This reverts commit 544d9128.

Currently, the MaxPool2D operation in the TOSA MLIR dialect does not
accept half-precision Fp16 and Bf16 tensors, converse to what is stated
in the [TOSA
Specification](https://www.mlplatform.org/tosa/tosa_spec.html#_max_pool2d).

This patch fixes the verifier to accept the two datatypes for
input/output tensors, and adds related LIT test cases in Tosa/ops.mlir

This reverts commit 659a48f2.

This caused another test to fail on Windows AArch64 and x86_64, and AArch64 Linux.

The image intrinsic optimizer pass was neglecting to check any arguments
of the load intrinsic after the VAddr arguments. For example multiple
loads from different resources should not have been combined but were,
because the pass was not checking the resource argument.

This matches the behaviour of the other SVE ACLE types.

This avoids link failures on other platorms that don't (yet) have an
implementation of __aarch64_sme_accessible.

This adds actual test cases for all the cases that are listed in a code
comment in the implementation of this function; having such test
coverage eases doing further modifications to the function.

Otherwise we create collisions, e.g. a struct named Foo inside an
anonymous
namespace will get the same USR no matter what the surrounding
decl-context is.

Fixes #69408

[DAGCombiner] Transform `(icmp eq/ne (and X,C0),(shift X,C1))` to use rotate or to getter constants.

If `C0` is a mask and `C1` shifts out all the masked bits (to
essentially compare two subsets of `X`), we can arbitrarily re-order
shift as `srl` or `shl`.

If `C1` (shift amount) is a power of 2, we can replace the and+shift
with a rotate.

Otherwise, based on target preference we can arbitrarily swap `shl`
and `shl` in/out to get better constants.

On x86 we can use this re-ordering to:
    1) get better `and` constants for `C0` (zero extended moves or
       avoid imm64).
    2) covert `srl` to `shl` if `shl` will be implementable with `lea`
       or `add` (both of which can be preferable).

Proofs: https://alive2.llvm.org/ce/z/qzGM_w

Reviewed By: RKSimon

Differential Revision: https://reviews.llvm.org/D152116

Differential Revision: https://reviews.llvm.org/D152115

With expensive checks enabled but without this patch, std::pop_heap
triggers an assertion failure.  This is because:

  updateAndCheckDecreased(Heap.front())

updates the priority associated with Heap.front(), so Heap may no
longer be a valid heap.  The libstdc++ version of std::pop_heap
requires that the entire range be a valid heap even though the core
task of std::pop_heap is just to swap the Heap.front() and
Heap.back().

This patch fixes the problem by:

- calling std::pop_heap to swap Heap.front() and Heap.back(),
- updating the priority of Heap.back(), and
- inserting Heap.back() back into the heap.

We could reduce the number of calls to updateAndCheckDecreased or the
number of elements being moved, but I think it's important to fix the
crash first.

Credit to Ivan Kosarev for identifying the problem and Liqiang Tao for
the analysis and initial version of the patch.

The last use was removed by:

  commit ee08897f
  Author: Reid Kleckner <reid@kleckner.net>
  Date:   Tue Dec 10 18:27:32 2013 +0000

This has very little effect on codegen in practice, but is a nice to
have I think.

See #68315

MaskRay suggested removal of the identifer in a review for another
ToolChain so doing the same for the rest.

[CodeLayout] CDSortImpl: remove HotChains and remove linear-time erase_value from mergeChains (#69276)

After mergeChainPairs initializes a priority queue, HotChains is unused
except a HotChains.size() use in LLVM_DEBUG. Optimize it out.

This patch is a bandage to get G_FRAME_INDEX working. We could import
the SelectionDAG patterns for the ComplexPattern FrameAddrRegImm, and
perhaps we will do that in the future. For now we just select it as an
addition with 0.

https://github.com/llvm/llvm-project/pull/68012/files

 added new data
formatters for LibStdC++ std::variant.

However, this formatter can crash if std::variant's index field has
invalid value (exceeds the number of template arguments).
This can happen if the current IP stops at a place std::variant is not
initialized yet.

This patch fixes the crash by ensuring the index is a valid value.

---------

Co-authored-by: jeffreytan81 <jeffreytan@fb.com>

This section is quite outdated at this point and no longer relevant
given LLVM (and CPUs) has evolved significantly over the past 20+ years.

This is also aimed at serving as a test of the llvm docs CI that was
recently added in.

46cb8d9a unconditionally changed the
mmap range to 2^48 for all riscv sanitizers. This changes the allocator
tunings for the 32-bit allocator for riscv and led to a severe
performance regression for our lsan tests. This effectively revers the
tuning change but only for Fuchsia.

Once we enable the 64-bit allocator for everything riscv, this value
will be irrelevant since it's only relevant for the 32-bit allocator.

The change to print this was made in 2014 in
35fc363c but apparently the LangRef was
never updated.

[unittest] Add lit option to allow disabling sharding in unittest

By default, googletest based unit tests uses sharding to speed up the
testing. However, when these unit tests are run through wrapper program
on a remote platform with large round trip time, the sharding will increase
the time cost dramatically. This patch adds a LLVM LIT option
"--disable-gtest-sharding" to allow sharding on gtest based unittest to
be disabled.

Some user hit this condition, where a crash isn't very friendly.

CompilerType constructors rely on the NDEBUG macro, so it's better to move them to their cpp file so that the header doesn't get confused when this macro is used differently for other compilation units.

This function always expects the pointer to be valid, a reference seems
more appropriate.

This adds back a namespace to the value of a #define statement. This is needed to prevent collision with other types with the same name.

One use of setSymbolAndType (related to https://reviews.llvm.org/D53864
"Do not crash when -r output uses linker script with `/DISCARD/`")
is no longer needed after commit 1981b1b6
demotes symbols in discarded sections to Undefined.

Outline some often used common code to dedicated variables in order
to make code compact. Rename variables to more accurately reflect
their purpose. Apply const qualifier where appropriate.
Fix and add bit more explanation comment for the existing code.

Relax the parser to accept scalar integer expression in addition to
scalar logical expression. The parser now accepts scalar expression and
the semantic checks its type.