Progress:
In progressChanges

Changes from Git (git http://labmaster3.local/git/llvm-project.git)

Summary

  1. [globalopt] Change so that emitting fragments doesn't use the type size of DIVariables (details)
  2. Revert "[Instruction] Add updateLocationAfterHoist helper" (details)
  3. [libcxx/variant] Add a few benchmarks for `std::visit`. (details)
  4. [flang] Fix interpretation of intrinsic names as arguments (details)
  5. [Diagnostics] Reworked -Wstring-concatenation (details)
  6. [WebAssembly][AsmParser] Name missing features in error message (details)
  7. [VectorCombine] add safety check for 0-width register (details)
Commit 54b6cca0f28484395ae43bcda4c9f929bc51cfe3 by akhuang
[globalopt] Change so that emitting fragments doesn't use the type size of DIVariables

When turning on -debug-info-kind=constructor we ran into a "fragment covers
entire variable" error during thinlto. The fragment is currently always
emitted if there is no type size, but sometimes the variable has a
forward declared struct type which doesn't have a size.

This changes the code to get the type size from the GlobalVariable instead.

Differential Revision: https://reviews.llvm.org/D85572
The file was addedllvm/test/DebugInfo/Generic/global-sra-struct-fwd-decl.ll
The file was modifiedllvm/lib/Transforms/IPO/GlobalOpt.cpp
Commit 30c1633386e7cfb01c0a54b31ccf4c3a3873e71b by Vedant Kumar
Revert "[Instruction] Add updateLocationAfterHoist helper"

This reverts commit 4a646ca9e2caf70d6312714770f516fb83b7e3cb.

This is causing some bots to fail with "!dbg attachment points at wrong
subprogram for function", like:

http://lab.llvm.org:8011/builders/sanitizer-windows/builds/67958/steps/stage%201%20check/logs/stdio
The file was modifiedllvm/test/Transforms/LICM/hoisting-preheader-debugloc.ll
The file was modifiedllvm/test/Transforms/GVN/PRE/phi-translate.ll
The file was modifiedllvm/unittests/IR/InstructionsTest.cpp
The file was modifiedllvm/lib/IR/DebugInfo.cpp
The file was modifiedllvm/lib/Transforms/Scalar/LICM.cpp
The file was modifiedllvm/test/DebugInfo/Generic/licm-hoist-debug-loc.ll
The file was modifiedllvm/include/llvm/IR/Instruction.h
The file was modifiedllvm/lib/Transforms/Scalar/GVN.cpp
Commit c6f51377e12b343ca2f712b6af6908555f07ad04 by mcypark
[libcxx/variant] Add a few benchmarks for `std::visit`.

This patch adds a few `std::visit` benchmarks as a starting point.

Reviewed By: ldionne, #libc

Differential Revision: https://reviews.llvm.org/D85419
The file was addedlibcxx/benchmarks/variant_visit_3.bench.cpp
The file was addedlibcxx/benchmarks/VariantBenchmarks.h
The file was addedlibcxx/benchmarks/variant_visit_2.bench.cpp
The file was addedlibcxx/benchmarks/variant_visit_1.bench.cpp
Commit f5e4451e1f27e87d30c4eb0c3c99f2f8d07039e1 by tkeith
[flang] Fix interpretation of intrinsic names as arguments

If an unrestricted specific intrinsic function name is first encountered
as an actual argument, it should be interpreted as an object entity,
not a procedure entity.

Fix some tests that depended on the previous interpretation by adding
explicit INTRINSIC statements.

Differential Revision: https://reviews.llvm.org/D85792
The file was modifiedflang/test/Semantics/call09.f90
The file was modifiedflang/test/Semantics/call02.f90
The file was modifiedflang/lib/Semantics/resolve-names.cpp
Commit b9af72bffe5f2769f3a7858a785981f89137a0ce by Dávid Bolvanský
[Diagnostics] Reworked -Wstring-concatenation
The file was modifiedclang/lib/Sema/SemaDecl.cpp
The file was modifiedclang/test/Sema/string-concat.c
The file was modifiedclang/lib/Sema/SemaExpr.cpp
Commit 2985c02f798ba8248a9168a9f33a74d90c0c5445 by tlively
[WebAssembly][AsmParser] Name missing features in error message

Rather than just saying that some feature is missing, report the exact
features to make the error message more useful and actionable.

Differential Revision: https://reviews.llvm.org/D85795
The file was modifiedllvm/lib/Target/WebAssembly/AsmParser/WebAssemblyAsmParser.cpp
The file was addedllvm/test/MC/WebAssembly/missing-features.s
Commit b0b95dab1ce21d93f3d62bc37256da9f38cff616 by spatel
[VectorCombine] add safety check for 0-width register

Based on post-commit discussion in D81766, Hexagon sets this to "0".
I'll see if I can come up with a test, but making the obvious
code fix first to unblock that target.
The file was modifiedllvm/lib/Transforms/Vectorize/VectorCombine.cpp