Re: [patch 00/14] x86/irq: Plug various vector cleanup races

[Joe Lawrence]

On 01/18/2016 10:43 AM, Borislav Petkov wrote:
> On Mon, Jan 18, 2016 at 10:00:49AM -0500, Joe Lawrence wrote:
>> INFO: task kworker/0:1:1506 blocked for more than 120 seconds.
>>       Tainted: P           OE   4.3.0sra12+ #50
> 
> P as in Proprietary?
> 
> Can you trigger without that module?
> 

Hi Boris,

Unfortunately, yes to proprietary.

I had a Stratus proprietary driver in place that enables a test to
hotplug add/remove entire PCI subtrees at random points of
initialization.  That driver allows us to insert various hardware faults
to facilitate testing the Stratus ftServer platform and OS device
drivers against PCI master abort detection and cleanup.

I can try w/o the driver, though the test will be drastically different
-- polite sysfs PCI removal of a single device (in this case, the xHCI
host controller).  An interesting data point to gather nonetheless.

-- Joe