[llvm] [RISCV] Add tests for vmadd for VP intrinsics. NFC (PR #70042)
Luke Lau via llvm-commits
llvm-commits at lists.llvm.org
Tue Oct 24 06:44:02 PDT 2023
https://github.com/lukel97 created https://github.com/llvm/llvm-project/pull/70042
We have VP tests for vmacc but not vmadd. This copies the vmacc tests but swaps
the false operand of vp.merge to be the multiplicand instead of the addend.
This shows how we could fold the vmerge into the vmadd's mask if we commuted %a
and %b.
>From ce718ec22f230ab97d610eea45debe1ab19751b4 Mon Sep 17 00:00:00 2001
From: Luke Lau <luke at igalia.com>
Date: Sat, 21 Oct 2023 22:51:40 -0400
Subject: [PATCH] [RISCV] Add tests for vmadd for VP intrinsics. NFC
We have VP tests for vmacc but not vmadd. This copies the vmacc tests but swaps
the false operand of vp.merge to be the multiplicand instead of the addend.
This shows how we could fold the vmerge into the vmadd's mask if we commuted %a
and %b.
---
llvm/test/CodeGen/RISCV/rvv/vmadd-vp.ll | 2427 +++++++++++++++++++++++
1 file changed, 2427 insertions(+)
create mode 100644 llvm/test/CodeGen/RISCV/rvv/vmadd-vp.ll
diff --git a/llvm/test/CodeGen/RISCV/rvv/vmadd-vp.ll b/llvm/test/CodeGen/RISCV/rvv/vmadd-vp.ll
new file mode 100644
index 000000000000000..6a6d7d2a41424c5
--- /dev/null
+++ b/llvm/test/CodeGen/RISCV/rvv/vmadd-vp.ll
@@ -0,0 +1,2427 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=riscv32 -mattr=+d,+zfh,+zvfh,+v,+m -target-abi=ilp32d \
+; RUN: -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV32
+; RUN: llc -mtriple=riscv64 -mattr=+d,+zfh,+zvfh,+v,+m -target-abi=lp64d \
+; RUN: -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV64
+
+declare <vscale x 1 x i8> @llvm.vp.mul.nxv1i8(<vscale x 1 x i8>, <vscale x 1 x i8>, <vscale x 1 x i1>, i32)
+declare <vscale x 1 x i8> @llvm.vp.add.nxv1i8(<vscale x 1 x i8>, <vscale x 1 x i8>, <vscale x 1 x i1>, i32)
+declare <vscale x 1 x i8> @llvm.vp.merge.nxv1i8(<vscale x 1 x i1>, <vscale x 1 x i8>, <vscale x 1 x i8>, i32)
+declare <vscale x 1 x i8> @llvm.vp.select.nxv1i8(<vscale x 1 x i1>, <vscale x 1 x i8>, <vscale x 1 x i8>, i32)
+
+define <vscale x 1 x i8> @vmadd_vv_nxv1i8(<vscale x 1 x i8> %a, <vscale x 1 x i8> %b, <vscale x 1 x i8> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e8, mf8, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i8> @llvm.vp.mul.nxv1i8(<vscale x 1 x i8> %a, <vscale x 1 x i8> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i8> @llvm.vp.add.nxv1i8(<vscale x 1 x i8> %x, <vscale x 1 x i8> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i8> @llvm.vp.merge.nxv1i8(<vscale x 1 x i1> %m, <vscale x 1 x i8> %y, <vscale x 1 x i8> %a, i32 %evl)
+ ret <vscale x 1 x i8> %u
+}
+
+define <vscale x 1 x i8> @vmadd_vv_nxv1i8_unmasked(<vscale x 1 x i8> %a, <vscale x 1 x i8> %b, <vscale x 1 x i8> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e8, mf8, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i8> @llvm.vp.mul.nxv1i8(<vscale x 1 x i8> %a, <vscale x 1 x i8> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i8> @llvm.vp.add.nxv1i8(<vscale x 1 x i8> %x, <vscale x 1 x i8> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i8> @llvm.vp.merge.nxv1i8(<vscale x 1 x i1> %allones, <vscale x 1 x i8> %y, <vscale x 1 x i8> %a, i32 %evl)
+ ret <vscale x 1 x i8> %u
+}
+
+define <vscale x 1 x i8> @vmadd_vx_nxv1i8(<vscale x 1 x i8> %a, i8 %b, <vscale x 1 x i8> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv1i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v9, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i8> %elt.head, <vscale x 1 x i8> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i8> @llvm.vp.mul.nxv1i8(<vscale x 1 x i8> %a, <vscale x 1 x i8> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i8> @llvm.vp.add.nxv1i8(<vscale x 1 x i8> %x, <vscale x 1 x i8> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i8> @llvm.vp.merge.nxv1i8(<vscale x 1 x i1> %m, <vscale x 1 x i8> %y, <vscale x 1 x i8> %a, i32 %evl)
+ ret <vscale x 1 x i8> %u
+}
+
+define <vscale x 1 x i8> @vmadd_vx_nxv1i8_unmasked(<vscale x 1 x i8> %a, i8 %b, <vscale x 1 x i8> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv1i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v9
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i8> %elt.head, <vscale x 1 x i8> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i8> @llvm.vp.mul.nxv1i8(<vscale x 1 x i8> %a, <vscale x 1 x i8> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i8> @llvm.vp.add.nxv1i8(<vscale x 1 x i8> %x, <vscale x 1 x i8> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i8> @llvm.vp.merge.nxv1i8(<vscale x 1 x i1> %allones, <vscale x 1 x i8> %y, <vscale x 1 x i8> %a, i32 %evl)
+ ret <vscale x 1 x i8> %u
+}
+
+define <vscale x 1 x i8> @vmadd_vv_nxv1i8_ta(<vscale x 1 x i8> %a, <vscale x 1 x i8> %b, <vscale x 1 x i8> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, mf8, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i8> @llvm.vp.mul.nxv1i8(<vscale x 1 x i8> %a, <vscale x 1 x i8> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i8> @llvm.vp.add.nxv1i8(<vscale x 1 x i8> %x, <vscale x 1 x i8> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i8> @llvm.vp.select.nxv1i8(<vscale x 1 x i1> %m, <vscale x 1 x i8> %y, <vscale x 1 x i8> %a, i32 %evl)
+ ret <vscale x 1 x i8> %u
+}
+
+define <vscale x 1 x i8> @vmadd_vx_nxv1i8_ta(<vscale x 1 x i8> %a, i8 %b, <vscale x 1 x i8> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv1i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, mf8, ta, ma
+; CHECK-NEXT: vmacc.vx v9, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i8> %elt.head, <vscale x 1 x i8> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i8> @llvm.vp.mul.nxv1i8(<vscale x 1 x i8> %a, <vscale x 1 x i8> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i8> @llvm.vp.add.nxv1i8(<vscale x 1 x i8> %x, <vscale x 1 x i8> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i8> @llvm.vp.select.nxv1i8(<vscale x 1 x i1> %m, <vscale x 1 x i8> %y, <vscale x 1 x i8> %a, i32 %evl)
+ ret <vscale x 1 x i8> %u
+}
+
+declare <vscale x 2 x i8> @llvm.vp.mul.nxv2i8(<vscale x 2 x i8>, <vscale x 2 x i8>, <vscale x 2 x i1>, i32)
+declare <vscale x 2 x i8> @llvm.vp.add.nxv2i8(<vscale x 2 x i8>, <vscale x 2 x i8>, <vscale x 2 x i1>, i32)
+declare <vscale x 2 x i8> @llvm.vp.merge.nxv2i8(<vscale x 2 x i1>, <vscale x 2 x i8>, <vscale x 2 x i8>, i32)
+declare <vscale x 2 x i8> @llvm.vp.select.nxv2i8(<vscale x 2 x i1>, <vscale x 2 x i8>, <vscale x 2 x i8>, i32)
+
+define <vscale x 2 x i8> @vmadd_vv_nxv2i8(<vscale x 2 x i8> %a, <vscale x 2 x i8> %b, <vscale x 2 x i8> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e8, mf4, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i8> @llvm.vp.mul.nxv2i8(<vscale x 2 x i8> %a, <vscale x 2 x i8> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i8> @llvm.vp.add.nxv2i8(<vscale x 2 x i8> %x, <vscale x 2 x i8> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i8> @llvm.vp.merge.nxv2i8(<vscale x 2 x i1> %m, <vscale x 2 x i8> %y, <vscale x 2 x i8> %a, i32 %evl)
+ ret <vscale x 2 x i8> %u
+}
+
+define <vscale x 2 x i8> @vmadd_vv_nxv2i8_unmasked(<vscale x 2 x i8> %a, <vscale x 2 x i8> %b, <vscale x 2 x i8> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e8, mf4, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i8> @llvm.vp.mul.nxv2i8(<vscale x 2 x i8> %a, <vscale x 2 x i8> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i8> @llvm.vp.add.nxv2i8(<vscale x 2 x i8> %x, <vscale x 2 x i8> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i8> @llvm.vp.merge.nxv2i8(<vscale x 2 x i1> %allones, <vscale x 2 x i8> %y, <vscale x 2 x i8> %a, i32 %evl)
+ ret <vscale x 2 x i8> %u
+}
+
+define <vscale x 2 x i8> @vmadd_vx_nxv2i8(<vscale x 2 x i8> %a, i8 %b, <vscale x 2 x i8> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv2i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v9, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i8> %elt.head, <vscale x 2 x i8> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i8> @llvm.vp.mul.nxv2i8(<vscale x 2 x i8> %a, <vscale x 2 x i8> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i8> @llvm.vp.add.nxv2i8(<vscale x 2 x i8> %x, <vscale x 2 x i8> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i8> @llvm.vp.merge.nxv2i8(<vscale x 2 x i1> %m, <vscale x 2 x i8> %y, <vscale x 2 x i8> %a, i32 %evl)
+ ret <vscale x 2 x i8> %u
+}
+
+define <vscale x 2 x i8> @vmadd_vx_nxv2i8_unmasked(<vscale x 2 x i8> %a, i8 %b, <vscale x 2 x i8> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv2i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v9
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i8> %elt.head, <vscale x 2 x i8> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i8> @llvm.vp.mul.nxv2i8(<vscale x 2 x i8> %a, <vscale x 2 x i8> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i8> @llvm.vp.add.nxv2i8(<vscale x 2 x i8> %x, <vscale x 2 x i8> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i8> @llvm.vp.merge.nxv2i8(<vscale x 2 x i1> %allones, <vscale x 2 x i8> %y, <vscale x 2 x i8> %a, i32 %evl)
+ ret <vscale x 2 x i8> %u
+}
+
+define <vscale x 2 x i8> @vmadd_vv_nxv2i8_ta(<vscale x 2 x i8> %a, <vscale x 2 x i8> %b, <vscale x 2 x i8> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, mf4, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i8> @llvm.vp.mul.nxv2i8(<vscale x 2 x i8> %a, <vscale x 2 x i8> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i8> @llvm.vp.add.nxv2i8(<vscale x 2 x i8> %x, <vscale x 2 x i8> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i8> @llvm.vp.select.nxv2i8(<vscale x 2 x i1> %m, <vscale x 2 x i8> %y, <vscale x 2 x i8> %a, i32 %evl)
+ ret <vscale x 2 x i8> %u
+}
+
+define <vscale x 2 x i8> @vmadd_vx_nxv2i8_ta(<vscale x 2 x i8> %a, i8 %b, <vscale x 2 x i8> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv2i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, mf4, ta, ma
+; CHECK-NEXT: vmacc.vx v9, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i8> %elt.head, <vscale x 2 x i8> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i8> @llvm.vp.mul.nxv2i8(<vscale x 2 x i8> %a, <vscale x 2 x i8> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i8> @llvm.vp.add.nxv2i8(<vscale x 2 x i8> %x, <vscale x 2 x i8> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i8> @llvm.vp.select.nxv2i8(<vscale x 2 x i1> %m, <vscale x 2 x i8> %y, <vscale x 2 x i8> %a, i32 %evl)
+ ret <vscale x 2 x i8> %u
+}
+
+declare <vscale x 4 x i8> @llvm.vp.mul.nxv4i8(<vscale x 4 x i8>, <vscale x 4 x i8>, <vscale x 4 x i1>, i32)
+declare <vscale x 4 x i8> @llvm.vp.add.nxv4i8(<vscale x 4 x i8>, <vscale x 4 x i8>, <vscale x 4 x i1>, i32)
+declare <vscale x 4 x i8> @llvm.vp.merge.nxv4i8(<vscale x 4 x i1>, <vscale x 4 x i8>, <vscale x 4 x i8>, i32)
+declare <vscale x 4 x i8> @llvm.vp.select.nxv4i8(<vscale x 4 x i1>, <vscale x 4 x i8>, <vscale x 4 x i8>, i32)
+
+define <vscale x 4 x i8> @vmadd_vv_nxv4i8(<vscale x 4 x i8> %a, <vscale x 4 x i8> %b, <vscale x 4 x i8> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e8, mf2, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i8> @llvm.vp.mul.nxv4i8(<vscale x 4 x i8> %a, <vscale x 4 x i8> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i8> @llvm.vp.add.nxv4i8(<vscale x 4 x i8> %x, <vscale x 4 x i8> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i8> @llvm.vp.merge.nxv4i8(<vscale x 4 x i1> %m, <vscale x 4 x i8> %y, <vscale x 4 x i8> %a, i32 %evl)
+ ret <vscale x 4 x i8> %u
+}
+
+define <vscale x 4 x i8> @vmadd_vv_nxv4i8_unmasked(<vscale x 4 x i8> %a, <vscale x 4 x i8> %b, <vscale x 4 x i8> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e8, mf2, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i8> @llvm.vp.mul.nxv4i8(<vscale x 4 x i8> %a, <vscale x 4 x i8> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i8> @llvm.vp.add.nxv4i8(<vscale x 4 x i8> %x, <vscale x 4 x i8> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i8> @llvm.vp.merge.nxv4i8(<vscale x 4 x i1> %allones, <vscale x 4 x i8> %y, <vscale x 4 x i8> %a, i32 %evl)
+ ret <vscale x 4 x i8> %u
+}
+
+define <vscale x 4 x i8> @vmadd_vx_nxv4i8(<vscale x 4 x i8> %a, i8 %b, <vscale x 4 x i8> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv4i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v9, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i8> %elt.head, <vscale x 4 x i8> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i8> @llvm.vp.mul.nxv4i8(<vscale x 4 x i8> %a, <vscale x 4 x i8> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i8> @llvm.vp.add.nxv4i8(<vscale x 4 x i8> %x, <vscale x 4 x i8> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i8> @llvm.vp.merge.nxv4i8(<vscale x 4 x i1> %m, <vscale x 4 x i8> %y, <vscale x 4 x i8> %a, i32 %evl)
+ ret <vscale x 4 x i8> %u
+}
+
+define <vscale x 4 x i8> @vmadd_vx_nxv4i8_unmasked(<vscale x 4 x i8> %a, i8 %b, <vscale x 4 x i8> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv4i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v9
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i8> %elt.head, <vscale x 4 x i8> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i8> @llvm.vp.mul.nxv4i8(<vscale x 4 x i8> %a, <vscale x 4 x i8> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i8> @llvm.vp.add.nxv4i8(<vscale x 4 x i8> %x, <vscale x 4 x i8> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i8> @llvm.vp.merge.nxv4i8(<vscale x 4 x i1> %allones, <vscale x 4 x i8> %y, <vscale x 4 x i8> %a, i32 %evl)
+ ret <vscale x 4 x i8> %u
+}
+
+define <vscale x 4 x i8> @vmadd_vv_nxv4i8_ta(<vscale x 4 x i8> %a, <vscale x 4 x i8> %b, <vscale x 4 x i8> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, mf2, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i8> @llvm.vp.mul.nxv4i8(<vscale x 4 x i8> %a, <vscale x 4 x i8> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i8> @llvm.vp.add.nxv4i8(<vscale x 4 x i8> %x, <vscale x 4 x i8> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i8> @llvm.vp.select.nxv4i8(<vscale x 4 x i1> %m, <vscale x 4 x i8> %y, <vscale x 4 x i8> %a, i32 %evl)
+ ret <vscale x 4 x i8> %u
+}
+
+define <vscale x 4 x i8> @vmadd_vx_nxv4i8_ta(<vscale x 4 x i8> %a, i8 %b, <vscale x 4 x i8> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv4i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, mf2, ta, ma
+; CHECK-NEXT: vmacc.vx v9, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i8> %elt.head, <vscale x 4 x i8> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i8> @llvm.vp.mul.nxv4i8(<vscale x 4 x i8> %a, <vscale x 4 x i8> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i8> @llvm.vp.add.nxv4i8(<vscale x 4 x i8> %x, <vscale x 4 x i8> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i8> @llvm.vp.select.nxv4i8(<vscale x 4 x i1> %m, <vscale x 4 x i8> %y, <vscale x 4 x i8> %a, i32 %evl)
+ ret <vscale x 4 x i8> %u
+}
+
+declare <vscale x 8 x i8> @llvm.vp.mul.nxv8i8(<vscale x 8 x i8>, <vscale x 8 x i8>, <vscale x 8 x i1>, i32)
+declare <vscale x 8 x i8> @llvm.vp.add.nxv8i8(<vscale x 8 x i8>, <vscale x 8 x i8>, <vscale x 8 x i1>, i32)
+declare <vscale x 8 x i8> @llvm.vp.merge.nxv8i8(<vscale x 8 x i1>, <vscale x 8 x i8>, <vscale x 8 x i8>, i32)
+declare <vscale x 8 x i8> @llvm.vp.select.nxv8i8(<vscale x 8 x i1>, <vscale x 8 x i8>, <vscale x 8 x i8>, i32)
+
+define <vscale x 8 x i8> @vmadd_vv_nxv8i8(<vscale x 8 x i8> %a, <vscale x 8 x i8> %b, <vscale x 8 x i8> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e8, m1, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i8> @llvm.vp.mul.nxv8i8(<vscale x 8 x i8> %a, <vscale x 8 x i8> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i8> @llvm.vp.add.nxv8i8(<vscale x 8 x i8> %x, <vscale x 8 x i8> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i8> @llvm.vp.merge.nxv8i8(<vscale x 8 x i1> %m, <vscale x 8 x i8> %y, <vscale x 8 x i8> %a, i32 %evl)
+ ret <vscale x 8 x i8> %u
+}
+
+define <vscale x 8 x i8> @vmadd_vv_nxv8i8_unmasked(<vscale x 8 x i8> %a, <vscale x 8 x i8> %b, <vscale x 8 x i8> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e8, m1, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i8> @llvm.vp.mul.nxv8i8(<vscale x 8 x i8> %a, <vscale x 8 x i8> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i8> @llvm.vp.add.nxv8i8(<vscale x 8 x i8> %x, <vscale x 8 x i8> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i8> @llvm.vp.merge.nxv8i8(<vscale x 8 x i1> %allones, <vscale x 8 x i8> %y, <vscale x 8 x i8> %a, i32 %evl)
+ ret <vscale x 8 x i8> %u
+}
+
+define <vscale x 8 x i8> @vmadd_vx_nxv8i8(<vscale x 8 x i8> %a, i8 %b, <vscale x 8 x i8> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv8i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v9, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i8> %elt.head, <vscale x 8 x i8> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i8> @llvm.vp.mul.nxv8i8(<vscale x 8 x i8> %a, <vscale x 8 x i8> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i8> @llvm.vp.add.nxv8i8(<vscale x 8 x i8> %x, <vscale x 8 x i8> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i8> @llvm.vp.merge.nxv8i8(<vscale x 8 x i1> %m, <vscale x 8 x i8> %y, <vscale x 8 x i8> %a, i32 %evl)
+ ret <vscale x 8 x i8> %u
+}
+
+define <vscale x 8 x i8> @vmadd_vx_nxv8i8_unmasked(<vscale x 8 x i8> %a, i8 %b, <vscale x 8 x i8> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv8i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v9
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i8> %elt.head, <vscale x 8 x i8> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i8> @llvm.vp.mul.nxv8i8(<vscale x 8 x i8> %a, <vscale x 8 x i8> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i8> @llvm.vp.add.nxv8i8(<vscale x 8 x i8> %x, <vscale x 8 x i8> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i8> @llvm.vp.merge.nxv8i8(<vscale x 8 x i1> %allones, <vscale x 8 x i8> %y, <vscale x 8 x i8> %a, i32 %evl)
+ ret <vscale x 8 x i8> %u
+}
+
+define <vscale x 8 x i8> @vmadd_vv_nxv8i8_ta(<vscale x 8 x i8> %a, <vscale x 8 x i8> %b, <vscale x 8 x i8> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i8> @llvm.vp.mul.nxv8i8(<vscale x 8 x i8> %a, <vscale x 8 x i8> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i8> @llvm.vp.add.nxv8i8(<vscale x 8 x i8> %x, <vscale x 8 x i8> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i8> @llvm.vp.select.nxv8i8(<vscale x 8 x i1> %m, <vscale x 8 x i8> %y, <vscale x 8 x i8> %a, i32 %evl)
+ ret <vscale x 8 x i8> %u
+}
+
+define <vscale x 8 x i8> @vmadd_vx_nxv8i8_ta(<vscale x 8 x i8> %a, i8 %b, <vscale x 8 x i8> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv8i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m1, ta, ma
+; CHECK-NEXT: vmacc.vx v9, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i8> %elt.head, <vscale x 8 x i8> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i8> @llvm.vp.mul.nxv8i8(<vscale x 8 x i8> %a, <vscale x 8 x i8> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i8> @llvm.vp.add.nxv8i8(<vscale x 8 x i8> %x, <vscale x 8 x i8> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i8> @llvm.vp.select.nxv8i8(<vscale x 8 x i1> %m, <vscale x 8 x i8> %y, <vscale x 8 x i8> %a, i32 %evl)
+ ret <vscale x 8 x i8> %u
+}
+
+declare <vscale x 16 x i8> @llvm.vp.mul.nxv16i8(<vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i1>, i32)
+declare <vscale x 16 x i8> @llvm.vp.add.nxv16i8(<vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i1>, i32)
+declare <vscale x 16 x i8> @llvm.vp.merge.nxv16i8(<vscale x 16 x i1>, <vscale x 16 x i8>, <vscale x 16 x i8>, i32)
+declare <vscale x 16 x i8> @llvm.vp.select.nxv16i8(<vscale x 16 x i1>, <vscale x 16 x i8>, <vscale x 16 x i8>, i32)
+
+define <vscale x 16 x i8> @vmadd_vv_nxv16i8(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv16i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vsetvli zero, zero, e8, m2, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i8> @llvm.vp.mul.nxv16i8(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i8> @llvm.vp.add.nxv16i8(<vscale x 16 x i8> %x, <vscale x 16 x i8> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i8> @llvm.vp.merge.nxv16i8(<vscale x 16 x i1> %m, <vscale x 16 x i8> %y, <vscale x 16 x i8> %a, i32 %evl)
+ ret <vscale x 16 x i8> %u
+}
+
+define <vscale x 16 x i8> @vmadd_vv_nxv16i8_unmasked(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv16i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vsetvli zero, zero, e8, m2, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v10
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i8> @llvm.vp.mul.nxv16i8(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i8> @llvm.vp.add.nxv16i8(<vscale x 16 x i8> %x, <vscale x 16 x i8> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i8> @llvm.vp.merge.nxv16i8(<vscale x 16 x i1> %allones, <vscale x 16 x i8> %y, <vscale x 16 x i8> %a, i32 %evl)
+ ret <vscale x 16 x i8> %u
+}
+
+define <vscale x 16 x i8> @vmadd_vx_nxv16i8(<vscale x 16 x i8> %a, i8 %b, <vscale x 16 x i8> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv16i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v10, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 16 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 16 x i8> %elt.head, <vscale x 16 x i8> poison, <vscale x 16 x i32> zeroinitializer
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i8> @llvm.vp.mul.nxv16i8(<vscale x 16 x i8> %a, <vscale x 16 x i8> %vb, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i8> @llvm.vp.add.nxv16i8(<vscale x 16 x i8> %x, <vscale x 16 x i8> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i8> @llvm.vp.merge.nxv16i8(<vscale x 16 x i1> %m, <vscale x 16 x i8> %y, <vscale x 16 x i8> %a, i32 %evl)
+ ret <vscale x 16 x i8> %u
+}
+
+define <vscale x 16 x i8> @vmadd_vx_nxv16i8_unmasked(<vscale x 16 x i8> %a, i8 %b, <vscale x 16 x i8> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv16i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v10
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 16 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 16 x i8> %elt.head, <vscale x 16 x i8> poison, <vscale x 16 x i32> zeroinitializer
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i8> @llvm.vp.mul.nxv16i8(<vscale x 16 x i8> %a, <vscale x 16 x i8> %vb, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i8> @llvm.vp.add.nxv16i8(<vscale x 16 x i8> %x, <vscale x 16 x i8> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i8> @llvm.vp.merge.nxv16i8(<vscale x 16 x i1> %allones, <vscale x 16 x i8> %y, <vscale x 16 x i8> %a, i32 %evl)
+ ret <vscale x 16 x i8> %u
+}
+
+define <vscale x 16 x i8> @vmadd_vv_nxv16i8_ta(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i8> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv16i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i8> @llvm.vp.mul.nxv16i8(<vscale x 16 x i8> %a, <vscale x 16 x i8> %b, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i8> @llvm.vp.add.nxv16i8(<vscale x 16 x i8> %x, <vscale x 16 x i8> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i8> @llvm.vp.select.nxv16i8(<vscale x 16 x i1> %m, <vscale x 16 x i8> %y, <vscale x 16 x i8> %a, i32 %evl)
+ ret <vscale x 16 x i8> %u
+}
+
+define <vscale x 16 x i8> @vmadd_vx_nxv16i8_ta(<vscale x 16 x i8> %a, i8 %b, <vscale x 16 x i8> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv16i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m2, ta, ma
+; CHECK-NEXT: vmacc.vx v10, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 16 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 16 x i8> %elt.head, <vscale x 16 x i8> poison, <vscale x 16 x i32> zeroinitializer
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i8> @llvm.vp.mul.nxv16i8(<vscale x 16 x i8> %a, <vscale x 16 x i8> %vb, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i8> @llvm.vp.add.nxv16i8(<vscale x 16 x i8> %x, <vscale x 16 x i8> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i8> @llvm.vp.select.nxv16i8(<vscale x 16 x i1> %m, <vscale x 16 x i8> %y, <vscale x 16 x i8> %a, i32 %evl)
+ ret <vscale x 16 x i8> %u
+}
+
+declare <vscale x 32 x i8> @llvm.vp.mul.nxv32i8(<vscale x 32 x i8>, <vscale x 32 x i8>, <vscale x 32 x i1>, i32)
+declare <vscale x 32 x i8> @llvm.vp.add.nxv32i8(<vscale x 32 x i8>, <vscale x 32 x i8>, <vscale x 32 x i1>, i32)
+declare <vscale x 32 x i8> @llvm.vp.merge.nxv32i8(<vscale x 32 x i1>, <vscale x 32 x i8>, <vscale x 32 x i8>, i32)
+declare <vscale x 32 x i8> @llvm.vp.select.nxv32i8(<vscale x 32 x i1>, <vscale x 32 x i8>, <vscale x 32 x i8>, i32)
+
+define <vscale x 32 x i8> @vmadd_vv_nxv32i8(<vscale x 32 x i8> %a, <vscale x 32 x i8> %b, <vscale x 32 x i8> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv32i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e8, m4, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i8> @llvm.vp.mul.nxv32i8(<vscale x 32 x i8> %a, <vscale x 32 x i8> %b, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i8> @llvm.vp.add.nxv32i8(<vscale x 32 x i8> %x, <vscale x 32 x i8> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i8> @llvm.vp.merge.nxv32i8(<vscale x 32 x i1> %m, <vscale x 32 x i8> %y, <vscale x 32 x i8> %a, i32 %evl)
+ ret <vscale x 32 x i8> %u
+}
+
+define <vscale x 32 x i8> @vmadd_vv_nxv32i8_unmasked(<vscale x 32 x i8> %a, <vscale x 32 x i8> %b, <vscale x 32 x i8> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv32i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e8, m4, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v12
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i8> @llvm.vp.mul.nxv32i8(<vscale x 32 x i8> %a, <vscale x 32 x i8> %b, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i8> @llvm.vp.add.nxv32i8(<vscale x 32 x i8> %x, <vscale x 32 x i8> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i8> @llvm.vp.merge.nxv32i8(<vscale x 32 x i1> %allones, <vscale x 32 x i8> %y, <vscale x 32 x i8> %a, i32 %evl)
+ ret <vscale x 32 x i8> %u
+}
+
+define <vscale x 32 x i8> @vmadd_vx_nxv32i8(<vscale x 32 x i8> %a, i8 %b, <vscale x 32 x i8> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv32i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v12, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 32 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 32 x i8> %elt.head, <vscale x 32 x i8> poison, <vscale x 32 x i32> zeroinitializer
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i8> @llvm.vp.mul.nxv32i8(<vscale x 32 x i8> %a, <vscale x 32 x i8> %vb, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i8> @llvm.vp.add.nxv32i8(<vscale x 32 x i8> %x, <vscale x 32 x i8> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i8> @llvm.vp.merge.nxv32i8(<vscale x 32 x i1> %m, <vscale x 32 x i8> %y, <vscale x 32 x i8> %a, i32 %evl)
+ ret <vscale x 32 x i8> %u
+}
+
+define <vscale x 32 x i8> @vmadd_vx_nxv32i8_unmasked(<vscale x 32 x i8> %a, i8 %b, <vscale x 32 x i8> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv32i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v12
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 32 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 32 x i8> %elt.head, <vscale x 32 x i8> poison, <vscale x 32 x i32> zeroinitializer
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i8> @llvm.vp.mul.nxv32i8(<vscale x 32 x i8> %a, <vscale x 32 x i8> %vb, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i8> @llvm.vp.add.nxv32i8(<vscale x 32 x i8> %x, <vscale x 32 x i8> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i8> @llvm.vp.merge.nxv32i8(<vscale x 32 x i1> %allones, <vscale x 32 x i8> %y, <vscale x 32 x i8> %a, i32 %evl)
+ ret <vscale x 32 x i8> %u
+}
+
+define <vscale x 32 x i8> @vmadd_vv_nxv32i8_ta(<vscale x 32 x i8> %a, <vscale x 32 x i8> %b, <vscale x 32 x i8> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv32i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e8, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i8> @llvm.vp.mul.nxv32i8(<vscale x 32 x i8> %a, <vscale x 32 x i8> %b, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i8> @llvm.vp.add.nxv32i8(<vscale x 32 x i8> %x, <vscale x 32 x i8> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i8> @llvm.vp.select.nxv32i8(<vscale x 32 x i1> %m, <vscale x 32 x i8> %y, <vscale x 32 x i8> %a, i32 %evl)
+ ret <vscale x 32 x i8> %u
+}
+
+define <vscale x 32 x i8> @vmadd_vx_nxv32i8_ta(<vscale x 32 x i8> %a, i8 %b, <vscale x 32 x i8> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv32i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m4, ta, ma
+; CHECK-NEXT: vmacc.vx v12, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 32 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 32 x i8> %elt.head, <vscale x 32 x i8> poison, <vscale x 32 x i32> zeroinitializer
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i8> @llvm.vp.mul.nxv32i8(<vscale x 32 x i8> %a, <vscale x 32 x i8> %vb, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i8> @llvm.vp.add.nxv32i8(<vscale x 32 x i8> %x, <vscale x 32 x i8> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i8> @llvm.vp.select.nxv32i8(<vscale x 32 x i1> %m, <vscale x 32 x i8> %y, <vscale x 32 x i8> %a, i32 %evl)
+ ret <vscale x 32 x i8> %u
+}
+
+declare <vscale x 64 x i8> @llvm.vp.mul.nxv64i8(<vscale x 64 x i8>, <vscale x 64 x i8>, <vscale x 64 x i1>, i32)
+declare <vscale x 64 x i8> @llvm.vp.add.nxv64i8(<vscale x 64 x i8>, <vscale x 64 x i8>, <vscale x 64 x i1>, i32)
+declare <vscale x 64 x i8> @llvm.vp.merge.nxv64i8(<vscale x 64 x i1>, <vscale x 64 x i8>, <vscale x 64 x i8>, i32)
+declare <vscale x 64 x i8> @llvm.vp.select.nxv64i8(<vscale x 64 x i1>, <vscale x 64 x i8>, <vscale x 64 x i8>, i32)
+
+define <vscale x 64 x i8> @vmadd_vv_nxv64i8(<vscale x 64 x i8> %a, <vscale x 64 x i8> %b, <vscale x 64 x i8> %c, <vscale x 64 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv64i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8r.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e8, m8, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 64 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 64 x i1> %splat, <vscale x 64 x i1> poison, <vscale x 64 x i32> zeroinitializer
+ %x = call <vscale x 64 x i8> @llvm.vp.mul.nxv64i8(<vscale x 64 x i8> %a, <vscale x 64 x i8> %b, <vscale x 64 x i1> %allones, i32 %evl)
+ %y = call <vscale x 64 x i8> @llvm.vp.add.nxv64i8(<vscale x 64 x i8> %x, <vscale x 64 x i8> %c, <vscale x 64 x i1> %allones, i32 %evl)
+ %u = call <vscale x 64 x i8> @llvm.vp.merge.nxv64i8(<vscale x 64 x i1> %m, <vscale x 64 x i8> %y, <vscale x 64 x i8> %a, i32 %evl)
+ ret <vscale x 64 x i8> %u
+}
+
+define <vscale x 64 x i8> @vmadd_vv_nxv64i8_unmasked(<vscale x 64 x i8> %a, <vscale x 64 x i8> %b, <vscale x 64 x i8> %c, <vscale x 64 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv64i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8r.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e8, m8, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v24
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 64 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 64 x i1> %splat, <vscale x 64 x i1> poison, <vscale x 64 x i32> zeroinitializer
+ %x = call <vscale x 64 x i8> @llvm.vp.mul.nxv64i8(<vscale x 64 x i8> %a, <vscale x 64 x i8> %b, <vscale x 64 x i1> %allones, i32 %evl)
+ %y = call <vscale x 64 x i8> @llvm.vp.add.nxv64i8(<vscale x 64 x i8> %x, <vscale x 64 x i8> %c, <vscale x 64 x i1> %allones, i32 %evl)
+ %u = call <vscale x 64 x i8> @llvm.vp.merge.nxv64i8(<vscale x 64 x i1> %allones, <vscale x 64 x i8> %y, <vscale x 64 x i8> %a, i32 %evl)
+ ret <vscale x 64 x i8> %u
+}
+
+define <vscale x 64 x i8> @vmadd_vx_nxv64i8(<vscale x 64 x i8> %a, i8 %b, <vscale x 64 x i8> %c, <vscale x 64 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv64i8:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v16, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 64 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 64 x i8> %elt.head, <vscale x 64 x i8> poison, <vscale x 64 x i32> zeroinitializer
+ %splat = insertelement <vscale x 64 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 64 x i1> %splat, <vscale x 64 x i1> poison, <vscale x 64 x i32> zeroinitializer
+ %x = call <vscale x 64 x i8> @llvm.vp.mul.nxv64i8(<vscale x 64 x i8> %a, <vscale x 64 x i8> %vb, <vscale x 64 x i1> %allones, i32 %evl)
+ %y = call <vscale x 64 x i8> @llvm.vp.add.nxv64i8(<vscale x 64 x i8> %x, <vscale x 64 x i8> %c, <vscale x 64 x i1> %allones, i32 %evl)
+ %u = call <vscale x 64 x i8> @llvm.vp.merge.nxv64i8(<vscale x 64 x i1> %m, <vscale x 64 x i8> %y, <vscale x 64 x i8> %a, i32 %evl)
+ ret <vscale x 64 x i8> %u
+}
+
+define <vscale x 64 x i8> @vmadd_vx_nxv64i8_unmasked(<vscale x 64 x i8> %a, i8 %b, <vscale x 64 x i8> %c, <vscale x 64 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv64i8_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v16
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 64 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 64 x i8> %elt.head, <vscale x 64 x i8> poison, <vscale x 64 x i32> zeroinitializer
+ %splat = insertelement <vscale x 64 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 64 x i1> %splat, <vscale x 64 x i1> poison, <vscale x 64 x i32> zeroinitializer
+ %x = call <vscale x 64 x i8> @llvm.vp.mul.nxv64i8(<vscale x 64 x i8> %a, <vscale x 64 x i8> %vb, <vscale x 64 x i1> %allones, i32 %evl)
+ %y = call <vscale x 64 x i8> @llvm.vp.add.nxv64i8(<vscale x 64 x i8> %x, <vscale x 64 x i8> %c, <vscale x 64 x i1> %allones, i32 %evl)
+ %u = call <vscale x 64 x i8> @llvm.vp.merge.nxv64i8(<vscale x 64 x i1> %allones, <vscale x 64 x i8> %y, <vscale x 64 x i8> %a, i32 %evl)
+ ret <vscale x 64 x i8> %u
+}
+
+define <vscale x 64 x i8> @vmadd_vv_nxv64i8_ta(<vscale x 64 x i8> %a, <vscale x 64 x i8> %b, <vscale x 64 x i8> %c, <vscale x 64 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv64i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8r.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 64 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 64 x i1> %splat, <vscale x 64 x i1> poison, <vscale x 64 x i32> zeroinitializer
+ %x = call <vscale x 64 x i8> @llvm.vp.mul.nxv64i8(<vscale x 64 x i8> %a, <vscale x 64 x i8> %b, <vscale x 64 x i1> %allones, i32 %evl)
+ %y = call <vscale x 64 x i8> @llvm.vp.add.nxv64i8(<vscale x 64 x i8> %x, <vscale x 64 x i8> %c, <vscale x 64 x i1> %allones, i32 %evl)
+ %u = call <vscale x 64 x i8> @llvm.vp.select.nxv64i8(<vscale x 64 x i1> %m, <vscale x 64 x i8> %y, <vscale x 64 x i8> %a, i32 %evl)
+ ret <vscale x 64 x i8> %u
+}
+
+define <vscale x 64 x i8> @vmadd_vx_nxv64i8_ta(<vscale x 64 x i8> %a, i8 %b, <vscale x 64 x i8> %c, <vscale x 64 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv64i8_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e8, m8, ta, ma
+; CHECK-NEXT: vmacc.vx v16, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 64 x i8> poison, i8 %b, i32 0
+ %vb = shufflevector <vscale x 64 x i8> %elt.head, <vscale x 64 x i8> poison, <vscale x 64 x i32> zeroinitializer
+ %splat = insertelement <vscale x 64 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 64 x i1> %splat, <vscale x 64 x i1> poison, <vscale x 64 x i32> zeroinitializer
+ %x = call <vscale x 64 x i8> @llvm.vp.mul.nxv64i8(<vscale x 64 x i8> %a, <vscale x 64 x i8> %vb, <vscale x 64 x i1> %allones, i32 %evl)
+ %y = call <vscale x 64 x i8> @llvm.vp.add.nxv64i8(<vscale x 64 x i8> %x, <vscale x 64 x i8> %c, <vscale x 64 x i1> %allones, i32 %evl)
+ %u = call <vscale x 64 x i8> @llvm.vp.select.nxv64i8(<vscale x 64 x i1> %m, <vscale x 64 x i8> %y, <vscale x 64 x i8> %a, i32 %evl)
+ ret <vscale x 64 x i8> %u
+}
+
+declare <vscale x 1 x i16> @llvm.vp.mul.nxv1i16(<vscale x 1 x i16>, <vscale x 1 x i16>, <vscale x 1 x i1>, i32)
+declare <vscale x 1 x i16> @llvm.vp.add.nxv1i16(<vscale x 1 x i16>, <vscale x 1 x i16>, <vscale x 1 x i1>, i32)
+declare <vscale x 1 x i16> @llvm.vp.merge.nxv1i16(<vscale x 1 x i1>, <vscale x 1 x i16>, <vscale x 1 x i16>, i32)
+declare <vscale x 1 x i16> @llvm.vp.select.nxv1i16(<vscale x 1 x i1>, <vscale x 1 x i16>, <vscale x 1 x i16>, i32)
+
+define <vscale x 1 x i16> @vmadd_vv_nxv1i16(<vscale x 1 x i16> %a, <vscale x 1 x i16> %b, <vscale x 1 x i16> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i16> @llvm.vp.mul.nxv1i16(<vscale x 1 x i16> %a, <vscale x 1 x i16> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i16> @llvm.vp.add.nxv1i16(<vscale x 1 x i16> %x, <vscale x 1 x i16> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i16> @llvm.vp.merge.nxv1i16(<vscale x 1 x i1> %m, <vscale x 1 x i16> %y, <vscale x 1 x i16> %a, i32 %evl)
+ ret <vscale x 1 x i16> %u
+}
+
+define <vscale x 1 x i16> @vmadd_vv_nxv1i16_unmasked(<vscale x 1 x i16> %a, <vscale x 1 x i16> %b, <vscale x 1 x i16> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, mf4, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i16> @llvm.vp.mul.nxv1i16(<vscale x 1 x i16> %a, <vscale x 1 x i16> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i16> @llvm.vp.add.nxv1i16(<vscale x 1 x i16> %x, <vscale x 1 x i16> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i16> @llvm.vp.merge.nxv1i16(<vscale x 1 x i1> %allones, <vscale x 1 x i16> %y, <vscale x 1 x i16> %a, i32 %evl)
+ ret <vscale x 1 x i16> %u
+}
+
+define <vscale x 1 x i16> @vmadd_vx_nxv1i16(<vscale x 1 x i16> %a, i16 %b, <vscale x 1 x i16> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv1i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v9, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i16> %elt.head, <vscale x 1 x i16> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i16> @llvm.vp.mul.nxv1i16(<vscale x 1 x i16> %a, <vscale x 1 x i16> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i16> @llvm.vp.add.nxv1i16(<vscale x 1 x i16> %x, <vscale x 1 x i16> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i16> @llvm.vp.merge.nxv1i16(<vscale x 1 x i1> %m, <vscale x 1 x i16> %y, <vscale x 1 x i16> %a, i32 %evl)
+ ret <vscale x 1 x i16> %u
+}
+
+define <vscale x 1 x i16> @vmadd_vx_nxv1i16_unmasked(<vscale x 1 x i16> %a, i16 %b, <vscale x 1 x i16> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv1i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v9
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i16> %elt.head, <vscale x 1 x i16> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i16> @llvm.vp.mul.nxv1i16(<vscale x 1 x i16> %a, <vscale x 1 x i16> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i16> @llvm.vp.add.nxv1i16(<vscale x 1 x i16> %x, <vscale x 1 x i16> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i16> @llvm.vp.merge.nxv1i16(<vscale x 1 x i1> %allones, <vscale x 1 x i16> %y, <vscale x 1 x i16> %a, i32 %evl)
+ ret <vscale x 1 x i16> %u
+}
+
+define <vscale x 1 x i16> @vmadd_vv_nxv1i16_ta(<vscale x 1 x i16> %a, <vscale x 1 x i16> %b, <vscale x 1 x i16> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, mf4, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i16> @llvm.vp.mul.nxv1i16(<vscale x 1 x i16> %a, <vscale x 1 x i16> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i16> @llvm.vp.add.nxv1i16(<vscale x 1 x i16> %x, <vscale x 1 x i16> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i16> @llvm.vp.select.nxv1i16(<vscale x 1 x i1> %m, <vscale x 1 x i16> %y, <vscale x 1 x i16> %a, i32 %evl)
+ ret <vscale x 1 x i16> %u
+}
+
+define <vscale x 1 x i16> @vmadd_vx_nxv1i16_ta(<vscale x 1 x i16> %a, i16 %b, <vscale x 1 x i16> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv1i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, mf4, ta, ma
+; CHECK-NEXT: vmacc.vx v9, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i16> %elt.head, <vscale x 1 x i16> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i16> @llvm.vp.mul.nxv1i16(<vscale x 1 x i16> %a, <vscale x 1 x i16> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i16> @llvm.vp.add.nxv1i16(<vscale x 1 x i16> %x, <vscale x 1 x i16> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i16> @llvm.vp.select.nxv1i16(<vscale x 1 x i1> %m, <vscale x 1 x i16> %y, <vscale x 1 x i16> %a, i32 %evl)
+ ret <vscale x 1 x i16> %u
+}
+
+declare <vscale x 2 x i16> @llvm.vp.mul.nxv2i16(<vscale x 2 x i16>, <vscale x 2 x i16>, <vscale x 2 x i1>, i32)
+declare <vscale x 2 x i16> @llvm.vp.add.nxv2i16(<vscale x 2 x i16>, <vscale x 2 x i16>, <vscale x 2 x i1>, i32)
+declare <vscale x 2 x i16> @llvm.vp.merge.nxv2i16(<vscale x 2 x i1>, <vscale x 2 x i16>, <vscale x 2 x i16>, i32)
+declare <vscale x 2 x i16> @llvm.vp.select.nxv2i16(<vscale x 2 x i1>, <vscale x 2 x i16>, <vscale x 2 x i16>, i32)
+
+define <vscale x 2 x i16> @vmadd_vv_nxv2i16(<vscale x 2 x i16> %a, <vscale x 2 x i16> %b, <vscale x 2 x i16> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i16> @llvm.vp.mul.nxv2i16(<vscale x 2 x i16> %a, <vscale x 2 x i16> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i16> @llvm.vp.add.nxv2i16(<vscale x 2 x i16> %x, <vscale x 2 x i16> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i16> @llvm.vp.merge.nxv2i16(<vscale x 2 x i1> %m, <vscale x 2 x i16> %y, <vscale x 2 x i16> %a, i32 %evl)
+ ret <vscale x 2 x i16> %u
+}
+
+define <vscale x 2 x i16> @vmadd_vv_nxv2i16_unmasked(<vscale x 2 x i16> %a, <vscale x 2 x i16> %b, <vscale x 2 x i16> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, mf2, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i16> @llvm.vp.mul.nxv2i16(<vscale x 2 x i16> %a, <vscale x 2 x i16> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i16> @llvm.vp.add.nxv2i16(<vscale x 2 x i16> %x, <vscale x 2 x i16> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i16> @llvm.vp.merge.nxv2i16(<vscale x 2 x i1> %allones, <vscale x 2 x i16> %y, <vscale x 2 x i16> %a, i32 %evl)
+ ret <vscale x 2 x i16> %u
+}
+
+define <vscale x 2 x i16> @vmadd_vx_nxv2i16(<vscale x 2 x i16> %a, i16 %b, <vscale x 2 x i16> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv2i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v9, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i16> %elt.head, <vscale x 2 x i16> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i16> @llvm.vp.mul.nxv2i16(<vscale x 2 x i16> %a, <vscale x 2 x i16> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i16> @llvm.vp.add.nxv2i16(<vscale x 2 x i16> %x, <vscale x 2 x i16> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i16> @llvm.vp.merge.nxv2i16(<vscale x 2 x i1> %m, <vscale x 2 x i16> %y, <vscale x 2 x i16> %a, i32 %evl)
+ ret <vscale x 2 x i16> %u
+}
+
+define <vscale x 2 x i16> @vmadd_vx_nxv2i16_unmasked(<vscale x 2 x i16> %a, i16 %b, <vscale x 2 x i16> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv2i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v9
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i16> %elt.head, <vscale x 2 x i16> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i16> @llvm.vp.mul.nxv2i16(<vscale x 2 x i16> %a, <vscale x 2 x i16> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i16> @llvm.vp.add.nxv2i16(<vscale x 2 x i16> %x, <vscale x 2 x i16> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i16> @llvm.vp.merge.nxv2i16(<vscale x 2 x i1> %allones, <vscale x 2 x i16> %y, <vscale x 2 x i16> %a, i32 %evl)
+ ret <vscale x 2 x i16> %u
+}
+
+define <vscale x 2 x i16> @vmadd_vv_nxv2i16_ta(<vscale x 2 x i16> %a, <vscale x 2 x i16> %b, <vscale x 2 x i16> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, mf2, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i16> @llvm.vp.mul.nxv2i16(<vscale x 2 x i16> %a, <vscale x 2 x i16> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i16> @llvm.vp.add.nxv2i16(<vscale x 2 x i16> %x, <vscale x 2 x i16> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i16> @llvm.vp.select.nxv2i16(<vscale x 2 x i1> %m, <vscale x 2 x i16> %y, <vscale x 2 x i16> %a, i32 %evl)
+ ret <vscale x 2 x i16> %u
+}
+
+define <vscale x 2 x i16> @vmadd_vx_nxv2i16_ta(<vscale x 2 x i16> %a, i16 %b, <vscale x 2 x i16> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv2i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, mf2, ta, ma
+; CHECK-NEXT: vmacc.vx v9, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i16> %elt.head, <vscale x 2 x i16> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i16> @llvm.vp.mul.nxv2i16(<vscale x 2 x i16> %a, <vscale x 2 x i16> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i16> @llvm.vp.add.nxv2i16(<vscale x 2 x i16> %x, <vscale x 2 x i16> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i16> @llvm.vp.select.nxv2i16(<vscale x 2 x i1> %m, <vscale x 2 x i16> %y, <vscale x 2 x i16> %a, i32 %evl)
+ ret <vscale x 2 x i16> %u
+}
+
+declare <vscale x 4 x i16> @llvm.vp.mul.nxv4i16(<vscale x 4 x i16>, <vscale x 4 x i16>, <vscale x 4 x i1>, i32)
+declare <vscale x 4 x i16> @llvm.vp.add.nxv4i16(<vscale x 4 x i16>, <vscale x 4 x i16>, <vscale x 4 x i1>, i32)
+declare <vscale x 4 x i16> @llvm.vp.merge.nxv4i16(<vscale x 4 x i1>, <vscale x 4 x i16>, <vscale x 4 x i16>, i32)
+declare <vscale x 4 x i16> @llvm.vp.select.nxv4i16(<vscale x 4 x i1>, <vscale x 4 x i16>, <vscale x 4 x i16>, i32)
+
+define <vscale x 4 x i16> @vmadd_vv_nxv4i16(<vscale x 4 x i16> %a, <vscale x 4 x i16> %b, <vscale x 4 x i16> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i16> @llvm.vp.mul.nxv4i16(<vscale x 4 x i16> %a, <vscale x 4 x i16> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i16> @llvm.vp.add.nxv4i16(<vscale x 4 x i16> %x, <vscale x 4 x i16> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i16> @llvm.vp.merge.nxv4i16(<vscale x 4 x i1> %m, <vscale x 4 x i16> %y, <vscale x 4 x i16> %a, i32 %evl)
+ ret <vscale x 4 x i16> %u
+}
+
+define <vscale x 4 x i16> @vmadd_vv_nxv4i16_unmasked(<vscale x 4 x i16> %a, <vscale x 4 x i16> %b, <vscale x 4 x i16> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e16, m1, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i16> @llvm.vp.mul.nxv4i16(<vscale x 4 x i16> %a, <vscale x 4 x i16> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i16> @llvm.vp.add.nxv4i16(<vscale x 4 x i16> %x, <vscale x 4 x i16> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i16> @llvm.vp.merge.nxv4i16(<vscale x 4 x i1> %allones, <vscale x 4 x i16> %y, <vscale x 4 x i16> %a, i32 %evl)
+ ret <vscale x 4 x i16> %u
+}
+
+define <vscale x 4 x i16> @vmadd_vx_nxv4i16(<vscale x 4 x i16> %a, i16 %b, <vscale x 4 x i16> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv4i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v9, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i16> %elt.head, <vscale x 4 x i16> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i16> @llvm.vp.mul.nxv4i16(<vscale x 4 x i16> %a, <vscale x 4 x i16> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i16> @llvm.vp.add.nxv4i16(<vscale x 4 x i16> %x, <vscale x 4 x i16> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i16> @llvm.vp.merge.nxv4i16(<vscale x 4 x i1> %m, <vscale x 4 x i16> %y, <vscale x 4 x i16> %a, i32 %evl)
+ ret <vscale x 4 x i16> %u
+}
+
+define <vscale x 4 x i16> @vmadd_vx_nxv4i16_unmasked(<vscale x 4 x i16> %a, i16 %b, <vscale x 4 x i16> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv4i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v9
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i16> %elt.head, <vscale x 4 x i16> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i16> @llvm.vp.mul.nxv4i16(<vscale x 4 x i16> %a, <vscale x 4 x i16> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i16> @llvm.vp.add.nxv4i16(<vscale x 4 x i16> %x, <vscale x 4 x i16> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i16> @llvm.vp.merge.nxv4i16(<vscale x 4 x i1> %allones, <vscale x 4 x i16> %y, <vscale x 4 x i16> %a, i32 %evl)
+ ret <vscale x 4 x i16> %u
+}
+
+define <vscale x 4 x i16> @vmadd_vv_nxv4i16_ta(<vscale x 4 x i16> %a, <vscale x 4 x i16> %b, <vscale x 4 x i16> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i16> @llvm.vp.mul.nxv4i16(<vscale x 4 x i16> %a, <vscale x 4 x i16> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i16> @llvm.vp.add.nxv4i16(<vscale x 4 x i16> %x, <vscale x 4 x i16> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i16> @llvm.vp.select.nxv4i16(<vscale x 4 x i1> %m, <vscale x 4 x i16> %y, <vscale x 4 x i16> %a, i32 %evl)
+ ret <vscale x 4 x i16> %u
+}
+
+define <vscale x 4 x i16> @vmadd_vx_nxv4i16_ta(<vscale x 4 x i16> %a, i16 %b, <vscale x 4 x i16> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv4i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m1, ta, ma
+; CHECK-NEXT: vmacc.vx v9, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i16> %elt.head, <vscale x 4 x i16> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i16> @llvm.vp.mul.nxv4i16(<vscale x 4 x i16> %a, <vscale x 4 x i16> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i16> @llvm.vp.add.nxv4i16(<vscale x 4 x i16> %x, <vscale x 4 x i16> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i16> @llvm.vp.select.nxv4i16(<vscale x 4 x i1> %m, <vscale x 4 x i16> %y, <vscale x 4 x i16> %a, i32 %evl)
+ ret <vscale x 4 x i16> %u
+}
+
+declare <vscale x 8 x i16> @llvm.vp.mul.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i1>, i32)
+declare <vscale x 8 x i16> @llvm.vp.add.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i1>, i32)
+declare <vscale x 8 x i16> @llvm.vp.merge.nxv8i16(<vscale x 8 x i1>, <vscale x 8 x i16>, <vscale x 8 x i16>, i32)
+declare <vscale x 8 x i16> @llvm.vp.select.nxv8i16(<vscale x 8 x i1>, <vscale x 8 x i16>, <vscale x 8 x i16>, i32)
+
+define <vscale x 8 x i16> @vmadd_vv_nxv8i16(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b, <vscale x 8 x i16> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i16> @llvm.vp.mul.nxv8i16(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i16> @llvm.vp.add.nxv8i16(<vscale x 8 x i16> %x, <vscale x 8 x i16> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i16> @llvm.vp.merge.nxv8i16(<vscale x 8 x i1> %m, <vscale x 8 x i16> %y, <vscale x 8 x i16> %a, i32 %evl)
+ ret <vscale x 8 x i16> %u
+}
+
+define <vscale x 8 x i16> @vmadd_vv_nxv8i16_unmasked(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b, <vscale x 8 x i16> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vsetvli zero, zero, e16, m2, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v10
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i16> @llvm.vp.mul.nxv8i16(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i16> @llvm.vp.add.nxv8i16(<vscale x 8 x i16> %x, <vscale x 8 x i16> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i16> @llvm.vp.merge.nxv8i16(<vscale x 8 x i1> %allones, <vscale x 8 x i16> %y, <vscale x 8 x i16> %a, i32 %evl)
+ ret <vscale x 8 x i16> %u
+}
+
+define <vscale x 8 x i16> @vmadd_vx_nxv8i16(<vscale x 8 x i16> %a, i16 %b, <vscale x 8 x i16> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv8i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v10, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i16> %elt.head, <vscale x 8 x i16> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i16> @llvm.vp.mul.nxv8i16(<vscale x 8 x i16> %a, <vscale x 8 x i16> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i16> @llvm.vp.add.nxv8i16(<vscale x 8 x i16> %x, <vscale x 8 x i16> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i16> @llvm.vp.merge.nxv8i16(<vscale x 8 x i1> %m, <vscale x 8 x i16> %y, <vscale x 8 x i16> %a, i32 %evl)
+ ret <vscale x 8 x i16> %u
+}
+
+define <vscale x 8 x i16> @vmadd_vx_nxv8i16_unmasked(<vscale x 8 x i16> %a, i16 %b, <vscale x 8 x i16> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv8i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v10
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i16> %elt.head, <vscale x 8 x i16> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i16> @llvm.vp.mul.nxv8i16(<vscale x 8 x i16> %a, <vscale x 8 x i16> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i16> @llvm.vp.add.nxv8i16(<vscale x 8 x i16> %x, <vscale x 8 x i16> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i16> @llvm.vp.merge.nxv8i16(<vscale x 8 x i1> %allones, <vscale x 8 x i16> %y, <vscale x 8 x i16> %a, i32 %evl)
+ ret <vscale x 8 x i16> %u
+}
+
+define <vscale x 8 x i16> @vmadd_vv_nxv8i16_ta(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b, <vscale x 8 x i16> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i16> @llvm.vp.mul.nxv8i16(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i16> @llvm.vp.add.nxv8i16(<vscale x 8 x i16> %x, <vscale x 8 x i16> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i16> @llvm.vp.select.nxv8i16(<vscale x 8 x i1> %m, <vscale x 8 x i16> %y, <vscale x 8 x i16> %a, i32 %evl)
+ ret <vscale x 8 x i16> %u
+}
+
+define <vscale x 8 x i16> @vmadd_vx_nxv8i16_ta(<vscale x 8 x i16> %a, i16 %b, <vscale x 8 x i16> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv8i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m2, ta, ma
+; CHECK-NEXT: vmacc.vx v10, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i16> %elt.head, <vscale x 8 x i16> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i16> @llvm.vp.mul.nxv8i16(<vscale x 8 x i16> %a, <vscale x 8 x i16> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i16> @llvm.vp.add.nxv8i16(<vscale x 8 x i16> %x, <vscale x 8 x i16> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i16> @llvm.vp.select.nxv8i16(<vscale x 8 x i1> %m, <vscale x 8 x i16> %y, <vscale x 8 x i16> %a, i32 %evl)
+ ret <vscale x 8 x i16> %u
+}
+
+declare <vscale x 16 x i16> @llvm.vp.mul.nxv16i16(<vscale x 16 x i16>, <vscale x 16 x i16>, <vscale x 16 x i1>, i32)
+declare <vscale x 16 x i16> @llvm.vp.add.nxv16i16(<vscale x 16 x i16>, <vscale x 16 x i16>, <vscale x 16 x i1>, i32)
+declare <vscale x 16 x i16> @llvm.vp.merge.nxv16i16(<vscale x 16 x i1>, <vscale x 16 x i16>, <vscale x 16 x i16>, i32)
+declare <vscale x 16 x i16> @llvm.vp.select.nxv16i16(<vscale x 16 x i1>, <vscale x 16 x i16>, <vscale x 16 x i16>, i32)
+
+define <vscale x 16 x i16> @vmadd_vv_nxv16i16(<vscale x 16 x i16> %a, <vscale x 16 x i16> %b, <vscale x 16 x i16> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv16i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e16, m4, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i16> @llvm.vp.mul.nxv16i16(<vscale x 16 x i16> %a, <vscale x 16 x i16> %b, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i16> @llvm.vp.add.nxv16i16(<vscale x 16 x i16> %x, <vscale x 16 x i16> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i16> @llvm.vp.merge.nxv16i16(<vscale x 16 x i1> %m, <vscale x 16 x i16> %y, <vscale x 16 x i16> %a, i32 %evl)
+ ret <vscale x 16 x i16> %u
+}
+
+define <vscale x 16 x i16> @vmadd_vv_nxv16i16_unmasked(<vscale x 16 x i16> %a, <vscale x 16 x i16> %b, <vscale x 16 x i16> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv16i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e16, m4, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v12
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i16> @llvm.vp.mul.nxv16i16(<vscale x 16 x i16> %a, <vscale x 16 x i16> %b, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i16> @llvm.vp.add.nxv16i16(<vscale x 16 x i16> %x, <vscale x 16 x i16> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i16> @llvm.vp.merge.nxv16i16(<vscale x 16 x i1> %allones, <vscale x 16 x i16> %y, <vscale x 16 x i16> %a, i32 %evl)
+ ret <vscale x 16 x i16> %u
+}
+
+define <vscale x 16 x i16> @vmadd_vx_nxv16i16(<vscale x 16 x i16> %a, i16 %b, <vscale x 16 x i16> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv16i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v12, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 16 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 16 x i16> %elt.head, <vscale x 16 x i16> poison, <vscale x 16 x i32> zeroinitializer
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i16> @llvm.vp.mul.nxv16i16(<vscale x 16 x i16> %a, <vscale x 16 x i16> %vb, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i16> @llvm.vp.add.nxv16i16(<vscale x 16 x i16> %x, <vscale x 16 x i16> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i16> @llvm.vp.merge.nxv16i16(<vscale x 16 x i1> %m, <vscale x 16 x i16> %y, <vscale x 16 x i16> %a, i32 %evl)
+ ret <vscale x 16 x i16> %u
+}
+
+define <vscale x 16 x i16> @vmadd_vx_nxv16i16_unmasked(<vscale x 16 x i16> %a, i16 %b, <vscale x 16 x i16> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv16i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v12
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 16 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 16 x i16> %elt.head, <vscale x 16 x i16> poison, <vscale x 16 x i32> zeroinitializer
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i16> @llvm.vp.mul.nxv16i16(<vscale x 16 x i16> %a, <vscale x 16 x i16> %vb, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i16> @llvm.vp.add.nxv16i16(<vscale x 16 x i16> %x, <vscale x 16 x i16> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i16> @llvm.vp.merge.nxv16i16(<vscale x 16 x i1> %allones, <vscale x 16 x i16> %y, <vscale x 16 x i16> %a, i32 %evl)
+ ret <vscale x 16 x i16> %u
+}
+
+define <vscale x 16 x i16> @vmadd_vv_nxv16i16_ta(<vscale x 16 x i16> %a, <vscale x 16 x i16> %b, <vscale x 16 x i16> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv16i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e16, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i16> @llvm.vp.mul.nxv16i16(<vscale x 16 x i16> %a, <vscale x 16 x i16> %b, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i16> @llvm.vp.add.nxv16i16(<vscale x 16 x i16> %x, <vscale x 16 x i16> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i16> @llvm.vp.select.nxv16i16(<vscale x 16 x i1> %m, <vscale x 16 x i16> %y, <vscale x 16 x i16> %a, i32 %evl)
+ ret <vscale x 16 x i16> %u
+}
+
+define <vscale x 16 x i16> @vmadd_vx_nxv16i16_ta(<vscale x 16 x i16> %a, i16 %b, <vscale x 16 x i16> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv16i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m4, ta, ma
+; CHECK-NEXT: vmacc.vx v12, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 16 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 16 x i16> %elt.head, <vscale x 16 x i16> poison, <vscale x 16 x i32> zeroinitializer
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i16> @llvm.vp.mul.nxv16i16(<vscale x 16 x i16> %a, <vscale x 16 x i16> %vb, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i16> @llvm.vp.add.nxv16i16(<vscale x 16 x i16> %x, <vscale x 16 x i16> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i16> @llvm.vp.select.nxv16i16(<vscale x 16 x i1> %m, <vscale x 16 x i16> %y, <vscale x 16 x i16> %a, i32 %evl)
+ ret <vscale x 16 x i16> %u
+}
+
+declare <vscale x 32 x i16> @llvm.vp.mul.nxv32i16(<vscale x 32 x i16>, <vscale x 32 x i16>, <vscale x 32 x i1>, i32)
+declare <vscale x 32 x i16> @llvm.vp.add.nxv32i16(<vscale x 32 x i16>, <vscale x 32 x i16>, <vscale x 32 x i1>, i32)
+declare <vscale x 32 x i16> @llvm.vp.merge.nxv32i16(<vscale x 32 x i1>, <vscale x 32 x i16>, <vscale x 32 x i16>, i32)
+declare <vscale x 32 x i16> @llvm.vp.select.nxv32i16(<vscale x 32 x i1>, <vscale x 32 x i16>, <vscale x 32 x i16>, i32)
+
+define <vscale x 32 x i16> @vmadd_vv_nxv32i16(<vscale x 32 x i16> %a, <vscale x 32 x i16> %b, <vscale x 32 x i16> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv32i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8re16.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e16, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e16, m8, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i16> @llvm.vp.mul.nxv32i16(<vscale x 32 x i16> %a, <vscale x 32 x i16> %b, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i16> @llvm.vp.add.nxv32i16(<vscale x 32 x i16> %x, <vscale x 32 x i16> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i16> @llvm.vp.merge.nxv32i16(<vscale x 32 x i1> %m, <vscale x 32 x i16> %y, <vscale x 32 x i16> %a, i32 %evl)
+ ret <vscale x 32 x i16> %u
+}
+
+define <vscale x 32 x i16> @vmadd_vv_nxv32i16_unmasked(<vscale x 32 x i16> %a, <vscale x 32 x i16> %b, <vscale x 32 x i16> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv32i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8re16.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e16, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e16, m8, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v24
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i16> @llvm.vp.mul.nxv32i16(<vscale x 32 x i16> %a, <vscale x 32 x i16> %b, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i16> @llvm.vp.add.nxv32i16(<vscale x 32 x i16> %x, <vscale x 32 x i16> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i16> @llvm.vp.merge.nxv32i16(<vscale x 32 x i1> %allones, <vscale x 32 x i16> %y, <vscale x 32 x i16> %a, i32 %evl)
+ ret <vscale x 32 x i16> %u
+}
+
+define <vscale x 32 x i16> @vmadd_vx_nxv32i16(<vscale x 32 x i16> %a, i16 %b, <vscale x 32 x i16> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv32i16:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m8, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v16, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 32 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 32 x i16> %elt.head, <vscale x 32 x i16> poison, <vscale x 32 x i32> zeroinitializer
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i16> @llvm.vp.mul.nxv32i16(<vscale x 32 x i16> %a, <vscale x 32 x i16> %vb, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i16> @llvm.vp.add.nxv32i16(<vscale x 32 x i16> %x, <vscale x 32 x i16> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i16> @llvm.vp.merge.nxv32i16(<vscale x 32 x i1> %m, <vscale x 32 x i16> %y, <vscale x 32 x i16> %a, i32 %evl)
+ ret <vscale x 32 x i16> %u
+}
+
+define <vscale x 32 x i16> @vmadd_vx_nxv32i16_unmasked(<vscale x 32 x i16> %a, i16 %b, <vscale x 32 x i16> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv32i16_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m8, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v16
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 32 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 32 x i16> %elt.head, <vscale x 32 x i16> poison, <vscale x 32 x i32> zeroinitializer
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i16> @llvm.vp.mul.nxv32i16(<vscale x 32 x i16> %a, <vscale x 32 x i16> %vb, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i16> @llvm.vp.add.nxv32i16(<vscale x 32 x i16> %x, <vscale x 32 x i16> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i16> @llvm.vp.merge.nxv32i16(<vscale x 32 x i1> %allones, <vscale x 32 x i16> %y, <vscale x 32 x i16> %a, i32 %evl)
+ ret <vscale x 32 x i16> %u
+}
+
+define <vscale x 32 x i16> @vmadd_vv_nxv32i16_ta(<vscale x 32 x i16> %a, <vscale x 32 x i16> %b, <vscale x 32 x i16> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv32i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8re16.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e16, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i16> @llvm.vp.mul.nxv32i16(<vscale x 32 x i16> %a, <vscale x 32 x i16> %b, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i16> @llvm.vp.add.nxv32i16(<vscale x 32 x i16> %x, <vscale x 32 x i16> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i16> @llvm.vp.select.nxv32i16(<vscale x 32 x i1> %m, <vscale x 32 x i16> %y, <vscale x 32 x i16> %a, i32 %evl)
+ ret <vscale x 32 x i16> %u
+}
+
+define <vscale x 32 x i16> @vmadd_vx_nxv32i16_ta(<vscale x 32 x i16> %a, i16 %b, <vscale x 32 x i16> %c, <vscale x 32 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv32i16_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e16, m8, ta, ma
+; CHECK-NEXT: vmacc.vx v16, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 32 x i16> poison, i16 %b, i32 0
+ %vb = shufflevector <vscale x 32 x i16> %elt.head, <vscale x 32 x i16> poison, <vscale x 32 x i32> zeroinitializer
+ %splat = insertelement <vscale x 32 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 32 x i1> %splat, <vscale x 32 x i1> poison, <vscale x 32 x i32> zeroinitializer
+ %x = call <vscale x 32 x i16> @llvm.vp.mul.nxv32i16(<vscale x 32 x i16> %a, <vscale x 32 x i16> %vb, <vscale x 32 x i1> %allones, i32 %evl)
+ %y = call <vscale x 32 x i16> @llvm.vp.add.nxv32i16(<vscale x 32 x i16> %x, <vscale x 32 x i16> %c, <vscale x 32 x i1> %allones, i32 %evl)
+ %u = call <vscale x 32 x i16> @llvm.vp.select.nxv32i16(<vscale x 32 x i1> %m, <vscale x 32 x i16> %y, <vscale x 32 x i16> %a, i32 %evl)
+ ret <vscale x 32 x i16> %u
+}
+
+declare <vscale x 1 x i32> @llvm.vp.mul.nxv1i32(<vscale x 1 x i32>, <vscale x 1 x i32>, <vscale x 1 x i1>, i32)
+declare <vscale x 1 x i32> @llvm.vp.add.nxv1i32(<vscale x 1 x i32>, <vscale x 1 x i32>, <vscale x 1 x i1>, i32)
+declare <vscale x 1 x i32> @llvm.vp.merge.nxv1i32(<vscale x 1 x i1>, <vscale x 1 x i32>, <vscale x 1 x i32>, i32)
+declare <vscale x 1 x i32> @llvm.vp.select.nxv1i32(<vscale x 1 x i1>, <vscale x 1 x i32>, <vscale x 1 x i32>, i32)
+
+define <vscale x 1 x i32> @vmadd_vv_nxv1i32(<vscale x 1 x i32> %a, <vscale x 1 x i32> %b, <vscale x 1 x i32> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i32> @llvm.vp.mul.nxv1i32(<vscale x 1 x i32> %a, <vscale x 1 x i32> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i32> @llvm.vp.add.nxv1i32(<vscale x 1 x i32> %x, <vscale x 1 x i32> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i32> @llvm.vp.merge.nxv1i32(<vscale x 1 x i1> %m, <vscale x 1 x i32> %y, <vscale x 1 x i32> %a, i32 %evl)
+ ret <vscale x 1 x i32> %u
+}
+
+define <vscale x 1 x i32> @vmadd_vv_nxv1i32_unmasked(<vscale x 1 x i32> %a, <vscale x 1 x i32> %b, <vscale x 1 x i32> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, mf2, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i32> @llvm.vp.mul.nxv1i32(<vscale x 1 x i32> %a, <vscale x 1 x i32> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i32> @llvm.vp.add.nxv1i32(<vscale x 1 x i32> %x, <vscale x 1 x i32> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i32> @llvm.vp.merge.nxv1i32(<vscale x 1 x i1> %allones, <vscale x 1 x i32> %y, <vscale x 1 x i32> %a, i32 %evl)
+ ret <vscale x 1 x i32> %u
+}
+
+define <vscale x 1 x i32> @vmadd_vx_nxv1i32(<vscale x 1 x i32> %a, i32 %b, <vscale x 1 x i32> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv1i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v9, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i32> %elt.head, <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i32> @llvm.vp.mul.nxv1i32(<vscale x 1 x i32> %a, <vscale x 1 x i32> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i32> @llvm.vp.add.nxv1i32(<vscale x 1 x i32> %x, <vscale x 1 x i32> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i32> @llvm.vp.merge.nxv1i32(<vscale x 1 x i1> %m, <vscale x 1 x i32> %y, <vscale x 1 x i32> %a, i32 %evl)
+ ret <vscale x 1 x i32> %u
+}
+
+define <vscale x 1 x i32> @vmadd_vx_nxv1i32_unmasked(<vscale x 1 x i32> %a, i32 %b, <vscale x 1 x i32> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv1i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v9
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i32> %elt.head, <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i32> @llvm.vp.mul.nxv1i32(<vscale x 1 x i32> %a, <vscale x 1 x i32> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i32> @llvm.vp.add.nxv1i32(<vscale x 1 x i32> %x, <vscale x 1 x i32> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i32> @llvm.vp.merge.nxv1i32(<vscale x 1 x i1> %allones, <vscale x 1 x i32> %y, <vscale x 1 x i32> %a, i32 %evl)
+ ret <vscale x 1 x i32> %u
+}
+
+define <vscale x 1 x i32> @vmadd_vv_nxv1i32_ta(<vscale x 1 x i32> %a, <vscale x 1 x i32> %b, <vscale x 1 x i32> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, mf2, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i32> @llvm.vp.mul.nxv1i32(<vscale x 1 x i32> %a, <vscale x 1 x i32> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i32> @llvm.vp.add.nxv1i32(<vscale x 1 x i32> %x, <vscale x 1 x i32> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i32> @llvm.vp.select.nxv1i32(<vscale x 1 x i1> %m, <vscale x 1 x i32> %y, <vscale x 1 x i32> %a, i32 %evl)
+ ret <vscale x 1 x i32> %u
+}
+
+define <vscale x 1 x i32> @vmadd_vx_nxv1i32_ta(<vscale x 1 x i32> %a, i32 %b, <vscale x 1 x i32> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv1i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, mf2, ta, ma
+; CHECK-NEXT: vmacc.vx v9, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i32> %elt.head, <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i32> @llvm.vp.mul.nxv1i32(<vscale x 1 x i32> %a, <vscale x 1 x i32> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i32> @llvm.vp.add.nxv1i32(<vscale x 1 x i32> %x, <vscale x 1 x i32> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i32> @llvm.vp.select.nxv1i32(<vscale x 1 x i1> %m, <vscale x 1 x i32> %y, <vscale x 1 x i32> %a, i32 %evl)
+ ret <vscale x 1 x i32> %u
+}
+
+declare <vscale x 2 x i32> @llvm.vp.mul.nxv2i32(<vscale x 2 x i32>, <vscale x 2 x i32>, <vscale x 2 x i1>, i32)
+declare <vscale x 2 x i32> @llvm.vp.add.nxv2i32(<vscale x 2 x i32>, <vscale x 2 x i32>, <vscale x 2 x i1>, i32)
+declare <vscale x 2 x i32> @llvm.vp.merge.nxv2i32(<vscale x 2 x i1>, <vscale x 2 x i32>, <vscale x 2 x i32>, i32)
+declare <vscale x 2 x i32> @llvm.vp.select.nxv2i32(<vscale x 2 x i1>, <vscale x 2 x i32>, <vscale x 2 x i32>, i32)
+
+define <vscale x 2 x i32> @vmadd_vv_nxv2i32(<vscale x 2 x i32> %a, <vscale x 2 x i32> %b, <vscale x 2 x i32> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m1, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i32> @llvm.vp.mul.nxv2i32(<vscale x 2 x i32> %a, <vscale x 2 x i32> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i32> @llvm.vp.add.nxv2i32(<vscale x 2 x i32> %x, <vscale x 2 x i32> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i32> @llvm.vp.merge.nxv2i32(<vscale x 2 x i1> %m, <vscale x 2 x i32> %y, <vscale x 2 x i32> %a, i32 %evl)
+ ret <vscale x 2 x i32> %u
+}
+
+define <vscale x 2 x i32> @vmadd_vv_nxv2i32_unmasked(<vscale x 2 x i32> %a, <vscale x 2 x i32> %b, <vscale x 2 x i32> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e32, m1, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i32> @llvm.vp.mul.nxv2i32(<vscale x 2 x i32> %a, <vscale x 2 x i32> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i32> @llvm.vp.add.nxv2i32(<vscale x 2 x i32> %x, <vscale x 2 x i32> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i32> @llvm.vp.merge.nxv2i32(<vscale x 2 x i1> %allones, <vscale x 2 x i32> %y, <vscale x 2 x i32> %a, i32 %evl)
+ ret <vscale x 2 x i32> %u
+}
+
+define <vscale x 2 x i32> @vmadd_vx_nxv2i32(<vscale x 2 x i32> %a, i32 %b, <vscale x 2 x i32> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv2i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v9, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i32> %elt.head, <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i32> @llvm.vp.mul.nxv2i32(<vscale x 2 x i32> %a, <vscale x 2 x i32> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i32> @llvm.vp.add.nxv2i32(<vscale x 2 x i32> %x, <vscale x 2 x i32> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i32> @llvm.vp.merge.nxv2i32(<vscale x 2 x i1> %m, <vscale x 2 x i32> %y, <vscale x 2 x i32> %a, i32 %evl)
+ ret <vscale x 2 x i32> %u
+}
+
+define <vscale x 2 x i32> @vmadd_vx_nxv2i32_unmasked(<vscale x 2 x i32> %a, i32 %b, <vscale x 2 x i32> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv2i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v9
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i32> %elt.head, <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i32> @llvm.vp.mul.nxv2i32(<vscale x 2 x i32> %a, <vscale x 2 x i32> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i32> @llvm.vp.add.nxv2i32(<vscale x 2 x i32> %x, <vscale x 2 x i32> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i32> @llvm.vp.merge.nxv2i32(<vscale x 2 x i1> %allones, <vscale x 2 x i32> %y, <vscale x 2 x i32> %a, i32 %evl)
+ ret <vscale x 2 x i32> %u
+}
+
+define <vscale x 2 x i32> @vmadd_vv_nxv2i32_ta(<vscale x 2 x i32> %a, <vscale x 2 x i32> %b, <vscale x 2 x i32> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i32> @llvm.vp.mul.nxv2i32(<vscale x 2 x i32> %a, <vscale x 2 x i32> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i32> @llvm.vp.add.nxv2i32(<vscale x 2 x i32> %x, <vscale x 2 x i32> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i32> @llvm.vp.select.nxv2i32(<vscale x 2 x i1> %m, <vscale x 2 x i32> %y, <vscale x 2 x i32> %a, i32 %evl)
+ ret <vscale x 2 x i32> %u
+}
+
+define <vscale x 2 x i32> @vmadd_vx_nxv2i32_ta(<vscale x 2 x i32> %a, i32 %b, <vscale x 2 x i32> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv2i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m1, ta, ma
+; CHECK-NEXT: vmacc.vx v9, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i32> %elt.head, <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i32> @llvm.vp.mul.nxv2i32(<vscale x 2 x i32> %a, <vscale x 2 x i32> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i32> @llvm.vp.add.nxv2i32(<vscale x 2 x i32> %x, <vscale x 2 x i32> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i32> @llvm.vp.select.nxv2i32(<vscale x 2 x i1> %m, <vscale x 2 x i32> %y, <vscale x 2 x i32> %a, i32 %evl)
+ ret <vscale x 2 x i32> %u
+}
+
+declare <vscale x 4 x i32> @llvm.vp.mul.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i1>, i32)
+declare <vscale x 4 x i32> @llvm.vp.add.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i1>, i32)
+declare <vscale x 4 x i32> @llvm.vp.merge.nxv4i32(<vscale x 4 x i1>, <vscale x 4 x i32>, <vscale x 4 x i32>, i32)
+declare <vscale x 4 x i32> @llvm.vp.select.nxv4i32(<vscale x 4 x i1>, <vscale x 4 x i32>, <vscale x 4 x i32>, i32)
+
+define <vscale x 4 x i32> @vmadd_vv_nxv4i32(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b, <vscale x 4 x i32> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i32> @llvm.vp.mul.nxv4i32(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i32> @llvm.vp.add.nxv4i32(<vscale x 4 x i32> %x, <vscale x 4 x i32> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i32> @llvm.vp.merge.nxv4i32(<vscale x 4 x i1> %m, <vscale x 4 x i32> %y, <vscale x 4 x i32> %a, i32 %evl)
+ ret <vscale x 4 x i32> %u
+}
+
+define <vscale x 4 x i32> @vmadd_vv_nxv4i32_unmasked(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b, <vscale x 4 x i32> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vsetvli zero, zero, e32, m2, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v10
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i32> @llvm.vp.mul.nxv4i32(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i32> @llvm.vp.add.nxv4i32(<vscale x 4 x i32> %x, <vscale x 4 x i32> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i32> @llvm.vp.merge.nxv4i32(<vscale x 4 x i1> %allones, <vscale x 4 x i32> %y, <vscale x 4 x i32> %a, i32 %evl)
+ ret <vscale x 4 x i32> %u
+}
+
+define <vscale x 4 x i32> @vmadd_vx_nxv4i32(<vscale x 4 x i32> %a, i32 %b, <vscale x 4 x i32> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv4i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v10, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i32> %elt.head, <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i32> @llvm.vp.mul.nxv4i32(<vscale x 4 x i32> %a, <vscale x 4 x i32> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i32> @llvm.vp.add.nxv4i32(<vscale x 4 x i32> %x, <vscale x 4 x i32> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i32> @llvm.vp.merge.nxv4i32(<vscale x 4 x i1> %m, <vscale x 4 x i32> %y, <vscale x 4 x i32> %a, i32 %evl)
+ ret <vscale x 4 x i32> %u
+}
+
+define <vscale x 4 x i32> @vmadd_vx_nxv4i32_unmasked(<vscale x 4 x i32> %a, i32 %b, <vscale x 4 x i32> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv4i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v10
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i32> %elt.head, <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i32> @llvm.vp.mul.nxv4i32(<vscale x 4 x i32> %a, <vscale x 4 x i32> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i32> @llvm.vp.add.nxv4i32(<vscale x 4 x i32> %x, <vscale x 4 x i32> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i32> @llvm.vp.merge.nxv4i32(<vscale x 4 x i1> %allones, <vscale x 4 x i32> %y, <vscale x 4 x i32> %a, i32 %evl)
+ ret <vscale x 4 x i32> %u
+}
+
+define <vscale x 4 x i32> @vmadd_vv_nxv4i32_ta(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b, <vscale x 4 x i32> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i32> @llvm.vp.mul.nxv4i32(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i32> @llvm.vp.add.nxv4i32(<vscale x 4 x i32> %x, <vscale x 4 x i32> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i32> @llvm.vp.select.nxv4i32(<vscale x 4 x i1> %m, <vscale x 4 x i32> %y, <vscale x 4 x i32> %a, i32 %evl)
+ ret <vscale x 4 x i32> %u
+}
+
+define <vscale x 4 x i32> @vmadd_vx_nxv4i32_ta(<vscale x 4 x i32> %a, i32 %b, <vscale x 4 x i32> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv4i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m2, ta, ma
+; CHECK-NEXT: vmacc.vx v10, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i32> %elt.head, <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i32> @llvm.vp.mul.nxv4i32(<vscale x 4 x i32> %a, <vscale x 4 x i32> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i32> @llvm.vp.add.nxv4i32(<vscale x 4 x i32> %x, <vscale x 4 x i32> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i32> @llvm.vp.select.nxv4i32(<vscale x 4 x i1> %m, <vscale x 4 x i32> %y, <vscale x 4 x i32> %a, i32 %evl)
+ ret <vscale x 4 x i32> %u
+}
+
+declare <vscale x 8 x i32> @llvm.vp.mul.nxv8i32(<vscale x 8 x i32>, <vscale x 8 x i32>, <vscale x 8 x i1>, i32)
+declare <vscale x 8 x i32> @llvm.vp.add.nxv8i32(<vscale x 8 x i32>, <vscale x 8 x i32>, <vscale x 8 x i1>, i32)
+declare <vscale x 8 x i32> @llvm.vp.merge.nxv8i32(<vscale x 8 x i1>, <vscale x 8 x i32>, <vscale x 8 x i32>, i32)
+declare <vscale x 8 x i32> @llvm.vp.select.nxv8i32(<vscale x 8 x i1>, <vscale x 8 x i32>, <vscale x 8 x i32>, i32)
+
+define <vscale x 8 x i32> @vmadd_vv_nxv8i32(<vscale x 8 x i32> %a, <vscale x 8 x i32> %b, <vscale x 8 x i32> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i32> @llvm.vp.mul.nxv8i32(<vscale x 8 x i32> %a, <vscale x 8 x i32> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i32> @llvm.vp.add.nxv8i32(<vscale x 8 x i32> %x, <vscale x 8 x i32> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i32> @llvm.vp.merge.nxv8i32(<vscale x 8 x i1> %m, <vscale x 8 x i32> %y, <vscale x 8 x i32> %a, i32 %evl)
+ ret <vscale x 8 x i32> %u
+}
+
+define <vscale x 8 x i32> @vmadd_vv_nxv8i32_unmasked(<vscale x 8 x i32> %a, <vscale x 8 x i32> %b, <vscale x 8 x i32> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m4, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v12
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i32> @llvm.vp.mul.nxv8i32(<vscale x 8 x i32> %a, <vscale x 8 x i32> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i32> @llvm.vp.add.nxv8i32(<vscale x 8 x i32> %x, <vscale x 8 x i32> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i32> @llvm.vp.merge.nxv8i32(<vscale x 8 x i1> %allones, <vscale x 8 x i32> %y, <vscale x 8 x i32> %a, i32 %evl)
+ ret <vscale x 8 x i32> %u
+}
+
+define <vscale x 8 x i32> @vmadd_vx_nxv8i32(<vscale x 8 x i32> %a, i32 %b, <vscale x 8 x i32> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv8i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v12, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i32> %elt.head, <vscale x 8 x i32> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i32> @llvm.vp.mul.nxv8i32(<vscale x 8 x i32> %a, <vscale x 8 x i32> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i32> @llvm.vp.add.nxv8i32(<vscale x 8 x i32> %x, <vscale x 8 x i32> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i32> @llvm.vp.merge.nxv8i32(<vscale x 8 x i1> %m, <vscale x 8 x i32> %y, <vscale x 8 x i32> %a, i32 %evl)
+ ret <vscale x 8 x i32> %u
+}
+
+define <vscale x 8 x i32> @vmadd_vx_nxv8i32_unmasked(<vscale x 8 x i32> %a, i32 %b, <vscale x 8 x i32> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv8i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v12
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i32> %elt.head, <vscale x 8 x i32> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i32> @llvm.vp.mul.nxv8i32(<vscale x 8 x i32> %a, <vscale x 8 x i32> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i32> @llvm.vp.add.nxv8i32(<vscale x 8 x i32> %x, <vscale x 8 x i32> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i32> @llvm.vp.merge.nxv8i32(<vscale x 8 x i1> %allones, <vscale x 8 x i32> %y, <vscale x 8 x i32> %a, i32 %evl)
+ ret <vscale x 8 x i32> %u
+}
+
+define <vscale x 8 x i32> @vmadd_vv_nxv8i32_ta(<vscale x 8 x i32> %a, <vscale x 8 x i32> %b, <vscale x 8 x i32> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e32, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i32> @llvm.vp.mul.nxv8i32(<vscale x 8 x i32> %a, <vscale x 8 x i32> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i32> @llvm.vp.add.nxv8i32(<vscale x 8 x i32> %x, <vscale x 8 x i32> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i32> @llvm.vp.select.nxv8i32(<vscale x 8 x i1> %m, <vscale x 8 x i32> %y, <vscale x 8 x i32> %a, i32 %evl)
+ ret <vscale x 8 x i32> %u
+}
+
+define <vscale x 8 x i32> @vmadd_vx_nxv8i32_ta(<vscale x 8 x i32> %a, i32 %b, <vscale x 8 x i32> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv8i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m4, ta, ma
+; CHECK-NEXT: vmacc.vx v12, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i32> %elt.head, <vscale x 8 x i32> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i32> @llvm.vp.mul.nxv8i32(<vscale x 8 x i32> %a, <vscale x 8 x i32> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i32> @llvm.vp.add.nxv8i32(<vscale x 8 x i32> %x, <vscale x 8 x i32> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i32> @llvm.vp.select.nxv8i32(<vscale x 8 x i1> %m, <vscale x 8 x i32> %y, <vscale x 8 x i32> %a, i32 %evl)
+ ret <vscale x 8 x i32> %u
+}
+
+declare <vscale x 16 x i32> @llvm.vp.mul.nxv16i32(<vscale x 16 x i32>, <vscale x 16 x i32>, <vscale x 16 x i1>, i32)
+declare <vscale x 16 x i32> @llvm.vp.add.nxv16i32(<vscale x 16 x i32>, <vscale x 16 x i32>, <vscale x 16 x i1>, i32)
+declare <vscale x 16 x i32> @llvm.vp.merge.nxv16i32(<vscale x 16 x i1>, <vscale x 16 x i32>, <vscale x 16 x i32>, i32)
+declare <vscale x 16 x i32> @llvm.vp.select.nxv16i32(<vscale x 16 x i1>, <vscale x 16 x i32>, <vscale x 16 x i32>, i32)
+
+define <vscale x 16 x i32> @vmadd_vv_nxv16i32(<vscale x 16 x i32> %a, <vscale x 16 x i32> %b, <vscale x 16 x i32> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv16i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8re32.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i32> @llvm.vp.mul.nxv16i32(<vscale x 16 x i32> %a, <vscale x 16 x i32> %b, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i32> @llvm.vp.add.nxv16i32(<vscale x 16 x i32> %x, <vscale x 16 x i32> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i32> @llvm.vp.merge.nxv16i32(<vscale x 16 x i1> %m, <vscale x 16 x i32> %y, <vscale x 16 x i32> %a, i32 %evl)
+ ret <vscale x 16 x i32> %u
+}
+
+define <vscale x 16 x i32> @vmadd_vv_nxv16i32_unmasked(<vscale x 16 x i32> %a, <vscale x 16 x i32> %b, <vscale x 16 x i32> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv16i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8re32.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e32, m8, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v24
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i32> @llvm.vp.mul.nxv16i32(<vscale x 16 x i32> %a, <vscale x 16 x i32> %b, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i32> @llvm.vp.add.nxv16i32(<vscale x 16 x i32> %x, <vscale x 16 x i32> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i32> @llvm.vp.merge.nxv16i32(<vscale x 16 x i1> %allones, <vscale x 16 x i32> %y, <vscale x 16 x i32> %a, i32 %evl)
+ ret <vscale x 16 x i32> %u
+}
+
+define <vscale x 16 x i32> @vmadd_vx_nxv16i32(<vscale x 16 x i32> %a, i32 %b, <vscale x 16 x i32> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv16i32:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, tu, mu
+; CHECK-NEXT: vmadd.vx v8, a0, v16, v0.t
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 16 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 16 x i32> %elt.head, <vscale x 16 x i32> poison, <vscale x 16 x i32> zeroinitializer
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i32> @llvm.vp.mul.nxv16i32(<vscale x 16 x i32> %a, <vscale x 16 x i32> %vb, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i32> @llvm.vp.add.nxv16i32(<vscale x 16 x i32> %x, <vscale x 16 x i32> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i32> @llvm.vp.merge.nxv16i32(<vscale x 16 x i1> %m, <vscale x 16 x i32> %y, <vscale x 16 x i32> %a, i32 %evl)
+ ret <vscale x 16 x i32> %u
+}
+
+define <vscale x 16 x i32> @vmadd_vx_nxv16i32_unmasked(<vscale x 16 x i32> %a, i32 %b, <vscale x 16 x i32> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv16i32_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, tu, ma
+; CHECK-NEXT: vmadd.vx v8, a0, v16
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 16 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 16 x i32> %elt.head, <vscale x 16 x i32> poison, <vscale x 16 x i32> zeroinitializer
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i32> @llvm.vp.mul.nxv16i32(<vscale x 16 x i32> %a, <vscale x 16 x i32> %vb, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i32> @llvm.vp.add.nxv16i32(<vscale x 16 x i32> %x, <vscale x 16 x i32> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i32> @llvm.vp.merge.nxv16i32(<vscale x 16 x i1> %allones, <vscale x 16 x i32> %y, <vscale x 16 x i32> %a, i32 %evl)
+ ret <vscale x 16 x i32> %u
+}
+
+define <vscale x 16 x i32> @vmadd_vv_nxv16i32_ta(<vscale x 16 x i32> %a, <vscale x 16 x i32> %b, <vscale x 16 x i32> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv16i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8re32.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i32> @llvm.vp.mul.nxv16i32(<vscale x 16 x i32> %a, <vscale x 16 x i32> %b, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i32> @llvm.vp.add.nxv16i32(<vscale x 16 x i32> %x, <vscale x 16 x i32> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i32> @llvm.vp.select.nxv16i32(<vscale x 16 x i1> %m, <vscale x 16 x i32> %y, <vscale x 16 x i32> %a, i32 %evl)
+ ret <vscale x 16 x i32> %u
+}
+
+define <vscale x 16 x i32> @vmadd_vx_nxv16i32_ta(<vscale x 16 x i32> %a, i32 %b, <vscale x 16 x i32> %c, <vscale x 16 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vx_nxv16i32_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a1, e32, m8, ta, ma
+; CHECK-NEXT: vmacc.vx v16, a0, v8
+; CHECK-NEXT: vmerge.vvm v8, v8, v16, v0
+; CHECK-NEXT: ret
+ %elt.head = insertelement <vscale x 16 x i32> poison, i32 %b, i32 0
+ %vb = shufflevector <vscale x 16 x i32> %elt.head, <vscale x 16 x i32> poison, <vscale x 16 x i32> zeroinitializer
+ %splat = insertelement <vscale x 16 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 16 x i1> %splat, <vscale x 16 x i1> poison, <vscale x 16 x i32> zeroinitializer
+ %x = call <vscale x 16 x i32> @llvm.vp.mul.nxv16i32(<vscale x 16 x i32> %a, <vscale x 16 x i32> %vb, <vscale x 16 x i1> %allones, i32 %evl)
+ %y = call <vscale x 16 x i32> @llvm.vp.add.nxv16i32(<vscale x 16 x i32> %x, <vscale x 16 x i32> %c, <vscale x 16 x i1> %allones, i32 %evl)
+ %u = call <vscale x 16 x i32> @llvm.vp.select.nxv16i32(<vscale x 16 x i1> %m, <vscale x 16 x i32> %y, <vscale x 16 x i32> %a, i32 %evl)
+ ret <vscale x 16 x i32> %u
+}
+
+declare <vscale x 1 x i64> @llvm.vp.mul.nxv1i64(<vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i1>, i32)
+declare <vscale x 1 x i64> @llvm.vp.add.nxv1i64(<vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i1>, i32)
+declare <vscale x 1 x i64> @llvm.vp.merge.nxv1i64(<vscale x 1 x i1>, <vscale x 1 x i64>, <vscale x 1 x i64>, i32)
+declare <vscale x 1 x i64> @llvm.vp.select.nxv1i64(<vscale x 1 x i1>, <vscale x 1 x i64>, <vscale x 1 x i64>, i32)
+
+define <vscale x 1 x i64> @vmadd_vv_nxv1i64(<vscale x 1 x i64> %a, <vscale x 1 x i64> %b, <vscale x 1 x i64> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i64> @llvm.vp.mul.nxv1i64(<vscale x 1 x i64> %a, <vscale x 1 x i64> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i64> @llvm.vp.add.nxv1i64(<vscale x 1 x i64> %x, <vscale x 1 x i64> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i64> @llvm.vp.merge.nxv1i64(<vscale x 1 x i1> %m, <vscale x 1 x i64> %y, <vscale x 1 x i64> %a, i32 %evl)
+ ret <vscale x 1 x i64> %u
+}
+
+define <vscale x 1 x i64> @vmadd_vv_nxv1i64_unmasked(<vscale x 1 x i64> %a, <vscale x 1 x i64> %b, <vscale x 1 x i64> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vsetvli zero, zero, e64, m1, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v9
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i64> @llvm.vp.mul.nxv1i64(<vscale x 1 x i64> %a, <vscale x 1 x i64> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i64> @llvm.vp.add.nxv1i64(<vscale x 1 x i64> %x, <vscale x 1 x i64> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i64> @llvm.vp.merge.nxv1i64(<vscale x 1 x i1> %allones, <vscale x 1 x i64> %y, <vscale x 1 x i64> %a, i32 %evl)
+ ret <vscale x 1 x i64> %u
+}
+
+define <vscale x 1 x i64> @vmadd_vx_nxv1i64(<vscale x 1 x i64> %a, i64 %b, <vscale x 1 x i64> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv1i64:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m1, ta, ma
+; RV32-NEXT: vlse64.v v10, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m1, ta, ma
+; RV32-NEXT: vmadd.vv v10, v8, v9
+; RV32-NEXT: vsetvli zero, zero, e64, m1, tu, ma
+; RV32-NEXT: vmerge.vvm v8, v8, v10, v0
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv1i64:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m1, tu, mu
+; RV64-NEXT: vmadd.vx v8, a0, v9, v0.t
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i64> %elt.head, <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i64> @llvm.vp.mul.nxv1i64(<vscale x 1 x i64> %a, <vscale x 1 x i64> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i64> @llvm.vp.add.nxv1i64(<vscale x 1 x i64> %x, <vscale x 1 x i64> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i64> @llvm.vp.merge.nxv1i64(<vscale x 1 x i1> %m, <vscale x 1 x i64> %y, <vscale x 1 x i64> %a, i32 %evl)
+ ret <vscale x 1 x i64> %u
+}
+
+define <vscale x 1 x i64> @vmadd_vx_nxv1i64_unmasked(<vscale x 1 x i64> %a, i64 %b, <vscale x 1 x i64> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv1i64_unmasked:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m1, ta, ma
+; RV32-NEXT: vlse64.v v10, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m1, ta, ma
+; RV32-NEXT: vmadd.vv v10, v8, v9
+; RV32-NEXT: vsetvli zero, zero, e64, m1, tu, ma
+; RV32-NEXT: vmv.v.v v8, v10
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv1i64_unmasked:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m1, tu, ma
+; RV64-NEXT: vmadd.vx v8, a0, v9
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i64> %elt.head, <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i64> @llvm.vp.mul.nxv1i64(<vscale x 1 x i64> %a, <vscale x 1 x i64> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i64> @llvm.vp.add.nxv1i64(<vscale x 1 x i64> %x, <vscale x 1 x i64> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i64> @llvm.vp.merge.nxv1i64(<vscale x 1 x i1> %allones, <vscale x 1 x i64> %y, <vscale x 1 x i64> %a, i32 %evl)
+ ret <vscale x 1 x i64> %u
+}
+
+define <vscale x 1 x i64> @vmadd_vv_nxv1i64_ta(<vscale x 1 x i64> %a, <vscale x 1 x i64> %b, <vscale x 1 x i64> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv1i64_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e64, m1, ta, ma
+; CHECK-NEXT: vmadd.vv v9, v8, v10
+; CHECK-NEXT: vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i64> @llvm.vp.mul.nxv1i64(<vscale x 1 x i64> %a, <vscale x 1 x i64> %b, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i64> @llvm.vp.add.nxv1i64(<vscale x 1 x i64> %x, <vscale x 1 x i64> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i64> @llvm.vp.select.nxv1i64(<vscale x 1 x i1> %m, <vscale x 1 x i64> %y, <vscale x 1 x i64> %a, i32 %evl)
+ ret <vscale x 1 x i64> %u
+}
+
+define <vscale x 1 x i64> @vmadd_vx_nxv1i64_ta(<vscale x 1 x i64> %a, i64 %b, <vscale x 1 x i64> %c, <vscale x 1 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv1i64_ta:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m1, ta, ma
+; RV32-NEXT: vlse64.v v10, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m1, ta, ma
+; RV32-NEXT: vmadd.vv v10, v8, v9
+; RV32-NEXT: vmerge.vvm v8, v8, v10, v0
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv1i64_ta:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m1, ta, ma
+; RV64-NEXT: vmacc.vx v9, a0, v8
+; RV64-NEXT: vmerge.vvm v8, v8, v9, v0
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 1 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 1 x i64> %elt.head, <vscale x 1 x i64> poison, <vscale x 1 x i32> zeroinitializer
+ %splat = insertelement <vscale x 1 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 1 x i1> %splat, <vscale x 1 x i1> poison, <vscale x 1 x i32> zeroinitializer
+ %x = call <vscale x 1 x i64> @llvm.vp.mul.nxv1i64(<vscale x 1 x i64> %a, <vscale x 1 x i64> %vb, <vscale x 1 x i1> %allones, i32 %evl)
+ %y = call <vscale x 1 x i64> @llvm.vp.add.nxv1i64(<vscale x 1 x i64> %x, <vscale x 1 x i64> %c, <vscale x 1 x i1> %allones, i32 %evl)
+ %u = call <vscale x 1 x i64> @llvm.vp.select.nxv1i64(<vscale x 1 x i1> %m, <vscale x 1 x i64> %y, <vscale x 1 x i64> %a, i32 %evl)
+ ret <vscale x 1 x i64> %u
+}
+
+declare <vscale x 2 x i64> @llvm.vp.mul.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i1>, i32)
+declare <vscale x 2 x i64> @llvm.vp.add.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i1>, i32)
+declare <vscale x 2 x i64> @llvm.vp.merge.nxv2i64(<vscale x 2 x i1>, <vscale x 2 x i64>, <vscale x 2 x i64>, i32)
+declare <vscale x 2 x i64> @llvm.vp.select.nxv2i64(<vscale x 2 x i1>, <vscale x 2 x i64>, <vscale x 2 x i64>, i32)
+
+define <vscale x 2 x i64> @vmadd_vv_nxv2i64(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b, <vscale x 2 x i64> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i64> @llvm.vp.mul.nxv2i64(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i64> @llvm.vp.add.nxv2i64(<vscale x 2 x i64> %x, <vscale x 2 x i64> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i64> @llvm.vp.merge.nxv2i64(<vscale x 2 x i1> %m, <vscale x 2 x i64> %y, <vscale x 2 x i64> %a, i32 %evl)
+ ret <vscale x 2 x i64> %u
+}
+
+define <vscale x 2 x i64> @vmadd_vv_nxv2i64_unmasked(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b, <vscale x 2 x i64> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vsetvli zero, zero, e64, m2, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v10
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i64> @llvm.vp.mul.nxv2i64(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i64> @llvm.vp.add.nxv2i64(<vscale x 2 x i64> %x, <vscale x 2 x i64> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i64> @llvm.vp.merge.nxv2i64(<vscale x 2 x i1> %allones, <vscale x 2 x i64> %y, <vscale x 2 x i64> %a, i32 %evl)
+ ret <vscale x 2 x i64> %u
+}
+
+define <vscale x 2 x i64> @vmadd_vx_nxv2i64(<vscale x 2 x i64> %a, i64 %b, <vscale x 2 x i64> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv2i64:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m2, ta, ma
+; RV32-NEXT: vlse64.v v12, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m2, ta, ma
+; RV32-NEXT: vmadd.vv v12, v8, v10
+; RV32-NEXT: vsetvli zero, zero, e64, m2, tu, ma
+; RV32-NEXT: vmerge.vvm v8, v8, v12, v0
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv2i64:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m2, tu, mu
+; RV64-NEXT: vmadd.vx v8, a0, v10, v0.t
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i64> %elt.head, <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i64> @llvm.vp.mul.nxv2i64(<vscale x 2 x i64> %a, <vscale x 2 x i64> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i64> @llvm.vp.add.nxv2i64(<vscale x 2 x i64> %x, <vscale x 2 x i64> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i64> @llvm.vp.merge.nxv2i64(<vscale x 2 x i1> %m, <vscale x 2 x i64> %y, <vscale x 2 x i64> %a, i32 %evl)
+ ret <vscale x 2 x i64> %u
+}
+
+define <vscale x 2 x i64> @vmadd_vx_nxv2i64_unmasked(<vscale x 2 x i64> %a, i64 %b, <vscale x 2 x i64> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv2i64_unmasked:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m2, ta, ma
+; RV32-NEXT: vlse64.v v12, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m2, ta, ma
+; RV32-NEXT: vmadd.vv v12, v8, v10
+; RV32-NEXT: vsetvli zero, zero, e64, m2, tu, ma
+; RV32-NEXT: vmv.v.v v8, v12
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv2i64_unmasked:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m2, tu, ma
+; RV64-NEXT: vmadd.vx v8, a0, v10
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i64> %elt.head, <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i64> @llvm.vp.mul.nxv2i64(<vscale x 2 x i64> %a, <vscale x 2 x i64> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i64> @llvm.vp.add.nxv2i64(<vscale x 2 x i64> %x, <vscale x 2 x i64> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i64> @llvm.vp.merge.nxv2i64(<vscale x 2 x i1> %allones, <vscale x 2 x i64> %y, <vscale x 2 x i64> %a, i32 %evl)
+ ret <vscale x 2 x i64> %u
+}
+
+define <vscale x 2 x i64> @vmadd_vv_nxv2i64_ta(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b, <vscale x 2 x i64> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv2i64_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e64, m2, ta, ma
+; CHECK-NEXT: vmadd.vv v10, v8, v12
+; CHECK-NEXT: vmerge.vvm v8, v8, v10, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i64> @llvm.vp.mul.nxv2i64(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i64> @llvm.vp.add.nxv2i64(<vscale x 2 x i64> %x, <vscale x 2 x i64> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i64> @llvm.vp.select.nxv2i64(<vscale x 2 x i1> %m, <vscale x 2 x i64> %y, <vscale x 2 x i64> %a, i32 %evl)
+ ret <vscale x 2 x i64> %u
+}
+
+define <vscale x 2 x i64> @vmadd_vx_nxv2i64_ta(<vscale x 2 x i64> %a, i64 %b, <vscale x 2 x i64> %c, <vscale x 2 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv2i64_ta:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m2, ta, ma
+; RV32-NEXT: vlse64.v v12, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m2, ta, ma
+; RV32-NEXT: vmadd.vv v12, v8, v10
+; RV32-NEXT: vmerge.vvm v8, v8, v12, v0
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv2i64_ta:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m2, ta, ma
+; RV64-NEXT: vmacc.vx v10, a0, v8
+; RV64-NEXT: vmerge.vvm v8, v8, v10, v0
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 2 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 2 x i64> %elt.head, <vscale x 2 x i64> poison, <vscale x 2 x i32> zeroinitializer
+ %splat = insertelement <vscale x 2 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 2 x i1> %splat, <vscale x 2 x i1> poison, <vscale x 2 x i32> zeroinitializer
+ %x = call <vscale x 2 x i64> @llvm.vp.mul.nxv2i64(<vscale x 2 x i64> %a, <vscale x 2 x i64> %vb, <vscale x 2 x i1> %allones, i32 %evl)
+ %y = call <vscale x 2 x i64> @llvm.vp.add.nxv2i64(<vscale x 2 x i64> %x, <vscale x 2 x i64> %c, <vscale x 2 x i1> %allones, i32 %evl)
+ %u = call <vscale x 2 x i64> @llvm.vp.select.nxv2i64(<vscale x 2 x i1> %m, <vscale x 2 x i64> %y, <vscale x 2 x i64> %a, i32 %evl)
+ ret <vscale x 2 x i64> %u
+}
+
+declare <vscale x 4 x i64> @llvm.vp.mul.nxv4i64(<vscale x 4 x i64>, <vscale x 4 x i64>, <vscale x 4 x i1>, i32)
+declare <vscale x 4 x i64> @llvm.vp.add.nxv4i64(<vscale x 4 x i64>, <vscale x 4 x i64>, <vscale x 4 x i1>, i32)
+declare <vscale x 4 x i64> @llvm.vp.merge.nxv4i64(<vscale x 4 x i1>, <vscale x 4 x i64>, <vscale x 4 x i64>, i32)
+declare <vscale x 4 x i64> @llvm.vp.select.nxv4i64(<vscale x 4 x i1>, <vscale x 4 x i64>, <vscale x 4 x i64>, i32)
+
+define <vscale x 4 x i64> @vmadd_vv_nxv4i64(<vscale x 4 x i64> %a, <vscale x 4 x i64> %b, <vscale x 4 x i64> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i64> @llvm.vp.mul.nxv4i64(<vscale x 4 x i64> %a, <vscale x 4 x i64> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i64> @llvm.vp.add.nxv4i64(<vscale x 4 x i64> %x, <vscale x 4 x i64> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i64> @llvm.vp.merge.nxv4i64(<vscale x 4 x i1> %m, <vscale x 4 x i64> %y, <vscale x 4 x i64> %a, i32 %evl)
+ ret <vscale x 4 x i64> %u
+}
+
+define <vscale x 4 x i64> @vmadd_vv_nxv4i64_unmasked(<vscale x 4 x i64> %a, <vscale x 4 x i64> %b, <vscale x 4 x i64> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m4, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v12
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i64> @llvm.vp.mul.nxv4i64(<vscale x 4 x i64> %a, <vscale x 4 x i64> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i64> @llvm.vp.add.nxv4i64(<vscale x 4 x i64> %x, <vscale x 4 x i64> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i64> @llvm.vp.merge.nxv4i64(<vscale x 4 x i1> %allones, <vscale x 4 x i64> %y, <vscale x 4 x i64> %a, i32 %evl)
+ ret <vscale x 4 x i64> %u
+}
+
+define <vscale x 4 x i64> @vmadd_vx_nxv4i64(<vscale x 4 x i64> %a, i64 %b, <vscale x 4 x i64> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv4i64:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m4, ta, ma
+; RV32-NEXT: vlse64.v v16, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m4, ta, ma
+; RV32-NEXT: vmadd.vv v16, v8, v12
+; RV32-NEXT: vsetvli zero, zero, e64, m4, tu, ma
+; RV32-NEXT: vmerge.vvm v8, v8, v16, v0
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv4i64:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m4, tu, mu
+; RV64-NEXT: vmadd.vx v8, a0, v12, v0.t
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i64> %elt.head, <vscale x 4 x i64> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i64> @llvm.vp.mul.nxv4i64(<vscale x 4 x i64> %a, <vscale x 4 x i64> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i64> @llvm.vp.add.nxv4i64(<vscale x 4 x i64> %x, <vscale x 4 x i64> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i64> @llvm.vp.merge.nxv4i64(<vscale x 4 x i1> %m, <vscale x 4 x i64> %y, <vscale x 4 x i64> %a, i32 %evl)
+ ret <vscale x 4 x i64> %u
+}
+
+define <vscale x 4 x i64> @vmadd_vx_nxv4i64_unmasked(<vscale x 4 x i64> %a, i64 %b, <vscale x 4 x i64> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv4i64_unmasked:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m4, ta, ma
+; RV32-NEXT: vlse64.v v16, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m4, ta, ma
+; RV32-NEXT: vmadd.vv v16, v8, v12
+; RV32-NEXT: vsetvli zero, zero, e64, m4, tu, ma
+; RV32-NEXT: vmv.v.v v8, v16
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv4i64_unmasked:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m4, tu, ma
+; RV64-NEXT: vmadd.vx v8, a0, v12
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i64> %elt.head, <vscale x 4 x i64> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i64> @llvm.vp.mul.nxv4i64(<vscale x 4 x i64> %a, <vscale x 4 x i64> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i64> @llvm.vp.add.nxv4i64(<vscale x 4 x i64> %x, <vscale x 4 x i64> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i64> @llvm.vp.merge.nxv4i64(<vscale x 4 x i1> %allones, <vscale x 4 x i64> %y, <vscale x 4 x i64> %a, i32 %evl)
+ ret <vscale x 4 x i64> %u
+}
+
+define <vscale x 4 x i64> @vmadd_vv_nxv4i64_ta(<vscale x 4 x i64> %a, <vscale x 4 x i64> %b, <vscale x 4 x i64> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv4i64_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vsetvli zero, a0, e64, m4, ta, ma
+; CHECK-NEXT: vmadd.vv v12, v8, v16
+; CHECK-NEXT: vmerge.vvm v8, v8, v12, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i64> @llvm.vp.mul.nxv4i64(<vscale x 4 x i64> %a, <vscale x 4 x i64> %b, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i64> @llvm.vp.add.nxv4i64(<vscale x 4 x i64> %x, <vscale x 4 x i64> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i64> @llvm.vp.select.nxv4i64(<vscale x 4 x i1> %m, <vscale x 4 x i64> %y, <vscale x 4 x i64> %a, i32 %evl)
+ ret <vscale x 4 x i64> %u
+}
+
+define <vscale x 4 x i64> @vmadd_vx_nxv4i64_ta(<vscale x 4 x i64> %a, i64 %b, <vscale x 4 x i64> %c, <vscale x 4 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv4i64_ta:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m4, ta, ma
+; RV32-NEXT: vlse64.v v16, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m4, ta, ma
+; RV32-NEXT: vmadd.vv v16, v8, v12
+; RV32-NEXT: vmerge.vvm v8, v8, v16, v0
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv4i64_ta:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m4, ta, ma
+; RV64-NEXT: vmacc.vx v12, a0, v8
+; RV64-NEXT: vmerge.vvm v8, v8, v12, v0
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 4 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 4 x i64> %elt.head, <vscale x 4 x i64> poison, <vscale x 4 x i32> zeroinitializer
+ %splat = insertelement <vscale x 4 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 4 x i1> %splat, <vscale x 4 x i1> poison, <vscale x 4 x i32> zeroinitializer
+ %x = call <vscale x 4 x i64> @llvm.vp.mul.nxv4i64(<vscale x 4 x i64> %a, <vscale x 4 x i64> %vb, <vscale x 4 x i1> %allones, i32 %evl)
+ %y = call <vscale x 4 x i64> @llvm.vp.add.nxv4i64(<vscale x 4 x i64> %x, <vscale x 4 x i64> %c, <vscale x 4 x i1> %allones, i32 %evl)
+ %u = call <vscale x 4 x i64> @llvm.vp.select.nxv4i64(<vscale x 4 x i1> %m, <vscale x 4 x i64> %y, <vscale x 4 x i64> %a, i32 %evl)
+ ret <vscale x 4 x i64> %u
+}
+
+declare <vscale x 8 x i64> @llvm.vp.mul.nxv8i64(<vscale x 8 x i64>, <vscale x 8 x i64>, <vscale x 8 x i1>, i32)
+declare <vscale x 8 x i64> @llvm.vp.add.nxv8i64(<vscale x 8 x i64>, <vscale x 8 x i64>, <vscale x 8 x i1>, i32)
+declare <vscale x 8 x i64> @llvm.vp.merge.nxv8i64(<vscale x 8 x i1>, <vscale x 8 x i64>, <vscale x 8 x i64>, i32)
+declare <vscale x 8 x i64> @llvm.vp.select.nxv8i64(<vscale x 8 x i1>, <vscale x 8 x i64>, <vscale x 8 x i64>, i32)
+
+define <vscale x 8 x i64> @vmadd_vv_nxv8i64(<vscale x 8 x i64> %a, <vscale x 8 x i64> %b, <vscale x 8 x i64> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i64:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8re64.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, tu, ma
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i64> @llvm.vp.mul.nxv8i64(<vscale x 8 x i64> %a, <vscale x 8 x i64> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i64> @llvm.vp.add.nxv8i64(<vscale x 8 x i64> %x, <vscale x 8 x i64> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i64> @llvm.vp.merge.nxv8i64(<vscale x 8 x i1> %m, <vscale x 8 x i64> %y, <vscale x 8 x i64> %a, i32 %evl)
+ ret <vscale x 8 x i64> %u
+}
+
+define <vscale x 8 x i64> @vmadd_vv_nxv8i64_unmasked(<vscale x 8 x i64> %a, <vscale x 8 x i64> %b, <vscale x 8 x i64> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i64_unmasked:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8re64.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vsetvli zero, zero, e64, m8, tu, ma
+; CHECK-NEXT: vmv.v.v v8, v24
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i64> @llvm.vp.mul.nxv8i64(<vscale x 8 x i64> %a, <vscale x 8 x i64> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i64> @llvm.vp.add.nxv8i64(<vscale x 8 x i64> %x, <vscale x 8 x i64> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i64> @llvm.vp.merge.nxv8i64(<vscale x 8 x i1> %allones, <vscale x 8 x i64> %y, <vscale x 8 x i64> %a, i32 %evl)
+ ret <vscale x 8 x i64> %u
+}
+
+define <vscale x 8 x i64> @vmadd_vx_nxv8i64(<vscale x 8 x i64> %a, i64 %b, <vscale x 8 x i64> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv8i64:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vmadd.vv v24, v8, v16
+; RV32-NEXT: vsetvli zero, zero, e64, m8, tu, ma
+; RV32-NEXT: vmerge.vvm v8, v8, v24, v0
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv8i64:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m8, tu, mu
+; RV64-NEXT: vmadd.vx v8, a0, v16, v0.t
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i64> %elt.head, <vscale x 8 x i64> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i64> @llvm.vp.mul.nxv8i64(<vscale x 8 x i64> %a, <vscale x 8 x i64> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i64> @llvm.vp.add.nxv8i64(<vscale x 8 x i64> %x, <vscale x 8 x i64> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i64> @llvm.vp.merge.nxv8i64(<vscale x 8 x i1> %m, <vscale x 8 x i64> %y, <vscale x 8 x i64> %a, i32 %evl)
+ ret <vscale x 8 x i64> %u
+}
+
+define <vscale x 8 x i64> @vmadd_vx_nxv8i64_unmasked(<vscale x 8 x i64> %a, i64 %b, <vscale x 8 x i64> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv8i64_unmasked:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vmadd.vv v24, v8, v16
+; RV32-NEXT: vsetvli zero, zero, e64, m8, tu, ma
+; RV32-NEXT: vmv.v.v v8, v24
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv8i64_unmasked:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m8, tu, ma
+; RV64-NEXT: vmadd.vx v8, a0, v16
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i64> %elt.head, <vscale x 8 x i64> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i64> @llvm.vp.mul.nxv8i64(<vscale x 8 x i64> %a, <vscale x 8 x i64> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i64> @llvm.vp.add.nxv8i64(<vscale x 8 x i64> %x, <vscale x 8 x i64> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i64> @llvm.vp.merge.nxv8i64(<vscale x 8 x i1> %allones, <vscale x 8 x i64> %y, <vscale x 8 x i64> %a, i32 %evl)
+ ret <vscale x 8 x i64> %u
+}
+
+define <vscale x 8 x i64> @vmadd_vv_nxv8i64_ta(<vscale x 8 x i64> %a, <vscale x 8 x i64> %b, <vscale x 8 x i64> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; CHECK-LABEL: vmadd_vv_nxv8i64_ta:
+; CHECK: # %bb.0:
+; CHECK-NEXT: vl8re64.v v24, (a0)
+; CHECK-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; CHECK-NEXT: vmacc.vv v24, v8, v16
+; CHECK-NEXT: vmerge.vvm v8, v8, v24, v0
+; CHECK-NEXT: ret
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i64> @llvm.vp.mul.nxv8i64(<vscale x 8 x i64> %a, <vscale x 8 x i64> %b, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i64> @llvm.vp.add.nxv8i64(<vscale x 8 x i64> %x, <vscale x 8 x i64> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i64> @llvm.vp.select.nxv8i64(<vscale x 8 x i1> %m, <vscale x 8 x i64> %y, <vscale x 8 x i64> %a, i32 %evl)
+ ret <vscale x 8 x i64> %u
+}
+
+define <vscale x 8 x i64> @vmadd_vx_nxv8i64_ta(<vscale x 8 x i64> %a, i64 %b, <vscale x 8 x i64> %c, <vscale x 8 x i1> %m, i32 zeroext %evl) {
+; RV32-LABEL: vmadd_vx_nxv8i64_ta:
+; RV32: # %bb.0:
+; RV32-NEXT: addi sp, sp, -16
+; RV32-NEXT: .cfi_def_cfa_offset 16
+; RV32-NEXT: sw a1, 12(sp)
+; RV32-NEXT: sw a0, 8(sp)
+; RV32-NEXT: addi a0, sp, 8
+; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, ma
+; RV32-NEXT: vlse64.v v24, (a0), zero
+; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
+; RV32-NEXT: vmadd.vv v24, v8, v16
+; RV32-NEXT: vmerge.vvm v8, v8, v24, v0
+; RV32-NEXT: addi sp, sp, 16
+; RV32-NEXT: ret
+;
+; RV64-LABEL: vmadd_vx_nxv8i64_ta:
+; RV64: # %bb.0:
+; RV64-NEXT: vsetvli zero, a1, e64, m8, ta, ma
+; RV64-NEXT: vmacc.vx v16, a0, v8
+; RV64-NEXT: vmerge.vvm v8, v8, v16, v0
+; RV64-NEXT: ret
+ %elt.head = insertelement <vscale x 8 x i64> poison, i64 %b, i32 0
+ %vb = shufflevector <vscale x 8 x i64> %elt.head, <vscale x 8 x i64> poison, <vscale x 8 x i32> zeroinitializer
+ %splat = insertelement <vscale x 8 x i1> poison, i1 -1, i32 0
+ %allones = shufflevector <vscale x 8 x i1> %splat, <vscale x 8 x i1> poison, <vscale x 8 x i32> zeroinitializer
+ %x = call <vscale x 8 x i64> @llvm.vp.mul.nxv8i64(<vscale x 8 x i64> %a, <vscale x 8 x i64> %vb, <vscale x 8 x i1> %allones, i32 %evl)
+ %y = call <vscale x 8 x i64> @llvm.vp.add.nxv8i64(<vscale x 8 x i64> %x, <vscale x 8 x i64> %c, <vscale x 8 x i1> %allones, i32 %evl)
+ %u = call <vscale x 8 x i64> @llvm.vp.select.nxv8i64(<vscale x 8 x i1> %m, <vscale x 8 x i64> %y, <vscale x 8 x i64> %a, i32 %evl)
+ ret <vscale x 8 x i64> %u
+}
More information about the llvm-commits
mailing list