doc: document the MAC block size getter

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15427)

Add fipsinstall option to run self test KATS on module load

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15149)

Remove engine_table_select_int

Add missing file/line args and call it engine_table_select

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15419)

DOCS: Don't mention internal functions in public documentation

This time noticed in OSSL_trace_set_channel.pod, and it turned out to
be easy to mention the public functions affected instead.

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15422)

Fix warning in gf_serialize

Compiling under -Werror fails in gf_serialize:

crypto/ec/curve448/f_generic.c:21:27: error: argument 1 of type 'uint8_t[56]' {aka 'unsigned char[56]'} with mismatched bound [-Werror=array-parameter=]
   21 | void gf_serialize(uint8_t serial[SER_BYTES], const gf x, int with_hibit)
      |                   ~~~~~~~~^~~~~~~~~~~~~~~~~
In file included from crypto/ec/curve448/f_generic.c:12:
crypto/ec/curve448/field.h:65:28: note: previously declared as 'uint8_t *' {aka 'unsigned char *'}
void gf_serialize(uint8_t *serial, const gf x, int with_highbit);
                  ~~~~~~~~~^~~~~~
Changed parameter to pointer to fix this warning.

Signed-off-by: Juergen Christ <jchrist@linux.ibm.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15376)

TEST: Avoid using just 'example.com'  - test_cmp_http

We have reports that some are using example.com in their /etc/hosts
for testing purposes, so we can't necessarily assume that those will
fail.

We fix it by using "random" hosts in that domain.

Fixes #15395

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15398)

openssl srp: make index.txt parsing error more verbose

If index.txt exists but has some problems (like for example
consisting of a single \n character or number of fields wrong in one of the lines)
then openssl will just exit. This fixes it by printing an error when
load_index returns null.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15360)

openssl ocsp: make index.txt parsing error more verbose

If index.txt exists but has some problems (like for example consisting of a single \n character in it,
or some field-number error in one of the lines) openssl will just exit without any error message.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15360)

openssl ca: make index.txt parsing error more verbose

If index.txt exists but has some problems (like for example a single \n character in it) openssl will just exit without any error message.

Bug at least expirienced twice: https://superuser.com/questions/1327848/openssl-ca-fails-after-password-without-error-message

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15360)

Don't try the same decoder multiple times

The function collect_decoder decides whether a given decoder should be
tried or not. It loops through all the names for matching keymgmts to
see if any are a match or not. If there is a match then the decoder gets
added. However, each keymgmt may have multiple aliases and a decoder was
being added for each one. For example DHX has 4 alias names, and therefore
4 instances of the DHX decoder were added and being tried.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15404)

Windows CI: use nasm on 32bit and 64bit shared builds

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15349)

Add some basic Windows builds to the Windows CI workflow

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15349)

Add a test for PKCS5_PBE_keyivgen()

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14326)

Added PKCS5_PBE_keyivgen_ex() to allow PBKDF1 algorithms to be fetched for a specific library context

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14326)

Add PBKDF1 to the legacy provider

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14326)

Remove some perl 5.14 use from rsaz-avx512.pl

The non-destructive substitution syntax (s///r), was introduced in perl
5.14. We need to support 5.10 and above.

Fixes #15378

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/15379)

Rename the field 'provctx and data' to 'algctx' inside some objects containing
pointers to provider size algorithm contexts.

Fixes #14284

The gettable_ctx_params methods were confusingly passing a 'provctx' and
a provider context which are completely different objects.
Some objects such as EVP_KDF used 'data' while others such as EVP_MD used 'provctx'.

For libcrypto this 'ctx' is an opaque ptr returned when a providers algorithm
implementation creates an internal context using a new_ctx() method.
Hence the new name 'algctx'.

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/15275)

fips: default to running self tests when starting the fips provider

Also add a C++ constructor as per note 7 of IG 9.10 if no DEP is available and
C++ is being used.

Fixes #15322

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15324)

doc: update core_thread_start() documentation

It is now passed an arugment to pass to the callback

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15278)

Fix a memleak in the FIPS provider

If the DRBG is used within the scope of the FIPS OSSL_provider_init
function then it attempts to register a thread callback via c_thread_start.
However the implementation of c_thread_start assumed that the provider's
provctx was already present. However because OSSL_provider_init is still
running it was actually NULL. This means the thread callback fail to work
correctly and a memory leak resulted.

Instead of having c_thread_start use the provctx as the callback argument
we change the definition of c_thread_start to have an explicit callback
argument to use.

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15278)

test: fix typo in comment in threadstest.c

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15278)

core: condition out more in FIPS builds

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15278)

test: add test case to reliably reproduce RAND leak during POST

The FIPS provider leaks a RAND if the POST is run at initialisation time.
This test case reliably reproduces this event.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15278)

80-test_cmp_http: Invert and correct the logic of success vs. failure exit

This makes the logic more intuitive and corrects the interpretation for NonStop.
Fixes #15386

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15402)

apps/cms.c: Correct -sign output and -verify input with -binary

Also add related warnings on irrelevant use of -nodetach and -content options.

Fixes #15347

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15358)

Disable loader_attic by default on VMS

The reason is that it currently doesn't build properly, due to the of
pvkfmt.c, causing multiply defined symbols since libcrypto exports
them as well.  At the same time, it can't do without that source file,
or it won't have access to certain internal symbols from there.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15320)

Make it possible to disable the loader_attic engine

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15320)

VMS: Fix run of generic generator programs in descrip.mms.tmpl

For a generic program, always go through the MCR utility.

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15397)

Configurations/descrip.mms.tmpl: rework the inclusion hacks

Because VMS C has some trouble with recursive inclusion of header
files, we have had to help it out for object files where there is such
an inclusion structure.

Previously, we did so with temporary logical names that were the same
as the first directory in an inclusion, so for example, to enable this
inclusion (found in ssl/ssl_local.h), we created the logical name
"record" when building any of the object files in the ssl/
subdirectories:

    #include "record/record.h"

However, there is another way with the VMS C compiler, to selectively
specify extra include directories in Unix form directly to the
compiler.  The logic is that from the directory where the source file
to compile is located, the specified inclusion directory merged with
the inclusion string should be able to access to specified header
file.

So for example, when a file in ssl/record/ is compiled, the following
inclusion is found:

    #include "../ssl_local.h"

So far so good, VMS C handles it properly.  However, the recursive
inclusion of "record/record.h" fails.  However, if the compiler is
helped out a little bit, with the following extra qualifier, then it
works:

    /INCLUDE="../"

The reason is that the compiler merges "../" and "record/record.h"
into "../record/record.h", which is the correct path to that header
file from the directory of the source file being compiled.

All that remained was to figure out all places where this trouble may
occur, and specify extra Unix formatted inclusion directories to
specify on per object file basis.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15369)

Fix upgrading docs for RSA_private_encrypt/RSA_public_decrypt

Despite the name, these functions manipulate signatures, which means
that their replacements are the EVP_PKEY_sign/EVP_PKEY_verify family.

Signed-off-by: Robbie Harwood <rharwood@redhat.com>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Dmitry Belyavskiy <beldmit@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/15359)

apps/cms: Simplify handling of encerts; add warning if they are ignored

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14843)

configurations: update template makefiles to install documentation images

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15399)

configure: build list of image files

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15399)

doc: process images when installing

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15399)

doc: rereference img locations into subdirectory

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15399)

doc: move images into their own subdirectory

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15399)

property: convert integers to strings properly.

The int64_t type was converted to int (truncation).
Negative values were not handled at all.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15396)

test/params_conversion_test.c: fix the use of strtoumax and strtoimax on VMS

We do this by making them aliases for strtoull and strtoll, since long
long is the current largest integer that have this sort of routine on
VMS.

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15366)

Include "internal/numbers.h" in test programs using SIZE_MAX

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15366)

VMS: don't use app_malloc() in apps/lib/vms_decc_argv.c

The reason being that it would otherwise force test programs to link
with all of libapps.a, which unfortunately causes multiple symbol
definition issues.

The quick and dirty fix is to use OPENSSL_malloc() instead of
app_malloc() in apps/lib/vms_decc_argv.c, and clean up libapps.a
later.

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15368)

Cleanup the peer point formats on regotiation

Fixes #14875

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15383)

DOCS: Fixups of the migration guide and the FIPS module manual

The markup needed a few touch-ups

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/15377)

HMAC doesn't work with a default digest

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15371)

Properly restore XMM registers in ChaCha20's AVX-512(VL) assembly

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15315)

PROV: Relegate most of the FIPS provider code to libfips.a

provider/fips/fipsprov.c contains a number of symbols that get used by
anything that's included in libfips.a, at least on Unix.
Unfortunately, there are platforms that do not support resolving
symbols to things that are already included in the end product (module
in this case) being built; they only support resolving symbols with
what comes next in the linking process.

The offending symbols in this case are FIPS_security_check_enabled,
c_thread_start and ossl_fips_intern_provider_init.

We resolve this by placing provider/fips/fipsprov.c in libfips.a along
with everything else there.  That takes care of the offending symbols.
What remains is to ensure that there is an entry point in an object
file used directly when linking the module, providers/fips/fips_entry.c

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15370)

Cleanup the missing*.txt files

One macro existed that was added since 1.1.1 and was undocumented. This
had been added to missingmacro.txt. This is the wrong approach and so
has been removed from there.

There were some entries in missingcrypto.txt that don't exist as functions
at all. There were also some which were in fact documented.

Additionally 2 entries from missingcrypto.txt have been moved to
missingmacro.txt. These entries existed in 1.1.1 and were undocumented. In
master they have been deprecated and compatibility macros for them
implemented. The replacement functions have been documented.

An entry in missingcrypto111.txt was not in alphabetical order (and was
also) duplicated, but the equivalent entry in missingcrypto.txt was in the
correct place. This has been corrected to make comparisons between the files
easier.

Finally a function has been added to missingcrypto111.txt. This function
did exist in 1.1.1 and was undocumented. Its unclear why this wasn't in
missingcrypto111.txt to start with.

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15333)

apps: Cleanup useless bio_open_default() calls for key input

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/15331)

FIPS label CI: Save PR number and use it

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15345)

Clean up the "fips" option to Configure

Don't die if someone says "fips" instead of "enable-fips"

Reviewed-by: Richard Levitte <levitte@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/15346)

BIO_s_accept.pod: Document port auto-selection feature of BIO_set_accept_port()

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15354)

EVP_DigestSignInit.pod: Clarification in EVP_DigestSignFinal() parameter 'sig'

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15354)

Exchange no-siv and no-ec2m between daily and ci workflows

The no-ec2m with ec enabled is much more likely to show
regressions such as #15170 than the no-siv build.

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15355)

pem_read_bio_key_legacy: Do not obscure real error if there is one

Fixes #15170

Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15355)

Add ordinal numbers to the .num files

Now that our next release is expected to be a beta release, "make update"
wants to see ordinal numbers in the .num files.

Run make update to add them.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15384)

apps/cms: Clean up order of options in help output and documentation

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15126)

APPS: Allow duplicate entries in options list, marking them OPT_DUP

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15126)

APPS: Allow non-option parameters appear anywhere in list, marking them OPT_PARAM

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15126)

check-format.pl: Rename '*-cmt' options '*-comment'

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15077)

check-format.pl: Rename 'one-letter' to 'single-letter', do not report 'l'

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15077)

check-format.pl: Allow extra space before end-of-line comments unless -e|--eol-cmt given

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15077)

check-format.pl: Replace 'SPC' and 'spc' by 'space' in reports and option names

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15077)

check-format.pl: Fix false positive on struct/union/enum in func return type

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15077)

check-format.pl: Fix false positive "no SPC before binary '*'" for '!*'

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15077)

check-format.pl: Report needless intermediate multiple SPC only on -e or --extra-spc

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15077)

check-format.pl: Add check for constant left of comparison operator

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15077)

VMS: Copy __DECC_INCLUDE_{PROLOGUE,EPILOGUE}.H to more places

Every inclusion directory related to a library we build need these two
files.  That signals to any other module using anything from these
libraries what to expect in terms of case sensitivity as well as how
long symbol names are dealt with.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15341)

util/find-doc-nits: Improve helpstr pattern matching

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15299)

DOC: Fix nits found by new check on SYNOPSIS and OPTIONS consistency

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15299)

find-doc-nits: Check that man1 SYNOPSIS and OPTIONS contain same options

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15299)

find-doc-nits: Add -m option allowing to select on which of man1,man3,man5,man7 to focus on

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15299)

find-doc-nits: Minor improvements of help and diagnostic output

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15299)

CMP test server: move apps/{,lib/}cmp_mock_srv.c and apps/{,include/}cmp_mock_srv.h

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15343)

apps/cmp.c: Move CMP server code portion to separate function

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15343)

Prepare for 3.0 beta 1

Reviewed-by: Richard Levitte <levitte@openssl.org>

Prepare for release of 3.0 alpha 17

Reviewed-by: Richard Levitte <levitte@openssl.org>

make update

Reviewed-by: Richard Levitte <levitte@openssl.org>

Update copyright year

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15381)

EVP: Modify EVP_PKEY_export() to handle legacy EVP_PKEYs

We use a fake EVP_KEYMGMT import function with the newly modified
EVP_PKEY_ASN1_METHOD export_to function to pass the exported
OSSL_PARAM array directly to the EVP_PKEY_export() callback instead of
exporting to an actual provided key and then getting the OSSL_PARAM
array from there, just to throw away that key again.

Fixes #15290

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15293)

Modify EVP_PKEY_ASN1_METHOD's export_to function to take an importer

We previously took an EVP_KEYMGMT pointer, but now found it necessary
to use a different import function in some cases.  Since that's the
only thing we use from EVP_KEYMGMT, we might as well pass the import
function directly, allowing for some flexibility in how export_to is
used.

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15293)

test/evp_extra_test2.c: Try EVP_PKEY_export() with a legacy RSA key

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15293)

Refer to the migration guide rather than the wiki in our announcements

We now have a migration guide which should be the definitive source of
information for upgrading from a previous version of OpenSSL.

Fixes #15186

Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15373)

Create symlinks when installing man pages

In 1.1.1 when installing the man pages we created symlinks to the base
page for all functions described on the page. We need to continue doing
this.

Fixes #14846

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15312)

app: add a -store_loaders option to list.

Fixes #15307

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15323)

Complete 'no-sock' guards in apps/ocsp.c

Modern compilers complain about variable being set but otherwise not used.

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15339)

Update documentation for global properties mirroring

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15242)

Ensure mirroring of properties works for subsequent updates

If the global properties are updated after a provider with a child libctx
has already started we need to make sure those updates are mirrored in
that child.

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15242)

Documentation updates for mirroring of global properties

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15242)

Test that properties are mirrored as we expect

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15242)

Add a callback for providers to know about global properties changes

Where a child libctx is in use it needs to know what the current global
properties are.

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15242)

Add a test for converting a property list to a string

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15242)

Implement the ability to convert a PROPERTY_LIST to a string

We have the ability to parse a string into a PROPERTY_LIST already. Now
we have the ability to go the other way.

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15242)

todo: remove TODO(3.0) from the sources.

Almost all were notes about wanting to deprecate CTRLs/utility functions.

Fixes #15325

Reviewed-by: Tomas Mraz <tomas@openssl.org>
Reviewed-by: Shane Lontis <shane.lontis@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/15328)

Test d2i_PrivateKey_bio() does not add errors to stack when decoding a X25519 key sucessfully.

This confirms that another merge has addressed this issue.

Fixes #14996

Reviewed-by: Paul Dale <pauli@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15303)

Revert "ARM assembly pack: translate bit-sliced AES implementation to AArch64"

This reverts commit da51566b256e0c0536d5b986e676863b0526bf5e.

Fixes #15321

Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/15364)

Add migration guide for 3.0

Reviewed-by: Paul Dale <pauli@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14710)

Update SSL_new_session_ticket() manual for triggered send

Document the recently added functionality.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14817)

Test new SSL_new_session_ticket() functionality

Now that we can become "in init" directly after the call, test the
various scenarios where explicit SSL_do_handshake() calls can come
into play.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14817)

Let SSL_new_session_ticket() enter init immediately

The initial implementation always deferred the generation of the
requested ticket(s) until the next application write, but this
is not a great fit for what it actually does, architecturally wise.
A request to send a session ticket means entering back into the
handshake state machine (or "in init", as it's known in the
implementation).  The state machine transition is not something that
only occurs at an application-data write, and in general could occur at
any time.  The only constraint is that we can't enter "init" while in
the middle of writing application data.  In such cases we will need to
wait until the next TLS record boundary to enter the state machine,
as is currently done.

However, there is no reason why we cannot enter the handshake state
machine immediately in SSL_new_session_ticket() if there are no
application writes pending.  Doing so provides a cleaner API surface to
the application, as then calling SSL_do_handshake() suffices to drive
the actual ticket generation.  In the previous state of affairs a dummy
zero-length SSL_write() would be needed to trigger the ticket
generation, which is a logical mismatch in the type of operation being
performed.

This commit should only change whether SSL_do_handshake() vs zero-length
SSL_write() is needed to immediately generate a ticket after the
SSL_new_session_ticket() call -- the default behavior is still to defer
the actual write until there is other application data to write, unless
the application requests otherwise.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14817)

danetest.c: Improve code formatting

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14422)

X509_STORE_CTX_get1_issuer(): Simplify code, reducing risk of failure

Reviewed-by: Tomas Mraz <tomas@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/14422)