From: Kees Cook <keescook@chromium.org>
To: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>,
Paul Mackerras <paulus@samba.org>,
Michael Ellerman <mpe@ellerman.id.au>,
Andrew Morton <akpm@linux-foundation.org>,
"James E.J. Bottomley" <James.Bottomley@hansenpartnership.com>,
Helge Deller <deller@gmx.de>, Arnd Bergmann <arnd@arndb.de>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
linux-kernel@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
linux-ia64@vger.kernel.org, linux-parisc@vger.kernel.org,
linux-arch@vger.kernel.org, linux-mm@kvack.org
Subject: Re: [PATCH v1 10/10] lkdtm: Fix execute_[user]_location()
Date: Wed, 13 Oct 2021 00:16:32 -0700 [thread overview]
Message-ID: <202110130012.4608FFD38E@keescook> (raw)
In-Reply-To: <c551f3f4b803d1a4a184b0fa94475d405ba436d8.1633964380.git.christophe.leroy@csgroup.eu>
On Mon, Oct 11, 2021 at 05:25:37PM +0200, Christophe Leroy wrote:
> execute_location() and execute_user_location() intent
> to copy do_nothing() text and execute it at a new location.
> However, at the time being it doesn't copy do_nothing() function
> but do_nothing() function descriptor which still points to the
> original text. So at the end it still executes do_nothing() at
> its original location allthough using a copied function descriptor.
>
> So, fix that by really copying do_nothing() text and build a new
> function descriptor by copying do_nothing() function descriptor and
> updating the target address with the new location.
>
> Also fix the displayed addresses by dereferencing do_nothing()
> function descriptor.
>
> Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
> ---
> drivers/misc/lkdtm/perms.c | 45 +++++++++++++++++++++++++++++++-------
> 1 file changed, 37 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/misc/lkdtm/perms.c b/drivers/misc/lkdtm/perms.c
> index da16564e1ecd..1d03cd44c21d 100644
> --- a/drivers/misc/lkdtm/perms.c
> +++ b/drivers/misc/lkdtm/perms.c
> @@ -44,19 +44,42 @@ static noinline void do_overwritten(void)
> return;
> }
>
> +static void *setup_function_descriptor(funct_descr_t *fdesc, void *dst)
> +{
> + int err;
> +
> + if (!__is_defined(HAVE_DEREFERENCE_FUNCTION_DESCRIPTOR))
> + return dst;
> +
> + err = copy_from_kernel_nofault(fdesc, do_nothing, sizeof(*fdesc));
> + if (err < 0)
> + return ERR_PTR(err);
> +
> + fdesc->addr = (unsigned long)dst;
> + barrier();
> +
> + return fdesc;
> +}
> +
> static noinline void execute_location(void *dst, bool write)
> {
> - void (*func)(void) = dst;
> + void (*func)(void);
> + funct_descr_t fdesc;
> + void *do_nothing_text = dereference_symbol_descriptor(do_nothing);
>
> - pr_info("attempting ok execution at %px\n", do_nothing);
> + pr_info("attempting ok execution at %px\n", do_nothing_text);
> do_nothing();
>
> if (write == CODE_WRITE) {
> - memcpy(dst, do_nothing, EXEC_SIZE);
> + memcpy(dst, do_nothing_text, EXEC_SIZE);
> flush_icache_range((unsigned long)dst,
> (unsigned long)dst + EXEC_SIZE);
> }
> - pr_info("attempting bad execution at %px\n", func);
> + func = setup_function_descriptor(&fdesc, dst);
> + if (IS_ERR(func))
> + return;
I think this error case should complain with some details. :) Maybe:
pr_err("FAIL: could not build function descriptor for %px\n", dst);
> +
> + pr_info("attempting bad execution at %px\n", dst);
And then leave this pr_info() as it was, before the
setup_function_descriptor() call.
> func();
> pr_err("FAIL: func returned\n");
> }
> @@ -66,16 +89,22 @@ static void execute_user_location(void *dst)
> int copied;
>
> /* Intentionally crossing kernel/user memory boundary. */
> - void (*func)(void) = dst;
> + void (*func)(void);
> + funct_descr_t fdesc;
> + void *do_nothing_text = dereference_symbol_descriptor(do_nothing);
>
> - pr_info("attempting ok execution at %px\n", do_nothing);
> + pr_info("attempting ok execution at %px\n", do_nothing_text);
> do_nothing();
>
> - copied = access_process_vm(current, (unsigned long)dst, do_nothing,
> + copied = access_process_vm(current, (unsigned long)dst, do_nothing_text,
> EXEC_SIZE, FOLL_WRITE);
> if (copied < EXEC_SIZE)
> return;
> - pr_info("attempting bad execution at %px\n", func);
> + func = setup_function_descriptor(&fdesc, dst);
> + if (IS_ERR(func))
> + return;
> +
> + pr_info("attempting bad execution at %px\n", dst);
Same here.
> func();
> pr_err("FAIL: func returned\n");
> }
> --
> 2.31.1
>
--
Kees Cook
next prev parent reply other threads:[~2021-10-13 7:16 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-11 15:25 [PATCH v1 00/10] Fix LKDTM for PPC64/IA64/PARISC Christophe Leroy
2021-10-11 15:25 ` [PATCH v1 01/10] powerpc: Move 'struct ppc64_opd_entry' back into asm/elf.h Christophe Leroy
2021-10-12 7:10 ` Michael Ellerman
2021-10-12 8:02 ` Arnd Bergmann
2021-10-13 6:59 ` Kees Cook
2021-10-11 15:25 ` [PATCH v1 02/10] powerpc: Rename 'funcaddr' to 'addr' in 'struct ppc64_opd_entry' Christophe Leroy
2021-10-13 6:59 ` Kees Cook
2021-10-11 15:25 ` [PATCH v1 03/10] ia64: Rename 'ip' to 'addr' in 'struct fdesc' Christophe Leroy
2021-10-13 6:59 ` Kees Cook
2021-10-11 15:25 ` [PATCH v1 04/10] asm-generic: Use HAVE_DEREFERENCE_FUNCTION_DESCRIPTOR to define associated stubs Christophe Leroy
2021-10-12 6:02 ` Helge Deller
2021-10-12 6:11 ` Christophe Leroy
2021-10-12 6:48 ` Helge Deller
2021-10-13 7:00 ` Kees Cook
2021-10-14 7:20 ` Christophe Leroy
2021-10-11 15:25 ` [PATCH v1 05/10] asm-generic: Define 'funct_descr_t' to commonly describe function descriptors Christophe Leroy
2021-10-13 7:01 ` Kees Cook
2021-10-13 7:23 ` Christophe Leroy
2021-10-13 7:27 ` Kees Cook
2021-10-11 15:25 ` [PATCH v1 06/10] asm-generic: Refactor dereference_[kernel]_function_descriptor() Christophe Leroy
2021-10-13 7:02 ` Kees Cook
2021-10-13 11:20 ` Christophe Leroy
2021-10-13 11:34 ` Arnd Bergmann
2021-10-11 15:25 ` [PATCH v1 07/10] lkdtm: Force do_nothing() out of line Christophe Leroy
2021-10-13 7:02 ` Kees Cook
2021-10-11 15:25 ` [PATCH v1 08/10] lkdtm: Really write into kernel text in WRITE_KERN Christophe Leroy
2021-10-13 7:05 ` Kees Cook
2021-10-13 7:29 ` Christophe Leroy
2021-10-11 15:25 ` [PATCH v1 09/10] lkdtm: Fix lkdtm_EXEC_RODATA() Christophe Leroy
2021-10-13 7:09 ` Kees Cook
2021-10-13 7:35 ` Christophe Leroy
2021-10-13 7:23 ` Kees Cook
2021-10-13 7:39 ` Christophe Leroy
2021-10-13 7:48 ` Christophe Leroy
2021-10-13 12:45 ` Christophe Leroy
2021-10-11 15:25 ` [PATCH v1 10/10] lkdtm: Fix execute_[user]_location() Christophe Leroy
2021-10-13 7:16 ` Kees Cook [this message]
2021-10-13 12:00 ` Christophe Leroy
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=202110130012.4608FFD38E@keescook \
--to=keescook@chromium.org \
--cc=James.Bottomley@hansenpartnership.com \
--cc=akpm@linux-foundation.org \
--cc=arnd@arndb.de \
--cc=benh@kernel.crashing.org \
--cc=christophe.leroy@csgroup.eu \
--cc=deller@gmx.de \
--cc=gregkh@linuxfoundation.org \
--cc=linux-arch@vger.kernel.org \
--cc=linux-ia64@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-parisc@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mpe@ellerman.id.au \
--cc=paulus@samba.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).