llvm-for-llvmta/test/CodeGen/RISCV/rvv/vzext-rv64.ll

1151 lines
34 KiB
LLVM
Raw Permalink Normal View History

2022-04-25 10:02:23 +02:00
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=riscv64 -mattr=+experimental-v -verify-machineinstrs \
; RUN: --riscv-no-aliases < %s | FileCheck %s
declare <vscale x 1 x i64> @llvm.riscv.vzext.nxv1i64.nxv1i8(
<vscale x 1 x i8>,
i64);
define <vscale x 1 x i64> @intrinsic_vzext_vf8_nxv1i64(<vscale x 1 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf8_nxv1i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m1,ta,mu
; CHECK-NEXT: vzext.vf8 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i64> @llvm.riscv.vzext.nxv1i64.nxv1i8(
<vscale x 1 x i8> %0,
i64 %1)
ret <vscale x 1 x i64> %a
}
declare <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i8(
<vscale x 1 x i64>,
<vscale x 1 x i8>,
<vscale x 1 x i1>,
i64);
define <vscale x 1 x i64> @intrinsic_vzext_mask_vf8_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf8_nxv1i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m1,tu,mu
; CHECK-NEXT: vzext.vf8 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i8(
<vscale x 1 x i64> %1,
<vscale x 1 x i8> %2,
<vscale x 1 x i1> %0,
i64 %3)
ret <vscale x 1 x i64> %a
}
declare <vscale x 2 x i64> @llvm.riscv.vzext.nxv2i64.nxv2i8(
<vscale x 2 x i8>,
i64);
define <vscale x 2 x i64> @intrinsic_vzext_vf8_nxv2i64(<vscale x 2 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf8_nxv2i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m2,ta,mu
; CHECK-NEXT: vzext.vf8 v26, v8
; CHECK-NEXT: vmv2r.v v8, v26
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i64> @llvm.riscv.vzext.nxv2i64.nxv2i8(
<vscale x 2 x i8> %0,
i64 %1)
ret <vscale x 2 x i64> %a
}
declare <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i8(
<vscale x 2 x i64>,
<vscale x 2 x i8>,
<vscale x 2 x i1>,
i64);
define <vscale x 2 x i64> @intrinsic_vzext_mask_vf8_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf8_nxv2i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m2,tu,mu
; CHECK-NEXT: vzext.vf8 v8, v10, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i8(
<vscale x 2 x i64> %1,
<vscale x 2 x i8> %2,
<vscale x 2 x i1> %0,
i64 %3)
ret <vscale x 2 x i64> %a
}
declare <vscale x 4 x i64> @llvm.riscv.vzext.nxv4i64.nxv4i8(
<vscale x 4 x i8>,
i64);
define <vscale x 4 x i64> @intrinsic_vzext_vf8_nxv4i64(<vscale x 4 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf8_nxv4i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m4,ta,mu
; CHECK-NEXT: vzext.vf8 v28, v8
; CHECK-NEXT: vmv4r.v v8, v28
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i64> @llvm.riscv.vzext.nxv4i64.nxv4i8(
<vscale x 4 x i8> %0,
i64 %1)
ret <vscale x 4 x i64> %a
}
declare <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i8(
<vscale x 4 x i64>,
<vscale x 4 x i8>,
<vscale x 4 x i1>,
i64);
define <vscale x 4 x i64> @intrinsic_vzext_mask_vf8_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf8_nxv4i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m4,tu,mu
; CHECK-NEXT: vzext.vf8 v8, v12, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i8(
<vscale x 4 x i64> %1,
<vscale x 4 x i8> %2,
<vscale x 4 x i1> %0,
i64 %3)
ret <vscale x 4 x i64> %a
}
declare <vscale x 8 x i64> @llvm.riscv.vzext.nxv8i64.nxv8i8(
<vscale x 8 x i8>,
i64);
define <vscale x 8 x i64> @intrinsic_vzext_vf8_nxv8i64(<vscale x 8 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf8_nxv8i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m8,ta,mu
; CHECK-NEXT: vzext.vf8 v16, v8
; CHECK-NEXT: vmv8r.v v8, v16
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i64> @llvm.riscv.vzext.nxv8i64.nxv8i8(
<vscale x 8 x i8> %0,
i64 %1)
ret <vscale x 8 x i64> %a
}
declare <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i8(
<vscale x 8 x i64>,
<vscale x 8 x i8>,
<vscale x 8 x i1>,
i64);
define <vscale x 8 x i64> @intrinsic_vzext_mask_vf8_nxv8i64(<vscale x 8 x i1> %0, <vscale x 8 x i64> %1, <vscale x 8 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf8_nxv8i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m8,tu,mu
; CHECK-NEXT: vzext.vf8 v8, v16, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i8(
<vscale x 8 x i64> %1,
<vscale x 8 x i8> %2,
<vscale x 8 x i1> %0,
i64 %3)
ret <vscale x 8 x i64> %a
}
declare <vscale x 1 x i64> @llvm.riscv.vzext.nxv1i64.nxv1i16(
<vscale x 1 x i16>,
i64);
define <vscale x 1 x i64> @intrinsic_vzext_vf4_nxv1i64(<vscale x 1 x i16> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf4_nxv1i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m1,ta,mu
; CHECK-NEXT: vzext.vf4 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i64> @llvm.riscv.vzext.nxv1i64.nxv1i16(
<vscale x 1 x i16> %0,
i64 %1)
ret <vscale x 1 x i64> %a
}
declare <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i16(
<vscale x 1 x i64>,
<vscale x 1 x i16>,
<vscale x 1 x i1>,
i64);
define <vscale x 1 x i64> @intrinsic_vzext_mask_vf4_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i16> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf4_nxv1i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m1,tu,mu
; CHECK-NEXT: vzext.vf4 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i16(
<vscale x 1 x i64> %1,
<vscale x 1 x i16> %2,
<vscale x 1 x i1> %0,
i64 %3)
ret <vscale x 1 x i64> %a
}
declare <vscale x 2 x i64> @llvm.riscv.vzext.nxv2i64.nxv2i16(
<vscale x 2 x i16>,
i64);
define <vscale x 2 x i64> @intrinsic_vzext_vf4_nxv2i64(<vscale x 2 x i16> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf4_nxv2i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m2,ta,mu
; CHECK-NEXT: vzext.vf4 v26, v8
; CHECK-NEXT: vmv2r.v v8, v26
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i64> @llvm.riscv.vzext.nxv2i64.nxv2i16(
<vscale x 2 x i16> %0,
i64 %1)
ret <vscale x 2 x i64> %a
}
declare <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i16(
<vscale x 2 x i64>,
<vscale x 2 x i16>,
<vscale x 2 x i1>,
i64);
define <vscale x 2 x i64> @intrinsic_vzext_mask_vf4_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i16> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf4_nxv2i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m2,tu,mu
; CHECK-NEXT: vzext.vf4 v8, v10, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i16(
<vscale x 2 x i64> %1,
<vscale x 2 x i16> %2,
<vscale x 2 x i1> %0,
i64 %3)
ret <vscale x 2 x i64> %a
}
declare <vscale x 4 x i64> @llvm.riscv.vzext.nxv4i64.nxv4i16(
<vscale x 4 x i16>,
i64);
define <vscale x 4 x i64> @intrinsic_vzext_vf4_nxv4i64(<vscale x 4 x i16> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf4_nxv4i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m4,ta,mu
; CHECK-NEXT: vzext.vf4 v28, v8
; CHECK-NEXT: vmv4r.v v8, v28
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i64> @llvm.riscv.vzext.nxv4i64.nxv4i16(
<vscale x 4 x i16> %0,
i64 %1)
ret <vscale x 4 x i64> %a
}
declare <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i16(
<vscale x 4 x i64>,
<vscale x 4 x i16>,
<vscale x 4 x i1>,
i64);
define <vscale x 4 x i64> @intrinsic_vzext_mask_vf4_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i16> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf4_nxv4i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m4,tu,mu
; CHECK-NEXT: vzext.vf4 v8, v12, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i16(
<vscale x 4 x i64> %1,
<vscale x 4 x i16> %2,
<vscale x 4 x i1> %0,
i64 %3)
ret <vscale x 4 x i64> %a
}
declare <vscale x 8 x i64> @llvm.riscv.vzext.nxv8i64.nxv8i16(
<vscale x 8 x i16>,
i64);
define <vscale x 8 x i64> @intrinsic_vzext_vf4_nxv8i64(<vscale x 8 x i16> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf4_nxv8i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m8,ta,mu
; CHECK-NEXT: vzext.vf4 v16, v8
; CHECK-NEXT: vmv8r.v v8, v16
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i64> @llvm.riscv.vzext.nxv8i64.nxv8i16(
<vscale x 8 x i16> %0,
i64 %1)
ret <vscale x 8 x i64> %a
}
declare <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i16(
<vscale x 8 x i64>,
<vscale x 8 x i16>,
<vscale x 8 x i1>,
i64);
define <vscale x 8 x i64> @intrinsic_vzext_mask_vf4_nxv8i64(<vscale x 8 x i1> %0, <vscale x 8 x i64> %1, <vscale x 8 x i16> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf4_nxv8i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m8,tu,mu
; CHECK-NEXT: vzext.vf4 v8, v16, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i16(
<vscale x 8 x i64> %1,
<vscale x 8 x i16> %2,
<vscale x 8 x i1> %0,
i64 %3)
ret <vscale x 8 x i64> %a
}
declare <vscale x 1 x i32> @llvm.riscv.vzext.nxv1i32.nxv1i8(
<vscale x 1 x i8>,
i64);
define <vscale x 1 x i32> @intrinsic_vzext_vf4_nxv1i32(<vscale x 1 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf4_nxv1i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,mf2,ta,mu
; CHECK-NEXT: vzext.vf4 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i32> @llvm.riscv.vzext.nxv1i32.nxv1i8(
<vscale x 1 x i8> %0,
i64 %1)
ret <vscale x 1 x i32> %a
}
declare <vscale x 1 x i32> @llvm.riscv.vzext.mask.nxv1i32.nxv1i8(
<vscale x 1 x i32>,
<vscale x 1 x i8>,
<vscale x 1 x i1>,
i64);
define <vscale x 1 x i32> @intrinsic_vzext_mask_vf4_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf4_nxv1i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,mf2,tu,mu
; CHECK-NEXT: vzext.vf4 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i32> @llvm.riscv.vzext.mask.nxv1i32.nxv1i8(
<vscale x 1 x i32> %1,
<vscale x 1 x i8> %2,
<vscale x 1 x i1> %0,
i64 %3)
ret <vscale x 1 x i32> %a
}
declare <vscale x 2 x i32> @llvm.riscv.vzext.nxv2i32.nxv2i8(
<vscale x 2 x i8>,
i64);
define <vscale x 2 x i32> @intrinsic_vzext_vf4_nxv2i32(<vscale x 2 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf4_nxv2i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m1,ta,mu
; CHECK-NEXT: vzext.vf4 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i32> @llvm.riscv.vzext.nxv2i32.nxv2i8(
<vscale x 2 x i8> %0,
i64 %1)
ret <vscale x 2 x i32> %a
}
declare <vscale x 2 x i32> @llvm.riscv.vzext.mask.nxv2i32.nxv2i8(
<vscale x 2 x i32>,
<vscale x 2 x i8>,
<vscale x 2 x i1>,
i64);
define <vscale x 2 x i32> @intrinsic_vzext_mask_vf4_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf4_nxv2i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m1,tu,mu
; CHECK-NEXT: vzext.vf4 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i32> @llvm.riscv.vzext.mask.nxv2i32.nxv2i8(
<vscale x 2 x i32> %1,
<vscale x 2 x i8> %2,
<vscale x 2 x i1> %0,
i64 %3)
ret <vscale x 2 x i32> %a
}
declare <vscale x 4 x i32> @llvm.riscv.vzext.nxv4i32.nxv4i8(
<vscale x 4 x i8>,
i64);
define <vscale x 4 x i32> @intrinsic_vzext_vf4_nxv4i32(<vscale x 4 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf4_nxv4i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m2,ta,mu
; CHECK-NEXT: vzext.vf4 v26, v8
; CHECK-NEXT: vmv2r.v v8, v26
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i32> @llvm.riscv.vzext.nxv4i32.nxv4i8(
<vscale x 4 x i8> %0,
i64 %1)
ret <vscale x 4 x i32> %a
}
declare <vscale x 4 x i32> @llvm.riscv.vzext.mask.nxv4i32.nxv4i8(
<vscale x 4 x i32>,
<vscale x 4 x i8>,
<vscale x 4 x i1>,
i64);
define <vscale x 4 x i32> @intrinsic_vzext_mask_vf4_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf4_nxv4i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m2,tu,mu
; CHECK-NEXT: vzext.vf4 v8, v10, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i32> @llvm.riscv.vzext.mask.nxv4i32.nxv4i8(
<vscale x 4 x i32> %1,
<vscale x 4 x i8> %2,
<vscale x 4 x i1> %0,
i64 %3)
ret <vscale x 4 x i32> %a
}
declare <vscale x 8 x i32> @llvm.riscv.vzext.nxv8i32.nxv8i8(
<vscale x 8 x i8>,
i64);
define <vscale x 8 x i32> @intrinsic_vzext_vf4_nxv8i32(<vscale x 8 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf4_nxv8i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m4,ta,mu
; CHECK-NEXT: vzext.vf4 v28, v8
; CHECK-NEXT: vmv4r.v v8, v28
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i32> @llvm.riscv.vzext.nxv8i32.nxv8i8(
<vscale x 8 x i8> %0,
i64 %1)
ret <vscale x 8 x i32> %a
}
declare <vscale x 8 x i32> @llvm.riscv.vzext.mask.nxv8i32.nxv8i8(
<vscale x 8 x i32>,
<vscale x 8 x i8>,
<vscale x 8 x i1>,
i64);
define <vscale x 8 x i32> @intrinsic_vzext_mask_vf4_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf4_nxv8i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m4,tu,mu
; CHECK-NEXT: vzext.vf4 v8, v12, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i32> @llvm.riscv.vzext.mask.nxv8i32.nxv8i8(
<vscale x 8 x i32> %1,
<vscale x 8 x i8> %2,
<vscale x 8 x i1> %0,
i64 %3)
ret <vscale x 8 x i32> %a
}
declare <vscale x 16 x i32> @llvm.riscv.vzext.nxv16i32.nxv16i8(
<vscale x 16 x i8>,
i64);
define <vscale x 16 x i32> @intrinsic_vzext_vf4_nxv16i32(<vscale x 16 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf4_nxv16i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m8,ta,mu
; CHECK-NEXT: vzext.vf4 v16, v8
; CHECK-NEXT: vmv8r.v v8, v16
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 16 x i32> @llvm.riscv.vzext.nxv16i32.nxv16i8(
<vscale x 16 x i8> %0,
i64 %1)
ret <vscale x 16 x i32> %a
}
declare <vscale x 16 x i32> @llvm.riscv.vzext.mask.nxv16i32.nxv16i8(
<vscale x 16 x i32>,
<vscale x 16 x i8>,
<vscale x 16 x i1>,
i64);
define <vscale x 16 x i32> @intrinsic_vzext_mask_vf4_nxv16i32(<vscale x 16 x i1> %0, <vscale x 16 x i32> %1, <vscale x 16 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf4_nxv16i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m8,tu,mu
; CHECK-NEXT: vzext.vf4 v8, v16, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 16 x i32> @llvm.riscv.vzext.mask.nxv16i32.nxv16i8(
<vscale x 16 x i32> %1,
<vscale x 16 x i8> %2,
<vscale x 16 x i1> %0,
i64 %3)
ret <vscale x 16 x i32> %a
}
declare <vscale x 1 x i64> @llvm.riscv.vzext.nxv1i64.nxv1i32(
<vscale x 1 x i32>,
i64);
define <vscale x 1 x i64> @intrinsic_vzext_vf2_nxv1i64(<vscale x 1 x i32> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv1i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m1,ta,mu
; CHECK-NEXT: vzext.vf2 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i64> @llvm.riscv.vzext.nxv1i64.nxv1i32(
<vscale x 1 x i32> %0,
i64 %1)
ret <vscale x 1 x i64> %a
}
declare <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i32(
<vscale x 1 x i64>,
<vscale x 1 x i32>,
<vscale x 1 x i1>,
i64);
define <vscale x 1 x i64> @intrinsic_vzext_mask_vf2_nxv1i64(<vscale x 1 x i1> %0, <vscale x 1 x i64> %1, <vscale x 1 x i32> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv1i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m1,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i64> @llvm.riscv.vzext.mask.nxv1i64.nxv1i32(
<vscale x 1 x i64> %1,
<vscale x 1 x i32> %2,
<vscale x 1 x i1> %0,
i64 %3)
ret <vscale x 1 x i64> %a
}
declare <vscale x 2 x i64> @llvm.riscv.vzext.nxv2i64.nxv2i32(
<vscale x 2 x i32>,
i64);
define <vscale x 2 x i64> @intrinsic_vzext_vf2_nxv2i64(<vscale x 2 x i32> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv2i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m2,ta,mu
; CHECK-NEXT: vzext.vf2 v26, v8
; CHECK-NEXT: vmv2r.v v8, v26
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i64> @llvm.riscv.vzext.nxv2i64.nxv2i32(
<vscale x 2 x i32> %0,
i64 %1)
ret <vscale x 2 x i64> %a
}
declare <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i32(
<vscale x 2 x i64>,
<vscale x 2 x i32>,
<vscale x 2 x i1>,
i64);
define <vscale x 2 x i64> @intrinsic_vzext_mask_vf2_nxv2i64(<vscale x 2 x i1> %0, <vscale x 2 x i64> %1, <vscale x 2 x i32> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv2i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m2,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v10, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i64> @llvm.riscv.vzext.mask.nxv2i64.nxv2i32(
<vscale x 2 x i64> %1,
<vscale x 2 x i32> %2,
<vscale x 2 x i1> %0,
i64 %3)
ret <vscale x 2 x i64> %a
}
declare <vscale x 4 x i64> @llvm.riscv.vzext.nxv4i64.nxv4i32(
<vscale x 4 x i32>,
i64);
define <vscale x 4 x i64> @intrinsic_vzext_vf2_nxv4i64(<vscale x 4 x i32> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv4i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m4,ta,mu
; CHECK-NEXT: vzext.vf2 v28, v8
; CHECK-NEXT: vmv4r.v v8, v28
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i64> @llvm.riscv.vzext.nxv4i64.nxv4i32(
<vscale x 4 x i32> %0,
i64 %1)
ret <vscale x 4 x i64> %a
}
declare <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i32(
<vscale x 4 x i64>,
<vscale x 4 x i32>,
<vscale x 4 x i1>,
i64);
define <vscale x 4 x i64> @intrinsic_vzext_mask_vf2_nxv4i64(<vscale x 4 x i1> %0, <vscale x 4 x i64> %1, <vscale x 4 x i32> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv4i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m4,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v12, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i64> @llvm.riscv.vzext.mask.nxv4i64.nxv4i32(
<vscale x 4 x i64> %1,
<vscale x 4 x i32> %2,
<vscale x 4 x i1> %0,
i64 %3)
ret <vscale x 4 x i64> %a
}
declare <vscale x 8 x i64> @llvm.riscv.vzext.nxv8i64.nxv8i32(
<vscale x 8 x i32>,
i64);
define <vscale x 8 x i64> @intrinsic_vzext_vf2_nxv8i64(<vscale x 8 x i32> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv8i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m8,ta,mu
; CHECK-NEXT: vzext.vf2 v16, v8
; CHECK-NEXT: vmv8r.v v8, v16
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i64> @llvm.riscv.vzext.nxv8i64.nxv8i32(
<vscale x 8 x i32> %0,
i64 %1)
ret <vscale x 8 x i64> %a
}
declare <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i32(
<vscale x 8 x i64>,
<vscale x 8 x i32>,
<vscale x 8 x i1>,
i64);
define <vscale x 8 x i64> @intrinsic_vzext_mask_vf2_nxv8i64(<vscale x 8 x i1> %0, <vscale x 8 x i64> %1, <vscale x 8 x i32> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv8i64:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e64,m8,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v16, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i64> @llvm.riscv.vzext.mask.nxv8i64.nxv8i32(
<vscale x 8 x i64> %1,
<vscale x 8 x i32> %2,
<vscale x 8 x i1> %0,
i64 %3)
ret <vscale x 8 x i64> %a
}
declare <vscale x 1 x i32> @llvm.riscv.vzext.nxv1i32.nxv1i16(
<vscale x 1 x i16>,
i64);
define <vscale x 1 x i32> @intrinsic_vzext_vf2_nxv1i32(<vscale x 1 x i16> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv1i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,mf2,ta,mu
; CHECK-NEXT: vzext.vf2 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i32> @llvm.riscv.vzext.nxv1i32.nxv1i16(
<vscale x 1 x i16> %0,
i64 %1)
ret <vscale x 1 x i32> %a
}
declare <vscale x 1 x i32> @llvm.riscv.vzext.mask.nxv1i32.nxv1i16(
<vscale x 1 x i32>,
<vscale x 1 x i16>,
<vscale x 1 x i1>,
i64);
define <vscale x 1 x i32> @intrinsic_vzext_mask_vf2_nxv1i32(<vscale x 1 x i1> %0, <vscale x 1 x i32> %1, <vscale x 1 x i16> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv1i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,mf2,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i32> @llvm.riscv.vzext.mask.nxv1i32.nxv1i16(
<vscale x 1 x i32> %1,
<vscale x 1 x i16> %2,
<vscale x 1 x i1> %0,
i64 %3)
ret <vscale x 1 x i32> %a
}
declare <vscale x 2 x i32> @llvm.riscv.vzext.nxv2i32.nxv2i16(
<vscale x 2 x i16>,
i64);
define <vscale x 2 x i32> @intrinsic_vzext_vf2_nxv2i32(<vscale x 2 x i16> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv2i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m1,ta,mu
; CHECK-NEXT: vzext.vf2 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i32> @llvm.riscv.vzext.nxv2i32.nxv2i16(
<vscale x 2 x i16> %0,
i64 %1)
ret <vscale x 2 x i32> %a
}
declare <vscale x 2 x i32> @llvm.riscv.vzext.mask.nxv2i32.nxv2i16(
<vscale x 2 x i32>,
<vscale x 2 x i16>,
<vscale x 2 x i1>,
i64);
define <vscale x 2 x i32> @intrinsic_vzext_mask_vf2_nxv2i32(<vscale x 2 x i1> %0, <vscale x 2 x i32> %1, <vscale x 2 x i16> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv2i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m1,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i32> @llvm.riscv.vzext.mask.nxv2i32.nxv2i16(
<vscale x 2 x i32> %1,
<vscale x 2 x i16> %2,
<vscale x 2 x i1> %0,
i64 %3)
ret <vscale x 2 x i32> %a
}
declare <vscale x 4 x i32> @llvm.riscv.vzext.nxv4i32.nxv4i16(
<vscale x 4 x i16>,
i64);
define <vscale x 4 x i32> @intrinsic_vzext_vf2_nxv4i32(<vscale x 4 x i16> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv4i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m2,ta,mu
; CHECK-NEXT: vzext.vf2 v26, v8
; CHECK-NEXT: vmv2r.v v8, v26
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i32> @llvm.riscv.vzext.nxv4i32.nxv4i16(
<vscale x 4 x i16> %0,
i64 %1)
ret <vscale x 4 x i32> %a
}
declare <vscale x 4 x i32> @llvm.riscv.vzext.mask.nxv4i32.nxv4i16(
<vscale x 4 x i32>,
<vscale x 4 x i16>,
<vscale x 4 x i1>,
i64);
define <vscale x 4 x i32> @intrinsic_vzext_mask_vf2_nxv4i32(<vscale x 4 x i1> %0, <vscale x 4 x i32> %1, <vscale x 4 x i16> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv4i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m2,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v10, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i32> @llvm.riscv.vzext.mask.nxv4i32.nxv4i16(
<vscale x 4 x i32> %1,
<vscale x 4 x i16> %2,
<vscale x 4 x i1> %0,
i64 %3)
ret <vscale x 4 x i32> %a
}
declare <vscale x 8 x i32> @llvm.riscv.vzext.nxv8i32.nxv8i16(
<vscale x 8 x i16>,
i64);
define <vscale x 8 x i32> @intrinsic_vzext_vf2_nxv8i32(<vscale x 8 x i16> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv8i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m4,ta,mu
; CHECK-NEXT: vzext.vf2 v28, v8
; CHECK-NEXT: vmv4r.v v8, v28
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i32> @llvm.riscv.vzext.nxv8i32.nxv8i16(
<vscale x 8 x i16> %0,
i64 %1)
ret <vscale x 8 x i32> %a
}
declare <vscale x 8 x i32> @llvm.riscv.vzext.mask.nxv8i32.nxv8i16(
<vscale x 8 x i32>,
<vscale x 8 x i16>,
<vscale x 8 x i1>,
i64);
define <vscale x 8 x i32> @intrinsic_vzext_mask_vf2_nxv8i32(<vscale x 8 x i1> %0, <vscale x 8 x i32> %1, <vscale x 8 x i16> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv8i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m4,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v12, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i32> @llvm.riscv.vzext.mask.nxv8i32.nxv8i16(
<vscale x 8 x i32> %1,
<vscale x 8 x i16> %2,
<vscale x 8 x i1> %0,
i64 %3)
ret <vscale x 8 x i32> %a
}
declare <vscale x 16 x i32> @llvm.riscv.vzext.nxv16i32.nxv16i16(
<vscale x 16 x i16>,
i64);
define <vscale x 16 x i32> @intrinsic_vzext_vf2_nxv16i32(<vscale x 16 x i16> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv16i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m8,ta,mu
; CHECK-NEXT: vzext.vf2 v16, v8
; CHECK-NEXT: vmv8r.v v8, v16
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 16 x i32> @llvm.riscv.vzext.nxv16i32.nxv16i16(
<vscale x 16 x i16> %0,
i64 %1)
ret <vscale x 16 x i32> %a
}
declare <vscale x 16 x i32> @llvm.riscv.vzext.mask.nxv16i32.nxv16i16(
<vscale x 16 x i32>,
<vscale x 16 x i16>,
<vscale x 16 x i1>,
i64);
define <vscale x 16 x i32> @intrinsic_vzext_mask_vf2_nxv16i32(<vscale x 16 x i1> %0, <vscale x 16 x i32> %1, <vscale x 16 x i16> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv16i32:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e32,m8,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v16, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 16 x i32> @llvm.riscv.vzext.mask.nxv16i32.nxv16i16(
<vscale x 16 x i32> %1,
<vscale x 16 x i16> %2,
<vscale x 16 x i1> %0,
i64 %3)
ret <vscale x 16 x i32> %a
}
declare <vscale x 1 x i16> @llvm.riscv.vzext.nxv1i16.nxv1i8(
<vscale x 1 x i8>,
i64);
define <vscale x 1 x i16> @intrinsic_vzext_vf2_nxv1i16(<vscale x 1 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv1i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,mf4,ta,mu
; CHECK-NEXT: vzext.vf2 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i16> @llvm.riscv.vzext.nxv1i16.nxv1i8(
<vscale x 1 x i8> %0,
i64 %1)
ret <vscale x 1 x i16> %a
}
declare <vscale x 1 x i16> @llvm.riscv.vzext.mask.nxv1i16.nxv1i8(
<vscale x 1 x i16>,
<vscale x 1 x i8>,
<vscale x 1 x i1>,
i64);
define <vscale x 1 x i16> @intrinsic_vzext_mask_vf2_nxv1i16(<vscale x 1 x i1> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv1i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,mf4,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 1 x i16> @llvm.riscv.vzext.mask.nxv1i16.nxv1i8(
<vscale x 1 x i16> %1,
<vscale x 1 x i8> %2,
<vscale x 1 x i1> %0,
i64 %3)
ret <vscale x 1 x i16> %a
}
declare <vscale x 2 x i16> @llvm.riscv.vzext.nxv2i16.nxv2i8(
<vscale x 2 x i8>,
i64);
define <vscale x 2 x i16> @intrinsic_vzext_vf2_nxv2i16(<vscale x 2 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv2i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,mf2,ta,mu
; CHECK-NEXT: vzext.vf2 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i16> @llvm.riscv.vzext.nxv2i16.nxv2i8(
<vscale x 2 x i8> %0,
i64 %1)
ret <vscale x 2 x i16> %a
}
declare <vscale x 2 x i16> @llvm.riscv.vzext.mask.nxv2i16.nxv2i8(
<vscale x 2 x i16>,
<vscale x 2 x i8>,
<vscale x 2 x i1>,
i64);
define <vscale x 2 x i16> @intrinsic_vzext_mask_vf2_nxv2i16(<vscale x 2 x i1> %0, <vscale x 2 x i16> %1, <vscale x 2 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv2i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,mf2,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 2 x i16> @llvm.riscv.vzext.mask.nxv2i16.nxv2i8(
<vscale x 2 x i16> %1,
<vscale x 2 x i8> %2,
<vscale x 2 x i1> %0,
i64 %3)
ret <vscale x 2 x i16> %a
}
declare <vscale x 4 x i16> @llvm.riscv.vzext.nxv4i16.nxv4i8(
<vscale x 4 x i8>,
i64);
define <vscale x 4 x i16> @intrinsic_vzext_vf2_nxv4i16(<vscale x 4 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv4i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,m1,ta,mu
; CHECK-NEXT: vzext.vf2 v25, v8
; CHECK-NEXT: vmv1r.v v8, v25
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i16> @llvm.riscv.vzext.nxv4i16.nxv4i8(
<vscale x 4 x i8> %0,
i64 %1)
ret <vscale x 4 x i16> %a
}
declare <vscale x 4 x i16> @llvm.riscv.vzext.mask.nxv4i16.nxv4i8(
<vscale x 4 x i16>,
<vscale x 4 x i8>,
<vscale x 4 x i1>,
i64);
define <vscale x 4 x i16> @intrinsic_vzext_mask_vf2_nxv4i16(<vscale x 4 x i1> %0, <vscale x 4 x i16> %1, <vscale x 4 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv4i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,m1,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v9, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 4 x i16> @llvm.riscv.vzext.mask.nxv4i16.nxv4i8(
<vscale x 4 x i16> %1,
<vscale x 4 x i8> %2,
<vscale x 4 x i1> %0,
i64 %3)
ret <vscale x 4 x i16> %a
}
declare <vscale x 8 x i16> @llvm.riscv.vzext.nxv8i16.nxv8i8(
<vscale x 8 x i8>,
i64);
define <vscale x 8 x i16> @intrinsic_vzext_vf2_nxv8i16(<vscale x 8 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv8i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,m2,ta,mu
; CHECK-NEXT: vzext.vf2 v26, v8
; CHECK-NEXT: vmv2r.v v8, v26
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i16> @llvm.riscv.vzext.nxv8i16.nxv8i8(
<vscale x 8 x i8> %0,
i64 %1)
ret <vscale x 8 x i16> %a
}
declare <vscale x 8 x i16> @llvm.riscv.vzext.mask.nxv8i16.nxv8i8(
<vscale x 8 x i16>,
<vscale x 8 x i8>,
<vscale x 8 x i1>,
i64);
define <vscale x 8 x i16> @intrinsic_vzext_mask_vf2_nxv8i16(<vscale x 8 x i1> %0, <vscale x 8 x i16> %1, <vscale x 8 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv8i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,m2,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v10, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 8 x i16> @llvm.riscv.vzext.mask.nxv8i16.nxv8i8(
<vscale x 8 x i16> %1,
<vscale x 8 x i8> %2,
<vscale x 8 x i1> %0,
i64 %3)
ret <vscale x 8 x i16> %a
}
declare <vscale x 16 x i16> @llvm.riscv.vzext.nxv16i16.nxv16i8(
<vscale x 16 x i8>,
i64);
define <vscale x 16 x i16> @intrinsic_vzext_vf2_nxv16i16(<vscale x 16 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv16i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,m4,ta,mu
; CHECK-NEXT: vzext.vf2 v28, v8
; CHECK-NEXT: vmv4r.v v8, v28
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 16 x i16> @llvm.riscv.vzext.nxv16i16.nxv16i8(
<vscale x 16 x i8> %0,
i64 %1)
ret <vscale x 16 x i16> %a
}
declare <vscale x 16 x i16> @llvm.riscv.vzext.mask.nxv16i16.nxv16i8(
<vscale x 16 x i16>,
<vscale x 16 x i8>,
<vscale x 16 x i1>,
i64);
define <vscale x 16 x i16> @intrinsic_vzext_mask_vf2_nxv16i16(<vscale x 16 x i1> %0, <vscale x 16 x i16> %1, <vscale x 16 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv16i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,m4,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v12, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 16 x i16> @llvm.riscv.vzext.mask.nxv16i16.nxv16i8(
<vscale x 16 x i16> %1,
<vscale x 16 x i8> %2,
<vscale x 16 x i1> %0,
i64 %3)
ret <vscale x 16 x i16> %a
}
declare <vscale x 32 x i16> @llvm.riscv.vzext.nxv32i16.nxv32i8(
<vscale x 32 x i8>,
i64);
define <vscale x 32 x i16> @intrinsic_vzext_vf2_nxv32i16(<vscale x 32 x i8> %0, i64 %1) nounwind {
; CHECK-LABEL: intrinsic_vzext_vf2_nxv32i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,m8,ta,mu
; CHECK-NEXT: vzext.vf2 v16, v8
; CHECK-NEXT: vmv8r.v v8, v16
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 32 x i16> @llvm.riscv.vzext.nxv32i16.nxv32i8(
<vscale x 32 x i8> %0,
i64 %1)
ret <vscale x 32 x i16> %a
}
declare <vscale x 32 x i16> @llvm.riscv.vzext.mask.nxv32i16.nxv32i8(
<vscale x 32 x i16>,
<vscale x 32 x i8>,
<vscale x 32 x i1>,
i64);
define <vscale x 32 x i16> @intrinsic_vzext_mask_vf2_nxv32i16(<vscale x 32 x i1> %0, <vscale x 32 x i16> %1, <vscale x 32 x i8> %2, i64 %3) nounwind {
; CHECK-LABEL: intrinsic_vzext_mask_vf2_nxv32i16:
; CHECK: # %bb.0: # %entry
; CHECK-NEXT: vsetvli a0, a0, e16,m8,tu,mu
; CHECK-NEXT: vzext.vf2 v8, v16, v0.t
; CHECK-NEXT: jalr zero, 0(ra)
entry:
%a = call <vscale x 32 x i16> @llvm.riscv.vzext.mask.nxv32i16.nxv32i8(
<vscale x 32 x i16> %1,
<vscale x 32 x i8> %2,
<vscale x 32 x i1> %0,
i64 %3)
ret <vscale x 32 x i16> %a
}