[all-commits] [llvm/llvm-project] 90ba33: [InstCombine] Canonicalize constant GEPs to i8 sou...

Wed Jan 24 06:25:42 PST 2024

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 90ba33099cbb17e7c159e9ebc5a512037db99d6d
      https://github.com/llvm/llvm-project/commit/90ba33099cbb17e7c159e9ebc5a512037db99d6d
  Author: Nikita Popov <npopov at redhat.com>
  Date:   2024-01-24 (Wed, 24 Jan 2024)

  Changed paths:
    M clang/test/CodeGen/PowerPC/builtins-ppc-pair-mma.c
    M clang/test/CodeGen/aarch64-ls64-inline-asm.c
    M clang/test/CodeGen/attr-arm-sve-vector-bits-bitcast.c
    M clang/test/CodeGen/attr-counted-by.c
    M clang/test/CodeGen/attr-riscv-rvv-vector-bits-bitcast.c
    M clang/test/CodeGen/cleanup-destslot-simple.c
    M clang/test/CodeGen/hexagon-brev-ld-ptr-incdec.c
    M clang/test/CodeGen/ms-intrinsics.c
    M clang/test/CodeGen/nofpclass.c
    M clang/test/CodeGen/union-tbaa1.c
    M clang/test/CodeGenCXX/RelativeVTablesABI/dynamic-cast.cpp
    M clang/test/CodeGenCXX/RelativeVTablesABI/type-info.cpp
    M clang/test/CodeGenCXX/microsoft-abi-dynamic-cast.cpp
    M clang/test/CodeGenCXX/microsoft-abi-typeid.cpp
    M clang/test/CodeGenObjC/arc-foreach.m
    M clang/test/CodeGenObjCXX/arc-cxx11-init-list.mm
    M clang/test/Headers/__clang_hip_math.hip
    M clang/test/OpenMP/bug57757.cpp
    M flang/test/HLFIR/no-block-merging.fir
    M llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
    M llvm/test/Analysis/BasicAA/featuretest.ll
    M llvm/test/CodeGen/AMDGPU/vector-alloca-bitcast.ll
    M llvm/test/CodeGen/BPF/preserve-static-offset/load-inline.ll
    M llvm/test/CodeGen/BPF/preserve-static-offset/load-unroll-inline.ll
    M llvm/test/CodeGen/BPF/preserve-static-offset/load-unroll.ll
    M llvm/test/CodeGen/BPF/preserve-static-offset/store-unroll-inline.ll
    M llvm/test/CodeGen/Hexagon/autohvx/vector-align-tbaa.ll
    M llvm/test/Transforms/Coroutines/coro-async.ll
    M llvm/test/Transforms/Coroutines/coro-retcon-alloca-opaque-ptr.ll
    M llvm/test/Transforms/Coroutines/coro-retcon-alloca.ll
    M llvm/test/Transforms/Coroutines/coro-retcon-once-value.ll
    M llvm/test/Transforms/Coroutines/coro-retcon-resume-values.ll
    M llvm/test/Transforms/Coroutines/coro-swifterror.ll
    M llvm/test/Transforms/InstCombine/2007-03-25-BadShiftMask.ll
    M llvm/test/Transforms/InstCombine/2009-01-08-AlignAlloca.ll
    M llvm/test/Transforms/InstCombine/2009-02-20-InstCombine-SROA.ll
    M llvm/test/Transforms/InstCombine/X86/x86-addsub-inseltpoison.ll
    M llvm/test/Transforms/InstCombine/X86/x86-addsub.ll
    M llvm/test/Transforms/InstCombine/add3.ll
    M llvm/test/Transforms/InstCombine/array.ll
    M llvm/test/Transforms/InstCombine/assume.ll
    M llvm/test/Transforms/InstCombine/cast_phi.ll
    M llvm/test/Transforms/InstCombine/catchswitch-phi.ll
    M llvm/test/Transforms/InstCombine/compare-alloca.ll
    M llvm/test/Transforms/InstCombine/extractvalue.ll
    M llvm/test/Transforms/InstCombine/gep-addrspace.ll
    M llvm/test/Transforms/InstCombine/gep-canonicalize-constant-indices.ll
    M llvm/test/Transforms/InstCombine/gep-combine-loop-invariant.ll
    M llvm/test/Transforms/InstCombine/gep-custom-dl.ll
    M llvm/test/Transforms/InstCombine/gep-merge-constant-indices.ll
    M llvm/test/Transforms/InstCombine/gep-vector-indices.ll
    M llvm/test/Transforms/InstCombine/gep-vector.ll
    M llvm/test/Transforms/InstCombine/gepphigep.ll
    M llvm/test/Transforms/InstCombine/getelementptr.ll
    M llvm/test/Transforms/InstCombine/icmp-custom-dl.ll
    M llvm/test/Transforms/InstCombine/icmp-gep.ll
    M llvm/test/Transforms/InstCombine/indexed-gep-compares.ll
    M llvm/test/Transforms/InstCombine/intptr1.ll
    M llvm/test/Transforms/InstCombine/intptr2.ll
    M llvm/test/Transforms/InstCombine/intptr3.ll
    M llvm/test/Transforms/InstCombine/intptr4.ll
    M llvm/test/Transforms/InstCombine/intptr5.ll
    M llvm/test/Transforms/InstCombine/intptr7.ll
    M llvm/test/Transforms/InstCombine/load-store-forward.ll
    M llvm/test/Transforms/InstCombine/load.ll
    M llvm/test/Transforms/InstCombine/loadstore-metadata.ll
    M llvm/test/Transforms/InstCombine/memchr-5.ll
    M llvm/test/Transforms/InstCombine/memchr-9.ll
    M llvm/test/Transforms/InstCombine/memcmp-3.ll
    M llvm/test/Transforms/InstCombine/memcmp-4.ll
    M llvm/test/Transforms/InstCombine/memcmp-5.ll
    M llvm/test/Transforms/InstCombine/memcmp-6.ll
    M llvm/test/Transforms/InstCombine/memcmp-7.ll
    M llvm/test/Transforms/InstCombine/memcpy_alloca.ll
    M llvm/test/Transforms/InstCombine/memrchr-5.ll
    M llvm/test/Transforms/InstCombine/memset2.ll
    M llvm/test/Transforms/InstCombine/multi-size-address-space-pointer.ll
    M llvm/test/Transforms/InstCombine/non-integral-pointers.ll
    M llvm/test/Transforms/InstCombine/opaque-ptr.ll
    M llvm/test/Transforms/InstCombine/phi-equal-incoming-pointers.ll
    M llvm/test/Transforms/InstCombine/phi-timeout.ll
    M llvm/test/Transforms/InstCombine/phi.ll
    M llvm/test/Transforms/InstCombine/pr39908.ll
    M llvm/test/Transforms/InstCombine/pr44242.ll
    M llvm/test/Transforms/InstCombine/pr58901.ll
    M llvm/test/Transforms/InstCombine/ptr-replace-alloca.ll
    M llvm/test/Transforms/InstCombine/select-cmp-br.ll
    M llvm/test/Transforms/InstCombine/select-gep.ll
    M llvm/test/Transforms/InstCombine/shift.ll
    M llvm/test/Transforms/InstCombine/sink_sideeffecting_instruction.ll
    M llvm/test/Transforms/InstCombine/sprintf-2.ll
    M llvm/test/Transforms/InstCombine/statepoint-cleanup.ll
    M llvm/test/Transforms/InstCombine/str-int-3.ll
    M llvm/test/Transforms/InstCombine/str-int-4.ll
    M llvm/test/Transforms/InstCombine/str-int-5.ll
    M llvm/test/Transforms/InstCombine/str-int.ll
    M llvm/test/Transforms/InstCombine/strcall-bad-sig.ll
    M llvm/test/Transforms/InstCombine/strcall-no-nul.ll
    M llvm/test/Transforms/InstCombine/strlen-7.ll
    M llvm/test/Transforms/InstCombine/strlen-9.ll
    M llvm/test/Transforms/InstCombine/strncmp-4.ll
    M llvm/test/Transforms/InstCombine/strncmp-5.ll
    M llvm/test/Transforms/InstCombine/strncmp-6.ll
    M llvm/test/Transforms/InstCombine/sub.ll
    M llvm/test/Transforms/InstCombine/unpack-fca.ll
    M llvm/test/Transforms/InstCombine/vec_demanded_elts-inseltpoison.ll
    M llvm/test/Transforms/InstCombine/vec_demanded_elts.ll
    M llvm/test/Transforms/InstCombine/vec_gep_scalar_arg-inseltpoison.ll
    M llvm/test/Transforms/InstCombine/vec_gep_scalar_arg.ll
    M llvm/test/Transforms/InstCombine/vscale_gep.ll
    M llvm/test/Transforms/InstCombine/wcslen-5.ll
    M llvm/test/Transforms/LoopUnroll/ARM/upperbound.ll
    M llvm/test/Transforms/LoopUnroll/peel-loop.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/deterministic-type-shrinkage.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/intrinsiccost.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/sve-cond-inv-loads.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/sve-interleaved-accesses.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/sve-widen-phi.ll
    M llvm/test/Transforms/LoopVectorize/AArch64/vector-reverse-mask4.ll
    M llvm/test/Transforms/LoopVectorize/AMDGPU/packed-math.ll
    M llvm/test/Transforms/LoopVectorize/ARM/mve-qabs.ll
    M llvm/test/Transforms/LoopVectorize/ARM/mve-reductions.ll
    M llvm/test/Transforms/LoopVectorize/ARM/mve-selectandorcost.ll
    M llvm/test/Transforms/LoopVectorize/ARM/pointer_iv.ll
    M llvm/test/Transforms/LoopVectorize/X86/float-induction-x86.ll
    M llvm/test/Transforms/LoopVectorize/X86/interleaving.ll
    M llvm/test/Transforms/LoopVectorize/X86/intrinsiccost.ll
    M llvm/test/Transforms/LoopVectorize/X86/invariant-store-vectorization.ll
    M llvm/test/Transforms/LoopVectorize/X86/metadata-enable.ll
    M llvm/test/Transforms/LoopVectorize/X86/pr23997.ll
    M llvm/test/Transforms/LoopVectorize/X86/small-size.ll
    M llvm/test/Transforms/LoopVectorize/X86/x86-interleaved-store-accesses-with-gaps.ll
    M llvm/test/Transforms/LoopVectorize/consecutive-ptr-uniforms.ll
    M llvm/test/Transforms/LoopVectorize/extract-last-veclane.ll
    M llvm/test/Transforms/LoopVectorize/float-induction.ll
    M llvm/test/Transforms/LoopVectorize/induction.ll
    M llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll
    M llvm/test/Transforms/LoopVectorize/reduction-inloop-uf4.ll
    M llvm/test/Transforms/LoopVectorize/runtime-check.ll
    M llvm/test/Transforms/LoopVectorize/scalar_after_vectorization.ll
    M llvm/test/Transforms/LoopVectorize/vector-geps.ll
    M llvm/test/Transforms/LowerMatrixIntrinsics/multiply-fused-dominance.ll
    M llvm/test/Transforms/LowerMatrixIntrinsics/multiply-fused-loops.ll
    M llvm/test/Transforms/LowerMatrixIntrinsics/multiply-fused-multiple-blocks.ll
    M llvm/test/Transforms/LowerMatrixIntrinsics/multiply-fused.ll
    M llvm/test/Transforms/LowerMatrixIntrinsics/multiply-minimal.ll
    M llvm/test/Transforms/PhaseOrdering/AArch64/hoisting-sinking-required-for-vectorization.ll
    M llvm/test/Transforms/PhaseOrdering/AArch64/peel-multiple-unreachable-exits-for-vectorization.ll
    M llvm/test/Transforms/PhaseOrdering/AArch64/quant_4x4.ll
    M llvm/test/Transforms/PhaseOrdering/AArch64/sinking-vs-if-conversion.ll
    M llvm/test/Transforms/PhaseOrdering/ARM/arm_mult_q15.ll
    M llvm/test/Transforms/PhaseOrdering/X86/excessive-unrolling.ll
    M llvm/test/Transforms/PhaseOrdering/X86/hoist-load-of-baseptr.ll
    M llvm/test/Transforms/PhaseOrdering/X86/pixel-splat.ll
    M llvm/test/Transforms/PhaseOrdering/X86/pr48844-br-to-switch-vectorization.ll
    M llvm/test/Transforms/PhaseOrdering/X86/pr50555.ll
    M llvm/test/Transforms/PhaseOrdering/X86/speculation-vs-tbaa.ll
    M llvm/test/Transforms/PhaseOrdering/X86/spurious-peeling.ll
    M llvm/test/Transforms/PhaseOrdering/X86/vdiv.ll
    M llvm/test/Transforms/PhaseOrdering/X86/vec-shift.ll
    M llvm/test/Transforms/PhaseOrdering/basic.ll
    M llvm/test/Transforms/PhaseOrdering/loop-access-checks.ll
    M llvm/test/Transforms/PhaseOrdering/pr39282.ll
    M llvm/test/Transforms/PhaseOrdering/simplifycfg-options.ll
    M llvm/test/Transforms/PhaseOrdering/switch_with_geps.ll
    M llvm/test/Transforms/SLPVectorizer/AArch64/gather-cost.ll
    M llvm/test/Transforms/SLPVectorizer/AArch64/gather-reduce.ll
    M llvm/test/Transforms/SLPVectorizer/AArch64/loadorder.ll
    M llvm/test/Transforms/SLPVectorizer/WebAssembly/no-vectorize-rotate.ll
    M llvm/test/Transforms/SLPVectorizer/X86/operandorder.ll
    M llvm/test/Transforms/SLPVectorizer/X86/opt.ll
    M llvm/test/Transforms/SLPVectorizer/X86/pr46983.ll
    M llvm/test/Transforms/SLPVectorizer/X86/pr47629-inseltpoison.ll
    M llvm/test/Transforms/SLPVectorizer/X86/pr47629.ll
    M llvm/test/Transforms/SampleProfile/pseudo-probe-instcombine.ll
    M llvm/test/Transforms/Util/strip-gc-relocates.ll

  Log Message:
  -----------
  [InstCombine] Canonicalize constant GEPs to i8 source element type (#68882)

This patch canonicalizes getelementptr instructions with constant
indices to use the `i8` source element type. This makes it easier for
optimizations to recognize that two GEPs are identical, because they
don't need to see past many different ways to express the same offset.

This is a first step towards
https://discourse.llvm.org/t/rfc-replacing-getelementptr-with-ptradd/68699.
This is limited to constant GEPs only for now, as they have a clear
canonical form, while we're not yet sure how exactly to deal with
variable indices.

The test llvm/test/Transforms/PhaseOrdering/switch_with_geps.ll gives
two representative examples of the kind of optimization improvement we
expect from this change. In the first test SimplifyCFG can now realize
that all switch branches are actually the same. In the second test it
can convert it into simple arithmetic. These are representative of
common optimization failures we see in Rust.

Fixes https://github.com/llvm/llvm-project/issues/69841.