mbox series

[v7,0/4] Add support for x509 certs with NIST p256 and p192 keys

Message ID 20210201151910.1465705-1-stefanb@linux.ibm.com
Headers show
Series Add support for x509 certs with NIST p256 and p192 keys | expand

Message

Stefan Berger Feb. 1, 2021, 3:19 p.m. UTC
This series of patches adds support for x509 certificates signed by a CA
that uses NIST p256 or p192 keys for signing. It also adds support for
certificates where the public key is a NIST p256 or p192 key. The math
for ECDSA signature verification is also added.

Since self-signed certificates are verified upon loading, the following
script can be used for testing:

k=$(keyctl newring test @u)

while :; do
	for hash in sha1 sha224 sha256 sha384 sha512; do
		openssl req \
			-x509 \
			-${hash} \
			-newkey ec \
			-pkeyopt ec_paramgen_curve:prime256v1 \
			-keyout key.pem \
			-days 365 \
			-subj '/CN=test' \
			-nodes \
			-outform der \
			-out cert.der
		keyctl padd asymmetric testkey $k < cert.der
		if [ $? -ne 0 ]; then
			echo "ERROR"
			exit 1
		fi
	done
done

It also works with restricted keyrings where an RSA key is used to sign
a NIST P256/P192 key. Scripts for testing are here:

https://github.com/stefanberger/eckey-testing

The ECDSA signature verification will be used by IMA Appraisal where ECDSA
file signatures stored in RPM packages will use substantially less space
than if RSA signatures were to be used.

   Stefan

v6->v7:
  - Moved some OID defintions to patch 1 for bisectability
  - Applied R-b's
  
v5->v6:
  - moved ecdsa code into its own module ecdsa_generic built from ecdsa.c
  - added script-generated test vectors for NIST P256 & P192 and all hashes
  - parsing of OID that contain header with new parse_oid()

v4->v5:
  - registering crypto support under names ecdsa-nist-p256/p192 following
    Hubert Xu's suggestion in other thread
  - appended IMA ECDSA support patch

v3->v4:
  - split off of ecdsa crypto part; registering akcipher as "ecdsa" and
    deriving used curve from digits in parsed key

v2->v3:
  - patch 2 now includes linux/scatterlist.h

v1->v2:
  - using faster vli_sub rather than newly added vli_mod_fast to 'reduce'
    result
  - rearranged switch statements to follow after RSA
  - 3rd patch from 1st posting is now 1st patch


Stefan Berger (4):
  crypto: Add support for ECDSA signature verification
  x509: Detect sm2 keys by their parameters OID
  x509: Add support for parsing x509 certs with ECDSA keys
  ima: Support EC keys for signature verification

 crypto/Kconfig                            |  10 +
 crypto/Makefile                           |   6 +
 crypto/asymmetric_keys/public_key.c       |  19 ++
 crypto/asymmetric_keys/x509_cert_parser.c |  44 ++-
 crypto/ecc.c                              |  13 +-
 crypto/ecc.h                              |  28 ++
 crypto/ecdsa.c                            | 361 ++++++++++++++++++++++
 crypto/ecdsasignature.asn1                |   4 +
 crypto/testmgr.c                          |  12 +
 crypto/testmgr.h                          | 267 ++++++++++++++++
 include/keys/asymmetric-type.h            |   6 +
 include/linux/oid_registry.h              |   7 +
 lib/oid_registry.c                        |  13 +
 security/integrity/digsig_asymmetric.c    |  29 +-
 14 files changed, 790 insertions(+), 29 deletions(-)
 create mode 100644 crypto/ecdsa.c
 create mode 100644 crypto/ecdsasignature.asn1

Comments

David Howells Feb. 1, 2021, 4:13 p.m. UTC | #1
Stefan Berger <stefanb@linux.ibm.com> wrote:

> v6->v7:
>   - Moved some OID defintions to patch 1 for bisectability
>   - Applied R-b's

But I can't now apply 2-4 without patch 1.

David
David Howells Feb. 1, 2021, 4:36 p.m. UTC | #2
Stefan Berger <stefanb@linux.ibm.com> wrote:

> 1) the whole series goes through the crypto tree
> 
> 2) I make the OIDs addition patch 1 that both keyrings and crypto take
> separately?

The first might be easiest, but 2 is okay also.  You'll just need to give
myself and Herbert separate branches to pull, rooted on the same commit.

Btw, what do patches 2-4 do if patch 1 isn't applied?


David
Herbert Xu Feb. 2, 2021, 3:59 a.m. UTC | #3
On Mon, Feb 01, 2021 at 11:45:16AM -0500, Stefan Berger wrote:
>

> With the crypto module missing in the kernel you will get an error trying to

> load an x509 certificate that needs the algorithm to verify the self-signed

> signature.

> 

> Before I post yet another series I hope that Herbert can say whether option

> 1) would work for him.


Please be patient.  We need to make sure that whatever scheme you
use for your algorithm also works for the driver authors who are
working in the same area.  Because if we end up having to change
the scheme then that'll just create more churn for you and David.

Thanks,
-- 
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt