* Port libAfterImage on Win64

Add `libAfterImage/CMakeLists.txt` to facilitate the port of libAfterImage on Win64

* Update CMake minimal version (thanks Oksana)

* First step to port ROOT on Windows 64 bit

 - Introduce the Win64 option and flags in the CMake build system
 - Introduce the new `Longptr_t` and `ULongptr_t` types to hold a pointer (4 bytes on Win32 and 8 bytes on Win64), aimed to replace the `Long_t` and `ULong_t` used to cast pointers, and being both 4 bytes on Windows. Their values will not change on other platforms (`Long_t` and `ULong_t`) for backward compatibility
 - Add the 64 bit export symbols in `core/metacling/src/CMakeLists.txt`
 - Add the 64 bit version of lzma (xz-5.2.4-win64.tar.gz)

* Add forgotten semicolons

Using target_compile_options to set INTERFACE items on IMPORTED targets
is only supported in cmake 3.11 and up (see similar PR #5396).

The option `dataframe_distpyspark` is superseded by `test_distrdf_pyspark`, to clarify that finding pyspark and its dependencies (like Java) during configuration is strictly needed only for the tests.

Generally speaking, pyspark (and Java) are optional runtime dependencies that interest only users of the distributed RDataFrame python module with the Spark backend.

Add new option dataframe_distpyspark. When the option is ON, the distributed dataframe module will also support distributed execution with pyspark.
Enabling the option also triggers checks for pyspark dependencies (e.g. Java, py4j) on the system during configuration.

Cling needs to be built and installed as part of llvm/clang and
then we need to specify paths as we do for builtin_clang=Off.

Add some missing snippets for CodeEditor

* [cmake] Patch tbb for Apple M1 (issue #6903).

Apply patch from https://github.com/oneapi-src/oneTBB/pull/258

Some of VC methods are marked as deprectaed and causing compiler
warnings (or errors) even when methods are not used

For developer builds one could use symbolic link on ${OPENUI5DIR}

Let install full openui5 library version from default location
While original openui5 library is big (~60 MB now) and
constantly growing, include in ROOT repository only small subset
which is used by ROOT components

When configure_file used with COPYONLY, it does not check if
file was updated and does not copy it again. Add special rule
similar to install root headers

Move existing nlohmann/json.hpp include to builtin location
and use it by default. When builtin_nlohmannjson is OFF,
cmake tries to find it in default location.

silence warning "include <gsl/... - you meant <GSL/...?" by renaming source dir.

* Move opengl related section in SearchInstalledSoftware.cmake

* Adapt wording sugestions by Axel about opengl

Co-authored-by: Axel Naumann <Axel.Naumann@cern.ch>

* Adapt wording sugestions by Axel about status message for disabling opengl

Co-authored-by: Axel Naumann <Axel.Naumann@cern.ch>

Co-authored-by: Axel Naumann <Axel.Naumann@cern.ch>

 * xrootd does not have anymore the -Werror flag in release mode.
   Therefore, we can remove the -w flag and use the standard
   ROOT_EXTERNAL_CXX_FLAGS for the builtin xrootd.
 * Explicitely set the build type of xrootd to Release. This switches
   from the default RelWithDebInfo to plain Release and ensures that
   -Werror is not set.

Just specifying -Dpyroot=ON does not prevent pyroot_experimental from being set.
And the "deprecated" warning will be issues as long as pyroot_experimental is set.
So tell people to set it to OFF, which will get rid of the warning.

The old client library libXrdClient is removed.
Also XrdSysDNS is removed.

Do not build the legacy XProofD file server and client (already
disabled by default - give error if someone enables it while building
with xrootd 5).

Do not build the old libNetx.

Build the new libNetNG without XrdSysDNS, use XrdNetAddr instead.

The replacements removes the -Werror option for externals, which is
intended. However, it also replaces e.g. -Werror=format-security with
=format-security, which results in compilation failures due to the
unknown option =format-security.

Had to make sure that the GSL_CBLAS_LIBRARY variable is set in the
same way in which FindGSL.cmake would set it, and that TMVA would
explicitly wait for the completion of the GSL build.

We see authentication issues with the builtin client, see https://cern.service-now.com/service-portal?id=ticket&table=incident&sys_id=8cfde578db0d50d0006fd9f9f49619e7&view=sp
The hope is that recent xrootd changes might solve those.