netdev - Re: [RFC PATCH] kbuild: re-implement detection of CONFIG options leaked to user-space

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]

Date:   Thu, 8 Aug 2019 00:45:07 -0700
From:   Christoph Hellwig <hch@...radead.org>
To:     Arnd Bergmann <arnd@...db.de>
Cc:     Masahiro Yamada <yamada.masahiro@...ionext.com>,
        Linux Kbuild mailing list <linux-kbuild@...r.kernel.org>,
        Sam Ravnborg <sam@...nborg.org>,
        Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        Martin KaFai Lau <kafai@...com>,
        Song Liu <songliubraving@...com>, Yonghong Song <yhs@...com>,
        bpf@...r.kernel.org,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Networking <netdev@...r.kernel.org>,
        Amit Pundir <amit.pundir@...aro.org>,
        David Howells <dhowells@...hat.com>
Subject: Re: [RFC PATCH] kbuild: re-implement detection of CONFIG options
 leaked to user-space

On Tue, Aug 06, 2019 at 11:00:19AM +0200, Arnd Bergmann wrote:
> > I was playing with sed yesterday, but the resulted code might be unreadable.
> >
> > Sed scripts tend to be somewhat unreadable.
> > I just wondered which language is appropriate for this?
> > Maybe perl, or what else? I am not good at perl, though.
> 
> I like the sed version, in particular as it seems to do the job and
> I'm not volunteering to write it in anything else.

Did anyone not like sed?  I have to say I do like scripts using sed and
awk because they are fairly readable and avoid dependencies on "big"
scripting language and their optional modules that sooner or later get
pulled in.

> This one is nontrivial, since it defines two incompatible layouts for
> this structure,
> and the fdpic version is currently not usable at all from user space. Also,
> the definition breaks configurations that have both CONFIG_BINFMT_ELF
> and CONFIG_BINFMT_ELF_FDPIC enabled, which has become possible
> with commit 382e67aec6a7 ("ARM: enable elf_fdpic on systems with an MMU").
> 
> The best way forward I see is to duplicate the structure definition, adding
> a new 'struct elf_fdpic_prstatus', and using that in fs/binfmt_elf_fdpic.c.
> The same change is required in include/linux/elfcore-compat.h.

Yeah, this is a mess.  David Howells suggested something similar when
I brought the issue to his attention last time.