Implicit rejection of PKCS#1 v1.5 RSA

[David Fernandez Gonzalez]

OpenSSL 3.2.0 introduced a change on PKCS#1 v1.5 RSA to return random output instead of an exception when detecting wrong padding (https://github.com/openssl/openssl/pull/13817).

There are available backports already:

* 3.0 https://gitlab.com/redhat/centos-stream/rpms/openssl/-/blob/c9s/0120-RSA-PKCS15-implicit-rejection.patch?ref_type=heads

* 1.1.1 https://gitlab.com/redhat/centos-stream/rpms/openssl/-/blob/c8s/openssl-1.1.1-pkcs1-implicit-rejection.patch?ref_type=heads

This change is needed to fix CVE-2023-50782.

[Launchpad Janitor (janitor)]

This bug was fixed in the package openssl - 3.0.10-1ubuntu2.3

---------------
openssl (3.0.10-1ubuntu2.3) mantic-security; urgency=medium

  * SECURITY UPDATE: Implicit rejection for RSA PKCS#1 (LP: #2054090)
    - debian/patches/openssl-pkcs1-implicit-rejection.patch:
      Return deterministic random output instead of an error in case
      there is a padding error in crypto/cms/cms_env.c,
      crypto/evp/ctrl_params_translate.c, crypto/pkcs7/pk7_doit.c,
      crypto/rsa/rsa_ossl.c, crypto/rsa/rsa_pk1.c,
      crypto/rsa/rsa_pmeth.c, doc/man1/openssl-pkeyutl.pod.in,
      doc/man1/openssl-rsautl.pod.in, doc/man3/EVP_PKEY_CTX_ctrl.pod,
      doc/man3/EVP_PKEY_decrypt.pod,
      doc/man3/RSA_padding_add_PKCS1_type_1.pod,
      doc/man3/RSA_public_encrypt.pod, doc/man7/provider-asym_cipher.pod,
      include/crypto/rsa.h, include/openssl/core_names.h,
      include/openssl/rsa.h,
      providers/implementations/asymciphers/rsa_enc.c and
      test/recipes/30-test_evp_data/evppkey_rsa_common.txt.

 -- David Fernandez Gonzalez <email address hidden> Wed, 21 Feb 2024 11:45:39 +0100

[Launchpad Janitor (janitor)]

This bug was fixed in the package openssl - 1.1.1f-1ubuntu2.22

---------------
openssl (1.1.1f-1ubuntu2.22) focal-security; urgency=medium

  * SECURITY UPDATE: Implicit rejection for RSA PKCS#1 (LP: #2054090)
    - debian/patches/openssl-1.1.1-pkcs1-implicit-rejection.patch:
      Return deterministic random output instead of an error in case
      there is a padding error in crypto/cms/cms_env.c,
      crypto/pkcs7/pk7_doit.c, crypto/rsa/rsa_local.h,
      crypto/rsa/rsa_ossl.c, crypto/rsa/rsa_pk1.c, crypto/rsa/rsa_pmeth.c,
      doc/man1/pkeyutl.pod, doc/man1/rsautl.pod,
      doc/man3/EVP_PKEY_CTX_ctrl.pod, doc/man3/EVP_PKEY_decrypt.pod,
      doc/man3/RSA_padding_add_PKCS1_type_1.pod,
      doc/man3/RSA_public_encrypt.pod, include/openssl/rsa.h and
      test/recipes/30-test_evp_data/evppkey.txt.

 -- David Fernandez Gonzalez <email address hidden> Fri, 16 Feb 2024 16:41:31 +0100

[Launchpad Janitor (janitor)]

This bug was fixed in the package openssl - 3.0.2-0ubuntu1.15

---------------
openssl (3.0.2-0ubuntu1.15) jammy-security; urgency=medium

  * SECURITY UPDATE: Implicit rejection for RSA PKCS#1 (LP: #2054090)
    - debian/patches/openssl-pkcs1-implicit-rejection.patch:
      Return deterministic random output instead of an error in case
      there is a padding error in crypto/cms/cms_env.c,
      crypto/evp/ctrl_params_translate.c, crypto/pkcs7/pk7_doit.c,
      crypto/rsa/rsa_ossl.c, crypto/rsa/rsa_pk1.c,
      crypto/rsa/rsa_pmeth.c, doc/man1/openssl-pkeyutl.pod.in,
      doc/man1/openssl-rsautl.pod.in, doc/man3/EVP_PKEY_CTX_ctrl.pod,
      doc/man3/EVP_PKEY_decrypt.pod,
      doc/man3/RSA_padding_add_PKCS1_type_1.pod,
      doc/man3/RSA_public_encrypt.pod, doc/man7/provider-asym_cipher.pod,
      include/crypto/rsa.h, include/openssl/core_names.h,
      include/openssl/rsa.h,
      providers/implementations/asymciphers/rsa_enc.c and
      test/recipes/30-test_evp_data/evppkey_rsa_common.txt.

 -- David Fernandez Gonzalez <email address hidden> Fri, 16 Feb 2024 09:51:30 +0100

[David Fernandez Gonzalez (litios)]

Bionic released in ESM Infra, version 1.1.1-1ubuntu2.1~18.04.23+esm5

[David Fernandez Gonzalez (litios)]

https://ubuntu.com/security/notices/USN-6663-1

[David Fernandez Gonzalez (litios)]

Fixed released for Xenial ESM: 1.0.2g-1ubuntu4.20+esm12

https://ubuntu.com/security/notices/USN-6663-2

[David Fernandez Gonzalez (litios)]

Trusty would require a significant backport, marking it as won't fix to prevent possible regressions.