Bug #13107
closed
Updated by [email protected] (Thomas Walpole) over 8 years ago
Forgot to add the location reported for the error
/home/travis/.rvm/rubies/ruby-2.4.0/lib/ruby/2.4.0/forwardable.rb:228:in `last_response'
Updated by eugeneius (Eugene Kenny) over 8 years ago
The mongo gem is also affected by this bug (https://jira.mongodb.org/browse/RUBY-1191). Running the tests in the spec/mongo/grid/stream/read_spec.rb file reliably reproduces it.
I bisected the changes to forwardable.rb between v2.3.0 and v2.4.0, and determined that the bug was introduced in r55376.
I've attached a that fixes the problem, although it effectively reverts the optimisation from r55376, so there may be a performance impact.
Updated by eugeneius (Eugene Kenny) over 8 years ago
Here's another that also fixes the problem, while keeping the optimisation from r55376.
I'll admit that I have no idea what this option does or how it could cause this bug - I noticed that the "portable" implementation of forwardable/impl.rb wasn't affected, and fiddled with the MRI-specific one until it worked. Please review this carefully if you intend to use it!
Removing tailcall_optimization also appears to work, although I think we need that option to get a sensible stack trace. Maybe trace_instruction: false, tailcall_optimization: true is an invalid combination?
Updated by nobu (Nobuyoshi Nakada) over 8 years ago
- Status changed from Open to Feedback
Enabling trace instruction disables tail call optimization, too.
It seems like a bug of the optimization and GC.
How can I reproduce it briefly?
Updated by nobu (Nobuyoshi Nakada) over 8 years ago
- Description updated (diff)
Updated by nobu (Nobuyoshi Nakada) over 8 years ago
- Status changed from Feedback to Closed
Applied in changeset r57293.
vm_insnhelper.c: block argument at tailcall
- vm_insnhelper.c (vm_call_iseq_setup_tailcall): check interrupts
after set up the new frame, not the passed block to be clobbered
by invoked finalizers and so on. [ruby-core:78981] [Bug ]
Updated by nobu (Nobuyoshi Nakada) over 8 years ago
When interrupted by a finalizer during setting up tailcall frame, a block argument on the stack was overwritten by the magic number in the finalizer frame which looks an Integer.
I tried in vain to make a test case, but a failure in mongo-ruby-driver seems fixed.
Could you try the latest trunk?
Updated by [email protected] (Thomas Walpole) over 8 years ago
This appears to be fixed in latest trunk now. Is there any info on when a 2.4.1 will be released - seems like a pretty serious issue.
Updated by vo.x (Vit Ondruch) over 8 years ago
- Backport changed from 2.2: UNKNOWN, 2.3: UNKNOWN, 2.4: UNKNOWN to 2.2: UNKNOWN, 2.3: UNKNOWN, 2.4: REQUIRED
Updated by vo.x (Vit Ondruch) over 8 years ago
Just FTR, this is causing failures in Mongo gem test suite [1]. Applying this fixed the issues.
Updated by vo.x (Vit Ondruch) over 8 years ago
And I can confirm that also rack-test test suite is passing with this [1].
[1] https://apps.fedoraproject.org/koschei/package/rubygem-rack-test?collection=f26
Updated by naruse (Yui NARUSE) over 8 years ago
- Backport changed from 2.2: UNKNOWN, 2.3: UNKNOWN, 2.4: REQUIRED to 2.2: UNKNOWN, 2.3: UNKNOWN, 2.4: DONE
ruby_2_4 r57853 merged revision(s) 57293.
Updated by Benoit_Tigeot (Benoit Tigeot) over 8 years ago
[email protected] (Thomas Walpole) wrote:
This appears to be fixed in latest trunk now. Is there any info on when a 2.4.1 will be released - seems like a pretty serious issue.
I am in the same situation actually. Locked because of this issue. Any idea when 2.4.1 will be released?