.h
M pp_sys.c
M proto.h
M t/lib/warnings/7fatal
M t/lib/warnings/doio
M t/lib/warnings/pp_sys
-----------------------------------------------------------------------
--
Perl5 Master Repository
