AMDGPU: Make v2f16 minimum/maximum legal for gfx950

llvm · Nov 26, 2024 · 7f94002 · 7f94002
1 parent f5e92eb
commit 7f94002
Show file tree

Hide file tree

Showing 6 changed files with 1,850 additions and 1,782 deletions.
diff --git a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp
@@ -859,6 +859,9 @@ SITargetLowering::SITargetLowering(const TargetMachine &TM,
     // FIXME: For nnan fmaximum, emit the fmaximum3 instead of fmaxnum
     if (Subtarget->hasMinimum3Maximum3F32())
       setOperationAction({ISD::FMAXIMUM, ISD::FMINIMUM}, MVT::f32, Legal);
+
+    if (Subtarget->hasMinimum3Maximum3PKF16())
+      setOperationAction({ISD::FMAXIMUM, ISD::FMINIMUM}, MVT::v2f16, Legal);
   }
 
   setOperationAction(ISD::INTRINSIC_WO_CHAIN,

diff --git a/llvm/lib/Target/AMDGPU/VOP3PInstructions.td b/llvm/lib/Target/AMDGPU/VOP3PInstructions.td
@@ -255,6 +255,17 @@ multiclass MadFmaMixPats<SDPatternOperator fma_like,
   >;
 }
 
+class MinimumMaximumByMinimum3Maximum3VOP3P<SDPatternOperator node,
+                                            Instruction inst> : GCNPat<
+  (v2f16 (node (VOP3PMods v2f16:$src0, i32:$src0_mods), (VOP3PMods v2f16:$src1, i32:$src1_mods))),
+  (inst $src0_mods, $src0, $src1_mods, $src1, $src1_mods, $src1)
+>;
+
+let SubtargetPredicate = HasMinimum3Maximum3PKF16 in {
+def : MinimumMaximumByMinimum3Maximum3VOP3P<fminimum, V_PK_MINIMUM3_F16>;
+def : MinimumMaximumByMinimum3Maximum3VOP3P<fmaximum, V_PK_MAXIMUM3_F16>;
+}
+
 let SubtargetPredicate = HasMadMixInsts, OtherPredicates = [NoFP32Denormals] in {
 
 // These are VOP3a-like opcodes which accept no omod.