netdev - Re: [PATCH net-next 1/5] net: bridge: provide shim definition for br_vlan

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: <20211027195442.mwlyvrokj6ygpv7m@skbuf>
Date:   Wed, 27 Oct 2021 19:54:43 +0000
From:   Vladimir Oltean <vladimir.oltean@....com>
To:     Nikolay Aleksandrov <nikolay@...dia.com>
CC:     "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        Jakub Kicinski <kuba@...nel.org>,
        "David S. Miller" <davem@...emloft.net>,
        Roopa Prabhu <roopa@...dia.com>, Jiri Pirko <jiri@...dia.com>,
        Ido Schimmel <idosch@...dia.com>
Subject: Re: [PATCH net-next 1/5] net: bridge: provide shim definition for
 br_vlan_flags

On Wed, Oct 27, 2021 at 10:50:47PM +0300, Nikolay Aleksandrov wrote:
> On 27/10/2021 22:45, Vladimir Oltean wrote:
> > On Wed, Oct 27, 2021 at 10:28:12PM +0300, Nikolay Aleksandrov wrote:
> >> On 27/10/2021 19:21, Vladimir Oltean wrote:
> >>> br_vlan_replay() needs this, and we're preparing to move it to
> >>> br_switchdev.c, which will be compiled regardless of whether or not
> >>> CONFIG_BRIDGE_VLAN_FILTERING is enabled.
> >>>
> >>> Signed-off-by: Vladimir Oltean <vladimir.oltean@....com>
> >>> ---
> >>>  net/bridge/br_private.h | 5 +++++
> >>>  1 file changed, 5 insertions(+)
> >>>
> >>> diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h
> >>> index 3c9327628060..cc31c3fe1e02 100644
> >>> --- a/net/bridge/br_private.h
> >>> +++ b/net/bridge/br_private.h
> >>> @@ -1708,6 +1708,11 @@ static inline bool br_vlan_can_enter_range(const struct net_bridge_vlan *v_curr,
> >>>  	return true;
> >>>  }
> >>>  
> >>> +static inline u16 br_vlan_flags(const struct net_bridge_vlan *v, u16 pvid)
> >>> +{
> >>> +	return 0;
> >>> +}
> >>> +
> >>>  static inline int br_vlan_replay(struct net_device *br_dev,
> >>>  				 struct net_device *dev, const void *ctx,
> >>>  				 bool adding, struct notifier_block *nb,
> >>>
> >>
> >> hm, shouldn't the vlan replay be a shim if bridge vlans are not defined?
> >> I.e. shouldn't this rather be turned into br_vlan_replay's shim?
> >>
> >> TBH, I haven't looked into the details just wonder why we would compile all that vlan
> >> code if bridge vlan filtering is not enabled.
> > 
> > The main reason is that I would like to avoid #ifdef if possible. If you
> > have a strong opinion otherwise I can follow suit.
> > 
> 
> Well, I see that we add ifdefs for IGMP, so I don't see a reason why not
> to ifdef out the vlan replay in the same way too.
> 
> I don't have a strong preference either way, end result is the same.

Since the caller and the callee are in the same C file, shimming out is
not as clean as providing a static inline function definition with an
empty body, and if I could avoid doing what I did for

br_mdb_replay()
{
#ifdef CONFIG_BRIDGE_IGMP_SNOOPING
	<some other variables>;
	int err;

	err = <body>;

	if (err)
		return err;
#endif

	return 0;
}

I'd do it. For br_vlan_replay() I could avoid it, so I left it at that.