If an error occurs after calling mux_state_select(), mux_state_deselect()
should be called as already done in the remove function.

Fixes: b6ef830c60b6 ("i2c: omap: Add support for setting mux")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
---
I'm not 100% sure of the error handling path.

Should pm_runtime_dont_use_autosuspend() be called after the err_disable_pm
label? (to match the calling order)

Also, should errors from omap_i2c_init() be handled?
---
 drivers/i2c/busses/i2c-omap.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index 876791d20ed5..5e46dc2cbbd7 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -1461,13 +1461,13 @@ omap_i2c_probe(struct platform_device *pdev)
 		if (IS_ERR(mux_state)) {
 			r = PTR_ERR(mux_state);
 			dev_dbg(&pdev->dev, "failed to get I2C mux: %d\n", r);
-			goto err_disable_pm;
+			goto err_put_pm;
 		}
 		omap->mux_state = mux_state;
 		r = mux_state_select(omap->mux_state);
 		if (r) {
 			dev_err(&pdev->dev, "failed to select I2C mux: %d\n", r);
-			goto err_disable_pm;
+			goto err_put_pm;
 		}
 	}
 
@@ -1515,6 +1515,9 @@ omap_i2c_probe(struct platform_device *pdev)
 
 err_unuse_clocks:
 	omap_i2c_write_reg(omap, OMAP_I2C_CON_REG, 0);
+	if (omap->mux_state)
+		mux_state_deselect(omap->mux_state);
+err_put_pm:
 	pm_runtime_dont_use_autosuspend(omap->dev);
 	pm_runtime_put_sync(omap->dev);
 err_disable_pm:
-- 
2.49.0

Hi Christophe,

On Sat, Jun 14, 2025 at 04:59:26PM +0200, Christophe JAILLET wrote:
> If an error occurs after calling mux_state_select(), mux_state_deselect()
> should be called as already done in the remove function.
> 
> Fixes: b6ef830c60b6 ("i2c: omap: Add support for setting mux")
> Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>

merged to i2c/i2c-host-fixes. Thanks!

> ---
> I'm not 100% sure of the error handling path.
> 
> Should pm_runtime_dont_use_autosuspend() be called after the err_disable_pm
> label? (to match the calling order)

Yes, I think you are right here.

> Also, should errors from omap_i2c_init() be handled?

Yes, if it fails it should be handled.

Thanks for the two reports. Do you have time to fix them or
can I go ahead and do it?

Thanks,
Andi

Le 26/06/2025 à 00:10, Andi Shyti a écrit :
> Hi Christophe,
> 
> On Sat, Jun 14, 2025 at 04:59:26PM +0200, Christophe JAILLET wrote:
>> If an error occurs after calling mux_state_select(), mux_state_deselect()
>> should be called as already done in the remove function.
>>
>> Fixes: b6ef830c60b6 ("i2c: omap: Add support for setting mux")
>> Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
> 
> merged to i2c/i2c-host-fixes. Thanks!
> 
>> ---
>> I'm not 100% sure of the error handling path.
>>
>> Should pm_runtime_dont_use_autosuspend() be called after the err_disable_pm
>> label? (to match the calling order)
> 
> Yes, I think you are right here.
> 
>> Also, should errors from omap_i2c_init() be handled?
> 
> Yes, if it fails it should be handled.
> 
> Thanks for the two reports. Do you have time to fix them or
> can I go ahead and do it?

I've sent a patch for each issue.

Feel free to take the lead if not correct. As said in the cover letter, 
I'm not completely confident with them.

CJ

> 
> Thanks,
> Andi
> 
>