Father Chrysostomos [Thu, 16 Aug 2012 05:27:54 +0000 (22:27 -0700)]
Use PADLIST in more places
Much code relies on the fact that PADLIST is typedeffed as AV.
PADLIST should be treated as a distinct type.
Father Chrysostomos [Thu, 16 Aug 2012 05:11:46 +0000 (22:11 -0700)]
Move PAD(LIST) typedefs to perl.h
otherwise they can only be used in some header files.
Father Chrysostomos [Tue, 21 Aug 2012 23:39:10 +0000 (16:39 -0700)]
[Merge] Enter inline.h
This is a home for static inline functions that cannot go in other
headers because they depend on proto.h or struct definitions.
This allows us to avoid repeating macros with GCC and non-GCC ver-
sions. It also makes it easier to avoid evaluating macro argu-
ments twice.
I’ve moved just enough things into it to offset the additional lines
added by the comments at the top. The ‘net code removal’ of this
branch is 4 lines.
Father Chrysostomos [Sat, 18 Aug 2012 20:16:31 +0000 (13:16 -0700)]
Move S_CvDEPTHp from cv.h to inline.h; shrink macros
This allows us to use assert() inside S_CvDEPTHp, so we no longer need
GCC and non-GCC variants of the macro that calls it.
Father Chrysostomos [Sat, 18 Aug 2012 19:58:38 +0000 (12:58 -0700)]
Static inline functions for SvPADTMP and SvPADSTALE
This allows non-GCC compilers to have assertions and avoids
repeating the macros.
Father Chrysostomos [Sat, 18 Aug 2012 19:39:40 +0000 (12:39 -0700)]
Use fast SvREFCNT_dec for non-GCC
Father Chrysostomos [Sat, 18 Aug 2012 19:34:33 +0000 (12:34 -0700)]
Use static inline functions for SvREFCNT_inc
This avoids the need to repeat the macros in GCC and non-GCC versions.
For non-GCC compilers capable of inlining, this should speed things up
slightly, too, as PL_Sv is no longer needed.
Father Chrysostomos [Fri, 17 Aug 2012 04:54:53 +0000 (21:54 -0700)]
[perl #113718] Add inline.h
We can put static inline functions here, and they can depend on
function prototypes and struct definitions from other header
files.
Chris 'BinGOs' Williams [Tue, 21 Aug 2012 22:55:41 +0000 (23:55 +0100)]
Sync Module-CoreList in Maintainers.pl for CPAN release
Chris 'BinGOs' Williams [Tue, 21 Aug 2012 22:46:13 +0000 (23:46 +0100)]
Update Changes fr Module-CoreList and bump to version 2.72
Father Chrysostomos [Tue, 21 Aug 2012 21:13:02 +0000 (14:13 -0700)]
[Merge] Here-doc parsing
I was waiting for 5.17.3 to be released, before merging my work on
padlists (which is blocking lexical subs), since I thought it would be
mean to inflict it on blead at the last minute before a release.
So, in the mean time, I decided to fix a small here-doc parsing bug,
that prevented them from occurring inside regexp code blocks.
As often happens, it turned out to be more involved than that....
I ended up writing a history of here-doc parsing, which you can find
in the commit message for
5097bf9b8d, which shows that the way they
have interacted with other quote-like operators (or other here-docs)
has changed over time in interesting ways.
While I was fixing those, I started to find other bugs. Since I was
modifying the code, I decided to try applying David Nicol’s patch that
allows a here-doc terminator with no newline after it, to avoid creat-
ing more conflicts through my changes. The patch didn’t work. And
while I was resolving what conflicts there were, I figured out a sim-
pler approach. So, instead of trying to investigate into why the
patch didn’t work, I just wrote my own version, which used less code.
Instead of working back on error to try to see whether we could have
accepted a terminator without a newline, we can just tack a newline on
the string buffer at EOF and let the rest of the code handle it the
usual way.
I continued to find more bugs as I went, till my ‘Yay, another bug!’
started to become ‘What? *Another* bug?’.
In the end:
• I fixed here-doc parsing, such that the body starts on the line fol-
lowing the <<foo marker, regardless of whether it is inside quotes,
string evals, or what have you (but see remaining bugs below). This
was contrary to the documentation, but the documentation was actu-
ally wrong half the time, so I corrected it.
• Here-doc terminators no longer require a final newline at EOF.
• You no longer get crashes with edge cases.
• Nulls in comments no longer confuse the here-doc parser.
And, finally, one bug that I fixed was not related to here-docs per
se, but got in the way. It deserves its own JAPH:
s/${s|||, \""}Just another Perl hacker,
/anything/;
print
There are still two bugs remaining:
• Here-docs whose markers occur in single-line s/// patterns where the
replacement part is multi-line or starts on a subsequent line are
still screwed.
• CR and CR LF line terminators are treated inconsistently inside and
outside of string evals.
I’ve decided to set those aside for later and merge what I’ve
done so far.
Father Chrysostomos [Tue, 21 Aug 2012 21:09:51 +0000 (14:09 -0700)]
perlop.pod: Update here-doc-in-quotes parsing rules
Father Chrysostomos [Tue, 21 Aug 2012 08:11:34 +0000 (01:11 -0700)]
smoke-me diag
nt,hun
Father Chrysostomos [Tue, 21 Aug 2012 08:45:15 +0000 (01:45 -0700)]
toke.c:scan_heredoc: Use PL_tokenbuf less
When scanning for a heredoc terminator in a string eval or quote-like
operator, the first character we are looking for is always a newline.
So instead of setting term to *PL_tokenbuf in those two code paths,
we can just hard-code '\n'.
Father Chrysostomos [Tue, 21 Aug 2012 06:58:59 +0000 (23:58 -0700)]
Fix substitution in substitution pattern
Guess what this prints:
s/${s|||, \""}Just another Perl hacker,
/anything/;
print
And look at this:
$ perl5.6.2 -e 's/${s|||;\""}/foo\n/; print;'
$ perl5.16.0 -e 's/${s|||;\""}/foo\n/; print;'
$ perl5.17.2 -e 's/${s|||;\""}/foo\n/; print;'
Bus error
$ ./miniperl -e 's/${s|||;\""}/foo\n/; print;'
Bus error
The first two gave no output, though they should have shown "foo".
And bleadperl now crashes.
When the lexer parses a quote-like operator, it begins by extracting
what is between the quotes. It puts it in an SV stored in the varia-
ble PL_lex_stuff. Then, if it is y/// or s///, it scans the replace-
ment part and puts it in an SV in PL_lex_repl. When it finishes with
it, it sets PL_lex_repl to NULL.
Now, if you put s/// in the pattern part of s/// (or y in s), the
inner s/// will clobber PL_lex_repl with its own replacement string.
So, when the outer s/// finish parsing its pattern and wants its
replacement string. If it is not there, it assumes it has already
parsed it (whether PL_lex_repl is set is how it remembers which half
of s/// it is parsing), and proceeds to feed bad code to the parser,
resulting in a bad op tree.
PL_lex_repl needs to be localised when a quote-like operator is
parsed. Since localisation for quote-like operators happens in a sep-
arate yylex call (yylex calls sublex_push, which does it) after the
string delimiters are found, at which point PL_lex_repl has already
been set (clobbering the previous value), we change the delim-
iter-scanning code (scan_{str,trans,subst}) to use the new
PL_sublex_info.repl, which sublex_push now copies into PL_lex_repl
after localising the latter.
Father Chrysostomos [Tue, 21 Aug 2012 02:08:57 +0000 (19:08 -0700)]
Fix here-docs in nested quote-like operators
When the lexer encounters a quote-like operator, it extracts the con-
tents of the quotes and starts an inner lexing scope.
To handle eval "s//<<FOO/e\n...", the here-doc parser peeks into the
outer lexing scope’s PL_linestr (current line buffer, which inside an
eval contains the entire string of code being parsed; for quote-like
operators, that is where the contents of the quote are stored). It
only does this inside a string eval. When parsing a file, the input
comes in one line at a time. So the here-doc parser steals lines from
the input stream for s//<<FOO/e outside an eval.
This approach fails in this case, as the peekee is the linestr for
s///, not for the eval:
eval ' s//"${\<<END}"/e; print
Just another Perl hacker,
END
'or die $@
__END__
Can't find string terminator "END" anywhere before EOF at (eval 1) line 1.
We also need to do this peeking stuff outside of a string eval, to
solve this:
s//"${\<<END}"
Just another Perl hacker,
END
/e; print
__END__
Can't find string terminator "END" anywhere before EOF at - line 1.
In the first example above, we need to look not in the parent lexing
scope’s linestr, but in that of the grandparent.
To solve the second example, we need to check whether the outer lexing
scope is a quote-like operator when we are not in an eval.
For parsing here-docs in quotes in eval, we currently store two
things, the former buffer pointer and the former linestr, in
PL_sublex_info.super_{bufp,lines}tr. The values for upper scopes are
stashed away on the savestack somewhere.
We need to be able to iterate through the outer lexer scopes till we
find one with multiple lines. Retrieving the information from the
savestack would be too complex and error-prone.
Since PL_linestr is an SV, we can abuse a couple of fields in it.
Upgrading it to PVNV gives it both IVX and NVX fields, which are big
enough to store pointers.
IVX is already used to hold an op number. So for the innermost quoted
scope we still need to use PL_sublex_info.super_bufptr. When entering
a new lexing scope (in sublex_push), we can localise the IVX field of
the outer PL_linestr SV and set it to what PL_sublex_info.super_bufptr
was in that scope. SvIVX(linestr) is only used for an op number when
that linestr’s lexing scope is the innermost one.
PL_sublex_info.super_linestr can be eliminated and replaced with
SvNVX(PL_linestr).
Father Chrysostomos [Tue, 21 Aug 2012 01:06:41 +0000 (18:06 -0700)]
Don’t use strchr when scanning for newline after <<foo
The code that uses this is specifically for parsing <<foo inside a
quote-like operator inside a string eval.
This prints bar:
eval "s//<<foo/e
bar
foo
";
print $_ || $@;
This prints Can't find string terminator blah blah blah:
eval "s//<<foo/e #\0
bar
foo
";
print $_ || $@;
Nulls in comments are allowed elsewhere. This prints bar:
eval "\$_ = <<foo #\0
bar
foo
";
print $_ || $@;
The problem with strchr is that it is specifically for scanning null-
terminated strings. If embedded nulls are permitted (and should be in
this case), memchr should be used.
This code was added by
0244c3a403.
David Nicol [Mon, 20 Aug 2012 23:22:15 +0000 (16:22 -0700)]
[perl #65838] perlop: remove caveat here-doc without newline
Father Chrysostomos [Mon, 20 Aug 2012 21:55:09 +0000 (14:55 -0700)]
here-doc in quotes in multiline s//.../e in eval
When <<END occurs on the last line of a quote-like operator inside a
string eval ("${\<<END}"), it peeks into the linestr buffer of the
parent lexing scope (quote-like operators start a new lexing scope
with the linestr buffer containing what is between the quotes) to find
the body of the here-doc. It modifies that buffer, stealing however
much it needs.
It was not leaving things in the consistent state that s///e checks
for when it finishes parsing the replacement (to make sure s//}+{/
doesn’t ‘work’). Specifically, it was not shrinking the parent buf-
fer, so when PL_bufend was reset in sublex_done to the end of the par-
ent buffer, it was pointing to the wrong spot.
Father Chrysostomos [Mon, 20 Aug 2012 19:57:29 +0000 (12:57 -0700)]
heredoc after "" in s/// in eval
This works fine:
eval ' s//<<END.""/e; print
Just another Perl hacker,
END
'or die $@
__END__
Just another Perl hacker,
But this doesn’t:
eval ' s//"$1".<<END/e; print
Just another Perl hacker,
END
'or die $@
__END__
Can't find string terminator "END" anywhere before EOF at (eval 1) line 1.
It fails because PL_sublex_info.super_buf*, added by commit
0244c3a403, are not localised, so, after the "", s/// sees its own
buffer pointers in those variables, instead of its parent string eval.
This used to happen only with s///e inside s///e, but that was because
here-docs would peek inside the parent linestr buffer only inside
s///e, and not other quote-like operators. That was fixed in
recent commits.
Simply moving the assignment of super_buf* into sublex_push does solve
the bug for a simple "", as "" does sublex_start, but not sublex_push.
We do need to localise those variables for "${\''}", however.
David Nicol [Mon, 20 Aug 2012 06:05:40 +0000 (23:05 -0700)]
toke.c:S_scan_heredoc: Add comment about <<\FOO
Father Chrysostomos [Mon, 20 Aug 2012 06:05:06 +0000 (23:05 -0700)]
[perl #65838] Allow here-doc with no final newline
When reading a line of input while scanning a here-doc, if the line
does not end in \n, then we know we have reached the end of input. By
simply tacking a \n on to the buffer, we can meet the expectations of
the rest of the here-doc parsing code. If it turns out the delimiter
is not found on that line, it does not matter that we modified it, as
we will croak anyway.
I had to add a new flag to lex_next_chunk. Before commit
f0e67a1d2,
S_scan_heredoc would read from the stream itself, without closing any
handles. So the next time through yylex, the eof code would supply
the final implicit semicolon.
Since
f0e67a1d2, S_scan_heredoc has been calling lex_next_chunk, which
takes care of reading from the stream an supply any final ; at eof.
The here-doc parser will just get confused as a result (<<';' would
work without any terminator). The new flag tells lex_next_chunk not
to do anything at eof (not even closing handles and resetting the
parser state), but to return false and leave everything as it was.
Father Chrysostomos [Mon, 20 Aug 2012 05:41:08 +0000 (22:41 -0700)]
heredoc.t: Suppress deprecation warnings
Michael G. Schwern [Fri, 12 Jun 2009 22:35:00 +0000 (15:35 -0700)]
Clean up heredoc.t
* Made the tests more independent, mostly by decoupling the use of
a single $string. This will make it easier to expand on the test file
later.
* Replace ok( $foo eq $bar ) with is() for better diagnostics
* Remove unnecessary STDERR redirection. fresh_perl does that for you.
* fix fresh_perl to honor progfile and stderr arguments passed in
rather than just blowing over them
David Nicol [Mon, 20 Aug 2012 05:16:13 +0000 (22:16 -0700)]
[perl #65838] Tests for here-docs without final newlines
and a few error cases
Father Chrysostomos [Sun, 19 Aug 2012 09:45:38 +0000 (02:45 -0700)]
[perl #114040] Parse here-docs correctly in quoted constructs
When parsing code outside a string eval or quoted construct, the lexer
reads one line at a time into PL_linestr.
To parse a here-doc (hereinafter ‘deer hock’, because I spike lunar-
isms), the lexer has to pull extra lines out of the input stream ahead
of the current line, the value of PL_linestr remaining the same.
In a string eval, the entire piece of code being parsed is in
PL_linestr.
To parse a deer hock inside a string eval, the lexer has to fiddle
with the contents of PL_linestr, scanning for newline characters.
Originally, S_scan_heredoc just followed those two approaches.
When the lexer encounters a quoted construct, it looks for the end-
ing delimiter (reading from the input stream if necessary), puts the
entire quoted thing (minus quotes) in PL_linestr, and then starts an
inner lexing scope.
This means that deer hocks would not nest properly outside of a string
eval, because the body of the inner deer hock would be pulled out of
the input stream *after* the outer deer hock.
Larry Wall fixed that in commit
fd2d095329 (Jan. 1997), so that this
would work:
<<foo
${\<<bar}
ber
bar
foo
He did so by following the string eval approach (looking for the deer
hock body in PL_linestr) if the deer hock was inside another quoted
construct.
Later, commit
a2c066523a (Mar. 1998) fixed this:
s/^not /substr(<<EOF, 0, 0)/e;
Ignored
EOF
by following the string eval approach only if the deer hock was inside
another non-backtick deer hock, not just any quoted construct.
The problem with the string eval approach inside a substitu-
tion is that it only looks in PL_linestr, which only contains
‘substr(<<EOF, 0, 0)’ when the lexer is handling the second part of
the s/// operator.
But that unfortunately broke this:
s/^not /substr(<<EOF, 0, 0)
Ignored
EOF
/e;
and this:
print <<`EOF`;
${\<<EOG}
echo stuff
EOG
EOF
reverting it to the pre-
fd2d095329 behaviour, because the outer quoted
construct was treated as one line.
Later on, commit
0244c3a403 (Mar. 1999) fixed this:
eval 's/.../<<FOO/e
stuff
FOO
';
which required a new approach not used before. When the replacement
part of the s/// is being parsed, PL_linestr contains ‘<<FOO’. The
body of the deer hock is not in the input stream (there isn’t one),
but in what was the previous value of PL_linestr before the lexer
encountered s///.
So
0244c3a403 fixed that by recording pointers into the outer string
and using them in S_scan_heredoc. That commit, for some reason, was
written such that it applied only to substitutions, and not to other
quoted constructs.
It also failed to take interpolation into account, and did not record
the outer buffer position, but then tried to use it anyway, resulting
in crashes in both these cases:
eval 's/${ <<END }//';
eval 's//${ <<END }//';
It also failed to take multiline s///’s into account, resulting in
neither of these working, because it lost track of the current cursor,
leaving it at 'D' instead of the line break following it:
eval '
s//<<END
/e;
blah blah blah
END
;1' or die $@;
eval '
s//<<END
blah blah blah
END
/e;
;1' or die $@;
S_scan_heredoc currently positions the cursor (s) at the last charac-
ter of <<END if there is a line break on the same line. There is an
s++ later on to account, but the code added by
0244c3a403 bypassed it.
So, in the end, deer hocks could only be nested in other quoted con-
structs if the outer construct was in a string eval and was not s///,
or was a non-backtick deer hock.
This commit hopefully fixes most of the problems. :-)
The s///-in-eval case is a little tricky. We have to see whether the
deer hock label is on the last line of the s///. If it is, we have
to peek into the outer buffer. Otherwise, we have to treat it like a
string eval.
This commit does not deal with <<END inside the pattern of a multi-
line s/// or in nested quotes.
Father Chrysostomos [Sun, 19 Aug 2012 06:54:02 +0000 (23:54 -0700)]
[perl #70836] Fix err msg for unterminated here-doc in eval
$ perl -e '<<foo'
Can't find string terminator "foo" anywhere before EOF at -e line 1.
$ perl -e 'eval "<<foo"; die $@'
Can't find string terminator "
foo" anywhere before EOF at (eval 1) line 1.
An internal implementation detail is leaking out.
When the lexer happens to have a multiline string in its line buffer
(in a string eval or quoted construct), it looks for "\nfoo" instead
of "foo". It was passing that same string to the error-reporting code
(S_missingterm), resulting in that extraneous newline.
Father Chrysostomos [Tue, 21 Aug 2012 15:25:13 +0000 (08:25 -0700)]
Increase $Module::CoreList::TieHashDelta::VERSION to 2.72
Father Chrysostomos [Tue, 21 Aug 2012 15:24:16 +0000 (08:24 -0700)]
[rt.cpan.org #79109] Avoid each $scalar in TieHashDelta.pm
This is dual-life, after all.
David Mitchell [Tue, 21 Aug 2012 09:55:00 +0000 (10:55 +0100)]
utf8 pos cache: always keep most recent value
UTF-8 strings may have magic attached that caches up to two byte position
to char position (or vice versa) mappings.
When a third position has been calculated (e.g. via sv_pos_b2u()), the
code has to decide how to update the cache: i.e. which value to discard.
Currently for each of the three possibilities, it looks at what would be
the remaining two values, and calculates the RMS sum of the three
distances between ^ ... cache A .. cache B ... $. Whichever permutation
gives the lowest result is picked. Note that this means that the most
recently calculated value may be discarded.
This makes sense if the next position request will be for a random part of
the string; however in reality, the next request is more likely to be for
the same position, or one a bit further along. Consider the following
innocuous code:
$_ = "\x{100}" x 1_000_000;
$p = pos while /./g;
This goes quadratic, and takes 150s on my system. The fix is is to always
keep the newest value, and use the RMS calculation only to decide which of
the two older values to discard. With this fix, the above code takes 0.4s.
The test suite takes the same time in both cases, so there's no obvious
slowdown elsewhere with this change.
Chris 'BinGOs' Williams [Tue, 21 Aug 2012 06:53:01 +0000 (07:53 +0100)]
Restore MANIFEST entry for Module::CoreList, sync with CPAN version
Craig A. Berry [Tue, 21 Aug 2012 00:15:23 +0000 (19:15 -0500)]
Consistent unixy path handling in File::Find::_find_opt.
Back in
a1ccf0c4149b we converted the current working directory to
Unix format on VMS, but neglected to change what later gets pasted
onto it with a hard-coded slash delimiter. The resulting mongrel
filespec was invalid and of course would not appear to exist even
if the file did exist under a properly assembled name.
So this commit makes the use of Unix-style paths on VMS within
_find_opt consistent.
The bug was tickled by a recent change to Module::Pluggable, whose
tests and the tests of other modules that depend on it started
failing en masse.
jkeenan [Sat, 11 Aug 2012 00:22:13 +0000 (20:22 -0400)]
Implement name change in POD example; Chris Waggoner++.
For: RT #114314.
Steve Hay [Mon, 20 Aug 2012 16:48:47 +0000 (17:48 +0100)]
RMG - update commit reference for version bump change
Still refer to a .0 bump since it contains more than other bumps, but
update it to 5.17.0's bump rather than 5.15.0's
Steve Hay [Mon, 20 Aug 2012 16:20:23 +0000 (17:20 +0100)]
Bump version to 5.17.4
Steve Hay [Mon, 20 Aug 2012 16:03:22 +0000 (17:03 +0100)]
RMG - update commit reference for new perldelta change
Refer to one that doesn't mention pod.lst since that is gone now.
Steve Hay [Mon, 20 Aug 2012 15:57:26 +0000 (16:57 +0100)]
Make new perldelta for 5.17.4
Steve Hay [Mon, 20 Aug 2012 15:30:52 +0000 (16:30 +0100)]
Undo VERSION bump for undone code
Commit
78ed4cf4d6 undid the accidental effect of
eb578fdb55 on OS2::REXX
but forgot to revert the accompanying VERSION bump, which is not otherwise
required since nothing else has changed.
Steve Hay [Mon, 20 Aug 2012 15:08:37 +0000 (16:08 +0100)]
Correct announcement date for 5.17.2's epigraph
Steve Hay [Mon, 20 Aug 2012 15:07:56 +0000 (16:07 +0100)]
Add epigraph for 5.17.3
Steve Hay [Mon, 20 Aug 2012 14:31:24 +0000 (15:31 +0100)]
Merge branch 'release-5.17.3' into blead
H.Merijn Brand [Mon, 20 Aug 2012 13:25:02 +0000 (15:25 +0200)]
Add the new smoke report test site
Better late than never!
Steve Hay [Mon, 20 Aug 2012 11:09:23 +0000 (12:09 +0100)]
Fix Module::CoreList test - TieHashDelta is to be expected too now
Steve Hay [Mon, 20 Aug 2012 10:36:53 +0000 (11:36 +0100)]
Add 5.17.3 to perlhist
Steve Hay [Mon, 20 Aug 2012 10:34:32 +0000 (11:34 +0100)]
Upgrade Module-CoreList to 2.71
Jesse Luehrs [Mon, 20 Aug 2012 10:18:00 +0000 (05:18 -0500)]
fix accidentally modified comment
Steve Hay [Mon, 20 Aug 2012 10:12:03 +0000 (11:12 +0100)]
perldelta - finalize with acknowledgements for 5.17.3
Steve Hay [Mon, 20 Aug 2012 10:08:48 +0000 (11:08 +0100)]
perldelta - Fix unescaped <>
Steve Hay [Mon, 20 Aug 2012 09:21:35 +0000 (10:21 +0100)]
Update RMG - note sync-with-cpan is untested on Windows
Steve Hay [Mon, 20 Aug 2012 08:32:19 +0000 (09:32 +0100)]
perldelta - Remove XXX sections ready for 5.17.3 release
Steve Hay [Sun, 19 Aug 2012 11:53:47 +0000 (12:53 +0100)]
Upgrade to Sys-Syslog-0.31
Steve Hay [Sun, 19 Aug 2012 11:27:06 +0000 (12:27 +0100)]
Corrections to Maintainers.pl and perldelta.pod for Text-Tabs+Wrap
Steve Hay [Sun, 19 Aug 2012 10:51:52 +0000 (11:51 +0100)]
Upgrade to Text-Tabs+Wrap-2012.0818
This incorporates earlier blead customizations to t/fill.t and t/tabs.t
Steve Hay [Sun, 19 Aug 2012 10:31:51 +0000 (11:31 +0100)]
Upgrade Module-Metadata to 1.000011
Steve Hay [Sun, 19 Aug 2012 10:24:53 +0000 (11:24 +0100)]
Upgrade Module-Build to 0.4003
Karl Williamson [Thu, 16 Aug 2012 16:50:14 +0000 (10:50 -0600)]
Omnibus removal of register declarations
This removes most register declarations in C code (and accompanying
documentation) in the Perl core. Retained are those in the ext
directory, Configure, and those that are associated with assembly
language.
See:
http://stackoverflow.com/questions/314994/whats-a-good-example-of-register-variable-usage-in-c
which says, in part:
There is no good example of register usage when using modern compilers
(read: last 10+ years) because it almost never does any good and can do
some bad. When you use register, you are telling the compiler "I know
how to optimize my code better than you do" which is almost never the
case. One of three things can happen when you use register:
The compiler ignores it, this is most likely. In this case the only
harm is that you cannot take the address of the variable in the
code.
The compiler honors your request and as a result the code runs slower.
The compiler honors your request and the code runs faster, this is the least likely scenario.
Even if one compiler produces better code when you use register, there
is no reason to believe another will do the same. If you have some
critical code that the compiler is not optimizing well enough your best
bet is probably to use assembler for that part anyway but of course do
the appropriate profiling to verify the generated code is really a
problem first.
Steve Hay [Sat, 18 Aug 2012 13:10:03 +0000 (14:10 +0100)]
Tweaks to RMG
Use the simpler syntax for starting the CPAN shell. Remove notes about
needing Unix tools on Windows for CPAN and CPANPLUS when LWP is not
installed: these are not required since the likes of Net::FTP and
HTTP::Tiny are used instead.
Steve Hay [Sat, 18 Aug 2012 11:28:32 +0000 (12:28 +0100)]
Don't use /dev/tty if it happens to exist on Windows
This fixes CPAN RT#79001 and CPAN RT#79064.
Steve Hay [Sat, 18 Aug 2012 09:39:56 +0000 (10:39 +0100)]
We don't support compilers other than MS VC++ and MinGW/gcc on Windows
Steve Hay [Sat, 18 Aug 2012 09:36:12 +0000 (10:36 +0100)]
Remove two unused #defines
Steve Hay [Sat, 18 Aug 2012 09:33:13 +0000 (10:33 +0100)]
We don't support MS VC++ < 6.0
Father Chrysostomos [Sat, 18 Aug 2012 06:20:53 +0000 (23:20 -0700)]
parser.t: Correct test count
Why do I keep making these mistakes? :-(
Father Chrysostomos [Fri, 17 Aug 2012 23:52:50 +0000 (16:52 -0700)]
sv.h: Don’t repeat _XPV_HEAD
Father Chrysostomos [Fri, 17 Aug 2012 23:54:40 +0000 (16:54 -0700)]
write.t: Eek! debugging code
Father Chrysostomos [Fri, 17 Aug 2012 23:44:57 +0000 (16:44 -0700)]
perldelta entries
Father Chrysostomos [Fri, 17 Aug 2012 21:45:29 +0000 (14:45 -0700)]
[perl #114040] Allow pod in quoted constructs
When the case = in toke.c:yylex is reached and PL_lex_state is
not LEX_NORMAL, that means we are in some sort of quoted construct,
and the entire construct’s content is in the current line buffer
(which, consequently contains more than one line). So we need to check
that when encountering pod. Quoted constructs need to be treated the
same way as string eval, which also puts all the code in the line
buffer.
Father Chrysostomos [Fri, 17 Aug 2012 21:24:05 +0000 (14:24 -0700)]
Don’t leak formats defined inside subs
I made them leak inadvertently in 5.17.2 with commit
e09ac076a1da.
This was unfortunately backported to 5.16.1 (as
3149499832) without
anybody noticing the bug.
Father Chrysostomos [Fri, 17 Aug 2012 20:39:27 +0000 (13:39 -0700)]
pad.c: Document pad_add_anon’s refcounting
Father Chrysostomos [Fri, 17 Aug 2012 20:28:46 +0000 (13:28 -0700)]
perldelta: Clarify note about B::PADLIST
Father Chrysostomos [Fri, 17 Aug 2012 06:44:11 +0000 (23:44 -0700)]
pp_ctl.c:pp_dbstate: Don’t adjust CvDEPTH for XSUBs
Commit
c127bd3aaa5c5 made XS DB::DB subs work. Before that,
pp_dbstate assumed DB::DB was written it perl. It adjusts CvDEPTH
when calling the XSUB, which serves no purpose. It was presumably
just copied from the pure-Perl-calling code. pp_entersub does-
n’t do this.
Steve Hay [Fri, 17 Aug 2012 23:33:53 +0000 (00:33 +0100)]
Restore VC++ 6 build on Windows
Commits
bb02a38feb and
1bd3586145 resulted in VC++ 6 complaining "error
C2099: initializer is not a constant" when initializing bodies_by_type in
sv.c. Workaround the apparent compiler bug using a patch from Jan Dubois,
amended to be compiler-specific as suggested by Nicholas Clark since
anonymous unions are not valid C89.
Date: Wed, 15 Aug 2012 00:55:06 -0700
Message-ID: <
005a01cd7abb$
498294e0$
dc87bea0$@activestate.com>
Steve Hay [Fri, 17 Aug 2012 17:41:02 +0000 (18:41 +0100)]
Upgrade to B::Debug 1.18
Craig A. Berry [Fri, 17 Aug 2012 16:05:14 +0000 (11:05 -0500)]
newCONSTSUB needs its own CV.
It had been using one called simply C<cv> but that name is already
taken in the (opaque) argument list generated by the XS_EUPXS
wrapper around the function name. And that cv is actually used
by boilerplate code generated from PPCODE, but only when there is
an ALIAS section present, which there wasn't before
c0810f8ef84,
but now is.
So declare and use our own CV and leave the one passed in alone.
Nicholas Clark [Tue, 14 Aug 2012 09:54:48 +0000 (11:54 +0200)]
Remove the UTS port.
UTS was a mainframe version of System V created by Amdahl, subsequently sold
to UTS Global. The port has not been touched since before 5.8.0, and UTS
Global is now defunct.
Steve Hay [Fri, 17 Aug 2012 08:20:50 +0000 (09:20 +0100)]
perldelta - Add remaining changes not yet documented
Also some other general tidying and wrapping.
Steve Hay [Thu, 16 Aug 2012 13:36:32 +0000 (14:36 +0100)]
CGI's t/url.t is no longer customized
The upgrade to 3.60 contained the changes which were in blead.
Steve Hay [Thu, 16 Aug 2012 08:25:35 +0000 (09:25 +0100)]
perldelta - Fix mistakes
Steve Hay [Thu, 16 Aug 2012 08:11:00 +0000 (09:11 +0100)]
perldelta - Document Windows-specific changes
Steve Hay [Thu, 16 Aug 2012 07:30:57 +0000 (08:30 +0100)]
Upgrade to CGI 3.60
There were already no t/lib/Test or cgi-lib_porting.html files, so these
can be removed from EXCLUDED.
Father Chrysostomos [Thu, 16 Aug 2012 03:34:04 +0000 (20:34 -0700)]
perldelta for
a444d2d4f37/#114368
Father Chrysostomos [Thu, 16 Aug 2012 03:32:50 +0000 (20:32 -0700)]
perldelta for
fc33dad25ea/#114020
This was not the main topic of #114020, but all the discussion
surrounding the change is in that ticket.
Father Chrysostomos [Thu, 16 Aug 2012 03:30:47 +0000 (20:30 -0700)]
perldelta: Tweak wording
I don’t know what (or whether) I was thinking when I wrote that.
Father Chrysostomos [Thu, 16 Aug 2012 03:29:15 +0000 (20:29 -0700)]
perldelta: missing bug number
Father Chrysostomos [Thu, 16 Aug 2012 03:27:55 +0000 (20:27 -0700)]
perldelta: more format stuff
Steve Hay [Wed, 15 Aug 2012 23:31:32 +0000 (00:31 +0100)]
Upgrade Module-Pluggable to 4.3
There is also no need to list Build.PL as EXCLUDED since it is IGNORABLE
anyway.
Steve Hay [Wed, 15 Aug 2012 23:12:09 +0000 (00:12 +0100)]
Upgrade perlfaq to 5.0150041
Steve Hay [Wed, 15 Aug 2012 22:59:29 +0000 (23:59 +0100)]
Upgrade Socket to 2.004
Steve Hay [Wed, 15 Aug 2012 16:56:13 +0000 (17:56 +0100)]
Upgrade Socket from 2.002 to 2.003
Steve Hay [Wed, 15 Aug 2012 16:54:23 +0000 (17:54 +0100)]
perldelta - Wrap to 79 columns
Steve Hay [Wed, 15 Aug 2012 16:52:43 +0000 (17:52 +0100)]
perldelta - Document Module::Pluggable changes
Nicholas Clark [Tue, 14 Aug 2012 10:22:45 +0000 (12:22 +0200)]
Tweak the test from
35f7559499c4a614 to work with PERL_UNICODE set.
Steve Hay [Wed, 15 Aug 2012 13:51:09 +0000 (14:51 +0100)]
perldelta
Remove duplicate note about B::Concise, fix a typo and note that
Module-Pluggable is now upgraded.
Steve Hay [Wed, 15 Aug 2012 13:04:18 +0000 (14:04 +0100)]
Upgrade to Module-Pluggable 4.2
The core build process cannot use Build.PL since Module::Build and/or its
prerequisites may not have been built yet, so EXCLUDE that and retain our
(already CUSTOMIZED) Makefile.PL instead for now.
Steve Hay [Wed, 15 Aug 2012 08:23:07 +0000 (09:23 +0100)]
Add new Win32 test script to MANIFEST
The file was added by
c3c06741ad.
Steve Hay [Wed, 15 Aug 2012 08:02:12 +0000 (09:02 +0100)]
Fix t/op/magic.t on Windows
These tests have been failing since they were added by
613c63b465, but
we can now fix them using new Win32 APIs.
Steve Hay [Wed, 15 Aug 2012 07:56:50 +0000 (08:56 +0100)]
perldelta - note changes in Win32 0.45
Steve Hay [Wed, 15 Aug 2012 07:52:27 +0000 (08:52 +0100)]
perldelta - note changes in Encode 2.47
Steve Hay [Wed, 15 Aug 2012 07:25:20 +0000 (08:25 +0100)]
Upgrade to Sys-Syslog 0.30
This now incorporates blead's customization of t/syslog.t. Also,
win32/PerlLog_RES.uu shouldn't be listed as EXCLUDED since it is
actually included!
Jan Dubois [Tue, 14 Aug 2012 23:34:32 +0000 (16:34 -0700)]
Update to Win32-0.45 from CPAN
Chris 'BinGOs' Williams [Wed, 15 Aug 2012 06:47:38 +0000 (07:47 +0100)]
Update Encode to CPAN version 2.47
[DELTA]
$Revision: 2.47 $ $Date: 2012/08/15 05:36:16 $
! Encode.pm
POD Fixes: Copyright and mail address
! Makefile.PL
Added LICENSE => 'perl'
! lib/Encode/GSM0338.pm t/gsm0338.t
REALLY fixed RT#75670: Wrong decoding for GSM 3.38 character \x09
ucm/gsm0338.ucm is dropped from MANIFEST since 2.25
but I was fixing the wrong file!
https://rt.cpan.org/Ticket/Display.html?id=75670