NACK: Xenial SRU - UbuntuKVM guest crashed while running I/O stress test with Ubuntu kernel 4.4.0-47-generic

On 04/10/2017 10:40 AM, Seth Forshee wrote:
> On Fri, Apr 07, 2017 at 03:37:46PM +0100, Tim Gardner wrote:
>> https://bugs.launchpad.net/bugs/1659111
>>
>> See attached.
>>
>> Though this looks like a big scary patch set, the 2 big prerequisite
>> backports are largely mechanical changes with no functional changes.
>> Subsequent patches are clean cherry-picks and righteous bug fixes (or
>> minor backports). I've run some stress-ng and ecryptfs stress tests
>> without any errors. I/O stress testing from IBM is also positive.
> 
> It's difficult to really review some of those without knowing the code
> better than I do. The backports do look correct though, and most of the
> scarier changes are clean cherry picks. I did find a couple of minor
> problems though.
> 
> It looks like you accidentally added a new file in your backport of
> dc3b17cc8bf2 "block: Use pointer to backing_dev_info from
> request_queue". It's a file that was touched in the upstream commit but
> not present in 4.4.
> 
> It also looks like you need to s/cherry picked/backported/ for many of
> the commits.
> 
> Seth
> 

Refreshed the branch to address your comments. See attached.

On Tue, Apr 11, 2017 at 12:02:35PM -0600, Tim Gardner wrote:
> On 04/10/2017 10:40 AM, Seth Forshee wrote:
> > On Fri, Apr 07, 2017 at 03:37:46PM +0100, Tim Gardner wrote:
> >> https://bugs.launchpad.net/bugs/1659111
> >>
> >> See attached.
> >>
> >> Though this looks like a big scary patch set, the 2 big prerequisite
> >> backports are largely mechanical changes with no functional changes.
> >> Subsequent patches are clean cherry-picks and righteous bug fixes (or
> >> minor backports). I've run some stress-ng and ecryptfs stress tests
> >> without any errors. I/O stress testing from IBM is also positive.
> > 
> > It's difficult to really review some of those without knowing the code
> > better than I do. The backports do look correct though, and most of the
> > scarier changes are clean cherry picks. I did find a couple of minor
> > problems though.
> > 
> > It looks like you accidentally added a new file in your backport of
> > dc3b17cc8bf2 "block: Use pointer to backing_dev_info from
> > request_queue". It's a file that was touched in the upstream commit but
> > not present in 4.4.
> > 
> > It also looks like you need to s/cherry picked/backported/ for many of
> > the commits.
> > 
> > Seth
> > 
> 
> Refreshed the branch to address your comments. See attached.

Looks better, and positive test results.

Acked-by: Seth Forshee <seth.forshee@canonical.com>