of_i2c_register_device already contains error message, so remove
the redundant dev_err message
Signed-off-by: tangbin <[email protected]>
---
drivers/i2c/i2c-core-of.c | 6 +-----
1 file changed, 1 insertion(+), 5 deletions(-)
diff --git a/drivers/i2c/i2c-core-of.c b/drivers/i2c/i2c-core-of.c
index 6787c1f71..7b0a786d3 100644
--- a/drivers/i2c/i2c-core-of.c
+++ b/drivers/i2c/i2c-core-of.c
@@ -103,9 +103,7 @@ void of_i2c_register_devices(struct i2c_adapter *adap)
client = of_i2c_register_device(adap, node);
if (IS_ERR(client)) {
- dev_err(&adap->dev,
- "Failed to create I2C device for %pOF\n",
- node);
+ return PTR_ERR(client);
of_node_clear_flag(node, OF_POPULATED);
}
}
@@ -246,8 +244,6 @@ static int of_i2c_notify(struct notifier_block *nb, unsigned long action,
client = of_i2c_register_device(adap, rd->dn);
if (IS_ERR(client)) {
- dev_err(&adap->dev, "failed to create client for '%pOF'\n",
- rd->dn);
put_device(&adap->dev);
of_node_clear_flag(rd->dn, OF_POPULATED);
return notifier_from_errno(PTR_ERR(client));
--
2.20.1.windows.1
On 2/26/20 11:39 AM, tangbin wrote:
> of_i2c_register_device already contains error message, so remove
> the redundant dev_err message
>
> Signed-off-by: tangbin <[email protected]>
> ---
> drivers/i2c/i2c-core-of.c | 6 +-----
> 1 file changed, 1 insertion(+), 5 deletions(-)
>
> diff --git a/drivers/i2c/i2c-core-of.c b/drivers/i2c/i2c-core-of.c
> index 6787c1f71..7b0a786d3 100644
> --- a/drivers/i2c/i2c-core-of.c
> +++ b/drivers/i2c/i2c-core-of.c
> @@ -103,9 +103,7 @@ void of_i2c_register_devices(struct i2c_adapter *adap)
>
> client = of_i2c_register_device(adap, node);
> if (IS_ERR(client)) {
> - dev_err(&adap->dev,
> - "Failed to create I2C device for %pOF\n",
> - node);
> + return PTR_ERR(client);
This looks like an unrelated (and wrong) change. Why would you alter the
semantics of of_i2c_register_devices()? Besides, this function doesn't
have a return value.
On Wed, Feb 26, 2020 at 11:58:41AM +0100, Francesco Lavra wrote:
> On 2/26/20 11:39 AM, tangbin wrote:
> > of_i2c_register_device already contains error message, so remove
> > the redundant dev_err message
> >
> > Signed-off-by: tangbin <[email protected]>
> > ---
> > drivers/i2c/i2c-core-of.c | 6 +-----
> > 1 file changed, 1 insertion(+), 5 deletions(-)
> >
> > diff --git a/drivers/i2c/i2c-core-of.c b/drivers/i2c/i2c-core-of.c
> > index 6787c1f71..7b0a786d3 100644
> > --- a/drivers/i2c/i2c-core-of.c
> > +++ b/drivers/i2c/i2c-core-of.c
> > @@ -103,9 +103,7 @@ void of_i2c_register_devices(struct i2c_adapter *adap)
> > client = of_i2c_register_device(adap, node);
> > if (IS_ERR(client)) {
> > - dev_err(&adap->dev,
> > - "Failed to create I2C device for %pOF\n",
> > - node);
> > + return PTR_ERR(client);
>
> This looks like an unrelated (and wrong) change. Why would you alter the
> semantics of of_i2c_register_devices()? Besides, this function doesn't have
> a return value.
Right. This is not correct.
In general, tangbin has a point, the error reporting is doubled. Lower
layers already report, so both(!) callers of of_i2c_register_device do
not need to. Since I am refactoring all this anyhow in "[RFC PATCH 5/7]
i2c: of: error message unification", I think I will just drop error
reporting in the callers there when resending the series (giving tanbin
credits for the removal).
Is this okay with everyone?