SHA256: more SHA256 support

Support the notion of a home directory separately from global configuration directory

 Define correct off64_t for AIX

#6422: handle dangling symbolic refs gracefully

Fixes issue #6156.

Signed-off-by: Sven Strickroth <email@cs-ware.de>

Fix build failure with -DEMBED_SSH_PATH

diff_file: Fix crash when freeing a patch representing an empty untracked file

Support OpenSSL 3 in dynamic loading mode

Add new PHP bindings project to language bindings section of README.md

Updates the README.md language binding section to reference a new
project providing PHP bindings at https://github.com/RogerGee/php-git2.

This prevents GIT_MMAP_VALIDATE from failing.
This also prevents git_diff_file_content__unload from attempting to free git_str__initstr.

The git included with xenial is ancient, and lacks sha256 support.

The longest path within a git repository is now a SHA256 format
packfile. Adjust our max length checking to match.

This is much of the plumbing for the object database to support SHA256,
and for objects to be able to parse SHA256 versions of themselves.

Teach the packfile machinery to cope with SHA256.

In clar, an underscore is meaningful; avoid using it incorrectly /
unnecessarily.

This is a conversion of the testrepo.git to SHA256 support.

Ensure that we maintain the `core.repositoryFormatVersion` value instead
of always overwriting it with the default.

Provide an internal function to set the repository's `objectformat`,
both in the internal object and in the configuration.

Teach the repository about the `objectformat` extension, supporting
`sha1` always and `sha256` when the experimental sha256 support is
active.

Try to load OpenSSL 3 libraries when compiled with OpenSSL-Dynamic
support.

Handle the deprecated symbol renaming of SSL_get_peer_certificate to
SSL_get1_peer_certificate -- try to load the old name and if it fails,
use the new one.

When we fail to initialize the OpenSSL subsystem, don't assume that
we're "initialized". Subsequent initialization requests -- if there are
any -- should replay initialization and fail again.

We provide `BUILD_PATH` to our build script; provide it and mutate
`PATH` when running our tests as well.

We were previously using `cygpath` to try to convert a _list_ of Windows
paths into cygwin / Unix style `PATH` format. This does not work -- it
treats the path list as a single path (with semicolons -- understandably
as those are allowed characters in a Windows path).

For example, `C:\One;C:\Two;C:\Three` is converted to
`/c/one;c:/two;c:/three`.

Add a new function to convert path lists, so that paths are split by
semicolon and fed to `cygpath` independently, then re-joined with a
colon. This means that our example `C:\One;C:\Two;C:\Three` is correctly
converted to `/c/one:/c/two:/c/three`.