Failure in stdlib/tst-arc4random-thread testing glibc 2.41 for release.

[Florian Weimer]

* Adhemerval Zanella Netto:

>> I think I found it: the len updates are not synchronized with fork, and
>> neither is the transfer from the states array to a thread and back.
>> After adding barriers, the states array may still contain null pointers,
>> so we should check for that and skip over them.
>
> Do you mean adding atomic_thread_fence_seq_cst() on 'len' updates? Yeah
> it makes sense and it explains why this only happens intermittently. 

Yes, around the len updates.  We need to make sure that the state
pointer is never in two places at the same time.

Should I send a patch?

Thanks,
Florian