Create one bitcode library for AMD and NVPTX

[MartinWehking]

Enable compilation of libdevice for AMD by adding AMDGCN to macro
guarded code parts in libdevice for enabling e.g. standard library
math function.
Add compilation workflow to SYCLLibdevice.cmake for AMD.

Create single bitcode libraries for AMD and NVPTX by compiling each
libdev file into bitcode first, linking these together and running
opt on them.
Strip away metadata by reusing prepare_builtins from libclc and making
the resulting library architecture independent.

Remove NVPTX bundles from the libdev object files and remove any
unbundling action spawned by the Clang driver for the SYCL toolchain
when compiling for the NVPTX backend.

Make the driver link against the single bitcode libraries for AMD and
NVPTX for the SYCL toolchain when device library linkage is not
excluded.

Do not select builtin LLVM intrinsics for AMDGCN by default to ensure
that stdlib functions can be found when linking against libdevice.

Ensure that the clang tests check for the correctness of the new
clang driver actions and check if the driver still links the device
code against the itt device libraries when device library linkage has
been excluded.

Refactor SYCLLibdevice.cmake by creating functions and grouping
e.g. the same compilation flags for a filetype together in one variable.
Reuse these variables and call functions to remove redundancies.

Fix a compilation error of Intel math function libraries for MSVC
when targeting AMD. Include "device.h" before including "device_imf.hpp"
to avoid the inclusion of <type_traits>, which failed with a
redefinition of symbols error.

[MartinWehking]

This has been superseded by #15048 and #15055