[PATCH 2/4] uml: helper.c warning corrections

[PATCH 2/4] uml: helper.c warning corrections

[Vitaliy Ivanov]

>From 7296c5b9770e95cd6ad4e9e71d2d14c972abdfe1 Mon Sep 17 00:00:00 2001
From: Vitaliy Ivanov <vitalivanov@gmail.com>
Date: Tue, 5 Jul 2011 02:03:06 +0300
Subject: [PATCH 2/4] uml: helper.c warning corrections
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

arch/um/os-Linux/helper.c: In function ‘helper_child’:
arch/um/os-Linux/helper.c:38:7: warning: ignoring return value of ‘write’, declared with attribute warn_unused_result

Signed-off-by: Vitaliy Ivanov <vitalivanov@gmail.com>
---
 arch/um/os-Linux/helper.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/um/os-Linux/helper.c b/arch/um/os-Linux/helper.c
index b6b1096..feff22d 100644
--- a/arch/um/os-Linux/helper.c
+++ b/arch/um/os-Linux/helper.c
@@ -28,14 +28,14 @@ static int helper_child(void *arg)
 {
 	struct helper_data *data = arg;
 	char **argv = data->argv;
-	int err;
+	int err, ret;
 
 	if (data->pre_exec != NULL)
 		(*data->pre_exec)(data->pre_data);
 	err = execvp_noalloc(data->buf, argv[0], argv);
 
 	/* If the exec succeeds, we don't get here */
-	write(data->fd, &err, sizeof(err));
+	CATCH_EINTR(ret = write(data->fd, &err, sizeof(err)));
 
 	return 0;
 }
-- 
1.7.4.1

Re: [PATCH 2/4] uml: helper.c warning corrections

[Richard Weinberger]

Am Dienstag 05 Juli 2011, 01:15:47 schrieb Vitaliy Ivanov:
> From 7296c5b9770e95cd6ad4e9e71d2d14c972abdfe1 Mon Sep 17 00:00:00 2001
> From: Vitaliy Ivanov <vitalivanov@gmail.com>
> Date: Tue, 5 Jul 2011 02:03:06 +0300
> Subject: [PATCH 2/4] uml: helper.c warning corrections
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
> 
> arch/um/os-Linux/helper.c: In function ‘helper_child’:
> arch/um/os-Linux/helper.c:38:7: warning: ignoring return value of ‘write’,
> declared with attribute warn_unused_result

Same question as before...

> Signed-off-by: Vitaliy Ivanov <vitalivanov@gmail.com>
> ---
>  arch/um/os-Linux/helper.c |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/um/os-Linux/helper.c b/arch/um/os-Linux/helper.c
> index b6b1096..feff22d 100644
> --- a/arch/um/os-Linux/helper.c
> +++ b/arch/um/os-Linux/helper.c
> @@ -28,14 +28,14 @@ static int helper_child(void *arg)
>  {
>  	struct helper_data *data = arg;
>  	char **argv = data->argv;
> -	int err;
> +	int err, ret;
> 
>  	if (data->pre_exec != NULL)
>  		(*data->pre_exec)(data->pre_data);
>  	err = execvp_noalloc(data->buf, argv[0], argv);
> 
>  	/* If the exec succeeds, we don't get here */
> -	write(data->fd, &err, sizeof(err));
> +	CATCH_EINTR(ret = write(data->fd, &err, sizeof(err)));

Is there really a realistic chance that this write() can be interrupted?
helper_child() gets called via clone() in run_helper().

Thanks,
//richard

Re: [PATCH 2/4] uml: helper.c warning corrections

[Vitaliy Ivanov]

Richard,

On Tue, Jul 5, 2011 at 12:42 PM, Richard Weinberger <richard@nod.at> wrote:
> Am Dienstag 05 Juli 2011, 01:15:47 schrieb Vitaliy Ivanov:
>> From 7296c5b9770e95cd6ad4e9e71d2d14c972abdfe1 Mon Sep 17 00:00:00 2001
>> From: Vitaliy Ivanov <vitalivanov@gmail.com>
>> Date: Tue, 5 Jul 2011 02:03:06 +0300
>> Subject: [PATCH 2/4] uml: helper.c warning corrections
>> MIME-Version: 1.0
>> Content-Type: text/plain; charset=UTF-8
>> Content-Transfer-Encoding: 8bit
>>
>> arch/um/os-Linux/helper.c: In function ‘helper_child’:
>> arch/um/os-Linux/helper.c:38:7: warning: ignoring return value of ‘write’,
>> declared with attribute warn_unused_result
>
> Same question as before...

Same answer is here:)

>> Signed-off-by: Vitaliy Ivanov <vitalivanov@gmail.com>
>> ---
>>  arch/um/os-Linux/helper.c |    4 ++--
>>  1 files changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/arch/um/os-Linux/helper.c b/arch/um/os-Linux/helper.c
>> index b6b1096..feff22d 100644
>> --- a/arch/um/os-Linux/helper.c
>> +++ b/arch/um/os-Linux/helper.c
>> @@ -28,14 +28,14 @@ static int helper_child(void *arg)
>>  {
>>       struct helper_data *data = arg;
>>       char **argv = data->argv;
>> -     int err;
>> +     int err, ret;
>>
>>       if (data->pre_exec != NULL)
>>               (*data->pre_exec)(data->pre_data);
>>       err = execvp_noalloc(data->buf, argv[0], argv);
>>
>>       /* If the exec succeeds, we don't get here */
>> -     write(data->fd, &err, sizeof(err));
>> +     CATCH_EINTR(ret = write(data->fd, &err, sizeof(err)));
>
> Is there really a realistic chance that this write() can be interrupted?
> helper_child() gets called via clone() in run_helper().

It's always better not to think whether it's possible but simply kill
all the chances for this to happen.
In this case that's pretty trivial thing and also it solves issue with warning.

Vitaliy

Re: [PATCH 2/4] uml: helper.c warning corrections

[Richard Weinberger]

Am Dienstag 05 Juli 2011, 13:45:06 schrieb Vitaliy Ivanov:
> Richard,
> 
> On Tue, Jul 5, 2011 at 12:42 PM, Richard Weinberger <richard@nod.at> wrote:
> > Am Dienstag 05 Juli 2011, 01:15:47 schrieb Vitaliy Ivanov:
> >> From 7296c5b9770e95cd6ad4e9e71d2d14c972abdfe1 Mon Sep 17 00:00:00 2001
> >> From: Vitaliy Ivanov <vitalivanov@gmail.com>
> >> Date: Tue, 5 Jul 2011 02:03:06 +0300
> >> Subject: [PATCH 2/4] uml: helper.c warning corrections
> >> MIME-Version: 1.0
> >> Content-Type: text/plain; charset=UTF-8
> >> Content-Transfer-Encoding: 8bit
> >> 
> >> arch/um/os-Linux/helper.c: In function ‘helper_child’:
> >> arch/um/os-Linux/helper.c:38:7: warning: ignoring return value of
> >> ‘write’, declared with attribute warn_unused_result
> > 
> > Same question as before...
> 
> Same answer is here:)
> 
> >> Signed-off-by: Vitaliy Ivanov <vitalivanov@gmail.com>
> >> ---
> >>  arch/um/os-Linux/helper.c |    4 ++--
> >>  1 files changed, 2 insertions(+), 2 deletions(-)
> >> 
> >> diff --git a/arch/um/os-Linux/helper.c b/arch/um/os-Linux/helper.c
> >> index b6b1096..feff22d 100644
> >> --- a/arch/um/os-Linux/helper.c
> >> +++ b/arch/um/os-Linux/helper.c
> >> @@ -28,14 +28,14 @@ static int helper_child(void *arg)
> >>  {
> >>       struct helper_data *data = arg;
> >>       char **argv = data->argv;
> >> -     int err;
> >> +     int err, ret;
> >> 
> >>       if (data->pre_exec != NULL)
> >>               (*data->pre_exec)(data->pre_data);
> >>       err = execvp_noalloc(data->buf, argv[0], argv);
> >> 
> >>       /* If the exec succeeds, we don't get here */
> >> -     write(data->fd, &err, sizeof(err));
> >> +     CATCH_EINTR(ret = write(data->fd, &err, sizeof(err)));
> > 
> > Is there really a realistic chance that this write() can be interrupted?
> > helper_child() gets called via clone() in run_helper().
> 
> It's always better not to think whether it's possible but simply kill
> all the chances for this to happen.
> In this case that's pretty trivial thing and also it solves issue with
> warning.

Defensive programming is not a solution for everything.
CPU time and memory are not for free, especially in kernel land.

Anyway, when this patch removes a real warning, it's fine for me.
Applied.

Thanks,
//richard

Re: [PATCH 2/4] uml: helper.c warning corrections

[Vitaliy Ivanov]

On Tue, Jul 5, 2011 at 3:02 PM, Richard Weinberger <richard@nod.at> wrote:
> Am Dienstag 05 Juli 2011, 13:45:06 schrieb Vitaliy Ivanov:
>> It's always better not to think whether it's possible but simply kill
>> all the chances for this to happen.
>> In this case that's pretty trivial thing and also it solves issue with
>> warning.
>
> Defensive programming is not a solution for everything.
> CPU time and memory are not for free, especially in kernel land.
>
> Anyway, when this patch removes a real warning, it's fine for me.
> Applied.

Thanks Richard.