From: David Gibson <david@gibson.dropbear.id.au>
To: "Pierre-Clément Tosi" <ptosi@google.com>
Cc: devicetree-compiler@vger.kernel.org
Subject: Re: [PATCH] libdft: fdt_next_tag: Harden offset overflow check
Date: Thu, 12 Oct 2023 15:23:44 +1100 [thread overview]
Message-ID: <ZSd00IlSvvV2JeyK@zatzit> (raw)
In-Reply-To: <20231011172427.g4tlsew3wsjtddil@google.com>
[-- Attachment #1: Type: text/plain, Size: 2101 bytes --]
On Wed, Oct 11, 2023 at 06:24:27PM +0100, Pierre-Clément Tosi wrote:
> As 'offset' is obtained through various paths within the function by
> adding user-provided values to 'startoffset' and as we validate its
> final value by substracting it from the initial one, there is a risk
> that one of the paths might have lead to an overflow, making the check
> validate a "negative" (wrong) length, potentially causing fdt_next_tag()
> to report an invalid offset as valid through 'nextoffset'.
>
> For example, when parsing an FDT_PROP, we currently validate that
>
> offset = startoffset + len + FDT_TAGSIZE
>
> doesn't overflow but then assign
>
> offset = startoffset + len + sizeof(struct fdt_property)
>
> so harden all paths by validating the offset in the very last check.
>
> Signed-off-by: Pierre-Clément Tosi <ptosi@google.com>
The intended purpose of fdt_offset_ptr() is that you can give it
otherwise unvetted values, and it will only return non-NULL if those
correspond to a sane range of bytes within the dtb. If the test
you're adding here is necessary, then fdt_offset_ptr() isn't
accomplishing that job.
So, it seems like it's fdt_offset_ptr() that should be hardened
instead, which may fix other places as well as this one.
> ---
> libfdt/fdt.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/libfdt/fdt.c b/libfdt/fdt.c
> index 20c6415..dda342d 100644
> --- a/libfdt/fdt.c
> +++ b/libfdt/fdt.c
> @@ -213,7 +213,8 @@ uint32_t fdt_next_tag(const void *fdt, int startoffset, int *nextoffset)
> return FDT_END;
> }
>
> - if (!fdt_offset_ptr(fdt, startoffset, offset - startoffset))
> + if (!can_assume(VALID_DTB) && (offset <= startoffset
> + || !fdt_offset_ptr(fdt, startoffset, offset - startoffset)))
> return FDT_END; /* premature end */
>
> *nextoffset = FDT_TAGALIGN(offset);
--
David Gibson | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http://www.ozlabs.org/~dgibson
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
prev parent reply other threads:[~2023-10-12 4:36 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-10-11 17:24 [PATCH] libdft: fdt_next_tag: Harden offset overflow check Pierre-Clément Tosi
2023-10-12 4:23 ` David Gibson [this message]
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=ZSd00IlSvvV2JeyK@zatzit \
--to=david@gibson.dropbear.id.au \
--cc=devicetree-compiler@vger.kernel.org \
--cc=ptosi@google.com \
/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 read-only IMAP folder(s) and NNTP newsgroup(s).