RFR: JDK-8129626: G1: set_in_progress() and clear_started() needs a barrier on non-TSO platforms

[Bengt Rutisson]

Hi Per,

Thanks for looking at this!

On 2015-06-24 10:09, Per Liden wrote:
> Hi Bengt,
>
> On 2015-06-24 09:28, Bengt Rutisson wrote:
>>
>>
>> Hi Vitaly,
>>
>> On 2015-06-23 23:53, Vitaly Davidovich wrote:
>>>
>>> Naive question - could this be converted into a numeric state field
>>> that indicates the lifecycle (e.g 1 = in progress, 2 = started)?
>>>
>>
>> Good question! That's a much simpler and more stable solution.
>>
>> New webrev:
>> http://cr.openjdk.java.net/~brutisso/8129626/webrev.02/
>
> This looks much nicer, and I can't think of any reason why that 
> wouldn't work. One little request though, can we name the states to 
> match the function names, like Idle/Started/InProgress? And 
> clear_in_progress() should probably be set_idle() to align with the rest.

Yes, I was struggling a bit with the naming. What do you think about this?

cr.openjdk.java.net/~brutisso/8129626/webrev.03/

Thanks,
Bengt
>
> cheers,
> /Per
>
>>
>> Thanks,
>> Bengt
>>
>>
>>
>>> sent from my phone
>>>
>>> On Jun 23, 2015 5:42 PM, "bill pittore" <bill.pittore at oracle.com
>>> <mailto:bill.pittore at oracle.com>> wrote:
>>>
>>>     Generally when you have a storestore on the write side you need a
>>>     loadload on the read side to prevent the second read from floating
>>>     above the first one. The reading thread could read in_progress as
>>>     0 before it reads starting. Meanwhile the write thread writes 1 to
>>>     in_progress, issues storestore, clears starting. Reading thread
>>>     then reads starting as 0. I don't know if the CGC mutex somehow
>>>     eliminates this issue as I'm not familiar with the code in detail.
>>>
>>>     bill
>>>
>>>     On 6/23/2015 4:25 PM, Bengt Rutisson wrote:
>>>>
>>>>     Hi everyone,
>>>>
>>>>     Could I have a couple of reviews for this change?
>>>>
>>>>     https://bugs.openjdk.java.net/browse/JDK-8129626
>>>>     http://cr.openjdk.java.net/~brutisso/8129626/webrev.00/
>>>> <http://cr.openjdk.java.net/%7Ebrutisso/8129626/webrev.00/>
>>>>
>>>>     We need to add a barrier between the calls to set_in_progress()
>>>>     and clear_started() to make sure that other threads sees the
>>>>     correct value when they use ConcurrentMarkThread::during_cycle().
>>>>
>>>>     Thanks to Per and Bertrand for helping out identifying and
>>>>     sorting this out.
>>>>
>>>>     From the bug report:
>>>>
>>>>     ConcurrentMarkThread::during_cycle() is implemented as:
>>>>
>>>>     bool during_cycle() { return started() || in_progress(); }
>>>>
>>>>     So, it checks both ConcurrentMarkThread::_started and
>>>>     ConcurrentMarkThread::_in_progress and they are meant to overlap.
>>>>     That is, we should not set _started to false until after we have
>>>>     set _in_progress to true. This is done in sleepBeforeNextCycle():
>>>>
>>>>     void ConcurrentMarkThread::sleepBeforeNextCycle() {
>>>>       // We join here because we don't want to do the
>>>>     "shouldConcurrentMark()"
>>>>       // below while the world is otherwise stopped.
>>>>       assert(!in_progress(), "should have been cleared");
>>>>
>>>>       MutexLockerEx x(CGC_lock, Mutex::_no_safepoint_check_flag);
>>>>       while (!started() && !_should_terminate) {
>>>>         CGC_lock->wait(Mutex::_no_safepoint_check_flag);
>>>>       }
>>>>
>>>>       if (started()) {
>>>>         set_in_progress();
>>>>         clear_started();
>>>>       }
>>>>     }
>>>>
>>>>     On non-TSO platforms there is a risk that the write to
>>>>     _in_progress (from set_in_progress()) is seen by other threads
>>>>     after the write to _started (in clear_started()). In that case
>>>>     there is a window when during_cycle() may return false even
>>>>     though we are in a concurrent cycle.
>>>>
>>>>
>>>>     Thanks,
>>>>     Bengt
>>>
>>