[all-commits] [llvm/llvm-project] 1f7ddb: [NFC][Offload][OMPT] Improve readability of liboff...

Marco Elver via All-commits all-commits at lists.llvm.org
Wed Oct 22 02:24:47 PDT 2025


  Branch: refs/heads/users/melver/spr/main.clang-implement-constexpr-evaluation-for-__builtin_infer_alloc_token
  Home:   https://github.com/llvm/llvm-project
  Commit: 1f7ddb61b368c56838bb7301ddef4547d0be9682
      https://github.com/llvm/llvm-project/commit/1f7ddb61b368c56838bb7301ddef4547d0be9682
  Author: Kaloyan Ignatov <kaloyan.ignatov at outlook.com>
  Date:   2025-10-22 (Wed, 22 Oct 2025)

  Changed paths:
    M offload/test/ompt/callbacks.h
    M offload/test/ompt/omp_api.c
    M offload/test/ompt/target_memcpy.c
    M offload/test/ompt/target_memcpy_emi.c
    M offload/test/ompt/veccopy.c
    M offload/test/ompt/veccopy_data.c
    M offload/test/ompt/veccopy_disallow_both.c
    M offload/test/ompt/veccopy_emi.c
    M offload/test/ompt/veccopy_emi_map.c
    M offload/test/ompt/veccopy_map.c
    M offload/test/ompt/veccopy_no_device_init.c
    M offload/test/ompt/veccopy_wrong_return.c

  Log Message:
  -----------
  [NFC][Offload][OMPT] Improve readability of liboffload OMPT tests (#163181)

- ompt_target_data_op_t, ompt_scope_endpoint_t and ompt_target_t are now
printed as strings instead of just numbers to ease debugging
- some missing clang-format clauses have been added


  Commit: f1c1063acb12c7ce7724a88bf93351dd76b561c4
      https://github.com/llvm/llvm-project/commit/f1c1063acb12c7ce7724a88bf93351dd76b561c4
  Author: Arthur Eubanks <aeubanks at google.com>
  Date:   2025-10-22 (Wed, 22 Oct 2025)

  Changed paths:
    M llvm/lib/Analysis/InstructionSimplify.cpp
    R llvm/test/Transforms/InstCombine/select_with_identical_phi.ll

  Log Message:
  -----------
  Revert "[InstCombinePHI] Enhance PHI CSE to remove redundant phis" (#164520)

Reverts llvm/llvm-project#163453

Causes crashes, see
https://github.com/llvm/llvm-project/pull/163453#issuecomment-3429922732


  Commit: 954a7190561e12802fd143bbaf58e73f54fe58eb
      https://github.com/llvm/llvm-project/commit/954a7190561e12802fd143bbaf58e73f54fe58eb
  Author: Timm Baeder <tbaeder at redhat.com>
  Date:   2025-10-22 (Wed, 22 Oct 2025)

  Changed paths:
    M clang/lib/AST/ByteCode/Compiler.cpp
    M clang/lib/AST/ByteCode/Interp.h
    M clang/lib/AST/ByteCode/Opcodes.td
    M clang/test/AST/ByteCode/bitfields.cpp

  Log Message:
  -----------
  [clang][bytecode] Fix unsigned wraparound behavior with bitfields (#164445)

and inc/dec operations. We need to truncate the new value to the
bitfield bitwidth.


  Commit: 5afe29e0faaee44c44f1daa3e2693fc1ab212530
      https://github.com/llvm/llvm-project/commit/5afe29e0faaee44c44f1daa3e2693fc1ab212530
  Author: Luo Yuanke <lyk_03 at hotmail.com>
  Date:   2025-10-22 (Wed, 22 Oct 2025)

  Changed paths:
    M llvm/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp
    A llvm/test/CodeGen/X86/bf16-fast-isel.ll

  Log Message:
  -----------
  [FastIsel] Get the right register type for call instruction (#164565)

When switch from fast isel to dag isel the input value is from llvm IR
instruction.
If the instruction is call we should get the calling convention of the
callee and
pass it to RegsForValue::getCopyFromRegs, so that it can deduce the
right RegisterVT
of the returned value of the callee.

---------

Co-authored-by: Yuanke Luo <ykluo at birentech.com>


  Commit: 29623f1e98f4970d5b9aa62875cda5c2c66694a8
      https://github.com/llvm/llvm-project/commit/29623f1e98f4970d5b9aa62875cda5c2c66694a8
  Author: Marco Elver <elver at google.com>
  Date:   2025-10-22 (Wed, 22 Oct 2025)

  Changed paths:
    M llvm/include/llvm/Support/AllocToken.h
    M llvm/include/llvm/Transforms/Instrumentation/AllocToken.h
    M llvm/lib/Passes/PassBuilder.cpp
    M llvm/lib/Passes/PassRegistry.def
    M llvm/lib/Support/AllocToken.cpp
    M llvm/lib/Transforms/Instrumentation/AllocToken.cpp
    M llvm/test/Instrumentation/AllocToken/basic.ll
    M llvm/test/Instrumentation/AllocToken/basic32.ll
    M llvm/test/Instrumentation/AllocToken/fast.ll
    M llvm/test/Instrumentation/AllocToken/intrinsic.ll
    M llvm/test/Instrumentation/AllocToken/intrinsic32.ll
    M llvm/test/Instrumentation/AllocToken/invoke.ll
    M llvm/test/Instrumentation/AllocToken/nonlibcalls.ll
    M llvm/test/Instrumentation/AllocToken/typehashpointersplit.ll

  Log Message:
  -----------
  [AllocToken] Make token mode a pass parameter (#163634)

Refactor the AllocToken pass to accept the mode via pass options rather
than LLVM cl::opt. This is both cleaner, but also required to make the
mode frontend-driven and avoid potential inconsistencies.


  Commit: e2c0ed7accb3308dc20f05efa3ed4649a5c7fad4
      https://github.com/llvm/llvm-project/commit/e2c0ed7accb3308dc20f05efa3ed4649a5c7fad4
  Author: Marco Elver <elver at google.com>
  Date:   2025-10-22 (Wed, 22 Oct 2025)

  Changed paths:
    M clang/lib/AST/ByteCode/Compiler.cpp
    M clang/lib/AST/ByteCode/Interp.h
    M clang/lib/AST/ByteCode/Opcodes.td
    M clang/test/AST/ByteCode/bitfields.cpp
    M llvm/lib/Analysis/InstructionSimplify.cpp
    M llvm/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp
    A llvm/test/CodeGen/X86/bf16-fast-isel.ll
    R llvm/test/Transforms/InstCombine/select_with_identical_phi.ll
    M offload/test/ompt/callbacks.h
    M offload/test/ompt/omp_api.c
    M offload/test/ompt/target_memcpy.c
    M offload/test/ompt/target_memcpy_emi.c
    M offload/test/ompt/veccopy.c
    M offload/test/ompt/veccopy_data.c
    M offload/test/ompt/veccopy_disallow_both.c
    M offload/test/ompt/veccopy_emi.c
    M offload/test/ompt/veccopy_emi_map.c
    M offload/test/ompt/veccopy_map.c
    M offload/test/ompt/veccopy_no_device_init.c
    M offload/test/ompt/veccopy_wrong_return.c

  Log Message:
  -----------
  [𝘀𝗽𝗿] changes introduced through rebase

Created using spr 1.3.8-beta.1

[skip ci]


Compare: https://github.com/llvm/llvm-project/compare/d46e27506b42...e2c0ed7accb3

To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications


More information about the All-commits mailing list