Re: [PATCH] driver core: flush async calls before testing driver removal

[Vladimir Zapolskiy]

Hello Dmitry,

On 12/10/2016 03:59 AM, Dmitry Torokhov wrote:
> On Fri, Dec 9, 2016 at 4:15 PM, Vladimir Zapolskiy <vz@xxxxxxxxx> wrote:
>> If CONFIG_DEBUG_TEST_DRIVER_REMOVE option is enabled a number of false
>> positives are reported for ATA controller drivers, because ATA port
>> probes are done asynchronously, and the same problem may also touch
>> other asynchronously probed drivers.
>>
>> To reduce the rate of false reports on boot call async_synchronize_full()
>> before attempting to remove a driver, the same is done in delete_module()
>> syscall for all possible drivers and in __device_release_driver() function
>> for asynchronously probed drivers.
> 
> I'd say CONFIG_DEBUG_TEST_DRIVER_REMOVE did what it was supposed to do
> and uncovered a big in ATA drivers. Since driver core did not
> asynchronously scheduled those actions it should not wait for their
> completion either, but either ATA core or drivers should wait for
> probing to complete before allowing remove() methods to run.
> 

can you please share the idea why?

I haven't managed to find any problems with ATA subsystem and drivers,
my fuzz testing by inserting delays to postpone scheduled execution
of async_port_probe() don't show any problems also.

So I believe the problem is with the test itself.

--
With best wishes,
Vladimir