(TestOldThreadSelect#test_old_select_read_timeout): if the machine
is fast enough, the time used by code around IO.select may be smaller
than Time implement threshold.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33131 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
typo.
* test/-ext-/old_thread_select/test_old_thread_select.rb
(TestOldThreadSelect#test_old_select_signal_safe): use SIGINT instead
of SIGUSR1 because the former is general and the latter is platform
dependent.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33130 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
rb_thread_fd_select(). old one is EINTR unsafe.
Patch by Eric Wong. [Bug #5229] [ruby-core:39102]
* test/-ext-/old_thread_select/test_old_thread_select.rb:
a testcase for rb_thread_select().
* ext/-test-/old_thread_select/old_thread_select.c: ditto.
* ext/-test-/old_thread_select/depend: ditto.
* ext/-test-/old_thread_select/extconf.rb: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33117 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
bitwise arithmetic with a Fixnum. #1792
* test/ruby/test_fixnum.rb: add tests for the above change.
* bignum.c (bit_coerce): A Fixnum and a Bignum are only permitted for
bitwise arithmetic with a Bignum. #1792
* test/ruby/test_bignum.rb: add tests for the above change.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33108 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
the encoding of the format. On Windows (at least Japanese Windows),
Time#strftime("%Z") includes non ASCII in locale encoding (CP932).
So convert locale to default internal. [ruby-core:39092] [Bug #5226]
* strftime.c (rb_strftime): ditto.
* strftime.c (rb_strftime_timespec): ditto.
* internal.h (rb_strftime_timespec): follow above.
* time.c (rb_strftime_alloc): ditto.
* time.c (strftimev): ditto.
* time.c (time_strftime): ditto.
* time.c (time_to_s): the resulted string of Time#to_s is always
ascii only, so this should be US-ASCII.
* time.c (time_asctime): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33094 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
What's the problem?
autoload is thread unsafe. When we define a constant to be
autoloaded, we expect the constant construction is invariant. But
current autoload implementation allows other threads to access the
constant while the first thread is loading a file.
What's happening inside?
The current implementation uses Qundef as a marker of autoload in
Constant table. Once the first thread find Qundef as a value at
constant lookup, it starts loading a defined feature. Generally a
loaded file overrides the Qundef in Constant table by module/class
declaration at very beginning lines of the file, so other threads
can see the new Module/Class object before feature loading is
finished. It breaks invariant construction.
How to solve?
To ensure invariant constant construction, we need to override
Qundef with defined Object after the feature loading. For keeping
Qundef in Constant table, I expanded autoload_data struct in
Module to have a slot for keeping the defined object while feature
loading. And changed Module's constant lookup/update logic a
little so that the slot is only visible from the thread which
invokes feature loading. (== the first thread which accessed the
autoload constant)
Evaluation?
All test passes (bootstrap test, test-all and RubySpec) and added
8 tests for threading behavior. Extra logics are executed only
when Qundef is found, so no perf drop should happen except
autoloading.
* variable.c (rb_autoload): Prepare new autoload_data struct.
* variable.c (rb_autoload_load): Load feature and update Constant
table after feature loading is finished.
* variable.c (rb_const_get_0): When the fetched constant is under
autoloading, it returns the object only for the thread which starts
autoloading.
* variable.c (rb_const_defined_0): Ditto.
* variable.c (rb_const_set): When the specified constant is under
autoloading, it sets the object only for the thread which starts
autoloading. Otherwise, simply overrides Qundef with constant
override warning.
* vm_insnhelper.c (vm_get_ev_const): Apply same change as
rb_const_get_0 in variable.c.
* test/ruby/test_autoload.rb: Added tests for threading behavior.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33078 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
creating ruby-format gemspecs. Fixes Gem.dir not being at the front
of Gem.path to fix uninstall and cleanup commands. Fixes gem
uninstall stopping on the first missing gem.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33074 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
should be manipulated because master pty may not be a tty on some
environment (e.g. Solaris). [Bug:#5222] [ruby-dev:44420]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33049 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
A patch by Magnus Holm <judofyr@gmail.com>. Thanks!
* test/ruby/test_variable.rb: add a test for above.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33045 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
HTTPS image paths to be turned into <img> tags. Prevents special
markup inside <tt> from being processed.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33043 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
line is out of the way. rb_sourceline() is unreliable when source
code is big. [ruby-dev:44413]
* test/coverage/test_coverage.rb: add a test for above.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33030 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
gzip header is readable to avoid SEGV.
* test/zlib/test_zlib.rb (test_corrupted_header): Test it.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33023 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* test/ruby/test_method.rb (test_body): add a test for the above change.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33013 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
is explicitly given as a buffer ([Bug #5207]), otherwise set the
encoding. also removed dead code.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33010 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
connection is closed. based on the patch by Hugo Barauna.
[Bug #5190] [ruby-core:38930]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33007 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
based on the patch by art lussos. [Bug #5163] [ruby-core:38820]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@33001 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
into Complex if the argument is a Complex with non-zero imaginary
part. [Bug #5020] [ruby-dev:44088]
* test/ruby/test_rational.rb (test_coerce): test for the above change.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32994 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* test/openssl/test_ssl.rb: Add a test for it.
Thanks to Eric Wong for the patch.
[Ruby 1.9 - Feature #5183] [ruby-core:38911]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32973 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
namespace by Commands. Remove unused variable and debugging
statement in tests.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32948 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
imaginary part is exactly zero. The patch is made by Marc-Andre
Lafortune. fixes [Bug #5178] [ruby-core:38885]
* test/ruby/test_complex.rb (test_rationalize): add a test for the
above change.
* complex.c (nucomp_to_r): fix RDoc comment. The patch is made by
Marc-Andre Lafortune.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32903 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* test/dl/test_base.rb (libc_so, libm_so): on Solaris, remove libc
and libm version numbers for detecting default libc and libm.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32892 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Patch by Nobuyoshi Nakada. [Bug #4457] [ruby-dev:43285]
* test/ruby/test_time.rb (class TestTime): add a test for the
above change.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32885 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Added missing require for `gem uninstall --format-executable`.
The correct name of the executable being uninstalled is now displayed
with --format-executable.
Fixed `gem unpack uninstalled_gem` default version picker.
RubyGems no longer claims a nonexistent gem can be uninstalled.
`gem which` no longer claims directories are requirable files.
`gem cleanup` continues cleaning up gems if one can't be uninstalled
due to permissions. Issue #82.
Gem repository directories are no longer created world-writable.
Patch by Sakuro OZAWA. [Ruby 1.9 - Bug #4930]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32852 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
OpenBSD seems not to detach the controlling terminal, when the
argument noclose is non-zero. ref: [Bug #5135]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32848 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
'BindAddress: var' => ':BindAddress => var' according to other configurations.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32828 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
[Bug #5153] [ruby-core:38736]
* io.c (argf_next_argv): remove the call of above function.
* ext/-test-/add_suffix, test/-ext-/test_add_suffix.rb: remove the test
extension module because this is only for testsing ruby_add_suffix().
* LEGAL: remove the mention about a part of util.c, because now we
removed the part.
* io.c (argf_next_argv): now the new filename is not guranteed to
use, so should check the return value of rename(2).
* test/ruby/test_argf.rb (TestArgf#test_inplace_rename_impossible):
now we expect same result with other platforms on no_safe_rename
platforms (=Windows).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32826 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
