On Wed,  3 Nov 2021 21:48:27 +1000
Alistair Francis [off-list ref] wrote:

[...]

+static int cyttsp5_probe(struct device *dev, struct regmap *regmap, int irq,
+			 const char *name)
+{
+	struct cyttsp5 *ts;
+	struct cyttsp5_sysinfo *si;
+	int rc = 0, i;
+
+	ts = devm_kzalloc(dev, sizeof(*ts), GFP_KERNEL);
+	if (!ts)
+		return -ENOMEM;
+
+	/* Initialize device info */
+	ts->regmap = regmap;
+	ts->dev = dev;
+	si = &ts->sysinfo;
+	dev_set_drvdata(dev, ts);
+
+	/* Initialize mutexes and spinlocks */
+	mutex_init(&ts->system_lock);
+
+	/* Initialize wait queue */
+	init_waitqueue_head(&ts->wait_q);
+
+	/* Power up the device */
+	ts->vdd = regulator_get(dev, "vdd");
+	if (IS_ERR(ts->vdd)) {
+		rc = PTR_ERR(ts->vdd);
+		dev_set_drvdata(dev, NULL);
+		kfree(ts);
+		return rc;
+	}
+
+	rc = regulator_enable(ts->vdd);
+	if (rc) {
+		regulator_put(ts->vdd);
+		dev_set_drvdata(dev, NULL);
+		kfree(ts);
+		return rc;
+	}
+
+	/* Reset the gpio to be in a reset state */
+	ts->reset_gpio = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_LOW);

        so we deassert reset for

+	if (IS_ERR(ts->reset_gpio)) {
+		rc = PTR_ERR(ts->reset_gpio);	
+		dev_err(dev, "Failed to request reset gpio, error %d\n", rc);
+		return rc;
+	}

	
	almost no time 

+	gpiod_set_value(ts->reset_gpio, 1);

        and then assert it, keeping the chip in reset

+
+	/* Need a delay to have device up */
+	msleep(20);
+

	and why it should wake up?

	I reversed the logic here to test. I have 

        reset-gpios = <&gpio5 13 GPIO_ACTIVE_LOW>;

	in my tests.

+	rc = devm_request_threaded_irq(dev, irq, NULL, cyttsp5_handle_irq,
+				       IRQF_TRIGGER_FALLING | IRQF_ONESHOT,
+				       name, ts);

        falling or level low (according to the example in the
        dt schema)?	

Regards,
Andreas