Use tablegen generic tables to get the index of image intrinsic
arguments.
Before, the computation of which image intrinsic argument is at which
index was scattered in a few places, tablegen, the SDag instruction
selection and GlobalISel. This patch changes that, so only tablegen
contains code to compute indices and the ImageDimIntrinsicInfo table
provides these information.

Differential Revision: https://reviews.llvm.org/D86270

Support register and frame-index pair correctly as operands of
generic load/store instrucitons, e.g. LD1BZXrri, STLrri, and etc.
Add regression tests also.

Differential Revision: https://reviews.llvm.org/D88779

While affine maps are part of the builtin memref type, there is very
limited support for manipulating them in the standard dialect. Add
transpose to the set of ops to complement the existing view/subview ops.
This is a metadata transformation that encodes the transpose into the
strides of a memref.

I'm planning to use this when lowering operations on strided memrefs,
using the transpose to remove the stride without adding a dependency on
linalg dialect.

Differential Revision: https://reviews.llvm.org/D88651

This tends to increase code size but more importantly it reduces vgpr
usage, and could avoid costly readfirstlanes if the result needs to be
in an sgpr.

Differential Revision: https://reviews.llvm.org/D88580

This is in preparation for making the GCN patterns divergence-aware.
NFC.

Differential Revision: https://reviews.llvm.org/D88579

When nesting INSERT_SUBREG and EXTRACT_SUBREG, GlobalISelEmitter would
fail to find the register class of the nested node. This patch fixes
that for registers with subregs.

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D88487

The error-bit was missing, if a DeclRefExpr (which refers to a VarDecl
with a contains-errors initializer).

It could cause different violations in clang -- the DeclRefExpr is value-dependent,
but not contains-errors, `ABC<DeclRefExpr>` could produce a non-error
and non-dependent type in non-template context, which will lead to
crashes in constexpr evaluation.

Reviewed By: sammccall

Differential Revision: https://reviews.llvm.org/D86048

Use SCEV to salvage additional @llvm.dbg.value that have turned into
referencing undef after transformation (and traditional
salvageDebugInfo). Before transformation compute SCEV for each
@llvm.dbg.value in the loop body and store it (along side its current
DIExpression). After transformation update those @llvm.dbg.value now
referencing undef by comparing its stored SCEV to the SCEV of the
current loop-header PHI-nodes. Allow match with offset by inserting
compensation code in the DIExpression.

Fixes : PR38815

Differential Revision: https://reviews.llvm.org/D87494

[11/11] patch series to port ASAN for riscv64

These changes allow using ASAN on RISCV64 architecture.
The majority of existing tests are passing. With few exceptions (see below).
Tests we run on qemu and on "HiFive Unleashed" board.

Tests run:

```
Asan-riscv64-inline-Test  - pass
Asan-riscv64-inline-Noinst-Test  - pass
Asan-riscv64-calls-Noinst-Test  - pass
Asan-riscv64-calls-Test  - pass
```

Lit tests:

```
RISCV64LinuxConfig (282 supported, few failures)
RISCV64LinuxDynamicConfig (289 supported, few failures)
```

Lit failures:

```
TestCases/malloc_context_size.cpp - asan works, but backtrace misses some calls
TestCases/Linux/malloc_delete_mismatch.cpp - asan works, but backtrace misses some calls
TestCases/Linux/static_tls.cpp - "Can't guess glibc version" (under debugging)
TestCases/asan_and_llvm_coverage_test.cpp - missing libclang_rt.profile-riscv64.a
```

These failures are under debugging currently and shall be addressed in a
subsequent commits.

Depends On D87581

Reviewed By: eugenis, vitalybuka

Differential Revision: https://reviews.llvm.org/D87582

Rename the DwarfFile class in DWARFLinker to DWARFFile. This is
consistent with the other DWARF classes and avoids a ODR violation with
the DwarfFile class in AsmPrinter.

Darwin seems to use stmmN instead of stN. Use a regex to accept both.

Also try to actually clear st(7).

Differential revision: https://reviews.llvm.org/D88795

RBX was copied to a virtual register before this instruction
was created. And the EBX input for the final MWAITX is still
in a virtual register. So EBX isn't read by this pseudo.

It's not possible to do this in complete generality - a CU using a
sec_offset DW_AT_ranges has no way of knowing where its rnglists
contribution starts, so should not attempt to parse any full rnglist
table/header to do so. And even using FORM_rnglistx there's no need to
parse the header - the offset can be computed using the CU's DWARF
format (32 or 64) to compute offset entry sizes, and then the list
parsed at that offset without ever trying to find a rnglist contribution
header immediately prior to the rnglists_base.

The option needs to be passed to both host and device compilation.

Differential Revision: https://reviews.llvm.org/D88550

By convention the default output file for -E is "-" (stdout).
This is expected by tools like ccache, which uses output
of -E to determine if a file and its dependence has changed.

Currently clang does not use stdout as default output file for -E
for HIP, which causes ccache not working.

This patch fixes that.

Differential Revision: https://reviews.llvm.org/D88730

recommit 64f7790e after
fixing hip-device-libs.hip.

This reverts commit 64f7790e due
to regression in hip-device-libs.hip.

Add an option --gpu-instrument-lib= to allow users to specify
an instrument device library. This is for supporting -finstrument
in device code for debugging/profiling tools.

Differential Revision: https://reviews.llvm.org/D88557

[X86] Remove MWAITX_SAVE_EBX pseudo instruction. Always save/restore the full %rbx register even in gnux32.

ebx/rbx only needs to be saved when 64-bit registers are supported
anyway. It should be fine to save/restore the whole rbx register
even in gnux32 where the base is technically just ebx.

This matches what we do for cmpxchg16b where rbx is saved/restored
regardless of gnux32.

This is one of the reason for extra invalidations in D84959. In
practice, I don't think we have use cases needing this. This simplifies
the pipeline a bit and prune corner cases when considering
invalidations.

Reviewed By: asbirlea

Differential Revision: https://reviews.llvm.org/D85676

All uses of the file undef the macro already.

Few places did this already - refactor them all into a common helper.

i.e. they cannot be unreachable from the entry (which usually indicate usage errors).
This change allows the removal of some nullptr checks.

Reviewed By: kuhar

Differential Revision: https://reviews.llvm.org/D88758

MWAITX doesn't touch EFLAGS so no pseudos should def EFLAGS.

The SAVE_EBX/RBX pseudos only needs to def the EBX register that
the expansion overwrites. The EAX and ECX registers are only read.

The pseudo emitted during isel that is used by the custom inserter
shouldn't have any implicit defs or uses since everything is in
vregs.

These are now emitted by a CustomInserter rather than using a custom
inserter themselves.

This reverts commit e9b87f43.

There are issues with macros generating macros without an obvious simple fix
so I'm going to revert this and try something different.

Some new function parameter attributes are derived under NPM.

Reviewed By: rjmccall

Differential Revision: https://reviews.llvm.org/D88760

[NFC][SCEV] Add a test with some patterns where we could treat inttoptr/ptrtoint as semi-transparent

Make the dumper & API a bit more informative by using the new tombstone
addresses to filter out or otherwise render more explicitly dead code
ranges.

This aligns the behavior with the standard call as well as the LLVM verifier.

Reviewed By: ftynse, dcaballe

Differential Revision: https://reviews.llvm.org/D88362

Apparently querying dereferenceability of array allocations is
being intentionally penalized (https://reviews.llvm.org/D41398),
so avoid using them in tests.

This file seems to have been accidentally added as part of commit
413577a8.

For the wide builtins, pass a single input and output pointer to
the builtins. Emit the GEPs and input loads from CGBuiltin.

We were taking multiple pointer arguments in the builtin.
gcc accepts a single void*.

The cast from void* to _m128i* caused the IR generation to assume
the pointer was aligned.

Instead make the builtin take a single void*, emit i8* GEPs to
adjust then cast to <2 x i64>* and perform a store with align of 1.