linux-kernel - Re: [PATCH v2 2/2] nfp: flower: Remove usage of the deprecated ida_simple

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: <0ad7ff2c-a5ad-1e5f-b186-0a43ce55057c@wanadoo.fr>
Date:   Thu, 9 Jun 2022 07:11:14 +0200
From:   Christophe JAILLET <christophe.jaillet@...adoo.fr>
To:     Simon Horman <simon.horman@...igine.com>,
        Jakub Kicinski <kuba@...nel.org>,
        "David S. Miller" <davem@...emloft.net>
Cc:     linux-kernel@...r.kernel.org, kernel-janitors@...r.kernel.org,
        oss-drivers@...igine.com, netdev@...r.kernel.org
Subject: Re: [PATCH v2 2/2] nfp: flower: Remove usage of the deprecated
 ida_simple_xxx API

Le 10/02/2022 à 23:35, Christophe JAILLET a écrit :
> Use ida_alloc_xxx()/ida_free() instead to
> ida_simple_get()/ida_simple_remove().
> The latter is deprecated and more verbose.
> 
> Signed-off-by: Christophe JAILLET <christophe.jaillet@...adoo.fr>
> ---
>   .../net/ethernet/netronome/nfp/flower/tunnel_conf.c    | 10 +++++-----
>   1 file changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c b/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c
> index 9244b35e3855..c71bd555f482 100644
> --- a/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c
> +++ b/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c
> @@ -942,8 +942,8 @@ nfp_tunnel_add_shared_mac(struct nfp_app *app, struct net_device *netdev,
>   	if (!nfp_mac_idx) {
>   		/* Assign a global index if non-repr or MAC is now shared. */
>   		if (entry || !port) {
> -			ida_idx = ida_simple_get(&priv->tun.mac_off_ids, 0,
> -						 NFP_MAX_MAC_INDEX, GFP_KERNEL);
> +			ida_idx = ida_alloc_max(&priv->tun.mac_off_ids,
> +						NFP_MAX_MAC_INDEX, GFP_KERNEL);
>   			if (ida_idx < 0)
>   				return ida_idx;
>   
> @@ -998,7 +998,7 @@ nfp_tunnel_add_shared_mac(struct nfp_app *app, struct net_device *netdev,
>   	kfree(entry);
>   err_free_ida:
>   	if (ida_idx != -1)
> -		ida_simple_remove(&priv->tun.mac_off_ids, ida_idx);
> +		ida_free(&priv->tun.mac_off_ids, ida_idx);
>   
>   	return err;
>   }
> @@ -1061,7 +1061,7 @@ nfp_tunnel_del_shared_mac(struct nfp_app *app, struct net_device *netdev,
>   		}
>   
>   		ida_idx = nfp_tunnel_get_ida_from_global_mac_idx(entry->index);
> -		ida_simple_remove(&priv->tun.mac_off_ids, ida_idx);
> +		ida_free(&priv->tun.mac_off_ids, ida_idx);
>   		entry->index = nfp_mac_idx;
>   		return 0;
>   	}
> @@ -1081,7 +1081,7 @@ nfp_tunnel_del_shared_mac(struct nfp_app *app, struct net_device *netdev,
>   	/* If MAC has global ID then extract and free the ida entry. */
>   	if (nfp_tunnel_is_mac_idx_global(nfp_mac_idx)) {
>   		ida_idx = nfp_tunnel_get_ida_from_global_mac_idx(entry->index);
> -		ida_simple_remove(&priv->tun.mac_off_ids, ida_idx);
> +		ida_free(&priv->tun.mac_off_ids, ida_idx);
>   	}
>   
>   	kfree(entry);

Hi,

This has been merged in -next in commit 432509013f66 but for some reason 
I looked at it again.


I just wanted to point out that this patch DOES change the behavior of 
the driver because ida_simple_get() is exclusive of the upper bound, 
while ida_alloc_max() is inclusive.

So, knowing that NFP_MAX_MAC_INDEX = 0xff = 255, with the previous code 
'ida_idx' was 0 ... 254.
Now it is 0 ... 255.

This still looks good to me, because NFP_MAX_MAC_INDEX is still not a 
power of 2.


But if 255 is a reserved value for whatever reason, then this patch has 
introduced a bug (apologies for it).

The change of behavior should have been mentioned in the commit 
description. So I wanted to make sure you was aware in case a follow-up 
fix is needed.

CJ