Re: [PATCH v2] reset: Add KUnit test cases

[Tzung-Bi Shih]

On Tue, Nov 03, 2026 at 11:21:53AM +0100, Philipp Zabel wrote:
> On Mo, 2026-11-02 at 23:52 +0800, Tzung-Bi Shih wrote:
> > +	/* supplier0 unregisters. */
> > +	complete(&ctx[0].enter);
> > +	ret = read_poll_timeout(reset_control_status, val, val == -ENODEV,
> > +				1000, 100000, false, rtc->rstc);
> > +	if (ret) {
> > +		for (i = 2; i >= 0; --i)
> > +			kthread_stop_put(ctx[i].thread);
> > +		KUNIT_ASSERT_EQ(test, ret, 0);
> > +	}
> > +	/* supplier0 can't exit.  It's waiting for all consumers. */
> 
> This comment looks like an untested assertion. For all we know,
> reset_controller_unregister() may have already returned by now.
> 
> How about signalling a completion "ctx->exited" after
> reset_controller_unregister() returns, calling
> wait_for_completion_timeout() here, and asserting it times out?

Fixed in v3[1].

[1] https://lore.kernel.org/all/20261104160123.6784-1-tzungbi@xxxxxxxxxx