netdev - Re: [PATCH net-next 4/6] octeontx2-vf: Ethtool support

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: <20200311101250.GI4215@unreal>
Date:   Wed, 11 Mar 2020 12:12:50 +0200
From:   Leon Romanovsky <leon@...nel.org>
To:     Sunil Kovvuri <sunil.kovvuri@...il.com>
Cc:     Andrew Lunn <andrew@...n.ch>,
        Linux Netdev List <netdev@...r.kernel.org>,
        "David S. Miller" <davem@...emloft.net>,
        Tomasz Duszynski <tduszynski@...vell.com>,
        Sunil Goutham <sgoutham@...vell.com>
Subject: Re: [PATCH net-next 4/6] octeontx2-vf: Ethtool support

On Wed, Mar 11, 2020 at 12:48:13PM +0530, Sunil Kovvuri wrote:
> On Wed, Mar 11, 2020 at 12:35 PM Leon Romanovsky <leon@...nel.org> wrote:
> >
> > On Wed, Mar 11, 2020 at 12:09:45PM +0530, Sunil Kovvuri wrote:
> > > On Wed, Mar 11, 2020 at 12:51 AM Andrew Lunn <andrew@...n.ch> wrote:
> > > >
> > > > On Wed, Mar 11, 2020 at 12:17:23AM +0530, sunil.kovvuri@...il.com wrote:
> > > > > +int __weak otx2vf_open(struct net_device *netdev)
> > > > > +{
> > > > > +     return 0;
> > > > > +}
> > > > > +
> > > > > +int __weak otx2vf_stop(struct net_device *netdev)
> > > > > +{
> > > > > +     return 0;
> > > > > +}
> > > >
> > > > Hi Sunil
> > > >
> > > > weak symbols are very unusual in a driver. Why are they required?
> > > >
> > > > Thanks
> > > >         Andrew
> > >
> > > For ethtool configs which need interface reinitialization of interface
> > > we need to either call PF or VF open/close fn()s.
> > > If VF driver is not compiled in, then PF driver compilation will fail
> > > without these weak symbols.
> > > They are there just for compilation purpose, no other use.
> >
> > It doesn't make sense, your PF driver should be changed to allow
> > compilation with those empty functions.
> >
> > Thanks
> >
>
> I didn't get, if VF driver is not compiled in then there are no
> otx2vf_open/stop fn()s defined.
> Either i have add weak fn()s or add empty ones based on VF CONFIG
> option, anyother option ?

I have no access to your implementation. But in general case:
If those fn() are accessed through pointer ->fn(), you need add
checks like this "if (..>fn) ..->fn()". If you call directly, add
empty declarations in the .h and enable them in case VF is not compiled.

Thanks

>
> Thanks,
> Sunil.