Linux PCI subsystem development
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: alistair23@gmail.com
Cc: gary@garyguo.net, linux-cxl@vger.kernel.org, ojeda@kernel.org,
	linux-pci@vger.kernel.org
Subject: Re: [PATCH v4 19/22] lib: rspdm: Support SPDM get_certificate
Date: Mon, 28 Sep 2026 01:43:43 +0000	[thread overview]
Message-ID: <20260928014344.261061F000FF@smtp.kernel.org> (raw)
In-Reply-To: <20260928013123.566184-1-alistair.francis@wdc.com>

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] rspdm: SPDM cert chain truncation and underflow in SpdmCertChain::validate
- [Medium] rspdm: integer underflow in get_certificate when transport_sz is small
--

commit 4862644196f20f9c3e54555a8df7d0d01825e781
Author: Alistair Francis <alistair@alistair23.me>

lib: rspdm: Support SPDM get_certificate

This commit introduces support for the GET_CERTIFICATE SPDM command. It
sends requests to the responder and iterates over the returned certificates
for later validation.

> diff --git a/lib/rspdm/state.rs b/lib/rspdm/state.rs
> index 502d99eb13a66..780b9d459abf9 100644
> --- a/lib/rspdm/state.rs
> +++ b/lib/rspdm/state.rs
[ ... ]
> @@ -654,4 +659,107 @@ pub(crate) fn get_digests(&mut self) -> Result<(), Error> {
[ ... ]
> +    pub(crate) fn get_certificate(&mut self, slot: u8) -> Result<(), Error> {
> +        let mut request = GetCertificateReq::default();
> +        request.header.version = self.version;
> +        request.header.param1 = slot;
> +
> +        let rsp_sz =
> +            (GET_CERTIFICATE_RSP_HDR_SZ as u32 + u16::MAX as u32).min(self.transport_sz) as usize;
> +
> +        request.offset = 0;
> +        request.length = (rsp_sz - GET_CERTIFICATE_RSP_HDR_SZ) as u16;

[Severity: Medium]
What happens if the caller provides a `transport_sz` that is less than 8?

Since `GET_CERTIFICATE_RSP_HDR_SZ` is 8, `rsp_sz` could be less than 8,
causing this subtraction to underflow `usize` and panic.

> +
> +        let mut response_vec: KVec<u8> = KVec::from_elem(0u8, rsp_sz, GFP_KERNEL)?;
> +
> +        let mut request_buf = request.to_bytes()?;
> +        let response = self.get_cert_exchange(request_buf.as_mut_slice(), &mut response_vec)?;
[ ... ]
> +        let mut offset: u16 = portion_length;
> +        let mut remainder_length = rem_length as usize;
> +
> +        while remainder_length > 0 {
> +            request.offset = offset;
> +            request.length = (remainder_length.min(rsp_sz - GET_CERTIFICATE_RSP_HDR_SZ)) as u16;

[Severity: Medium]
Could this subtraction also underflow under the same conditions where `rsp_sz`
is smaller than `GET_CERTIFICATE_RSP_HDR_SZ`?

> +
> +            let mut request_buf = request.to_bytes()?;
[ ... ]
> diff --git a/lib/rspdm/validator.rs b/lib/rspdm/validator.rs
> index c058dde314401..f4e9485179d01 100644
> --- a/lib/rspdm/validator.rs
> +++ b/lib/rspdm/validator.rs
[ ... ]
> @@ -794,3 +795,154 @@ fn validate(unvalidated: &[u8], context: &'c SpdmState<'c>) -> Result<Self, Self
[ ... ]
> +impl<'a, 'c> Validate<'c, Untrusted<&'a [u8]>> for SpdmCertChain {
> +    type Err = Error;
> +
> +    type Context = &'c SpdmState<'c>;
> +
> +    fn validate(unvalidated: &[u8], context: &'c SpdmState<'c>) -> Result<Self, Self::Err> {
[ ... ]
> +        let root_hash_end = SPDM_CERT_CHAIN_HDR_SZ + context.hash_len;
> +        let mut root_hash = KVec::new();
> +        root_hash.extend_from_slice(
> +            unvalidated
> +                .get(SPDM_CERT_CHAIN_HDR_SZ..root_hash_end)
> +                .ok_or(EIO)?,
> +            GFP_KERNEL,
> +        )?;
> +
> +        let cert_chain_end = length as usize - root_hash_end;

[Severity: High]
Does this calculation safely handle payloads where `length` is smaller than
`root_hash_end`?

It appears this subtraction will underflow and trigger a panic if the untrusted
payload provides a short length.

> +        let mut certificates = KVec::new();
> +        certificates.extend_from_slice(
> +            unvalidated.get(root_hash_end..cert_chain_end).ok_or(EIO)?,

[Severity: High]
Is it intentional to use the subtracted value as the exclusive slice bound here?

Because `cert_chain_end` is calculated as a relative length but used as an
absolute upper bound, it seems to unconditionally truncate the valid certificate
stream by `root_hash_end` bytes.

> +            GFP_KERNEL,
> +        )?;
> +
> +        Ok(SpdmCertChain {
> +            length,
> +            root_hash,
> +            certificates,
> +        })
> +    }
> +}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260928011123.450800-1-alistair.francis@wdc.com?part=19

  parent reply	other threads:[~2026-09-28  1:43 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-28  1:11 [PATCH v4 00/22] lib: Rust implementation of SPDM alistair23
2026-09-28  1:11 ` [PATCH v4 01/22] virt: coco: change tsm_register() to use a const struct alistair23
2026-09-28  1:41   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 02/22] rust: transmute: add `cast_slice[_mut]` functions alistair23
2026-09-28  1:35   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 03/22] rust: create basic untrusted data API alistair23
2026-09-28  1:41   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 04/22] rust: validate: add `Validate` trait alistair23
2026-09-28  1:41   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 05/22] X.509: Make certificate parser public alistair23
2026-09-28  1:39   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 06/22] X.509: Parse Subject Alternative Name in certificates alistair23
2026-09-28  1:39   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 07/22] X.509: Move certificate length retrieval into new helper alistair23
2026-09-28  1:36   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 08/22] rust: add bindings for hash.h alistair23
2026-09-28  1:43   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 09/22] rust: error: impl From<FromBytesWithNulError> for Kernel Error alistair23
2026-09-28  1:36   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 10/22] lib: rspdm: Initial commit of Rust SPDM alistair23
2026-09-28  1:45   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 11/22] PCI/TSM: Rename pf0 to host alistair23
2026-09-28  1:41   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 12/22] PCI/TSM: Support connecting to PCIe CMA devices alistair23
2026-09-28  1:43   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 13/22] PCI/CMA: Add a PCI TSM CMA driver using SPDM alistair23
2026-09-28  1:43   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 14/22] PCI/CMA: Validate Subject Alternative Name in certificates alistair23
2026-09-28  1:44   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 15/22] lib: rspdm: Support SPDM get_version alistair23
2026-09-28  1:39   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 16/22] lib: rspdm: Support SPDM get_capabilities alistair23
2026-09-28  1:41   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 17/22] lib: rspdm: Support SPDM negotiate_algorithms alistair23
2026-09-28  1:45   ` sashiko-bot
2026-09-28  1:11 ` [PATCH v4 18/22] lib: rspdm: Support SPDM get_digests alistair23
2026-09-28  1:45   ` sashiko-bot
2026-09-28  1:31 ` [PATCH v4 19/22] lib: rspdm: Support SPDM get_certificate alistair23
2026-09-28  1:31   ` [PATCH v4 20/22] lib: rspdm: Support SPDM certificate validation alistair23
2026-09-28  1:47     ` sashiko-bot
2026-09-28  1:31   ` [PATCH v4 21/22] rust: allow extracting the buffer from a CString alistair23
2026-09-28  1:41     ` sashiko-bot
2026-09-28  1:31   ` [PATCH v4 22/22] lib: rspdm: Support SPDM challenge alistair23
2026-09-28  1:46     ` sashiko-bot
2026-09-28  1:43   ` sashiko-bot [this message]
  -- strict thread matches above, loose matches on Subject: below --
2026-09-28  1:20 [PATCH v4 19/22] lib: rspdm: Support SPDM get_certificate alistair23

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=20260928014344.261061F000FF@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=alistair23@gmail.com \
    --cc=gary@garyguo.net \
    --cc=linux-cxl@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=ojeda@kernel.org \
    --cc=sashiko-reviews@lists.linux.dev \
    /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