All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dmitry Vyukov <dvyukov@google.com>
To: Arvind Sankar <nivedita@alum.mit.edu>
Cc: Kees Cook <keescook@chromium.org>, Marco Elver <elver@google.com>,
	"the arch/x86 maintainers" <x86@kernel.org>,
	kasan-dev <kasan-dev@googlegroups.com>,
	LKML <linux-kernel@vger.kernel.org>,
	Andrey Konovalov <andreyknvl@google.com>,
	Alexander Potapenko <glider@google.com>
Subject: Re: [RFC PATCH 1/2] lib/string: Disable instrumentation
Date: Wed, 9 Sep 2020 07:20:24 +0200	[thread overview]
Message-ID: <CACT4Y+aCa0Y8t198GSwEFShUPuOsqFV5eP8GY_7TK8fi_pML_Q@mail.gmail.com> (raw)
In-Reply-To: <20200908184003.GA4164124@rani.riverdale.lan>

On Tue, Sep 8, 2020 at 8:40 PM Arvind Sankar <nivedita@alum.mit.edu> wrote:
>
> On Tue, Sep 08, 2020 at 10:21:32AM -0700, Kees Cook wrote:
> > On Tue, Sep 08, 2020 at 11:39:11AM +0200, Marco Elver wrote:
> > > On Sun, 6 Sep 2020 at 00:23, Arvind Sankar <nivedita@alum.mit.edu> wrote:
> > > >
> > > > String functions can be useful in early boot, but using instrumented
> > > > versions can be problematic: eg on x86, some of the early boot code is
> > > > executing out of an identity mapping rather than the kernel virtual
> > > > addresses. Accessing any global variables at this point will lead to a
> > > > crash.
> > > >
> > >
> > > Ouch.
> > >
> > > We have found manifestations of bugs in lib/string.c functions, e.g.:
> > >   https://groups.google.com/forum/#!msg/syzkaller-bugs/atbKWcFqE9s/x7AtoVoBAgAJ
> > >   https://groups.google.com/forum/#!msg/syzkaller-bugs/iGBUm-FDhkM/chl05uEgBAAJ
> > >
> > > Is there any way this can be avoided?
> >
> > Agreed: I would like to keep this instrumentation; it's a common place
> > to find bugs, security issues, etc.
> >
> > --
> > Kees Cook
>
> Ok, understood. I'll revise to open-code the strscpy instead.
>
> Is instrumentation supported on x86-32? load_ucode_bsp() on 32-bit is
> called before paging is enabled, and load_ucode_bsp() itself, along with
> eg lib/earlycpio and lib/string that it uses, don't have anything to
> disable instrumentation. kcov, kasan, kcsan are unsupported already on
> 32-bit, but the others like gcov and PROFILE_ALL_BRANCHES look like they
> would just cause a crash if microcode loading is enabled.

I agree we should not disable instrumentation of such common functions.

Instead of open-coding these functions maybe we could produce both
instrumented and non-instrumented versions from the same source
implementation. Namely, place implementation in a header function with
always_inline attribute and include it from 2 source files, one with
instrumentation enabled and another with instrumentation disabled.
This way we could produce strscpy (instrumented) and __strscpy
(non-instrumented) from the same source.

  reply	other threads:[~2020-09-09  5:20 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-05 22:23 [RFC PATCH 0/2] Allow use of lib/string in early boot Arvind Sankar
2020-09-05 22:23 ` [RFC PATCH 1/2] lib/string: Disable instrumentation Arvind Sankar
2020-09-08  9:39   ` Marco Elver
2020-09-08 17:21     ` Kees Cook
2020-09-08 18:40       ` Arvind Sankar
2020-09-09  5:20         ` Dmitry Vyukov [this message]
2020-09-05 22:23 ` [RFC PATCH 2/2] x86/cmdline: Use strscpy to initialize boot_command_line Arvind Sankar
2020-09-05 22:59   ` Randy Dunlap
2020-09-05 23:16     ` Arvind Sankar

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CACT4Y+aCa0Y8t198GSwEFShUPuOsqFV5eP8GY_7TK8fi_pML_Q@mail.gmail.com \
    --to=dvyukov@google.com \
    --cc=andreyknvl@google.com \
    --cc=elver@google.com \
    --cc=glider@google.com \
    --cc=kasan-dev@googlegroups.com \
    --cc=keescook@chromium.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=nivedita@alum.mit.edu \
    --cc=x86@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.