This is unsafe, as the runtime PM callbacks are called from the PM
workqueue, so this may deadlock when handling an i2c attached clock,
which may already hold the clk_prepare mutex from another context.
Signed-off-by: Alexander Stein <alexander.stein@xxxxxxxxxxxxxxx>
---
This is the imx-lpi2c equivalent to commit d9a22d713acb ("i2c: imx: avoid
taking clk_prepare mutex in PM callbacks"). This patch needs to be used
together with [1] in order to fix the deadlock conditions.
Commit message is shamlessly taken from commit d9a22d713acb.
Best regards,
Alexander
[1] https://patchwork.ozlabs.org/project/linux-i2c/patch/20230310130815.562418-1-alexander.sverdlin@xxxxxxxxxxx/
drivers/i2c/busses/i2c-imx-lpi2c.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/i2c/busses/i2c-imx-lpi2c.c b/drivers/i2c/busses/i2c-imx-lpi2c.c
index a49b14d52a98..1af0a637d7f1 100644
--- a/drivers/i2c/busses/i2c-imx-lpi2c.c
+++ b/drivers/i2c/busses/i2c-imx-lpi2c.c
@@ -639,7 +639,7 @@ static int __maybe_unused lpi2c_runtime_suspend(struct device *dev)
{
struct lpi2c_imx_struct *lpi2c_imx = dev_get_drvdata(dev);
- clk_bulk_disable_unprepare(lpi2c_imx->num_clks, lpi2c_imx->clks);
+ clk_bulk_disable(lpi2c_imx->num_clks, lpi2c_imx->clks);
pinctrl_pm_select_sleep_state(dev);
return 0;
@@ -651,7 +651,7 @@ static int __maybe_unused lpi2c_runtime_resume(struct device *dev)
int ret;
pinctrl_pm_select_default_state(dev);
- ret = clk_bulk_prepare_enable(lpi2c_imx->num_clks, lpi2c_imx->clks);
+ ret = clk_bulk_enable(lpi2c_imx->num_clks, lpi2c_imx->clks);
if (ret) {
dev_err(dev, "failed to enable I2C clock, ret=%d\n", ret);
return ret;
--
2.34.1
