2019-06-03 17:55:10

by Colin King

Subject: [PATCH][next] bpf: remove redundant assignment to err

From: Colin Ian King <[email protected]>

The variable err is assigned with the value -EINVAL that is never
read and it is re-assigned a new value later on. The assignment is
redundant and can be removed.

Addresses-Coverity: ("Unused value")
Signed-off-by: Colin Ian King <[email protected]>
---
kernel/bpf/devmap.c | 2 +-
kernel/bpf/xskmap.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
index 5ae7cce5ef16..a76cc6412fc4 100644
--- a/kernel/bpf/devmap.c
+++ b/kernel/bpf/devmap.c
@@ -88,7 +88,7 @@ static u64 dev_map_bitmap_size(const union bpf_attr *attr)
static struct bpf_map *dev_map_alloc(union bpf_attr *attr)
{
struct bpf_dtab *dtab;
- int err = -EINVAL;
+ int err;
u64 cost;

if (!capable(CAP_NET_ADMIN))
diff --git a/kernel/bpf/xskmap.c b/kernel/bpf/xskmap.c
index 22066c28ba61..26859c6c9491 100644
--- a/kernel/bpf/xskmap.c
+++ b/kernel/bpf/xskmap.c
@@ -17,7 +17,7 @@ struct xsk_map {

static struct bpf_map *xsk_map_alloc(union bpf_attr *attr)
{
- int cpu, err = -EINVAL;
+ int cpu, err;
struct xsk_map *m;
u64 cost;

--
2.20.1

2019-06-03 17:57:46

by Jakub Kicinski

Subject: Re: [PATCH][next] bpf: remove redundant assignment to err

On Mon, 3 Jun 2019 18:39:16 +0100, Colin Ian King wrote:
> On 03/06/2019 18:21, Jakub Kicinski wrote:
> > On Mon, 3 Jun 2019 18:02:47 +0100, Colin King wrote:
> >> From: Colin Ian King <[email protected]>
> >>
> >> The variable err is assigned with the value -EINVAL that is never
> >> read and it is re-assigned a new value later on. The assignment is
> >> redundant and can be removed.
> >>
> >> Addresses-Coverity: ("Unused value")
> >> Signed-off-by: Colin Ian King <[email protected]>
> >> ---
> >> kernel/bpf/devmap.c | 2 +-
> >> kernel/bpf/xskmap.c | 2 +-
> >> 2 files changed, 2 insertions(+), 2 deletions(-)
> >>
> >> diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
> >> index 5ae7cce5ef16..a76cc6412fc4 100644
> >> --- a/kernel/bpf/devmap.c
> >> +++ b/kernel/bpf/devmap.c
> >> @@ -88,7 +88,7 @@ static u64 dev_map_bitmap_size(const union bpf_attr *attr)
> >> static struct bpf_map *dev_map_alloc(union bpf_attr *attr)
> >> {
> >> struct bpf_dtab *dtab;
> >> - int err = -EINVAL;
> >> + int err;
> >> u64 cost;
> >
> > Perhaps keep the variables ordered longest to shortest?
>
> Is that a required coding standard?

For networking code, yes. Just look around the files you're changing
and see for yourself.

> >> if (!capable(CAP_NET_ADMIN))
> >> diff --git a/kernel/bpf/xskmap.c b/kernel/bpf/xskmap.c
> >> index 22066c28ba61..26859c6c9491 100644
> >> --- a/kernel/bpf/xskmap.c
> >> +++ b/kernel/bpf/xskmap.c
> >> @@ -17,7 +17,7 @@ struct xsk_map {
> >>
> >> static struct bpf_map *xsk_map_alloc(union bpf_attr *attr)
> >> {
> >> - int cpu, err = -EINVAL;
> >> + int cpu, err;
> >> struct xsk_map *m;
> >> u64 cost;
> >
> > And here.

2019-06-03 18:08:46

by Colin King

Subject: Re: [PATCH][next] bpf: remove redundant assignment to err

On 03/06/2019 18:49, Jakub Kicinski wrote:
> On Mon, 3 Jun 2019 18:39:16 +0100, Colin Ian King wrote:
>> On 03/06/2019 18:21, Jakub Kicinski wrote:
>>> On Mon, 3 Jun 2019 18:02:47 +0100, Colin King wrote:
>>>> From: Colin Ian King <[email protected]>
>>>>
>>>> The variable err is assigned with the value -EINVAL that is never
>>>> read and it is re-assigned a new value later on. The assignment is
>>>> redundant and can be removed.
>>>>
>>>> Addresses-Coverity: ("Unused value")
>>>> Signed-off-by: Colin Ian King <[email protected]>
>>>> ---
>>>> kernel/bpf/devmap.c | 2 +-
>>>> kernel/bpf/xskmap.c | 2 +-
>>>> 2 files changed, 2 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
>>>> index 5ae7cce5ef16..a76cc6412fc4 100644
>>>> --- a/kernel/bpf/devmap.c
>>>> +++ b/kernel/bpf/devmap.c
>>>> @@ -88,7 +88,7 @@ static u64 dev_map_bitmap_size(const union bpf_attr *attr)
>>>> static struct bpf_map *dev_map_alloc(union bpf_attr *attr)
>>>> {
>>>> struct bpf_dtab *dtab;
>>>> - int err = -EINVAL;
>>>> + int err;
>>>> u64 cost;
>>>
>>> Perhaps keep the variables ordered longest to shortest?
>>
>> Is that a required coding standard?
>
> For networking code, yes. Just look around the files you're changing
> and see for yourself.

Ah, informal coding standards. Great. Won't this end up with more diff
churn?

>
>>>> if (!capable(CAP_NET_ADMIN))
>>>> diff --git a/kernel/bpf/xskmap.c b/kernel/bpf/xskmap.c
>>>> index 22066c28ba61..26859c6c9491 100644
>>>> --- a/kernel/bpf/xskmap.c
>>>> +++ b/kernel/bpf/xskmap.c
>>>> @@ -17,7 +17,7 @@ struct xsk_map {
>>>>
>>>> static struct bpf_map *xsk_map_alloc(union bpf_attr *attr)
>>>> {
>>>> - int cpu, err = -EINVAL;
>>>> + int cpu, err;
>>>> struct xsk_map *m;
>>>> u64 cost;
>>>
>>> And here.

2019-06-06 16:28:09

by Dan Carpenter

Subject: Re: [PATCH][next] bpf: remove redundant assignment to err

On Mon, Jun 03, 2019 at 07:07:20PM +0100, Colin Ian King wrote:
> On 03/06/2019 18:49, Jakub Kicinski wrote:
> > On Mon, 3 Jun 2019 18:39:16 +0100, Colin Ian King wrote:
> >> On 03/06/2019 18:21, Jakub Kicinski wrote:
> >>> On Mon, 3 Jun 2019 18:02:47 +0100, Colin King wrote:
> >>>> From: Colin Ian King <[email protected]>
> >>>>
> >>>> The variable err is assigned with the value -EINVAL that is never
> >>>> read and it is re-assigned a new value later on. The assignment is
> >>>> redundant and can be removed.
> >>>>
> >>>> Addresses-Coverity: ("Unused value")
> >>>> Signed-off-by: Colin Ian King <[email protected]>
> >>>> ---
> >>>> kernel/bpf/devmap.c | 2 +-
> >>>> kernel/bpf/xskmap.c | 2 +-
> >>>> 2 files changed, 2 insertions(+), 2 deletions(-)
> >>>>
> >>>> diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
> >>>> index 5ae7cce5ef16..a76cc6412fc4 100644
> >>>> --- a/kernel/bpf/devmap.c
> >>>> +++ b/kernel/bpf/devmap.c
> >>>> @@ -88,7 +88,7 @@ static u64 dev_map_bitmap_size(const union bpf_attr *attr)
> >>>> static struct bpf_map *dev_map_alloc(union bpf_attr *attr)
> >>>> {
> >>>> struct bpf_dtab *dtab;
> >>>> - int err = -EINVAL;
> >>>> + int err;
> >>>> u64 cost;
> >>>
> >>> Perhaps keep the variables ordered longest to shortest?
> >>
> >> Is that a required coding standard?
> >
> > For networking code, yes. Just look around the files you're changing
> > and see for yourself.
>
> Ah, informal coding standards. Great. Won't this end up with more diff
> churn?

Everyone knows that netdev uses reverse Christmas tree declarations...

regards,
dan carpenter