LLVM 18.1.0rc Release Notes

Introduction

This document contains the release notes for the LLVM Compiler Infrastructure, release 18.1.0rc. Here we describe the status of LLVM, including major improvements from the previous release, improvements in various subprojects of LLVM, and some of the current users of the code. All LLVM releases may be downloaded from the LLVM releases web site.

For more information about LLVM, including information about the latest release, please check out the main LLVM web site. If you have questions or comments, the Discourse forums is a good place to ask them.

Note that if you are reading this file from a Git checkout or the main LLVM web page, this document applies to the next release, not the current one. To see the release notes for a specific release, please see the releases page.

Non-comprehensive list of changes in this release

Update on required toolchains to build LLVM

Changes to the LLVM IR

  • The llvm.stacksave and llvm.stackrestore intrinsics now use an overloaded pointer type to support non-0 address spaces.

  • The constant expression variants of the following instructions have been removed:

    • and

    • or

    • lshr

    • ashr

    • zext

    • sext

    • fptrunc

    • fpext

    • fptoui

    • fptosi

    • uitofp

    • sitofp

  • Added llvm.exp10 intrinsic.

  • Added a code_model attribute for the global variable.

Changes to LLVM infrastructure

  • Minimum Clang version to build LLVM in C++20 configuration has been updated to clang-17.0.6.

Changes to building LLVM

Changes to TableGen

Changes to Interprocedural Optimizations

Changes to the AArch64 Backend

  • Added support for Cortex-A520, Cortex-A720 and Cortex-X4 CPUs.

  • Neoverse-N2 was incorrectly marked as an Armv8.5a core. This has been changed to an Armv9.0a core. However, crypto options are not enabled by default for Armv9 cores, so -mcpu=neoverse-n2+crypto is now required to enable crypto for this core. As far as the compiler is concerned, Armv9.0a has the same features enabled as Armv8.5a, with the exception of crypto.

  • Assembler/disassembler support has been added for 2023 architecture extensions.

  • Support has been added for Stack Clash Protection. During function frame creation and dynamic stack allocations, the compiler will issue memory accesses at reguilar intervals so that a guard area at the top of the stack can’t be skipped over.

Changes to the AMDGPU Backend

  • llvm.sqrt.f32 is now lowered correctly. Use llvm.amdgcn.sqrt.f32 for raw instruction access.

  • Implemented llvm.stacksave and llvm.stackrestore intrinsics.

  • Implemented llvm.get.rounding

  • The default AMDHSA code object version is now 5.

Changes to the ARM Backend

  • Added support for Cortex-M52 CPUs.

  • Added execute-only support for Armv6-M.

Changes to the AVR Backend

Changes to the DirectX Backend

Changes to the Hexagon Backend

Changes to the LoongArch Backend

  • Added intrinsics support for all LSX (128-bits SIMD) and LASX (256-bits SIMD) instructions.

  • Added definition and intrinsics support for new instructions that were introduced in LoongArch Reference Manual V1.10.

  • Emitted adjacent pcaddu18i+jirl instrunction sequence with one relocation R_LARCH_CALL36 instead of pcalau12i+jirl with two relocations R_LARCH_PCALA_{HI20,LO12} for function call in medium code model.

  • The code model of global variables can now be overridden by means of the newly added LLVM IR attribute, code_model.

  • Added support for the llvm.is.fpclass intrinsic.

  • mulodi4 and muloti4 libcalls were disabled due to absence in libgcc.

  • Added initial support for auto vectorization.

  • Added initial support for linker relaxation.

  • Assorted codegen improvements.

Changes to the MIPS Backend

Changes to the PowerPC Backend

  • LLJIT’s JIT linker now defaults to JITLink on 64-bit ELFv2 targets.

  • Initial-exec TLS model is supported on AIX.

  • Implemented new resource based scheduling model of POWER7 and POWER8.

  • frexp libcall now references correct symbol name for fp128.

  • Optimized materialization of 64-bit immediates, code generation of vec_promote and atomics.

  • Global constant strings are pooled in the TOC under one entry to reduce the number of entries in the TOC.

  • Added a number of missing Power10 extended mnemonics.

  • Added the SCV instruction.

  • Fixed register class for the paddi instruction.

  • Optimize VPERM and fix code order for swapping vector operands on LE.

  • Added various bug fixes and code gen improvements.

AIX Support/improvements:

  • Support for a non-TOC-based access sequence for the local-exec TLS model (called small local-exec).

  • XCOFF toc-data peephole optimization and bug fixes.

  • Move less often used __ehinfo TOC entries to the end of the TOC section.

  • Fixed problems when the AIX libunwind unwinds starting from a signal handler and the function that raised the signal happens to be a leaf function that shares the stack frame with its caller or a leaf function that does not store the stack frame backchain.

Changes to the RISC-V Backend

  • The Zfa extension version was upgraded to 1.0 and is no longer experimental.

  • Zihintntl extension version was upgraded to 1.0 and is no longer experimental.

  • Intrinsics were added for Zk*, Zbb, and Zbc. See https://github.com/riscv-non-isa/riscv-c-api-doc/blob/master/riscv-c-api.md#scalar-bit-manipulation-extension-intrinsics

  • Default ABI with F but without D was changed to ilp32f for RV32 and to lp64f for RV64.

  • The Zvbb, Zvbc, Zvkb, Zvkg, Zvkn, Zvknc, Zvkned, Zvkng, Zvknha, Zvknhb, Zvks, Zvksc, Zvksed, Zvksg, Zvksh, and Zvkt extension version was upgraded to 1.0 and is no longer experimental. However, the C intrinsics for these extensions are still experimental. To use the C intrinsics for these extensions, -menable-experimental-extensions needs to be passed to Clang.

  • XSfcie extension and SiFive CSRs and instructions that were associated with it have been removed. None of these CSRs and instructions were part of “SiFive Custom Instruction Extension” as SiFive defines it. The LLVM project needs to work with SiFive to define and document real extension names for individual CSRs and instructions.

  • -mcpu=sifive-p450 was added.

  • CodeGen of RV32E/RV64E was supported experimentally.

  • CodeGen of ilp32e/lp64e was supported experimentally.

  • Support was added for the Ziccif, Ziccrse, Ziccamoa, Zicclsm, Za64rs, Za128rs and Zic64b extensions which were introduced as a part of the RISC-V Profiles specification.

  • The Smepmp 1.0 extension is now supported.

  • -mcpu=sifive-p670 was added.

  • Support for the Zicond extension is no longer experimental.

Changes to the WebAssembly Backend

Changes to the Windows Target

Changes to the X86 Backend

  • The i128 type now matches GCC and clang’s __int128 type. This mainly benefits external projects such as Rust which aim to be binary compatible with C, but also fixes code generation where LLVM already assumed that the type matched and called into libgcc helper functions.

  • Support ISA of USER_MSR.

  • Support ISA of AVX10.1-256 and AVX10.1-512.

  • -mcpu=pantherlake and -mcpu=clearwaterforest are now supported.

  • -mapxf is supported.

  • Marking global variables with code_model = "small"/"large" in the IR now overrides the global code model to allow 32-bit relocations or require 64-bit relocations to the global variable.

  • The medium code model’s code generation was audited to be more similar to the small code model where possible.

Changes to the OCaml bindings

Changes to the Python bindings

  • The python bindings have been removed.

Changes to the C API

  • Added LLVMGetTailCallKind and LLVMSetTailCallKind to allow getting and setting tail, musttail, and notail attributes on call instructions.

  • The following functions for creating constant expressions have been removed, because the underlying constant expressions are no longer supported. Instead, an instruction should be created using the LLVMBuildXYZ APIs, which will constant fold the operands if possible and create an instruction otherwise:

    • LLVMConstAnd

    • LLVMConstOr

    • LLVMConstLShr

    • LLVMConstAShr

    • LLVMConstZExt

    • LLVMConstSExt

    • LLVMConstZExtOrBitCast

    • LLVMConstSExtOrBitCast

    • LLVMConstIntCast

    • LLVMConstFPTrunc

    • LLVMConstFPExt

    • LLVMConstFPToUI

    • LLVMConstFPToSI

    • LLVMConstUIToFP

    • LLVMConstSIToFP

    • LLVMConstFPCast

  • Added LLVMCreateTargetMachineWithOptions, along with helper functions for an opaque option structure, as an alternative to LLVMCreateTargetMachine. The option structure exposes an additional setting (i.e., the target ABI) and provides default values for unspecified settings.

  • Added LLVMGetNNeg and LLVMSetNNeg for getting/setting the new nneg flag on zext instructions, and LLVMGetIsDisjoint and LLVMSetIsDisjoint for getting/setting the new disjoint flag on or instructions.

  • Added the following functions for manipulating operand bundles, as well as building call and invoke instructions that use operand bundles:

    • LLVMBuildCallWithOperandBundles

    • LLVMBuildInvokeWithOperandBundles

    • LLVMCreateOperandBundle

    • LLVMDisposeOperandBundle

    • LLVMGetNumOperandBundles

    • LLVMGetOperandBundleAtIndex

    • LLVMGetNumOperandBundleArgs

    • LLVMGetOperandBundleArgAtIndex

    • LLVMGetOperandBundleTag

  • Added LLVMGetFastMathFlags and LLVMSetFastMathFlags for getting/setting the fast-math flags of an instruction, as well as LLVMCanValueUseFastMathFlags for checking if an instruction can use such flags

Changes to the CodeGen infrastructure

  • A new debug type isel-dump is added to show only the SelectionDAG dumps after each ISel phase (i.e. -debug-only=isel-dump). This new debug type can be filtered by function names using -filter-print-funcs=<function names>, the same flag used to filter IR dumps after each Pass. Note that the existing -debug-only=isel will take precedence over the new behavior and print SelectionDAG dumps of every single function regardless of -filter-print-funcs’s values.

  • PrologEpilogInserter no longer supports register scavenging during forwards frame index elimination. Targets should use backwards frame index elimination instead.

  • RegScavenger no longer supports forwards register scavenging. Clients should use backwards register scavenging instead, which is preferred because it does not depend on accurate kill flags.

Changes to the Metadata Info

  • Added a new loop metadata !{!”llvm.loop.align”, i32 64}

Changes to the Debug Info

Changes to the LLVM tools

  • llvm-symbolizer now treats invalid input as an address for which source information is not found.

  • Fixed big-endian support in llvm-symbolizer’s DWARF location parser.

  • llvm-readelf now supports --extra-sym-info (-X) to display extra information (section name) when showing symbols.

  • llvm-nm now supports the --line-numbers (-l) option to use debugging information to print symbols’ filenames and line numbers.

  • llvm-symbolizer and llvm-addr2line now support addresses specified as symbol names.

  • llvm-objcopy now supports --gap-fill and --pad-to options, for ELF input and binary output files only.

  • Supported parsing XCOFF auxiliary symbols in obj2yaml.

  • llvm-ranlib now supports -X on AIX to specify the type of object file ranlib should examine.

  • llvm-nm now supports --export-symbol to ignore the import symbol file.

  • llvm-rc and llvm-windres now accept file path references in .rc files concatenated from multiple string literals.

  • The llvm-windres option --preprocessor now resolves its argument in the PATH environment variable as expected, and options passed with --preprocessor-arg are placed before the input file as they should be.

  • The llvm-windres option --preprocessor has been updated with the breaking behaviour change from GNU windres from binutils 2.36, where the whole argument is considered as one path, not considered as a sequence of tool name and parameters.

Changes to LLDB

  • SBWatchpoint::GetHardwareIndex is deprecated and now returns -1 to indicate the index is unavailable.

  • Methods in SBHostOS related to threads have had their implementations removed. These methods will return a value indicating failure.

  • SBType::FindDirectNestedType function is added. It’s useful for formatters to quickly find directly nested type when it’s known where to search for it, avoiding more expensive global search via SBTarget::FindFirstType.

  • lldb-vscode was renamed to lldb-dap and and its installation instructions have been updated to reflect this. The underlying functionality remains unchanged.

  • The mte_ctrl register can now be read from AArch64 Linux core files.

  • LLDB on AArch64 Linux now supports debugging the Scalable Matrix Extension (SME) and Scalable Matrix Extension 2 (SME2) for both live processes and core files. For details refer to the AArch64 Linux documentation.

  • LLDB now supports symbol and binary acquisition automatically using the DEBUFINFOD protocol. The standard mechanism of specifying DEBUFINOD servers in the DEBUGINFOD_URLS environment variable is used by default. In addition, users can specify servers to request symbols from using the LLDB setting plugin.symbol-locator.debuginfod.server_urls, override or adding to the environment variable.

  • When running on AArch64 Linux, lldb-server now provides register field information for the following registers: cpsr, fpcr, fpsr, svcr and mte_ctrl.

    (lldb) register read cpsr
          cpsr = 0x80001000
               = (N = 1, Z = 0, C = 0, V = 0, SS = 0, IL = 0, <...>
    

    This is only available when lldb is built with XML support. Where possible the CPU’s capabilities are used to decide which fields are present, however this is not always possible or entirely accurate. If in doubt, refer to the numerical value.

  • On Windows, LLDB can now read the thread names.

Changes to Sanitizers

  • HWASan now defaults to detecting use-after-scope bugs.

Changes to the Profile Runtime

  • Public header profile/instr_prof_interface.h is added to declare four API functions to fine tune profile collection.

Other Changes

  • The Flags field of llvm::opt::Option has been split into Flags and Visibility to simplify option sharing between various drivers (such as clang, clang-cl, or flang) that rely on Clang’s Options.td. Overloads of llvm::opt::OptTable that use FlagsToInclude have been deprecated. There is a script and instructions on how to resolve conflicts - see https://reviews.llvm.org/D157150 and https://reviews.llvm.org/D157151 for details.

  • On Linux, FreeBSD, and NetBSD, setting the environment variable LLVM_ENABLE_SYMBOLIZER_MARKUP causes tools to print stacktraces using Symbolizer Markup. This works even if the tools have no embedded symbol information (i.e. are fully stripped); llvm-symbolizer can symbolize the markup afterwards using debuginfod.

External Open Source Projects Using LLVM 15

  • A project…

Additional Information

A wide variety of additional information is available on the LLVM web page, in particular in the documentation section. The web page also contains versions of the API documentation which is up-to-date with the Git version of the source code. You can access versions of these documents specific to this release by going into the llvm/docs/ directory in the LLVM tree.

If you have any questions or comments about LLVM, please feel free to contact us via the Discourse forums.