derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
1918 lines
119 KiB
LLVM
1918 lines
119 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc -mtriple=riscv32 -mattr=+v,+m -target-abi=ilp32d \
|
|
; RUN: -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV32
|
|
; RUN: llc -mtriple=riscv64 -mattr=+v,+m -target-abi=lp64d \
|
|
; RUN: -verify-machineinstrs < %s | FileCheck %s --check-prefixes=CHECK,RV64
|
|
|
|
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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 1 x i8> @llvm.vp.merge.nxv1i8(<vscale x 1 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 1 x i8> @llvm.vp.merge.nxv1i8(<vscale x 1 x i1> splat (i1 -1), <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 a0, zero, e8, mf8, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e8, mf8, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 2 x i8> @llvm.vp.merge.nxv2i8(<vscale x 2 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 2 x i8> @llvm.vp.merge.nxv2i8(<vscale x 2 x i1> splat (i1 -1), <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 a0, zero, e8, mf4, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e8, mf4, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 4 x i8> @llvm.vp.merge.nxv4i8(<vscale x 4 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 4 x i8> @llvm.vp.merge.nxv4i8(<vscale x 4 x i1> splat (i1 -1), <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 a0, zero, e8, mf2, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e8, mf2, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 8 x i8> @llvm.vp.merge.nxv8i8(<vscale x 8 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 8 x i8> @llvm.vp.merge.nxv8i8(<vscale x 8 x i1> splat (i1 -1), <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 a0, zero, e8, m1, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e8, m1, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 16 x i8> @llvm.vp.merge.nxv16i8(<vscale x 16 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 16 x i8> @llvm.vp.merge.nxv16i8(<vscale x 16 x i1> splat (i1 -1), <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 a0, zero, e8, m2, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e8, m2, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 32 x i8> @llvm.vp.merge.nxv32i8(<vscale x 32 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 32 x i8> @llvm.vp.merge.nxv32i8(<vscale x 32 x i1> splat (i1 -1), <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 a0, zero, e8, m4, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e8, m4, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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: vsetvli zero, a1, e8, m8, ta, ma
|
|
; CHECK-NEXT: vle8.v v24, (a0)
|
|
; CHECK-NEXT: vsetvli zero, zero, e8, m8, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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: vsetvli zero, a1, e8, m8, ta, ma
|
|
; CHECK-NEXT: vle8.v v24, (a0)
|
|
; CHECK-NEXT: vsetvli zero, zero, e8, m8, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 64 x i8> @llvm.vp.merge.nxv64i8(<vscale x 64 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 64 x i8> @llvm.vp.merge.nxv64i8(<vscale x 64 x i1> splat (i1 -1), <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 a0, zero, e8, m8, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e8, m8, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 1 x i16> @llvm.vp.merge.nxv1i16(<vscale x 1 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 1 x i16> @llvm.vp.merge.nxv1i16(<vscale x 1 x i1> splat (i1 -1), <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 a0, zero, e16, mf4, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e16, mf4, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 2 x i16> @llvm.vp.merge.nxv2i16(<vscale x 2 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 2 x i16> @llvm.vp.merge.nxv2i16(<vscale x 2 x i1> splat (i1 -1), <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 a0, zero, e16, mf2, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e16, mf2, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 4 x i16> @llvm.vp.merge.nxv4i16(<vscale x 4 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 4 x i16> @llvm.vp.merge.nxv4i16(<vscale x 4 x i1> splat (i1 -1), <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 a0, zero, e16, m1, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e16, m1, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 8 x i16> @llvm.vp.merge.nxv8i16(<vscale x 8 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 8 x i16> @llvm.vp.merge.nxv8i16(<vscale x 8 x i1> splat (i1 -1), <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 a0, zero, e16, m2, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e16, m2, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 16 x i16> @llvm.vp.merge.nxv16i16(<vscale x 16 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 16 x i16> @llvm.vp.merge.nxv16i16(<vscale x 16 x i1> splat (i1 -1), <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 a0, zero, e16, m4, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e16, m4, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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: vsetvli zero, a1, e16, m8, ta, ma
|
|
; CHECK-NEXT: vle16.v v24, (a0)
|
|
; CHECK-NEXT: vsetvli zero, zero, e16, m8, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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: vsetvli zero, a1, e16, m8, ta, ma
|
|
; CHECK-NEXT: vle16.v v24, (a0)
|
|
; CHECK-NEXT: vsetvli zero, zero, e16, m8, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 32 x i16> @llvm.vp.merge.nxv32i16(<vscale x 32 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 32 x i16> @llvm.vp.merge.nxv32i16(<vscale x 32 x i1> splat (i1 -1), <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 a0, zero, e16, m8, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e16, m8, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 1 x i32> @llvm.vp.merge.nxv1i32(<vscale x 1 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 1 x i32> @llvm.vp.merge.nxv1i32(<vscale x 1 x i1> splat (i1 -1), <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 a0, zero, e32, mf2, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e32, mf2, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 2 x i32> @llvm.vp.merge.nxv2i32(<vscale x 2 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 2 x i32> @llvm.vp.merge.nxv2i32(<vscale x 2 x i1> splat (i1 -1), <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 a0, zero, e32, m1, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e32, m1, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 4 x i32> @llvm.vp.merge.nxv4i32(<vscale x 4 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 4 x i32> @llvm.vp.merge.nxv4i32(<vscale x 4 x i1> splat (i1 -1), <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 a0, zero, e32, m2, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e32, m2, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 8 x i32> @llvm.vp.merge.nxv8i32(<vscale x 8 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 8 x i32> @llvm.vp.merge.nxv8i32(<vscale x 8 x i1> splat (i1 -1), <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 a0, zero, e32, m4, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e32, m4, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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: vsetvli zero, a1, e32, m8, ta, ma
|
|
; CHECK-NEXT: vle32.v v24, (a0)
|
|
; CHECK-NEXT: vsetvli zero, zero, e32, m8, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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: vsetvli zero, a1, e32, m8, ta, ma
|
|
; CHECK-NEXT: vle32.v v24, (a0)
|
|
; CHECK-NEXT: vsetvli zero, zero, e32, m8, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 16 x i32> @llvm.vp.merge.nxv16i32(<vscale x 16 x i1> splat (i1 -1), <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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 16 x i32> @llvm.vp.merge.nxv16i32(<vscale x 16 x i1> splat (i1 -1), <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 a0, zero, e32, m8, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a1, zero, e32, m8, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 1 x i64> @llvm.vp.merge.nxv1i64(<vscale x 1 x i1> splat (i1 -1), <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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli zero, a2, e64, m1, ta, ma
|
|
; RV32-NEXT: vlse64.v v10, (a0), zero
|
|
; RV32-NEXT: vsetvli zero, zero, e64, m1, tu, mu
|
|
; RV32-NEXT: vmadd.vv v8, v10, v9, v0.t
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli zero, a2, e64, m1, ta, ma
|
|
; RV32-NEXT: vlse64.v v10, (a0), zero
|
|
; RV32-NEXT: vsetvli zero, zero, e64, m1, tu, ma
|
|
; RV32-NEXT: vmadd.vv v8, v10, v9
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 1 x i64> @llvm.vp.merge.nxv1i64(<vscale x 1 x i1> splat (i1 -1), <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 a0, zero, e64, m1, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v9, v10, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli a1, zero, e64, m1, ta, mu
|
|
; RV32-NEXT: vlse64.v v10, (a0), zero
|
|
; RV32-NEXT: vmadd.vv v8, v10, v9, v0.t
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; RV32-NEXT: ret
|
|
;
|
|
; RV64-LABEL: vmadd_vx_nxv1i64_ta:
|
|
; RV64: # %bb.0:
|
|
; RV64-NEXT: vsetvli a1, zero, e64, m1, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 2 x i64> @llvm.vp.merge.nxv2i64(<vscale x 2 x i1> splat (i1 -1), <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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli zero, a2, e64, m2, ta, ma
|
|
; RV32-NEXT: vlse64.v v12, (a0), zero
|
|
; RV32-NEXT: vsetvli zero, zero, e64, m2, tu, mu
|
|
; RV32-NEXT: vmadd.vv v8, v12, v10, v0.t
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli zero, a2, e64, m2, ta, ma
|
|
; RV32-NEXT: vlse64.v v12, (a0), zero
|
|
; RV32-NEXT: vsetvli zero, zero, e64, m2, tu, ma
|
|
; RV32-NEXT: vmadd.vv v8, v12, v10
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 2 x i64> @llvm.vp.merge.nxv2i64(<vscale x 2 x i1> splat (i1 -1), <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 a0, zero, e64, m2, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v10, v12, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli a1, zero, e64, m2, ta, mu
|
|
; RV32-NEXT: vlse64.v v12, (a0), zero
|
|
; RV32-NEXT: vmadd.vv v8, v12, v10, v0.t
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; RV32-NEXT: ret
|
|
;
|
|
; RV64-LABEL: vmadd_vx_nxv2i64_ta:
|
|
; RV64: # %bb.0:
|
|
; RV64-NEXT: vsetvli a1, zero, e64, m2, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 4 x i64> @llvm.vp.merge.nxv4i64(<vscale x 4 x i1> splat (i1 -1), <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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli zero, a2, e64, m4, ta, ma
|
|
; RV32-NEXT: vlse64.v v16, (a0), zero
|
|
; RV32-NEXT: vsetvli zero, zero, e64, m4, tu, mu
|
|
; RV32-NEXT: vmadd.vv v8, v16, v12, v0.t
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli zero, a2, e64, m4, ta, ma
|
|
; RV32-NEXT: vlse64.v v16, (a0), zero
|
|
; RV32-NEXT: vsetvli zero, zero, e64, m4, tu, ma
|
|
; RV32-NEXT: vmadd.vv v8, v16, v12
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 4 x i64> @llvm.vp.merge.nxv4i64(<vscale x 4 x i1> splat (i1 -1), <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 a0, zero, e64, m4, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v12, v16, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli a1, zero, e64, m4, ta, mu
|
|
; RV32-NEXT: vlse64.v v16, (a0), zero
|
|
; RV32-NEXT: vmadd.vv v8, v16, v12, v0.t
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; RV32-NEXT: ret
|
|
;
|
|
; RV64-LABEL: vmadd_vx_nxv4i64_ta:
|
|
; RV64: # %bb.0:
|
|
; RV64-NEXT: vsetvli a1, zero, e64, m4, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 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: vsetvli zero, a1, e64, m8, ta, ma
|
|
; CHECK-NEXT: vle64.v v24, (a0)
|
|
; CHECK-NEXT: vsetvli zero, zero, e64, m8, tu, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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: vsetvli zero, a1, e64, m8, ta, ma
|
|
; CHECK-NEXT: vle64.v v24, (a0)
|
|
; CHECK-NEXT: vsetvli zero, zero, e64, m8, tu, ma
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 8 x i64> @llvm.vp.merge.nxv8i64(<vscale x 8 x i1> splat (i1 -1), <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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
|
|
; RV32-NEXT: vlse64.v v24, (a0), zero
|
|
; RV32-NEXT: vsetvli zero, zero, e64, m8, tu, mu
|
|
; RV32-NEXT: vmadd.vv v8, v24, v16, v0.t
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli zero, a2, e64, m8, ta, ma
|
|
; RV32-NEXT: vlse64.v v24, (a0), zero
|
|
; RV32-NEXT: vsetvli zero, zero, e64, m8, tu, ma
|
|
; RV32-NEXT: vmadd.vv v8, v24, v16
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), i32 %evl)
|
|
%u = call <vscale x 8 x i64> @llvm.vp.merge.nxv8i64(<vscale x 8 x i1> splat (i1 -1), <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 a0, zero, e64, m8, ta, mu
|
|
; CHECK-NEXT: vmadd.vv v8, v16, v24, v0.t
|
|
; CHECK-NEXT: ret
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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 a0, 8(sp)
|
|
; RV32-NEXT: sw a1, 12(sp)
|
|
; RV32-NEXT: addi a0, sp, 8
|
|
; RV32-NEXT: vsetvli a1, zero, e64, m8, ta, mu
|
|
; RV32-NEXT: vlse64.v v24, (a0), zero
|
|
; RV32-NEXT: vmadd.vv v8, v24, v16, v0.t
|
|
; RV32-NEXT: addi sp, sp, 16
|
|
; RV32-NEXT: .cfi_def_cfa_offset 0
|
|
; RV32-NEXT: ret
|
|
;
|
|
; RV64-LABEL: vmadd_vx_nxv8i64_ta:
|
|
; RV64: # %bb.0:
|
|
; RV64-NEXT: vsetvli a1, zero, e64, m8, ta, 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
|
|
%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> splat (i1 -1), 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> splat (i1 -1), 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
|
|
}
|