test: add error reasons to TLS 1.2 PRF tests

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25135)

test: add error reasons to Single Step KDF tests

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25135)

test: add error reasons to SSHKDF tests

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25135)

test: add error reasons to PBKDF2 tests

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25135)

test: add error reasons to HKDF tests

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25135)

test: add error reasons to KMAC tests

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25135)

test: add error reasons to RSA tests

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25135)

fips: change from function call to macro in rsa_enc.c

Use of the function instead of the macro for the indicator unapproved check was
noted in: https://github.com/openssl/openssl/pull/25070#discussion_r1706564363
Fix things to use the macro properly.

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25134)

amend! fixup! limit bignums to 512 bytes

fixup! limit bignums to 512 bytes

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25013)

limit bignums to 128 bytes

Keep us from spinning forever doing huge amounts of math in the fuzzer

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25013)

Add "no-fips-post" configure option.

Using this option disables the OpenSSL FIPS provider
self tests.
This is intended for debugging purposes only,
as it breaks FIPS compliance.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25063)

Fix FIPS indicator defines for larger indicies.

A newer PR is using setable3 now so these indicies should be fixed.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25118)

test: add negative tests for KBKDF key size check under FIPS

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25095)

fips: add kbkdf key length check as per SP 800-131a revision 2

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25095)

fips: add kbkdf key check checking function

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25095)

doc: docment key-check param for kbkdf

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25095)

doc: document kbkdf key check argument for fipsinstall

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25095)

fips: install with the kbkdf key check option set

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25095)

params: add kbkdf key check param

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25095)

fipsinstall: add kbkdf key check option

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25095)

Update BN_add.pod documentation so it is consistent with header declarations

CLA: trivial

Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24215)

Remove duplicate colon in otherName display

The colon is already added in X509V3_EXT_val_prn(). In fact, the other
branches from i2v_GENERAL_NAME() do not include a trailing colon.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23428)

Add OIDs id-kp-wisun-fan-device and id-on-hardwareModule

Sub-OIDs for {iso(1) identified-organization(3) dod(6) internet(1)
private(4) enterprise(1) 45605} are recorded in the document "Wi-SUN
Assigned Value Registry" (WAVR).

OID id-on-hardwareModule is defined in RFC 4108.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/23428)

Remove the event queue code

PR #18345 added some code for an event queue. It also added a test for it.
Unfortunately this event queue code has never been used for anything.
Additionally the test was never integrated into a test recipe, so it never
actually gets invoked via "make test". This makes the code entirely dead,
unnecessarily bloats the size of libssl and causes a decrease in our
testing code coverage value.

We remove the dead code.

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25100)

dh_kmgmt.c: Avoid expensive public key validation for known safe-prime groups

The partial validation is fully sufficient to check the key validity.

Thanks to Szilárd Pfeiffer for reporting the issue.

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25088)

ossl_print_attribute_value(): use a sequence value only if type is a sequence

Move the switch to print a distinguished name inside the
switch by the printed attribute type, otherwise a malformed
attribute will cause a crash.

Updated the fuzz corpora with the testcase

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25087)

rsa_pss_compute_saltlen(): Avoid integer overflows and check MD and RSA sizes

Fixes Coverity 1604651

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25085)

do_print_ex(): Avoid possible integer overflow

Fixes Coverity 1604657
Fixes openssl/project#780

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25084)

test: update SSL API test in light of PKCS#1 version 1.5 padding change under FIPS

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

test: update SSL old test in light of PKCS#1 version 1.5 padding change under FIPS

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

sslapitest: add meaningful skip messages

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

cms: fix tests in light of PKCS#1 version 1.5 padding check

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

fips: add PKCS#1 version 1.5 padding check option

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

rsa: disallow PKCS#1 version 1.5 padding for encrpytion under FIPS.

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

fips: add option to disable PKCS#1 version 1.5 padding

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

param: add ASYM_CIPHER_PARAM_PKCS15_PADDING_DISABLED

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

doc: document pkcs15-padding-disabled param for RSA

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

doc: fix typos

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

doc: document the fipsintsall option to disallow PKCS#1 version 1.5 padding for key agreement & transport

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

fipsinstall: add option to disable RSA PKCS#1 version 1.5 padding

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25070)

Add a test for a missing supported_versions extension in the HRR

Confirm that we correctly fail if supported_versions is missing from an
HRR.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25068)

Check that a supported_versions extension is present in an HRR

If an HRR is sent then it MUST contain supported_versions according to the
RFC. We were sanity checking any supported_versions extension that was sent
but failed to verify that it was actually present.

Fixes #25041

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25068)

Add logging support for early data

-trace option didn't cover early data message which resulted in
misleading logging.

CLA: trivial

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25026)

ssl: factorize and improved hex conversion code

Add inline qualifier to avoid exporting a function for one unique use

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24968)

crypto: factorize to hex chars conversion code.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24968)

ssl: remove stdio.h and sprintf use from libssl.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24968)

apps: add missing entry to tls extension label list

noticed by @sftcd

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24968)

Fix '--strict-warnings' build breakage

... due to a missing const.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24968)

Fix typos found by codespell

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24949)

coveralls.yml: Do not run tests in parallel

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Neil Horman <nhorman@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25105)

Fix Potential NULL pointer dereference

CLA: trivial

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Paul Yang <kaishen.yy@antfin.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25065)

Use parent directory instead of index.html

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25073)

Update links in CONTRIBUTING.md

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25073)

Fix some small typos

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25073)

Test vectors from rfc9579 and creation tests

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24577)

Implementation of the RFC 9579, PBMAC1 in PKCS#12 - documentation

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24577)

Implementation of the RFC 9579, PBMAC1 in PKCS#12

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24577)

test: omit failing test when using a FIPS provider that supports eddsa_no_verify_digested

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25032)

test: update EdDSA tests for FIPS and no pre-hashing

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25032)

EdDSA: disallow verification from a pregenerated hash when in FIPS

Also includes an indicator and the capability to bypass via configuration
or params.

Fixes #24937

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25032)

fips: add no digested option to FIPS provider

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25032)

param: add OSSL_SIGNATURE_PARAM_EDDSA_VERIFY_DIGESTED

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25032)

doc: add verify-digested param for EdDSA

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25032)

doc: add documentation for -eddsa_no_verify_digested fipsinstall option

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25032)

fipsinstall: add ed_no_verify_digested option

FIPS doesn't permit message hashes to be processed by thee algorithms.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/25032)

Add FIPS DSA Keygen tests

Adjust the existing tests to disable DSA keygen in FIPS mode.

Allow evp_test to load DSA 'KeyParams' that can then be used to
perform a DSA KeyGen.

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24978)

Disallow DSA Keygen in the FIPS provider

This uses a FIPS indicator.
Since DSA KeyGen is only useful for DSA signing,
it reuses the DSA signing FIPS configuration option and settable ctx name.

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24978)

Add KeyManagement keygen parameter getter/gettable functions.

Added OSSL_FUNC_keymgmt_gen_get_params() and
OSSL_FUNC_keymgmt_gen_gettable_params()

This will allow a FIPS indicator parameter to be queried after keygen.

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24978)

Fix evp_test HKDF failure in crosstest 3.1.2 FIPS provider with master

Fixes #25089
The test to check if the FIPS indicator was correct failed in 3.1.2
since EVP_PKEY_CTX_get_params() returns 0 if there is no
gettable/getter.

The code has been modified to return 1 if there is no gettable.
Manually reproduced and tested by copying the 3.1.2 FIPS provider to master.

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25093)

github: fix quoting in github workflow for jitter tests

Nested quoting got ignore previously. And this way one can specify
string name directly.

Successfully run with Jitter at
https://github.com/xnox/openssl/actions/runs/10223149419/job/28289017013

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25053)

Add new configurable item `pbkdf2-lower-bound-check`

Since FIPS provider performs lower bound check by default from v3.0, the
default value for new configurable item will be one.

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24120)

Add explicit indicator for PBKDF2

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24120)

Do not implicitly start connection with SSL_handle_events() or SSL_poll()

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25069)

Return infinity time from SSL_get_event_timeout when the connection is not started

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25069)

Do not falsely start the connection through SSL_pending()/_has_pending()

Fixes #25054

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25069)

fix: util/mkinstallvars.pl mistreated LDLIBS on Unix (and Windows)

Don't do comma separation on those platforms.

Fixes #24986

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25018)

ssl_evp_cipher_fetch(): Avoid using 3DES from the FIPS provider

Avoid using a fetched cipher that is decrypt-only
which is the case for 3DES from the fips provider.

Add a decrypt-only parameter to the EVP_CIPHER and test it
in libssl when fetching.

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25028)

3DES ciphersuites are not allowed in FIPS anymore

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25028)

Add enable-weak-ssl-ciphers to full_featured CI job

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25028)

Speed up SSL_add_{file,dir}_cert_subjects_to_stack

The X509_NAME comparison function converts its arguments to DER using
i2d_X509_NAME before comparing the results using memcmp(). For every
invocation of the comparison function (of which there are many when
loading many certificates), it allocates two buffers of the appropriate
size for the DER encoding.

Switching to static buffers (possibly of X509_NAME_MAX size as defined
in crypto/x509/x_name.c) would not work with multithreaded use, e.g.,
when two threads sort two separate STACK_OF(X509_NAME)s at the same
time. A suitable re-usable buffer could have been added to the
STACK_OF(X509_NAME) if sk_X509_NAME_compfunc did have a void* argument,
or a pointer to the STACK_OF(X509_NAME) – but it does not.

Instead, copy the solution chosen in SSL_load_client_CA_file() by
filling an LHASH_OF(X509_NAME) with all existing names in the stack and
using that to deduplicate, rather than relying on sk_X509_NAME_find(),
which ends up being very slow.

Adjust SSL_add_dir_cert_subjects_to_stack() to keep a local
LHASH_OF(X509_NAME)s over the complete directory it is processing.

In a small benchmark that calls SSL_add_dir_cert_subjects_to_stack()
twice, once on a directory with one entry, and once with a directory
with 1000 certificates, and repeats this in a loop 10 times, this change
yields a speed-up of 5.32:

| Benchmark 1: ./bench 10 dir-1 dir-1000
|   Time (mean ± σ):      6.685 s ±  0.017 s    [User: 6.402 s, System: 0.231 s]
|   Range (min … max):    6.658 s …  6.711 s    10 runs
|
| Benchmark 2: LD_LIBRARY_PATH=. ./bench 10 dir-1 dir-1000
|   Time (mean ± σ):      1.256 s ±  0.013 s    [User: 1.034 s, System: 0.212 s]
|   Range (min … max):    1.244 s …  1.286 s    10 runs
|
| Summary
|   LD_LIBRARY_PATH=. ./bench 10 dir-1 dir-1000 ran
|    5.32 ± 0.06 times faster than ./bench 10 dir-1 dir-1000

In the worst case scenario where many entries are added to a stack that
is then repeatedly used to add more certificates, and with a larger test
size, the speedup is still very significant. With 15000 certificates,
a single pass to load them, followed by attempting to load a subset of
1000 of these 15000 certificates, followed by a single certificate, the
new approach is ~85 times faster:

| Benchmark 1: ./bench 1 dir-15000 dir-1000 dir-1
|   Time (mean ± σ):     176.295 s ±  4.147 s    [User: 174.593 s, System: 0.448 s]
|   Range (min … max):   173.774 s … 185.594 s    10 runs
|
| Benchmark 2: LD_LIBRARY_PATH=. ./bench 1 dir-15000 dir-1000 dir-1
|   Time (mean ± σ):      2.087 s ±  0.034 s    [User: 1.679 s, System: 0.393 s]
|   Range (min … max):    2.057 s …  2.167 s    10 runs
|
| Summary
|   LD_LIBRARY_PATH=. ./bench 1 dir-15000 dir-1000 dir-1 ran
|    84.48 ± 2.42 times faster than ./bench 1 dir-15000 dir-1000 dir-1

Signed-off-by: Clemens Lang <cllang@redhat.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25056)

Free fetched digest in show_digests

Fixes #24892

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Neil Horman <nhorman@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25046)

Update documentation

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/23900)

Call key_check_passed in set_ctx_params

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/23900)

Add FIPS indicator tests for KDFs

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/23900)

Restrict the length of key-derivation key used in KDFs

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/23900)

CMS_get1_crls(): Remove redundant check for NULL crls

Fixes Coverity 1610380

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/24928)

disable rwlocks on nonstop klt model

It appears nonstops new threading model defines some level of rwlock
pthread api, but its not working properly.  Disable rwlocks for
_KLT_MODEL_ for now

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/24969)

Add error checking to CRYPTO_atomic_[load|store] calls

Noted that we didn't check return codes of the atomic loads/stores in
the new hashtable, and they can fail

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/24969)

Make ossl_ht_delete use read-once semantics

To ensure that the value of h->md doesn't get recomputed during a delete
operation use ossl_rcu_deref on it

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/24969)

Fix CRYPTO_atomic_store

If the implementation of this function falls to using a pthread lock to
update a value, it should be a write lock, not a read lock

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/24969)

evp_get_digest/cipherbyname_ex(): Try to fetch if not found

If the name is not found in namemap, we need
to try to fetch the algorithm and query the
namemap again.

Fixes #19338

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/24940)

Avoid leaking *ba_ret on reconnections

Also fixes Coverity 1604639
There is no point in checking ba_ret as it can never be NULL.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/24931)

jitter: retry intermittent failures

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24844)

JITTER: implement error handling from jitter library

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24844)

JITTER: excercise all tests in CI with JITTER seed source under certain build configuration

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24844)

JITTER: add documentation

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24844)

jitter: add a new provider containing a jitter entropy source alone

This entropy source can be used instead of SEED-SRC. Sample
openssl.cnf configuration is provided. It is built as a separate
provider, because it is likely to require less frequent updates than
fips provider. The same build likely can span multiple generations of
FIPS 140 standard revisions.

Note that rand-instances currently chain from public/private instances
to primary, prior to consuming the seed. Thus currently a unique ESV
needs to be obtained, and resue of jitterentropy.a certificate is not
possible as is. Separately a patch will be sent to allow for
unchaining public/private RAND instances for the purpose of reusing
ESV.

Also I do wonder if it makes sense to create a fips variant of stock
SEED-SRC entropy source, which in addition to using getrandom() also
verifies that the kernel is operating in FIPS mode and thus is likely
a validated entropy source. As in on Linux, check that
/proc/sys/crypto/fips_enabled is set to 1, and similar checks on
Windows / MacOS and so on.

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/24844)

Add FIPS indicator to CMAC.

There is a issue currently related to CMAC TDES, when the new provider
is tested against older branches.

The new strict check caused backwards compatibility issues when
using old branch with the new FIPS provider.

To get around this CMAC now allows TDES by default, but it can be either
enabled via config or a settable. (i.e it uses an indicator)

Where the TDES cipher check can be done turned out to be problematic.
Shifting the check in the TDES cipherout of the init doesnt work because
ciphers can run thru either final or cipher (and checking on every
cipher call seemed bad). This means it needs to stay in the cipher init.
So the check needs to be done in CMAC BEFORE the underlying TDES cipher
does it check.
When using an indicator the TDES cipher needs its "encrypt-check" set
so that needs to be propagated from the CMAC object. This requires
the ability to set the param at the time the cipher ctx is inited.
An internal function was required in order to pass params to CMAC_Init.

Note also that the check was done where it is, because EVP_Q_mac() calls
EVP_MAC_CTX_set_params(ctx, cipher_param)
EVP_MAC_CTX_set_params(ctx, params)
EVP_MAC_init(ctx, key, keylen, params)
Where the second call to set_params would set up "encrypt-check" after
"cipher".

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25022)

Add negative tests for SHAKE with PBKDF2 and X9.42 KDF

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/24862)

Disallow SHAKE when using PBKDF2 and X9.42 KDF

The operation is non-sensical.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/24862)