| Age | Commit message (Collapse) | Author |
|---|
|
|
|
Fixed FD leaks
---
test/socket/test_tcp.rb | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
|
|
[Bug #18264] Fix memory leak in TracePoint
TracePoint leaks memory because it allocates a `rb_tp_t` struct
without ever freeing it (it is created with `RUBY_TYPED_NEVER_FREE`).
---
test/ruby/test_settracefunc.rb | 10 ++++++++++
vm_trace.c | 2 +-
2 files changed, 11 insertions(+), 1 deletion(-)
|
|
|
|
|
|
https://github.com/ruby/drb/commit/7edf67654c
|
|
https://github.com/ruby/fcntl/commit/0bcc0c4518
|
|
|
|
|
|
|
|
Bring the local copy of ruby/openssl in sync with the upstream gem
release v2.2.1. The commits happened in the upstream repository can be
found at:
https://github.com/ruby/openssl/compare/v2.2.0...v2.2.1
Note that many of these have already been applied to ruby.git and don't
appear in the file changes of this commit.
|
|
Fix Ractor.make_shareable changing locals for Procs
env_copy() uses rb_ary_delete_at() with a loop counting up while
iterating through the list of read only locals. rb_ary_delete_at() can
shift elements in the array to an index lesser than the loop index,
causing locals to be missed and set to Qfalse in the returned
environment.
Iterate through the locals in reverse instead, this way the shifting
never happens for locals that are yet to be visited and we process all
the locals in the array.
[Bug #18023]
---
bootstraptest/test_ractor.rb | 22 ++++++++++++++++++++++
vm.c | 2 +-
2 files changed, 23 insertions(+), 1 deletion(-)
|
|
Dump outer variables tables when dumping an iseq to binary
This commit dumps the outer variables table when dumping an iseq to
binary. This fixes a case where Ractors aren't able to tell what outer
variables belong to a lambda after the lambda is loaded via ISeq.load_from_binary
[Bug #18232] [ruby-core:105504]
---
compile.c | 55 +++++++++++++++++++++++++++++++++++++++++++++++++-
test/ruby/test_iseq.rb | 10 +++++++++
2 files changed, 64 insertions(+), 1 deletion(-)
|
|
Avoid race condition in Regexp#match
In certain conditions, Regexp#match could return a MatchData with
missing captures. This seems to require at the least, multiple
threads calling a method that calls the same block/proc/lambda
which calls Regexp#match.
The race condition happens because the MatchData is passed from
indirectly via the backref, and other threads can modify the
backref.
Fix the issue by:
1. Not reusing the existing MatchData from the backref, and always
allocating a new MatchData.
2. Passing the MatchData directly to the caller using a VALUE*,
instead of indirectly through the backref.
It's likely that variants of this issue exist for other Regexp
methods. Anywhere that MatchData is passed implicitly through
the backref is probably vulnerable to this issue.
Fixes [Bug #17507]
---
re.c | 46 +++++++++++++++++++---------------------------
test/ruby/test_regexp.rb | 21 +++++++++++++++++++++
2 files changed, 40 insertions(+), 27 deletions(-)
|
|
60d0421ca861944459f52292d65dbf0ece26e38a,b6534691a16d751d59fc572d5dddebcaeb21f007,409dbc951b9875d27bd73748c88e15386473cffb,842b0008c132dd587f09766a228041afb7fed24f: [Backport #18191]
Fix the encoding of loaded feature names [Bug #18191]
The feature names loaded from the default load paths should also
be in the file system encoding.
---
ruby.c | 12 +++++++++++-
test/ruby/test_require.rb | 22 ++++++++++++++++++++++
2 files changed, 33 insertions(+), 1 deletion(-)
Copy path strings as interned strings
---
ruby.c | 12 ++++++++++--
1 file changed, 10 insertions(+), 2 deletions(-)
Replace expanded load path only when modified
---
ruby.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
Skip broken strings as the locale encoding
---
internal/string.h | 1 +
ruby.c | 11 +++++++----
string.c | 6 ++++++
3 files changed, 14 insertions(+), 4 deletions(-)
|
|
89242279e61b023a81c58065c62a82de8829d0b3,529fc204af84f825f98f83c34b004acbaa802615: [Backport #18141]
Marshal.load: do not call the proc until strings have their encoding
Ref: https://bugs.ruby-lang.org/issues/18141
---
marshal.c | 7 +++-
spec/ruby/core/marshal/shared/load.rb | 62 +++++++++++++++++++++++------------
test/ruby/test_marshal.rb | 17 ++++++++++
3 files changed, 64 insertions(+), 22 deletions(-)
marshal.c: don't call the proc with partially initialized objects.
(#4866)
For cyclic objects, it requires to keep a st_table of the partially
initialized objects.
---
marshal.c | 75 ++++++++++++++++++++---------------
spec/ruby/core/marshal/shared/load.rb | 75 ++++++++++++++++++++---------------
test/ruby/test_marshal.rb | 12 ++++++
3 files changed, 97 insertions(+), 65 deletions(-)
|
|
7c0230b05d0978958f89434c84ddd9c82419c1a5,552728a23aeab0df598b356b19a573259e297d14,49af9012a20a824542cf588e55e5488895553e09: [Backport #18184]
Check the entire name as `ruby2_keywords_flag` [Bug #18184]
---
marshal.c | 2 +-
test/ruby/test_marshal.rb | 8 +++++++-
2 files changed, 8 insertions(+), 2 deletions(-)
Check the encoding of `ruby2_keywords_flag` [Bug #18184]
---
marshal.c | 1 +
test/ruby/test_marshal.rb | 6 +++++-
2 files changed, 6 insertions(+), 1 deletion(-)
Prohibit invalid encoding symbols [Bug #18184]
---
marshal.c | 8 +++++++-
test/ruby/test_marshal.rb | 10 +++++++---
2 files changed, 14 insertions(+), 4 deletions(-)
|
|
[Bug #18173] Update loaded_features_index
If $LOADED_FEATURES is changed in the just required file, also the
index table needs to be updated before loaded_features_snapshot is
reset. If the snapshot is reset without updating the table, the
name of the added feature will not be found.
---
load.c | 1 +
test/ruby/test_require.rb | 17 +++++++++++++++++
2 files changed, 18 insertions(+)
|
|
Use `%printer` directive for Bison 3.8
---
parse.y | 74 ++++++++++++++++++++++++++---------------------------------------
1 file changed, 29 insertions(+), 45 deletions(-)
|
|
|
|
If the thread termination invokes user code after `th->status` becomes
`THREAD_KILLED`, and the user unblock function causes that `th->status` to
become something else (e.g. `THREAD_RUNNING`), threads waiting in
`thread_join_sleep` will hang forever. We move the unblock function call
to before the thread status is updated, and allow threads to join as soon
as `th->value` becomes defined.
This reverts commit 6505c77501f1924571b2fe620c5c7b31ede0cd22.
|
|
|
|
|
|
* Wake up join list within thread EC context.
* Consume items from join list so that they are not re-executed.
If `rb_fiber_scheduler_unblock` raises an exception, it can result in a
segfault if `rb_threadptr_join_list_wakeup` is not within a valid EC. This
change moves `rb_threadptr_join_list_wakeup` into the thread's top level EC
which initially caused an infinite loop because on exception will retry. We
explicitly remove items from the thread's join list to avoid this situation.
* Verify the required scheduler interface.
* Test several scheduler hooks methods with broken `unblock` implementation.
|
|
Revert "Force recycle intermediate collection in Hash#transform_keys!
[Bug #17735]"
This reverts commit 522d4cd32f7727886f4fcbc28ed29c08d361ee20.
---
hash.c | 1 -
1 file changed, 1 deletion(-)
|
|
Preserve the encoding of the argument in IndexError [Bug #18160]
---
re.c | 20 ++++++++++----------
test/ruby/test_regexp.rb | 7 ++++++-
2 files changed, 16 insertions(+), 11 deletions(-)
|
|
[ruby/strscan] Fix segmentation fault of `StringScanner#charpos` when
`String#byteslice` returns non string value [Bug #17756] (#20)
https://github.com/ruby/strscan/commit/92961cde2b
---
ext/strscan/strscan.c | 5 +----
test/strscan/test_stringscanner.rb | 17 +++++++++++++++++
2 files changed, 18 insertions(+), 4 deletions(-)
|
|
Check if closed after each yield [Bug #17661]
---
io.c | 4 +++-
test/ruby/test_io.rb | 36 ++++++++++++++++++++++++++++++++++++
2 files changed, 39 insertions(+), 1 deletion(-)
|
|
[Bug #18154] Fix memory leak in String#initialize
String#initialize can leak memory when called on a string that is marked
with STR_NOFREE because it does not unset the STR_NOFREE flag.
---
string.c | 2 +-
test/ruby/test_string.rb | 10 ++++++++++
2 files changed, 11 insertions(+), 1 deletion(-)
|
|
Use free instead of xfree to free altstack
The altstack memory of a thread may be free'ed even after the VM is
destructed. After that, GC is no longer available, so calling xfree
may lead to a segfault.
This changeset uses the bare free function to free the altstack memory
instead of xfree. [Bug #18126]
---
signal.c | 5 ++++-
vm_core.h | 2 +-
2 files changed, 5 insertions(+), 2 deletions(-)
|
|
Avoid `free(3)`ing invalid pointer
Fixes [Bug #17794]
---
addr2line.c | 2 ++
1 file changed, 2 insertions(+)
|
|
addr2line.c: DW_LNS_fixed_advance_pc takes a single uhalf operand
Fixes [Bug #17609]
Co-Authored-By: xtkoba (Tee KOBAYASHI) <xtkoba+ruby@gmail.com>
---
addr2line.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
|
|
undefine alloc functions for C extensions
per guidance in doc/extension.rdoc, these classes now undefine their
alloc functions:
- ObjectSpace::InternalObjectWrapper
- Socket::Ifaddr
---
ext/objspace/objspace.c | 1 +
ext/socket/ifaddr.c | 1 +
2 files changed, 2 insertions(+)
|
|
[Bug #18117] Fix Ractor race condition with GC
rb_objspace_reachable_objects_from requires that the GC not be active.
Since the Ractor barrier is not executed for incremental sweeping,
Ractor may call rb_objspace_reachable_objects_from after sweeping
has started to share objects. This causes a crash that looks like
the following:
```
<internal:ractor>:627: [BUG] rb_objspace_reachable_objects_from() is not supported while during_gc == true
```
Co-authored-by: Vinicius Stock <vinicius.stock@shopify.com>
---
bootstraptest/test_ractor.rb | 15 +++++++++++++++
ractor.c | 12 ++++++++++--
2 files changed, 25 insertions(+), 2 deletions(-)
|
|
|
|
cd4f5b13228879d954fa97b6aa479c4a5ef4fb0a,8db269edb3550a85dfab9b193ea115ca36912ced,ab63f6d8543903f177c46634f38e5428655f003b: [Backport #18140]
Guard array when appending
This prevents early collection of the array. The GC doesn't see the
array on the stack when Ruby is compiled with optimizations enabled
[ruby-core:105099] [Bug #18140]
---
array.c | 1 +
test/ruby/test_array.rb | 6 ++++++
2 files changed, 7 insertions(+)
Guard array when appending
This prevents early collection of the array. The GC doesn't see the
array on the stack when Ruby is compiled with optimizations enabled
Thanks @jhaberman for the test case
[ruby-core:105099] [Bug #18140]
---
ext/-test-/array/concat/depend | 321 ++++++++++++++++++++++++++++++++
ext/-test-/array/concat/extconf.rb | 2 +
ext/-test-/array/concat/to_ary_conact.c | 64 +++++++
test/-ext-/array/test_to_ary_concat.rb | 20 ++
4 files changed, 407 insertions(+)
create mode 100644 ext/-test-/array/concat/depend
create mode 100644 ext/-test-/array/concat/extconf.rb
create mode 100644 ext/-test-/array/concat/to_ary_conact.c
create mode 100644 test/-ext-/array/test_to_ary_concat.rb
Refined test [Bug #18140]
---
ext/-test-/array/concat/to_ary_conact.c | 48 +++++++--------------------------
test/ruby/test_array.rb | 5 +++-
2 files changed, 13 insertions(+), 40 deletions(-)
|
|
Co-authored-by: nagachika <nagachika@ruby-lang.org>
|
|
d43279edacd09edf3a43e02d62f5be475e7c3bcb,5dc36ddcd00fc556c04c15ce9770c5a84d7d43dc,523bf31564f160f899f8cf9f73540d6a6f687f17: [Backport #18138]
Fix length calculation for Array#slice!
Commit 4f24255 introduced a bug which allows a length to be passed to
rb_ary_new4 which is too large, resulting in invalid memory access.
For example:
(1..1000).to_a.slice!(-2, 1000)
---
array.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Add out of range tests for Array#slice!
---
test/ruby/test_array.rb | 13 +++++++++++++
1 file changed, 13 insertions(+)
Add negative position tests [Bug #18138]
---
test/ruby/test_array.rb | 4 ++++
1 file changed, 4 insertions(+)
|
|
|
|
* Bundle RBS 1.4.0
* Bundle typeprof 0.15.2
|
|
|
|
|
|
|
|
|
|
|
|
|
|
164f50dea918e7019847f578c3cffb079993d626,1985a3a77fac64cda177c74113a9348e36233630:"
This reverts commit 09d90c0ed861e74e58a59bc413bc39bcf9775db8.
|
|
164f50dea918e7019847f578c3cffb079993d626,1985a3a77fac64cda177c74113a9348e36233630:
Update TypeProf to 0.15.1
---
.github/workflows/ubuntu.yml | 2 +-
gems/bundled_gems | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
Bundle RBS 1.4.0 & typeprof 0.15.2 (#4753)
* Bundle RBS 1.4.0
* Bundle typeprof 0.15.2
---
gems/bundled_gems | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
|
|
|
|
This fixes multiple bugs found in the partial backtrace
optimization added in 3b24b79.
These bugs occurs when passing a start argument to caller where
the start argument lands on a iseq frame without a pc.
Before this commit, the following code results in the same
line being printed twice, both for the #each method.
```
def a; [1].group_by { b } end
def b; puts(caller(2, 1).first, caller(3, 1).first) end
a
```
After this commit and in Ruby 2.7, the lines are different,
with the first line being for each and the second for group_by.
Before this commit, the following code can either segfault or
result in an infinite loop:
```
def foo
caller_locations(2, 1).inspect # segfault
caller_locations(2, 1)[0].path # infinite loop
end
1.times.map { 1.times.map { foo } }
```
After this commit, this code works correctly.
In terms of the implementation, this correctly skips iseq frames
without pc that occur before the number of frames the caller
requested to skip.
This rewrites the algorithm used for handling the partial
backtraces. It scans from the current frame outward to the
earliest frame, until it has found the desired number of frames.
It records that frame as the start frame. If needed, it continues
scanning backwards until arg->prev_cfp is set, as that is needed
to set the location of the first frame. Due to the fact that arg
is a void pointer, it's not possible to check this directly, but
this calls the iter_skip function in a situation where it knows
it will set arg->prev_cfp, and then breaks out of the loop.
Fixes [Bug #18053]
|
