EVP_PKEY_keygen_init: fix return check

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/17031)

EVP_PKEY_paramgen_init: fix return check

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/17030)

EVP_DigestVerifyFinal: fix test function and invocation

Signed-off-by: Peiwei Hu <jlu.hpw@foxmail.com>
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/17029)

EVP_Cipher: fix the incomplete return check

Signed-off-by: Peiwei Hu <jlu.hpw@foxmail.com>
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/17027)

Add RSA encrypt demo

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16283)

Abstract out policy and extensions in CA.pl

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/13819)

Add a test for creating ECX private keys that are too short

We expect attempting to create such short keys to fail

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17041)

Don't create an ECX key with short keys

If an ECX key is created and the private key is too short, a fromdata
call would create the key, and then later detect the error and report it
after freeing the key. However freeing the key was calling
OPENSSL_secure_clear_free() and assuming that the private key was of the
correct length. If it was actually too short this will write over memory
that it shouldn't.

Fixes #17017

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17041)

RAND_bytes_ex: fix return check

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17032)

Clarify the PEM docs

Make it clear how the existing PEM functions can be used to create an
X509, X509_REQ or X509_CRL object with an associated libctx/propq.

Fixes #16966

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17012)

Reset the rwstate before calling ASYNC_start_job()

If an async job pauses while processing a TLS connection then the
rwstate gets set to SSL_ASYNC_PAUSED. When resuming the job we should
reset the rwstate back to SSL_NOTHING. In fact we can do this
unconditionally since if we're about to call ASYNC_start_job() then either
we are about to start the async job for the first time (in which case the
rwstate should already by SSL_NOTHING), or we are restarting it after a
pause (in which case reseting it to SSL_NOTHING is the correct action).

Fixes #16809

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17013)

BIO_s_connect(): Enable BIO_gets()

Fixes #16028

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16030)

Add null digest implementation to the default provider

This is necessary to keep compatibility with 1.1.1.

Fixes #16660

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17016)

Extend the test_multi_load() test

Run more threads and load the legacy provider (which uses a child lib ctx)
in order to hit more possible thread failures.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Hold the flag_lock when calling child callbacks

Not holding the flag lock when creating/removing child providers can
confuse the activation counts if the parent provider is loaded/unloaded
at the same time.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Use a write lock during ossl_provider_find()

A "find" operation on a stack can end up sorting the underlying stack. In
this case it is necessary to use a "write" lock to synchronise access to
the stack across multiple threads.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Correctly activate the provider in OSSL_PROVIDER_try_load

If during OSSL_PROVIDER_try_load() we attempt to load a provider, but
adding to the store gives back a different provider, then we need to
ensure this different provider has its activation count increased.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Stop receiving child callbacks in a child libctx when appropriate

We should stop receiving child callbacks if we're about to free up
the child libctx. Otherwise we can get callbacks when the libctx is half
freed up.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Don't bail out during provider deactivation if we don't have store

A provider may have been activated, but failed when being added to
the store. At this point we still need to deactivate it.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Don't try and do ossl_provider_find in ossl_provider_new

We leave it to the caller to confirm that the provider does not exist
in the store. If it does exist then later adding it to the store will
fail.

It is possible that the provider could be added to the store in
between the caller checking, and the caller calling ossl_provider_new.
We leave it to the caller to properly handle the failure when it
attempts to add the provider to the store. This is simpler than
having ossl_provider_new try to handle it.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Remove the isinited variable from child_prov_globals

This variable might have made sense at some point but it not longer does
so. It was being used to check whether we are still initing or not. If we
are still initing then the assumption was that we already hold the lock.
That assumption was untrue. We need to always take the lock.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Avoid a race in init_thread_stop()

init_thread_stop() is called when a thread is stopping. It calls all
the callbacks that need to know about the demise of this thread. However,
the list of callbacks is also available globally and may be updated by
other threads so we need to make sure we use the right lock.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Don't attempt to deactive child providers if we don't need to

If a provider doesn't have any child providers then there is no need
to attempt to remove them - so we should not do so. This removes some
potentialy thread races.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

Don't write to the globals ossl_property_true and ossl_property_false

These global variables were previously overwritten with the same value
every time we created a new OSSL_LIB_CTX. Instead we preinitialise them
with the correct values, and then confirm that settings for each
OSSL_LIB_CTX agree with the preinitialised values.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16980)

doc: Document outcome of multiple digestsign/digestverify calls

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16964)

evp_extra_test: Add SIPHASH MAC digestsign test with reinitialization

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16964)

providers: Allow possible reinitialization in all signature algorithms

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16964)

test: Add testing of reinitialization via EVP_DigestSignInit()

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16964)

do_sigver_init: Allow reinitialization of an existing operation.

Fixes #16936

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16964)

Fix return value checking of BN_check_prime invocations

Negative return value indicates an error so we bail out.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16975)

Add return value NULL checks that were missing

Issues located by Brian Carpenter of Geeknik's Farm.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/17007)

x509: use safe maths calls for overflow detection

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16930)

run-checker: add CI to test safe_math without compiler support.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16930)

stack: increase the reallocation ratio

This change increases the reallocation ratio from 1.5 to 1.6.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16930)

test: add unit tests for integer overflow helpers

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16930)

doc: document the internal integer overflow helpers

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16930)

header: add integer overflow helper functions

Define a number of helper functions that ease the difficulty of detecting
integer overflows.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16930)

25-test_req.t: Add systematic SKID+AKID tests for self-issued (incl. self-signed) certs

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/16342)

X509: Fix handling of AKID and SKID extensions according to configuration

Fixes #16300

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/16342)

Add OID for RPKI id-ct-ASPA

References: draft-ietf-sidrops-aspa-profile
"A Profile for Autonomous System Provider Authorization" (ASPA)

OID permanently assigned under 'SMI Security for S/MIME CMS Content Type (1.2.840.113549.1.9.16.1)'
https://www.iana.org/assignments/smi-numbers/smi-numbers.xhtml#security-smime-1

CLA: trivial

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/17002)

sha/asm/keccak1600-ppc64.pl: Load data in 8 byte chunks on little endian

We currently load data byte by byte in order to byteswap it on big
endian. On little endian we can just do 8 byte loads.

A SHAKE128 benchmark runs 10% faster on POWER9 with this patch applied.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8455)

prov: remove unused field `flag_fallback` and function `ossl_provider_set_fallback`

These are legacy of older versions of the code and are currently not used
anywhere.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16985)

Fix sigsize usage in apps/speed.c

In a recent upstream change
(43da9a14f0e73f42f28ae34219929b44df5d1a11)
the parameter sigsize become a read/write input in
EVP_PKEY_sign(), and after signing, sigsize will be overwritten with
the actual size and used in the verify step. As the speed program
calls EVP_PKEY_sign() on the same context repeatedly, sigsize value is
no longer the initial available size, and may fail in later buffer
size checks.

This fix adds a new buflen member in struct loopargs (which is only
used within apps/speed.c), to save available buffer size and
to be used as sigsize input in EVP_PKEY_sign() calls.
Sigsize still contains the signature size for the verify step.

Signed-off-by: Xiaofei Bai <xiaofei.bai@arm.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16991)

x509: remove dead call to strlen()

The condition `userlen == -1` isn't possible because this is already checked
on line 159 above and the subsequent strlen(3) call guarantees that it's value
is positive.

Reviewed-by: Paul Yang <kaishen.yy@antfin.com>
(Merged from https://github.com/openssl/openssl/pull/16987)

Fix: invoking X509_self_signed improperly

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16976)

coverity: add a daily coverity build

The weekly build got lost when we stopped using Travis.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16988)

Fix DER encoder implementations for output structures "EC" and "SM2"

These DER encoder implementations are supposed to be aliases for the
"type-specific" output structure, but were made different in so far
that they would output a "type specific" public key, which turns out
to be garbage (it called i2o_ECPublicKey()).  The "type-specific"
output structure doesn't support that, and shouldn't.

Fixes #16977

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16983)

(cherry picked from commit 2cb802e16fff3fb2c57ae664baa7bd9ce3e33805)

Fix: invoking x509_name_cannon improperly

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16974)

KTLS: use EVP_CIPHER_is_a instead of nid

Signed-off-by: Tianjia Zhang <tianjia.zhang@linux.alibaba.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Yang <kaishen.yy@antfin.com>
(Merged from https://github.com/openssl/openssl/pull/16963)

APPS/x509: Fix generation of AKID via v2i_AUTHORITY_KEYID()

Fixes #16300

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16442)

APPS/cmp: make the -sans option support email addresses (type rfc822Name)

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16960)

Fix errors in EVP_PKEY_fromdata examples

The EVP_PKEY_fromdata man page has some code examples with various
errors in them. This fixes those errors.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16973)

Fix data race setting `default_DSO_meth`

The global variable `default_DSO_meth` was potentially set multiple times by
different threads.  It turns out that it could only be set to a single value
so the race is harmless but still better avoided.  The fix here simply removes
the global and accesses the value it was set to via the `DSO_METHOD_openssl()`
call.

Problem discovered via #16970, but this does not resolve that issue because
there are other concerns.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16972)

Address Coverity 1493362 resource leak

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16962)

Address coverity 1493382 argument cannot be negative

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16962)

Address Coverity 1493387 Logically dead code

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16962)

Fix coverity 1493364 & 1493375: unchecked return value

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16962)

Fix incorrect return check of BN_bn2binpad

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16942)

avoid a NULL dereference when getting digest

Fixes #16961

Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/16969)

Fix tests to check for negative results when calling EVP_PKEY_fromdata_init

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16801)

Fix for a segv interrupt that occurs when fix_dh_rfc5114 is called with
ctx->p2 being a null pointer.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16913)

(cherry picked from commit 07e6c857364770f6e986b5d8ceb9fbe296f3c6d0)

Add unit tests for weak key and key parity checks

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/16953)

Convert the weak key and key parity tests to be constant time.

Fixes #16944
Fixes #16859

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/16953)

Remove redundant tests

Removed the three checks of type against NULL since type cannot be NULL for
any of them.

Moved a check of ->engine inside a CPP guard for engines.

Didn't address the teardown and rebuild of the provider context.

Partially fixes #16947

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16952)

Fix a memory leak in ssl_create_cipher_list

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16954)

Fix a memory leak in tls_parse_stoc_key_share

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16956)

Fix builds on Armv8 systems without AArch64

This fixes "undefined reference to `aes_gcm_dec_128_kernel' in function
`armv8_aes_gcm_decrypt'" and similar

Fixes #16949

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16951)

check the return value of BN_new() and BN_dup()

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16948)

DES_set_key(): return values as DES_set_key_checked() but always set

This avoids using accidentally uninitialized key schedule in
applications that use DES_set_key() not expecting it to check the key
which is the default on OpenSSL <= 1.1.1

Fixes #16859

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16944)

Revise s_client and s_server verbiage re secure renegotiation.

Since TLS v1.3 eschews renegotiation entirely it’s misleading to have
these apps say it’s “not supported” when in fact the TLS version is
new enough not to need renegotiation at all.

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
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/16937)

Add missing check according to SM2 Digital Signature generation algorithm

The process should be conforming to clause 6.1 and 6.2 of GMT 0003.2-2012.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16931)

update doc: BN_bn2lebinpad() and BN_bn2nativepad()

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16943)

Fix incorrect return check of BN_bn2nativepad

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16943)

Remove redundant RAND_get0_private() call

The test called this twice which doesn't hurt but isn't ideal.

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/16921)

Fix return value error in doc, and an error test

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16889)

[edited to remove end of line whitespace and wrap lines to eighty columns]

fix some code with obvious wrong coding style

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16918)

Configurations/windows-makefile.tmpl: obj2bin(): use the resource file too

When remaking how programs were linked, the variable `$ress` was forgotten.
Unfortunately, perl treats this with silence.

Fixes #16870
Fixes #16667

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16875)

Fix a gcc 11.2.0 warning

gcc 11.2.0 is the default on Ubuntu 21.10. It emits a (spurious) warning
when compiling test/packettest.c, which causes --strict-warnings builds
to fail. A simple fix avoids the warning.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/16887)

Clarify the documentation for the "byname" functions

Make it clear that the cipher/digest objects returned from
EVP_get_cipherbyname() and EVP_get_digestbyname() functions have no
associated implementation fetched from a provider.

Fixes #16864

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16893)

test: fetching proper signature provider for non-exportable keys

Co-author: Selva Nair <selva.nair@gmail.com>

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16725)

CORE: Encure that cached fetches can be done per provider

This mostly entails passing around a provider pointer, and handling
queries that includes a pointer to a provider, where NULL means "any".

This also means that there's a need to pass the provider pointer, not
just down to the cache functions, but also be able to get it from
ossl_method_store_fetch().  To this end, that function's OSSL_PROVIDER
pointer argument is modified to be a pointer reference, so the
function can answer back what provider the method comes from.

Test added.

Fixes #16614

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16725)

EVP: For all operations that use an EVP_PKEY, check that there is one

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16725)

EVP: Allow a fallback for operations that work with an EVP_PKEY

Functions like EVP_PKEY_sign_init() do an implicit fetch of the
operation implementation (EVP_SIGNATURE in this case), then get the
KEYMGMT from the same provider, and tries to export the key there if
necessary.

If an export of the key isn't possible (because the provider that
holds the key is an HSM and therefore can't export), we would simply
fail without looking any further.

This change modifies the behaviour a bit by trying a second fetch of
the operation implementation, but specifically from the provider of
the EVP_PKEY that's being used.  This is done with the same properties
that were used with the initial operation implementation fetch, and
should therefore be safe, allowing only what those properties allow.

Fixes #16614

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16725)

EVP: Add internal functions to fetch type specific EVP methods from provider

Added functions:

evp_signature_fetch_from_prov(), evp_asym_cipher_fetch_from_prov(),
evp_keyexch_fetch_from_prov(), evp_kem_fetch_from_prov()

These are all like the public conterparts, except they all take a
provider instead of a library context as first argument.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16725)

EVP: Reverse the fetch logic in all pkey using functionality

In all initializing functions for functionality that use an EVP_PKEY, the
coded logic was to find an KEYMGMT implementation first, and then try to
find the operation method (for example, SIGNATURE implementation) in the
same provider.

This implies that in providers where there is a KEYMGMT implementation,
there must also be a SIGNATURE implementation, along with a KEYEXCH,
ASYM_CIPHER, etc implementation.

The intended design was, however, the opposite implication, i.e. that
where there is a SIGNATURE implementation, there must also be KEYMGMT.

This change reverses the logic of the code to be closer to the intended
design.

There is a consequence; we now use the query_operation_name function from
the KEYMGMT of the EVP_PKEY given by the EVP_PKEY_CTX (ultimately given by
the application).  Previously, we used the query_operation_name function
from the KEYMGMT found alongside the SIGNATURE implementation.

Another minor consequence is that the |keymgmt| field in EVP_PKEY_CTX
is now always a reference to the KEYMGMT of the |pkey| field if that
one is given (|pkey| isn't NULL) and is provided (|pkey->keymgmt|
isn't NULL).

Fixes #16614

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16725)

EVP: Add evp_keymgmt_fetch_from_prov()

This is an internal function to fetch a keymgmt method from a specific
provider.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16725)

EVP: Add the internal function evp_generic_fetch_from_prov()

This function leverages the generic possibility to fetch EVP methods
from a specific provider.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16725)

CORE: add a provider argument to ossl_method_construct()

This makes it possible to limit the search of methods to that
particular provider.  This uses already available possibilities in
ossl_algorithm_do_all().

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16725)

test/ssl_old_test.c: Do NULL pointer check before its use

In openssl-3.0.0 and system provided, it is not reasonable to
check null pointer after use. The order was accidentally reversed.
Therefore, it is better to correct it.

CLA: trivial

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16915)

free the Post-Handshake Auth digest when there is an error saving the digest

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16917)

Remove unused variable 'sctx'

In the openssl-3.0.0 and system provided, the variable 'sctx' is unused in test_509_dup_cert.
Therefore, it might be better to remove the definition and operation of it.

CLA: trivial

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16916)

Test that a key is usable after an EVP_PKEY_fromdata call

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16911)

Don't crash encoding a public key with no public key value

If asked to encode an EC_KEY public key, but no public key value is present
in the structure, we should fail rather than crash.

Fixes the crash seen here:
https://mta.openssl.org/pipermail/openssl-users/2021-October/014479.html

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16911)

add checks for the return values of BN_new(), sk_RSA_PRIME_INFO_new_reserve(),
EVP_PKEY_CTX_new_from_pkey() and EVP_CIPHER_CTX_new().
Otherwise may result in memory errors.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16892)

speed: range check the argument given to -multi

For machines where sizeof(size_t) == sizeof(int) there is a possible overflow
which could cause a crash.

For machines where sizeof(size_t) > sizeof(int), the existing checks adequately
detect the situation.

Fixes #16899

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16904)

Doc: replace `NULL` terminated with `NUL`

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16885)

doc: Fix some function signature errors

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16897)

test-rand: return failure on not enough data, allow parent

The test-rand RNG was returning success when it had some but insufficient data.
Now, it returns failure and doesn't advance the data pointer.

The test-rand RNG was failing when a parent was specified.  This case is now
ignored.

Fixes #16785

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16905)

Note that SHA1 and MD5 x509 signatures are also forbidden at security level 1

The exclusion of SHA1 for X509 signatures is not obvious as the "intuative"
idea is that SHA1 should have 80 security bits. However the security bits
of SHA1 are explicitly set to 63 to avoid the it being strong enough for
security level 1. x509_set.c has the comment:

    /*
     * SHA1 and MD5 are known to be broken. Reduce security bits so that
     * they're no longer accepted at security level 1.
     * The real values don't really matter as long as they're lower than 80,
     * which is our security level 1.
     */

Signed-off-by: Arne Schwabe <arne@rfc2549.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16895)

migration_guide: Mention ERR_GET_FUNC() and function code removal

Fixes #16817

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/16879)