On Tue, Jun 28, 2022 at 09:21:12AM +0200, Fabrice Gasnier wrote:
> On 6/27/22 15:17, Heikki Krogerus wrote:
> > Hi,
> >
> > On Fri, Jun 24, 2022 at 05:54:11PM +0200, Fabrice Gasnier wrote:
> >> +static int ucsi_stm32g0_probe(struct i2c_client *client, const struct i2c_device_id *id)
> >> +{
> >> + struct device *dev = &client->dev;
> >> + struct ucsi_stm32g0 *g0;
> >> + int ret;
> >> +
> >> + g0 = devm_kzalloc(dev, sizeof(*g0), GFP_KERNEL);
> >> + if (!g0)
> >> + return -ENOMEM;
> >> +
> >> + g0->dev = dev;
> >> + g0->client = client;
> >> + init_completion(&g0->complete);
> >> + i2c_set_clientdata(client, g0);
> >> +
> >> + g0->ucsi = ucsi_create(dev, &ucsi_stm32g0_ops);
> >> + if (IS_ERR(g0->ucsi))
> >> + return PTR_ERR(g0->ucsi);
> >> +
> >> + ucsi_set_drvdata(g0->ucsi, g0);
> >> +
> >> + /* Request alert interrupt */
> >> + ret = request_threaded_irq(client->irq, NULL, ucsi_stm32g0_irq_handler, IRQF_ONESHOT,
> >> + dev_name(&client->dev), g0);
> >> + if (ret) {
> >> + dev_err_probe(dev, ret, "request IRQ failed\n");
> >> + goto destroy;
> >> + }
> >> +
> >> + ret = ucsi_register(g0->ucsi);
> >> + if (ret) {
> >> + dev_err_probe(dev, ret, "ucsi_register failed\n");
> >> + goto freeirq;
> >> + }
> >
> > If there isn't UCSI firmware, then ucsi_register() will always safely
> > fail here, right?
>
> Hi Heikki,
>
> Yes, in such a case, the first i2c read (UCSI_VERSION) in
> ucsi_register() will return an error and safely fail here.
Okay, thanks.
--
heikki
