Regression in RCU subsystem in latest mainline kernel

[Steven Rostedt]

On Fri, 2013-06-14 at 14:46 +0200, Rojhalat Ibrahim wrote:
> On Friday 14 June 2013 05:28:00 Paul E. McKenney wrote:
> > On Fri, Jun 14, 2013 at 01:47:00PM +0200, Rojhalat Ibrahim wrote:
> > > Hi,
> > > 
> > > the current mainline kernel from git reproducibly hangs on my Freescale
> > > PowerPC P5020DS development system. Most of the time the hang occurs
> > > during
> > > boot (kernel just stops booting, system hangs), sometimes shortly after
> > > booting (system hangs, no message). Very often booting stops at the
> > > initialization of the Intel PRO 1000 PCIe network device (e1000e).

I was pretty much able to reproduce this on my PA Semi PPC box. Funny
thing is, when I type on the console, it makes progress. Anyway, it
seems that powerpc has an issue with irq_work(). I'll try to get some
time either tonight or next week to figure it out.

Thanks!

-- Steve

> > > 
> > > I used git bisect and arrived at this commit:
> > > 
> > > 016a8d5be6ddcc72ef0432d82d9f6fa34f61b907
> > > rcu: Don't call wakeup() with rcu_node structure ->lock held
> > > 
> > > With versions before that commit the system runs normally.
> > 
> > Odd, given that this fix was intended to prevent hangs.  Could you
> > please send your .config?
> > 
> > 							Thanx, Paul