[all-commits] [llvm/llvm-project] 982017: [amdgpu] Reimplement LDS lowering

[Jon Chesterfield via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 982017240d7f25a8a6969b8b73dc51f9ac5b93ed
      https://github.com/llvm/llvm-project/commit/982017240d7f25a8a6969b8b73dc51f9ac5b93ed
  Author: Jon Chesterfield <jonathanchesterfield at gmail.com>
  Date:   2022-12-06 (Tue, 06 Dec 2022)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUISelLowering.cpp
    M llvm/lib/Target/AMDGPU/AMDGPULowerModuleLDSPass.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUMCInstLower.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUMachineFunction.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUMachineFunction.h
    M llvm/lib/Target/AMDGPU/SIISelLowering.cpp
    M llvm/test/CodeGen/AMDGPU/GlobalISel/dropped_debug_info_assert.ll
    M llvm/test/CodeGen/AMDGPU/GlobalISel/irtranslator-inline-asm.ll
    M llvm/test/CodeGen/AMDGPU/addrspacecast-known-non-null.ll
    M llvm/test/CodeGen/AMDGPU/ds_write2.ll
    M llvm/test/CodeGen/AMDGPU/hsa.ll
    M llvm/test/CodeGen/AMDGPU/lds-size.ll
    M llvm/test/CodeGen/AMDGPU/local-memory.amdgcn.ll
    M llvm/test/CodeGen/AMDGPU/local-memory.r600.ll
    M llvm/test/CodeGen/AMDGPU/lower-kernel-and-module-lds.ll
    M llvm/test/CodeGen/AMDGPU/lower-kernel-lds-constexpr.ll
    M llvm/test/CodeGen/AMDGPU/lower-kernel-lds-super-align.ll
    M llvm/test/CodeGen/AMDGPU/lower-kernel-lds.ll
    M llvm/test/CodeGen/AMDGPU/lower-lds-struct-aa-memcpy.ll
    M llvm/test/CodeGen/AMDGPU/lower-lds-struct-aa-merge.ll
    M llvm/test/CodeGen/AMDGPU/lower-module-lds-constantexpr-phi.ll
    M llvm/test/CodeGen/AMDGPU/lower-module-lds-constantexpr.ll
    M llvm/test/CodeGen/AMDGPU/lower-module-lds-inactive.ll
    M llvm/test/CodeGen/AMDGPU/lower-module-lds-offsets.ll
    A llvm/test/CodeGen/AMDGPU/lower-module-lds-single-var-ambiguous.ll
    A llvm/test/CodeGen/AMDGPU/lower-module-lds-single-var-unambiguous.ll
    M llvm/test/CodeGen/AMDGPU/lower-module-lds-used-list.ll
    A llvm/test/CodeGen/AMDGPU/lower-module-lds-via-table.ll
    M llvm/test/CodeGen/AMDGPU/lower-module-lds.ll
    M llvm/test/CodeGen/AMDGPU/module-lds-false-sharing.ll
    M llvm/test/CodeGen/AMDGPU/noclobber-barrier.ll
    M llvm/test/CodeGen/AMDGPU/promote-alloca-stored-pointer-value.ll

  Log Message:
  -----------
  [amdgpu] Reimplement LDS lowering

Renames the current lowering scheme to "module" and introduces two new
ones, "kernel" and "table", plus a "hybrid" that chooses between those three
on a per-variable basis.

Unit tests are set up to pass with the default lowering of "module" or "hybrid"
with this patch defaulting to "module", which will be a less dramatic codegen
change relative to the current. This reflects the sparsity of test coverage for
the table lowering method. Hybrid is better than module in every respect and
will be default in a subsequent patch.

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D139433