5.4-stable review patch. If anyone has any objections, please let me know. ------------------ From: Wenhua Lin <Wenhua.Lin@xxxxxxxxxx> [ Upstream commit 26d9e5640d2130ee16df7b1fb6a908f460ab004c ] The drivers uses a mutex and I2C bus access in its PMIC EIC chip get implementation. This means these functions can sleep and the PMIC EIC chip should set the can_sleep property to true. This will ensure that a warning is printed when trying to get the value from a context that potentially can't sleep. Fixes: 348f3cde84ab ("gpio: Add Spreadtrum PMIC EIC driver support") Signed-off-by: Wenhua Lin <Wenhua.Lin@xxxxxxxxxx> Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@xxxxxxxxxx> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx> --- drivers/gpio/gpio-pmic-eic-sprd.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpio/gpio-pmic-eic-sprd.c b/drivers/gpio/gpio-pmic-eic-sprd.c index 05000cace9b24..abe01518bf19f 100644 --- a/drivers/gpio/gpio-pmic-eic-sprd.c +++ b/drivers/gpio/gpio-pmic-eic-sprd.c @@ -338,6 +338,7 @@ static int sprd_pmic_eic_probe(struct platform_device *pdev) pmic_eic->chip.set_config = sprd_pmic_eic_set_config; pmic_eic->chip.set = sprd_pmic_eic_set; pmic_eic->chip.get = sprd_pmic_eic_get; + pmic_eic->chip.can_sleep = true; pmic_eic->intc.name = dev_name(&pdev->dev); pmic_eic->intc.irq_mask = sprd_pmic_eic_irq_mask; -- 2.40.1