[all-commits] [llvm/llvm-project] 6de65b: [AMDGPU] Add getLDSAllocGranule to TargetParser

[Chinmay Deshpande via All-commits]

  Branch: refs/heads/users/chinmaydd/lds-alloc-granule
  Home:   https://github.com/llvm/llvm-project
  Commit: 6de65b83222b02ea3aa9e5c49ee6c25a1ab70807
      https://github.com/llvm/llvm-project/commit/6de65b83222b02ea3aa9e5c49ee6c25a1ab70807
  Author: Chinmay Deshpande <chdeshpa at amd.com>
  Date:   2026-09-18 (Fri, 18 Sep 2026)

  Changed paths:
    M llvm/include/llvm/TargetParser/AMDGPUTargetParser.h
    M llvm/lib/Target/AMDGPU/AMDGPU.td
    M llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp
    M llvm/lib/Target/AMDGPU/AMDGPUFeatures.td
    M llvm/lib/Target/AMDGPU/GCNSubtarget.cpp
    M llvm/lib/Target/AMDGPU/R600Processors.td
    M llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp
    M llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h
    M llvm/lib/TargetParser/AMDGPUTargetParser.cpp
    A llvm/test/CodeGen/AMDGPU/lds-size-gfx9-4-generic.ll
    A llvm/test/TableGen/AMDGPUTargetDefLDSAllocGranularity.td
    M llvm/unittests/TargetParser/TargetParserTest.cpp

  Log Message:
  -----------
  [AMDGPU] Add getLDSAllocGranule to TargetParser

Model LDS allocation granularity with subtarget features derived from
AMDGPUGenericAnyFeature. Generic targets can select a granularity present
on any covered GPU, independently of their addressable LDS size.

Expose the byte-valued query for GPUKind and subarch and consolidate
backend users on it. gfx9-4-generic uses gfx950's 1280-byte granule while
retaining its 64 KiB addressable LDS capacity.

Test feature membership, granularity assignments, query overloads and
fallbacks, and generic-target LDS block rounding.

Change-Id: Ic0c9345e7657ec3c6978a646628598cb7608b390



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