QUIC RX: Fix QRX packet handling refactor w.r.t. list refactor

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19703)

QUIC RX: Support refcounted packets and eliminate wrapper

Previously, the QRX filled in a OSSL_QRX_PKT structure provided by the
caller. This necessitated the caller managing reference counting itself
using a OSSL_QRX_PKT_WRAP structure. The need for this structure has
been eliminated by adding refcounting support to the QRX itself. The QRX
now outputs a pointer to an OSSL_QRX_PKT instead of filling in a
structure provided by the caller. The OSSL_QRX_PKT_WRAP structure has
been eliminated.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19703)

QUIC: Dummy Handshake Layer for Prototyping

This disables -Wtype-limits /
-Wtautological-constant-out-of-range-compare. Since it generates
warnings for valid and reasonable code, IMO this actually encourages
people to write worse code.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19703)

QUIC DEMUX: Allow BIO to be changed

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19703)

QUIC Congestion Control: API to determine deadline at which more credit will be available

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19703)

QUIC ACKM: Add support for psuedo-loss

This is required to support retries during connection establishment.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19703)

Support all five EdDSA instances from RFC 8032

Fixes #6277

Description:
Make each of the five EdDSA instances defined in RFC 8032 -- Ed25519,
Ed25519ctx, Ed25519ph, Ed448, Ed448ph -- available via the EVP APIs.

The desired EdDSA instance is specified via an OSSL_PARAM.

All instances, except for Ed25519, allow context strings as input.
Context strings are passed via an OSSL_PARAM.  For Ed25519ctx, the
context string must be nonempty.

Ed25519, Ed25519ctx, Ed448 are PureEdDSA instances, which means that
the full message (not a digest) must be passed to sign and verify
operations.

Ed25519ph, Ed448ph are HashEdDSA instances, which means that the input
message is hashed before sign and verify.

Testing:
All 21 test vectors from RFC 8032 have been added to evppkey_ecx.txt
(thanks to Shane Lontis for showing how to do that).  Those 21 test
vectors are exercised by evp_test.c and cover all five instances.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/19705)

fips: make EdDSA unapproved for FIPS

Likewise for the related ECX key exchanges.

NIST is mandating this until FIPS 186-5 is finalised.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/20020)

Make RSA_generate_multi_prime_key() not segfault if e is NULL.

This is not a big problem for higher level keygen, as these set e
beforehand to a default value. But the logic at the lower level is
incorrect since it was doing a NULL check in one place but then
segfaulting during a later BN_copy().

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Todd Short <todd.short@me.com>
(Merged from https://github.com/openssl/openssl/pull/20025)

remove unused macro in rc2_local.h and rc5_local.h

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/20015)

SSKDF with KMAC should return SIZE_MAX when EVP_KDF_CTX_get_kdf_size()
is used.

Fixes #19934

The existing code was looking for the digest size, and then returned
zero.

The example code in EVP_KDF-SS.pod has been corrected to not use a
digest.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19935)

fix manpage of `d2i_X509(3)`

* capitalize `X509_NAME`
* add missing suffixes to `i2d_TYPE`

CLA: trivial

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/20016)

Do not check definition of a macro and use it in a single condition

The condition evaluation in #if conditions does not tolerate this
if the macro is not defined.

Fixes #19628

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/20018)

OSSL_PARAM_BLD and BIGNUM; ensure at least one byte is allocated

A zero BIGNUM contains zero bytes, while OSSL_PARAMs with an INTEGER (or
UNSIGNED INTEGER) data type are expected to have at least one data byte
allocated, containing a zero.  This wasn't handled correctly.

Fixes #20011

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/20013)

In OSSL_PARAM_set_BN(), make sure that the data_size field is at least 1

This way, we guarantee that a zero is represented with one byte of data
that's set to zero.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/20013)

test/param_build_test.c: test zero BIGNUM

We also add tests where the zero bignum is the only parameter, to test what
that does with the allocated blocks that the OSSL_PARAM_BLD functionality
handles.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/20013)

Add empty migration guide for 3.1

Fixes #19953

CLA: trivial

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/20008)

Doc: Update history section of EC_GROUP API's.

Fixes #8630

The remaining functions are at least as old as 0.9.8 so it is
not worth documenting this.

Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19956)

Documentation for EVP_PKEY_CTX_get0_pkey() and EVP_PKEY_CTX_get0_peerkey().

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19962)

info.c: Fix typos in seed macro name and description string

Fixes: #19996
CLA: trivial

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/20001)

rsaz_exp_x2.c: Remove unused ALIGN64 macro

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19597)

rsaz_exp_x2.c: Avoid potential undefined behavior with strict aliasing

Fixes #19584

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19597)

Revert "Fix an occasional CI failure due to unaligned access"

This reverts commit 8511520842b744d1794ea794c032ce5f78cd874b.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19597)

Fix SM4 test failures on big-endian ARM processors

Signed-off-by: Xu Yizhou <xuyizhou1@huawei.com>
Reviewed-by: Paul Yang <kaishen.yy@antfin.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19910)

remove extra define for __NR_getrandom and add some comments

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19985)

Write SSL_R alerts to error state to keep updated strings

Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19950)

80-test_cms.t: Fix rsapssSaltlen check on MinGW

Fixes #19907

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19957)

25-test_pkcs8.t: Make text comparison ignore extra CR characters

This is needed to pass the test on MinGW.

Fixes #19921

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19957)

Cleanse internal BN_generate_dsa_nonce() buffers used to generate k.

Fixes #9205

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19954)

Fixed typos in documentation and comments

Fixed typo: accomodate -> accommodate
Fixed typo: analagous -> analogous
Fixed typo: auxilliary -> auxiliary
Fixed typo: eigth -> eighth
Fixed typo: explotation -> exploitation
Fixed typo: originaly -> originally
Fixed typo: simplier -> simpler
Fixed typo: sucessful -> successful
Fixed typo: recievers -> receivers

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19977)

Bump actions/setup-python from 4.3.1 to 4.4.0

Bumps [actions/setup-python](https://github.com/actions/setup-python) from 4.3.1 to 4.4.0.
- [Release notes](https://github.com/actions/setup-python/releases)
- [Commits](https://github.com/actions/setup-python/compare/v4.3.1...v4.4.0)

---
updated-dependencies:
- dependency-name: actions/setup-python
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
CLA: trivial

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19961)

INSTALL.md: Remove trailing space

Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/19960)

Support multibin to allow multiple binary models to co-exist.

This change parallels the implementation of multilib and initially
only applies to the NonStop platform's DLL loader limitations.

Fixes: #16460
Signed-off-by: Randall S. Becker <rsbecker@nexbridge.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16472)

Docs: Move deprecated ECDSA_ functions into a separate file.

Fixes #19829

Examples added for setting/getting ECDSA SIG related r and s values

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19936)

Change HKDF to alloc the info buffer.

Fixes #19909

I have enforced a maximum bound still but it is much higher.
Note also that TLS13 still uses the 2048 buffer size.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19923)

ssl3_mac(): Fix possible divide by zero bug

CLA: trivial

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19912)

ec_kmgmt.c: check the return of BN_CTX_get() in time.

If x and y are all NULL, then it is unnecessary to do subsequent operations.

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19905)

INSTALL.md: Fix typo

CLA:trivial

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Yang <kaishen.yy@antfin.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19882)

Add Demos for DSA params/DSA keygen.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19853)

Fix possible UB in init_info_strings

"openssl version -c" may create undefined behavior in the shift:

crypto/info.c:42:50: runtime error: left shift of 4275712515
by 32 places cannot be represented in type 'long long int'

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19668)

Add testcase for OSSL_trace_set_callback()

Also test the OSSL_TRACE_CATEGORY_TRACE tracing - this fails
on address sanitizer runs without the fix for #19915

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/19917)

Avoid ifdefs in trace categories

The trace code assumes all categories are present and
the category numbers are equal to the index in the table.

Fixes #19915

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/19917)

Add a CMS test for a bad encryption algorithm

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19918)

Ensure ossl_cms_EncryptedContent_init_bio() reports an error on no OID

If the cipher being used in ossl_cms_EncryptedContent_init_bio() has no
associated OID then we should report an error rather than continuing on
regardless. Continuing on still ends up failing - but later on and with a
more cryptic error message.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19918)

Fix BIO_f_asn1() to properly report some errors

Some things that may go wrong in asn1_bio_write() are serious errors
that should be reported as -1, rather than 0 (which just means "we wrote
no data").

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19918)

Fix SMIME_crlf_copy() to properly report an error

If the BIO unexpectedly fails to flush then SMIME_crlf_copy() was not
correctly reporting the error. We modify it to properly propagate the
error condition.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19918)

Fix BIO_f_cipher() flushing

If an error occurs during a flush on a BIO_f_cipher() then in some cases
we could get into an infinite loop. We add a check to make sure we are
making progress during flush and exit if not.

This issue was reported by Octavio Galland who also demonstrated an
infinite loop in CMS encryption as a result of this bug.

The security team has assessed this issue as not a CVE. This occurs on
*encryption* only which is typically processing trusted data. We are not
aware of a way to trigger this with untrusted data.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19918)

Fix a logic flaw in test_mod_exp_zero

Due to the logic flaw, possible test failures
in this test case might be ignored.

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19929)

Add CHANGES.md entry for support for KMAC in KBKDF

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Release: yes
(Merged from https://github.com/openssl/openssl/pull/19943)

(cherry picked from commit c75203021bd37571c6259317b1d0e63dc7857c45)

19607 No need to link explicitly with libpthread on Solaris

CLA: trivial

Reviewed-by: Zdenek.Kotal@oracle.com
Reviewed-by: Ali.Bahrami@oracle.com
Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19611)

Update FIPS related build instructions.

This also links back to the new location that lists the cert and
security policy.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19617)

Refine the documents of several APIs

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19816)

Address coverity issue CID 1517105

The code path for this resource leak indicates that this is a false
positive (if you look at the callers).
Rather than ignoring the warning an extra check has been added, in case
future callers do the wrong thing.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19799)

Update HMAC() documentation.

Fixes #19782

Clarify that EVP_Q_MAC() can be used as an alternative that allows
setting of the libctx.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19855)

Update pyca-cryptography submodule to fix CI

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19926)

Add KMAC support to KBKDF.

Now that ACVP test vectors exist, support has been added for this mode.
See https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-108r1.pdf

Note that the test vectors used fairly large values for the input key
and the context, so the contraints for these has been increased from
256 to 512 bytes.

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19916)

timing_load_creds requires POSIX1.2001 due to rusage

Fixes #19838

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/19839)

BIO_s_dgram: add documentation and hazard warnings

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19795)

Coverity: fix 272011 resource leak

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/19900)

timing_load_creds: Fix typos in the timersub macro

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19865)

(cherry picked from commit f1b104953af8e3a82f7c5ee33f0403fc067c8516)

Do not build P10-specific AES-GCM assembler on AIX

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19865)

(cherry picked from commit 5c92ac52c282dbcb07e3880c2edd3a1bce452396)

Do not build P10-specific Chacha20 assembler on AIX

Fixes #18145

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19865)

(cherry picked from commit cdcc439aa0acba8a50b5e3444fb12d6e1157b991)

test: cmp_vfy_test: fix defined but unused

Building with

  ./config -DFUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION \
           -DPEDANTIC -Wall -Werror -pedantic

fails since the following test cases are excluded when
FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION is defined:

 - test_validate_msg_signature_srvcert_missing
 - test_validate_msg_mac_alg_protection_wrong
 - test_validate_msg_mac_alg_protection_missing

Guard the test cases by the corresponding preprocessor conditionals.

Signed-off-by: Čestmír Kalina <ckalina@redhat.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19868)

Fix openssl storeutl to allow serial + issuer

storeutl wants to enforce the use of issuer and serial together,
however the current code prevents to use them together and returns an
error if only one of them is specified.

Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19856)

rsa: fix merge CI runs with old FIPS providers

Since the fips provider version isn't frozen at 3.0.0, and the first
planned release with the fix in the fips provider is in 3.2.0,
we need to skip all the tests that expect implicit rejection
in all versions below 3.2.0

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19890)

rsa: fix version of rsa implicit rejection introduction

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19889)

Allow OBJ_create() to create an OBJ and NID with a NULL OID

We already permit this in crypto/objects/objects.txt, but not programatically,
although being able to do so programatically would be beneficial.

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19876)

Make error reason for disallowed legacy sigalg more specific

The internal error reason is confusing and indicating an error
in OpenSSL and not a configuration problem.

Fixes #19867

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19875)

Fix SM4-CBC regression on Armv8

Fixes #19858

During decryption, the last ciphertext is not fed to next block
correctly when the number of input blocks is exactly 4. Fix this
and add the corresponding test cases.

Thanks xu-yi-zhou for reporting this issue and proposing the fix.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19872)

Fix-up to f3090fc7

Define OSSL_SIGNATURE_PARAM_NONCE_TYPE as "nonce-type" (rather than
"nonce_type") so that it is consistent with the documentation.

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19883)

crypto/err: expand on error code generation

Signed-off-by: Čestmír Kalina <ckalina@redhat.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19823)

rsa: add implicit rejection CHANGES entry

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/13817)

smime/pkcs7: disable the Bleichenbacher workaround

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/13817)

rsa: add test for the option to disable implicit rejection

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/13817)

rsa: Skip the synthethic plaintext test with old FIPS provider

since the 3.0.0 FIPS provider doesn't implement the Bleichenbacher
workaround, the decryption fails instead of providing a synthetic
plaintext, so skip them then

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/13817)

rsa: Add option to disable implicit rejection

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/13817)

rsa: add test vectors for the implicit rejection in RSA PKCS#1 v1.5

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/13817)

rsa: add implicit rejection in PKCS#1 v1.5

The RSA decryption as implemented before required very careful handling
of both the exit code returned by OpenSSL and the potentially returned
ciphertext. Looking at the recent security vulnerabilities
(CVE-2020-25659 and CVE-2020-25657) it is unlikely that most users of
OpenSSL do it correctly.

Given that correct code requires side channel secure programming in
application code, we can classify the existing RSA decryption methods
as CWE-676, which in turn likely causes CWE-208 and CWE-385 in
application code.

To prevent that, we can use a technique called "implicit rejection".
For that we generate a random message to be returned in case the
padding check fails. We generate the message based on static secret
data (the private exponent) and the provided ciphertext (so that the
attacker cannot determine that the returned value is randomly generated
instead of result of decryption and de-padding). We return it in case
any part of padding check fails.

The upshot of this approach is that then not only is the length of the
returned message useless as the Bleichenbacher oracle, so are the
actual bytes of the returned message. So application code doesn't have
to perform any operations on the returned message in side-channel free
way to remain secure against Bleichenbacher attacks.

Note: this patch implements a specific algorithm, shared with Mozilla
NSS, so that the attacker cannot use one library as an oracle against the
other in heterogeneous environments.

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/13817)

Run-checker merge CI: Memleak test does not work without ubsan

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19874)

Revert "Run-checker merge CI: Replace no-shared with no-modules"

This reverts commit d5696547e46e9ea85fcb7581b9d49c58b7c24eeb.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19874)

unbuffer stdin before get passwd from stdin

commond LD_LIBRARY_PATH= openssl rsa -aes256 -passout stdin <<< "xxxxxx” will get pass(fun app_get_pass()) from stdin first, and then load key(fun load_key()). but it unbuffer stdin before load key, this will cause the load key to fail.

now unbuffer stdin before get pass, this will solve https://github.com/openssl/openssl/issues/19835

CLA: trivial

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19851)

Run-checker merge CI: Replace no-shared with no-modules

ASAN otherwise fails to detect memleaks.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19860)

Cross compiles CI: Disable stringop-overflow warning on s390x and m68k

These warnings trigger on false positives on these platforms
with recent compiler update.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19860)

Fuzz checker CI: Use more generic include dir for fuzzer includes

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19860)

Bump actions/setup-python from 4.3.0 to 4.3.1

Bumps [actions/setup-python](https://github.com/actions/setup-python) from 4.3.0 to 4.3.1.
- [Release notes](https://github.com/actions/setup-python/releases)
- [Commits](https://github.com/actions/setup-python/compare/v4.3.0...v4.3.1)

---
updated-dependencies:
- dependency-name: actions/setup-python
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
CLA: trivial

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19866)

Fix `no-ec enable-ktls` build

The KTLS test uses a TLSv1.2 cipher that uses ECDHE

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19841)

test: add test case for deadlock reported in #19643

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19652)

x509: fix double locking problem

This reverts commit 9aa4be691f5c73eb3c68606d824c104550c053f7 and removed the
redundant flag setting.

Fixes #19643

Fixes LOW CVE-2022-3996

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19652)

signature: Clamp PSS salt len to MD len

FIPS 186-4 section 5 "The RSA Digital Signature Algorithm", subsection
5.5 "PKCS #1" says: "For RSASSA-PSS […] the length (in bytes) of the
salt (sLen) shall satisfy 0 <= sLen <= hLen, where hLen is the length of
the hash function output block (in bytes)."

Introduce a new option RSA_PSS_SALTLEN_AUTO_DIGEST_MAX and make it the
default. The new value will behave like RSA_PSS_SALTLEN_AUTO, but will
not use more than the digest length when signing, so that FIPS 186-4 is
not violated. This value has two advantages when compared with
RSA_PSS_SALTLEN_DIGEST: (1) It will continue to do auto-detection when
verifying signatures for maximum compatibility, where
RSA_PSS_SALTLEN_DIGEST would fail for other digest sizes. (2) It will
work for combinations where the maximum salt length is smaller than the
digest size, which typically happens with large digest sizes (e.g.,
SHA-512) and small RSA keys.

J.-S. Coron shows in "Optimal Security Proofs for PSS and Other
Signature Schemes. Advances in Cryptology – Eurocrypt 2002, volume 2332
of Lecture Notes in Computer Science, pp. 272 – 287. Springer Verlag,
2002." that longer salts than the output size of modern hash functions
do not increase security: "For example,for an application in which at
most one billion signatures will be generated, k0 = 30 bits of random
salt are actually sufficient to guarantee the same level of security as
RSA, and taking a larger salt does not increase the security level."

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/19724)

Obtain PSS salt length from provider

Rather than computing the PSS salt length again in core using
ossl_rsa_ctx_to_pss_string, which calls rsa_ctx_to_pss and computes the
salt length, obtain it from the provider using the
OSSL_SIGNATURE_PARAM_ALGORITHM_ID param to handle the case where the
interpretation of the magic constants in the provider differs from that
of OpenSSL core.

Add tests that verify that the rsa_pss_saltlen:max,
rsa_pss_saltlen:<integer> and rsa_pss_saltlen:digest options work and
put the computed digest length into the CMS_ContentInfo struct when
using CMS. Do not add a test for the salt length generated by a provider
when no specific rsa_pss_saltlen option is defined, since that number
could change between providers and provider versions, and we want to
preserve compatibility with older providers.

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/19724)

prevent HPKE sender setting seq unwisely

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19840)

OSSL_CMP_validate_msg(): make sure to reject protection type mismatch

Do not accept password-based if expected signature-based and no secret is available and
do not accept signature-based if expected password-based and no trust anchors available.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/19729)

Replace some boldened types with a corresponding man page link

The types OSSL_DISPATCH, OSSL_ITEM, OSSL_ALGORITHM, OSSL_PARAM,
OSSL_CALLBACK, and OSSL_PASSPHRASE_CALLBACK are described in their own
manual page, so we change every mention of them to links to those pages.

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19842)

Move the description of the core types into their own pages

This expands on some of the core type descriptions, and also makes it
easier to find the documentation for each type, at least on Unix, with
a simple call like "man OSSL_ALGORITHM".

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19842)

Better sorting of util/other.syms

Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19842)

Fix treatment of BUILD_METADATA

According to documentation [^1], the BUILD_METADATA from VERSION.dat should
be prefixed with a plus sign when used.  It is given this treatment in
Configure, but not in all other scripts that use VERSION.dat directly.
This change fixes that.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Todd Short <todd.short@me.com>
(Merged from https://github.com/openssl/openssl/pull/19815)

Compensate for CMP-related TODOs removed by PR #15539

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
Reviewed-by: David von Oheimb <david.von.oheimb@siemens.com>
(Merged from https://github.com/openssl/openssl/pull/16006)

Replace "a RSA" with "an RSA"

Fixes openssl#19771

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19787)

Implement OSSL_PROVIDER_get0_default_search_path, add docs and tests.

Reviewed-by: Todd Short <todd.short@me.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19752)

timing_load_creds: Add timersub macro for platforms where it is missing

Fixes #19812

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Hugo Landau <hlandau@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/19817)