[RISCV] Support vector strict rounding operations.
The patch basically models custom lowering of base rounding operations to expand
rounding by coverting to ingter and coverting back to FP. The other one thing
the patch does is to covert sNan of the source to qNan.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D148519
diff --git a/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ftrunc-constrained-sdnode.ll b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ftrunc-constrained-sdnode.ll
new file mode 100644
index 0000000..7c43e28
--- /dev/null
+++ b/llvm/test/CodeGen/RISCV/rvv/fixed-vectors-ftrunc-constrained-sdnode.ll
@@ -0,0 +1,306 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=riscv32 -mattr=+d,+zfh,+experimental-zvfh,+v -target-abi=ilp32d \
+; RUN: -verify-machineinstrs < %s | FileCheck %s
+; RUN: llc -mtriple=riscv64 -mattr=+d,+zfh,+experimental-zvfh,+v -target-abi=lp64d \
+; RUN: -verify-machineinstrs < %s | FileCheck %s
+
+define <1 x half> @trunc_v1f16(<1 x half> %x) {
+; CHECK-LABEL: trunc_v1f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 1, e16, mf4, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI0_0)
+; CHECK-NEXT: flh fa5, %lo(.LCPI0_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v9, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <1 x half> @llvm.experimental.constrained.trunc.v1f16(<1 x half> %x, metadata !"fpexcept.strict")
+ ret <1 x half> %a
+}
+declare <1 x half> @llvm.experimental.constrained.trunc.v1f16(<1 x half>, metadata)
+
+define <2 x half> @trunc_v2f16(<2 x half> %x) {
+; CHECK-LABEL: trunc_v2f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 2, e16, mf4, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI1_0)
+; CHECK-NEXT: flh fa5, %lo(.LCPI1_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v9, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <2 x half> @llvm.experimental.constrained.trunc.v2f16(<2 x half> %x, metadata !"fpexcept.strict")
+ ret <2 x half> %a
+}
+declare <2 x half> @llvm.experimental.constrained.trunc.v2f16(<2 x half>, metadata)
+
+define <4 x half> @trunc_v4f16(<4 x half> %x) {
+; CHECK-LABEL: trunc_v4f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 4, e16, mf2, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI2_0)
+; CHECK-NEXT: flh fa5, %lo(.LCPI2_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v9, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <4 x half> @llvm.experimental.constrained.trunc.v4f16(<4 x half> %x, metadata !"fpexcept.strict")
+ ret <4 x half> %a
+}
+declare <4 x half> @llvm.experimental.constrained.trunc.v4f16(<4 x half>, metadata)
+
+define <8 x half> @trunc_v8f16(<8 x half> %x) {
+; CHECK-LABEL: trunc_v8f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 8, e16, m1, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI3_0)
+; CHECK-NEXT: flh fa5, %lo(.LCPI3_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v9, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <8 x half> @llvm.experimental.constrained.trunc.v8f16(<8 x half> %x, metadata !"fpexcept.strict")
+ ret <8 x half> %a
+}
+declare <8 x half> @llvm.experimental.constrained.trunc.v8f16(<8 x half>, metadata)
+
+define <16 x half> @trunc_v16f16(<16 x half> %x) {
+; CHECK-LABEL: trunc_v16f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 16, e16, m2, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI4_0)
+; CHECK-NEXT: flh fa5, %lo(.LCPI4_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v10, v8
+; CHECK-NEXT: vmflt.vf v0, v10, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v10, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <16 x half> @llvm.experimental.constrained.trunc.v16f16(<16 x half> %x, metadata !"fpexcept.strict")
+ ret <16 x half> %a
+}
+declare <16 x half> @llvm.experimental.constrained.trunc.v16f16(<16 x half>, metadata)
+
+define <32 x half> @trunc_v32f16(<32 x half> %x) {
+; CHECK-LABEL: trunc_v32f16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: li a0, 32
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI5_0)
+; CHECK-NEXT: flh fa5, %lo(.LCPI5_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v12, v8
+; CHECK-NEXT: vmflt.vf v0, v12, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v12, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e16, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <32 x half> @llvm.experimental.constrained.trunc.v32f16(<32 x half> %x, metadata !"fpexcept.strict")
+ ret <32 x half> %a
+}
+declare <32 x half> @llvm.experimental.constrained.trunc.v32f16(<32 x half>, metadata)
+
+define <1 x float> @trunc_v1f32(<1 x float> %x) {
+; CHECK-LABEL: trunc_v1f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 1, e32, mf2, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: lui a0, 307200
+; CHECK-NEXT: fmv.w.x fa5, a0
+; CHECK-NEXT: vmflt.vf v0, v9, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v9, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <1 x float> @llvm.experimental.constrained.trunc.v1f32(<1 x float> %x, metadata !"fpexcept.strict")
+ ret <1 x float> %a
+}
+declare <1 x float> @llvm.experimental.constrained.trunc.v1f32(<1 x float>, metadata)
+
+define <2 x float> @trunc_v2f32(<2 x float> %x) {
+; CHECK-LABEL: trunc_v2f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 2, e32, mf2, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: lui a0, 307200
+; CHECK-NEXT: fmv.w.x fa5, a0
+; CHECK-NEXT: vmflt.vf v0, v9, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v9, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <2 x float> @llvm.experimental.constrained.trunc.v2f32(<2 x float> %x, metadata !"fpexcept.strict")
+ ret <2 x float> %a
+}
+declare <2 x float> @llvm.experimental.constrained.trunc.v2f32(<2 x float>, metadata)
+
+define <4 x float> @trunc_v4f32(<4 x float> %x) {
+; CHECK-LABEL: trunc_v4f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 4, e32, m1, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: lui a0, 307200
+; CHECK-NEXT: fmv.w.x fa5, a0
+; CHECK-NEXT: vmflt.vf v0, v9, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v9, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <4 x float> @llvm.experimental.constrained.trunc.v4f32(<4 x float> %x, metadata !"fpexcept.strict")
+ ret <4 x float> %a
+}
+declare <4 x float> @llvm.experimental.constrained.trunc.v4f32(<4 x float>, metadata)
+
+define <8 x float> @trunc_v8f32(<8 x float> %x) {
+; CHECK-LABEL: trunc_v8f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v10, v8
+; CHECK-NEXT: lui a0, 307200
+; CHECK-NEXT: fmv.w.x fa5, a0
+; CHECK-NEXT: vmflt.vf v0, v10, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v10, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <8 x float> @llvm.experimental.constrained.trunc.v8f32(<8 x float> %x, metadata !"fpexcept.strict")
+ ret <8 x float> %a
+}
+declare <8 x float> @llvm.experimental.constrained.trunc.v8f32(<8 x float>, metadata)
+
+define <16 x float> @trunc_v16f32(<16 x float> %x) {
+; CHECK-LABEL: trunc_v16f32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 16, e32, m4, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v12, v8
+; CHECK-NEXT: lui a0, 307200
+; CHECK-NEXT: fmv.w.x fa5, a0
+; CHECK-NEXT: vmflt.vf v0, v12, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v12, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <16 x float> @llvm.experimental.constrained.trunc.v16f32(<16 x float> %x, metadata !"fpexcept.strict")
+ ret <16 x float> %a
+}
+declare <16 x float> @llvm.experimental.constrained.trunc.v16f32(<16 x float>, metadata)
+
+define <1 x double> @trunc_v1f64(<1 x double> %x) {
+; CHECK-LABEL: trunc_v1f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 1, e64, m1, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI11_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI11_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v9, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <1 x double> @llvm.experimental.constrained.trunc.v1f64(<1 x double> %x, metadata !"fpexcept.strict")
+ ret <1 x double> %a
+}
+declare <1 x double> @llvm.experimental.constrained.trunc.v1f64(<1 x double>, metadata)
+
+define <2 x double> @trunc_v2f64(<2 x double> %x) {
+; CHECK-LABEL: trunc_v2f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 2, e64, m1, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI12_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI12_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v9, v8
+; CHECK-NEXT: vmflt.vf v0, v9, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v9, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v9, v9, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v9, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <2 x double> @llvm.experimental.constrained.trunc.v2f64(<2 x double> %x, metadata !"fpexcept.strict")
+ ret <2 x double> %a
+}
+declare <2 x double> @llvm.experimental.constrained.trunc.v2f64(<2 x double>, metadata)
+
+define <4 x double> @trunc_v4f64(<4 x double> %x) {
+; CHECK-LABEL: trunc_v4f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI13_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI13_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v10, v8
+; CHECK-NEXT: vmflt.vf v0, v10, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v10, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v10, v10, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v10, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <4 x double> @llvm.experimental.constrained.trunc.v4f64(<4 x double> %x, metadata !"fpexcept.strict")
+ ret <4 x double> %a
+}
+declare <4 x double> @llvm.experimental.constrained.trunc.v4f64(<4 x double>, metadata)
+
+define <8 x double> @trunc_v8f64(<8 x double> %x) {
+; CHECK-LABEL: trunc_v8f64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetivli zero, 8, e64, m4, ta, ma
+; CHECK-NEXT: vmfne.vv v0, v8, v8
+; CHECK-NEXT: lui a0, %hi(.LCPI14_0)
+; CHECK-NEXT: fld fa5, %lo(.LCPI14_0)(a0)
+; CHECK-NEXT: vfadd.vv v8, v8, v8, v0.t
+; CHECK-NEXT: vfabs.v v12, v8
+; CHECK-NEXT: vmflt.vf v0, v12, fa5
+; CHECK-NEXT: vfcvt.rtz.x.f.v v12, v8, v0.t
+; CHECK-NEXT: vfcvt.f.x.v v12, v12, v0.t
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, ta, mu
+; CHECK-NEXT: vfsgnj.vv v8, v12, v8, v0.t
+; CHECK-NEXT: ret
+ %a = call <8 x double> @llvm.experimental.constrained.trunc.v8f64(<8 x double> %x, metadata !"fpexcept.strict")
+ ret <8 x double> %a
+}
+declare <8 x double> @llvm.experimental.constrained.trunc.v8f64(<8 x double>, metadata)
