test/CodeGen/AArch64/sve-fixed-length-int-rem.ll

; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -aarch64-sve-vector-bits-min=128  < %s | FileCheck %s -check-prefixes=CHECK,VBITS_GE_128
; RUN: llc -aarch64-sve-vector-bits-min=256  < %s | FileCheck %s -check-prefixes=CHECK,VBITS_GE_256
; RUN: llc -aarch64-sve-vector-bits-min=512  < %s | FileCheck %s -check-prefixes=CHECK,VBITS_GE_512
; RUN: llc -aarch64-sve-vector-bits-min=2048 < %s | FileCheck %s -check-prefixes=CHECK,VBITS_GE_512

target triple = "aarch64-unknown-linux-gnu"

;
; SREM
;

; Vector vXi8 sdiv are not legal for NEON so use SVE when available.
; FIXME: We should be able to improve the codegen for >= 256 bits here.
define <8 x i8> @srem_v8i8(<8 x i8> %op1, <8 x i8> %op2) #0 {
; VBITS_GE_128-LABEL: srem_v8i8:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    sshll v2.8h, v1.8b, #0
; VBITS_GE_128-NEXT:    sshll v3.8h, v0.8b, #0
; VBITS_GE_128-NEXT:    sshll2 v4.4s, v2.8h, #0
; VBITS_GE_128-NEXT:    sshll2 v5.4s, v3.8h, #0
; VBITS_GE_128-NEXT:    sshll v2.4s, v2.4h, #0
; VBITS_GE_128-NEXT:    sshll v3.4s, v3.4h, #0
; VBITS_GE_128-NEXT:    sdivr z4.s, p0/m, z4.s, z5.s
; VBITS_GE_128-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    uzp1 v2.8h, v2.8h, v4.8h
; VBITS_GE_128-NEXT:    xtn v2.8b, v2.8h
; VBITS_GE_128-NEXT:    mls v0.8b, v2.8b, v1.8b
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: srem_v8i8:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    // kill: def $d1 killed $d1 def $z1
; VBITS_GE_256-NEXT:    // kill: def $d0 killed $d0 def $z0
; VBITS_GE_256-NEXT:    sunpklo z2.h, z1.b
; VBITS_GE_256-NEXT:    sunpklo z3.h, z0.b
; VBITS_GE_256-NEXT:    ptrue p0.s, vl8
; VBITS_GE_256-NEXT:    sunpklo z2.s, z2.h
; VBITS_GE_256-NEXT:    sunpklo z3.s, z3.h
; VBITS_GE_256-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_256-NEXT:    umov w8, v2.h[0]
; VBITS_GE_256-NEXT:    umov w9, v2.h[1]
; VBITS_GE_256-NEXT:    fmov s3, w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[2]
; VBITS_GE_256-NEXT:    mov v3.b[1], w9
; VBITS_GE_256-NEXT:    mov v3.b[2], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[3]
; VBITS_GE_256-NEXT:    mov v3.b[3], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[4]
; VBITS_GE_256-NEXT:    mov v3.b[4], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[5]
; VBITS_GE_256-NEXT:    mov v3.b[5], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[6]
; VBITS_GE_256-NEXT:    mov v3.b[6], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[7]
; VBITS_GE_256-NEXT:    mov v3.b[7], w8
; VBITS_GE_256-NEXT:    mls v0.8b, v3.8b, v1.8b
; VBITS_GE_256-NEXT:    // kill: def $d0 killed $d0 killed $z0
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: srem_v8i8:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    // kill: def $d1 killed $d1 def $z1
; VBITS_GE_512-NEXT:    // kill: def $d0 killed $d0 def $z0
; VBITS_GE_512-NEXT:    sunpklo z2.h, z1.b
; VBITS_GE_512-NEXT:    sunpklo z3.h, z0.b
; VBITS_GE_512-NEXT:    ptrue p0.s, vl8
; VBITS_GE_512-NEXT:    sunpklo z2.s, z2.h
; VBITS_GE_512-NEXT:    sunpklo z3.s, z3.h
; VBITS_GE_512-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_512-NEXT:    umov w8, v2.h[0]
; VBITS_GE_512-NEXT:    umov w9, v2.h[1]
; VBITS_GE_512-NEXT:    fmov s3, w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[2]
; VBITS_GE_512-NEXT:    mov v3.b[1], w9
; VBITS_GE_512-NEXT:    mov v3.b[2], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[3]
; VBITS_GE_512-NEXT:    mov v3.b[3], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[4]
; VBITS_GE_512-NEXT:    mov v3.b[4], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[5]
; VBITS_GE_512-NEXT:    mov v3.b[5], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[6]
; VBITS_GE_512-NEXT:    mov v3.b[6], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[7]
; VBITS_GE_512-NEXT:    mov v3.b[7], w8
; VBITS_GE_512-NEXT:    mls v0.8b, v3.8b, v1.8b
; VBITS_GE_512-NEXT:    // kill: def $d0 killed $d0 killed $z0
; VBITS_GE_512-NEXT:    ret
  %res = srem <8 x i8> %op1, %op2
  ret <8 x i8> %res
}

define <16 x i8> @srem_v16i8(<16 x i8> %op1, <16 x i8> %op2) #0 {
; VBITS_GE_128-LABEL: srem_v16i8:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    sshll2 v2.8h, v1.16b, #0
; VBITS_GE_128-NEXT:    sshll2 v3.8h, v0.16b, #0
; VBITS_GE_128-NEXT:    sshll2 v4.4s, v2.8h, #0
; VBITS_GE_128-NEXT:    sshll2 v5.4s, v3.8h, #0
; VBITS_GE_128-NEXT:    sshll v2.4s, v2.4h, #0
; VBITS_GE_128-NEXT:    sshll v3.4s, v3.4h, #0
; VBITS_GE_128-NEXT:    sdivr z4.s, p0/m, z4.s, z5.s
; VBITS_GE_128-NEXT:    sshll v5.8h, v0.8b, #0
; VBITS_GE_128-NEXT:    sshll2 v7.4s, v5.8h, #0
; VBITS_GE_128-NEXT:    sshll v5.4s, v5.4h, #0
; VBITS_GE_128-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    sshll v3.8h, v1.8b, #0
; VBITS_GE_128-NEXT:    sshll2 v6.4s, v3.8h, #0
; VBITS_GE_128-NEXT:    sshll v3.4s, v3.4h, #0
; VBITS_GE_128-NEXT:    sdivr z6.s, p0/m, z6.s, z7.s
; VBITS_GE_128-NEXT:    uzp1 v2.8h, v2.8h, v4.8h
; VBITS_GE_128-NEXT:    sdivr z3.s, p0/m, z3.s, z5.s
; VBITS_GE_128-NEXT:    uzp1 v3.8h, v3.8h, v6.8h
; VBITS_GE_128-NEXT:    uzp1 v2.16b, v3.16b, v2.16b
; VBITS_GE_128-NEXT:    mls v0.16b, v2.16b, v1.16b
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: srem_v16i8:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    // kill: def $q1 killed $q1 def $z1
; VBITS_GE_256-NEXT:    // kill: def $q0 killed $q0 def $z0
; VBITS_GE_256-NEXT:    sunpklo z2.h, z1.b
; VBITS_GE_256-NEXT:    sunpklo z3.h, z0.b
; VBITS_GE_256-NEXT:    ptrue p0.s, vl8
; VBITS_GE_256-NEXT:    sunpklo z4.s, z2.h
; VBITS_GE_256-NEXT:    sunpklo z5.s, z3.h
; VBITS_GE_256-NEXT:    ext z2.b, z2.b, z2.b, #16
; VBITS_GE_256-NEXT:    ext z3.b, z3.b, z3.b, #16
; VBITS_GE_256-NEXT:    sunpklo z2.s, z2.h
; VBITS_GE_256-NEXT:    sunpklo z3.s, z3.h
; VBITS_GE_256-NEXT:    sdivr z4.s, p0/m, z4.s, z5.s
; VBITS_GE_256-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    ptrue p0.h, vl8
; VBITS_GE_256-NEXT:    uzp1 z3.h, z4.h, z4.h
; VBITS_GE_256-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_256-NEXT:    splice z3.h, p0, z3.h, z2.h
; VBITS_GE_256-NEXT:    uzp1 z2.b, z3.b, z3.b
; VBITS_GE_256-NEXT:    mls v0.16b, v2.16b, v1.16b
; VBITS_GE_256-NEXT:    // kill: def $q0 killed $q0 killed $z0
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: srem_v16i8:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    // kill: def $q1 killed $q1 def $z1
; VBITS_GE_512-NEXT:    // kill: def $q0 killed $q0 def $z0
; VBITS_GE_512-NEXT:    sunpklo z2.h, z1.b
; VBITS_GE_512-NEXT:    sunpklo z3.h, z0.b
; VBITS_GE_512-NEXT:    ptrue p0.s, vl16
; VBITS_GE_512-NEXT:    sunpklo z2.s, z2.h
; VBITS_GE_512-NEXT:    sunpklo z3.s, z3.h
; VBITS_GE_512-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_512-NEXT:    uzp1 z2.b, z2.b, z2.b
; VBITS_GE_512-NEXT:    mls v0.16b, v2.16b, v1.16b
; VBITS_GE_512-NEXT:    // kill: def $q0 killed $q0 killed $z0
; VBITS_GE_512-NEXT:    ret
  %res = srem <16 x i8> %op1, %op2
  ret <16 x i8> %res
}

define void @srem_v32i8(ptr %a, ptr %b) vscale_range(8,0) #0 {
; CHECK-LABEL: srem_v32i8:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.b, vl32
; CHECK-NEXT:    ptrue p1.s, vl32
; CHECK-NEXT:    ld1b { z0.b }, p0/z, [x0]
; CHECK-NEXT:    ld1b { z1.b }, p0/z, [x1]
; CHECK-NEXT:    sunpklo z2.h, z1.b
; CHECK-NEXT:    sunpklo z3.h, z0.b
; CHECK-NEXT:    sunpklo z2.s, z2.h
; CHECK-NEXT:    sunpklo z3.s, z3.h
; CHECK-NEXT:    sdivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    uzp1 z2.b, z2.b, z2.b
; CHECK-NEXT:    mls z0.b, p0/m, z2.b, z1.b
; CHECK-NEXT:    st1b { z0.b }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <32 x i8>, ptr %a
  %op2 = load <32 x i8>, ptr %b
  %res = srem <32 x i8> %op1, %op2
  store <32 x i8> %res, ptr %a
  ret void
}

define void @srem_v64i8(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: srem_v64i8:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.b, vl64
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1b { z0.b }, p0/z, [x0]
; CHECK-NEXT:    ld1b { z1.b }, p0/z, [x1]
; CHECK-NEXT:    sunpklo z2.h, z1.b
; CHECK-NEXT:    sunpklo z3.h, z0.b
; CHECK-NEXT:    sunpklo z2.s, z2.h
; CHECK-NEXT:    sunpklo z3.s, z3.h
; CHECK-NEXT:    sdivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    uzp1 z2.b, z2.b, z2.b
; CHECK-NEXT:    mls z0.b, p0/m, z2.b, z1.b
; CHECK-NEXT:    st1b { z0.b }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <64 x i8>, ptr %a
  %op2 = load <64 x i8>, ptr %b
  %res = srem <64 x i8> %op1, %op2
  store <64 x i8> %res, ptr %a
  ret void
}

define void @srem_v128i8(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: srem_v128i8:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.b, vl128
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1b { z0.b }, p0/z, [x0]
; CHECK-NEXT:    ld1b { z1.b }, p0/z, [x1]
; CHECK-NEXT:    sunpklo z2.h, z1.b
; CHECK-NEXT:    sunpklo z3.h, z0.b
; CHECK-NEXT:    sunpklo z4.s, z2.h
; CHECK-NEXT:    sunpklo z5.s, z3.h
; CHECK-NEXT:    ext z2.b, z2.b, z2.b, #128
; CHECK-NEXT:    ext z3.b, z3.b, z3.b, #128
; CHECK-NEXT:    sunpklo z2.s, z2.h
; CHECK-NEXT:    sunpklo z3.s, z3.h
; CHECK-NEXT:    sdivr z4.s, p1/m, z4.s, z5.s
; CHECK-NEXT:    sdivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    ptrue p1.h, vl64
; CHECK-NEXT:    uzp1 z3.h, z4.h, z4.h
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    splice z3.h, p1, z3.h, z2.h
; CHECK-NEXT:    uzp1 z2.b, z3.b, z3.b
; CHECK-NEXT:    mls z0.b, p0/m, z2.b, z1.b
; CHECK-NEXT:    st1b { z0.b }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <128 x i8>, ptr %a
  %op2 = load <128 x i8>, ptr %b
  %res = srem <128 x i8> %op1, %op2
  store <128 x i8> %res, ptr %a
  ret void
}

define void @srem_v256i8(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: srem_v256i8:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.b, vl256
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1b { z0.b }, p0/z, [x0]
; CHECK-NEXT:    ld1b { z1.b }, p0/z, [x1]
; CHECK-NEXT:    sunpklo z2.h, z1.b
; CHECK-NEXT:    sunpklo z3.h, z0.b
; CHECK-NEXT:    sunpklo z4.s, z2.h
; CHECK-NEXT:    sunpklo z5.s, z3.h
; CHECK-NEXT:    ext z2.b, z2.b, z2.b, #128
; CHECK-NEXT:    ext z3.b, z3.b, z3.b, #128
; CHECK-NEXT:    sunpklo z2.s, z2.h
; CHECK-NEXT:    sunpklo z3.s, z3.h
; CHECK-NEXT:    sdivr z4.s, p1/m, z4.s, z5.s
; CHECK-NEXT:    mov z5.d, z0.d
; CHECK-NEXT:    ext z5.b, z5.b, z0.b, #128
; CHECK-NEXT:    sunpklo z5.h, z5.b
; CHECK-NEXT:    sunpklo z7.s, z5.h
; CHECK-NEXT:    ext z5.b, z5.b, z5.b, #128
; CHECK-NEXT:    sdivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    mov z3.d, z1.d
; CHECK-NEXT:    sunpklo z5.s, z5.h
; CHECK-NEXT:    ext z3.b, z3.b, z1.b, #128
; CHECK-NEXT:    uzp1 z4.h, z4.h, z4.h
; CHECK-NEXT:    sunpklo z3.h, z3.b
; CHECK-NEXT:    sunpklo z6.s, z3.h
; CHECK-NEXT:    ext z3.b, z3.b, z3.b, #128
; CHECK-NEXT:    sunpklo z3.s, z3.h
; CHECK-NEXT:    sdivr z6.s, p1/m, z6.s, z7.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    sdivr z3.s, p1/m, z3.s, z5.s
; CHECK-NEXT:    ptrue p1.h, vl64
; CHECK-NEXT:    uzp1 z5.h, z6.h, z6.h
; CHECK-NEXT:    splice z4.h, p1, z4.h, z2.h
; CHECK-NEXT:    uzp1 z2.b, z4.b, z4.b
; CHECK-NEXT:    uzp1 z3.h, z3.h, z3.h
; CHECK-NEXT:    splice z5.h, p1, z5.h, z3.h
; CHECK-NEXT:    ptrue p1.b, vl128
; CHECK-NEXT:    uzp1 z3.b, z5.b, z5.b
; CHECK-NEXT:    splice z2.b, p1, z2.b, z3.b
; CHECK-NEXT:    mls z0.b, p0/m, z2.b, z1.b
; CHECK-NEXT:    st1b { z0.b }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <256 x i8>, ptr %a
  %op2 = load <256 x i8>, ptr %b
  %res = srem <256 x i8> %op1, %op2
  store <256 x i8> %res, ptr %a
  ret void
}

; Vector vXi16 sdiv are not legal for NEON so use SVE when available.
; FIXME: We should be able to improve the codegen for >= 256 bits here.
define <4 x i16> @srem_v4i16(<4 x i16> %op1, <4 x i16> %op2) #0 {
; VBITS_GE_128-LABEL: srem_v4i16:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    sshll v2.4s, v1.4h, #0
; VBITS_GE_128-NEXT:    sshll v3.4s, v0.4h, #0
; VBITS_GE_128-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    xtn v2.4h, v2.4s
; VBITS_GE_128-NEXT:    mls v0.4h, v2.4h, v1.4h
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: srem_v4i16:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.s, vl4
; VBITS_GE_256-NEXT:    sshll v2.4s, v1.4h, #0
; VBITS_GE_256-NEXT:    sshll v3.4s, v0.4h, #0
; VBITS_GE_256-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    mov w8, v2.s[1]
; VBITS_GE_256-NEXT:    mov v3.16b, v2.16b
; VBITS_GE_256-NEXT:    mov w9, v2.s[2]
; VBITS_GE_256-NEXT:    mov v3.h[1], w8
; VBITS_GE_256-NEXT:    mov w8, v2.s[3]
; VBITS_GE_256-NEXT:    mov v3.h[2], w9
; VBITS_GE_256-NEXT:    mov v3.h[3], w8
; VBITS_GE_256-NEXT:    mls v0.4h, v3.4h, v1.4h
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: srem_v4i16:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.s, vl4
; VBITS_GE_512-NEXT:    sshll v2.4s, v1.4h, #0
; VBITS_GE_512-NEXT:    sshll v3.4s, v0.4h, #0
; VBITS_GE_512-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    mov w8, v2.s[1]
; VBITS_GE_512-NEXT:    mov v3.16b, v2.16b
; VBITS_GE_512-NEXT:    mov w9, v2.s[2]
; VBITS_GE_512-NEXT:    mov v3.h[1], w8
; VBITS_GE_512-NEXT:    mov w8, v2.s[3]
; VBITS_GE_512-NEXT:    mov v3.h[2], w9
; VBITS_GE_512-NEXT:    mov v3.h[3], w8
; VBITS_GE_512-NEXT:    mls v0.4h, v3.4h, v1.4h
; VBITS_GE_512-NEXT:    ret
  %res = srem <4 x i16> %op1, %op2
  ret <4 x i16> %res
}

define <8 x i16> @srem_v8i16(<8 x i16> %op1, <8 x i16> %op2) #0 {
; VBITS_GE_128-LABEL: srem_v8i16:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    sshll2 v2.4s, v1.8h, #0
; VBITS_GE_128-NEXT:    sshll2 v3.4s, v0.8h, #0
; VBITS_GE_128-NEXT:    sshll v4.4s, v0.4h, #0
; VBITS_GE_128-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    sshll v3.4s, v1.4h, #0
; VBITS_GE_128-NEXT:    sdivr z3.s, p0/m, z3.s, z4.s
; VBITS_GE_128-NEXT:    uzp1 v2.8h, v3.8h, v2.8h
; VBITS_GE_128-NEXT:    mls v0.8h, v2.8h, v1.8h
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: srem_v8i16:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.s, vl8
; VBITS_GE_256-NEXT:    // kill: def $q1 killed $q1 def $z1
; VBITS_GE_256-NEXT:    // kill: def $q0 killed $q0 def $z0
; VBITS_GE_256-NEXT:    sunpklo z2.s, z1.h
; VBITS_GE_256-NEXT:    sunpklo z3.s, z0.h
; VBITS_GE_256-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_256-NEXT:    mls v0.8h, v2.8h, v1.8h
; VBITS_GE_256-NEXT:    // kill: def $q0 killed $q0 killed $z0
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: srem_v8i16:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.s, vl8
; VBITS_GE_512-NEXT:    // kill: def $q1 killed $q1 def $z1
; VBITS_GE_512-NEXT:    // kill: def $q0 killed $q0 def $z0
; VBITS_GE_512-NEXT:    sunpklo z2.s, z1.h
; VBITS_GE_512-NEXT:    sunpklo z3.s, z0.h
; VBITS_GE_512-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_512-NEXT:    mls v0.8h, v2.8h, v1.8h
; VBITS_GE_512-NEXT:    // kill: def $q0 killed $q0 killed $z0
; VBITS_GE_512-NEXT:    ret
  %res = srem <8 x i16> %op1, %op2
  ret <8 x i16> %res
}

define void @srem_v16i16(ptr %a, ptr %b) #0 {
; VBITS_GE_128-LABEL: srem_v16i16:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    ldp q4, q1, [x1]
; VBITS_GE_128-NEXT:    ldr q0, [x0, #16]
; VBITS_GE_128-NEXT:    sshll2 v2.4s, v1.8h, #0
; VBITS_GE_128-NEXT:    sshll2 v3.4s, v0.8h, #0
; VBITS_GE_128-NEXT:    sshll2 v5.4s, v4.8h, #0
; VBITS_GE_128-NEXT:    sshll v16.4s, v0.4h, #0
; VBITS_GE_128-NEXT:    sdivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    ldr q3, [x0]
; VBITS_GE_128-NEXT:    sshll2 v6.4s, v3.8h, #0
; VBITS_GE_128-NEXT:    sshll v7.4s, v3.4h, #0
; VBITS_GE_128-NEXT:    sdivr z5.s, p0/m, z5.s, z6.s
; VBITS_GE_128-NEXT:    sshll v6.4s, v4.4h, #0
; VBITS_GE_128-NEXT:    sdivr z6.s, p0/m, z6.s, z7.s
; VBITS_GE_128-NEXT:    sshll v7.4s, v1.4h, #0
; VBITS_GE_128-NEXT:    sdivr z7.s, p0/m, z7.s, z16.s
; VBITS_GE_128-NEXT:    uzp1 v5.8h, v6.8h, v5.8h
; VBITS_GE_128-NEXT:    mls v3.8h, v5.8h, v4.8h
; VBITS_GE_128-NEXT:    uzp1 v2.8h, v7.8h, v2.8h
; VBITS_GE_128-NEXT:    mls v0.8h, v2.8h, v1.8h
; VBITS_GE_128-NEXT:    stp q3, q0, [x0]
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: srem_v16i16:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.h, vl16
; VBITS_GE_256-NEXT:    ptrue p1.s, vl8
; VBITS_GE_256-NEXT:    ld1h { z0.h }, p0/z, [x0]
; VBITS_GE_256-NEXT:    ld1h { z1.h }, p0/z, [x1]
; VBITS_GE_256-NEXT:    sunpklo z2.s, z1.h
; VBITS_GE_256-NEXT:    sunpklo z3.s, z0.h
; VBITS_GE_256-NEXT:    mov z4.d, z0.d
; VBITS_GE_256-NEXT:    ext z4.b, z4.b, z0.b, #16
; VBITS_GE_256-NEXT:    sdivr z2.s, p1/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    mov z3.d, z1.d
; VBITS_GE_256-NEXT:    sunpklo z4.s, z4.h
; VBITS_GE_256-NEXT:    ext z3.b, z3.b, z1.b, #16
; VBITS_GE_256-NEXT:    sunpklo z3.s, z3.h
; VBITS_GE_256-NEXT:    sdivr z3.s, p1/m, z3.s, z4.s
; VBITS_GE_256-NEXT:    ptrue p1.h, vl8
; VBITS_GE_256-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_256-NEXT:    uzp1 z3.h, z3.h, z3.h
; VBITS_GE_256-NEXT:    splice z2.h, p1, z2.h, z3.h
; VBITS_GE_256-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; VBITS_GE_256-NEXT:    st1h { z0.h }, p0, [x0]
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: srem_v16i16:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.h, vl16
; VBITS_GE_512-NEXT:    ptrue p1.s, vl16
; VBITS_GE_512-NEXT:    ld1h { z0.h }, p0/z, [x0]
; VBITS_GE_512-NEXT:    ld1h { z1.h }, p0/z, [x1]
; VBITS_GE_512-NEXT:    sunpklo z2.s, z1.h
; VBITS_GE_512-NEXT:    sunpklo z3.s, z0.h
; VBITS_GE_512-NEXT:    sdivr z2.s, p1/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_512-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; VBITS_GE_512-NEXT:    st1h { z0.h }, p0, [x0]
; VBITS_GE_512-NEXT:    ret
  %op1 = load <16 x i16>, ptr %a
  %op2 = load <16 x i16>, ptr %b
  %res = srem <16 x i16> %op1, %op2
  store <16 x i16> %res, ptr %a
  ret void
}

define void @srem_v32i16(ptr %a, ptr %b) vscale_range(8,0) #0 {
; CHECK-LABEL: srem_v32i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.h, vl32
; CHECK-NEXT:    ptrue p1.s, vl32
; CHECK-NEXT:    ld1h { z0.h }, p0/z, [x0]
; CHECK-NEXT:    ld1h { z1.h }, p0/z, [x1]
; CHECK-NEXT:    sunpklo z2.s, z1.h
; CHECK-NEXT:    sunpklo z3.s, z0.h
; CHECK-NEXT:    sdivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; CHECK-NEXT:    st1h { z0.h }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <32 x i16>, ptr %a
  %op2 = load <32 x i16>, ptr %b
  %res = srem <32 x i16> %op1, %op2
  store <32 x i16> %res, ptr %a
  ret void
}

define void @srem_v64i16(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: srem_v64i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.h, vl64
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1h { z0.h }, p0/z, [x0]
; CHECK-NEXT:    ld1h { z1.h }, p0/z, [x1]
; CHECK-NEXT:    sunpklo z2.s, z1.h
; CHECK-NEXT:    sunpklo z3.s, z0.h
; CHECK-NEXT:    sdivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; CHECK-NEXT:    st1h { z0.h }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <64 x i16>, ptr %a
  %op2 = load <64 x i16>, ptr %b
  %res = srem <64 x i16> %op1, %op2
  store <64 x i16> %res, ptr %a
  ret void
}

define void @srem_v128i16(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: srem_v128i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.h, vl128
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1h { z0.h }, p0/z, [x0]
; CHECK-NEXT:    ld1h { z1.h }, p0/z, [x1]
; CHECK-NEXT:    sunpklo z2.s, z1.h
; CHECK-NEXT:    sunpklo z3.s, z0.h
; CHECK-NEXT:    mov z4.d, z0.d
; CHECK-NEXT:    ext z4.b, z4.b, z0.b, #128
; CHECK-NEXT:    sdivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    mov z3.d, z1.d
; CHECK-NEXT:    sunpklo z4.s, z4.h
; CHECK-NEXT:    ext z3.b, z3.b, z1.b, #128
; CHECK-NEXT:    sunpklo z3.s, z3.h
; CHECK-NEXT:    sdivr z3.s, p1/m, z3.s, z4.s
; CHECK-NEXT:    ptrue p1.h, vl64
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    uzp1 z3.h, z3.h, z3.h
; CHECK-NEXT:    splice z2.h, p1, z2.h, z3.h
; CHECK-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; CHECK-NEXT:    st1h { z0.h }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <128 x i16>, ptr %a
  %op2 = load <128 x i16>, ptr %b
  %res = srem <128 x i16> %op1, %op2
  store <128 x i16> %res, ptr %a
  ret void
}

; Vector v2i32 sdiv are not legal for NEON so use SVE when available.
define <2 x i32> @srem_v2i32(<2 x i32> %op1, <2 x i32> %op2) vscale_range(1,0) #0 {
; CHECK-LABEL: srem_v2i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    // kill: def $d0 killed $d0 def $z0
; CHECK-NEXT:    // kill: def $d1 killed $d1 def $z1
; CHECK-NEXT:    ptrue p0.s, vl2
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls v0.2s, v2.2s, v1.2s
; CHECK-NEXT:    // kill: def $d0 killed $d0 killed $z0
; CHECK-NEXT:    ret
  %res = srem <2 x i32> %op1, %op2
  ret <2 x i32> %res
}

; Vector v4i32 sdiv are not legal for NEON so use SVE when available.
define <4 x i32> @srem_v4i32(<4 x i32> %op1, <4 x i32> %op2) vscale_range(1,0) #0 {
; CHECK-LABEL: srem_v4i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    // kill: def $q0 killed $q0 def $z0
; CHECK-NEXT:    // kill: def $q1 killed $q1 def $z1
; CHECK-NEXT:    ptrue p0.s, vl4
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls v0.4s, v2.4s, v1.4s
; CHECK-NEXT:    // kill: def $q0 killed $q0 killed $z0
; CHECK-NEXT:    ret
  %res = srem <4 x i32> %op1, %op2
  ret <4 x i32> %res
}

define void @srem_v8i32(ptr %a, ptr %b) vscale_range(2,0) #0 {
; CHECK-LABEL: srem_v8i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.s, vl8
; CHECK-NEXT:    ld1w { z0.s }, p0/z, [x0]
; CHECK-NEXT:    ld1w { z1.s }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls z0.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    st1w { z0.s }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <8 x i32>, ptr %a
  %op2 = load <8 x i32>, ptr %b
  %res = srem <8 x i32> %op1, %op2
  store <8 x i32> %res, ptr %a
  ret void
}

define void @srem_v16i32(ptr %a, ptr %b) #0 {
; VBITS_GE_128-LABEL: srem_v16i32:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ldp q0, q3, [x1]
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    ldp q1, q2, [x0]
; VBITS_GE_128-NEXT:    movprfx z4, z1
; VBITS_GE_128-NEXT:    sdiv z4.s, p0/m, z4.s, z0.s
; VBITS_GE_128-NEXT:    ldp q16, q5, [x0, #32]
; VBITS_GE_128-NEXT:    movprfx z19, z2
; VBITS_GE_128-NEXT:    sdiv z19.s, p0/m, z19.s, z3.s
; VBITS_GE_128-NEXT:    ldp q17, q6, [x1, #32]
; VBITS_GE_128-NEXT:    movprfx z7, z5
; VBITS_GE_128-NEXT:    sdiv z7.s, p0/m, z7.s, z6.s
; VBITS_GE_128-NEXT:    movprfx z18, z16
; VBITS_GE_128-NEXT:    sdiv z18.s, p0/m, z18.s, z17.s
; VBITS_GE_128-NEXT:    mls v1.4s, v4.4s, v0.4s
; VBITS_GE_128-NEXT:    mls v2.4s, v19.4s, v3.4s
; VBITS_GE_128-NEXT:    mls v16.4s, v18.4s, v17.4s
; VBITS_GE_128-NEXT:    mls v5.4s, v7.4s, v6.4s
; VBITS_GE_128-NEXT:    stp q1, q2, [x0]
; VBITS_GE_128-NEXT:    stp q16, q5, [x0, #32]
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: srem_v16i32:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.s, vl8
; VBITS_GE_256-NEXT:    mov x8, #8 // =0x8
; VBITS_GE_256-NEXT:    ld1w { z0.s }, p0/z, [x0, x8, lsl #2]
; VBITS_GE_256-NEXT:    ld1w { z1.s }, p0/z, [x0]
; VBITS_GE_256-NEXT:    ld1w { z2.s }, p0/z, [x1, x8, lsl #2]
; VBITS_GE_256-NEXT:    ld1w { z4.s }, p0/z, [x1]
; VBITS_GE_256-NEXT:    movprfx z3, z0
; VBITS_GE_256-NEXT:    sdiv z3.s, p0/m, z3.s, z2.s
; VBITS_GE_256-NEXT:    movprfx z5, z1
; VBITS_GE_256-NEXT:    sdiv z5.s, p0/m, z5.s, z4.s
; VBITS_GE_256-NEXT:    mls z0.s, p0/m, z3.s, z2.s
; VBITS_GE_256-NEXT:    mls z1.s, p0/m, z5.s, z4.s
; VBITS_GE_256-NEXT:    st1w { z0.s }, p0, [x0, x8, lsl #2]
; VBITS_GE_256-NEXT:    st1w { z1.s }, p0, [x0]
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: srem_v16i32:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.s, vl16
; VBITS_GE_512-NEXT:    ld1w { z0.s }, p0/z, [x0]
; VBITS_GE_512-NEXT:    ld1w { z1.s }, p0/z, [x1]
; VBITS_GE_512-NEXT:    movprfx z2, z0
; VBITS_GE_512-NEXT:    sdiv z2.s, p0/m, z2.s, z1.s
; VBITS_GE_512-NEXT:    mls z0.s, p0/m, z2.s, z1.s
; VBITS_GE_512-NEXT:    st1w { z0.s }, p0, [x0]
; VBITS_GE_512-NEXT:    ret
  %op1 = load <16 x i32>, ptr %a
  %op2 = load <16 x i32>, ptr %b
  %res = srem <16 x i32> %op1, %op2
  store <16 x i32> %res, ptr %a
  ret void
}

define void @srem_v32i32(ptr %a, ptr %b) vscale_range(8,0) #0 {
; CHECK-LABEL: srem_v32i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.s, vl32
; CHECK-NEXT:    ld1w { z0.s }, p0/z, [x0]
; CHECK-NEXT:    ld1w { z1.s }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls z0.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    st1w { z0.s }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <32 x i32>, ptr %a
  %op2 = load <32 x i32>, ptr %b
  %res = srem <32 x i32> %op1, %op2
  store <32 x i32> %res, ptr %a
  ret void
}

define void @srem_v64i32(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: srem_v64i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.s, vl64
; CHECK-NEXT:    ld1w { z0.s }, p0/z, [x0]
; CHECK-NEXT:    ld1w { z1.s }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls z0.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    st1w { z0.s }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <64 x i32>, ptr %a
  %op2 = load <64 x i32>, ptr %b
  %res = srem <64 x i32> %op1, %op2
  store <64 x i32> %res, ptr %a
  ret void
}

; Vector i64 sdiv are not legal for NEON so use SVE when available.
; FIXME: We should be able to improve the codegen for the 128 bits case here.
define <1 x i64> @srem_v1i64(<1 x i64> %op1, <1 x i64> %op2) vscale_range(1,0) #0 {
; CHECK-LABEL: srem_v1i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl1
; CHECK-NEXT:    // kill: def $d0 killed $d0 def $z0
; CHECK-NEXT:    // kill: def $d1 killed $d1 def $z1
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    // kill: def $d0 killed $d0 killed $z0
; CHECK-NEXT:    ret
  %res = srem <1 x i64> %op1, %op2
  ret <1 x i64> %res
}

; Vector i64 sdiv are not legal for NEON so use SVE when available.
; FIXME: We should be able to improve the codegen for the 128 bits case here.
define <2 x i64> @srem_v2i64(<2 x i64> %op1, <2 x i64> %op2) vscale_range(1,0) #0 {
; CHECK-LABEL: srem_v2i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl2
; CHECK-NEXT:    // kill: def $q0 killed $q0 def $z0
; CHECK-NEXT:    // kill: def $q1 killed $q1 def $z1
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    // kill: def $q0 killed $q0 killed $z0
; CHECK-NEXT:    ret
  %res = srem <2 x i64> %op1, %op2
  ret <2 x i64> %res
}

define void @srem_v4i64(ptr %a, ptr %b) vscale_range(2,0) #0 {
; CHECK-LABEL: srem_v4i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl4
; CHECK-NEXT:    ld1d { z0.d }, p0/z, [x0]
; CHECK-NEXT:    ld1d { z1.d }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    st1d { z0.d }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <4 x i64>, ptr %a
  %op2 = load <4 x i64>, ptr %b
  %res = srem <4 x i64> %op1, %op2
  store <4 x i64> %res, ptr %a
  ret void
}

define void @srem_v8i64(ptr %a, ptr %b) #0 {
; VBITS_GE_128-LABEL: srem_v8i64:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.d, vl2
; VBITS_GE_128-NEXT:    ldp q0, q3, [x1]
; VBITS_GE_128-NEXT:    ldp q1, q2, [x0]
; VBITS_GE_128-NEXT:    movprfx z4, z1
; VBITS_GE_128-NEXT:    sdiv z4.d, p0/m, z4.d, z0.d
; VBITS_GE_128-NEXT:    ldp q16, q5, [x0, #32]
; VBITS_GE_128-NEXT:    movprfx z19, z2
; VBITS_GE_128-NEXT:    sdiv z19.d, p0/m, z19.d, z3.d
; VBITS_GE_128-NEXT:    ldp q17, q6, [x1, #32]
; VBITS_GE_128-NEXT:    movprfx z7, z5
; VBITS_GE_128-NEXT:    sdiv z7.d, p0/m, z7.d, z6.d
; VBITS_GE_128-NEXT:    movprfx z18, z16
; VBITS_GE_128-NEXT:    sdiv z18.d, p0/m, z18.d, z17.d
; VBITS_GE_128-NEXT:    mls z16.d, p0/m, z18.d, z17.d
; VBITS_GE_128-NEXT:    mls z5.d, p0/m, z7.d, z6.d
; VBITS_GE_128-NEXT:    msb z0.d, p0/m, z4.d, z1.d
; VBITS_GE_128-NEXT:    movprfx z1, z2
; VBITS_GE_128-NEXT:    mls z1.d, p0/m, z19.d, z3.d
; VBITS_GE_128-NEXT:    stp q16, q5, [x0, #32]
; VBITS_GE_128-NEXT:    stp q0, q1, [x0]
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: srem_v8i64:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.d, vl4
; VBITS_GE_256-NEXT:    mov x8, #4 // =0x4
; VBITS_GE_256-NEXT:    ld1d { z0.d }, p0/z, [x0, x8, lsl #3]
; VBITS_GE_256-NEXT:    ld1d { z1.d }, p0/z, [x0]
; VBITS_GE_256-NEXT:    ld1d { z2.d }, p0/z, [x1, x8, lsl #3]
; VBITS_GE_256-NEXT:    ld1d { z4.d }, p0/z, [x1]
; VBITS_GE_256-NEXT:    movprfx z3, z0
; VBITS_GE_256-NEXT:    sdiv z3.d, p0/m, z3.d, z2.d
; VBITS_GE_256-NEXT:    movprfx z5, z1
; VBITS_GE_256-NEXT:    sdiv z5.d, p0/m, z5.d, z4.d
; VBITS_GE_256-NEXT:    mls z0.d, p0/m, z3.d, z2.d
; VBITS_GE_256-NEXT:    mls z1.d, p0/m, z5.d, z4.d
; VBITS_GE_256-NEXT:    st1d { z0.d }, p0, [x0, x8, lsl #3]
; VBITS_GE_256-NEXT:    st1d { z1.d }, p0, [x0]
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: srem_v8i64:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.d, vl8
; VBITS_GE_512-NEXT:    ld1d { z0.d }, p0/z, [x0]
; VBITS_GE_512-NEXT:    ld1d { z1.d }, p0/z, [x1]
; VBITS_GE_512-NEXT:    movprfx z2, z0
; VBITS_GE_512-NEXT:    sdiv z2.d, p0/m, z2.d, z1.d
; VBITS_GE_512-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; VBITS_GE_512-NEXT:    st1d { z0.d }, p0, [x0]
; VBITS_GE_512-NEXT:    ret
  %op1 = load <8 x i64>, ptr %a
  %op2 = load <8 x i64>, ptr %b
  %res = srem <8 x i64> %op1, %op2
  store <8 x i64> %res, ptr %a
  ret void
}

define void @srem_v16i64(ptr %a, ptr %b) vscale_range(8,0) #0 {
; CHECK-LABEL: srem_v16i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl16
; CHECK-NEXT:    ld1d { z0.d }, p0/z, [x0]
; CHECK-NEXT:    ld1d { z1.d }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    st1d { z0.d }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <16 x i64>, ptr %a
  %op2 = load <16 x i64>, ptr %b
  %res = srem <16 x i64> %op1, %op2
  store <16 x i64> %res, ptr %a
  ret void
}

define void @srem_v32i64(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: srem_v32i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl32
; CHECK-NEXT:    ld1d { z0.d }, p0/z, [x0]
; CHECK-NEXT:    ld1d { z1.d }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    sdiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    st1d { z0.d }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <32 x i64>, ptr %a
  %op2 = load <32 x i64>, ptr %b
  %res = srem <32 x i64> %op1, %op2
  store <32 x i64> %res, ptr %a
  ret void
}

;
; UREM
;

; Vector vXi8 udiv are not legal for NEON so use SVE when available.
; FIXME: We should be able to improve the codegen for >= 256 bits here.
define <8 x i8> @urem_v8i8(<8 x i8> %op1, <8 x i8> %op2) #0 {
; VBITS_GE_128-LABEL: urem_v8i8:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    ushll v2.8h, v1.8b, #0
; VBITS_GE_128-NEXT:    ushll v3.8h, v0.8b, #0
; VBITS_GE_128-NEXT:    ushll2 v4.4s, v2.8h, #0
; VBITS_GE_128-NEXT:    ushll2 v5.4s, v3.8h, #0
; VBITS_GE_128-NEXT:    ushll v2.4s, v2.4h, #0
; VBITS_GE_128-NEXT:    ushll v3.4s, v3.4h, #0
; VBITS_GE_128-NEXT:    udivr z4.s, p0/m, z4.s, z5.s
; VBITS_GE_128-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    uzp1 v2.8h, v2.8h, v4.8h
; VBITS_GE_128-NEXT:    xtn v2.8b, v2.8h
; VBITS_GE_128-NEXT:    mls v0.8b, v2.8b, v1.8b
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: urem_v8i8:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    // kill: def $d1 killed $d1 def $z1
; VBITS_GE_256-NEXT:    // kill: def $d0 killed $d0 def $z0
; VBITS_GE_256-NEXT:    uunpklo z2.h, z1.b
; VBITS_GE_256-NEXT:    uunpklo z3.h, z0.b
; VBITS_GE_256-NEXT:    ptrue p0.s, vl8
; VBITS_GE_256-NEXT:    uunpklo z2.s, z2.h
; VBITS_GE_256-NEXT:    uunpklo z3.s, z3.h
; VBITS_GE_256-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_256-NEXT:    umov w8, v2.h[0]
; VBITS_GE_256-NEXT:    umov w9, v2.h[1]
; VBITS_GE_256-NEXT:    fmov s3, w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[2]
; VBITS_GE_256-NEXT:    mov v3.b[1], w9
; VBITS_GE_256-NEXT:    mov v3.b[2], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[3]
; VBITS_GE_256-NEXT:    mov v3.b[3], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[4]
; VBITS_GE_256-NEXT:    mov v3.b[4], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[5]
; VBITS_GE_256-NEXT:    mov v3.b[5], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[6]
; VBITS_GE_256-NEXT:    mov v3.b[6], w8
; VBITS_GE_256-NEXT:    umov w8, v2.h[7]
; VBITS_GE_256-NEXT:    mov v3.b[7], w8
; VBITS_GE_256-NEXT:    mls v0.8b, v3.8b, v1.8b
; VBITS_GE_256-NEXT:    // kill: def $d0 killed $d0 killed $z0
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: urem_v8i8:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    // kill: def $d1 killed $d1 def $z1
; VBITS_GE_512-NEXT:    // kill: def $d0 killed $d0 def $z0
; VBITS_GE_512-NEXT:    uunpklo z2.h, z1.b
; VBITS_GE_512-NEXT:    uunpklo z3.h, z0.b
; VBITS_GE_512-NEXT:    ptrue p0.s, vl8
; VBITS_GE_512-NEXT:    uunpklo z2.s, z2.h
; VBITS_GE_512-NEXT:    uunpklo z3.s, z3.h
; VBITS_GE_512-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_512-NEXT:    umov w8, v2.h[0]
; VBITS_GE_512-NEXT:    umov w9, v2.h[1]
; VBITS_GE_512-NEXT:    fmov s3, w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[2]
; VBITS_GE_512-NEXT:    mov v3.b[1], w9
; VBITS_GE_512-NEXT:    mov v3.b[2], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[3]
; VBITS_GE_512-NEXT:    mov v3.b[3], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[4]
; VBITS_GE_512-NEXT:    mov v3.b[4], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[5]
; VBITS_GE_512-NEXT:    mov v3.b[5], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[6]
; VBITS_GE_512-NEXT:    mov v3.b[6], w8
; VBITS_GE_512-NEXT:    umov w8, v2.h[7]
; VBITS_GE_512-NEXT:    mov v3.b[7], w8
; VBITS_GE_512-NEXT:    mls v0.8b, v3.8b, v1.8b
; VBITS_GE_512-NEXT:    // kill: def $d0 killed $d0 killed $z0
; VBITS_GE_512-NEXT:    ret
  %res = urem <8 x i8> %op1, %op2
  ret <8 x i8> %res
}

define <16 x i8> @urem_v16i8(<16 x i8> %op1, <16 x i8> %op2) #0 {
; VBITS_GE_128-LABEL: urem_v16i8:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    ushll2 v2.8h, v1.16b, #0
; VBITS_GE_128-NEXT:    ushll2 v3.8h, v0.16b, #0
; VBITS_GE_128-NEXT:    ushll2 v4.4s, v2.8h, #0
; VBITS_GE_128-NEXT:    ushll2 v5.4s, v3.8h, #0
; VBITS_GE_128-NEXT:    ushll v2.4s, v2.4h, #0
; VBITS_GE_128-NEXT:    ushll v3.4s, v3.4h, #0
; VBITS_GE_128-NEXT:    udivr z4.s, p0/m, z4.s, z5.s
; VBITS_GE_128-NEXT:    ushll v5.8h, v0.8b, #0
; VBITS_GE_128-NEXT:    ushll2 v7.4s, v5.8h, #0
; VBITS_GE_128-NEXT:    ushll v5.4s, v5.4h, #0
; VBITS_GE_128-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    ushll v3.8h, v1.8b, #0
; VBITS_GE_128-NEXT:    ushll2 v6.4s, v3.8h, #0
; VBITS_GE_128-NEXT:    ushll v3.4s, v3.4h, #0
; VBITS_GE_128-NEXT:    udivr z6.s, p0/m, z6.s, z7.s
; VBITS_GE_128-NEXT:    uzp1 v2.8h, v2.8h, v4.8h
; VBITS_GE_128-NEXT:    udivr z3.s, p0/m, z3.s, z5.s
; VBITS_GE_128-NEXT:    uzp1 v3.8h, v3.8h, v6.8h
; VBITS_GE_128-NEXT:    uzp1 v2.16b, v3.16b, v2.16b
; VBITS_GE_128-NEXT:    mls v0.16b, v2.16b, v1.16b
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: urem_v16i8:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    // kill: def $q1 killed $q1 def $z1
; VBITS_GE_256-NEXT:    // kill: def $q0 killed $q0 def $z0
; VBITS_GE_256-NEXT:    uunpklo z2.h, z1.b
; VBITS_GE_256-NEXT:    uunpklo z3.h, z0.b
; VBITS_GE_256-NEXT:    ptrue p0.s, vl8
; VBITS_GE_256-NEXT:    uunpklo z4.s, z2.h
; VBITS_GE_256-NEXT:    uunpklo z5.s, z3.h
; VBITS_GE_256-NEXT:    ext z2.b, z2.b, z2.b, #16
; VBITS_GE_256-NEXT:    ext z3.b, z3.b, z3.b, #16
; VBITS_GE_256-NEXT:    uunpklo z2.s, z2.h
; VBITS_GE_256-NEXT:    uunpklo z3.s, z3.h
; VBITS_GE_256-NEXT:    udivr z4.s, p0/m, z4.s, z5.s
; VBITS_GE_256-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    ptrue p0.h, vl8
; VBITS_GE_256-NEXT:    uzp1 z3.h, z4.h, z4.h
; VBITS_GE_256-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_256-NEXT:    splice z3.h, p0, z3.h, z2.h
; VBITS_GE_256-NEXT:    uzp1 z2.b, z3.b, z3.b
; VBITS_GE_256-NEXT:    mls v0.16b, v2.16b, v1.16b
; VBITS_GE_256-NEXT:    // kill: def $q0 killed $q0 killed $z0
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: urem_v16i8:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    // kill: def $q1 killed $q1 def $z1
; VBITS_GE_512-NEXT:    // kill: def $q0 killed $q0 def $z0
; VBITS_GE_512-NEXT:    uunpklo z2.h, z1.b
; VBITS_GE_512-NEXT:    uunpklo z3.h, z0.b
; VBITS_GE_512-NEXT:    ptrue p0.s, vl16
; VBITS_GE_512-NEXT:    uunpklo z2.s, z2.h
; VBITS_GE_512-NEXT:    uunpklo z3.s, z3.h
; VBITS_GE_512-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_512-NEXT:    uzp1 z2.b, z2.b, z2.b
; VBITS_GE_512-NEXT:    mls v0.16b, v2.16b, v1.16b
; VBITS_GE_512-NEXT:    // kill: def $q0 killed $q0 killed $z0
; VBITS_GE_512-NEXT:    ret
  %res = urem <16 x i8> %op1, %op2
  ret <16 x i8> %res
}

define void @urem_v32i8(ptr %a, ptr %b) vscale_range(8,0) #0 {
; CHECK-LABEL: urem_v32i8:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.b, vl32
; CHECK-NEXT:    ptrue p1.s, vl32
; CHECK-NEXT:    ld1b { z0.b }, p0/z, [x0]
; CHECK-NEXT:    ld1b { z1.b }, p0/z, [x1]
; CHECK-NEXT:    uunpklo z2.h, z1.b
; CHECK-NEXT:    uunpklo z3.h, z0.b
; CHECK-NEXT:    uunpklo z2.s, z2.h
; CHECK-NEXT:    uunpklo z3.s, z3.h
; CHECK-NEXT:    udivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    uzp1 z2.b, z2.b, z2.b
; CHECK-NEXT:    mls z0.b, p0/m, z2.b, z1.b
; CHECK-NEXT:    st1b { z0.b }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <32 x i8>, ptr %a
  %op2 = load <32 x i8>, ptr %b
  %res = urem <32 x i8> %op1, %op2
  store <32 x i8> %res, ptr %a
  ret void
}

define void @urem_v64i8(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: urem_v64i8:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.b, vl64
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1b { z0.b }, p0/z, [x0]
; CHECK-NEXT:    ld1b { z1.b }, p0/z, [x1]
; CHECK-NEXT:    uunpklo z2.h, z1.b
; CHECK-NEXT:    uunpklo z3.h, z0.b
; CHECK-NEXT:    uunpklo z2.s, z2.h
; CHECK-NEXT:    uunpklo z3.s, z3.h
; CHECK-NEXT:    udivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    uzp1 z2.b, z2.b, z2.b
; CHECK-NEXT:    mls z0.b, p0/m, z2.b, z1.b
; CHECK-NEXT:    st1b { z0.b }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <64 x i8>, ptr %a
  %op2 = load <64 x i8>, ptr %b
  %res = urem <64 x i8> %op1, %op2
  store <64 x i8> %res, ptr %a
  ret void
}

define void @urem_v128i8(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: urem_v128i8:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.b, vl128
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1b { z0.b }, p0/z, [x0]
; CHECK-NEXT:    ld1b { z1.b }, p0/z, [x1]
; CHECK-NEXT:    uunpklo z2.h, z1.b
; CHECK-NEXT:    uunpklo z3.h, z0.b
; CHECK-NEXT:    uunpklo z4.s, z2.h
; CHECK-NEXT:    uunpklo z5.s, z3.h
; CHECK-NEXT:    ext z2.b, z2.b, z2.b, #128
; CHECK-NEXT:    ext z3.b, z3.b, z3.b, #128
; CHECK-NEXT:    uunpklo z2.s, z2.h
; CHECK-NEXT:    uunpklo z3.s, z3.h
; CHECK-NEXT:    udivr z4.s, p1/m, z4.s, z5.s
; CHECK-NEXT:    udivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    ptrue p1.h, vl64
; CHECK-NEXT:    uzp1 z3.h, z4.h, z4.h
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    splice z3.h, p1, z3.h, z2.h
; CHECK-NEXT:    uzp1 z2.b, z3.b, z3.b
; CHECK-NEXT:    mls z0.b, p0/m, z2.b, z1.b
; CHECK-NEXT:    st1b { z0.b }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <128 x i8>, ptr %a
  %op2 = load <128 x i8>, ptr %b
  %res = urem <128 x i8> %op1, %op2
  store <128 x i8> %res, ptr %a
  ret void
}

define void @urem_v256i8(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: urem_v256i8:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.b, vl256
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1b { z0.b }, p0/z, [x0]
; CHECK-NEXT:    ld1b { z1.b }, p0/z, [x1]
; CHECK-NEXT:    uunpklo z2.h, z1.b
; CHECK-NEXT:    uunpklo z3.h, z0.b
; CHECK-NEXT:    uunpklo z4.s, z2.h
; CHECK-NEXT:    uunpklo z5.s, z3.h
; CHECK-NEXT:    ext z2.b, z2.b, z2.b, #128
; CHECK-NEXT:    ext z3.b, z3.b, z3.b, #128
; CHECK-NEXT:    uunpklo z2.s, z2.h
; CHECK-NEXT:    uunpklo z3.s, z3.h
; CHECK-NEXT:    udivr z4.s, p1/m, z4.s, z5.s
; CHECK-NEXT:    mov z5.d, z0.d
; CHECK-NEXT:    ext z5.b, z5.b, z0.b, #128
; CHECK-NEXT:    uunpklo z5.h, z5.b
; CHECK-NEXT:    uunpklo z7.s, z5.h
; CHECK-NEXT:    ext z5.b, z5.b, z5.b, #128
; CHECK-NEXT:    udivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    mov z3.d, z1.d
; CHECK-NEXT:    uunpklo z5.s, z5.h
; CHECK-NEXT:    ext z3.b, z3.b, z1.b, #128
; CHECK-NEXT:    uzp1 z4.h, z4.h, z4.h
; CHECK-NEXT:    uunpklo z3.h, z3.b
; CHECK-NEXT:    uunpklo z6.s, z3.h
; CHECK-NEXT:    ext z3.b, z3.b, z3.b, #128
; CHECK-NEXT:    uunpklo z3.s, z3.h
; CHECK-NEXT:    udivr z6.s, p1/m, z6.s, z7.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    udivr z3.s, p1/m, z3.s, z5.s
; CHECK-NEXT:    ptrue p1.h, vl64
; CHECK-NEXT:    uzp1 z5.h, z6.h, z6.h
; CHECK-NEXT:    splice z4.h, p1, z4.h, z2.h
; CHECK-NEXT:    uzp1 z2.b, z4.b, z4.b
; CHECK-NEXT:    uzp1 z3.h, z3.h, z3.h
; CHECK-NEXT:    splice z5.h, p1, z5.h, z3.h
; CHECK-NEXT:    ptrue p1.b, vl128
; CHECK-NEXT:    uzp1 z3.b, z5.b, z5.b
; CHECK-NEXT:    splice z2.b, p1, z2.b, z3.b
; CHECK-NEXT:    mls z0.b, p0/m, z2.b, z1.b
; CHECK-NEXT:    st1b { z0.b }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <256 x i8>, ptr %a
  %op2 = load <256 x i8>, ptr %b
  %res = urem <256 x i8> %op1, %op2
  store <256 x i8> %res, ptr %a
  ret void
}

; Vector vXi16 udiv are not legal for NEON so use SVE when available.
; FIXME: We should be able to improve the codegen for >= 256 bits here.
define <4 x i16> @urem_v4i16(<4 x i16> %op1, <4 x i16> %op2) #0 {
; VBITS_GE_128-LABEL: urem_v4i16:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    ushll v2.4s, v1.4h, #0
; VBITS_GE_128-NEXT:    ushll v3.4s, v0.4h, #0
; VBITS_GE_128-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    xtn v2.4h, v2.4s
; VBITS_GE_128-NEXT:    mls v0.4h, v2.4h, v1.4h
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: urem_v4i16:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.s, vl4
; VBITS_GE_256-NEXT:    ushll v2.4s, v1.4h, #0
; VBITS_GE_256-NEXT:    ushll v3.4s, v0.4h, #0
; VBITS_GE_256-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    mov w8, v2.s[1]
; VBITS_GE_256-NEXT:    mov v3.16b, v2.16b
; VBITS_GE_256-NEXT:    mov w9, v2.s[2]
; VBITS_GE_256-NEXT:    mov v3.h[1], w8
; VBITS_GE_256-NEXT:    mov w8, v2.s[3]
; VBITS_GE_256-NEXT:    mov v3.h[2], w9
; VBITS_GE_256-NEXT:    mov v3.h[3], w8
; VBITS_GE_256-NEXT:    mls v0.4h, v3.4h, v1.4h
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: urem_v4i16:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.s, vl4
; VBITS_GE_512-NEXT:    ushll v2.4s, v1.4h, #0
; VBITS_GE_512-NEXT:    ushll v3.4s, v0.4h, #0
; VBITS_GE_512-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    mov w8, v2.s[1]
; VBITS_GE_512-NEXT:    mov v3.16b, v2.16b
; VBITS_GE_512-NEXT:    mov w9, v2.s[2]
; VBITS_GE_512-NEXT:    mov v3.h[1], w8
; VBITS_GE_512-NEXT:    mov w8, v2.s[3]
; VBITS_GE_512-NEXT:    mov v3.h[2], w9
; VBITS_GE_512-NEXT:    mov v3.h[3], w8
; VBITS_GE_512-NEXT:    mls v0.4h, v3.4h, v1.4h
; VBITS_GE_512-NEXT:    ret
  %res = urem <4 x i16> %op1, %op2
  ret <4 x i16> %res
}

define <8 x i16> @urem_v8i16(<8 x i16> %op1, <8 x i16> %op2) #0 {
; VBITS_GE_128-LABEL: urem_v8i16:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    ushll2 v2.4s, v1.8h, #0
; VBITS_GE_128-NEXT:    ushll2 v3.4s, v0.8h, #0
; VBITS_GE_128-NEXT:    ushll v4.4s, v0.4h, #0
; VBITS_GE_128-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    ushll v3.4s, v1.4h, #0
; VBITS_GE_128-NEXT:    udivr z3.s, p0/m, z3.s, z4.s
; VBITS_GE_128-NEXT:    uzp1 v2.8h, v3.8h, v2.8h
; VBITS_GE_128-NEXT:    mls v0.8h, v2.8h, v1.8h
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: urem_v8i16:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.s, vl8
; VBITS_GE_256-NEXT:    // kill: def $q1 killed $q1 def $z1
; VBITS_GE_256-NEXT:    // kill: def $q0 killed $q0 def $z0
; VBITS_GE_256-NEXT:    uunpklo z2.s, z1.h
; VBITS_GE_256-NEXT:    uunpklo z3.s, z0.h
; VBITS_GE_256-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_256-NEXT:    mls v0.8h, v2.8h, v1.8h
; VBITS_GE_256-NEXT:    // kill: def $q0 killed $q0 killed $z0
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: urem_v8i16:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.s, vl8
; VBITS_GE_512-NEXT:    // kill: def $q1 killed $q1 def $z1
; VBITS_GE_512-NEXT:    // kill: def $q0 killed $q0 def $z0
; VBITS_GE_512-NEXT:    uunpklo z2.s, z1.h
; VBITS_GE_512-NEXT:    uunpklo z3.s, z0.h
; VBITS_GE_512-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_512-NEXT:    mls v0.8h, v2.8h, v1.8h
; VBITS_GE_512-NEXT:    // kill: def $q0 killed $q0 killed $z0
; VBITS_GE_512-NEXT:    ret
  %res = urem <8 x i16> %op1, %op2
  ret <8 x i16> %res
}

define void @urem_v16i16(ptr %a, ptr %b) #0 {
; VBITS_GE_128-LABEL: urem_v16i16:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    ldp q4, q1, [x1]
; VBITS_GE_128-NEXT:    ldr q0, [x0, #16]
; VBITS_GE_128-NEXT:    ushll2 v2.4s, v1.8h, #0
; VBITS_GE_128-NEXT:    ushll2 v3.4s, v0.8h, #0
; VBITS_GE_128-NEXT:    ushll2 v5.4s, v4.8h, #0
; VBITS_GE_128-NEXT:    ushll v16.4s, v0.4h, #0
; VBITS_GE_128-NEXT:    udivr z2.s, p0/m, z2.s, z3.s
; VBITS_GE_128-NEXT:    ldr q3, [x0]
; VBITS_GE_128-NEXT:    ushll2 v6.4s, v3.8h, #0
; VBITS_GE_128-NEXT:    ushll v7.4s, v3.4h, #0
; VBITS_GE_128-NEXT:    udivr z5.s, p0/m, z5.s, z6.s
; VBITS_GE_128-NEXT:    ushll v6.4s, v4.4h, #0
; VBITS_GE_128-NEXT:    udivr z6.s, p0/m, z6.s, z7.s
; VBITS_GE_128-NEXT:    ushll v7.4s, v1.4h, #0
; VBITS_GE_128-NEXT:    udivr z7.s, p0/m, z7.s, z16.s
; VBITS_GE_128-NEXT:    uzp1 v5.8h, v6.8h, v5.8h
; VBITS_GE_128-NEXT:    mls v3.8h, v5.8h, v4.8h
; VBITS_GE_128-NEXT:    uzp1 v2.8h, v7.8h, v2.8h
; VBITS_GE_128-NEXT:    mls v0.8h, v2.8h, v1.8h
; VBITS_GE_128-NEXT:    stp q3, q0, [x0]
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: urem_v16i16:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.h, vl16
; VBITS_GE_256-NEXT:    ptrue p1.s, vl8
; VBITS_GE_256-NEXT:    ld1h { z0.h }, p0/z, [x0]
; VBITS_GE_256-NEXT:    ld1h { z1.h }, p0/z, [x1]
; VBITS_GE_256-NEXT:    uunpklo z2.s, z1.h
; VBITS_GE_256-NEXT:    uunpklo z3.s, z0.h
; VBITS_GE_256-NEXT:    mov z4.d, z0.d
; VBITS_GE_256-NEXT:    ext z4.b, z4.b, z0.b, #16
; VBITS_GE_256-NEXT:    udivr z2.s, p1/m, z2.s, z3.s
; VBITS_GE_256-NEXT:    mov z3.d, z1.d
; VBITS_GE_256-NEXT:    uunpklo z4.s, z4.h
; VBITS_GE_256-NEXT:    ext z3.b, z3.b, z1.b, #16
; VBITS_GE_256-NEXT:    uunpklo z3.s, z3.h
; VBITS_GE_256-NEXT:    udivr z3.s, p1/m, z3.s, z4.s
; VBITS_GE_256-NEXT:    ptrue p1.h, vl8
; VBITS_GE_256-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_256-NEXT:    uzp1 z3.h, z3.h, z3.h
; VBITS_GE_256-NEXT:    splice z2.h, p1, z2.h, z3.h
; VBITS_GE_256-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; VBITS_GE_256-NEXT:    st1h { z0.h }, p0, [x0]
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: urem_v16i16:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.h, vl16
; VBITS_GE_512-NEXT:    ptrue p1.s, vl16
; VBITS_GE_512-NEXT:    ld1h { z0.h }, p0/z, [x0]
; VBITS_GE_512-NEXT:    ld1h { z1.h }, p0/z, [x1]
; VBITS_GE_512-NEXT:    uunpklo z2.s, z1.h
; VBITS_GE_512-NEXT:    uunpklo z3.s, z0.h
; VBITS_GE_512-NEXT:    udivr z2.s, p1/m, z2.s, z3.s
; VBITS_GE_512-NEXT:    uzp1 z2.h, z2.h, z2.h
; VBITS_GE_512-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; VBITS_GE_512-NEXT:    st1h { z0.h }, p0, [x0]
; VBITS_GE_512-NEXT:    ret
  %op1 = load <16 x i16>, ptr %a
  %op2 = load <16 x i16>, ptr %b
  %res = urem <16 x i16> %op1, %op2
  store <16 x i16> %res, ptr %a
  ret void
}

define void @urem_v32i16(ptr %a, ptr %b) vscale_range(8,0) #0 {
; CHECK-LABEL: urem_v32i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.h, vl32
; CHECK-NEXT:    ptrue p1.s, vl32
; CHECK-NEXT:    ld1h { z0.h }, p0/z, [x0]
; CHECK-NEXT:    ld1h { z1.h }, p0/z, [x1]
; CHECK-NEXT:    uunpklo z2.s, z1.h
; CHECK-NEXT:    uunpklo z3.s, z0.h
; CHECK-NEXT:    udivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; CHECK-NEXT:    st1h { z0.h }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <32 x i16>, ptr %a
  %op2 = load <32 x i16>, ptr %b
  %res = urem <32 x i16> %op1, %op2
  store <32 x i16> %res, ptr %a
  ret void
}

define void @urem_v64i16(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: urem_v64i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.h, vl64
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1h { z0.h }, p0/z, [x0]
; CHECK-NEXT:    ld1h { z1.h }, p0/z, [x1]
; CHECK-NEXT:    uunpklo z2.s, z1.h
; CHECK-NEXT:    uunpklo z3.s, z0.h
; CHECK-NEXT:    udivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; CHECK-NEXT:    st1h { z0.h }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <64 x i16>, ptr %a
  %op2 = load <64 x i16>, ptr %b
  %res = urem <64 x i16> %op1, %op2
  store <64 x i16> %res, ptr %a
  ret void
}

define void @urem_v128i16(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: urem_v128i16:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.h, vl128
; CHECK-NEXT:    ptrue p1.s, vl64
; CHECK-NEXT:    ld1h { z0.h }, p0/z, [x0]
; CHECK-NEXT:    ld1h { z1.h }, p0/z, [x1]
; CHECK-NEXT:    uunpklo z2.s, z1.h
; CHECK-NEXT:    uunpklo z3.s, z0.h
; CHECK-NEXT:    mov z4.d, z0.d
; CHECK-NEXT:    ext z4.b, z4.b, z0.b, #128
; CHECK-NEXT:    udivr z2.s, p1/m, z2.s, z3.s
; CHECK-NEXT:    mov z3.d, z1.d
; CHECK-NEXT:    uunpklo z4.s, z4.h
; CHECK-NEXT:    ext z3.b, z3.b, z1.b, #128
; CHECK-NEXT:    uunpklo z3.s, z3.h
; CHECK-NEXT:    udivr z3.s, p1/m, z3.s, z4.s
; CHECK-NEXT:    ptrue p1.h, vl64
; CHECK-NEXT:    uzp1 z2.h, z2.h, z2.h
; CHECK-NEXT:    uzp1 z3.h, z3.h, z3.h
; CHECK-NEXT:    splice z2.h, p1, z2.h, z3.h
; CHECK-NEXT:    mls z0.h, p0/m, z2.h, z1.h
; CHECK-NEXT:    st1h { z0.h }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <128 x i16>, ptr %a
  %op2 = load <128 x i16>, ptr %b
  %res = urem <128 x i16> %op1, %op2
  store <128 x i16> %res, ptr %a
  ret void
}

; Vector v2i32 udiv are not legal for NEON so use SVE when available.
define <2 x i32> @urem_v2i32(<2 x i32> %op1, <2 x i32> %op2) vscale_range(1,0) #0 {
; CHECK-LABEL: urem_v2i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    // kill: def $d0 killed $d0 def $z0
; CHECK-NEXT:    // kill: def $d1 killed $d1 def $z1
; CHECK-NEXT:    ptrue p0.s, vl2
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls v0.2s, v2.2s, v1.2s
; CHECK-NEXT:    // kill: def $d0 killed $d0 killed $z0
; CHECK-NEXT:    ret
  %res = urem <2 x i32> %op1, %op2
  ret <2 x i32> %res
}

; Vector v4i32 udiv are not legal for NEON so use SVE when available.
define <4 x i32> @urem_v4i32(<4 x i32> %op1, <4 x i32> %op2) vscale_range(1,0) #0 {
; CHECK-LABEL: urem_v4i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    // kill: def $q0 killed $q0 def $z0
; CHECK-NEXT:    // kill: def $q1 killed $q1 def $z1
; CHECK-NEXT:    ptrue p0.s, vl4
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls v0.4s, v2.4s, v1.4s
; CHECK-NEXT:    // kill: def $q0 killed $q0 killed $z0
; CHECK-NEXT:    ret
  %res = urem <4 x i32> %op1, %op2
  ret <4 x i32> %res
}

define void @urem_v8i32(ptr %a, ptr %b) vscale_range(2,0) #0 {
; CHECK-LABEL: urem_v8i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.s, vl8
; CHECK-NEXT:    ld1w { z0.s }, p0/z, [x0]
; CHECK-NEXT:    ld1w { z1.s }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls z0.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    st1w { z0.s }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <8 x i32>, ptr %a
  %op2 = load <8 x i32>, ptr %b
  %res = urem <8 x i32> %op1, %op2
  store <8 x i32> %res, ptr %a
  ret void
}

define void @urem_v16i32(ptr %a, ptr %b) #0 {
; VBITS_GE_128-LABEL: urem_v16i32:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ldp q0, q3, [x1]
; VBITS_GE_128-NEXT:    ptrue p0.s, vl4
; VBITS_GE_128-NEXT:    ldp q1, q2, [x0]
; VBITS_GE_128-NEXT:    movprfx z4, z1
; VBITS_GE_128-NEXT:    udiv z4.s, p0/m, z4.s, z0.s
; VBITS_GE_128-NEXT:    ldp q16, q5, [x0, #32]
; VBITS_GE_128-NEXT:    movprfx z19, z2
; VBITS_GE_128-NEXT:    udiv z19.s, p0/m, z19.s, z3.s
; VBITS_GE_128-NEXT:    ldp q17, q6, [x1, #32]
; VBITS_GE_128-NEXT:    movprfx z7, z5
; VBITS_GE_128-NEXT:    udiv z7.s, p0/m, z7.s, z6.s
; VBITS_GE_128-NEXT:    movprfx z18, z16
; VBITS_GE_128-NEXT:    udiv z18.s, p0/m, z18.s, z17.s
; VBITS_GE_128-NEXT:    mls v1.4s, v4.4s, v0.4s
; VBITS_GE_128-NEXT:    mls v2.4s, v19.4s, v3.4s
; VBITS_GE_128-NEXT:    mls v16.4s, v18.4s, v17.4s
; VBITS_GE_128-NEXT:    mls v5.4s, v7.4s, v6.4s
; VBITS_GE_128-NEXT:    stp q1, q2, [x0]
; VBITS_GE_128-NEXT:    stp q16, q5, [x0, #32]
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: urem_v16i32:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.s, vl8
; VBITS_GE_256-NEXT:    mov x8, #8 // =0x8
; VBITS_GE_256-NEXT:    ld1w { z0.s }, p0/z, [x0, x8, lsl #2]
; VBITS_GE_256-NEXT:    ld1w { z1.s }, p0/z, [x0]
; VBITS_GE_256-NEXT:    ld1w { z2.s }, p0/z, [x1, x8, lsl #2]
; VBITS_GE_256-NEXT:    ld1w { z4.s }, p0/z, [x1]
; VBITS_GE_256-NEXT:    movprfx z3, z0
; VBITS_GE_256-NEXT:    udiv z3.s, p0/m, z3.s, z2.s
; VBITS_GE_256-NEXT:    movprfx z5, z1
; VBITS_GE_256-NEXT:    udiv z5.s, p0/m, z5.s, z4.s
; VBITS_GE_256-NEXT:    mls z0.s, p0/m, z3.s, z2.s
; VBITS_GE_256-NEXT:    mls z1.s, p0/m, z5.s, z4.s
; VBITS_GE_256-NEXT:    st1w { z0.s }, p0, [x0, x8, lsl #2]
; VBITS_GE_256-NEXT:    st1w { z1.s }, p0, [x0]
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: urem_v16i32:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.s, vl16
; VBITS_GE_512-NEXT:    ld1w { z0.s }, p0/z, [x0]
; VBITS_GE_512-NEXT:    ld1w { z1.s }, p0/z, [x1]
; VBITS_GE_512-NEXT:    movprfx z2, z0
; VBITS_GE_512-NEXT:    udiv z2.s, p0/m, z2.s, z1.s
; VBITS_GE_512-NEXT:    mls z0.s, p0/m, z2.s, z1.s
; VBITS_GE_512-NEXT:    st1w { z0.s }, p0, [x0]
; VBITS_GE_512-NEXT:    ret
  %op1 = load <16 x i32>, ptr %a
  %op2 = load <16 x i32>, ptr %b
  %res = urem <16 x i32> %op1, %op2
  store <16 x i32> %res, ptr %a
  ret void
}

define void @urem_v32i32(ptr %a, ptr %b) vscale_range(8,0) #0 {
; CHECK-LABEL: urem_v32i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.s, vl32
; CHECK-NEXT:    ld1w { z0.s }, p0/z, [x0]
; CHECK-NEXT:    ld1w { z1.s }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls z0.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    st1w { z0.s }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <32 x i32>, ptr %a
  %op2 = load <32 x i32>, ptr %b
  %res = urem <32 x i32> %op1, %op2
  store <32 x i32> %res, ptr %a
  ret void
}

define void @urem_v64i32(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: urem_v64i32:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.s, vl64
; CHECK-NEXT:    ld1w { z0.s }, p0/z, [x0]
; CHECK-NEXT:    ld1w { z1.s }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    mls z0.s, p0/m, z2.s, z1.s
; CHECK-NEXT:    st1w { z0.s }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <64 x i32>, ptr %a
  %op2 = load <64 x i32>, ptr %b
  %res = urem <64 x i32> %op1, %op2
  store <64 x i32> %res, ptr %a
  ret void
}

; Vector i64 udiv are not legal for NEON so use SVE when available.
; FIXME: We should be able to improve the codegen for the 128 bits case here.
define <1 x i64> @urem_v1i64(<1 x i64> %op1, <1 x i64> %op2) vscale_range(1,0) #0 {
; CHECK-LABEL: urem_v1i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl1
; CHECK-NEXT:    // kill: def $d0 killed $d0 def $z0
; CHECK-NEXT:    // kill: def $d1 killed $d1 def $z1
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    // kill: def $d0 killed $d0 killed $z0
; CHECK-NEXT:    ret
  %res = urem <1 x i64> %op1, %op2
  ret <1 x i64> %res
}

; Vector i64 udiv are not legal for NEON so use SVE when available.
; FIXME: We should be able to improve the codegen for the 128 bits case here.
define <2 x i64> @urem_v2i64(<2 x i64> %op1, <2 x i64> %op2) vscale_range(1,0) #0 {
; CHECK-LABEL: urem_v2i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl2
; CHECK-NEXT:    // kill: def $q0 killed $q0 def $z0
; CHECK-NEXT:    // kill: def $q1 killed $q1 def $z1
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    // kill: def $q0 killed $q0 killed $z0
; CHECK-NEXT:    ret
  %res = urem <2 x i64> %op1, %op2
  ret <2 x i64> %res
}

define void @urem_v4i64(ptr %a, ptr %b) vscale_range(2,0) #0 {
; CHECK-LABEL: urem_v4i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl4
; CHECK-NEXT:    ld1d { z0.d }, p0/z, [x0]
; CHECK-NEXT:    ld1d { z1.d }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    st1d { z0.d }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <4 x i64>, ptr %a
  %op2 = load <4 x i64>, ptr %b
  %res = urem <4 x i64> %op1, %op2
  store <4 x i64> %res, ptr %a
  ret void
}

define void @urem_v8i64(ptr %a, ptr %b) #0 {
; VBITS_GE_128-LABEL: urem_v8i64:
; VBITS_GE_128:       // %bb.0:
; VBITS_GE_128-NEXT:    ptrue p0.d, vl2
; VBITS_GE_128-NEXT:    ldp q0, q3, [x1]
; VBITS_GE_128-NEXT:    ldp q1, q2, [x0]
; VBITS_GE_128-NEXT:    movprfx z4, z1
; VBITS_GE_128-NEXT:    udiv z4.d, p0/m, z4.d, z0.d
; VBITS_GE_128-NEXT:    ldp q16, q5, [x0, #32]
; VBITS_GE_128-NEXT:    movprfx z19, z2
; VBITS_GE_128-NEXT:    udiv z19.d, p0/m, z19.d, z3.d
; VBITS_GE_128-NEXT:    ldp q17, q6, [x1, #32]
; VBITS_GE_128-NEXT:    movprfx z7, z5
; VBITS_GE_128-NEXT:    udiv z7.d, p0/m, z7.d, z6.d
; VBITS_GE_128-NEXT:    movprfx z18, z16
; VBITS_GE_128-NEXT:    udiv z18.d, p0/m, z18.d, z17.d
; VBITS_GE_128-NEXT:    mls z16.d, p0/m, z18.d, z17.d
; VBITS_GE_128-NEXT:    mls z5.d, p0/m, z7.d, z6.d
; VBITS_GE_128-NEXT:    msb z0.d, p0/m, z4.d, z1.d
; VBITS_GE_128-NEXT:    movprfx z1, z2
; VBITS_GE_128-NEXT:    mls z1.d, p0/m, z19.d, z3.d
; VBITS_GE_128-NEXT:    stp q16, q5, [x0, #32]
; VBITS_GE_128-NEXT:    stp q0, q1, [x0]
; VBITS_GE_128-NEXT:    ret
;
; VBITS_GE_256-LABEL: urem_v8i64:
; VBITS_GE_256:       // %bb.0:
; VBITS_GE_256-NEXT:    ptrue p0.d, vl4
; VBITS_GE_256-NEXT:    mov x8, #4 // =0x4
; VBITS_GE_256-NEXT:    ld1d { z0.d }, p0/z, [x0, x8, lsl #3]
; VBITS_GE_256-NEXT:    ld1d { z1.d }, p0/z, [x0]
; VBITS_GE_256-NEXT:    ld1d { z2.d }, p0/z, [x1, x8, lsl #3]
; VBITS_GE_256-NEXT:    ld1d { z4.d }, p0/z, [x1]
; VBITS_GE_256-NEXT:    movprfx z3, z0
; VBITS_GE_256-NEXT:    udiv z3.d, p0/m, z3.d, z2.d
; VBITS_GE_256-NEXT:    movprfx z5, z1
; VBITS_GE_256-NEXT:    udiv z5.d, p0/m, z5.d, z4.d
; VBITS_GE_256-NEXT:    mls z0.d, p0/m, z3.d, z2.d
; VBITS_GE_256-NEXT:    mls z1.d, p0/m, z5.d, z4.d
; VBITS_GE_256-NEXT:    st1d { z0.d }, p0, [x0, x8, lsl #3]
; VBITS_GE_256-NEXT:    st1d { z1.d }, p0, [x0]
; VBITS_GE_256-NEXT:    ret
;
; VBITS_GE_512-LABEL: urem_v8i64:
; VBITS_GE_512:       // %bb.0:
; VBITS_GE_512-NEXT:    ptrue p0.d, vl8
; VBITS_GE_512-NEXT:    ld1d { z0.d }, p0/z, [x0]
; VBITS_GE_512-NEXT:    ld1d { z1.d }, p0/z, [x1]
; VBITS_GE_512-NEXT:    movprfx z2, z0
; VBITS_GE_512-NEXT:    udiv z2.d, p0/m, z2.d, z1.d
; VBITS_GE_512-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; VBITS_GE_512-NEXT:    st1d { z0.d }, p0, [x0]
; VBITS_GE_512-NEXT:    ret
  %op1 = load <8 x i64>, ptr %a
  %op2 = load <8 x i64>, ptr %b
  %res = urem <8 x i64> %op1, %op2
  store <8 x i64> %res, ptr %a
  ret void
}

define void @urem_v16i64(ptr %a, ptr %b) vscale_range(8,0) #0 {
; CHECK-LABEL: urem_v16i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl16
; CHECK-NEXT:    ld1d { z0.d }, p0/z, [x0]
; CHECK-NEXT:    ld1d { z1.d }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    st1d { z0.d }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <16 x i64>, ptr %a
  %op2 = load <16 x i64>, ptr %b
  %res = urem <16 x i64> %op1, %op2
  store <16 x i64> %res, ptr %a
  ret void
}

define void @urem_v32i64(ptr %a, ptr %b) vscale_range(16,0) #0 {
; CHECK-LABEL: urem_v32i64:
; CHECK:       // %bb.0:
; CHECK-NEXT:    ptrue p0.d, vl32
; CHECK-NEXT:    ld1d { z0.d }, p0/z, [x0]
; CHECK-NEXT:    ld1d { z1.d }, p0/z, [x1]
; CHECK-NEXT:    movprfx z2, z0
; CHECK-NEXT:    udiv z2.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    mls z0.d, p0/m, z2.d, z1.d
; CHECK-NEXT:    st1d { z0.d }, p0, [x0]
; CHECK-NEXT:    ret
  %op1 = load <32 x i64>, ptr %a
  %op2 = load <32 x i64>, ptr %b
  %res = urem <32 x i64> %op1, %op2
  store <32 x i64> %res, ptr %a
  ret void
}

attributes #0 = { "target-features"="+sve" }