[PATCH v4 0/2] kobject: trival fix

[PATCH v4 0/2] kobject: trival fix

[Bo YU]

This is trival fix for lib/kobject_uevent.c file.

Bo YU (2):
  kobject: use pr_warn to replace printk
  kobject: drop newline from msg string

 lib/kobject_uevent.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

--
2.11.0

[PATCH v4 1/2] kobject: use pr_warn to replace printk

[Bo YU]

This is a fix to replace printk with pr_warn

Signed-off-by: Bo YU <tsu.yubo@gmail.com>
---
 lib/kobject_uevent.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c
index 27c6118afd1c..499dfcdafcb0 100644
--- a/lib/kobject_uevent.c
+++ b/lib/kobject_uevent.c
@@ -224,7 +224,7 @@ int kobject_synth_uevent(struct kobject *kobj, const char *buf, size_t count)
 out:
 	if (r) {
 		devpath = kobject_get_path(kobj, GFP_KERNEL);
-		printk(KERN_WARNING "synth uevent: %s: %s",
+		pr_warn("synth uevent: %s: %s",
 		       devpath ?: "unknown device",
 		       msg ?: "failed to send uevent");
 		kfree(devpath);
-- 
2.11.0

[PATCH v4 2/2] kobject: drop newline from msg string

[Bo YU]

There is currently a missing terminating newline in non-switch case
match,when msg == NULL

Signed-off-by: Bo YU <tsu.yubo@gmail.com>
---
Changes in V4:
Recovery  originly declaration, requested by Greg.

Changes in V3:
Improve the commit log, requested by rafael.

Changes in V2:
According to Joe's suggestion, drop newline from msg string
---
 lib/kobject_uevent.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c
index 499dfcdafcb0..118226a02707 100644
--- a/lib/kobject_uevent.c
+++ b/lib/kobject_uevent.c
@@ -200,7 +200,7 @@ int kobject_synth_uevent(struct kobject *kobj, const char *buf, size_t count)
 
 	r = kobject_action_type(buf, count, &action, &action_args);
 	if (r) {
-		msg = "unknown uevent action string\n";
+		msg = "unknown uevent action string";
 		goto out;
 	}
 
@@ -212,7 +212,7 @@ int kobject_synth_uevent(struct kobject *kobj, const char *buf, size_t count)
 	r = kobject_action_args(action_args,
 				count - (action_args - buf), &env);
 	if (r == -EINVAL) {
-		msg = "incorrect uevent action arguments\n";
+		msg = "incorrect uevent action arguments";
 		goto out;
 	}
 
@@ -224,7 +224,7 @@ int kobject_synth_uevent(struct kobject *kobj, const char *buf, size_t count)
 out:
 	if (r) {
 		devpath = kobject_get_path(kobj, GFP_KERNEL);
-		pr_warn("synth uevent: %s: %s",
+		pr_warn("synth uevent: %s: %s\n",
 		       devpath ?: "unknown device",
 		       msg ?: "failed to send uevent");
 		kfree(devpath);
-- 
2.11.0

Re: [PATCH v4 1/2] kobject: use pr_warn to replace printk

[Rafael J. Wysocki]

On Tue, Jan 8, 2019 at 4:11 AM Bo YU <tsu.yubo@gmail.com> wrote:
>
> This is a fix to replace printk with pr_warn

You call it a "fix", but what exactly is wrong with the code as is?

> Signed-off-by: Bo YU <tsu.yubo@gmail.com>
> ---
>  lib/kobject_uevent.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c
> index 27c6118afd1c..499dfcdafcb0 100644
> --- a/lib/kobject_uevent.c
> +++ b/lib/kobject_uevent.c
> @@ -224,7 +224,7 @@ int kobject_synth_uevent(struct kobject *kobj, const char *buf, size_t count)
>  out:
>         if (r) {
>                 devpath = kobject_get_path(kobj, GFP_KERNEL);
> -               printk(KERN_WARNING "synth uevent: %s: %s",
> +               pr_warn("synth uevent: %s: %s",
>                        devpath ?: "unknown device",
>                        msg ?: "failed to send uevent");
>                 kfree(devpath);
> --
> 2.11.0
>

Re: [PATCH v4 2/2] kobject: drop newline from msg string

[Rafael J. Wysocki]

On Tue, Jan 8, 2019 at 4:11 AM Bo YU <tsu.yubo@gmail.com> wrote:
>
> There is currently a missing terminating newline in non-switch case
> match,when msg == NULL
>
> Signed-off-by: Bo YU <tsu.yubo@gmail.com>

Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

> ---
> Changes in V4:
> Recovery  originly declaration, requested by Greg.
>
> Changes in V3:
> Improve the commit log, requested by rafael.
>
> Changes in V2:
> According to Joe's suggestion, drop newline from msg string
> ---
>  lib/kobject_uevent.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c
> index 499dfcdafcb0..118226a02707 100644
> --- a/lib/kobject_uevent.c
> +++ b/lib/kobject_uevent.c
> @@ -200,7 +200,7 @@ int kobject_synth_uevent(struct kobject *kobj, const char *buf, size_t count)
>
>         r = kobject_action_type(buf, count, &action, &action_args);
>         if (r) {
> -               msg = "unknown uevent action string\n";
> +               msg = "unknown uevent action string";
>                 goto out;
>         }
>
> @@ -212,7 +212,7 @@ int kobject_synth_uevent(struct kobject *kobj, const char *buf, size_t count)
>         r = kobject_action_args(action_args,
>                                 count - (action_args - buf), &env);
>         if (r == -EINVAL) {
> -               msg = "incorrect uevent action arguments\n";
> +               msg = "incorrect uevent action arguments";
>                 goto out;
>         }
>
> @@ -224,7 +224,7 @@ int kobject_synth_uevent(struct kobject *kobj, const char *buf, size_t count)
>  out:
>         if (r) {
>                 devpath = kobject_get_path(kobj, GFP_KERNEL);
> -               pr_warn("synth uevent: %s: %s",
> +               pr_warn("synth uevent: %s: %s\n",
>                        devpath ?: "unknown device",
>                        msg ?: "failed to send uevent");
>                 kfree(devpath);
> --
> 2.11.0
>

Re: [PATCH v4 1/2] kobject: use pr_warn to replace printk

[YU Bo]

Hi,
On Tue, Jan 08, 2019 at 10:39:57AM +0100, Rafael J. Wysocki wrote:
>On Tue, Jan 8, 2019 at 4:11 AM Bo YU <tsu.yubo@gmail.com> wrote:
>>
>> This is a fix to replace printk with pr_warn
>
>You call it a "fix", but what exactly is wrong with the code as is?
Is there any differentions between pr_* and printk?
If they were the same with functionial, i think it  would be better
to use pr_* to replace printk.There are two lines code with printk only
and others are pr_* in kobject_uevent.c.
In fact,as you know, the code is fine.I just want to keep consistency with
looks like
Thanks
>
>> Signed-off-by: Bo YU <tsu.yubo@gmail.com>
>> ---
>>  lib/kobject_uevent.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c
>> index 27c6118afd1c..499dfcdafcb0 100644
>> --- a/lib/kobject_uevent.c
>> +++ b/lib/kobject_uevent.c
>> @@ -224,7 +224,7 @@ int kobject_synth_uevent(struct kobject *kobj, const char *buf, size_t count)
>>  out:
>>         if (r) {
>>                 devpath = kobject_get_path(kobj, GFP_KERNEL);
>> -               printk(KERN_WARNING "synth uevent: %s: %s",
>> +               pr_warn("synth uevent: %s: %s",
>>                        devpath ?: "unknown device",
>>                        msg ?: "failed to send uevent");
>>                 kfree(devpath);
>> --
>> 2.11.0
>>

Re: [PATCH v4 1/2] kobject: use pr_warn to replace printk

[Rafael J. Wysocki]

On Tue, Jan 8, 2019 at 1:54 PM YU Bo <tsu.yubo@gmail.com> wrote:
>
> Hi,
> On Tue, Jan 08, 2019 at 10:39:57AM +0100, Rafael J. Wysocki wrote:
> >On Tue, Jan 8, 2019 at 4:11 AM Bo YU <tsu.yubo@gmail.com> wrote:
> >>
> >> This is a fix to replace printk with pr_warn
> >
> >You call it a "fix", but what exactly is wrong with the code as is?
> Is there any differentions between pr_* and printk?
> If they were the same with functionial, i think it  would be better
> to use pr_* to replace printk.There are two lines code with printk only
> and others are pr_* in kobject_uevent.c.
> In fact,as you know, the code is fine.I just want to keep consistency with
> looks like

OK, so please don't call it a "fix" as doing so suggests a functional
defect in the current code that is addressed by your patch which is
not the case.

You can write something like

"Replace printk() with pr_warn() in kobject_synth_uevent() to make it
consistent with the other code in kobject_uevent.c."

in the changelog which would be clear enough IMO.