[Cbe-oss-dev] [RFC] Cell: shutdown method for spu_sysdev_class

[Vivek Goyal]

On Sun, May 06, 2007 at 11:27:26PM -0300, Luke Browning wrote:
> On Mon, 2007-05-07 at 11:20 +1000, Michael Ellerman wrote:
> 
> > 
> > Nothing gets called in the kdump case, the kernel is panicked.
> > 
> > For kexec it makes sense to do any cleanup you can, but for kdump you
> > need to expect that the previous kernel has left everything screwed up.
> > 
> > One thing we should consider for kdump is stopping execution of the
> > spus. Although it increases the amount of code we're executing in the
> > panicked kernel, I think it might still be a good idea. We'd want to use
> > a special shutdown routine which walks as few data structures as
> > possible.
> > 
> > cheers
> 
> Hi Mike,
> 
> I got a little confused with your use of kexec and kdump, but I believe
> we are on the same page.
> 
> Lucio is adding logic to stop the spus in the exception path.  This is
> part of the default kernel which is compiled with kexec.  During the
> processing of the exception, the previously loaded kdump kernel is
> invoked.  It boots, the system dump file is saved, and the the kdump
> capture kernel is shutdown so that the default kernel is rebooted.  
> 
> We expect the new shutdown routines to be invoked in the kdump kernel,
> but not the kexec kernel, at least not in the case of system crash.
> 
> My question for Mike, Prasad, or Vivek is:
> 
> When the default kernel is rebooted, how is the kexec -p command run
> again.  Seems odd to expect the user to do it.  Is there some kind of
> init / rc script that performs this operation each boot?
> 

Hi Luke,

Yes, kdump kernel is automatically loaded (kexec -p) with the help of init
scripts when system boots and if user chose to enable kdump service at
installation time or some other time.

Thanks
Vivek

> Luke
> 
>