[LoongArch] Expand FSIN/FCOS/FSINCOS/FPOW/FREM

[xen0n]

llvm/test/CodeGen/LoongArch/fp-expand.ll

8–10

Do you mean float for these f32 intrinsics instead?

[gonglingqin]

llvm/test/CodeGen/LoongArch/fp-expand.ll

8–10

Yes, thank you for checking, I will modify it.

[gonglingqin]

Fix function declarations in test cases.

[xry111]

Straightforward enough.

llvm/lib/Target/LoongArch/LoongArchISelLowering.cpp

if (Subtarget.hasBasicF()) {
     setCondCodeAction(FPCCToExpand, MVT::f32, Expand);
     setOperationAction(ISD::SELECT_CC, MVT::f32, Expand);
     setOperationAction(ISD::BR_CC, MVT::f32, Expand);
     setOperationAction(ISD::FMA, MVT::f32, Legal);
     setOperationAction(ISD::FMINNUM_IEEE, MVT::f32, Legal);
     setOperationAction(ISD::FMAXNUM_IEEE, MVT::f32, Legal);
     setOperationAction(ISD::STRICT_FSETCCS, MVT::f32, Legal);
     setOperationAction(ISD::STRICT_FSETCC, MVT::f32, Legal);
+    setOperationAction(ISD::FSIN, MVT::f32, Expand);
+    setOperationAction(ISD::FCOS, MVT::f32, Expand);
+    setOperationAction(ISD::FSINCOS, MVT::f32, Expand);
+    setOperationAction(ISD::FPOW, MVT::f32, Expand);
+    setOperationAction(ISD::FREM, MVT::f32, Expand);
   }
   if (Subtarget.hasBasicD()) {
     setCondCodeAction(FPCCToExpand, MVT::f64, Expand);
     setOperationAction(ISD::SELECT_CC, MVT::f64, Expand);
     setOperationAction(ISD::BR_CC, MVT::f64, Expand);
     setOperationAction(ISD::STRICT_FSETCCS, MVT::f64, Legal);
     setOperationAction(ISD::STRICT_FSETCC, MVT::f64, Legal);
     setLoadExtAction(ISD::EXTLOAD, MVT::f64, MVT::f32, Expand);
     setLoadExtAction(ISD::EXTLOAD, MVT::f64, MVT::f32, Expand);
     setOperationAction(ISD::FMA, MVT::f64, Legal);
     setOperationAction(ISD::FMINNUM_IEEE, MVT::f64, Legal);
     setOperationAction(ISD::FMAXNUM_IEEE, MVT::f64, Legal);
+    setOperationAction(ISD::FSIN, MVT::f64, Expand);
+    setOperationAction(ISD::FCOS, MVT::f64, Expand);
+    setOperationAction(ISD::FSINCOS, MVT::f64, Expand);
+    setOperationAction(ISD::FPOW, MVT::f64, Expand);
+    setOperationAction(ISD::FREM, MVT::f64, Expand);
   }
 
   setOperationAction(ISD::BR_JT, MVT::Other, Expand);
 
   setOperationAction(ISD::BR_CC, GRLenVT, Expand);
   setOperationAction(ISD::SELECT_CC, GRLenVT, Expand);
   setOperationAction(ISD::SIGN_EXTEND_INREG, MVT::i1, Expand);
   setOperationAction({ISD::SMUL_LOHI, ISD::UMUL_LOHI}, GRLenVT, Expand);

llvm/test/CodeGen/LoongArch/fp-expand.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc --mtriple=loongarch32 --mattr=+d < %s | FileCheck %s --check-prefix=LA32
+; RUN: llc --mtriple=loongarch64 --mattr=+d < %s | FileCheck %s --check-prefix=LA64
+
+;; TODO: Add more test cases after ABI implementation for ilp32f and lp64f.
+
+declare float @llvm.sin.f32(float)
+declare float @llvm.cos.f32(float)
+declare float @llvm.pow.f32(float, float)
+declare double @llvm.sin.f64(double)
+declare double @llvm.cos.f64(double)
+declare double @llvm.pow.f64(double, double)
+
+define float @sin_f32(float %a) nounwind {
+; LA32-LABEL: sin_f32:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -16
+; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
+; LA32-NEXT:    bl %plt(sinf)
+; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 16
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: sin_f32:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -16
+; LA64-NEXT:    st.d $ra, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    bl %plt(sinf)
+; LA64-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 16
+; LA64-NEXT:    ret
+  %1 = call float @llvm.sin.f32(float %a)
+  ret float %1
+}
+
+define float @cos_f32(float %a) nounwind {
+; LA32-LABEL: cos_f32:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -16
+; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
+; LA32-NEXT:    bl %plt(cosf)
+; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 16
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: cos_f32:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -16
+; LA64-NEXT:    st.d $ra, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    bl %plt(cosf)
+; LA64-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 16
+; LA64-NEXT:    ret
+  %1 = call float @llvm.cos.f32(float %a)
+  ret float %1
+}
+
+define float @sincos_f32(float %a) nounwind {
+; LA32-LABEL: sincos_f32:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -32
+; LA32-NEXT:    st.w $ra, $sp, 28 # 4-byte Folded Spill
+; LA32-NEXT:    fst.d $fs0, $sp, 16 # 8-byte Folded Spill
+; LA32-NEXT:    fst.d $fs1, $sp, 8 # 8-byte Folded Spill
+; LA32-NEXT:    fmov.s $fs0, $fa0
+; LA32-NEXT:    bl %plt(sinf)
+; LA32-NEXT:    fmov.s $fs1, $fa0
+; LA32-NEXT:    fmov.s $fa0, $fs0
+; LA32-NEXT:    bl %plt(cosf)
+; LA32-NEXT:    fadd.s $fa0, $fs1, $fa0
+; LA32-NEXT:    fld.d $fs1, $sp, 8 # 8-byte Folded Reload
+; LA32-NEXT:    fld.d $fs0, $sp, 16 # 8-byte Folded Reload
+; LA32-NEXT:    ld.w $ra, $sp, 28 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 32
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: sincos_f32:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -32
+; LA64-NEXT:    st.d $ra, $sp, 24 # 8-byte Folded Spill
+; LA64-NEXT:    fst.d $fs0, $sp, 16 # 8-byte Folded Spill
+; LA64-NEXT:    fst.d $fs1, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    fmov.s $fs0, $fa0
+; LA64-NEXT:    bl %plt(sinf)
+; LA64-NEXT:    fmov.s $fs1, $fa0
+; LA64-NEXT:    fmov.s $fa0, $fs0
+; LA64-NEXT:    bl %plt(cosf)
+; LA64-NEXT:    fadd.s $fa0, $fs1, $fa0
+; LA64-NEXT:    fld.d $fs1, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    fld.d $fs0, $sp, 16 # 8-byte Folded Reload
+; LA64-NEXT:    ld.d $ra, $sp, 24 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 32
+; LA64-NEXT:    ret
+  %1 = call float @llvm.sin.f32(float %a)
+  %2 = call float @llvm.cos.f32(float %a)
+  %3 = fadd float %1, %2
+  ret float %3
+}
+
+define float @pow_f32(float %a, float %b) nounwind {
+; LA32-LABEL: pow_f32:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -16
+; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
+; LA32-NEXT:    bl %plt(powf)
+; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 16
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: pow_f32:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -16
+; LA64-NEXT:    st.d $ra, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    bl %plt(powf)
+; LA64-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 16
+; LA64-NEXT:    ret
+  %1 = call float @llvm.pow.f32(float %a, float %b)
+  ret float %1
+}
+
+define float @frem_f32(float %a, float %b) nounwind {
+; LA32-LABEL: frem_f32:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -16
+; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
+; LA32-NEXT:    bl %plt(fmodf)
+; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 16
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: frem_f32:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -16
+; LA64-NEXT:    st.d $ra, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    bl %plt(fmodf)
+; LA64-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 16
+; LA64-NEXT:    ret
+  %1 = frem float %a, %b
+  ret float %1
+}
+
+define double @sin_f64(double %a) nounwind {
+; LA32-LABEL: sin_f64:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -16
+; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
+; LA32-NEXT:    bl %plt(sin)
+; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 16
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: sin_f64:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -16
+; LA64-NEXT:    st.d $ra, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    bl %plt(sin)
+; LA64-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 16
+; LA64-NEXT:    ret
+  %1 = call double @llvm.sin.f64(double %a)
+  ret double %1
+}
+
+define double @cos_f64(double %a) nounwind {
+; LA32-LABEL: cos_f64:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -16
+; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
+; LA32-NEXT:    bl %plt(cos)
+; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 16
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: cos_f64:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -16
+; LA64-NEXT:    st.d $ra, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    bl %plt(cos)
+; LA64-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 16
+; LA64-NEXT:    ret
+  %1 = call double @llvm.cos.f64(double %a)
+  ret double %1
+}
+
+define double @sincos_f64(double %a) nounwind {
+; LA32-LABEL: sincos_f64:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -32
+; LA32-NEXT:    st.w $ra, $sp, 28 # 4-byte Folded Spill
+; LA32-NEXT:    fst.d $fs0, $sp, 16 # 8-byte Folded Spill
+; LA32-NEXT:    fst.d $fs1, $sp, 8 # 8-byte Folded Spill
+; LA32-NEXT:    fmov.d $fs0, $fa0
+; LA32-NEXT:    bl %plt(sin)
+; LA32-NEXT:    fmov.d $fs1, $fa0
+; LA32-NEXT:    fmov.d $fa0, $fs0
+; LA32-NEXT:    bl %plt(cos)
+; LA32-NEXT:    fadd.d $fa0, $fs1, $fa0
+; LA32-NEXT:    fld.d $fs1, $sp, 8 # 8-byte Folded Reload
+; LA32-NEXT:    fld.d $fs0, $sp, 16 # 8-byte Folded Reload
+; LA32-NEXT:    ld.w $ra, $sp, 28 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 32
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: sincos_f64:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -32
+; LA64-NEXT:    st.d $ra, $sp, 24 # 8-byte Folded Spill
+; LA64-NEXT:    fst.d $fs0, $sp, 16 # 8-byte Folded Spill
+; LA64-NEXT:    fst.d $fs1, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    fmov.d $fs0, $fa0
+; LA64-NEXT:    bl %plt(sin)
+; LA64-NEXT:    fmov.d $fs1, $fa0
+; LA64-NEXT:    fmov.d $fa0, $fs0
+; LA64-NEXT:    bl %plt(cos)
+; LA64-NEXT:    fadd.d $fa0, $fs1, $fa0
+; LA64-NEXT:    fld.d $fs1, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    fld.d $fs0, $sp, 16 # 8-byte Folded Reload
+; LA64-NEXT:    ld.d $ra, $sp, 24 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 32
+; LA64-NEXT:    ret
+  %1 = call double @llvm.sin.f64(double %a)
+  %2 = call double @llvm.cos.f64(double %a)
+  %3 = fadd double %1, %2
+  ret double %3
+}
+
+define double @pow_f64(double %a, double %b) nounwind {
+; LA32-LABEL: pow_f64:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -16
+; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
+; LA32-NEXT:    bl %plt(pow)
+; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 16
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: pow_f64:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -16
+; LA64-NEXT:    st.d $ra, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    bl %plt(pow)
+; LA64-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 16
+; LA64-NEXT:    ret
+  %1 = call double @llvm.pow.f64(double %a, double %b)
+  ret double %1
+}
+
+define double @frem_f64(double %a, double %b) nounwind {
+; LA32-LABEL: frem_f64:
+; LA32:       # %bb.0:
+; LA32-NEXT:    addi.w $sp, $sp, -16
+; LA32-NEXT:    st.w $ra, $sp, 12 # 4-byte Folded Spill
+; LA32-NEXT:    bl %plt(fmod)
+; LA32-NEXT:    ld.w $ra, $sp, 12 # 4-byte Folded Reload
+; LA32-NEXT:    addi.w $sp, $sp, 16
+; LA32-NEXT:    ret
+;
+; LA64-LABEL: frem_f64:
+; LA64:       # %bb.0:
+; LA64-NEXT:    addi.d $sp, $sp, -16
+; LA64-NEXT:    st.d $ra, $sp, 8 # 8-byte Folded Spill
+; LA64-NEXT:    bl %plt(fmod)
+; LA64-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
+; LA64-NEXT:    addi.d $sp, $sp, 16
+; LA64-NEXT:    ret
+  %1 = frem double %a, %b
+  ret double %1
+}