Re: [PATCH] SYSVIPC - Fix the ipc structures initialization

[Manfred Spraul <manfred@colorfullife.com>]

Andrew Morton wrote:
>> I don't understand what's going on... I hope I'm not the only one who
>> still get it because I doubt about this bug.
>>     
One thing is odd:
The test app uses fork(), two processes.
One process constantly does
    msgctl(0, IPC_STAT, &buf);

The other process creates a new msg queue, then it destroys it.
But: The probablity that the new queue has the msqid is 0 is tiny 
[impossible?].
That means the idr_find() call should [must?] fail.
But then the spin_lock() should never be executed. Why do you get an 
oops in that call?

Do you still use the test app that is on bugzilla?

>>     
>
> Time is starting to press on this one.  Is there something which we can
> revert which would fix this bug?
>
> Thanks.
>