[flang-commits] [clang] [flang] [llvm] [AMDGPU] Expose more subtarget features for MLIR (PR #223561)

[via flang-commits]

llvmorg-github-actions[bot] wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-flang-openmp

Author: Krzysztof Drewniak (krzysz00)

<details>
<summary>Changes</summary>

- `FeatureAtomitFaddNoRtnInsts` is needed to control when atomic emulation
for buffer operations happens
- `FeatureBackOffBarrier` controls an inline-assembly workaround for
LDS-only barriers on old gfx9
- `FeatureLfsBarrierArriveAtomic` is a precise flag compared to arch versions
- `FeatureORPFP8ConversionInsts` controls which MLIR FP8 types can be
lowered to LLVM
- `FeaturePermlane16Insts`is another feature check

Clang and flang tests are updated since they look at the feature set

AI disclosure: Claude made these changes, I wrote this message

---
Full diff: https://github.com/llvm/llvm-project/pull/223561.diff


3 Files Affected:

- (modified) clang/test/CodeGen/link-builtin-bitcode.c (+3-3) 
- (modified) flang/test/Lower/OpenMP/target_cpu_features.f90 (+1-1) 
- (modified) llvm/lib/Target/AMDGPU/AMDGPU.td (+8-6) 


``````````diff
diff --git a/clang/test/CodeGen/link-builtin-bitcode.c b/clang/test/CodeGen/link-builtin-bitcode.c
index 2cae7c027a4f3..9fcd0dccad12f 100644
--- a/clang/test/CodeGen/link-builtin-bitcode.c
+++ b/clang/test/CodeGen/link-builtin-bitcode.c
@@ -44,6 +44,6 @@ int bar() { return no_attr() + attr_in_target() + attr_not_in_target() + attr_in
 // CHECK-SAME: () #[[ATTR_INCOMPATIBLE:[0-9]+]] {
 
 // CHECK: attributes #[[ATTR_BAR]] = { {{.*}} "no-trapping-math"="true" {{.*}} }
-// CHECK: attributes #[[ATTR_COMPATIBLE]] = { {{.*}} "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+atomic-fmin-fmax-global-f64,+ci-insts,+cube-insts,+cvt-pknorm-vop2-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+flat-buffer-global-fadd-f64-inst,+flat-global-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gws,+image-insts,+lerp-inst,+mai-insts,+mqsad-insts,+mqsad-pk-insts,+msad-insts,+qsad-insts,+s-memrealtime,+s-memtime-inst,+sad-insts,+vmem-to-lds-load-insts,+wavefrontsize64" }
-// CHECK: attributes #[[ATTR_EXTEND]] = { {{.*}} "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+atomic-fmin-fmax-global-f64,+ci-insts,+cube-insts,+cvt-pknorm-vop2-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dot8-insts,+dpp,+flat-buffer-global-fadd-f64-inst,+flat-global-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gws,+image-insts,+lerp-inst,+mai-insts,+mqsad-insts,+mqsad-pk-insts,+msad-insts,+qsad-insts,+s-memrealtime,+s-memtime-inst,+sad-insts,+vmem-to-lds-load-insts,+wavefrontsize64" }
-// CHECK: attributes #[[ATTR_INCOMPATIBLE]] = { {{.*}} "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-rtn-insts,+atomic-fmin-fmax-global-f64,+ci-insts,+cube-insts,+cvt-pknorm-vop2-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+flat-buffer-global-fadd-f64-inst,+flat-global-insts,+gfx8-insts,+gfx90a-insts,+gws,+image-insts,+lerp-inst,+mai-insts,+mqsad-insts,+mqsad-pk-insts,+msad-insts,+qsad-insts,+s-memrealtime,+s-memtime-inst,+sad-insts,+vmem-to-lds-load-insts,+wavefrontsize64,-gfx9-insts" }
+// CHECK: attributes #[[ATTR_COMPATIBLE]] = { {{.*}} "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-no-rtn-insts,+atomic-fadd-rtn-insts,+atomic-fmin-fmax-global-f64,+back-off-barrier,+ci-insts,+cube-insts,+cvt-pknorm-vop2-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+flat-buffer-global-fadd-f64-inst,+flat-global-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gws,+image-insts,+lerp-inst,+mai-insts,+mqsad-insts,+mqsad-pk-insts,+msad-insts,+qsad-insts,+s-memrealtime,+s-memtime-inst,+sad-insts,+vmem-to-lds-load-insts,+wavefrontsize64" }
+// CHECK: attributes #[[ATTR_EXTEND]] = { {{.*}} "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-no-rtn-insts,+atomic-fadd-rtn-insts,+atomic-fmin-fmax-global-f64,+back-off-barrier,+ci-insts,+cube-insts,+cvt-pknorm-vop2-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dot8-insts,+dpp,+flat-buffer-global-fadd-f64-inst,+flat-global-insts,+gfx8-insts,+gfx9-insts,+gfx90a-insts,+gws,+image-insts,+lerp-inst,+mai-insts,+mqsad-insts,+mqsad-pk-insts,+msad-insts,+qsad-insts,+s-memrealtime,+s-memtime-inst,+sad-insts,+vmem-to-lds-load-insts,+wavefrontsize64" }
+// CHECK: attributes #[[ATTR_INCOMPATIBLE]] = { {{.*}} "target-features"="+16-bit-insts,+atomic-buffer-global-pk-add-f16-insts,+atomic-fadd-no-rtn-insts,+atomic-fadd-rtn-insts,+atomic-fmin-fmax-global-f64,+back-off-barrier,+ci-insts,+cube-insts,+cvt-pknorm-vop2-insts,+dl-insts,+dot1-insts,+dot10-insts,+dot2-insts,+dot3-insts,+dot4-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+flat-buffer-global-fadd-f64-inst,+flat-global-insts,+gfx8-insts,+gfx90a-insts,+gws,+image-insts,+lerp-inst,+mai-insts,+mqsad-insts,+mqsad-pk-insts,+msad-insts,+qsad-insts,+s-memrealtime,+s-memtime-inst,+sad-insts,+vmem-to-lds-load-insts,+wavefrontsize64,-gfx9-insts" }
diff --git a/flang/test/Lower/OpenMP/target_cpu_features.f90 b/flang/test/Lower/OpenMP/target_cpu_features.f90
index 78f29b23068af..96ac924afb488 100644
--- a/flang/test/Lower/OpenMP/target_cpu_features.f90
+++ b/flang/test/Lower/OpenMP/target_cpu_features.f90
@@ -8,7 +8,7 @@
 
 !AMDGCN: module attributes {
 !AMDGCN-SAME: fir.target_cpu = "gfx908"
-!AMDGCN-SAME: fir.target_features = #llvm.target_features<["+16-bit-insts", "+ci-insts",
+!AMDGCN-SAME: fir.target_features = #llvm.target_features<["+16-bit-insts", "+atomic-fadd-no-rtn-insts", "+ci-insts",
 !AMDGCN-SAME: "+dl-insts", "+dot1-insts", "+dot10-insts", "+dot2-insts", "+dot3-insts",
 !AMDGCN-SAME: "+dot4-insts", "+dot5-insts", "+dot6-insts", "+dot7-insts", "+dpp",
 !AMDGCN-SAME: "+gfx8-insts", "+gfx9-insts", "+gws", "+image-insts", "+lerp-inst", "+mai-insts",
diff --git a/llvm/lib/Target/AMDGPU/AMDGPU.td b/llvm/lib/Target/AMDGPU/AMDGPU.td
index 30db12b565082..b7b63146277ab 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPU.td
+++ b/llvm/lib/Target/AMDGPU/AMDGPU.td
@@ -3228,8 +3228,9 @@ def AMDGPUFrontendVisibleFeatures {
   Feature16BitInsts, FeatureAddMinMaxInsts, FeatureAshrPkInsts,
   FeatureAsyncLoadToLDSInsts, FeatureAsyncStoreFromLDSInsts, FeatureAsynccnt,
   FeatureAtomicBufferGlobalPkAddF16Insts, FeatureAtomicBufferPkAddBF16Inst, FeatureAtomicDsPkAdd16Insts,
-  FeatureAtomicFMinFMaxF32GlobalInsts, FeatureAtomicFMinFMaxF64GlobalInsts, FeatureAtomicFaddRtnInsts,
-  FeatureAtomicFlatPkAdd16Insts, FeatureAtomicGlobalPkAddBF16Inst, FeatureBF16ConversionInsts,
+  FeatureAtomicFMinFMaxF32GlobalInsts, FeatureAtomicFMinFMaxF64GlobalInsts, FeatureAtomicFaddNoRtnInsts,
+  FeatureAtomicFaddRtnInsts, FeatureAtomicFlatPkAdd16Insts, FeatureAtomicGlobalPkAddBF16Inst,
+  FeatureBF16ConversionInsts,
   FeatureBF16PackedInsts, FeatureBF16TransInsts, FeatureBF8ConversionScaleInsts,
   FeatureBVHRayTracingInsts, FeatureBitOp3Insts, FeatureCIInsts,
   FeatureClusters, FeatureCubeInsts, FeatureCvtPkNormVOP2Insts,
@@ -3247,9 +3248,10 @@ def AMDGPUFrontendVisibleFeatures {
   FeatureGFX1250Insts, FeatureGFX1251GEMMInsts, FeatureGFX12Insts,
   FeatureGFX13Insts, FeatureGFX8Insts, FeatureGFX90AInsts,
   FeatureGFX940Insts, FeatureGFX950Insts, FeatureGFX9Insts,
-  FeatureGWS, FeatureImageInsts, FeatureLerpInst,
-  FeatureMAIInsts, FeatureMcastLoadInsts, FeatureMqsadInsts,
-  FeatureMqsadPkInsts, FeatureMsadInsts, FeaturePermlane16Swap,
+  FeatureGWS, FeatureImageInsts, FeatureLdsBarrierArriveAtomic,
+  FeatureLerpInst, FeatureMAIInsts, FeatureMcastLoadInsts,
+  FeatureMqsadInsts, FeatureMqsadPkInsts, FeatureMsadInsts,
+  FeatureOCPFP8ConversionInsts, FeaturePermlane16Insts, FeaturePermlane16Swap,
   FeaturePermlane32Swap, FeaturePkAddMinMaxInsts, FeaturePrngInst,
   FeatureQsadInsts, FeatureSMemRealTime, FeatureSMemTimeInst,
   FeatureSWMMACGfx1200Insts, FeatureSWMMACGfx1250Insts, FeatureSWakeupBarrier,
@@ -3262,7 +3264,7 @@ def AMDGPUFrontendVisibleFeatures {
   FeatureSupportsXNACK, FeatureSupportsSRAMECC, FeatureXNACKOnOffModes,
   FeatureSGPRInitBug, FeatureApertureRegs, FeatureGetDoorbellID,
   FeatureAGPRAlloc, Feature1536VGPRs, Feature1024AddressableVGPRs,
-  FeatureHalfAddressablePhysicalLocalMemory,
+  FeatureHalfAddressablePhysicalLocalMemory, FeatureBackOffBarrier,
   ];
 }
 

``````````

</details>


https://github.com/llvm/llvm-project/pull/223561