This is a complete Options.td compiled from ld(1) dated 2018-03-07 and
cross checked with ld64 source code version 512.4 dated 2018-03-18.

This is the first in a series of diffs for argument handling. Follow-ups
will include switch cases for all the new instances of `OPT_foo`, and
parsing/validation of arguments attached to options, e.g., more code
akin to `OPT_platform_version` and associated `parsePlatformVersion()`.

Reviewed By: smeenai

Differential Revision: https://reviews.llvm.org/D80582

This wraps the uint8_t[12] type used in two places, because I
plan to introduce a third use of the same pattern.

The case the calculation works for is when r_length = 2.

Summary:
If a record has a mix of relative pointers and other fields they
wouldn't necessarily be the same.

Fallout from D77592.

rdar://64309883

Subscribers: cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D81857

Add selection support for ext via a new opcode, G_EXT and a post-legalizer
combine which matches it.

Add an `applyEXT` function, because the AArch64ext patterns require a register
for the immediate. So, we have to create a G_CONSTANT to get these without
writing new patterns or modifying the existing ones.

Tests are the same as arm64-ext.ll.

Also prevent ext from firing on the zip test. It has higher priority, so we
don't want it potentially getting in the way of mask tests.

Also fix up the shuffle-splat test, because ext is now selected there. The
test was incorrectly regbank selected before, which could cause a verifier
failure when you emit copies.

Differential Revision: https://reviews.llvm.org/D81436

This reverts commit 695c7d63.

Breaks windows (e.g.
http://lab.llvm.org:8011/builders/clang-x64-windows-msvc/builds/16497)

Likely to cause problems with XCode.

Pointed out by dblaikie.

Summary:
Currently, add_llvm_library would create an OBJECT library alongside
of a STATIC / SHARED library, but losing the link interface (its
elements would become dependencies instead). To support scenarios
where linking an object library also brings in its usage
requirements, this patch adds support for 'stand-alone' OBJECT
libraries - i.e. without an accompanying SHARED/STATIC library, and
maintaining the link interface defined by the user.

The support is via a new option, OBJECT_ONLY, to avoid breaking changes
- since just specifying "OBJECT" would currently imply also STATIC or
SHARED, depending on BUILD_SHARED_LIBS.

This is useful for cases where, for example, we want to build a part
of a component separately. Using a STATIC target would incur the risk
that symbols not referenced in the consumer would be dropped (which may
be undesirable).

The current application is the ML part of Analysis. It should be part
of the Analysis component, so it may reference other analyses; and (in
upcoming changes) it has dependencies on optional libraries.

Reviewers: karies, davidxl

Subscribers: mgorny, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D81447

This is a somewhat artifical limit, so avoid repeating it many places
in case it changes.

This is useful for checking runtime properties of the target system.
This is a partial re-application of 3ea9450b. This part was tested
to work on a Windows host with a SSH executor.

This reverts commit 32c757e4.

Broke the build bot:

******************** TEST 'MLIR :: Examples/standalone/test.toy' FAILED ********************
[...]
/tmp/ci-KIMiRFcVZt/lib/libMLIRLinalgToLLVM.a(LinalgToLLVM.cpp.o): In function `(anonymous namespace)::ConvertLinalgToLLVMPass::runOnOperation()':
LinalgToLLVM.cpp:(.text._ZN12_GLOBAL__N_123ConvertLinalgToLLVMPass14runOnOperationEv+0x100): undefined reference to `mlir::populateExpandTanhPattern(mlir::OwningRewritePatternList&, mlir::MLIRContext*)'

Change isa<> to a variadic function template, so that it can be used to test against one of multiple types as follows:
   isa<Type0, Type1, Type2>(Val)

Differential Revision: https://reviews.llvm.org/D81045

Pass in all discovered tests to report generators.

The XunitReport generator now creates testcase items for unexecuted
tests and documents why they have been skipped.  This makes it easier
to compare test runs with different filters or configurations, or across
platforms.

I don't know who is using the JsonReport generator and what the
expectations there are (it doesn't have tests), so decided to preserve
the old behavior by filtering out the unexecuted tests.

Reviewed By: jdenny

Differential Revision: https://reviews.llvm.org/D81316

Commit 498dd745 introduced a dependence on Core. This patch updates
LLVMbuild.txt to reflect this.

Summary:
This CL introduces an integration test directory for MLIR in general, with
vector dialect integration tests in particular as a first working suite. To
run all the integration tests (and currently just the vector suite):

$ cmake --build . --target check-mlir-integration
[0/1] Running the MLIR integration tests
Testing Time: 0.24s
Passed: 22

The general call is to contribute to this integration test directory with more
tests and other suites, running end-to-end examples that may be too heavy for
the regular test directory, but should be tested occasionally to verify the
health of MLIR.

Background discussion at:
https://llvm.discourse.group/t/vectorops-rfc-add-suite-of-integration-tests-for-vector-dialect-operations/1213/

Reviewers: nicolasvasilache, reidtatge, andydavis1, rriddle, ftynse, mehdi_amini, jpienaar, stephenneuendorffer

Reviewed By: nicolasvasilache, stephenneuendorffer

Subscribers: mgorny, mehdi_amini, rriddle, jpienaar, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, liufengdb, stephenneuendorffer, Joonsoo, grosul1, frgossen, Kayjukh, jurahul, msifontes

Tags: #mlir

Differential Revision: https://reviews.llvm.org/D81626

Fix the copy/paste mistake that caused it to fail previously

Port partial constant store merging logic to MemorySSA backed DSE. The
heavy lifting is done by the existing helper function. It is used in
context where we already ensured that the later instruction can
eliminate the earlier one, if it is a complete overwrite.

Summary:
Add a pattern for expanding tanh op into exp form.
A `tanh` is expanded into:
   1) 1-exp^{-2x} / 1+exp^{-2x}, if x => 0
   2) exp^{2x}-1 / exp^{2x}+1  , if x < 0.

Differential Revision: https://reviews.llvm.org/D81618

If a symbol name begins with the linker private global prefix (as
described by the DataLayout) then it should be treated as non-exported,
regardless of its LLVM IR visibility value.

This is necessary for standalone builds where the libc++ in use has a
custom configuration set up inside __config_site -- one needs to build
libc++abi against the installed headers of libc++ (which are properly
configured) instead of the ones inside libcxx/include.

See https://reviews.llvm.org/rGe619e9d#927848 for details.

This adds 4 new reloc types.

A lot of code that previously assumed any memory or offset values could be contained in a uint32_t (and often truncated results from functions returning 64-bit values) have been upgraded to uint64_t. This is not comprehensive: it is only the values that come in contact with the new relocation values and their dependents.

A new tablegen mapping was added to automatically upgrade loads/stores in the assembler, which otherwise has no way to select for these instructions (since they are indentical other than for the offset immediate). It follows a similar technique to https://reviews.llvm.org/D53307

Differential Revision: https://reviews.llvm.org/D81704

Summary:
The following intrinsics have been added, guarded by the macro
`__ARM_FEATURE_SVE_MATMUL_FP64`:

* svtrn1q[_*]
* svtrn2q[_*]
* svuzp1q[_*]
* svuzp2q[_*]
* svzip1q[_*]
* svzip2q[_*]

Supported types:

* svint[8|16|32|64]_t
* svuint[8|16|32|64]_t
* svfloat[16|32|64]_t

TODO: add support for svbfloat16_t

Reviewers: efriedma, sdesmalen, kmclaughlin, rengolin

Reviewed By: sdesmalen

Subscribers: tschuett, kristof.beyls, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D80851

Reviewed by: AndreyChurbanov

Differential Revision: https://reviews.llvm.org/D81804

[X86] Remove printanymem/printopaquemem from the InstPrinters. Just tell tablegen to printMemReference directly. NFC

Most of the wrappers exist to print the memory size in Intel syntax
and then call the printMemReference. But printanymem/printopaquemem
don't print anything extra in Intel syntax so just drop them.

Hoist the general logic into a new function, because it can be re-used
by the MemorySSA backed DSE as well.

This implements the following combines:

((0-A) + B) -> B-A
(A + (0-B)) -> A-B

Porting over the basic algebraic combines from the DAGCombiner. There are
several combines which fold adds away into subtracts. This is just the simplest
one.

I noticed that add combines are some of the most commonly hit across CTMark,
(via print statements when they fire), so I'm porting over some of the obvious
ones.

This gives some minor code size improvements on CTMark at -O3 on AArch64.

Differential Revision: https://reviews.llvm.org/D77453

Review comments addressed.

This patch remove the indentation before the command help output.
Supposedly it was meant to be aligned with the different subcommands.

Differential revision: https://reviews.llvm.org/D81783

Along that way, platform specific options to memcpy, memset and bzero
builds have been enclosed in conditionals. Also, the optimization level
has been set to -O2 for the memory function builds to actually see the
static functions inlined.

Reviewers: gchatelet

Differential Revision: https://reviews.llvm.org/D81621

Summary:
Adding intrinsics and codegen patterns for:

* trn1 <Zd>.q, <Zm>.q, <Zn>.q
* trn2 <Zd>.q, <Zm>.q, <Zn>.q
* zip1 <Zd>.q, <Zm>.q, <Zn>.q
* zip2 <Zd>.q, <Zm>.q, <Zn>.q
* uzp1 <Zd>.q, <Zm>.q, <Zn>.q
* uzp2 <Zd>.q, <Zm>.q, <Zn>.q

These instructions are defined in Armv8.6-A.

Reviewers: sdesmalen, efriedma, kmclaughlin

Reviewed By: sdesmalen

Subscribers: tschuett, kristof.beyls, hiraditya, rkruppe, psnobl, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D80850

Reviewers: sscalpone

Subscribers: llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D79507

This allows reporting issues early when creating feature names.

This is a follow-up to D81845.

Summary:
The initial implementation of typeHierarchy/resolve only supported
cases where an initial request was made for children, and then
typeHierarchy/resolve was used to get additional levels of children.

However, a client may also want to make an initial request for
parents, and then show other children of those parents, so support
typeHierarchy/resolve for items returned in response to a request
for parents as well.

Subscribers: ilya-biryukov, MaskRay, jkorous, arphaman, kadircet, usaxena95, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D81845