[Precise][SRU][PATCH 0/1] Revert upstream stable commit 73d63d03

[Leann Ogasawara]

On Tue, 2012-04-10 at 15:49 +0200, Stefan Bader wrote:
> On 10.04.2012 15:35, leann.ogasawara at canonical.com wrote:
> > From: Leann Ogasawara <leann.ogasawara at canonical.com>
> > 
> > === Precise SRU Justification ===
> > The latest Precise Kernel (linux-headers-3.2.0-22-generic 3.2.0-22.35)
> > does not boot in pvops mode due to a kernel NULL pointer dereference
> > oops generated in xen_irq_init.
> > 
> > === Fix ===
> > Revert commit 73d63d03 which was included in upstream stable v3.2.14.
> > Based on discusson upstream regarding this commit, it appears it should
> > not have been applied to upstream stable v3.2.y:
> > 
> > http://www.mail-archive.com/stable@vger.kernel.org/msg04782.html
> > 
> > commit 73d63d038ee9f769f5e5b46792d227fe20e442c5
> > Author: Suresh Siddha <suresh.b.siddha at intel.com>
> > Date: Mon Mar 12 11:36:33 2012 -0700
> > 
> >     x86/ioapic: Add register level checks to detect bogus io-apic
> > entries
> > 
> > Leann Ogasawara (1):
> >   Revert "x86/ioapic: Add register level checks to detect bogus io-apic
> >     entries"
> > 
> >  arch/x86/kernel/apic/io_apic.c |   40 ++++++++--------------------------------
> >  1 files changed, 8 insertions(+), 32 deletions(-)
> > 
> > === Testing ===
> > The bug reporter has tested a kernel with the above commit reverted.
> > They confirm the machine boots properly.
> > 
> > Note, the same patch also went into upstream stable v3.0.y and hit
> > Oneiric, commit f9021c45.  I do not have test confirmation for Oneiric,
> > but I believe the evidence to be sufficient enough to support reverting
> > in Oneiric as well.
> >  
> 
> This one seems to work around the issue in 3.4... was that tried?
> 
> 
> commit 2531d64b6fe2724dc432b67d8dc66bd45621da0b
> Author: Konrad Rzeszutek Wilk <konrad.wilk at oracle.com>
> Date:   Tue Mar 20 15:04:18 2012 -0400
> 
>     xen/x86: Workaround 'x86/ioapic: Add register level checks to detect bogus i
> 
>     The above mentioned patch checks the IOAPIC and if it contains...

The above patch was not tested, but indeed something I can help build a
test kernel for.  For the time being, I think the safest route is to
proceed with the revert, which has been tested, as we don't have much
time to make the upload window today.  I'm assuming the next upstream
stable v3.2.15 release will either contain this revert or whatever
additional fixes are necessary and we'll pull that in via our normal SRU
process.

Thanks,
Leann