test: the attributeDescriptor X.509v3 extension

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25429)

doc: the attributeDescriptor X.509v3 extension

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25429)

feat: support the attributeDescriptor X.509v3 extension

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25429)

txp_generate_stream_frames(): Set stream id in header early enough

Otherwise we will calculate an incorrect header
size for higher stream ids and won't fit the
frame into the packet.

Fixes #25417

Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25928)

APPS/load_key_certs_crls(): refactor to clean up the code a little and add clarifying comments

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
(Merged from https://github.com/openssl/openssl/pull/22528)

openssl-pkeyutl.pod.in: improve description of -rawin and -digest options

Fixes #25827

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25831)

openssl-pkeyutl.pod.in: add that -sign is default op, update claim on hash needed for -sign/-verify, etc.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25831)

APPS/pkeyutl: add missing high-level check for -verifyrecover being usable only with RSA

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25903)

APPS/pkeyutl: remove wrong check for -verifyrecover regarding too long sign/verify input

Fixed #25898

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25903)

Add a test for setting TLSv1.2 ciphersuites on a QUIC object

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25886)

Don't complain with "no cipher match" for QUIC objects

Calling the functions SSL_CTX_set_cipher_list() or SSL_set_cipher_list() will
return the error "no cipher match" if no TLSv1.2 (or below) ciphers are enabled
after calling them. However this is normal behaviour for QUIC objects which do
not support TLSv1.2 ciphers. Therefore we should suppress that error in this
case.

Fixes #25878

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25886)

Add a test for the new_session_cb from a QUIC object

Setting a new_session_cb should work for a QUIC object just as it does
with a normal TLS object.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25874)

Make sure we use the correct SSL object when making a callback

When processing a callback within libssl that applies to TLS the original
SSL object may have been created for TLS directly, or for QUIC. When making
the callback we must make sure that we use the correct SSL object. In the
case of QUIC we must not use the internal only SSL object.

Fixes #25788

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25874)

Keep hold of a reference to the user SSL in QUIC

In some cases a QUIC SSL_CONNECTION object needs to get hold of a reference
to the original SSL object as created by the user. We should keep a
reference to it.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25874)

Bugfixes for params to legacy control translations for EC parameters

param->ctrl translation: Fix fix_ecdh_cofactor()

In POST_PARAMS_TO_CTRL state the fix_ecdh_cofactor() function should
return value in ctx->p1

param->ctrl translation: fix evp_pkey_ctx_setget_params_to_ctrl
 return

Since some of the ctrl operations may return 0 as valid value
(e.g. ecdh_cofactor value 0 is valid setting), before colling
POST_PARAMS_TO_CTRL, we need to check return value for 0 as well
otherwise the evp_pkey_ctx_setget_params_to_ctrl function fails
without a chance to fix the return value

param->ctrl translation: Set ecdh_cofactor default action_type GET

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/22587)

Adjust naming authority formatting when printing out admission extension

Indent namingAuthority section with two spaces to match the parent
node.

Signed-off-by: oleg.hoefling <oleg.hoefling@gmail.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25814)

Fix memory leak on failure in copy_issuer()

When sk_GENERAL_NAME_reserve() fails, ialt is not freed.
Add the freeing operation in the common error path.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25876)

Remove unnecessary sk_GENERAL_NAME_free() calls on NULL

There are several calls to sk_GENERAL_NAME_free() where the argument is
actually NULL, there are not necessary.

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25877)

x509: add a newline after printing Full Name

We forget it in 58301e24f66aa74b13b85a171dd14e6088c35662.

Fixes #25853

CLA: trivial

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25854)

interop-tests.yml: Update to Fedora 40 and fix provisioning breakage

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25847)

Fix uses of `EVP_PKEY_Q_keygen` with `size_t` variadic argument

Fix cases where `int` argument was passed instead of `size_t`.

CLA: trivial

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25857)

Fix the default_md example in the ca docs

We should not have an example showing the default_md as md5.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25856)

Document extensibility of -groups parameter via providers

Also add brainpool curves

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25821)

Remove two unused union members from struct x509_object_st.

CLA: trivial

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25826)

Use correct alerts for some cert comp errors

Fixes #25471

Signed-off-by: Todd Short <todd.short@me.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25845)

Support DEFAULT keyword and '-' prefix in SSL_CTX_set1_groups_list()

Fixes #25790

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25813)

APPS/pkeyutl: strengthen error message on too long sign/verify input

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/22910)

APPS/pkeyutl: -digest implies -rawin and can only be used with -sign and -verify

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/22910)

APPS/pkeyutl: improve -rawin usability (implied by Ed25519 and Ed448) and doc

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/22910)

CHANGES.md: re-word 'app' to 'command' and 'apps' to 'commands' for consistency

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/22910)

apps/speed.c: Fix the benchmarking for AEAD ciphers

Fixed the benchmarking for the evp aead interface for ccm, gcm, ocb, and siv,
where decryption fails when executing
`openssl speed -evp aes-128-ccm -decrypt` and
`openssl speed -evp aes-128-gcm -decrypt`.

Related issues are [24686](https://github.com/openssl/openssl/issues/24686)
and [24250](https://github.com/openssl/openssl/issues/24250).
Now both encryption and decryption, with or without AAD, executes correctly
without issues.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25823)

apps: Change default cipher to aes-256-cbc for req, cms and smime apps

Update `CHANGES.md` and `NEWS.md`; remove `no-des` guard from req, cms,
and smime apps

Update MAN pages for default cipher; fix styling by removing braces around single statements

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25839)

Fix incorrect openssl-smime doc sample command for encrypt

Original documented sample command causes error. PEM recipient cert argument needs to go last.

CLA: trivial

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25829)

speed.c: Check for 0 block size

Although this cannot really happen check for 0 block size
to avoid division by 0.

Fixes Coverity 1633936

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25822)

Fix potential memory leak in OSSL_HPKE_CTX_new()

ctx->propq is a duplicated string, but the error code does not free
the duplicated string's memory. If e.g. EVP_CIPHER_fetch() fails then
we can leak the string's memory.

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25812)

test: Add test cases for req -cipher option

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25796)

req: Add -cipher option to specify private key encryption cipher

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25796)

Remove sslkeylog file setup from quic-hq-interop

Now that libcrypto supports the user of SSLKEYLOGFILE, the interop demo
attempts to open the same file based on the same env variable.

The hq-interop-demo code can just be removed, and it fixes the open
failure when both libcrypto and hq-interop attempt to open and write the
same file, which is causing the nightly failure

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25819)

fix: util/check-format-commit.sh - fix ending check

Look at the end result instead of the file name it's stored in

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25810)

output negotiated TLS1.3 group

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25808)

speed.c: Check block size before running EVP_Cipher_loop()

Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Todd Short <todd.short@me.com>
(Merged from https://github.com/openssl/openssl/pull/25792)

test_speed: Explicitly test the crashing command line on sparc

Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Todd Short <todd.short@me.com>
(Merged from https://github.com/openssl/openssl/pull/25792)

Fix builds on riscv64 using musl

Some environments using musl are reported to have the hwprobe.h include
file but not have the __NR_riscv_hwprobe define.

Fixes #25772

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Todd Short <todd.short@me.com>
(Merged from https://github.com/openssl/openssl/pull/25787)

Fix ambiguous output of Signature Algorithms

Signature Algorithms are printed in a SIG+HASH format.
In some cases this is ambiguous like brainpool and RSA-PSS.
And the name of ed25519 and ed448 must be spelled in lower case,
so that the output can be used as a -sigalgs parameter value.

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25797)

check-format-commit.sh: various improvements; check unstaged changes on empty commit range

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25666)

check-format-commit.sh: make awk patterns more robust for MacOS and certain flavors of Linux

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25666)

Dependabot update

CLA: trivial

(deps): Bump actions/setup-python

Bumps [actions/setup-python](https://github.com/actions/setup-python) from 5.2.0 to 5.3.0.
- [Release notes](https://github.com/actions/setup-python/releases)
- [Commits](https://github.com/actions/setup-python/compare/v5.2.0...v5.3.0)

---
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>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25795)

github: add fips configuration, with legacy transitions turned off

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25762)

fips no-des: compile out TDES KAT

FIPS provider correctly supports no-des build time option and doesn't
advertise DES related algorithms. However KAT test for DES is still
attempted to be executed and fails.

This prevents configuring FIPS provider without legacy behaviour as
defined in SP 800-131Arev2. Also see #25761 internal docs.

Fix `enable-fips no-des` build option, and add a daily checker for
"legacy-free" (as much as currently feasible) FIPS configuration.

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25762)

tls_common.c: Align the calculation of maximal alignment value

In tls_setup_write_buffer() and tls_setup_read_buffer() the calculation
is different. Make them the same.

Fixes #25746

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25764)

Do not confuse TAP::Parser by mixing up stderr with stdout.

This avoids false psotivie failures on FreeBSD-CI which
suffers most from this issue.

Fixes #23992

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/25613)

Dependabot: no longer set an "approval: otc review pending" label

The label doesn't exist anymore.

Reviewed-by: Paul Dale <ppzgs1@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/25706)

Fix missing sendmmsg/recvmmsg on AIX

This at least fixes the build failures on AIX

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25704)

TRACE: automatically respect disabled categories

by fixing OSSL_trace_begin() to return NULL when given category is not enabled

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25652)

Fix memory leaks in ossl_quic_calculate_retry_integrity_tag()

Fixes #25625

Several error paths return 0 directly instead of going to err to clean
up the objects.

CLA: trivial

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25636)

Improve documentation about duplicate algorithm registrations

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25624)

s390x: Don't probe crypto cards for ME/CRT offloading during initialization

Probing for crypto cards during initialization by issuing an ioctl to the
zcrypt device driver can cause a lot of traffic and overhead, because it
runs for each and every application that uses OpenSSL, regardless if that
application will later perform ME or CRT operations or not.

Fix this by performing no probing during initialization, but detect the
crypto card availability only at the first ME/CRT operation that is subject
to be offloaded. If the ioctl returns ENODEV, then no suitable crypto
card is available in the system, and we disable further offloading
attempts by setting flag OPENSSL_s390xcex_nodev to 1.

Setting the global flag OPENSSL_s390xcex_nodev in case of ENODEV is
intentionally not made in a thread save manner, because the only thing
that could happen is that another thread, that misses the flag update,
also issues an ioctl and gets ENODEV as well.

The file descriptor is not closed in such error cases, because this could
cause raise conditions where we would close a foreign file if the same
file descriptor got reused by another thread. The file descriptor is finally
closed during termination by the atexit handler.

In case the ioctl returns ENOTTY then this indicates that the file descriptor
was closed (e.g. by a sandbox), but in the meantime the same file descriptor
has been reused for another file. Do not use the file descriptor anymore,
and also do not close it during termination.

Fixes: https://github.com/openssl/openssl/commit/79040cf29e011c21789563d74da626b7465a0540
Signed-off-by: Ingo Franzki <ifranzki@linux.ibm.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25576)

work around oqsprovider out-of-source build bug

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25702)

Updated oqsprovider to v0.7.0

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25702)

Add CHANGES.md and NEWS.md entries for CVE-2024-9143

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/25734)

(cherry picked from commit 233034bc5a294b26d37186dc68d7d6d8357d889a)

ecdh_cofactor_derive_test(): Skip the test if the curve is not supported

It will not be supported if the fips provider was built with no-ec2m.

Fixes #25729

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25731)

run-checker-merge.yml: Combine no-ec2m with enable-fips

This can reveal more errors than just no-ec2m.

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25731)

update logo

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25725)

first cut at KEM & key management skeletons

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25640)

ci: re-organize external tests

Signed-off-by: Ondrej Moris <omoris@redhat.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25587)

test: clean-up README-external documentation

Signed-off-by: Ondrej Moris <omoris@redhat.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25587)

test: Add external test for pkcs11-provider

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25587)

add test coverage for #25298

Add test coverage for issue #25298, clean up the json file so
it uses consistent indentation

Signed-off-by: Alicja Kario <hkario@redhat.com>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
(Merged from https://github.com/openssl/openssl/pull/25329)

update tlsfuzzer to new version

Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
(Merged from https://github.com/openssl/openssl/pull/25329)

updating comments in test recipie

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

updating docs to reflect security risks for SSLKEYLOGFILE

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

Convert keylogging in response to comments

1) Convert failures in keylog setup to trace messages for a warning-like
   mechanism

2) Convert sslkeylogfile_cb to be a flag used to determine making a
   direct call to the internal logging function

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

Fix build break when OPENSSL_NO_SSLKEYLOGFILE is not defined

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

Update sslkeylog in response to comments

* instead of keeping an external reference count, just use the
  BIO_up_ref call, and the BIO's callback mechanism to detect the
  final free, for which we set keylog_bio to NULL

* Return an error from SSL_CTX_new_ex if the setup of the keylog file
  fails

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

Adding documentation to INSTALL

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

Enable sslkeylog for some CI builds for checking

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

Add test to confirm sslkeylogfile format

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

Add documentation for SSLKEYLOGFILE

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

Add sslkeylog config option and implementation

Add a config option for sslkeylog (disabled by default)

When enabled, SSL_CTX_new[_ex] becomes sensitive to the SSLKEYLOGFILE
environment variable.  It records keylog callback messages to the file
specified in the environment variable according to the format specified
in https://www.ietf.org/archive/id/draft-thomson-tls-keylogfile-00.html

Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Hugo Landau <hlandau@devever.net>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25297)

CHANGES.md: Support for PKCS#7 inner contents verification

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/22575)

Add support for p7->d.sign->contents with the V_ASN1_SEQUENCE type

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/22575)

Revert API change of OPENSSL_version()

There was an API change done as part of PR #24450.
This patch reverts it.

Fixes #25690

Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25692)

[DOC] Document the OPENSSL_WINCTX type for OpenSSL_version()

Reviewed-by: Matt Caswell <matt@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/25693)

fix: let util/mkbuildinf.pl use SOURCE_DATE_EPOCH, even if it's zero

Doing this allows reproducible builds, for those who want this.

Fixes #25475

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25699)

openssl-info.pod.in: Add windowscontext option to synopsis in doc

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/25712)

Add 'openssl info' item for the Windows install context

This information is already present as an 'openssl version' item.

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25694)

mem: Don't use posix_memalign() and friends with custom wrapper

If the application provides custom memory allocations functions via
CRYPTO_set_mem_functions() then those should be used instead something
else like posix_memalign(). The applications might verify alloc and free
calls and pointers from posix_memalign() were never returned by the
implementations.

At least stunnel4 complains here.

Use posix_memalign() or if aligned_alloc() only if the application did
not provide a custom malloc() implementation. In case of a custom
implementation use CRYPTO_malloc() and align the memory accordingly.

Fixes #25678

Signed-off-by: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25682)

Dependabot update

CLA: trivial

(deps): Bump coverallsapp/github-action

Bumps [coverallsapp/github-action](https://github.com/coverallsapp/github-action) from 2.3.0 to 2.3.2.
- [Release notes](https://github.com/coverallsapp/github-action/releases)
- [Commits](https://github.com/coverallsapp/github-action/compare/v2.3.0...v2.3.2)

---
updated-dependencies:
- dependency-name: coverallsapp/github-action
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25701)

Harden BN_GF2m_poly2arr against misuse.

The BN_GF2m_poly2arr() function converts characteristic-2 field
(GF_{2^m}) Galois polynomials from a representation as a BIGNUM bitmask,
to a compact array with just the exponents of the non-zero terms.

These polynomials are then used in BN_GF2m_mod_arr() to perform modular
reduction.  A precondition of calling BN_GF2m_mod_arr() is that the
polynomial must have a non-zero constant term (i.e. the array has `0` as
its final element).

Internally, callers of BN_GF2m_poly2arr() did not verify that
precondition, and binary EC curve parameters with an invalid polynomial
could lead to out of bounds memory reads and writes in BN_GF2m_mod_arr().

The precondition is always true for polynomials that arise from the
standard form of EC parameters for characteristic-two fields (X9.62).
See the "Finite Field Identification" section of:

    https://www.itu.int/ITU-T/formal-language/itu-t/x/x894/2018-cor1/ANSI-X9-62.html

The OpenSSL GF(2^m) code supports only the trinomial and pentanomial
basis X9.62 forms.

This commit updates BN_GF2m_poly2arr() to return `0` (failure) when
the constant term is zero (i.e. the input bitmask BIGNUM is not odd).

Additionally, the return value is made unambiguous when there is not
enough space to also pad the array with a final `-1` sentinel value.
The return value is now always the number of elements (including the
final `-1`) that would be filled when the output array is sufficiently
large.  Previously the same count was returned both when the array has
just enough room for the final `-1` and when it had only enough space
for non-sentinel values.

Finally, BN_GF2m_poly2arr() is updated to reject polynomials whose
degree exceeds `OPENSSL_ECC_MAX_FIELD_BITS`, this guards against
CPU exhausition attacks via excessively large inputs.

The above issues do not arise in processing X.509 certificates.  These
generally have EC keys from "named curves", and RFC5840 (Section 2.1.1)
disallows explicit EC parameters.  The TLS code in OpenSSL enforces this
constraint only after the certificate is decoded, but, even if explicit
parameters are specified, they are in X9.62 form, which cannot represent
problem values as noted above.

Initially reported as oss-fuzz issue 71623.

A closely related issue was earlier reported in
<https://github.com/openssl/openssl/issues/19826>.

Severity: Low, CVE-2024-9143

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25639)

Update documentation for SSL_SESSION_set_time_ex()

Fixes #24322

CLA: trivial

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
(Merged from https://github.com/openssl/openssl/pull/25665)

replace various calls to sprintf() by BiO_snprintf() to avoid compiler warnings, e.g., on MacOS

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Saša Nedvědický <sashan@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25534)

Fix memory leak in tls_parse_ctos_psk()

`sess` is not NULL at this point, and is freed on the success path, but
not on the error path. Fix this by going to the `err` label such that
`SSL_SESSION_free(sess)` is called.

CLA: trivial

Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25643)

fipsinstall: Save the 'status indicator' if the FIPS provider is 3.0.X.

Fixes #23400

The 3.1 FIPS provider no longer writes out the 'status indicator' by
default due to changes related to FIPS 140-3 requirements. For Backwards
compatability if the fipsinstall detects it is loading a 3.0.X FIPS
provider then it will save the 'status indicator' by default.

Disclaimer: Using a fipsinstall command line utility that is not supplied
with the FIPS provider tarball source is not recommended.

This PR deliberately does not attempt to exclude any additional options
that were added after 3.0.X. These additional options will be ignored by older
providers.

Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Hugo Landau <hlandau@devever.net>
(Merged from https://github.com/openssl/openssl/pull/23689)

Fix potential double free through SRP_user_pwd_set1_ids()

If SRP_user_pwd_set1_ids() fails during one of the duplications, or id
is NULL, then the old pointer values are still stored but they are now dangling.
Later when SRP_user_pwd_free() is called these are freed again,
leading to a double free.

Although there are no such uses in OpenSSL as far as I found,
it's still a public API.

CLA: trivial

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25655)

Use poll() in BIO_socket_wait() if available

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25055)

Avoid undefined behaviour with the <ctype.h> functions.

fix https://github.com/openssl/openssl/issues/25112

As defined in the C standard:

   In all cases the argument is an int, the value of which shall
   be representable as an unsigned char or shall equal the value
   of the macro EOF.  If the argument has any other value, the
   behavior is undefined.

This is because they're designed to work with the int values returned
by getc or fgetc; they need extra work to handle a char value.

If EOF is -1 (as it almost always is), with 8-bit bytes, the allowed
inputs to the ctype.h functions are:

   {-1, 0, 1, 2, 3, ..., 255}.

However, on platforms where char is signed, such as x86 with the
usual ABI, code like

   char *p = ...;
   ... isspace(*p) ...

may pass in values in the range:

   {-128, -127, -126, ..., -2, -1, 0, 1, ..., 127}.

This has two problems:

1. Inputs in the set {-128, -127, -126, ..., -2} are forbidden.

2. The non-EOF byte 0xff is conflated with the value EOF = -1, so
   even though the input is not forbidden, it may give the wrong
   answer.

Casting char inputs to unsigned char first works around this, by
mapping the (non-EOF character) range {-128, -127, ..., -1} to {128,
129, ..., 255}, leaving no collisions with EOF.  So the above
fragment needs to be:

   char *p = ...;
   ... isspace((unsigned char)*p) ...

This patch inserts unsigned char casts where necessary.  Most of the
cases I changed, I compile-tested using -Wchar-subscripts -Werror on
NetBSD, which defines the ctype.h functions as macros so that they
trigger the warning when the argument has type char.  The exceptions
are under #ifdef __VMS or #ifdef _WIN32.  I left alone calls where
the input is int where the cast would obviously be wrong; and I left
alone calls where the input is already unsigned char so the cast is
unnecessary.

Reviewed-by: Neil Horman <nhorman@openssl.org>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Paul Dale <ppzgs1@gmail.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25113)

Revert "TEMPORARY: run daily checks on PR"

This reverts commit 4c44603d555ece65cb635ebe191f4000c88bb429.

Reviewed-by: Saša Nedvědický <sashan@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25658)

improve and move text on OPENSSL_TRACE from doc/man1/openssl.pod to doc/man7/openssl-env.pod

Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25540)

Add utility function ossl_param_is_empty()

Changed all provider implementations that have a set_ctx_params()
to call this function instead of just testing (params == NULL).This
detects the case wherean OSSL_PARAM array contains just a terminator
entry.

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Tom Cosgrove <tom.cosgrove@arm.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25499)

APPS/storeutl: fix case where uri and outfile are the same

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/25559)