RFR(S): 8219517: assert(false) failed: infinite loop in PhaseIterGVN::optimize

[Nils Eliasson]

On 2019-03-08 19:27, Vladimir Kozlov wrote:
> Hi Nils,
>
> What if it is normal diamond shaped graph and no loop? 

It can't be. The loads memory edge is 'mem'. 'mem' is a phi, and we loop 
over its in-edges . The optimization only triggers when: mem 
==opt_mem_chain(mem->in(x)). That can only happen when there is a loop.

(And my addition is to block the optimization when both are loops: mem 
== opt_mem_chain(mem->in(1)) == opt_mem_chain(mem->in(2)).

Regards,

/ Nils

> You can have case when both branches point to the same memory too. May 
> be add is_Loop() check.
>
> Thanks,
> Vladimir
>
>
> On 3/8/19 2:52 AM, Nils Eliasson wrote:
>> Hi all,
>>
>> Background:
>>
>> We can get stuck in an infinite loop in IGVN. The method reproducing 
>> the problem is quite a big graph, and after some optimization, a huge 
>> loop will die. But since it is so big, it takes a while before it has 
>> been pruned.
>>
>> In an inner loop there is a phi on memory that gets reduced to a self 
>> looping heart, with a membar on each in edge. There is also a 
>> connected region that keeps it alive. (From the start there is other 
>> memory state coming into this loop, but it gets disconnected early 
>> when the loop dies.)
>>
>> +---+                 +---+
>> |   v                 v   |
>> | Membar +-+ +---+ Membar |
>> |          | |            |
>> |          v v            |
>> |          Phi            |
>> |          + + +          |
>> |          | | |          |
>> +----------+ | +----------+
>>               |
>>               v
>>             LoadN
>>
>> In IGVN, Ideal() will be called on the Load.
>>
>> On iteration 1 - A split_through_phi on one edge will be performed, 
>> because we can prove that other edge of the phi is a loop. Now the 
>> Load hangs of one of the membars.
>>
>> On iteration 2 - Optimize_memory_chain will suggest the in to the 
>> membar as a more ideal memory, and then the load get the phi back as 
>> the memory input.
>>
>> Repeat.
>>
>> I have gone great lengths to show that this code is part of a huge 
>> loop, that is dead, and will be eliminated in due time.
>>
>> My suggested solution to breaking the infinite loop, is to change the 
>> first case, by simply not perform the memory replacement when both 
>> inputs are self loops.
>>
>> https://bugs.openjdk.java.net/browse/JDK-8219517
>>
>> http://cr.openjdk.java.net/~neliasso/8219517/webrev.01/
>>
>> Regards,
>>
>> Nils
>>
>>
>>
>>