On 2017/8/28 8:25, Fabio Estevam wrote:
The reset GPIO can be connected to a I2C or SPI IO expander, which may
sleep, so it is safer to use the gpiod_set_value_cansleep() variant
instead.
Signed-off-by: Fabio Estevam <festevam@xxxxxxxxx>
---
Changes since v1:
- Resending with Shawn Lin added on Cc.
drivers/pci/host/pcie-rockchip.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
Acked-by: Shawn Lin <shawn.lin@xxxxxxxxxxxxxx>
diff --git a/drivers/pci/host/pcie-rockchip.c b/drivers/pci/host/pcie-rockchip.c
index 2eccd53..124b280 100644
--- a/drivers/pci/host/pcie-rockchip.c
+++ b/drivers/pci/host/pcie-rockchip.c
@@ -537,7 +537,7 @@ static int rockchip_pcie_init_port(struct rockchip_pcie *rockchip)
int err, i;
u32 status;
- gpiod_set_value(rockchip->ep_gpio, 0);
+ gpiod_set_value_cansleep(rockchip->ep_gpio, 0);
err = reset_control_assert(rockchip->aclk_rst);
if (err) {
@@ -682,7 +682,7 @@ static int rockchip_pcie_init_port(struct rockchip_pcie *rockchip)
rockchip_pcie_write(rockchip, PCIE_CLIENT_LINK_TRAIN_ENABLE,
PCIE_CLIENT_CONFIG);
- gpiod_set_value(rockchip->ep_gpio, 1);
+ gpiod_set_value_cansleep(rockchip->ep_gpio, 1);
/* 500ms timeout value should be enough for Gen1/2 training */
err = readl_poll_timeout(rockchip->apb_base + PCIE_CLIENT_BASIC_STATUS1,
