4989987419
Explicit conversion to boolean to suppress shorten-64-to-32 warnings
2020-07-10 13:02:31 +09:00
a82252df42
Fixed another typo
2020-07-10 12:48:47 +09:00
234f8eee33
Fixed typos
2020-07-10 12:32:48 +09:00
e4ee992099
vm_push_frame_debug_counter_inc: use branches
...
Ko1 doesn't like previous code.
2020-07-10 12:23:41 +09:00
0e276dc458
vm_push_frame: move assignments around
...
Struct assignment using a compound literal is more readable than before,
to me at least. It seems compilers reorder assignments anyways.
Neither speedup nor slowdown is observed on my machine.
2020-07-10 12:23:41 +09:00
215c6fa3d0
RUBY_CONST_ASSERT: use STATIC_ASSERT instead
...
Static assertions shall be done using STATIC_ASSERT these days.
2020-07-10 12:23:41 +09:00
4b8170ce80
vm_push_frame: move assertions out of the function
...
These assertions are purely static. Ned not be checked on-the-fly.
2020-07-10 12:23:41 +09:00
1d93705d6a
vm_push_frame: hoist out debug codes
...
Made it a bit readable.
2020-07-10 12:23:41 +09:00
db7f3496dd
nobody uses the return value of vm_push_frame
...
Surprised to see such a waste of time in this super duper hot path.
2020-07-10 12:23:41 +09:00
ba81bc24e6
Add instance_methods to class generated by DelegateClass
...
Also, make DelegateClass.instance_method fallback to superclass.
Fixes [Bug #16982 ]
2020-07-09 15:01:10 -07:00
7a479b30b6
* 2020-07-10 [ci skip]
2020-07-10 06:50:04 +09:00
e1bbb9ea9c
Document that Kernel#load will load relative to current directory [ci skip]
...
Update and format the Kernel#load documentation to separate the
three cases (absolute path, explicit relative path, other), and
also document that it raises LoadError on failure.
Fixes [Bug #16988 ]
2020-07-09 14:47:13 -07:00
addbb7d16e
Fix typos [ci skip]
2020-07-09 18:14:53 +09:00
cb0404cc51
* 2020-07-09 [ci skip]
2020-07-09 13:56:06 +09:00
a0a5eec0e7
Bail out when unsupported coroutine target
2020-07-09 13:53:20 +09:00
9fc25811d5
Enhanced RDoc for Array ( #3282 )
...
Methods:
reject!
reject
delete_if
zip
transpose
replace
clear
2020-07-08 09:27:10 -05:00
23655e6de5
Fix build error on Apple silicon macOS
...
This patch will fix following build error on macOS 11.0 + Xcode 12 beta.
```
$ ./configure
$ make V=1
BASERUBY = /Users/watson/.rbenv/shims/ruby --disable=gems
CC = clang -fdeclspec
LD = ld
LDSHARED = clang -fdeclspec -dynamiclib
CFLAGS = -O3 -ggdb3 -Wall -Wextra -Werror=deprecated-declarations -Werror=division-by-zero -Werror=implicit-function-declaration -Werror=implicit-int -Werror=misleading-indentation -Werror=pointer-arith -Werror=shorten-64-to-32 -Werror=write-strings -Wmissing-noreturn -Wno-constant-logical-operand -Wno-long-long -Wno-missing-field-initializers -Wno-overlength-strings -Wno-parentheses-equality -Wno-self-assign -Wno-tautological-compare -Wno-unused-parameter -Wno-unused-value -Wunused-variable -Werror=extra-tokens -std=gnu99 -pipe
XCFLAGS = -D_FORTIFY_SOURCE=2 -fstack-protector-strong -fno-strict-overflow -DRUBY_DEVEL=1 -fvisibility=hidden -DRUBY_EXPORT -fPIE -DCANONICALIZATION_FOR_MATHN -I. -I.ext/include/arm64-darwin20 -I./include -I. -I./enc/unicode/12.1.0
CPPFLAGS = -D_XOPEN_SOURCE -D_DARWIN_C_SOURCE -D_DARWIN_UNLIMITED_SELECT -D_REENTRANT
DLDFLAGS = -Wl,-undefined,dynamic_lookup -Wl,-multiply_defined,suppress -fstack-protector-strong -Wl,-pie -framework Security -framework Foundation
SOLIBS = -lpthread -ldl -lobjc
LANG = ja_JP.UTF-8
LC_ALL =
LC_CTYPE =
MFLAGS =
Apple clang version 12.0.0 (clang-1200.0.22.19)
Target: arm64-apple-darwin20.0.0
Thread model: posix
InstalledDir: /Applications/Xcode-beta.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin
clang -fdeclspec -O3 -ggdb3 -Wall -Wextra -Werror=deprecated-declarations -Werror=division-by-zero -Werror=implicit-function-declaration -Werror=implicit-int -Werror=misleading-indentation -Werror=pointer-arith -Werror=shorten-64-to-32 -Werror=write-strings -Wmissing-noreturn -Wno-constant-logical-operand -Wno-long-long -Wno-missing-field-initializers -Wno-overlength-strings -Wno-parentheses-equality -Wno-self-assign -Wno-tautological-compare -Wno-unused-parameter -Wno-unused-value -Wunused-variable -Werror=extra-tokens -std=gnu99 -pipe -D_FORTIFY_SOURCE=2 -fstack-protector-strong -fno-strict-overflow -DRUBY_DEVEL=1 -fvisibility=hidden -DRUBY_EXPORT -fPIE -DCANONICALIZATION_FOR_MATHN -I. -I.ext/include/arm64-darwin20 -I./include -I. -I./enc/unicode/12.1.0 -D_XOPEN_SOURCE -D_DARWIN_C_SOURCE -D_DARWIN_UNLIMITED_SELECT -D_REENTRANT -o cont.o -c cont.c
In file included from cont.c:19:
./coroutine/ucontext/Context.h:32:5: error: 'getcontext' is deprecated: first deprecated in macOS 10.6 [-Werror,-Wdeprecated-declarations]
getcontext(&context->state);
^
/Applications/Xcode-beta.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/include/ucontext.h:37:6: note: 'getcontext' has been
explicitly marked deprecated here
int getcontext(ucontext_t *) __OSX_AVAILABLE_BUT_DEPRECATED(__MAC_10_5, __MAC_10_6, __IPHONE_2_0, __IPHONE_2_0) __WATCHOS_PROHIBITED __TVOS_PROHIBITED;
^
In file included from cont.c:19:
./coroutine/ucontext/Context.h:51:5: error: 'makecontext' is deprecated: first deprecated in macOS 10.6 [-Werror,-Wdeprecated-declarations]
makecontext(&context->state, (void(*)(void))coroutine_trampoline, 2, (void*)start, (void*)context);
^
/Applications/Xcode-beta.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/include/ucontext.h:38:6: note: 'makecontext' has been
explicitly marked deprecated here
void makecontext(ucontext_t *, void (*)(), int, ...) __OSX_AVAILABLE_BUT_DEPRECATED(__MAC_10_5, __MAC_10_6, __IPHONE_2_0, __IPHONE_2_0) __WATCHOS_PROHI...
^
In file included from cont.c:19:
./coroutine/ucontext/Context.h:59:5: error: 'swapcontext' is deprecated: first deprecated in macOS 10.6 [-Werror,-Wdeprecated-declarations]
swapcontext(¤t->state, &target->state);
^
/Applications/Xcode-beta.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/include/ucontext.h:40:6: note: 'swapcontext' has been
explicitly marked deprecated here
int swapcontext(ucontext_t * __restrict, const ucontext_t * __restrict) __OSX_AVAILABLE_BUT_DEPRECATED(__MAC_10_5, __MAC_10_6, __IPHONE_2_0, __IPHONE_...
^
3 errors generated.
make: *** [cont.o] Error 1
```
2020-07-08 22:56:49 +09:00
11af12026e
Hoisted out functions for no name rest argument symbol
2020-07-08 18:35:46 +09:00
6a05532315
Constified NODE pointer in ASTNodeData
2020-07-08 18:26:12 +09:00
ff5e660340
Added `NODE_SPECIAL_EXCESSIVE_COMMA` info to `ARGS` of `RubyVM::AbstractSyntaxTree`.
2020-07-08 17:43:04 +09:00
57cde28c20
* 2020-07-08 [ci skip]
2020-07-08 10:00:30 +09:00
ecfc09d053
gc.c: Cast int literal "1" to bits_t
...
... because shifting by more than 31 bits has undefined behavior
(depending upon platform). Coverity Scan found this issue.
2020-07-08 09:58:48 +09:00
b06a4dc6f1
Expand heap pages to be exactly 16kb
...
This commit expands heap pages to be exactly 16KiB and eliminates the
`REQUIRED_SIZE_BY_MALLOC` constant.
I believe the goal of `REQUIRED_SIZE_BY_MALLOC` was to make the heap
pages consume some multiple of OS page size. 16KiB is convenient because
OS page size is typically 4KiB, so one Ruby page is four OS pages.
Do not guess how malloc works
=============================
We should not try to guess how `malloc` works and instead request (and
use) four OS pages.
Here is my reasoning:
1. Not all mallocs will store metadata in the same region as user requested
memory. jemalloc specifically states[1]:
> Information about the states of the runs is stored as a page map at the beginning of each chunk.
2. We're using `posix_memalign` to request memory. This means that the
first address must be divisible by the alignment. Our allocation is
page aligned, so if malloc is storing metadata *before* the page,
then we've already crossed page boundaries.
3. Some allocators like glibc will use the memory at the end of the
page. I am able to demonstrate that glibc will return pointers
within the page boundary that contains `heap_page_body`[2]. We
*expected* the allocation to look like this:
But since `heap_page` is allocated immediately after
`heap_page_body`[3], instead the layout looks like this:
This is not optimal because `heap_page` gets allocated immediately
after `heap_page_body`. We frequently write to `heap_page`, so the
bottom OS page of `heap_page_body` is very likely to be copied.
One more object per page
========================
In jemalloc, allocation requests are rounded to the nearest boundary,
which in this case is 16KiB[4], so `REQUIRED_SIZE_BY_MALLOC` space is
just wasted on jemalloc.
On glibc, the space is not wasted, but instead it is very likely to
cause page faults.
Instead of wasting space or causing page faults, lets just use the space
to store one more Ruby object. Using the space to store one more Ruby
object will prevent page faults, stop wasting space, decrease memory
usage, decrease GC time, etc.
1. https://people.freebsd.org/~jasone/jemalloc/bsdcan2006/jemalloc.pdf
2. 33390d15e7289a28e68f/gc.c (L1757-L1763)https://people.freebsd.org/~jasone/jemalloc/bsdcan2006/jemalloc.pdf page 4
Co-authored-by: John Hawthorn <john@hawthorn.email>
2020-07-06 14:17:54 -07:00
b02a9584a1
* 2020-07-07 [ci skip]
2020-07-07 00:26:58 +09:00
2c6512fe67
Get rid of the redundant stat() in rb_check_realpath_internal
2020-07-07 00:26:37 +09:00
167d139487
Inline builtin struct aref
...
We don't do this for aset because it might raise a FrozenError.
```
$ benchmark-driver -v --rbenv 'before;after;before --jit;after --jit' benchmark/mjit_struct_aref.yml --repeat-count=4
before: ruby 2.8.0dev (2020-07-06T01:47:11Z master d94ef7c6b6d94ef7c6b6
2020-07-06 00:14:00 -07:00
d94ef7c6b6
Run method_missing in the same execution context
2020-07-06 10:47:11 +09:00
efe851a0df
[ruby/bigdecimal] Version 2.0.1
...
https://github.com/ruby/bigdecimal/commit/3fa4f2ac67
2020-07-06 01:15:47 +09:00
6a826eb4b0
[ruby/bigdecimal] Return US-ASCII string from BigDecimal#to_s
...
Fixes #159
https://github.com/ruby/bigdecimal/commit/57ee92e700
2020-07-06 01:15:47 +09:00
e794d96ca4
[ruby/bigdecimal] Drop Ruby 2.3 support
...
https://github.com/ruby/bigdecimal/commit/79a819d205
2020-07-06 01:15:47 +09:00
40b82afe6a
[ruby/bigdecimal] Support Ruby < 2.6
...
https://github.com/ruby/bigdecimal/commit/61ec452599
2020-07-06 01:15:46 +09:00
03a33603c6
[ruby/bigdecimal] Add Complex#to_d
...
https://github.com/ruby/bigdecimal/commit/97e794ac97
2020-07-06 01:15:46 +09:00
6607212224
[ruby/bigdecimal] Make tests pass on Ruby 2.4
...
https://github.com/ruby/bigdecimal/commit/9d19e842ee
2020-07-06 01:15:46 +09:00
d36b197d7c
[ruby/bigdecimal] Remove definition of BigDecimal#initialize_copy
...
This leaves the default definition, which will raise FrozenError.
https://github.com/ruby/bigdecimal/commit/05e843d838
2020-07-06 01:15:46 +09:00
f00efef30a
[ruby/bigdecimal] Support a Complex in Kernel.BigDecimal()
...
https://github.com/ruby/bigdecimal/commit/00795cb01f
2020-07-06 01:15:46 +09:00
ceb9d516c6
[ruby/bigdecimal] Undef BigDecimal#initialize_copy
...
Both BigDecimal#clone and BigDecimal#dup return self, there is no
reason to have initialize_copy exposed as a Ruby method.
The same is true for initialize_clone and initialize_dup.
https://github.com/ruby/bigdecimal/commit/aaf237fa9e
2020-07-06 01:15:45 +09:00
d69510b6cd
Removed ext/bigdecimal/util/depend unused since dd0c75fdc2
2020-07-06 01:15:23 +09:00
1aed23c18d
tool/sync_default_gems.rb: Accept multiple commit ranges
2020-07-06 00:48:33 +09:00
8e189df32c
Add operator info to `OP_ASGN2` of `RubyVM::AbstractSyntaxTree`.
2020-07-06 00:48:15 +09:00
988b4bc9ea
* 2020-07-06 [ci skip]
2020-07-06 00:28:53 +09:00
41fef191d1
tool/sync_default_gems.rb: Added bigdecimal
2020-07-05 22:08:04 +09:00
6236bbf363
Removed duplicate test with test_BigDecimal_bug7522
2020-07-05 22:02:59 +09:00
4ea57c2a39
Escape `#` for GNU make 3
2020-07-05 19:15:03 +09:00
12b377a96c
Updated bundled gems
2020-07-05 19:00:46 +09:00
358f91bf6f
Skip comment and empty lines in gems/bundled_gems file
2020-07-05 18:59:06 +09:00
41168f69fb
Prefix export symbol prefix to coroutine_transfer
2020-07-05 17:27:12 +09:00
04c704c5c9
Removed trailing spaces [ci skip]
2020-07-05 17:12:20 +09:00
95f5fd9a55
fix up Primitive.cinit! code
...
Recent changes break Primitive.cinit!(c_code) so fix it.
2020-07-05 10:44:36 +09:00
e4f7eee009
Check ROBJECT_EMBED on guards-merged ivar access
...
Fix CI failure like
http://ci.rvm.jp/results/trunk-mjit-wait@silicon-docker/3043247
introduced by a69dd699ee
2020-07-04 16:02:46 -07:00
c146362555
Use static variables in Range#minmax
2020-07-04 13:34:38 -07:00
Nobuyoshi Nakada
卜部昌平
Masataka Pocke Kuwabara
git
Jeremy Evans
Kazuhiro NISHIYAMA
Burdette Lamar
Watson
manga_osyo
Yusuke Endoh
Aaron Patterson
Jean Boussier
Takashi Kokubun
Kenta Murata
Koichi Sasada
