This commit introduces the ability to serialize and deserialize the internal
state of SHA-2 digest contexts (SHA-256 and SHA-512 families).
This functionality is exposed via the new OSSL_DIGEST_SERIALIZATION parameter,
which can be used with EVP_MD_CTX_get_params() to retrieve the state and with
EVP_DigestInit_ex2() to restore it into a new context.
This allows an application to save the state of a hash operation and resume it
later, which is useful for process migration or for saving the state of long-
unning computations. A new test case has been added to verify this.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/28837)
This commit introduces two new functions, EVP_MD_CTX_serialize and
EVP_MD_CTX_deserialize, to the EVP digest API.
These functions allow an application to save the state of a digest
context (EVP_MD_CTX) and restore it later. This is useful for
checkpointing long-running computations, enabling them to be paused
and resumed without starting over.
The implementation adds the OSSL_FUNC_DIGEST_SERIALIZE and
OSSL_FUNC_DIGEST_DESERIALIZE dispatch functions for providers to
supply this functionality.
Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/28837)
ec_gen_set_params() can fail after some big numbers have already been
copied over. Those need to be cleaned to avoid a memory leak on failure.
This can be done with ec_gen_cleanup(), which is also consistent in how
the ecx_gen code does it.
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/29335)
DTLSv1_listen built the HelloVerifyRequest in wbuf but invoked
msg_callback with buf and DTLS1_RT_HEADER_LENGTH, and version 0.
That caused incorrect logging and could disclose the ClientHello
to write callbacks. Use wbuf and the actual record version for the
record header, and add a second callback that reports the handshake
message bytes. No change to on-wire behavior.
Signed-off-by: Joshua Rogers <MegaManSec@users.noreply.github.com>
Reviewed-by: Frederik Wedel-Heinen <fwh.openssl@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/28916)
Fixes#28758
When X509_V_FLAG_CRL_CHECK is not set, the man pages document that X509_V_FLAG_CRL_CHECK_ALL is ignored.
Prior to 3.6.0, this was indeed the case.
In 3.6.0, the behavior changed, and setting X509_V_FLAG_CRL_CHECK_ALL began to imply X509_V_FLAG_CRL_CHECK.
This unfortunately breaks the majority of ruby installations, which relied on the documented behavior.
For consistency, this commit applies the same logic to the new X509_V_FLAG_OCSP_RESP_CHECK and X509_V_FLAG_OCSP_RESP_CHECK_ALL flags,
which are still undocumented as of 3.6.0.
All existing tests continue to pass. They also make the assumption that the xxx_CHECK_ALL flags are irrelevant unless xxx_CHECK is set.
We could add a new test for this regression. I'll leave that to another commit.
CLA: trivial
Reviewed-by: Nikola Pajkovsky <nikolap@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/28797)
(cherry picked from commit cbaf28ce48)
clang-format sensibly thinks this is an arithmatic operation,
and formats the math. Sadly it does not know we eventually
stringify this behind several other layers of nested macros
and so putting spaces in here is bad.
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Norbert Pocs <norbertp@openssl.org>
Reviewed-by: Neil Horman <nhorman@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/29350)
Changes for SSL_listen_ex squashed creation of qlog objects even when
configured.
Fix that up so qlog objects are created regardless of weather we use
SSL_accept_connection or SSL_listen_ex
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/27397)
We recently found that the addition of a git config command in
util/find-doc-nits is broken in some cases, sepecifically because git
around version 2.46 broke command line compatibility, replacing the
--regexp option with the --get-regexp option. So to maintain usage of
this specific command to parse the .gitconfig file, we would need to do
some extra version detection to construct the proper command line.
However, find-doc-nits already has a fallback condition, which does some
pure perl parsing of the gitconfig file, which works perfectly well.
Instead of trying to do version matching to construct the right form of
the git config command line, just remove it all, and rely on the perl
parrse to do this work for us, which works currently in all cases.
Fixes#29197
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/29304)
In was premature to make OSSL_(EN|DE)CODER_CTX_[sg]et_finalized() be
public interfaces. Forunately, these have not yet appeared outside the
"master" branch, so we can still retract them.
Also, in the case of decoders, the implementation failed to take into
account that the context was duplicated before it was returned to the
user, and the duplicated copy failed to copy the "finalized" field.
This commit also renames "finalized" to "frozen", because
finalisation is a misleading term in this context, it suggests
resource reclamation during garbage collection or deallocation,
not marking a structure partly immutable.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Eugene Syromiatnikov <esyr@openssl.org>
Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/29206)
We support selection of ML-KEM and ML-DSA key formats on input and
output at the provider level, these are essentially global defaults,
in effect for the lifetime of the process.
Unfortunately, the JAVA interface in openssl-jostle needs to be able to
output a specific key in seed-only form. To that end, this PR
introduces a new "output-formats" PKEY encoding parameter, that can be used
with OSSL_ENCODER_CTX_set_params(3) when encoding a key to PKCS#8, after
using OSSL_ENCODER_CTX_new_for_key(3), rather than i2d_PrivateKey(3),
i2d_PKCS8PrivateKey(3) or PEM equivalents.
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Eugene Syromiatnikov <esyr@openssl.org>
Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/29206)
In a hypothetical scenario that jent_entropy_init_ex fails, or if
get_jitter_random-value fails, there are a few unexpected
posibilities.
If jent_entropy_init_ex fails, the seed initialisation may return NULL
and then DRBG will be initiated with NULL seed, which will
automatically fallback to os-seed, which will escape module boundary
(if this jitter rng is from the fips module), and call getrandom
syscall.
And separately if get_jitter_random_value fails, it may put DRBG in an
error state, but it might not put the FIPS module in error state, like
it should as per the ISO standard.
To instrument these things, I had to create tampered
jitterentropy-library that always returns errors for init_ex and
read_entropy apis, and then use gdb tracing on both libcrypto.so and
fips.so.
The most minimal solution to above hypothetical error code paths, is
to simply call ossl_set_error_state. It is either harmless, or in case
of fips-jitter will correctly put the FIPS module into error state and
prevent any further operation; and cruitially prevent silent fallback
to getrandom syscall.
Note it is unlikely that this ever was out of compliance, as often
enough getrandom syscall goes to a kernel with validated entropy
source; and openssl fips module still did reject sampling which is too
entropy source compliant.
Nonetheless it is good to fix this hypothetical error path, and
backport this to 3.5 and up.
This is similar / additional fixes, to this previous change:
- https://github.com/openssl/openssl/pull/25957
- b9886a6f34
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/29226)
