`uses_global_work_offset` attribute applies to a device function/lambda function or function call operator. Valid values are 0 and 1.
If value >1 is specified as an argument, we warn user that it will be adjusted to 1. As attribute is not ignored in this case and it does not match any existing diagnostic group, the new `AdjustedAttributes` group was added.

Signed-off-by: Viktoria Maksimova <viktoria.maksimova@intel.com>

…ups (intel#971)

The problem is that CheckSYCLCall used to get the caller incorrectly.
getCurFunctionDecl gets the wrong thing. It will get the function
containing the lambda at definition time, rather than the lambda operator().

Additional changes:
Cleaned up unnecessary calls of CheckSYCLCall function.
Fixed diagnosing of storage allocation through deferred diagnostics
system.
Renamed CheckSYCLCall with checkSYCLDeviceFunction since we don't really
check calls as CUDA/OpenMP does.
Removed unnecessary emitting of diagnostics from SemaSYCL.

Signed-off-by: Mariya Podchishchaeva <mariya.podchishchaeva@intel.com>

…tel#1025)

Signed-off-by: Joey Genfi <joey.genfi@intel.com>

Today we re-use OpenCL parsed attributes, but have separate SYCL address
space semantic attributes as current implementation of OpenCL semantics
breaks valid C++. This patch enables re-use of OpenCL semantic
attributes by allowing conversions between types qualified with OpenCL
address spaces and type w/o address space qualifiers. Clang compiler
(almost) always adds address space qualifiers in OpenCL mode, so it
should not affect OpenCL mode.

NOTE: this change also disables implicit conversion between the
unqualified types and types qualified with
`__attribute__((address_space(N)))`, enabled by one of the previous SYCL
patches.

Signed-off-by: Alexey Bader <alexey.bader@intel.com>