linux-kernel - Re: [RFC PATCH] UML: add support for KASAN under x86

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: <4f382794416c023b6711ed2ca645abe4fb17d6da.camel@sipsolutions.net>
Date:   Thu, 16 Jan 2020 08:57:22 +0100
From:   Johannes Berg <johannes@...solutions.net>
To:     Patricia Alfonso <trishalfonso@...gle.com>
Cc:     jdike@...toit.com, richard@....at, anton.ivanov@...bridgegreys.com,
        aryabinin@...tuozzo.com, dvyukov@...gle.com,
        David Gow <davidgow@...gle.com>,
        Brendan Higgins <brendanhiggins@...gle.com>,
        linux-um@...ts.infradead.org, linux-kernel@...r.kernel.org,
        kasan-dev@...glegroups.com
Subject: Re: [RFC PATCH] UML: add support for KASAN under x86_64

Hi,

> This seems like a good idea. I'll keep the #ifdef around
> KASAN_SHADOW_SIZE, but add "select HAVE_ARCH_KASAN if X86_64" as well.
> This will make extending it later easier.

Yeah, that makes a lot of sense.

I think once somebody (Anton? Richard?) start applying patches again,
they will pick up my revert for CONFIG_CONSTRUCTORS:

https://patchwork.ozlabs.org/patch/1204275/

(See there for why I had to revert it)

If I remember correctly, KASAN depends on CONSTRUCTORS, so that revert
will then break your patch here?

And if I remember from looking at KASAN, some of the constructors there
depended on initializing after the KASAN data structures were set up (or
at least allocated)? It may be that you solved that by allocating the
shadow so very early though.

In any case, I think you should pick up that revert of
CONFIG_CONSTRUCTORS and see what you have to do to make it still work,
if that's possible.

If not, then ... tricky, not sure what to do. Maybe then we could
somehow hook in and have our own constructor that's called even before
the compiler-emitted ASAN constructors, to allocate the necessary data
structures.

johannes