linux-kernel - Re: [PATCH net] fjes: Fix an error handling path in fjes

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <437145bf-d925-e91e-affd-835d272c55a0@wanadoo.fr>
Date:   Thu, 29 Dec 2022 17:34:22 +0100
From:   Christophe JAILLET <christophe.jaillet@...adoo.fr>
To:     Michal Swiatkowski <michal.swiatkowski@...ux.intel.com>
Cc:     "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>,
        Taku Izumi <izumi.taku@...fujitsu.com>,
        linux-kernel@...r.kernel.org, kernel-janitors@...r.kernel.org,
        netdev@...r.kernel.org
Subject: Re: [PATCH net] fjes: Fix an error handling path in fjes_probe()

Le 21/12/2022 à 10:59, Michal Swiatkowski a écrit :
> On Tue, Dec 20, 2022 at 09:57:06PM +0100, Christophe JAILLET wrote:
>> A netif_napi_add() call is hidden in fjes_sw_init(). It should be undone
>> by a corresponding netif_napi_del() call in the error handling path of the
>> probe, as already done inthe remove function.
>>
>> Fixes: 265859309a76 ("fjes: NAPI polling function")
>> Signed-off-by: Christophe JAILLET <christophe.jaillet@...adoo.fr>
>> ---
>>   drivers/net/fjes/fjes_main.c | 4 +++-
>>   1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/net/fjes/fjes_main.c b/drivers/net/fjes/fjes_main.c
>> index 2513be6d4e11..01b4c9c6adbd 100644
>> --- a/drivers/net/fjes/fjes_main.c
>> +++ b/drivers/net/fjes/fjes_main.c
>> @@ -1370,7 +1370,7 @@ static int fjes_probe(struct platform_device *plat_dev)
>>   	adapter->txrx_wq = alloc_workqueue(DRV_NAME "/txrx", WQ_MEM_RECLAIM, 0);
>>   	if (unlikely(!adapter->txrx_wq)) {
>>   		err = -ENOMEM;
>> -		goto err_free_netdev;
>> +		goto err_del_napi;
>>   	}
>>   
>>   	adapter->control_wq = alloc_workqueue(DRV_NAME "/control",
>> @@ -1431,6 +1431,8 @@ static int fjes_probe(struct platform_device *plat_dev)
>>   	destroy_workqueue(adapter->control_wq);
>>   err_free_txrx_wq:
>>   	destroy_workqueue(adapter->txrx_wq);
>> +err_del_napi:
>> +	netif_napi_del(&adapter->napi);
>>   err_free_netdev:
>>   	free_netdev(netdev);
>>   err_out:
> 
> Reviewed-by: Michal Swiatkowski <michal.swiatkowski@...ux.intel.com>
> 
> I wonder if it won't be better to have fjes_sw_deinit() instead or
> change fjes_sw_init to only netif_napi_add(). You know, to avoid another
> bug here when someone add sth to the fjes_sw_deinit(). This is only
> suggestion, patch looks fine.

hi,

based on Jakub's comment [1], free_netdev() already cleans up NAPIs (see 
[2]).

So would it make more sense to remove netif_napi_del() from the 
.remove() function instead?
The call looks useless to me now.

CJ

[1]: https://lore.kernel.org/all/20221221174043.1191996a@kernel.org/
[2]: https://elixir.bootlin.com/linux/v6.2-rc1/source/net/core/dev.c#L10710

> 
>> -- 
>> 2.34.1
>>
>