linux-hardening - Re: [PATCH net-next v2] netlink: split up copies in the ack construction

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]

Date:   Thu, 17 Nov 2022 14:35:32 -0800
From:   Kees Cook <keescook@...omium.org>
To:     Jakub Kicinski <kuba@...nel.org>
Cc:     Stephen Hemminger <stephen@...workplumber.org>,
        "Gustavo A. R. Silva" <gustavo@...eddedor.com>,
        David Ahern <dsahern@...nel.org>, davem@...emloft.net,
        netdev@...r.kernel.org, edumazet@...gle.com, pabeni@...hat.com,
        linux-hardening@...r.kernel.org
Subject: Re: [PATCH net-next v2] netlink: split up copies in the ack
 construction

On Thu, Nov 17, 2022 at 12:36:15PM -0800, Jakub Kicinski wrote:
> On Thu, 17 Nov 2022 08:25:56 -0800 Stephen Hemminger wrote:
> > > I was asking based on your own commit 1e6e9d0f4859 ("uapi: revert
> > > flexible-array conversions"). This is uAPI as well.  
> >  
> > Some of the flex-array conversions fixed build warnings that occur in
> > iproute2 when using Gcc 12 or later.
> 
> Alright, this is getting complicated. I'll post a patch to fix 
> the issue I've added and gently place my head back into the sand.

Thanks! I think the path forward is clear. I should not have suggested
adding a flex-array member to the "header" struct lo these many moons
ago. You and Gustavo are right: we need a separate struct with the header
at the beginning, just as iproute2 is doing itself.

As for testing, I can do that if you want -- the goal was to make sure
the final result doesn't trip FORTIFY when built with -fstrict-flex-arrays
(not yet in a released compiler version, but present in both GCC and Clang
truck builds) and with __builtin_dynamic_object_size() enabled (which
is not yet in -next, as it is waiting on the last of ksize() clean-ups).

-- 
Kees Cook