diff mbox series

net: mellanox: mlx5: fix error return code in mlx5_fpga_device_start()

Message ID 20210304141814.8508-1-baijiaju1990@gmail.com
State New
Headers show
Series net: mellanox: mlx5: fix error return code in mlx5_fpga_device_start() | expand

Commit Message

Jia-Ju Bai March 4, 2021, 2:18 p.m. UTC
When mlx5_is_fpga_lookaside() returns a non-zero value, no error 
return code is assigned.
To fix this bug, err is assigned with -EINVAL as error return code.

Reported-by: TOTE Robot <oslab@tsinghua.edu.cn>
Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

Comments

Leon Romanovsky March 7, 2021, 8:50 a.m. UTC | #1
On Thu, Mar 04, 2021 at 06:18:14AM -0800, Jia-Ju Bai wrote:
> When mlx5_is_fpga_lookaside() returns a non-zero value, no error

> return code is assigned.

> To fix this bug, err is assigned with -EINVAL as error return code.

>

> Reported-by: TOTE Robot <oslab@tsinghua.edu.cn>

> Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>

> ---

>  drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c | 4 +++-

>  1 file changed, 3 insertions(+), 1 deletion(-)


Like Heiner said, the current code has correct behavior.
The mlx5_fpga_device_load_check() has same mlx5_is_fpga_lookaside()
check and it is not an error if it returns true.

NAK: Leon Romanovsky <leonro@nvidia.com>

Thanks
Saeed Mahameed March 11, 2021, 10:51 p.m. UTC | #2
On Sun, 2021-03-07 at 10:50 +0200, Leon Romanovsky wrote:
> On Thu, Mar 04, 2021 at 06:18:14AM -0800, Jia-Ju Bai wrote:

> > When mlx5_is_fpga_lookaside() returns a non-zero value, no error

> > return code is assigned.

> > To fix this bug, err is assigned with -EINVAL as error return code.

> > 

> > Reported-by: TOTE Robot <oslab@tsinghua.edu.cn>

> > Signed-off-by: Jia-Ju Bai <baijiaju1990@gmail.com>

> > ---

> >  drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c | 4 +++-

> >  1 file changed, 3 insertions(+), 1 deletion(-)

> 

> Like Heiner said, the current code has correct behavior.

> The mlx5_fpga_device_load_check() has same mlx5_is_fpga_lookaside()

> check and it is not an error if it returns true.

> 

> NAK: Leon Romanovsky <leonro@nvidia.com>

> 

> Thanks


Agreed, apparently this robot is looking for "goto {out|*err*}"
statements and treats all of them as errors, this is very unreliable,
diff mbox series

Patch

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c b/drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c
index 2ce4241459ce..c9e6da97126f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fpga/core.c
@@ -198,8 +198,10 @@  int mlx5_fpga_device_start(struct mlx5_core_dev *mdev)
 	mlx5_fpga_info(fdev, "FPGA card %s:%u\n", mlx5_fpga_name(fpga_id), fpga_id);
 
 	/* No QPs if FPGA does not participate in net processing */
-	if (mlx5_is_fpga_lookaside(fpga_id))
+	if (mlx5_is_fpga_lookaside(fpga_id)) {
+		err = -EINVAL;
 		goto out;
+	}
 
 	mlx5_fpga_info(fdev, "%s(%d): image, version %u; SBU %06x:%04x version %d\n",
 		       mlx5_fpga_image_name(fdev->last_oper_image),