[PATCH v2 2/2] staging: greybus: loopback_test: Fix race preventing test completion

Axel Haslam ahaslam at baylibre.com
Tue Jan 3 12:24:14 UTC 2017


On Tue, Jan 3, 2017 at 12:51 PM, Bryan O'Donoghue
<pure.logic at nexus-software.ie> wrote:
> On 03/01/17 10:17, Axel Haslam wrote:
>> as i see it,  a successful test means:
>> 1- each interfaces should send an event upon completion.
> yes
>
>> 2- the iteration count should equal iteration_max on each of the interfaces
>
> yes
>
>> what  am i missing?
>
> count == max implies the kernel thread is still working and we should
> continue to sleep.

i think you mean count != max here.

>
> I am interested in why count != max when in user-space number_of_events
> == t->poll_count - I don't see how that's logically possible, though it

right, i think this is the issue.

> makes me wonder why the check is_complete() was put in place - you must
> have seen this yourself ?

as a general sanity check. in case the event count were wrong for some reason.
which seems to be happening here. i dont remember if i was actually seeing it
and under which case, i will give it a try and see if i can reproduce it.

looking at your patch it seems the count will eventually complete
and equal "max_iterations" maybe the event is somehow sent
early.

>
> As a fundamental though - unless count == max, the test is not done.


More information about the devel mailing list