[#82706] [Ruby trunk Bug#13851] getting "can't modify string; temporarily locked" on non-frozen instances — cardoso_tiago@...
Issue #13851 has been updated by chucke (Tiago Cardoso).
3 messages
2017/09/07
[#82853] [Ruby trunk Bug#13916] Race condition when sending a signal to a new fork — russell.davis@...
Issue #13916 has been reported by russelldavis (Russell Davis).
3 messages
2017/09/19
[#82892] [Ruby trunk Bug#13921] buffered read_nonblock doesn't work as expected using SSLSocket — cardoso_tiago@...
Issue #13921 has been updated by chucke (Tiago Cardoso).
3 messages
2017/09/20
[ruby-core:82832] [Ruby trunk Misc#13163] Uncaught exceptions may not be reported when Thread#report_on_exception=true and Thread#abort_on_exception=true
From:
eregontp@...
Date:
2017-09-16 14:47:26 UTC
List:
ruby-core #82832
Issue #13163 has been updated by Eregon (Benoit Daloze). +1 The patch looks good. The message of the assert_warn should be changed though. ---------------------------------------- Misc #13163: Uncaught exceptions may not be reported when Thread#report_on_exception=true and Thread#abort_on_exception=true https://bugs.ruby-lang.org/issues/13163#change-66719 * Author: tagomoris (Satoshi TAGOMORI) * Status: Feedback * Priority: Normal * Assignee: ---------------------------------------- When we set true on both of `Thread#report_on_exception` and `Thread#abort_on_exception`, it just works as `Thread#abort_on_exception` and raised error will be re-raised in main thread. But in our case, main thread sometimes in `begin`-`rescue` clause, or sometimes out of `begin`-`rescue` clause. When main thread in `begin`-`rescue` clause (and main thread code may raise errors in their own business), we cannot know that threads are dead by errors. What I expected by specifying both of `Thread#report_on_exception=true` and `Thread#abort_on_exception=true`: * "Thread terminated with exception" is reported always (even if `Thread#abort_on_exception` is set) * Raised error is re-raised in main thread if `Thread#abort_on_exception` is true How do you think about this idea? -- https://bugs.ruby-lang.org/ Unsubscribe: <mailto:ruby-core-request@ruby-lang.org?subject=unsubscribe> <http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>