55018 Commits

Author SHA1 Message Date
Urabe, Shyouhei
34e1079aef glibc says memcpy cannot take NULL
At least since 2004, glibc's <string.h> annotates memcpy as
__attribute__((__nonnull__)).  This basedir is passed to it.  When
LOAD_RELATIVE is not defined and MJIT_SEARCH_BUILD_DIR is not set,
this variable is never updated.  Should initialize with meaningful
default value.
2019-04-29 21:52:44 +09:00
Urabe, Shyouhei
e7b18ca6d9 glibc says memcpy cannot take NULL
At least since 2004, glibc's <string.h> annotates memcpy as
__attribute__((__nonnull__)).  On the other hand the argv here,
which is passed from rb_funcallv, may be NULL.  Practically this
should never be a serious problem but for maximum safety, let's
avoid passing NULL here.
2019-04-29 21:52:44 +09:00
Urabe, Shyouhei
a116f04cca suppress some UBSAN sanitizers
They are not "undefined".  UBSAN reports them because it thinks
they are "often unintentional".  We see the report rather annoying.
2019-04-29 21:52:44 +09:00
Urabe, Shyouhei
f4c68640d6 this variable is not guaranteed aligned
No problem for unaligned-ness because we never dereference.
2019-04-29 21:52:44 +09:00
Urabe, Shyouhei
7c0f513e97 fix typo 2019-04-29 21:52:44 +09:00
Nobuyoshi Nakada
6a3165e19d
Fixed HERETERM_LENGTH_MAX on IL32LLP64 2019-04-29 18:32:21 +09:00
Nobuyoshi Nakada
c7f780c155
Refactored rb_strterm_heredoc_t 2019-04-29 17:45:58 +09:00
Nobuyoshi Nakada
1432471a75
Disallow also CR in here-doc identifier
* parse.y (heredoc_identifier): CR in here-document identifier
  might or might not result in a syntax error, by the EOL code.
  make a syntax error regardless of the EOL code.
2019-04-29 13:47:20 +09:00
Nobuyoshi Nakada
23375c8b81
Make only mesg can be assigned with default fname 2019-04-29 13:42:50 +09:00
Nobuyoshi Nakada
330b376133
parse.y: fix here-doc identifier with newline
* parse.y (heredoc_identifier): quoted here-document identifier
  must end within the same line.

  the only corner case that here-document identifier can contain a
  newline is that the closing qoute is placed at the beginning of
  the next line, and has been warned since 2.4.

  ```ruby
  <<"EOS
  " # warning: here document identifier ends with a newline
  EOS
  ```
2019-04-29 12:49:59 +09:00
Kazuhiro NISHIYAMA
69cad44fac
Reduce debug output
because I found machine clock problem
see r67347
https://travis-ci.org/ruby/ruby/jobs/525784924
https://travis-ci.community/t/mtime-of-source-codes-are-sometimes-newer-than-build-time-clock-skew/3215
2019-04-29 12:21:50 +09:00
Nobuyoshi Nakada
812a438145
iseq.c: removed unnecessary zero-fills 2019-04-29 11:31:18 +09:00
Nobuyoshi Nakada
6033423d68
NEWS: Moved "Integer#[] with range" to "Core classes updates" 2019-04-29 10:11:44 +09:00
Benoit Daloze
daa0874056 CLOCK_MONOTONIC_RAW_APPROX seems less precise than advertised on macOS
* https://travis-ci.org/ruby/ruby/builds/525651487
2019-04-28 23:29:49 +02:00
Benoit Daloze
79671ec57e Update to ruby/spec@7de852d 2019-04-28 23:20:11 +02:00
Benoit Daloze
994833085a Update to ruby/mspec@c25d63d 2019-04-28 23:20:09 +02:00
Benoit Daloze
bbb93608a3 Move the check for CentOS CLOCK_MONOTONIC_RAW next to others
* Fixes "No behavior expectation was found in the example"
* https://rubyci.org/logs/rubyci.s3.amazonaws.com/centos6/ruby-trunk/log/20190428T093004Z.fail.html.gz
2019-04-28 23:17:42 +02:00
Takashi Kokubun
dfc0eeb0cf
Fully support Git in tool/merger.rb
as both backport source repository and backport destination repository.
2019-04-29 01:54:01 +09:00
Kazuhiro NISHIYAMA
f0776e3203
Add more debug print for random CI failure on osx Travis
and remove `git status` with noisy rvm trace log

see r67347
2019-04-29 01:23:20 +09:00
Yusuke Endoh
50cbb21ba5 Add "Integer#[] with range" to NEWS 2019-04-29 01:17:27 +09:00
git
75260d36b5 * remove trailing spaces. 2019-04-29 01:00:20 +09:00
Takashi Kokubun
991e32681e
tool/merger.rb: Avoid making too-deep indentation
because it's hard to understand what's going on when indentation depth
is too deep.

Sorry for polluting git blame, but most of the Merger's lines are
updated recently anyway.
2019-04-29 00:59:22 +09:00
Takashi Kokubun
2dddd370a1
Support tool/merger.rb removetag under Git
repository.
2019-04-29 00:53:38 +09:00
git
bbad0d05b3 * 2019-04-29 2019-04-29 00:22:48 +09:00
Takashi Kokubun
fad2825e42
Support tool/merger.rb tag under Git repository 2019-04-29 00:19:58 +09:00
git
555d1dda71 * expand tabs. 2019-04-28 23:42:46 +09:00
Yusuke Endoh
6bedbf4625 numeric.c: Extend Integer#[] to support range arguments
````
0b01001101[2, 4]  #=> 0b0011
0b01001100[2..5]  #=> 0b0011
0b01001100[2...6] #=> 0b0011
    ^^^^
````

[Feature #8842]
2019-04-28 23:40:57 +09:00
Yusuke Endoh
cb55024613 test/ruby/test_integer.rb: Add a sane test for Integer#[] 2019-04-28 23:23:00 +09:00
Takashi Kokubun
3bc810334c
Support tool/merger.rb up under Git repository
updating indentation (and slightly changing styles) for areas already
supporting Git.
2019-04-28 22:53:21 +09:00
Benoit Daloze
d906dd87b0 Skip CLOCK_UPTIME_RAW_APPROX since it seems less precise than advertised on macOS
* See https://travis-ci.org/ruby/ruby/jobs/525595997
2019-04-28 15:25:09 +02:00
Takashi Kokubun
392d84b490
Fix wrong svn options
for SVN as a backport source.

This was a mistake in de5378233b2ff5434f024ac66285e699794a321d...
2019-04-28 22:15:01 +09:00
Takashi Kokubun
1c8cefca31
Stop using global variable for repos
to avoid having impact from other places.
2019-04-28 22:02:34 +09:00
Benoit Daloze
9426da83c6 Exclude failing Process.clock_getres specs on AIX 2019-04-28 14:59:14 +02:00
Benoit Daloze
a27f7e499c Add missing platform guard 2019-04-28 14:55:30 +02:00
Benoit Daloze
9a0dbb3414 Skip problematic Process.clock_getres specs on ARM
* https://rubyci.org/logs/rubyci.s3.amazonaws.com/scw-9d6766/ruby-trunk/log/20190428T051708Z.fail.html.gz
* https://rubyci.org/logs/rubyci.s3.amazonaws.com/scw-ad7f67/ruby-trunk/log/20190428T045405Z.fail.html.gz
2019-04-28 14:55:30 +02:00
Takashi Kokubun
cb8eb37377
Now tool/merger.rb may use Git [ci skip] 2019-04-28 21:53:58 +09:00
Takashi Kokubun
f2d7ba6a74
make sync-default-gems GEM=irb
from e8e79d569e.

This colorizes Range object on IRB inspect.
2019-04-28 21:51:37 +09:00
Benoit Daloze
13abf5519a Workaround a CentOS bug in Process.clock_getres specs 2019-04-28 14:44:12 +02:00
Takashi Kokubun
14965c5f4b
Drop SVN-specific method from Merger module
to clarify it's not needed for Git support.
2019-04-28 21:41:02 +09:00
Takashi Kokubun
fcb1875513
Make Merger.version private to the module
because it's not used outside the Module.
2019-04-28 21:41:02 +09:00
Benoit Daloze
16695af0ef Refactor logic in Process.clock_gettime spec 2019-04-28 14:36:03 +02:00
Benoit Daloze
fc37a045a0 Fix typo in spec 2019-04-28 14:35:17 +02:00
Takashi Kokubun
d835ed4df1
Define merger.rb's methods under Merger namespace
so that we do not monkey-patch all classes by defining methods on
top-level (Object class).

Not arranging indentation in it to keep `git blame` for now.
2019-04-28 21:08:58 +09:00
Takashi Kokubun
8711f77a26
make sync-default-gems GEM=irb
from 96f05e7268.

Just syncing newer test changes so that conflicts do not happen when trunk is
modified and we need to backport that to ruby/irb.
2019-04-28 20:34:11 +09:00
Takashi Kokubun
588f212c26
make sync-default-gems GEM=irb
from 44301d3827.

This includes some fixes for string interpolation highlight fixes.
2019-04-28 20:18:45 +09:00
Takashi Kokubun
4d8ad48f7d
Support git as redmine-backporter's done destination 2019-04-28 20:17:20 +09:00
Benoit Daloze
5d24fba544 Skip the entire Process.clock_getres spec on FreeBSD
* Clocks don't match the reported precision.
* https://rubyci.org/logs/rubyci.s3.amazonaws.com/freebsd11zfs/ruby-trunk/log/20190428T093003Z.fail.html.gz
2019-04-28 12:37:29 +02:00
Takashi Kokubun
f40458e9dd
Specify VM_CHECK_MODE explicitly
In my understanding, `VM_CHECK_MODE` should be Integer and I'm not sure
how `-DVM_CHECK_MODE` would behave. To make the matters simple, let me
pass the mode explicitly.
2019-04-28 14:10:09 +09:00
Takashi Kokubun
766293fc49
Ruby 2.3 is EOL 2019-04-28 14:04:48 +09:00
Nobuyoshi Nakada
4c8f107898
Make the range to export as changelog optional
* `from` is defaulted to the beginning of the branch inclusively,
  otherwise the given revision is excluded as the previous.

* `to` is defaulted to the head.
2019-04-28 13:46:16 +09:00