2019-10-28 17:36:25

by Gustavo A. R. Silva

[permalink] [raw]
Subject: [PATCH net-next] net: aquantia: fix error handling in aq_nic_init

Fix currenty ignored returned error by properly error checking
aq_phy_init().

Addresses-Coverity-ID: 1487376 ("Unused value")
Fixes: dbcd6806af42 ("net: aquantia: add support for Phy access")
Signed-off-by: Gustavo A. R. Silva <[email protected]>
---
drivers/net/ethernet/aquantia/atlantic/aq_nic.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
index 433adc099e44..1914aa0a19d0 100644
--- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
+++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
@@ -341,7 +341,8 @@ int aq_nic_init(struct aq_nic_s *self)

if (self->aq_nic_cfg.aq_hw_caps->media_type == AQ_HW_MEDIA_TYPE_TP) {
self->aq_hw->phy_id = HW_ATL_PHY_ID_MAX;
- err = aq_phy_init(self->aq_hw);
+ if (!aq_phy_init(self->aq_hw))
+ goto err_exit;
}

for (i = 0U, aq_vec = self->aq_vec[0];
--
2.23.0


2019-10-28 19:53:40

by Igor Russkikh

[permalink] [raw]
Subject: Re: [EXT] [PATCH net-next] net: aquantia: fix error handling in aq_nic_init


> Fix currenty ignored returned error by properly error checking
> aq_phy_init().
>
> Addresses-Coverity-ID: 1487376 ("Unused value")
> Fixes: dbcd6806af42 ("net: aquantia: add support for Phy access")
> Signed-off-by: Gustavo A. R. Silva <[email protected]>
> ---
> drivers/net/ethernet/aquantia/atlantic/aq_nic.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
> index 433adc099e44..1914aa0a19d0 100644
> --- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
> +++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
> @@ -341,7 +341,8 @@ int aq_nic_init(struct aq_nic_s *self)
>
> if (self->aq_nic_cfg.aq_hw_caps->media_type == AQ_HW_MEDIA_TYPE_TP) {
> self->aq_hw->phy_id = HW_ATL_PHY_ID_MAX;
> - err = aq_phy_init(self->aq_hw);
> + if (!aq_phy_init(self->aq_hw))
> + goto err_exit;
> }
>
> for (i = 0U, aq_vec = self->aq_vec[0];
>

Hi Gustavo,

I'd say the intention here was to ignore the error, as driver may still live if
something unexpected happened on Phy access path.

Notice in the above fix you leave `err` as zero but do error path return -
that'll break the datapath.

I'd prefer to fix this with simple

(void)aq_phy_init(self->aq_hw);

--
Regards,
Igor

2019-10-29 01:36:34

by Gustavo A. R. Silva

[permalink] [raw]
Subject: Re: [EXT] [PATCH net-next] net: aquantia: fix error handling in aq_nic_init



On 10/28/19 05:53, Igor Russkikh wrote:
>
>> Fix currenty ignored returned error by properly error checking
>> aq_phy_init().
>>
>> Addresses-Coverity-ID: 1487376 ("Unused value")
>> Fixes: dbcd6806af42 ("net: aquantia: add support for Phy access")
>> Signed-off-by: Gustavo A. R. Silva <[email protected]>
>> ---
>> drivers/net/ethernet/aquantia/atlantic/aq_nic.c | 3 ++-
>> 1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
>> index 433adc099e44..1914aa0a19d0 100644
>> --- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
>> +++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c
>> @@ -341,7 +341,8 @@ int aq_nic_init(struct aq_nic_s *self)
>>
>> if (self->aq_nic_cfg.aq_hw_caps->media_type == AQ_HW_MEDIA_TYPE_TP) {
>> self->aq_hw->phy_id = HW_ATL_PHY_ID_MAX;
>> - err = aq_phy_init(self->aq_hw);
>> + if (!aq_phy_init(self->aq_hw))
>> + goto err_exit;
>> }
>>
>> for (i = 0U, aq_vec = self->aq_vec[0];
>>
>
> Hi Gustavo,
>

Hi!

> I'd say the intention here was to ignore the error, as driver may still live if
> something unexpected happened on Phy access path.
>

I see. Please, see my comments below...

> Notice in the above fix you leave `err` as zero but do error path return -
> that'll break the datapath.
>

You're right. I totally missed that.

> I'd prefer to fix this with simple
>
> (void)aq_phy_init(self->aq_hw);
>

Yep. This is much better. Otherwise, both static analyzers and developers
get confused. :p

Something even better is to add both a cast and a comment explaining why
the return value is not being checked in this case (as you do above).

Thanks
--
Gustavo