; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: sed 's/iXLen/i32/g' %s | llc -mtriple=riscv32 -mattr=+v,+zfh,+zvfh \
; RUN: -verify-machineinstrs -target-abi=ilp32d | FileCheck %s
; RUN: sed 's/iXLen/i64/g' %s | llc -mtriple=riscv64 -mattr=+v,+zfh,+zvfh \
; RUN: -verify-machineinstrs -target-abi=lp64d | FileCheck %s
declare <vscale x 1 x float> @llvm.riscv.vfwsub.w.nxv1f32.nxv1f16(
<vscale x 1 x float>,
<vscale x 1 x float>,
<vscale x 1 x half>,
iXLen, iXLen);
define <vscale x 1 x float> @intrinsic_vfwsub.w_wv_nxv1f32_nxv1f32_nxv1f16(<vscale x 1 x float> %0, <vscale x 1 x half> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_nxv1f32_nxv1f32_nxv1f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v8, v9
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x float> @llvm.riscv.vfwsub.w.nxv1f32.nxv1f16(
<vscale x 1 x float> undef,
<vscale x 1 x float> %0,
<vscale x 1 x half> %1,
iXLen 0, iXLen %2)
ret <vscale x 1 x float> %a
}
declare <vscale x 1 x float> @llvm.riscv.vfwsub.w.mask.nxv1f32.nxv1f16(
<vscale x 1 x float>,
<vscale x 1 x float>,
<vscale x 1 x half>,
<vscale x 1 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 1 x float> @intrinsic_vfwsub.w_mask_wv_nxv1f32_nxv1f32_nxv1f16(<vscale x 1 x float> %0, <vscale x 1 x float> %1, <vscale x 1 x half> %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_nxv1f32_nxv1f32_nxv1f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v9, v10, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x float> @llvm.riscv.vfwsub.w.mask.nxv1f32.nxv1f16(
<vscale x 1 x float> %0,
<vscale x 1 x float> %1,
<vscale x 1 x half> %2,
<vscale x 1 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 1 x float> %a
}
declare <vscale x 2 x float> @llvm.riscv.vfwsub.w.nxv2f32.nxv2f16(
<vscale x 2 x float>,
<vscale x 2 x float>,
<vscale x 2 x half>,
iXLen, iXLen);
define <vscale x 2 x float> @intrinsic_vfwsub.w_wv_nxv2f32_nxv2f32_nxv2f16(<vscale x 2 x float> %0, <vscale x 2 x half> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_nxv2f32_nxv2f32_nxv2f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v8, v9
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x float> @llvm.riscv.vfwsub.w.nxv2f32.nxv2f16(
<vscale x 2 x float> undef,
<vscale x 2 x float> %0,
<vscale x 2 x half> %1,
iXLen 0, iXLen %2)
ret <vscale x 2 x float> %a
}
declare <vscale x 2 x float> @llvm.riscv.vfwsub.w.mask.nxv2f32.nxv2f16(
<vscale x 2 x float>,
<vscale x 2 x float>,
<vscale x 2 x half>,
<vscale x 2 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 2 x float> @intrinsic_vfwsub.w_mask_wv_nxv2f32_nxv2f32_nxv2f16(<vscale x 2 x float> %0, <vscale x 2 x float> %1, <vscale x 2 x half> %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_nxv2f32_nxv2f32_nxv2f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v9, v10, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x float> @llvm.riscv.vfwsub.w.mask.nxv2f32.nxv2f16(
<vscale x 2 x float> %0,
<vscale x 2 x float> %1,
<vscale x 2 x half> %2,
<vscale x 2 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 2 x float> %a
}
declare <vscale x 4 x float> @llvm.riscv.vfwsub.w.nxv4f32.nxv4f16(
<vscale x 4 x float>,
<vscale x 4 x float>,
<vscale x 4 x half>,
iXLen, iXLen);
define <vscale x 4 x float> @intrinsic_vfwsub.w_wv_nxv4f32_nxv4f32_nxv4f16(<vscale x 4 x float> %0, <vscale x 4 x half> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_nxv4f32_nxv4f32_nxv4f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v8, v10
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x float> @llvm.riscv.vfwsub.w.nxv4f32.nxv4f16(
<vscale x 4 x float> undef,
<vscale x 4 x float> %0,
<vscale x 4 x half> %1,
iXLen 0, iXLen %2)
ret <vscale x 4 x float> %a
}
declare <vscale x 4 x float> @llvm.riscv.vfwsub.w.mask.nxv4f32.nxv4f16(
<vscale x 4 x float>,
<vscale x 4 x float>,
<vscale x 4 x half>,
<vscale x 4 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 4 x float> @intrinsic_vfwsub.w_mask_wv_nxv4f32_nxv4f32_nxv4f16(<vscale x 4 x float> %0, <vscale x 4 x float> %1, <vscale x 4 x half> %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_nxv4f32_nxv4f32_nxv4f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v10, v12, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x float> @llvm.riscv.vfwsub.w.mask.nxv4f32.nxv4f16(
<vscale x 4 x float> %0,
<vscale x 4 x float> %1,
<vscale x 4 x half> %2,
<vscale x 4 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 4 x float> %a
}
declare <vscale x 8 x float> @llvm.riscv.vfwsub.w.nxv8f32.nxv8f16(
<vscale x 8 x float>,
<vscale x 8 x float>,
<vscale x 8 x half>,
iXLen, iXLen);
define <vscale x 8 x float> @intrinsic_vfwsub.w_wv_nxv8f32_nxv8f32_nxv8f16(<vscale x 8 x float> %0, <vscale x 8 x half> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_nxv8f32_nxv8f32_nxv8f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v8, v12
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x float> @llvm.riscv.vfwsub.w.nxv8f32.nxv8f16(
<vscale x 8 x float> undef,
<vscale x 8 x float> %0,
<vscale x 8 x half> %1,
iXLen 0, iXLen %2)
ret <vscale x 8 x float> %a
}
declare <vscale x 8 x float> @llvm.riscv.vfwsub.w.mask.nxv8f32.nxv8f16(
<vscale x 8 x float>,
<vscale x 8 x float>,
<vscale x 8 x half>,
<vscale x 8 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 8 x float> @intrinsic_vfwsub.w_mask_wv_nxv8f32_nxv8f32_nxv8f16(<vscale x 8 x float> %0, <vscale x 8 x float> %1, <vscale x 8 x half> %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_nxv8f32_nxv8f32_nxv8f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v12, v16, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x float> @llvm.riscv.vfwsub.w.mask.nxv8f32.nxv8f16(
<vscale x 8 x float> %0,
<vscale x 8 x float> %1,
<vscale x 8 x half> %2,
<vscale x 8 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 8 x float> %a
}
declare <vscale x 16 x float> @llvm.riscv.vfwsub.w.nxv16f32.nxv16f16(
<vscale x 16 x float>,
<vscale x 16 x float>,
<vscale x 16 x half>,
iXLen, iXLen);
define <vscale x 16 x float> @intrinsic_vfwsub.w_wv_nxv16f32_nxv16f32_nxv16f16(<vscale x 16 x float> %0, <vscale x 16 x half> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_nxv16f32_nxv16f32_nxv16f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v8, v16
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 16 x float> @llvm.riscv.vfwsub.w.nxv16f32.nxv16f16(
<vscale x 16 x float> undef,
<vscale x 16 x float> %0,
<vscale x 16 x half> %1,
iXLen 0, iXLen %2)
ret <vscale x 16 x float> %a
}
declare <vscale x 16 x float> @llvm.riscv.vfwsub.w.mask.nxv16f32.nxv16f16(
<vscale x 16 x float>,
<vscale x 16 x float>,
<vscale x 16 x half>,
<vscale x 16 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 16 x float> @intrinsic_vfwsub.w_mask_wv_nxv16f32_nxv16f32_nxv16f16(<vscale x 16 x float> %0, <vscale x 16 x float> %1, <vscale x 16 x half> %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_nxv16f32_nxv16f32_nxv16f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vl4re16.v v24, (a0)
; CHECK-NEXT: fsrmi a0, 0
; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v16, v24, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: ret
entry:
%a = call <vscale x 16 x float> @llvm.riscv.vfwsub.w.mask.nxv16f32.nxv16f16(
<vscale x 16 x float> %0,
<vscale x 16 x float> %1,
<vscale x 16 x half> %2,
<vscale x 16 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 16 x float> %a
}
declare <vscale x 1 x double> @llvm.riscv.vfwsub.w.nxv1f64.nxv1f32(
<vscale x 1 x double>,
<vscale x 1 x double>,
<vscale x 1 x float>,
iXLen, iXLen);
define <vscale x 1 x double> @intrinsic_vfwsub.w_wv_nxv1f64_nxv1f64_nxv1f32(<vscale x 1 x double> %0, <vscale x 1 x float> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_nxv1f64_nxv1f64_nxv1f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v8, v9
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.nxv1f64.nxv1f32(
<vscale x 1 x double> undef,
<vscale x 1 x double> %0,
<vscale x 1 x float> %1,
iXLen 0, iXLen %2)
ret <vscale x 1 x double> %a
}
declare <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.nxv1f32(
<vscale x 1 x double>,
<vscale x 1 x double>,
<vscale x 1 x float>,
<vscale x 1 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 1 x double> @intrinsic_vfwsub.w_mask_wv_nxv1f64_nxv1f64_nxv1f32(<vscale x 1 x double> %0, <vscale x 1 x double> %1, <vscale x 1 x float> %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_nxv1f64_nxv1f64_nxv1f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v9, v10, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.nxv1f32(
<vscale x 1 x double> %0,
<vscale x 1 x double> %1,
<vscale x 1 x float> %2,
<vscale x 1 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 1 x double> %a
}
declare <vscale x 2 x double> @llvm.riscv.vfwsub.w.nxv2f64.nxv2f32(
<vscale x 2 x double>,
<vscale x 2 x double>,
<vscale x 2 x float>,
iXLen, iXLen);
define <vscale x 2 x double> @intrinsic_vfwsub.w_wv_nxv2f64_nxv2f64_nxv2f32(<vscale x 2 x double> %0, <vscale x 2 x float> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_nxv2f64_nxv2f64_nxv2f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v8, v10
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.nxv2f64.nxv2f32(
<vscale x 2 x double> undef,
<vscale x 2 x double> %0,
<vscale x 2 x float> %1,
iXLen 0, iXLen %2)
ret <vscale x 2 x double> %a
}
declare <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.nxv2f32(
<vscale x 2 x double>,
<vscale x 2 x double>,
<vscale x 2 x float>,
<vscale x 2 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 2 x double> @intrinsic_vfwsub.w_mask_wv_nxv2f64_nxv2f64_nxv2f32(<vscale x 2 x double> %0, <vscale x 2 x double> %1, <vscale x 2 x float> %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_nxv2f64_nxv2f64_nxv2f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v10, v12, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.nxv2f32(
<vscale x 2 x double> %0,
<vscale x 2 x double> %1,
<vscale x 2 x float> %2,
<vscale x 2 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 2 x double> %a
}
declare <vscale x 4 x double> @llvm.riscv.vfwsub.w.nxv4f64.nxv4f32(
<vscale x 4 x double>,
<vscale x 4 x double>,
<vscale x 4 x float>,
iXLen, iXLen);
define <vscale x 4 x double> @intrinsic_vfwsub.w_wv_nxv4f64_nxv4f64_nxv4f32(<vscale x 4 x double> %0, <vscale x 4 x float> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_nxv4f64_nxv4f64_nxv4f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v8, v12
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.nxv4f64.nxv4f32(
<vscale x 4 x double> undef,
<vscale x 4 x double> %0,
<vscale x 4 x float> %1,
iXLen 0, iXLen %2)
ret <vscale x 4 x double> %a
}
declare <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.nxv4f32(
<vscale x 4 x double>,
<vscale x 4 x double>,
<vscale x 4 x float>,
<vscale x 4 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 4 x double> @intrinsic_vfwsub.w_mask_wv_nxv4f64_nxv4f64_nxv4f32(<vscale x 4 x double> %0, <vscale x 4 x double> %1, <vscale x 4 x float> %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_nxv4f64_nxv4f64_nxv4f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v12, v16, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.nxv4f32(
<vscale x 4 x double> %0,
<vscale x 4 x double> %1,
<vscale x 4 x float> %2,
<vscale x 4 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 4 x double> %a
}
declare <vscale x 8 x double> @llvm.riscv.vfwsub.w.nxv8f64.nxv8f32(
<vscale x 8 x double>,
<vscale x 8 x double>,
<vscale x 8 x float>,
iXLen, iXLen);
define <vscale x 8 x double> @intrinsic_vfwsub.w_wv_nxv8f64_nxv8f64_nxv8f32(<vscale x 8 x double> %0, <vscale x 8 x float> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_nxv8f64_nxv8f64_nxv8f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vfwsub.wv v8, v8, v16
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.nxv8f64.nxv8f32(
<vscale x 8 x double> undef,
<vscale x 8 x double> %0,
<vscale x 8 x float> %1,
iXLen 0, iXLen %2)
ret <vscale x 8 x double> %a
}
declare <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.nxv8f32(
<vscale x 8 x double>,
<vscale x 8 x double>,
<vscale x 8 x float>,
<vscale x 8 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 8 x double> @intrinsic_vfwsub.w_mask_wv_nxv8f64_nxv8f64_nxv8f32(<vscale x 8 x double> %0, <vscale x 8 x double> %1, <vscale x 8 x float> %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_nxv8f64_nxv8f64_nxv8f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vl4re32.v v24, (a0)
; CHECK-NEXT: fsrmi a0, 0
; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v16, v24, v0.t
; CHECK-NEXT: fsrm a0
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.nxv8f32(
<vscale x 8 x double> %0,
<vscale x 8 x double> %1,
<vscale x 8 x float> %2,
<vscale x 8 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 8 x double> %a
}
declare <vscale x 1 x float> @llvm.riscv.vfwsub.w.nxv1f32.f16(
<vscale x 1 x float>,
<vscale x 1 x float>,
half,
iXLen, iXLen);
define <vscale x 1 x float> @intrinsic_vfwsub.w_wf_nxv1f32_nxv1f32_f16(<vscale x 1 x float> %0, half %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wf_nxv1f32_nxv1f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
; CHECK-NEXT: vfwsub.wf v8, v8, fa0
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x float> @llvm.riscv.vfwsub.w.nxv1f32.f16(
<vscale x 1 x float> undef,
<vscale x 1 x float> %0,
half %1,
iXLen 0, iXLen %2)
ret <vscale x 1 x float> %a
}
declare <vscale x 1 x float> @llvm.riscv.vfwsub.w.mask.nxv1f32.f16(
<vscale x 1 x float>,
<vscale x 1 x float>,
half,
<vscale x 1 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 1 x float> @intrinsic_vfwsub.w_mask_wf_nxv1f32_nxv1f32_f16(<vscale x 1 x float> %0, <vscale x 1 x float> %1, half %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_nxv1f32_nxv1f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v9, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x float> @llvm.riscv.vfwsub.w.mask.nxv1f32.f16(
<vscale x 1 x float> %0,
<vscale x 1 x float> %1,
half %2,
<vscale x 1 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 1 x float> %a
}
declare <vscale x 2 x float> @llvm.riscv.vfwsub.w.nxv2f32.f16(
<vscale x 2 x float>,
<vscale x 2 x float>,
half,
iXLen, iXLen);
define <vscale x 2 x float> @intrinsic_vfwsub.w_wf_nxv2f32_nxv2f32_f16(<vscale x 2 x float> %0, half %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wf_nxv2f32_nxv2f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
; CHECK-NEXT: vfwsub.wf v8, v8, fa0
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x float> @llvm.riscv.vfwsub.w.nxv2f32.f16(
<vscale x 2 x float> undef,
<vscale x 2 x float> %0,
half %1,
iXLen 0, iXLen %2)
ret <vscale x 2 x float> %a
}
declare <vscale x 2 x float> @llvm.riscv.vfwsub.w.mask.nxv2f32.f16(
<vscale x 2 x float>,
<vscale x 2 x float>,
half,
<vscale x 2 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 2 x float> @intrinsic_vfwsub.w_mask_wf_nxv2f32_nxv2f32_f16(<vscale x 2 x float> %0, <vscale x 2 x float> %1, half %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_nxv2f32_nxv2f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v9, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x float> @llvm.riscv.vfwsub.w.mask.nxv2f32.f16(
<vscale x 2 x float> %0,
<vscale x 2 x float> %1,
half %2,
<vscale x 2 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 2 x float> %a
}
declare <vscale x 4 x float> @llvm.riscv.vfwsub.w.nxv4f32.f16(
<vscale x 4 x float>,
<vscale x 4 x float>,
half,
iXLen, iXLen);
define <vscale x 4 x float> @intrinsic_vfwsub.w_wf_nxv4f32_nxv4f32_f16(<vscale x 4 x float> %0, half %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wf_nxv4f32_nxv4f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
; CHECK-NEXT: vfwsub.wf v8, v8, fa0
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x float> @llvm.riscv.vfwsub.w.nxv4f32.f16(
<vscale x 4 x float> undef,
<vscale x 4 x float> %0,
half %1,
iXLen 0, iXLen %2)
ret <vscale x 4 x float> %a
}
declare <vscale x 4 x float> @llvm.riscv.vfwsub.w.mask.nxv4f32.f16(
<vscale x 4 x float>,
<vscale x 4 x float>,
half,
<vscale x 4 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 4 x float> @intrinsic_vfwsub.w_mask_wf_nxv4f32_nxv4f32_f16(<vscale x 4 x float> %0, <vscale x 4 x float> %1, half %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_nxv4f32_nxv4f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v10, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x float> @llvm.riscv.vfwsub.w.mask.nxv4f32.f16(
<vscale x 4 x float> %0,
<vscale x 4 x float> %1,
half %2,
<vscale x 4 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 4 x float> %a
}
declare <vscale x 8 x float> @llvm.riscv.vfwsub.w.nxv8f32.f16(
<vscale x 8 x float>,
<vscale x 8 x float>,
half,
iXLen, iXLen);
define <vscale x 8 x float> @intrinsic_vfwsub.w_wf_nxv8f32_nxv8f32_f16(<vscale x 8 x float> %0, half %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wf_nxv8f32_nxv8f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; CHECK-NEXT: vfwsub.wf v8, v8, fa0
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x float> @llvm.riscv.vfwsub.w.nxv8f32.f16(
<vscale x 8 x float> undef,
<vscale x 8 x float> %0,
half %1,
iXLen 0, iXLen %2)
ret <vscale x 8 x float> %a
}
declare <vscale x 8 x float> @llvm.riscv.vfwsub.w.mask.nxv8f32.f16(
<vscale x 8 x float>,
<vscale x 8 x float>,
half,
<vscale x 8 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 8 x float> @intrinsic_vfwsub.w_mask_wf_nxv8f32_nxv8f32_f16(<vscale x 8 x float> %0, <vscale x 8 x float> %1, half %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_nxv8f32_nxv8f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v12, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x float> @llvm.riscv.vfwsub.w.mask.nxv8f32.f16(
<vscale x 8 x float> %0,
<vscale x 8 x float> %1,
half %2,
<vscale x 8 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 8 x float> %a
}
declare <vscale x 16 x float> @llvm.riscv.vfwsub.w.nxv16f32.f16(
<vscale x 16 x float>,
<vscale x 16 x float>,
half,
iXLen, iXLen);
define <vscale x 16 x float> @intrinsic_vfwsub.w_wf_nxv16f32_nxv16f32_f16(<vscale x 16 x float> %0, half %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wf_nxv16f32_nxv16f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
; CHECK-NEXT: vfwsub.wf v8, v8, fa0
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 16 x float> @llvm.riscv.vfwsub.w.nxv16f32.f16(
<vscale x 16 x float> undef,
<vscale x 16 x float> %0,
half %1,
iXLen 0, iXLen %2)
ret <vscale x 16 x float> %a
}
declare <vscale x 16 x float> @llvm.riscv.vfwsub.w.mask.nxv16f32.f16(
<vscale x 16 x float>,
<vscale x 16 x float>,
half,
<vscale x 16 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 16 x float> @intrinsic_vfwsub.w_mask_wf_nxv16f32_nxv16f32_f16(<vscale x 16 x float> %0, <vscale x 16 x float> %1, half %2, <vscale x 16 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_nxv16f32_nxv16f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v16, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 16 x float> @llvm.riscv.vfwsub.w.mask.nxv16f32.f16(
<vscale x 16 x float> %0,
<vscale x 16 x float> %1,
half %2,
<vscale x 16 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 16 x float> %a
}
declare <vscale x 1 x double> @llvm.riscv.vfwsub.w.nxv1f64.f32(
<vscale x 1 x double>,
<vscale x 1 x double>,
float,
iXLen, iXLen);
define <vscale x 1 x double> @intrinsic_vfwsub.w_wf_nxv1f64_nxv1f64_f32(<vscale x 1 x double> %0, float %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wf_nxv1f64_nxv1f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
; CHECK-NEXT: vfwsub.wf v8, v8, fa0
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.nxv1f64.f32(
<vscale x 1 x double> undef,
<vscale x 1 x double> %0,
float %1,
iXLen 0, iXLen %2)
ret <vscale x 1 x double> %a
}
declare <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.f32(
<vscale x 1 x double>,
<vscale x 1 x double>,
float,
<vscale x 1 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 1 x double> @intrinsic_vfwsub.w_mask_wf_nxv1f64_nxv1f64_f32(<vscale x 1 x double> %0, <vscale x 1 x double> %1, float %2, <vscale x 1 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_nxv1f64_nxv1f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v9, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.f32(
<vscale x 1 x double> %0,
<vscale x 1 x double> %1,
float %2,
<vscale x 1 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 1 x double> %a
}
declare <vscale x 2 x double> @llvm.riscv.vfwsub.w.nxv2f64.f32(
<vscale x 2 x double>,
<vscale x 2 x double>,
float,
iXLen, iXLen);
define <vscale x 2 x double> @intrinsic_vfwsub.w_wf_nxv2f64_nxv2f64_f32(<vscale x 2 x double> %0, float %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wf_nxv2f64_nxv2f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
; CHECK-NEXT: vfwsub.wf v8, v8, fa0
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.nxv2f64.f32(
<vscale x 2 x double> undef,
<vscale x 2 x double> %0,
float %1,
iXLen 0, iXLen %2)
ret <vscale x 2 x double> %a
}
declare <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.f32(
<vscale x 2 x double>,
<vscale x 2 x double>,
float,
<vscale x 2 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 2 x double> @intrinsic_vfwsub.w_mask_wf_nxv2f64_nxv2f64_f32(<vscale x 2 x double> %0, <vscale x 2 x double> %1, float %2, <vscale x 2 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_nxv2f64_nxv2f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v10, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.f32(
<vscale x 2 x double> %0,
<vscale x 2 x double> %1,
float %2,
<vscale x 2 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 2 x double> %a
}
declare <vscale x 4 x double> @llvm.riscv.vfwsub.w.nxv4f64.f32(
<vscale x 4 x double>,
<vscale x 4 x double>,
float,
iXLen, iXLen);
define <vscale x 4 x double> @intrinsic_vfwsub.w_wf_nxv4f64_nxv4f64_f32(<vscale x 4 x double> %0, float %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wf_nxv4f64_nxv4f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; CHECK-NEXT: vfwsub.wf v8, v8, fa0
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.nxv4f64.f32(
<vscale x 4 x double> undef,
<vscale x 4 x double> %0,
float %1,
iXLen 0, iXLen %2)
ret <vscale x 4 x double> %a
}
declare <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.f32(
<vscale x 4 x double>,
<vscale x 4 x double>,
float,
<vscale x 4 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 4 x double> @intrinsic_vfwsub.w_mask_wf_nxv4f64_nxv4f64_f32(<vscale x 4 x double> %0, <vscale x 4 x double> %1, float %2, <vscale x 4 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_nxv4f64_nxv4f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v12, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.f32(
<vscale x 4 x double> %0,
<vscale x 4 x double> %1,
float %2,
<vscale x 4 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 4 x double> %a
}
declare <vscale x 8 x double> @llvm.riscv.vfwsub.w.nxv8f64.f32(
<vscale x 8 x double>,
<vscale x 8 x double>,
float,
iXLen, iXLen);
define <vscale x 8 x double> @intrinsic_vfwsub.w_wf_nxv8f64_nxv8f64_f32(<vscale x 8 x double> %0, float %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wf_nxv8f64_nxv8f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vfwsub.wf v8, v8, fa0
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.nxv8f64.f32(
<vscale x 8 x double> undef,
<vscale x 8 x double> %0,
float %1,
iXLen 0, iXLen %2)
ret <vscale x 8 x double> %a
}
declare <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.f32(
<vscale x 8 x double>,
<vscale x 8 x double>,
float,
<vscale x 8 x i1>,
iXLen, iXLen, iXLen);
define <vscale x 8 x double> @intrinsic_vfwsub.w_mask_wf_nxv8f64_nxv8f64_f32(<vscale x 8 x double> %0, <vscale x 8 x double> %1, float %2, <vscale x 8 x i1> %3, iXLen %4) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_nxv8f64_nxv8f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v16, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.f32(
<vscale x 8 x double> %0,
<vscale x 8 x double> %1,
float %2,
<vscale x 8 x i1> %3,
iXLen 0, iXLen %4, iXLen 1)
ret <vscale x 8 x double> %a
}
define <vscale x 1 x float> @intrinsic_vfwsub.w_mask_wv_tie_nxv1f32_nxv1f32_nxv1f16(<vscale x 1 x float> %0, <vscale x 1 x half> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_tie_nxv1f32_nxv1f32_nxv1f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v8, v9, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x float> @llvm.riscv.vfwsub.w.mask.nxv1f32.nxv1f16(
<vscale x 1 x float> %0,
<vscale x 1 x float> %0,
<vscale x 1 x half> %1,
<vscale x 1 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 1 x float> %a
}
define <vscale x 2 x float> @intrinsic_vfwsub.w_mask_wv_tie_nxv2f32_nxv2f32_nxv2f16(<vscale x 2 x float> %0, <vscale x 2 x half> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_tie_nxv2f32_nxv2f32_nxv2f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v8, v9, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x float> @llvm.riscv.vfwsub.w.mask.nxv2f32.nxv2f16(
<vscale x 2 x float> %0,
<vscale x 2 x float> %0,
<vscale x 2 x half> %1,
<vscale x 2 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 2 x float> %a
}
define <vscale x 4 x float> @intrinsic_vfwsub.w_mask_wv_tie_nxv4f32_nxv4f32_nxv4f16(<vscale x 4 x float> %0, <vscale x 4 x half> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_tie_nxv4f32_nxv4f32_nxv4f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v8, v10, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x float> @llvm.riscv.vfwsub.w.mask.nxv4f32.nxv4f16(
<vscale x 4 x float> %0,
<vscale x 4 x float> %0,
<vscale x 4 x half> %1,
<vscale x 4 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 4 x float> %a
}
define <vscale x 8 x float> @intrinsic_vfwsub.w_mask_wv_tie_nxv8f32_nxv8f32_nxv8f16(<vscale x 8 x float> %0, <vscale x 8 x half> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_tie_nxv8f32_nxv8f32_nxv8f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v8, v12, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x float> @llvm.riscv.vfwsub.w.mask.nxv8f32.nxv8f16(
<vscale x 8 x float> %0,
<vscale x 8 x float> %0,
<vscale x 8 x half> %1,
<vscale x 8 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 8 x float> %a
}
define <vscale x 16 x float> @intrinsic_vfwsub.w_mask_wv_tie_nxv16f32_nxv16f32_nxv16f16(<vscale x 16 x float> %0, <vscale x 16 x half> %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_tie_nxv16f32_nxv16f32_nxv16f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v8, v16, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 16 x float> @llvm.riscv.vfwsub.w.mask.nxv16f32.nxv16f16(
<vscale x 16 x float> %0,
<vscale x 16 x float> %0,
<vscale x 16 x half> %1,
<vscale x 16 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 16 x float> %a
}
define <vscale x 1 x double> @intrinsic_vfwsub.w_mask_wv_tie_nxv1f64_nxv1f64_nxv1f32(<vscale x 1 x double> %0, <vscale x 1 x float> %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_tie_nxv1f64_nxv1f64_nxv1f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v8, v9, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.nxv1f32(
<vscale x 1 x double> %0,
<vscale x 1 x double> %0,
<vscale x 1 x float> %1,
<vscale x 1 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 1 x double> %a
}
define <vscale x 2 x double> @intrinsic_vfwsub.w_mask_wv_tie_nxv2f64_nxv2f64_nxv2f32(<vscale x 2 x double> %0, <vscale x 2 x float> %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_tie_nxv2f64_nxv2f64_nxv2f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v8, v10, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.nxv2f32(
<vscale x 2 x double> %0,
<vscale x 2 x double> %0,
<vscale x 2 x float> %1,
<vscale x 2 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 2 x double> %a
}
define <vscale x 4 x double> @intrinsic_vfwsub.w_mask_wv_tie_nxv4f64_nxv4f64_nxv4f32(<vscale x 4 x double> %0, <vscale x 4 x float> %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_tie_nxv4f64_nxv4f64_nxv4f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v8, v12, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.nxv4f32(
<vscale x 4 x double> %0,
<vscale x 4 x double> %0,
<vscale x 4 x float> %1,
<vscale x 4 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 4 x double> %a
}
define <vscale x 8 x double> @intrinsic_vfwsub.w_mask_wv_tie_nxv8f64_nxv8f64_nxv8f32(<vscale x 8 x double> %0, <vscale x 8 x float> %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wv_tie_nxv8f64_nxv8f64_nxv8f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vfwsub.wv v8, v8, v16, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.nxv8f32(
<vscale x 8 x double> %0,
<vscale x 8 x double> %0,
<vscale x 8 x float> %1,
<vscale x 8 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 8 x double> %a
}
define <vscale x 1 x float> @intrinsic_vfwsub.w_mask_wf_tie_nxv1f32_nxv1f32_f16(<vscale x 1 x float> %0, half %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_tie_nxv1f32_nxv1f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v8, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x float> @llvm.riscv.vfwsub.w.mask.nxv1f32.f16(
<vscale x 1 x float> %0,
<vscale x 1 x float> %0,
half %1,
<vscale x 1 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 1 x float> %a
}
define <vscale x 2 x float> @intrinsic_vfwsub.w_mask_wf_tie_nxv2f32_nxv2f32_f16(<vscale x 2 x float> %0, half %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_tie_nxv2f32_nxv2f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v8, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x float> @llvm.riscv.vfwsub.w.mask.nxv2f32.f16(
<vscale x 2 x float> %0,
<vscale x 2 x float> %0,
half %1,
<vscale x 2 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 2 x float> %a
}
define <vscale x 4 x float> @intrinsic_vfwsub.w_mask_wf_tie_nxv4f32_nxv4f32_f16(<vscale x 4 x float> %0, half %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_tie_nxv4f32_nxv4f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v8, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x float> @llvm.riscv.vfwsub.w.mask.nxv4f32.f16(
<vscale x 4 x float> %0,
<vscale x 4 x float> %0,
half %1,
<vscale x 4 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 4 x float> %a
}
define <vscale x 8 x float> @intrinsic_vfwsub.w_mask_wf_tie_nxv8f32_nxv8f32_f16(<vscale x 8 x float> %0, half %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_tie_nxv8f32_nxv8f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v8, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x float> @llvm.riscv.vfwsub.w.mask.nxv8f32.f16(
<vscale x 8 x float> %0,
<vscale x 8 x float> %0,
half %1,
<vscale x 8 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 8 x float> %a
}
define <vscale x 16 x float> @intrinsic_vfwsub.w_mask_wf_tie_nxv16f32_nxv16f32_f16(<vscale x 16 x float> %0, half %1, <vscale x 16 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_tie_nxv16f32_nxv16f32_f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v8, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 16 x float> @llvm.riscv.vfwsub.w.mask.nxv16f32.f16(
<vscale x 16 x float> %0,
<vscale x 16 x float> %0,
half %1,
<vscale x 16 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 16 x float> %a
}
define <vscale x 1 x double> @intrinsic_vfwsub.w_mask_wf_tie_nxv1f64_nxv1f64_f32(<vscale x 1 x double> %0, float %1, <vscale x 1 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_tie_nxv1f64_nxv1f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v8, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.mask.nxv1f64.f32(
<vscale x 1 x double> %0,
<vscale x 1 x double> %0,
float %1,
<vscale x 1 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 1 x double> %a
}
define <vscale x 2 x double> @intrinsic_vfwsub.w_mask_wf_tie_nxv2f64_nxv2f64_f32(<vscale x 2 x double> %0, float %1, <vscale x 2 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_tie_nxv2f64_nxv2f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v8, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.mask.nxv2f64.f32(
<vscale x 2 x double> %0,
<vscale x 2 x double> %0,
float %1,
<vscale x 2 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 2 x double> %a
}
define <vscale x 4 x double> @intrinsic_vfwsub.w_mask_wf_tie_nxv4f64_nxv4f64_f32(<vscale x 4 x double> %0, float %1, <vscale x 4 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_tie_nxv4f64_nxv4f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v8, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.mask.nxv4f64.f32(
<vscale x 4 x double> %0,
<vscale x 4 x double> %0,
float %1,
<vscale x 4 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 4 x double> %a
}
define <vscale x 8 x double> @intrinsic_vfwsub.w_mask_wf_tie_nxv8f64_nxv8f64_f32(<vscale x 8 x double> %0, float %1, <vscale x 8 x i1> %2, iXLen %3) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_mask_wf_tie_nxv8f64_nxv8f64_f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, mu
; CHECK-NEXT: vfwsub.wf v8, v8, fa0, v0.t
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.mask.nxv8f64.f32(
<vscale x 8 x double> %0,
<vscale x 8 x double> %0,
float %1,
<vscale x 8 x i1> %2,
iXLen 0, iXLen %3, iXLen 1)
ret <vscale x 8 x double> %a
}
define <vscale x 1 x float> @intrinsic_vfwsub.w_wv_untie_nxv1f32_nxv1f32_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x float> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_untie_nxv1f32_nxv1f32_nxv1f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
; CHECK-NEXT: vfwsub.wv v10, v9, v8
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x float> @llvm.riscv.vfwsub.w.nxv1f32.nxv1f16(
<vscale x 1 x float> undef,
<vscale x 1 x float> %1,
<vscale x 1 x half> %0,
iXLen 0, iXLen %2)
ret <vscale x 1 x float> %a
}
define <vscale x 2 x float> @intrinsic_vfwsub.w_wv_untie_nxv2f32_nxv2f32_nxv2f16(<vscale x 2 x half> %0, <vscale x 2 x float> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_untie_nxv2f32_nxv2f32_nxv2f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
; CHECK-NEXT: vfwsub.wv v10, v9, v8
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x float> @llvm.riscv.vfwsub.w.nxv2f32.nxv2f16(
<vscale x 2 x float> undef,
<vscale x 2 x float> %1,
<vscale x 2 x half> %0,
iXLen 0, iXLen %2)
ret <vscale x 2 x float> %a
}
define <vscale x 4 x float> @intrinsic_vfwsub.w_wv_untie_nxv4f32_nxv4f32_nxv4f16(<vscale x 4 x half> %0, <vscale x 4 x float> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_untie_nxv4f32_nxv4f32_nxv4f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
; CHECK-NEXT: vfwsub.wv v12, v10, v8
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: vmv2r.v v8, v12
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x float> @llvm.riscv.vfwsub.w.nxv4f32.nxv4f16(
<vscale x 4 x float> undef,
<vscale x 4 x float> %1,
<vscale x 4 x half> %0,
iXLen 0, iXLen %2)
ret <vscale x 4 x float> %a
}
define <vscale x 8 x float> @intrinsic_vfwsub.w_wv_untie_nxv8f32_nxv8f32_nxv8f16(<vscale x 8 x half> %0, <vscale x 8 x float> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_untie_nxv8f32_nxv8f32_nxv8f16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
; CHECK-NEXT: vfwsub.wv v16, v12, v8
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: vmv4r.v v8, v16
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x float> @llvm.riscv.vfwsub.w.nxv8f32.nxv8f16(
<vscale x 8 x float> undef,
<vscale x 8 x float> %1,
<vscale x 8 x half> %0,
iXLen 0, iXLen %2)
ret <vscale x 8 x float> %a
}
define <vscale x 1 x double> @intrinsic_vfwsub.w_wv_untie_nxv1f64_nxv1f64_nxv1f32(<vscale x 1 x float> %0, <vscale x 1 x double> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_untie_nxv1f64_nxv1f64_nxv1f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
; CHECK-NEXT: vfwsub.wv v10, v9, v8
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: vmv1r.v v8, v10
; CHECK-NEXT: ret
entry:
%a = call <vscale x 1 x double> @llvm.riscv.vfwsub.w.nxv1f64.nxv1f32(
<vscale x 1 x double> undef,
<vscale x 1 x double> %1,
<vscale x 1 x float> %0,
iXLen 0, iXLen %2)
ret <vscale x 1 x double> %a
}
define <vscale x 2 x double> @intrinsic_vfwsub.w_wv_untie_nxv2f64_nxv2f64_nxv2f32(<vscale x 2 x float> %0, <vscale x 2 x double> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_untie_nxv2f64_nxv2f64_nxv2f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
; CHECK-NEXT: vfwsub.wv v12, v10, v8
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: vmv2r.v v8, v12
; CHECK-NEXT: ret
entry:
%a = call <vscale x 2 x double> @llvm.riscv.vfwsub.w.nxv2f64.nxv2f32(
<vscale x 2 x double> undef,
<vscale x 2 x double> %1,
<vscale x 2 x float> %0,
iXLen 0, iXLen %2)
ret <vscale x 2 x double> %a
}
define <vscale x 4 x double> @intrinsic_vfwsub.w_wv_untie_nxv4f64_nxv4f64_nxv4f32(<vscale x 4 x float> %0, <vscale x 4 x double> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_untie_nxv4f64_nxv4f64_nxv4f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
; CHECK-NEXT: vfwsub.wv v16, v12, v8
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: vmv4r.v v8, v16
; CHECK-NEXT: ret
entry:
%a = call <vscale x 4 x double> @llvm.riscv.vfwsub.w.nxv4f64.nxv4f32(
<vscale x 4 x double> undef,
<vscale x 4 x double> %1,
<vscale x 4 x float> %0,
iXLen 0, iXLen %2)
ret <vscale x 4 x double> %a
}
define <vscale x 8 x double> @intrinsic_vfwsub.w_wv_untie_nxv8f64_nxv8f64_nxv8f32(<vscale x 8 x float> %0, <vscale x 8 x double> %1, iXLen %2) nounwind {
; CHECK-LABEL: intrinsic_vfwsub.w_wv_untie_nxv8f64_nxv8f64_nxv8f32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: fsrmi a1, 0
; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
; CHECK-NEXT: vfwsub.wv v24, v16, v8
; CHECK-NEXT: fsrm a1
; CHECK-NEXT: vmv8r.v v8, v24
; CHECK-NEXT: ret
entry:
%a = call <vscale x 8 x double> @llvm.riscv.vfwsub.w.nxv8f64.nxv8f32(
<vscale x 8 x double> undef,
<vscale x 8 x double> %1,
<vscale x 8 x float> %0,
iXLen 0, iXLen %2)
ret <vscale x 8 x double> %a
}
Codebase Browser
llvm