Virtualization - Nov 2022 - [PATCH] vdpa_sim: fix possible memory leak in vdpasim_net_init() and vdpasim_blk_init()

On Fri, Nov 4, 2022 at 4:41 PM ruanjinjie <ruanjinjie at huawei.com>
wrote:
>
> If device_register() returns error in vdpasim_net_init() or
> vdpasim_blk_init(), name of kobject which is allocated in dev_set_name()
> called in device_add() is leaked.
>
> As comment of device_add() says, it should call put_device() to drop
> the reference count that was set in device_initialize() when it fails,
> so the name can be freed in kobject_cleanup().
>
> Signed-off-by: ruanjinjie <ruanjinjie at huawei.com>
Acked-by: Jason Wang <jasowang at redhat.com>

Should we have a fixes tag and cc stable?

Thanks
> ---
>  drivers/vdpa/vdpa_sim/vdpa_sim_blk.c | 4 +++-
>  drivers/vdpa/vdpa_sim/vdpa_sim_net.c | 4 +++-
>  2 files changed, 6 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
> index c6db1a1baf76..f745926237a8 100644
> --- a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
> +++ b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
> @@ -427,8 +427,10 @@ static int __init vdpasim_blk_init(void)
>         int ret;
>
>         ret = device_register(&vdpasim_blk_mgmtdev);
> -       if (ret)
> +       if (ret) {
> +               put_device(&vdpasim_blk_mgmtdev);
>                 return ret;
> +       }
>
>         ret = vdpa_mgmtdev_register(&mgmt_dev);
>         if (ret)
> diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c
b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c
> index c3cb225ea469..11f5a121df24 100644
> --- a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c
> +++ b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c
> @@ -305,8 +305,10 @@ static int __init vdpasim_net_init(void)
>         int ret;
>
>         ret = device_register(&vdpasim_net_mgmtdev);
> -       if (ret)
> +       if (ret) {
> +               put_device(&vdpasim_net_mgmtdev);
>                 return ret;
> +       }
>
>         ret = vdpa_mgmtdev_register(&mgmt_dev);
>         if (ret)
> --
> 2.25.1
>

On Mon, Nov 07, 2022 at 03:41:42PM +0800, Jason Wang
wrote:
>On Fri, Nov 4, 2022 at 4:41 PM ruanjinjie <ruanjinjie at huawei.com>
wrote:
>>
>> If device_register() returns error in vdpasim_net_init() or
>> vdpasim_blk_init(), name of kobject which is allocated in
dev_set_name()
>> called in device_add() is leaked.
>>
>> As comment of device_add() says, it should call put_device() to drop
>> the reference count that was set in device_initialize() when it fails,
>> so the name can be freed in kobject_cleanup().
>>
>> Signed-off-by: ruanjinjie <ruanjinjie at huawei.com>
>
>Acked-by: Jason Wang <jasowang at redhat.com>
>
>Should we have a fixes tag and cc stable?
Yep, I think so.

The fixes tags should be:

Fixes: 899c4d187f6a ("vdpa_sim_blk: add support for vdpa management
tool")
Fixes: a3c06ae158dd ("vdpa_sim_net: Add support for user supported
devices")

With them:

Reviewed-by: Stefano Garzarella <sgarzare at redhat.com>

Thanks,
Stefano