On 01-01-21, 16:54, Yangtao Li wrote:
We should use dev_pm_opp_put_clkname() to free opp table each time dev_pm_opp_of_add_table() got error.
Signed-off-by: Yangtao Li tiny.windzz@gmail.com
drivers/tty/serial/qcom_geni_serial.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/drivers/tty/serial/qcom_geni_serial.c b/drivers/tty/serial/qcom_geni_serial.c index 291649f02821..5aada7ebae35 100644 --- a/drivers/tty/serial/qcom_geni_serial.c +++ b/drivers/tty/serial/qcom_geni_serial.c @@ -1438,9 +1438,12 @@ static int qcom_geni_serial_probe(struct platform_device *pdev) return PTR_ERR(port->se.opp_table); /* OPP table is optional */ ret = dev_pm_opp_of_add_table(&pdev->dev);
- if (ret && ret != -ENODEV) {
dev_err(&pdev->dev, "invalid OPP table in device tree\n");goto put_clkname;
if (ret) {
dev_pm_opp_put_clkname(port->se.opp_table);if (ret != -ENODEV) {dev_err(&pdev->dev, "invalid OPP table in device tree\n");return ret;}}
port->private_data.drv = drv;
@@ -1482,7 +1485,6 @@ static int qcom_geni_serial_probe(struct platform_device *pdev) return 0; err: dev_pm_opp_of_remove_table(&pdev->dev); -put_clkname: dev_pm_opp_put_clkname(port->se.opp_table); return ret; }
Since put_clkname is always done in remove(), I don't think there is any memleak here. Over that with your patch we will do put_clkname twice now, once in probe and once in remove. And that is a bug AFAICT.