linux-kernel - Re: [PATCH] fs/9p: Remove INET dependency

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: <ZFGdTDrPa1LD899y@codewreck.org>
Date:   Wed, 3 May 2023 08:31:24 +0900
From:   Dominique Martinet <asmadeus@...ewreck.org>
To:     Jason Andryuk <jandryuk@...il.com>
Cc:     Eric Van Hensbergen <ericvh@...nel.org>,
        Latchesar Ionkov <lucho@...kov.net>,
        Christian Schoenebeck <linux_oss@...debyte.com>,
        v9fs@...ts.linux.dev, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] fs/9p: Remove INET dependency

Jason Andryuk wrote on Tue, May 02, 2023 at 08:03:15AM -0400:
> > Good catch.
> > Could you add depends on INET to NET_9P_FD in net/9p/Kconfig?
> 
> trans_fd.o builds without INET and without UNIX.  Looks like trying to
> use either TCP or UNIX sockets would lead to runtime errors, but it
> also supports generic file descriptors.  That could be useful for
> someone.   Since it builds, would it be okay to leave this patch
> unchanged, or do you still want the depends added?  Leaving the patch
> unchanged provides more flexibility, though the runtime errors may
> surprise users.  While I built a Xen kernel without INET or UNIX, that
> is an unusual scenario.  Typically INET and UNIX sockets will be built
> in anyway even without a depends.

Hmm, good point; the socket interface is common between INET and
UNIX.. And you're right about the ability to use plain fd as well.

How about 'imply INET'? That will automatically select INET when
enabled, but allow someone to disable it explicitly.

It's rare enough that I didn't know about it though, just found it in
Documentation/kbuild/kconfig-language.rst ...
I guess this version is fine as is if you think it's too minor, I agree
kernels without INET ought to be rare enough and if it builds for
randconfig tests I'm fine with this.

-- 
Dominique Martinet | Asmadeus