[PATCH] powerpc: set stack limit properly in crit_transfer_to_handler

[Yoder Stuart-B08248]

> -----Original Message-----
> From: Kumar Gala [mailto:galak at kernel.crashing.org]
> Sent: Monday, July 16, 2012 3:36 PM
> To: Yoder Stuart-B08248
> Cc: benh at kernel.crashing.org; sfr at canb.auug.org.au; paulus at samba.org; linuxppc-dev at lists.ozlabs.org;
> agraf at suse.de
> Subject: Re: [PATCH] powerpc: set stack limit properly in crit_transfer_to_handler
> 
> 
> On Jul 16, 2012, at 2:06 PM, Stuart Yoder wrote:
> 
> > From: Stuart Yoder <stuart.yoder at freescale.com>
> >
> > without setting the stack limit like this there is the possibility
> > of stack overflow which corrupts the thread info but
> > is not detected by stack overflow detection
> >
> > Signed-off-by: Stuart Yoder <stuart.yoder at freescale.com>
> > ---
> > arch/powerpc/kernel/entry_32.S |   12 ++++++++++--
> > 1 files changed, 10 insertions(+), 2 deletions(-)
> 
> Do we need this for debug & mcheck ?

Maybe... I don't know the history of this code
and why this is only done in the critical exception
handlers.

But, this particular patch is to undo/fix something I
introduced in:
powerpc: Use CURRENT_THREAD_INFO instead of open coded assembly

Stuart