netdev - Re: [PATCH net-next] net: vertexcom: default to disabled on kbuild

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: <20220104071522.548024b2@kicinski-fedora-PC1C0HJN.hsd1.ca.comcast.net>
Date:   Tue, 4 Jan 2022 07:15:22 -0800
From:   Jakub Kicinski <kuba@...nel.org>
To:     Saeed Mahameed <saeed@...nel.org>
Cc:     "David S. Miller" <davem@...emloft.net>, netdev@...r.kernel.org,
        Saeed Mahameed <saeedm@...dia.com>,
        Stefan Wahren <stefan.wahren@...e.com>
Subject: Re: [PATCH net-next] net: vertexcom: default to disabled on kbuild

On Tue, 4 Jan 2022 00:25:55 -0800 Saeed Mahameed wrote:
> On Mon, Jan 03, 2022 at 08:43:21AM -0800, Jakub Kicinski wrote:
> >On Sun,  2 Jan 2022 14:11:26 -0800 Saeed Mahameed wrote:  
> >> Sorry for being rude but new vendors/drivers are supposed to be disabled
> >> by default, otherwise we will have to manually keep track of all vendors
> >> we are not interested in building.  
> >
> >Vendors default to y, drivers default to n. Vendors don't build
> >anything, hence the somewhat unusual convention. Are you saying
> >you want to change all the Kconfigs... including Mellanox?  
> 
> I didn't know about the vendor vs drivers thing! What's the point ? :) 
> Now as i think about it, Yes, why not change all vendors including Mellanox? 
> I don't see any point to visit a sub-tree where nothing is going to build,
> and waste a chunk of precious build time.
> 
> Anyhow, I can add the default 'n' to the driver and revert the
> vendor back to 'y', Please let me know if you'd want that.

I'm not sure what the origins of the convention are. Maybe when 
the Ethernet drivers got split by vendor we were concerned that 
if we default vendors to n someone doing olddefconfig will lose
drivers?

In any case, my weak preference is also to abandon the convention
unless someone can point out why it's good/useful. The fewer rules 
the better.

FWIW I think that if there is no default set explicitly in Kconfig
it will default to n so if you're writing the patch - you can remove
the overrides rather than setting them to n.