August 2011 - macruby-tickets - lists.macosforge.org

[MacRuby] #1102: segfault when trying const_get on unitialized constant with rspec2
by MacRuby 04 Aug '11

04 Aug '11

#1102: segfault when trying const_get on unitialized constant with rspec2 ------------------------------------+--------------------------------------- Reporter: mrada@… | Owner: lsansonetti@… Type: defect | Status: new Priority: blocker | Milestone: Component: MacRuby | Keywords: ------------------------------------+--------------------------------------- When I try to run const_get inside of a begin/rescue/end statement it should be safe, but when I try to spec the code in rspec it causes MacRuby to segfault instead of rescuing. I was using rspec 2.4 with the latest nightly build (Jan 6). This may be related to ticket #1095, I have attached a reduced script. -- Ticket URL: <http://www.macruby.org/trac/ticket/1102> MacRuby <http://macruby.org/>

1 3

[MacRuby] #1372: Exceptions thrown inside dispatch queue blocks propagate as uncaught objective-c exceptions
by MacRuby 04 Aug '11

04 Aug '11

#1372: Exceptions thrown inside dispatch queue blocks propagate as uncaught objective-c exceptions ------------------------------------+--------------------------------------- Reporter: mrada@… | Owner: lsansonetti@… Type: defect | Status: new Priority: blocker | Milestone: Component: MacRuby | Keywords: reduction ------------------------------------+--------------------------------------- Specifically, I had this problem because a test assertion inside of a block failed. Reduction: {{{ a = Dispatch::Queue.new 'org.macruby.test' a.sync { raise Exception, 'oh noses' } }}} -- Ticket URL: <http://www.macruby.org/trac/ticket/1372> MacRuby <http://macruby.org/>

1 5

[MacRuby] #1371: MacRuby evaluates the first method defined using define_method through #send
by MacRuby 01 Aug '11

01 Aug '11

#1371: MacRuby evaluates the first method defined using define_method through #send ------------------------------------+--------------------------------------- Reporter: mrada@… | Owner: lsansonetti@… Type: defect | Status: new Priority: blocker | Milestone: Component: MacRuby | Keywords: reduction ------------------------------------+--------------------------------------- Discovered this problem while on a crusade to eliminate string eval usage in hot cocoa: Reduction: {{{ a = Module.new a.send :define_method, :test do |value| puts 'first' end a.send :define_method, :something_else do |value| puts 'second' end }}} If you run the code, it will print 'first', but not second. If you then swap the order of the define_method calls and run it again, it will print 'second', but will not print 'first'. -- Ticket URL: <http://www.macruby.org/trac/ticket/1371> MacRuby <http://macruby.org/>

1 1

Re: [MacRuby] #542: Attempt to unlock a mutex which is not locked (ThreadError)
by MacRuby 01 Aug '11

01 Aug '11

#542: Attempt to unlock a mutex which is not locked (ThreadError) -------------------------------------+-------------------------------------- Reporter: valerii.hiora@… | Owner: lsansonetti@… Type: defect | Status: new Priority: major | Milestone: MacRuby 1.0 Component: MacRuby | Keywords: -------------------------------------+-------------------------------------- Comment(by jhemmelg@…): OK, I was looking further into this. It looks like the mutex is not re- locked when cond.wait returns in the main thread. The synchronize method then tries to unlock it and finds it already unlocked. The documentation for ConditionVariable::wait claims to re-lock the mutex on wakeup. I can't see where that is supposed to be happening. It may be some magic with the thread wakeup logic, but I can't see it in either the macruby code or the cruby 1.9.2 code. I'll try to keep digging but I wouldn't be disappointed if someone else figures it out first. -- Ticket URL: <http://www.macruby.org/trac/ticket/542#comment:10> MacRuby <http://macruby.org/>

1 0