Messages in this thread | | | Date | Tue, 24 Jun 2008 23:15:47 +0200 | From | Ingo Molnar <> | Subject | Re: [Bug #10815] 2.6.26-rc4: RIP find_pid_ns+0x6b/0xa0 |
| |
* Paul E. McKenney <paulmck@linux.vnet.ibm.com> wrote:
> > > Also Core 2 Duo, x86_64 [1] > > > > > > Race is wide -- 60 seconds of rcutorture is enough. > > This is rcutorture by itself, or in parallel with LTP/kernbench? > > (I have mostly been running on 4-CPU boxes without failure either way, > so will try a dual-CPU box.)
FYI, i've been running rcutorture for days on lots of testboxes ranging all across the x86 spectrum from single CPU, through dual-core, to dual-socket, dual-socket HT, 8-way and 16-way - both 32-bit and 64-bit x86.
Not a single failure has been detected in thousands of bootups of random kernels (rcu-preempt + rcutorture was a frequent combination tested). I added a WARN_ON() to rcutorture failures so it should show up very clearly.
This makes me suspect that it might be some special environment issue - gcc for example. I'm using gcc 4.2.2 on most of the testboxes.
Ingo
| |