Changes

Summary

  1. [mlir] Apply py::module_local() to a few more classes. (details)
  2. [test] precommit test for D109786 (details)
  3. [test] precommit anoter test for D109786 (details)
  4. NFC: DebugInfo: refactor pretty printing into a utility class (details)
  5. Remove unused llvm/Support/Parallel.h from MLIR (NFC) (details)
  6. [flang] Make 'this_image()' an intrinsic function (details)
  7. Attributor: Fix typos (details)
  8. AMDGPU: Rename attributor class for uniform-work-group-size (details)
  9. Attributor: Fix crash on undef in !callees (details)
  10. SeparateConstOffsetFromGEP: Fix stack overflow in unreachable code (details)
  11. [mlir][linalg] PadTensorOp vectorization: Avoid redundant FillOps (details)
Commit 8dca953dd39c0cd8c80decbeb38753f58a4de580 by silvasean
[mlir] Apply py::module_local() to a few more classes.

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D109776
The file was modifiedmlir/lib/Bindings/Python/IRModule.h
The file was modifiedmlir/lib/Bindings/Python/PybindUtils.h
The file was modifiedmlir/lib/Bindings/Python/DialectSparseTensor.cpp
Commit 162aed482403a4a7d54930d73a4d74e69e1e79db by listmail
[test] precommit test for D109786
The file was addedllvm/test/Analysis/ScalarEvolution/mul.ll
Commit baff4b41055fc2d75bcdcea7bb39edce38a40acf by listmail
[test] precommit anoter test for D109786
The file was modifiedllvm/test/Analysis/ScalarEvolution/mul.ll
Commit 4cabaf594aa272b02c0d227ae16a2aced687d796 by dblaikie
NFC: DebugInfo: refactor pretty printing into a utility class

Laying more foundation for full template name rebuilding - more complex
type printing benefits from an object to carry some state rather than
passing it around as parameters to every function.
The file was modifiedllvm/lib/DebugInfo/DWARF/DWARFDie.cpp
Commit 1a406cd5f2e2403b325f6d914b692459a2ab3b9b by joker.eph
Remove unused llvm/Support/Parallel.h from MLIR (NFC)

This header aren't needed anymore: MLIR is using a thread pool
injected in the context instead of a global one.
The file was modifiedmlir/lib/Pass/Pass.cpp
The file was modifiedmlir/lib/Pass/PassCrashRecovery.cpp
The file was modifiedmlir/lib/IR/Verifier.cpp
The file was modifiedmlir/lib/Transforms/Inliner.cpp
Commit 81f8ad1769665a569a235b749e0e9e69ce7dc65e by erasmus
[flang] Make 'this_image()' an intrinsic function

Added 'this_image()' to the list of functions that are evaluated as intrinsic.
Added IsCoarray functions to determine if an expression is a coarray (corank > 1).

Added save attribute to coarray variables in test file, this_image.f90.

reviewers: klausler, PeteSteinfeld

Differential Revision: https://reviews.llvm.org/D108059
The file was modifiedflang/include/flang/Evaluate/tools.h
The file was modifiedflang/lib/Evaluate/tools.cpp
The file was modifiedflang/test/Semantics/call10.f90
The file was modifiedflang/lib/Evaluate/intrinsics.cpp
The file was modifiedflang/docs/Intrinsics.md
The file was addedflang/test/Semantics/this_image.f90
Commit 99dfe90695a811f74fb7503703ffd52bd214dd2e by Matthew.Arsenault
Attributor: Fix typos
The file was modifiedllvm/include/llvm/Transforms/IPO/Attributor.h
Commit f12174204c639f9780f17cf7b8e910be703b6b8c by Matthew.Arsenault
AMDGPU: Rename attributor class for uniform-work-group-size

This isn't really an AMDGPU specific attribute and could be moved to
generic code. It's also important to include the word uniform in the
name.
The file was modifiedllvm/lib/Target/AMDGPU/AMDGPUAttributor.cpp
Commit fdd9761dd1a971c9f4d6776b511ea54d7765bfeb by Matthew.Arsenault
Attributor: Fix crash on undef in !callees
The file was modifiedllvm/lib/Transforms/IPO/AttributorAttributes.cpp
The file was modifiedllvm/test/Transforms/Attributor/callgraph.ll
Commit 88146230e1b21aa042da481e5fd702fab82408fc by Matthew.Arsenault
SeparateConstOffsetFromGEP: Fix stack overflow in unreachable code

ConstantOffsetExtractor::Find was infinitely recursing on the add
referencing itself.
The file was modifiedllvm/lib/Transforms/Scalar/SeparateConstOffsetFromGEP.cpp
The file was addedllvm/test/Transforms/SeparateConstOffsetFromGEP/crash-in-unreachable-code.ll
Commit 9adc0114bfeb704ca62d8c369fa52d0530179274 by springerm
[mlir][linalg] PadTensorOp vectorization: Avoid redundant FillOps

Do not generate FillOps when these would be entirely overwritten.

Differential Revision: https://reviews.llvm.org/D109741
The file was modifiedmlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp