[all-commits] [llvm/llvm-project] 937596: [TableGen][GlobalISel] Specialize more MatchTable ...

[Pierre van Houtryve via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 9375962ac9e8d8e83fa8e32c3b04bb6970e4b242
      https://github.com/llvm/llvm-project/commit/9375962ac9e8d8e83fa8e32c3b04bb6970e4b242
  Author: Pierre van Houtryve <pierre.vanhoutryve at amd.com>
  Date:   2024-04-24 (Wed, 24 Apr 2024)

  Changed paths:
    M llvm/include/llvm/CodeGen/GlobalISel/GIMatchTableExecutor.h
    M llvm/include/llvm/CodeGen/GlobalISel/GIMatchTableExecutorImpl.h
    M llvm/test/TableGen/ContextlessPredicates.td
    M llvm/test/TableGen/DefaultOpsGlobalISel.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/builtins/match-table-eraseroot.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/builtins/match-table-replacerreg.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/match-table-imms.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/match-table-intrinsics.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/match-table-miflags.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/match-table-operand-types.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/match-table-patfrag-root.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/match-table-permutations.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/match-table-temp-defs.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/match-table-typeof.td
    M llvm/test/TableGen/GlobalISelCombinerEmitter/match-table.td
    M llvm/test/TableGen/GlobalISelEmitter-atomic_store.td
    M llvm/test/TableGen/GlobalISelEmitter-immAllZeroOne.td
    M llvm/test/TableGen/GlobalISelEmitter-immarg-literal-pattern.td
    M llvm/test/TableGen/GlobalISelEmitter-input-discard.td
    M llvm/test/TableGen/GlobalISelEmitter-multiple-output-discard.td
    M llvm/test/TableGen/GlobalISelEmitter-multiple-output.td
    M llvm/test/TableGen/GlobalISelEmitter-nested-subregs.td
    M llvm/test/TableGen/GlobalISelEmitter-notype-output-pattern.td
    M llvm/test/TableGen/GlobalISelEmitter-output-discard.td
    M llvm/test/TableGen/GlobalISelEmitter-zero-reg.td
    M llvm/test/TableGen/GlobalISelEmitter.td
    M llvm/test/TableGen/GlobalISelEmitterCustomPredicate.td
    M llvm/test/TableGen/GlobalISelEmitterFlags.td
    M llvm/test/TableGen/GlobalISelEmitterHwModes.td
    M llvm/test/TableGen/GlobalISelEmitterMatchTableOptimizer.td
    M llvm/test/TableGen/GlobalISelEmitterMatchTableOptimizerSameOperand-invalid.td
    M llvm/test/TableGen/GlobalISelEmitterMatchTableOptimizerSameOperand.td
    M llvm/test/TableGen/GlobalISelEmitterOverloadedPtr.td
    M llvm/test/TableGen/GlobalISelEmitterRegSequence.td
    M llvm/test/TableGen/GlobalISelEmitterSubreg.td
    M llvm/test/TableGen/GlobalISelEmitterVariadic.td
    M llvm/test/TableGen/HasNoUse.td
    M llvm/test/TableGen/address-space-patfrags.td
    M llvm/test/TableGen/gisel-physreg-input.td
    M llvm/test/TableGen/immarg-predicated.td
    M llvm/test/TableGen/immarg.td
    M llvm/utils/TableGen/Common/GlobalISel/GlobalISelMatchTable.cpp
    M llvm/utils/TableGen/Common/GlobalISel/GlobalISelMatchTable.h

  Log Message:
  -----------
  [TableGen][GlobalISel] Specialize more MatchTable Opcodes (#89736)

The vast majority of the following (very common) opcodes were always
called with identical arguments:

- `GIM_CheckType` for the root
- `GIM_CheckRegBankForClass` for the root
- `GIR_Copy` between the old and new root
- `GIR_ConstrainSelectedInstOperands` on the new root
- `GIR_BuildMI` to create the new root

I added overloaded version of each opcode specialized for the root
instructions. It always saves between 1 and 2 bytes per instance
depending on the number of arguments specialized into the opcode. Some
of these opcodes had between 5 and 15k occurences in the AArch64
GlobalISel Match Table.

Additionally, the following opcodes are almost always used in the same
sequence:

- `GIR_EraseFromParent 0` + `GIR_Done` 
- `GIR_EraseRootFromParent_Done` has been created to do both. Saves 2
bytes per occurence.
- `GIR_IsSafeToFold` was *always* called for each InsnID except 0.
- Changed the opcode to take the number of instructions to check after
`MI[0]`

The savings from these are pretty neat. For `AArch64GenGlobalISel.inc`:
- `AArch64InstructionSelector.cpp.o` goes down from 772kb to 704kb (-10%
code size)
- Self-reported MatchTable size goes from 420380 bytes to 352426 bytes
(~ -17%)

A smaller match table means a faster match table because we spend less
time iterating and decoding.
I don't have a solid measurement methodology for GlobalISel performance
so I don't have precise numbers but I saw a few % of improvements in a
simple testcase.



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