| Age | Commit message (Collapse) | Author |
|
Windows paths do not start with a slash, so we add an extra slash to
separate the host from the path in file:// urls. Otherwise "D:" is
parsed as the host segment in the URI.
The path for those URLs now starts with "/", so we ignore that leading
character when using the URI's path.
This reduces Windows CI spec failures from 429 to 355.
https://github.com/bundler/bundler/commit/1b7e274cbc
|
|
https://github.com/bundler/bundler/commit/5946d62ad0
|
|
https://github.com/bundler/bundler/commit/f694fe7f67
|
|
So we can later add more tests with the same setup.
https://github.com/bundler/bundler/commit/21b4b6c49c
|
|
https://github.com/bundler/bundler/commit/d9d2bf6d52
|
|
https://github.com/bundler/bundler/commit/80217bfe0b
|
|
https://github.com/bundler/bundler/commit/ca28a04668
|
|
ANSI codes don't work there.
https://github.com/bundler/bundler/commit/4f05417ca7
|
|
For consistency.
https://github.com/bundler/bundler/commit/279603ab42
|
|
https://github.com/bundler/bundler/commit/a4d04506ab
|
|
Thor's base shell will be memoized the first time it is set. So if we
instantiate a no-color shell first, further instantiations of a bundler
shell will be initialized with a no-color shell by default. This is
caused some sources specs to fail, depending on the order they run.
See for example https://travis-ci.org/bundler/bundler/builds/500328994.
What we do to fix it is to reset the shell unless no-color is explicitly
specified. That way, further instantiations will rerun thor's internal
logic to choose the default shell.
https://github.com/bundler/bundler/commit/786b5d9894
|
|
The builtin thor's color shell already checks this before printing, and
if not a tty, prints just the string without added color codes.
https://github.com/bundler/bundler/commit/4f62611c87
|
|
https://github.com/bundler/bundler/commit/1892626858
|
|
e9e17cbc051e894dfd27eda5feca2939f65552db (enabling the warning by
default) caused a warning in test-spec:
```
/data/chkbuild/tmp/build/20190802T213005Z/ruby/spec/ruby/core/binding/irb_spec.rb
Binding#irb
- creates an IRB session with the binding in scope/data/chkbuild/tmp/build/20190802T213005Z/ruby/spec/ruby/core/binding/fixtures/irb.rb:3: warning: __FILE__ in eval may not return location in binding; use Binding#source_location instead
```
https://rubyci.org/logs/rubyci.s3.amazonaws.com/debian/ruby-master/log/20190802T213005Z.log.html.gz
ref: [Bug #4352]
|
|
|
|
|
|
to control erros on our own.
|
|
This reverts commit a1727e43af18c122c4d7adfa603f895fdc8f3df1.
|
|
This reverts commit a0980f2446c0db735b8ffeb37e241370c458a626.
Retry for macOS Mojave.
|
|
[Bug #4352]
|
|
This reverts commit e05f397f08e98fb170e3aca885d6028b344e833c.
Failed... https://travis-ci.org/ruby/ruby/jobs/566961111
We should talk to Travis people instead...
|
|
|
|
ref. https://github.com/ruby/logger/pull/35
|
|
|
|
|
|
|
|
add debug_counters to check the Hash object statistics.
|
|
This reverts commit 9faef3113fb4331524b81ba73005ba13fa0ef6c6.
It seemed to cause a failure on macOS Mojave, though I'm unsure how.
https://rubyci.org/logs/rubyci.s3.amazonaws.com/osx1014/ruby-master/log/20190802T034503Z.fail.html.gz
This tentative revert is to check if the issue is actually caused by the
change or not.
|
|
|
|
|
|
|
|
|
|
|
|
receiver
Previously, attr* methods could be private even if not in the
private section of a class/module block.
This uses the same approach that ruby started using for define_method
in 1fc33199736f316dd71d0c551edbf514528ddde6.
Fixes [Bug #4537]
|
|
__unused__ is unavailable on Sun C.
https://rubyci.org/logs/rubyci.s3.amazonaws.com/solaris11s-sunc/ruby-master/log/20190801T112505Z.fail.html.gz
|
|
Some VM frames (dummy and top pushed by `rb_vm_call_cfunc`) has
iseq but has no pc.
|
|
|
|
because the test seems to hang there forever:
https://travis-ci.org/ruby/ruby/jobs/566409880
|
|
|
|
|
|
|
|
It seems the compiler does not support VLAs.
See also: https://ci.appveyor.com/project/ruby/ruby/builds/26392589/job/px6nuiuw4e78weg1
|
|
Hash hint for ar_array is 1 byte (unsigned char). This patch introduce
ar_hint_t which represents hint type.
|
|
|
|
This function has a lot of assumptions. Should make them sure.
|
|
PC modification in gc_event_hook_body was careless. There are (so
to say) abnormal iseqs stored in the cfp. We have to check sanity
before we touch the PC.
This has not been fixed because there was no way to (ab)use the
setup from pure-Ruby. However by using our official C APIs it is
possible to touch such frame(s), resulting in SEGV.
Fixes [Bug #14834].
|
|
|
|
gcc 9.1.0 warns this.
```
signal.c:258:46: warning: '%.*s' directive argument is not a nul-terminated string [-Wformat-overflow=]
258 | rb_raise(rb_eArgError, "unsupported signal `%.*s%"PRIsVALUE"'",
| ^~~~
259 | prefix, signame_prefix, vsig);
| ~~~~~~~~~~~~~~
signal.c:200:19: note: referenced argument declared here
200 | static const char signame_prefix[3] = "SIG";
| ^~~~~~~~~~~~~~
```
https://rubyci.org/logs/rubyci.s3.amazonaws.com/ubuntu/ruby-master/log/20190801T033009Z.log.html.gz
|
|
"hash_iter_lev" can be exported by Marshal.dump and it will
introduce inconsistency. To avoid this issue, use internal_id
instead of normal ID. This issue is pointed out by Chikanaga-san.
|
|
|