1 files changed, 80 insertions, 11 deletions
diff --git a/llvm/lib/Target/AMDGPU/AMDGPU.td b/llvm/lib/Target/AMDGPU/AMDGPU.td
index d0d7a9dc1724..63d83346528a 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPU.td
+++ b/llvm/lib/Target/AMDGPU/AMDGPU.td
@@ -351,6 +351,7 @@ def FeatureGFX90AInsts : SubtargetFeature<"gfx90a-insts",
   "GFX90AInsts",
   "true",
   "Additional instructions for GFX90A+"
+  // [HasAtomicFMinFMaxF64GlobalInsts, HasAtomicFMinFMaxF64FlatInsts] // TODO
 >;
 
 def FeatureGFX940Insts : SubtargetFeature<"gfx940-insts",
@@ -711,6 +712,30 @@ def FeatureAtomicFaddRtnInsts : SubtargetFeature<"atomic-fadd-rtn-insts",
   [FeatureFlatGlobalInsts]
 >;
 
+def FeatureAtomicFMinFMaxF32GlobalInsts : SubtargetFeature<"atomic-fmin-fmax-global-f32",
+  "HasAtomicFMinFMaxF32GlobalInsts",
+  "true",
+  "Has global/buffer instructions for atomicrmw fmin/fmax for float"
+>;
+
+def FeatureAtomicFMinFMaxF64GlobalInsts : SubtargetFeature<"atomic-fmin-fmax-global-f64",
+  "HasAtomicFMinFMaxF64GlobalInsts",
+  "true",
+  "Has global/buffer instructions for atomicrmw fmin/fmax for float"
+>;
+
+def FeatureAtomicFMinFMaxF32FlatInsts : SubtargetFeature<"atomic-fmin-fmax-flat-f32",
+  "HasAtomicFMinFMaxF32FlatInsts",
+  "true",
+  "Has flat memory instructions for atomicrmw fmin/fmax for float"
+>;
+
+def FeatureAtomicFMinFMaxF64FlatInsts : SubtargetFeature<"atomic-fmin-fmax-flat-f64",
+  "HasAtomicFMinFMaxF64FlatInsts",
+  "true",
+  "Has flat memory instructions for atomicrmw fmin/fmax for double"
+>;
+
 def FeatureAtomicFaddNoRtnInsts : SubtargetFeature<"atomic-fadd-no-rtn-insts",
   "HasAtomicFaddNoRtnInsts",
   "true",
@@ -743,6 +768,12 @@ def FeatureAtomicGlobalPkAddBF16Inst : SubtargetFeature<"atomic-global-pk-add-bf
  [FeatureFlatGlobalInsts]
 >;
 
+def FeatureAtomicBufferPkAddBF16Inst : SubtargetFeature<"atomic-buffer-pk-add-bf16-inst",
+ "HasAtomicBufferPkAddBF16Inst",
+ "true",
+ "Has buffer_atomic_pk_add_bf16 instruction"
+>;
+
 def FeatureAtomicCSubNoRtnInsts : SubtargetFeature<"atomic-csub-no-rtn-insts",
   "HasAtomicCSubNoRtnInsts",
   "true",
@@ -1061,7 +1092,8 @@ def FeatureSouthernIslands : GCNSubtargetFeatureGeneration<"SOUTHERN_ISLANDS",
   FeatureWavefrontSize64, FeatureSMemTimeInst, FeatureMadMacF32Insts,
   FeatureDsSrc2Insts, FeatureLDSBankCount32, FeatureMovrel,
   FeatureTrigReducedRange, FeatureExtendedImageInsts, FeatureImageInsts,
-  FeatureGDS, FeatureGWS, FeatureDefaultComponentZero
+  FeatureGDS, FeatureGWS, FeatureDefaultComponentZero,
+  FeatureAtomicFMinFMaxF32GlobalInsts, FeatureAtomicFMinFMaxF64GlobalInsts
   ]
 >;
 
@@ -1072,7 +1104,9 @@ def FeatureSeaIslands : GCNSubtargetFeatureGeneration<"SEA_ISLANDS",
   FeatureCIInsts, FeatureMovrel, FeatureTrigReducedRange,
   FeatureGFX7GFX8GFX9Insts, FeatureSMemTimeInst, FeatureMadMacF32Insts,
   FeatureDsSrc2Insts, FeatureExtendedImageInsts, FeatureUnalignedBufferAccess,
-  FeatureImageInsts, FeatureGDS, FeatureGWS, FeatureDefaultComponentZero
+  FeatureImageInsts, FeatureGDS, FeatureGWS, FeatureDefaultComponentZero,
+  FeatureAtomicFMinFMaxF32GlobalInsts, FeatureAtomicFMinFMaxF64GlobalInsts,
+  FeatureAtomicFMinFMaxF32FlatInsts, FeatureAtomicFMinFMaxF64FlatInsts
   ]
 >;
 
@@ -1127,7 +1161,9 @@ def FeatureGFX10 : GCNSubtargetFeatureGeneration<"GFX10",
    FeatureA16, FeatureSMemTimeInst, FeatureFastDenormalF32, FeatureG16,
    FeatureUnalignedBufferAccess, FeatureUnalignedDSAccess, FeatureImageInsts,
    FeatureGDS, FeatureGWS, FeatureDefaultComponentZero,
-   FeatureMaxHardClauseLength63
+   FeatureMaxHardClauseLength63,
+   FeatureAtomicFMinFMaxF32GlobalInsts, FeatureAtomicFMinFMaxF64GlobalInsts,
+   FeatureAtomicFMinFMaxF32FlatInsts, FeatureAtomicFMinFMaxF64FlatInsts
   ]
 >;
 
@@ -1148,7 +1184,8 @@ def FeatureGFX11 : GCNSubtargetFeatureGeneration<"GFX11",
    FeatureA16, FeatureFastDenormalF32, FeatureG16,
    FeatureUnalignedBufferAccess, FeatureUnalignedDSAccess, FeatureGDS,
    FeatureGWS, FeatureDefaultComponentZero,
-   FeatureMaxHardClauseLength32
+   FeatureMaxHardClauseLength32,
+   FeatureAtomicFMinFMaxF32GlobalInsts, FeatureAtomicFMinFMaxF32FlatInsts
   ]
 >;
 
@@ -1169,7 +1206,8 @@ def FeatureGFX12 : GCNSubtargetFeatureGeneration<"GFX12",
    FeatureA16, FeatureFastDenormalF32, FeatureG16,
    FeatureUnalignedBufferAccess, FeatureUnalignedDSAccess,
    FeatureTrue16BitInsts, FeatureDefaultComponentBroadcast,
-   FeatureMaxHardClauseLength32
+   FeatureMaxHardClauseLength32,
+   FeatureAtomicFMinFMaxF32GlobalInsts, FeatureAtomicFMinFMaxF32FlatInsts
   ]
 >;
 
@@ -1332,7 +1370,10 @@ def FeatureISAVersion9_0_A : FeatureSet<
      FeaturePackedTID,
      FullRate64Ops,
      FeatureBackOffBarrier,
-     FeatureKernargPreload])>;
+     FeatureKernargPreload,
+     FeatureAtomicFMinFMaxF64GlobalInsts,
+     FeatureAtomicFMinFMaxF64FlatInsts
+     ])>;
 
 def FeatureISAVersion9_0_C : FeatureSet<
   !listconcat(FeatureISAVersion9_0_Consumer_Common.Features,
@@ -1372,7 +1413,10 @@ def FeatureISAVersion9_4_Common : FeatureSet<
    FeatureArchitectedFlatScratch,
    FullRate64Ops,
    FeatureBackOffBarrier,
-   FeatureKernargPreload]>;
+   FeatureKernargPreload,
+   FeatureAtomicFMinFMaxF64GlobalInsts,
+   FeatureAtomicFMinFMaxF64FlatInsts
+   ]>;
 
 def FeatureISAVersion9_4_0 : FeatureSet<
   !listconcat(FeatureISAVersion9_4_Common.Features,
@@ -1561,6 +1605,7 @@ def FeatureISAVersion12 : FeatureSet<
    FeatureAtomicFlatPkAdd16Insts,
    FeatureAtomicBufferGlobalPkAddF16Insts,
    FeatureAtomicGlobalPkAddBF16Inst,
+   FeatureAtomicBufferPkAddBF16Inst,
    FeatureFlatAtomicFaddF32Inst,
    FeatureImageInsts,
    FeatureExtendedImageInsts,
@@ -1572,7 +1617,9 @@ def FeatureISAVersion12 : FeatureSet<
    FeatureHasRestrictedSOffset,
    FeatureVGPRSingleUseHintInsts,
    FeatureScalarDwordx3Loads,
-   FeatureDPPSrc1SGPR]>;
+   FeatureDPPSrc1SGPR,
+   FeatureMaxHardClauseLength32,
+   Feature1_5xVGPRs]>;
 
 def FeatureISAVersion12_Generic: FeatureSet<
   !listconcat(FeatureISAVersion12.Features,
@@ -1862,9 +1909,28 @@ def isGFX12Plus :
 def HasFlatAddressSpace : Predicate<"Subtarget->hasFlatAddressSpace()">,
   AssemblerPredicate<(all_of FeatureFlatAddressSpace)>;
 
-def HasBufferFlatGlobalAtomicsF64 :
+
+def HasBufferFlatGlobalAtomicsF64 : // FIXME: Rename to show it's only for fadd
   Predicate<"Subtarget->hasBufferFlatGlobalAtomicsF64()">,
-  AssemblerPredicate<(any_of FeatureGFX90AInsts)>;
+  // FIXME: This is too coarse, and working around using pseudo's predicates on real instruction.
+  AssemblerPredicate<(any_of FeatureGFX90AInsts, FeatureGFX10Insts, FeatureSouthernIslands, FeatureSeaIslands)>;
+
+def HasAtomicFMinFMaxF32GlobalInsts :
+  Predicate<"Subtarget->hasAtomicFMinFMaxF32GlobalInsts()">,
+  AssemblerPredicate<(any_of FeatureAtomicFMinFMaxF32GlobalInsts)>;
+
+def HasAtomicFMinFMaxF64GlobalInsts :
+  Predicate<"Subtarget->hasAtomicFMinFMaxF64GlobalInsts()">,
+  AssemblerPredicate<(any_of FeatureAtomicFMinFMaxF64GlobalInsts)>;
+
+def HasAtomicFMinFMaxF32FlatInsts :
+  Predicate<"Subtarget->hasAtomicFMinFMaxF32FlatInsts()">,
+  AssemblerPredicate<(any_of FeatureAtomicFMinFMaxF32FlatInsts)>;
+
+def HasAtomicFMinFMaxF64FlatInsts :
+  Predicate<"Subtarget->hasAtomicFMinFMaxF64FlatInsts()">,
+  AssemblerPredicate<(any_of FeatureAtomicFMinFMaxF64FlatInsts)>;
+
 def HasLdsAtomicAddF64 :
   Predicate<"Subtarget->hasLdsAtomicAddF64()">,
   AssemblerPredicate<(any_of FeatureGFX90AInsts)>;
@@ -2118,7 +2184,10 @@ def HasAtomicBufferGlobalPkAddF16Insts
   AssemblerPredicate<(all_of FeatureAtomicBufferGlobalPkAddF16Insts)>;
 def HasAtomicGlobalPkAddBF16Inst
   : Predicate<"Subtarget->hasAtomicGlobalPkAddBF16Inst()">,
-  AssemblerPredicate<(all_of FeatureAtomicGlobalPkAddBF16Inst)>;
+    AssemblerPredicate<(all_of FeatureAtomicGlobalPkAddBF16Inst)>;
+def HasAtomicBufferPkAddBF16Inst
+  : Predicate<"Subtarget->hasAtomicBufferPkAddBF16Inst()">,
+    AssemblerPredicate<(all_of FeatureAtomicBufferPkAddBF16Inst)>;
 def HasFlatAtomicFaddF32Inst
   : Predicate<"Subtarget->hasFlatAtomicFaddF32Inst()">,
   AssemblerPredicate<(all_of FeatureFlatAtomicFaddF32Inst)>;