netdev - Re: [PATCH net 1/2] mpls: Make MPLS_IPTUNNEL select NET_MPLS

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] [day] [month] [year] [list]

Message-ID: <20201026090419.GA23071@linux.home>
Date:   Mon, 26 Oct 2020 10:04:19 +0100
From:   Guillaume Nault <gnault@...hat.com>
To:     David Ahern <dsahern@...il.com>
Cc:     Jakub Kicinski <kuba@...nel.org>,
        David Miller <davem@...emloft.net>, netdev@...r.kernel.org,
        Alexander Ovechkin <ovov@...dex-team.ru>
Subject: Re: [PATCH net 1/2] mpls: Make MPLS_IPTUNNEL select NET_MPLS_GSO

On Sun, Oct 25, 2020 at 08:48:38PM -0600, David Ahern wrote:
> On 10/25/20 3:43 PM, Jakub Kicinski wrote:
> > On Fri, 23 Oct 2020 20:48:16 +0200 Guillaume Nault wrote:
> >> On Fri, Oct 23, 2020 at 11:23:04AM -0700, Jakub Kicinski wrote:
> >>> On Fri, 23 Oct 2020 18:19:43 +0200 Guillaume Nault wrote:  
> >>>> Since commit b7c24497baea ("mpls: load mpls_gso after mpls_iptunnel"),
> >>>> mpls_iptunnel.ko has a softdep on mpls_gso.ko. For this to work, we
> >>>> need to ensure that mpls_gso.ko is built whenever MPLS_IPTUNNEL is set.  
> >>>
> >>> Does it generate an error or a warning? I don't know much about soft
> >>> dependencies, but I'd think it's optional.  
> >>
> >> Yes, it's optional from a softdep point of view. My point was that
> >> having a softdep isn't a complete solution, as a bad .config can still
> >> result in inability to properly transmit GSO packets.
> > 
> > IMO the combination of select and softdep feels pretty strange.
> > 
> > It's either a softdep and therefore optional, or we really don't 
> > want to be missing it in a correctly functioning system, and the
> > dependency should be made stronger.
> > 
> 
> That's why I liked the softdep solution - if the module is there, load it.

Okay, then I'll resubmit without the Kconfig bits.