Add support of linking files compiled into SPIR-V objects
using spirv-link.

Command line inteface examples:

clang --target=spirv64 test1.cl test2.cl

clang  --target=spirv64 test1.cl -o test1.o
clang  --target=spirv64 test1.o test2.cl -o test_app.out

This works independently from the SPIR-V generation method
(via an external tool or an internal backend) and applies
to either approach that is being used.

Differential Revision: https://reviews.llvm.org/D116266

By forwarding it to the host platform.

This reverts commit f62f47f5.

As reported in https://reviews.llvm.org/D116766#3233042

Make the test more robust by replacing undef by pointer arguments. This
ensures that the GEPs cannot be folded to undef.

It turns out this is conflating a few different PMU extensions. And on
Arm ended up breaking M-Profile code generation. Reverting for the
moment whilst we sort out the details.

This reverts commit d17fb46e.

If a class declares an instance property, and an inheritor class declares a class property with the same name, Clang Sema currently treats the latter as an overridden property, and compares the attributes of the two properties to check for a mismatch. The resulting diagnostics might be misleading, since neither of the properties actually overrides the another one.

rdar://86018435

Differential Revision: https://reviews.llvm.org/D116412

For certain negative indices passed to the VECTOR_SPLICE operation
we can actually directly use the SVE splice instruction by creating
the appropriate predicate. The predicate needs to be constructed in
such a way that all but the last -idx elements are false. We can do
this efficiently using a combination of 'ptrue' (with the appropriate
fixed pattern, e.g. vl1, vl2, etc.) and 'rev'. The advantage of using
these instructions to generate the predicate is they do not set any
flags, unlike the whilelo instruction. This is critical when the splice
operation is in a loop, since we want MachineLICM to hoist the
predicate generation out of the loop.

Differential Revision: https://reviews.llvm.org/D115863

The interface for these instructions changed with support for mandatory tail
calls, and now -1 indicates the CalleePopAmount argument is not valid.
Unfortunately I didn't realise FastISel or GISel did calls at the time so
didn't update them.

The pointer is always dereferenced immediately below, so assert the cast is correct instead of returning nullptr

This function is no longer used since
2cafbcb5.

This function is no longer used after
dcbc91f4.

For these "visible on unwind/ret" checks we only care about the
fact that no other code has access to the pointer (unless it
escapes). A noalias call is sufficient for this, it does not
have to be a known allocation function.

This is basically the same change as D116728, but for DSE rather
than LICM.

llvm/test/Transforms/LoopStrengthReduce/uglygep-address-space.ll has
exactly the same checks and input. Remove the duplicated test.

Previously, CallOps did not have any aliasing OpResult/OpOperand pairs. Therefore, CallOps were mostly ignored by the analysis and buffer copies were not inserted when necessary.

This commit introduces the following changes:
* Function bbArgs writable by default. A function can now be bufferized without inspecting its callers.
* Callers must introduce buffer copies of function arguments when necessary. If a function is external, the caller must conservatively assume that a function argument is modified by the callee after bufferization. If the function is not external, the caller inspects the callee to determine if a function argument is modified.

Differential Revision: https://reviews.llvm.org/D116457

Currently this is special-cased to TLI alloc functions only.

Reland 55d96ac3 and 37ec65e1 with a clang-tidy fix.

and use that to simplify MD5's hex string code which was previously
using a string stream, as well as Clang's
CGDebugInfo::computeChecksum().

Differential revision: https://reviews.llvm.org/D116960

If the allocation is not captured, then all the stores before the
ret are dead anyway.

A lookup table, which toHex() was using, seems like the better approach.
Having two implementations is redundant, so put the lookup table in
hexdigit() and make toHex() call that.

Differential revision: https://reviews.llvm.org/D116960

[SemaOverload] compareConversionFunctions - use castAs<> instead of getAs<> to avoid dereference of nullptr

The pointer is dereferenced immediately below, so assert the cast is correct instead of returning nullptr

The pointer is always dereferenced inside BuildSimilarlyQualifiedPointerType, so assert the cast is correct instead of returning nullptr

This reinstates a test that was temporarily removed in
e26bbae3, in a form that works on
Darwin.

Use -LD instead of -link as a linker argument that is unused when
compiling, that produces warnings normally. -LD can be placed anywhere
in the command line, so that the command line ends with "-- %s", making
paths starting with /U correctly interpreted as paths, not options.

Changed the remaining appearances of alloc to memref.alloc in several
documentation sections, since they lead to misunderstandings, if they
are used.

Differential Revision: https://reviews.llvm.org/D116999

The code can only address the whole RV32 address space or the lower 2 GiB
of the RV64 address space in small code model, so 32 bits entry is enough.
Cache hit ratio and code size have some improvements.

Reviewed By: asb

Differential Revision: https://reviews.llvm.org/D116435

Cover more of the code paths from LLVMTypeConverter::convertPointerLike
and LLVMTypeConverter::convertSequenceType.

Differential Revision: https://reviews.llvm.org/D116927

This saves about 10% of SelectionVisitor::pop().

Use PtrToInt instead Add when creating temporary instructions. The add
might get folded away with more sophisticated folding.

I've changed the definition of the experimental.vector.splice
instrinsic to reject indices that are known to be or possibly
out-of-bounds. In practice, this means changing the definition so that
the index is now only valid in the range [-VL, VL-1] where VL is the
known minimum vector length. We use the vscale_range attribute to
take the minimum vscale value into account so that we can permit
more indices when the attribute is present.

The splice intrinsic is currently only ever generated by the vectoriser,
which will never attempt to splice vectors with out-of-bounds values.
Changing the definition also makes things simpler for codegen since we
can always assume that the index is valid.

This patch was created in response to review comments on D115863

Differential Revision: https://reviews.llvm.org/D115933

This seems to be strictly faster in all cases. Before fixing D116978 it
was one of the hot paths, and may become one again.

These tests were recently added in
50ec1306. The clang-cl invocations
interpret the source path, %s, which begins with /Users, as a cl
option /U. Normally this is worked around by passing -- before
the arguments that must be interpreted as files, not options, but
in the current test, the -link option must be passed last, after any
file names.

This makes improvements in follow-up patches more obvious.

D115722 added a DL spec to GPU modules. It happens that the DL
default interface implementation is sensitive to the name of the
DL spec attribute. This patch is fixing the name of the attribute
to be the expected one.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D116956

When calling emitArrayDestroy(), the pointer will usually have
ConvertTypeForMem(EltType) as the element type, as one would expect.
However, globals with initializers sometimes don't use the same
types as values normally would, e.g. here the global uses
{ double, i32 } rather than %struct.T as element type.

Add an early cast to the global destruction path to avoid this
special case. The cast would happen lateron anyway, it only gets
moved to an earlier point.

Differential Revision: https://reviews.llvm.org/D116219

Loading constants inline is expensive on CSKY and it's in general better
to place the constant nearby in code space and then it can be loaded with a
simple 16/32 bit load instruction like lrw.

It needs lift or duplicates constant pool entry to make constant nearby so that lrw instruction can reach.

When `Zbt` is enabled, we can generate SELECT for division by power
of 2, so that there is no data dependency.

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D114856

[clang] Add --start-no-unused-arguments/--end-no-unused-arguments to silence some unused argument warnings

When passing a set of flags to configure defaults for a specific
target (similar to the cmake settings `CLANG_DEFAULT_RTLIB`,
`CLANG_DEFAULT_UNWINDLIB`, `CLANG_DEFAULT_CXX_STDLIB` and
`CLANG_DEFAULT_LINKER`, but without hardcoding them in the binary),
some of the flags may cause warnings (e.g. `-stdlib=` when compiling C
code). Allow requesting selectively ignoring unused arguments among
some of the arguments on the command line, without needing to resort
to `-Qunused-arguments` or `-Wno-unused-command-line-argument`.

Fix up the existing diagnostics.c testcase. It was added in
response to PR12181 to fix handling of
`-Werror=unused-command-line-argument`, but the command line option
in the test (`-fzyzzybalubah`) now triggers "error: unknown argument"
instead of the intended warning. Change it into a linker input
(`-lfoo`) which triggers the intended diagnostic. Extend the
existing test case to check more cases and make sure that it keeps
testing the intended case.

Add testing of the new option to this existing test.

Differential Revision: https://reviews.llvm.org/D116503