wasmtime 19.0.2 → 20.0.0

data/ext/cargo-vendor/cranelift-codegen-0.107.2/src/isa/s390x/inst/emit.rs

@@ -0,0 +1,3646 @@
+//! S390x ISA: binary code emission.
+
+use crate::binemit::StackMap;
+use crate::ir::{MemFlags, TrapCode};
+use crate::isa::s390x::inst::*;
+use crate::isa::s390x::settings as s390x_settings;
+use crate::trace;
+use cranelift_control::ControlPlane;
+use regalloc2::Allocation;
+
+/// Debug macro for testing that a regpair is valid: that the high register is even, and the low
+/// register is one higher than the high register.
+macro_rules! debug_assert_valid_regpair {
+    ($hi:expr, $lo:expr) => {
+        if cfg!(debug_assertions) {
+            match ($hi.to_real_reg(), $lo.to_real_reg()) {
+                (Some(hi), Some(lo)) => {
+                    assert!(
+                        hi.hw_enc() % 2 == 0,
+                        "High register is not even: {}",
+                        show_reg($hi)
+                    );
+                    assert_eq!(
+                        hi.hw_enc() + 1,
+                        lo.hw_enc(),
+                        "Low register is not valid: {}, {}",
+                        show_reg($hi),
+                        show_reg($lo)
+                    );
+                }
+
+                _ => {
+                    panic!(
+                        "Expected real registers for {} {}",
+                        show_reg($hi),
+                        show_reg($lo)
+                    );
+                }
+            }
+        }
+    };
+}
+
+/// Type(s) of memory instructions available for mem_finalize.
+pub struct MemInstType {
+    /// True if 12-bit unsigned displacement is supported.
+    pub have_d12: bool,
+    /// True if 20-bit signed displacement is supported.
+    pub have_d20: bool,
+    /// True if PC-relative addressing is supported (memory access).
+    pub have_pcrel: bool,
+    /// True if PC-relative addressing is supported (load address).
+    pub have_unaligned_pcrel: bool,
+    /// True if an index register is supported.
+    pub have_index: bool,
+}
+
+/// Memory addressing mode finalization: convert "special" modes (e.g.,
+/// generic arbitrary stack offset) into real addressing modes, possibly by
+/// emitting some helper instructions that come immediately before the use
+/// of this amode.
+pub fn mem_finalize(
+    mem: &MemArg,
+    state: &EmitState,
+    mi: MemInstType,
+) -> (SmallVec<[Inst; 4]>, MemArg) {
+    let mut insts = SmallVec::new();
+
+    // Resolve virtual addressing modes.
+    let mem = match mem {
+        &MemArg::RegOffset { off, .. }
+        | &MemArg::InitialSPOffset { off }
+        | &MemArg::NominalSPOffset { off } => {
+            let base = match mem {
+                &MemArg::RegOffset { reg, .. } => reg,
+                &MemArg::InitialSPOffset { .. } | &MemArg::NominalSPOffset { .. } => stack_reg(),
+                _ => unreachable!(),
+            };
+            let adj = match mem {
+                &MemArg::InitialSPOffset { .. } => {
+                    state.initial_sp_offset + state.virtual_sp_offset
+                }
+                &MemArg::NominalSPOffset { .. } => state.virtual_sp_offset,
+                _ => 0,
+            };
+            let off = off + adj;
+
+            if let Some(disp) = UImm12::maybe_from_u64(off as u64) {
+                MemArg::BXD12 {
+                    base,
+                    index: zero_reg(),
+                    disp,
+                    flags: mem.get_flags(),
+                }
+            } else if let Some(disp) = SImm20::maybe_from_i64(off) {
+                MemArg::BXD20 {
+                    base,
+                    index: zero_reg(),
+                    disp,
+                    flags: mem.get_flags(),
+                }
+            } else {
+                let tmp = writable_spilltmp_reg();
+                assert!(base != tmp.to_reg());
+                if let Ok(imm) = i16::try_from(off) {
+                    insts.push(Inst::Mov64SImm16 { rd: tmp, imm });
+                } else if let Ok(imm) = i32::try_from(off) {
+                    insts.push(Inst::Mov64SImm32 { rd: tmp, imm });
+                } else {
+                    // The offset must be smaller than the stack frame size,
+                    // which the ABI code limits to 128 MB.
+                    unreachable!();
+                }
+                MemArg::reg_plus_reg(base, tmp.to_reg(), mem.get_flags())
+            }
+        }
+        _ => mem.clone(),
+    };
+
+    // If this addressing mode cannot be handled by the instruction, use load-address.
+    let need_load_address = match &mem {
+        &MemArg::Label { .. } | &MemArg::Symbol { .. } if !mi.have_pcrel => true,
+        &MemArg::Symbol { flags, .. } if !mi.have_unaligned_pcrel && !flags.aligned() => true,
+        &MemArg::BXD20 { .. } if !mi.have_d20 => true,
+        &MemArg::BXD12 { index, .. } | &MemArg::BXD20 { index, .. } if !mi.have_index => {
+            index != zero_reg()
+        }
+        _ => false,
+    };
+    let mem = if need_load_address {
+        let flags = mem.get_flags();
+        let tmp = writable_spilltmp_reg();
+        insts.push(Inst::LoadAddr { rd: tmp, mem });
+        MemArg::reg(tmp.to_reg(), flags)
+    } else {
+        mem
+    };
+
+    // Convert 12-bit displacement to 20-bit if required.
+    let mem = match &mem {
+        &MemArg::BXD12 {
+            base,
+            index,
+            disp,
+            flags,
+        } if !mi.have_d12 => {
+            assert!(mi.have_d20);
+            MemArg::BXD20 {
+                base,
+                index,
+                disp: SImm20::from_uimm12(disp),
+                flags,
+            }
+        }
+        _ => mem,
+    };
+
+    (insts, mem)
+}
+
+pub fn mem_emit(
+    rd: Reg,
+    mem: &MemArg,
+    opcode_rx: Option<u16>,
+    opcode_rxy: Option<u16>,
+    opcode_ril: Option<u16>,
+    add_trap: bool,
+    sink: &mut MachBuffer<Inst>,
+    emit_info: &EmitInfo,
+    state: &mut EmitState,
+) {
+    let (mem_insts, mem) = mem_finalize(
+        mem,
+        state,
+        MemInstType {
+            have_d12: opcode_rx.is_some(),
+            have_d20: opcode_rxy.is_some(),
+            have_pcrel: opcode_ril.is_some(),
+            have_unaligned_pcrel: opcode_ril.is_some() && !add_trap,
+            have_index: true,
+        },
+    );
+    for inst in mem_insts.into_iter() {
+        inst.emit(&[], sink, emit_info, state);
+    }
+
+    if add_trap {
+        if let Some(trap_code) = mem.get_flags().trap_code() {
+            sink.add_trap(trap_code);
+        }
+    }
+
+    match &mem {
+        &MemArg::BXD12 {
+            base, index, disp, ..
+        } => {
+            put(
+                sink,
+                &enc_rx(opcode_rx.unwrap(), rd, base, index, disp.bits()),
+            );
+        }
+        &MemArg::BXD20 {
+            base, index, disp, ..
+        } => {
+            put(
+                sink,
+                &enc_rxy(opcode_rxy.unwrap(), rd, base, index, disp.bits()),
+            );
+        }
+        &MemArg::Label { target } => {
+            sink.use_label_at_offset(sink.cur_offset(), target, LabelUse::BranchRIL);
+            put(sink, &enc_ril_b(opcode_ril.unwrap(), rd, 0));
+        }
+        &MemArg::Symbol {
+            ref name, offset, ..
+        } => {
+            sink.add_reloc_at_offset(2, Reloc::S390xPCRel32Dbl, &**name, (offset + 2).into());
+            put(sink, &enc_ril_b(opcode_ril.unwrap(), rd, 0));
+        }
+        _ => unreachable!(),
+    }
+}
+
+pub fn mem_rs_emit(
+    rd: Reg,
+    rn: Reg,
+    mem: &MemArg,
+    opcode_rs: Option<u16>,
+    opcode_rsy: Option<u16>,
+    add_trap: bool,
+    sink: &mut MachBuffer<Inst>,
+    emit_info: &EmitInfo,
+    state: &mut EmitState,
+) {
+    let (mem_insts, mem) = mem_finalize(
+        mem,
+        state,
+        MemInstType {
+            have_d12: opcode_rs.is_some(),
+            have_d20: opcode_rsy.is_some(),
+            have_pcrel: false,
+            have_unaligned_pcrel: false,
+            have_index: false,
+        },
+    );
+    for inst in mem_insts.into_iter() {
+        inst.emit(&[], sink, emit_info, state);
+    }
+
+    if add_trap {
+        if let Some(trap_code) = mem.get_flags().trap_code() {
+            sink.add_trap(trap_code);
+        }
+    }
+
+    match &mem {
+        &MemArg::BXD12 {
+            base, index, disp, ..
+        } => {
+            assert!(index == zero_reg());
+            put(sink, &enc_rs(opcode_rs.unwrap(), rd, rn, base, disp.bits()));
+        }
+        &MemArg::BXD20 {
+            base, index, disp, ..
+        } => {
+            assert!(index == zero_reg());
+            put(
+                sink,
+                &enc_rsy(opcode_rsy.unwrap(), rd, rn, base, disp.bits()),
+            );
+        }
+        _ => unreachable!(),
+    }
+}
+
+pub fn mem_imm8_emit(
+    imm: u8,
+    mem: &MemArg,
+    opcode_si: u16,
+    opcode_siy: u16,
+    add_trap: bool,
+    sink: &mut MachBuffer<Inst>,
+    emit_info: &EmitInfo,
+    state: &mut EmitState,
+) {
+    let (mem_insts, mem) = mem_finalize(
+        mem,
+        state,
+        MemInstType {
+            have_d12: true,
+            have_d20: true,
+            have_pcrel: false,
+            have_unaligned_pcrel: false,
+            have_index: false,
+        },
+    );
+    for inst in mem_insts.into_iter() {
+        inst.emit(&[], sink, emit_info, state);
+    }
+
+    if add_trap {
+        if let Some(trap_code) = mem.get_flags().trap_code() {
+            sink.add_trap(trap_code);
+        }
+    }
+
+    match &mem {
+        &MemArg::BXD12 {
+            base, index, disp, ..
+        } => {
+            assert!(index == zero_reg());
+            put(sink, &enc_si(opcode_si, base, disp.bits(), imm));
+        }
+        &MemArg::BXD20 {
+            base, index, disp, ..
+        } => {
+            assert!(index == zero_reg());
+            put(sink, &enc_siy(opcode_siy, base, disp.bits(), imm));
+        }
+        _ => unreachable!(),
+    }
+}
+
+pub fn mem_imm16_emit(
+    imm: i16,
+    mem: &MemArg,
+    opcode_sil: u16,
+    add_trap: bool,
+    sink: &mut MachBuffer<Inst>,
+    emit_info: &EmitInfo,
+    state: &mut EmitState,
+) {
+    let (mem_insts, mem) = mem_finalize(
+        mem,
+        state,
+        MemInstType {
+            have_d12: true,
+            have_d20: false,
+            have_pcrel: false,
+            have_unaligned_pcrel: false,
+            have_index: false,
+        },
+    );
+    for inst in mem_insts.into_iter() {
+        inst.emit(&[], sink, emit_info, state);
+    }
+
+    if add_trap {
+        if let Some(trap_code) = mem.get_flags().trap_code() {
+            sink.add_trap(trap_code);
+        }
+    }
+
+    match &mem {
+        &MemArg::BXD12 {
+            base, index, disp, ..
+        } => {
+            assert!(index == zero_reg());
+            put(sink, &enc_sil(opcode_sil, base, disp.bits(), imm));
+        }
+        _ => unreachable!(),
+    }
+}
+
+pub fn mem_mem_emit(
+    dst: &MemArgPair,
+    src: &MemArgPair,
+    len_minus_one: u8,
+    opcode_ss: u8,
+    add_trap: bool,
+    sink: &mut MachBuffer<Inst>,
+    _state: &mut EmitState,
+) {
+    if add_trap {
+        if let Some(trap_code) = dst.flags.trap_code().or(src.flags.trap_code()) {
+            sink.add_trap(trap_code);
+        }
+    }
+
+    put(
+        sink,
+        &enc_ss_a(
+            opcode_ss,
+            dst.base,
+            dst.disp.bits(),
+            src.base,
+            src.disp.bits(),
+            len_minus_one,
+        ),
+    );
+}
+
+pub fn mem_vrx_emit(
+    rd: Reg,
+    mem: &MemArg,
+    opcode: u16,
+    m3: u8,
+    add_trap: bool,
+    sink: &mut MachBuffer<Inst>,
+    emit_info: &EmitInfo,
+    state: &mut EmitState,
+) {
+    let (mem_insts, mem) = mem_finalize(
+        mem,
+        state,
+        MemInstType {
+            have_d12: true,
+            have_d20: false,
+            have_pcrel: false,
+            have_unaligned_pcrel: false,
+            have_index: true,
+        },
+    );
+    for inst in mem_insts.into_iter() {
+        inst.emit(&[], sink, emit_info, state);
+    }
+
+    if add_trap {
+        if let Some(trap_code) = mem.get_flags().trap_code() {
+            sink.add_trap(trap_code);
+        }
+    }
+
+    match &mem {
+        &MemArg::BXD12 {
+            base, index, disp, ..
+        } => {
+            put(sink, &enc_vrx(opcode, rd, base, index, disp.bits(), m3));
+        }
+        _ => unreachable!(),
+    }
+}
+
+//=============================================================================
+// Instructions and subcomponents: emission
+
+fn machreg_to_gpr(m: Reg) -> u8 {
+    assert_eq!(m.class(), RegClass::Int);
+    u8::try_from(m.to_real_reg().unwrap().hw_enc()).unwrap()
+}
+
+fn machreg_to_vr(m: Reg) -> u8 {
+    assert_eq!(m.class(), RegClass::Float);
+    u8::try_from(m.to_real_reg().unwrap().hw_enc()).unwrap()
+}
+
+fn machreg_to_fpr(m: Reg) -> u8 {
+    assert!(is_fpr(m));
+    u8::try_from(m.to_real_reg().unwrap().hw_enc()).unwrap()
+}
+
+fn machreg_to_gpr_or_fpr(m: Reg) -> u8 {
+    let reg = u8::try_from(m.to_real_reg().unwrap().hw_enc()).unwrap();
+    assert!(reg < 16);
+    reg
+}
+
+fn rxb(v1: Option<Reg>, v2: Option<Reg>, v3: Option<Reg>, v4: Option<Reg>) -> u8 {
+    let mut rxb = 0;
+
+    let is_high_vr = |reg| -> bool {
+        if let Some(reg) = reg {
+            if !is_fpr(reg) {
+                return true;
+            }
+        }
+        false
+    };
+
+    if is_high_vr(v1) {
+        rxb = rxb | 8;
+    }
+    if is_high_vr(v2) {
+        rxb = rxb | 4;
+    }
+    if is_high_vr(v3) {
+        rxb = rxb | 2;
+    }
+    if is_high_vr(v4) {
+        rxb = rxb | 1;
+    }
+
+    rxb
+}
+
+/// E-type instructions.
+///
+///   15
+///   opcode
+///        0
+///
+fn enc_e(opcode: u16) -> [u8; 2] {
+    let mut enc: [u8; 2] = [0; 2];
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+
+    enc[0] = opcode1;
+    enc[1] = opcode2;
+    enc
+}
+
+/// RIa-type instructions.
+///
+///   31      23 19      15
+///   opcode1 r1 opcode2 i2
+///        24 20      16  0
+///
+fn enc_ri_a(opcode: u16, r1: Reg, i2: u16) -> [u8; 4] {
+    let mut enc: [u8; 4] = [0; 4];
+    let opcode1 = ((opcode >> 4) & 0xff) as u8;
+    let opcode2 = (opcode & 0xf) as u8;
+    let r1 = machreg_to_gpr(r1) & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | opcode2;
+    enc[2..].copy_from_slice(&i2.to_be_bytes());
+    enc
+}
+
+/// RIb-type instructions.
+///
+///   31      23 19      15
+///   opcode1 r1 opcode2 ri2
+///        24 20      16   0
+///
+fn enc_ri_b(opcode: u16, r1: Reg, ri2: i32) -> [u8; 4] {
+    let mut enc: [u8; 4] = [0; 4];
+    let opcode1 = ((opcode >> 4) & 0xff) as u8;
+    let opcode2 = (opcode & 0xf) as u8;
+    let r1 = machreg_to_gpr(r1) & 0x0f;
+    let ri2 = ((ri2 >> 1) & 0xffff) as u16;
+
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | opcode2;
+    enc[2..].copy_from_slice(&ri2.to_be_bytes());
+    enc
+}
+
+/// RIc-type instructions.
+///
+///   31      23 19      15
+///   opcode1 m1 opcode2 ri2
+///        24 20      16   0
+///
+fn enc_ri_c(opcode: u16, m1: u8, ri2: i32) -> [u8; 4] {
+    let mut enc: [u8; 4] = [0; 4];
+    let opcode1 = ((opcode >> 4) & 0xff) as u8;
+    let opcode2 = (opcode & 0xf) as u8;
+    let m1 = m1 & 0x0f;
+    let ri2 = ((ri2 >> 1) & 0xffff) as u16;
+
+    enc[0] = opcode1;
+    enc[1] = m1 << 4 | opcode2;
+    enc[2..].copy_from_slice(&ri2.to_be_bytes());
+    enc
+}
+
+/// RIEa-type instructions.
+///
+///   47      39 35 31 15 11 7
+///   opcode1 r1 -- i2 m3 -- opcode2
+///        40 36 32 16 12 8       0
+///
+fn enc_rie_a(opcode: u16, r1: Reg, i2: u16, m3: u8) -> [u8; 6] {
+    let mut enc: [u8; 6] = [0; 6];
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr(r1) & 0x0f;
+    let m3 = m3 & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = r1 << 4;
+    enc[2..4].copy_from_slice(&i2.to_be_bytes());
+    enc[4] = m3 << 4;
+    enc[5] = opcode2;
+    enc
+}
+
+/// RIEd-type instructions.
+///
+///   47      39 35 31 15 7
+///   opcode1 r1 r3 i2 -- opcode2
+///        40 36 32 16  8       0
+///
+fn enc_rie_d(opcode: u16, r1: Reg, r3: Reg, i2: u16) -> [u8; 6] {
+    let mut enc: [u8; 6] = [0; 6];
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr(r1) & 0x0f;
+    let r3 = machreg_to_gpr(r3) & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | r3;
+    enc[2..4].copy_from_slice(&i2.to_be_bytes());
+    enc[5] = opcode2;
+    enc
+}
+
+/// RIEf-type instructions.
+///
+///   47      39 35 31 23 15 7
+///   opcode1 r1 r2 i3 i4 i5 opcode2
+///        40 36 32 24 16  8       0
+///
+fn enc_rie_f(opcode: u16, r1: Reg, r2: Reg, i3: u8, i4: u8, i5: u8) -> [u8; 6] {
+    let mut enc: [u8; 6] = [0; 6];
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr(r1) & 0x0f;
+    let r2 = machreg_to_gpr(r2) & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | r2;
+    enc[2] = i3;
+    enc[3] = i4;
+    enc[4] = i5;
+    enc[5] = opcode2;
+    enc
+}
+
+/// RIEg-type instructions.
+///
+///   47      39 35 31 15 7
+///   opcode1 r1 m3 i2 -- opcode2
+///        40 36 32 16  8       0
+///
+fn enc_rie_g(opcode: u16, r1: Reg, i2: u16, m3: u8) -> [u8; 6] {
+    let mut enc: [u8; 6] = [0; 6];
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr(r1) & 0x0f;
+    let m3 = m3 & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | m3;
+    enc[2..4].copy_from_slice(&i2.to_be_bytes());
+    enc[5] = opcode2;
+    enc
+}
+
+/// RILa-type instructions.
+///
+///   47      39 35      31
+///   opcode1 r1 opcode2 i2
+///        40 36      32  0
+///
+fn enc_ril_a(opcode: u16, r1: Reg, i2: u32) -> [u8; 6] {
+    let mut enc: [u8; 6] = [0; 6];
+    let opcode1 = ((opcode >> 4) & 0xff) as u8;
+    let opcode2 = (opcode & 0xf) as u8;
+    let r1 = machreg_to_gpr(r1) & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | opcode2;
+    enc[2..].copy_from_slice(&i2.to_be_bytes());
+    enc
+}
+
+/// RILb-type instructions.
+///
+///   47      39 35      31
+///   opcode1 r1 opcode2 ri2
+///        40 36      32   0
+///
+fn enc_ril_b(opcode: u16, r1: Reg, ri2: u32) -> [u8; 6] {
+    let mut enc: [u8; 6] = [0; 6];
+    let opcode1 = ((opcode >> 4) & 0xff) as u8;
+    let opcode2 = (opcode & 0xf) as u8;
+    let r1 = machreg_to_gpr(r1) & 0x0f;
+    let ri2 = ri2 >> 1;
+
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | opcode2;
+    enc[2..].copy_from_slice(&ri2.to_be_bytes());
+    enc
+}
+
+/// RILc-type instructions.
+///
+///   47      39 35      31
+///   opcode1 m1 opcode2 i2
+///        40 36      32  0
+///
+fn enc_ril_c(opcode: u16, m1: u8, ri2: u32) -> [u8; 6] {
+    let mut enc: [u8; 6] = [0; 6];
+    let opcode1 = ((opcode >> 4) & 0xff) as u8;
+    let opcode2 = (opcode & 0xf) as u8;
+    let m1 = m1 & 0x0f;
+    let ri2 = ri2 >> 1;
+
+    enc[0] = opcode1;
+    enc[1] = m1 << 4 | opcode2;
+    enc[2..].copy_from_slice(&ri2.to_be_bytes());
+    enc
+}
+
+/// RR-type instructions.
+///
+///   15     7  3
+///   opcode r1 r2
+///        8  4  0
+///
+fn enc_rr(opcode: u16, r1: Reg, r2: Reg) -> [u8; 2] {
+    let mut enc: [u8; 2] = [0; 2];
+    let opcode = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr_or_fpr(r1) & 0x0f;
+    let r2 = machreg_to_gpr_or_fpr(r2) & 0x0f;
+
+    enc[0] = opcode;
+    enc[1] = r1 << 4 | r2;
+    enc
+}
+
+/// RRD-type instructions.
+///
+///   31     15 11 7  3
+///   opcode r1 -- r3 r2
+///       16 12  8 4  0
+///
+fn enc_rrd(opcode: u16, r1: Reg, r2: Reg, r3: Reg) -> [u8; 4] {
+    let mut enc: [u8; 4] = [0; 4];
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_fpr(r1) & 0x0f;
+    let r2 = machreg_to_fpr(r2) & 0x0f;
+    let r3 = machreg_to_fpr(r3) & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = opcode2;
+    enc[2] = r1 << 4;
+    enc[3] = r3 << 4 | r2;
+    enc
+}
+
+/// RRE-type instructions.
+///
+///   31     15 7  3
+///   opcode -- r1 r2
+///       16  8  4  0
+///
+fn enc_rre(opcode: u16, r1: Reg, r2: Reg) -> [u8; 4] {
+    let mut enc: [u8; 4] = [0; 4];
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr_or_fpr(r1) & 0x0f;
+    let r2 = machreg_to_gpr_or_fpr(r2) & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = opcode2;
+    enc[3] = r1 << 4 | r2;
+    enc
+}
+
+/// RRFa/b-type instructions.
+///
+///   31     15 11 7  3
+///   opcode r3 m4 r1 r2
+///       16 12  8  4  0
+///
+fn enc_rrf_ab(opcode: u16, r1: Reg, r2: Reg, r3: Reg, m4: u8) -> [u8; 4] {
+    let mut enc: [u8; 4] = [0; 4];
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr_or_fpr(r1) & 0x0f;
+    let r2 = machreg_to_gpr_or_fpr(r2) & 0x0f;
+    let r3 = machreg_to_gpr_or_fpr(r3) & 0x0f;
+    let m4 = m4 & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = opcode2;
+    enc[2] = r3 << 4 | m4;
+    enc[3] = r1 << 4 | r2;
+    enc
+}
+
+/// RRFc/d/e-type instructions.
+///
+///   31     15 11 7  3
+///   opcode m3 m4 r1 r2
+///       16 12  8  4  0
+///
+fn enc_rrf_cde(opcode: u16, r1: Reg, r2: Reg, m3: u8, m4: u8) -> [u8; 4] {
+    let mut enc: [u8; 4] = [0; 4];
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr_or_fpr(r1) & 0x0f;
+    let r2 = machreg_to_gpr_or_fpr(r2) & 0x0f;
+    let m3 = m3 & 0x0f;
+    let m4 = m4 & 0x0f;
+
+    enc[0] = opcode1;
+    enc[1] = opcode2;
+    enc[2] = m3 << 4 | m4;
+    enc[3] = r1 << 4 | r2;
+    enc
+}
+
+/// RS-type instructions.
+///
+///   31     23 19 15 11
+///   opcode r1 r3 b2 d2
+///       24 20 16 12  0
+///
+fn enc_rs(opcode: u16, r1: Reg, r3: Reg, b2: Reg, d2: u32) -> [u8; 4] {
+    let opcode = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr_or_fpr(r1) & 0x0f;
+    let r3 = machreg_to_gpr_or_fpr(r3) & 0x0f;
+    let b2 = machreg_to_gpr(b2) & 0x0f;
+    let d2_lo = (d2 & 0xff) as u8;
+    let d2_hi = ((d2 >> 8) & 0x0f) as u8;
+
+    let mut enc: [u8; 4] = [0; 4];
+    enc[0] = opcode;
+    enc[1] = r1 << 4 | r3;
+    enc[2] = b2 << 4 | d2_hi;
+    enc[3] = d2_lo;
+    enc
+}
+
+/// RSY-type instructions.
+///
+///   47      39 35 31 27  15  7
+///   opcode1 r1 r3 b2 dl2 dh2 opcode2
+///        40 36 32 28  16   8       0
+///
+fn enc_rsy(opcode: u16, r1: Reg, r3: Reg, b2: Reg, d2: u32) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr_or_fpr(r1) & 0x0f;
+    let r3 = machreg_to_gpr_or_fpr(r3) & 0x0f;
+    let b2 = machreg_to_gpr(b2) & 0x0f;
+    let dl2_lo = (d2 & 0xff) as u8;
+    let dl2_hi = ((d2 >> 8) & 0x0f) as u8;
+    let dh2 = ((d2 >> 12) & 0xff) as u8;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | r3;
+    enc[2] = b2 << 4 | dl2_hi;
+    enc[3] = dl2_lo;
+    enc[4] = dh2;
+    enc[5] = opcode2;
+    enc
+}
+
+/// RX-type instructions.
+///
+///   31     23 19 15 11
+///   opcode r1 x2 b2 d2
+///       24 20 16 12  0
+///
+fn enc_rx(opcode: u16, r1: Reg, b2: Reg, x2: Reg, d2: u32) -> [u8; 4] {
+    let opcode = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr_or_fpr(r1) & 0x0f;
+    let b2 = machreg_to_gpr(b2) & 0x0f;
+    let x2 = machreg_to_gpr(x2) & 0x0f;
+    let d2_lo = (d2 & 0xff) as u8;
+    let d2_hi = ((d2 >> 8) & 0x0f) as u8;
+
+    let mut enc: [u8; 4] = [0; 4];
+    enc[0] = opcode;
+    enc[1] = r1 << 4 | x2;
+    enc[2] = b2 << 4 | d2_hi;
+    enc[3] = d2_lo;
+    enc
+}
+
+/// RXY-type instructions.
+///
+///   47      39 35 31 27  15  7
+///   opcode1 r1 x2 b2 dl2 dh2 opcode2
+///        40 36 32 28  16   8       0
+///
+fn enc_rxy(opcode: u16, r1: Reg, b2: Reg, x2: Reg, d2: u32) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let r1 = machreg_to_gpr_or_fpr(r1) & 0x0f;
+    let b2 = machreg_to_gpr(b2) & 0x0f;
+    let x2 = machreg_to_gpr(x2) & 0x0f;
+    let dl2_lo = (d2 & 0xff) as u8;
+    let dl2_hi = ((d2 >> 8) & 0x0f) as u8;
+    let dh2 = ((d2 >> 12) & 0xff) as u8;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | x2;
+    enc[2] = b2 << 4 | dl2_hi;
+    enc[3] = dl2_lo;
+    enc[4] = dh2;
+    enc[5] = opcode2;
+    enc
+}
+
+/// SI-type instructions.
+///
+///   31     23 15 11
+///   opcode i2 b1 d1
+///       24 16 12  0
+///
+fn enc_si(opcode: u16, b1: Reg, d1: u32, i2: u8) -> [u8; 4] {
+    let opcode = (opcode & 0xff) as u8;
+    let b1 = machreg_to_gpr(b1) & 0x0f;
+    let d1_lo = (d1 & 0xff) as u8;
+    let d1_hi = ((d1 >> 8) & 0x0f) as u8;
+
+    let mut enc: [u8; 4] = [0; 4];
+    enc[0] = opcode;
+    enc[1] = i2;
+    enc[2] = b1 << 4 | d1_hi;
+    enc[3] = d1_lo;
+    enc
+}
+
+/// SIL-type instructions.
+///
+///   47     31 27 15
+///   opcode b1 d1 i2
+///       32 28 16  0
+///
+fn enc_sil(opcode: u16, b1: Reg, d1: u32, i2: i16) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let b1 = machreg_to_gpr(b1) & 0x0f;
+    let d1_lo = (d1 & 0xff) as u8;
+    let d1_hi = ((d1 >> 8) & 0x0f) as u8;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = opcode2;
+    enc[2] = b1 << 4 | d1_hi;
+    enc[3] = d1_lo;
+    enc[4..].copy_from_slice(&i2.to_be_bytes());
+    enc
+}
+
+/// SIY-type instructions.
+///
+///   47      39 31 27  15  7
+///   opcode1 i2 b1 dl1 dh1 opcode2
+///        40 32 28  16   8       0
+///
+fn enc_siy(opcode: u16, b1: Reg, d1: u32, i2: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let b1 = machreg_to_gpr(b1) & 0x0f;
+    let dl1_lo = (d1 & 0xff) as u8;
+    let dl1_hi = ((d1 >> 8) & 0x0f) as u8;
+    let dh1 = ((d1 >> 12) & 0xff) as u8;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = i2;
+    enc[2] = b1 << 4 | dl1_hi;
+    enc[3] = dl1_lo;
+    enc[4] = dh1;
+    enc[5] = opcode2;
+    enc
+}
+
+/// SSa-type instructions.
+///
+///   47     39 31 27 15 11
+///   opcode  l b1 d1 b2 d2
+///       40 32 28 16 12  0
+///
+///
+fn enc_ss_a(opcode: u8, b1: Reg, d1: u32, b2: Reg, d2: u32, l: u8) -> [u8; 6] {
+    let b1 = machreg_to_gpr(b1) & 0x0f;
+    let d1_lo = (d1 & 0xff) as u8;
+    let d1_hi = ((d1 >> 8) & 0x0f) as u8;
+    let b2 = machreg_to_gpr(b2) & 0x0f;
+    let d2_lo = (d2 & 0xff) as u8;
+    let d2_hi = ((d2 >> 8) & 0x0f) as u8;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode;
+    enc[1] = l;
+    enc[2] = b1 << 4 | d1_hi;
+    enc[3] = d1_lo;
+    enc[4] = b2 << 4 | d2_hi;
+    enc[5] = d2_lo;
+    enc
+}
+
+/// VRIa-type instructions.
+///
+///   47      39 35 31 15 11  7
+///   opcode1 v1 -  i2 m3 rxb opcode2
+///        40 36 32 16 12   8       0
+///
+fn enc_vri_a(opcode: u16, v1: Reg, i2: u16, m3: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), None, None, None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let m3 = m3 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4;
+    enc[2..4].copy_from_slice(&i2.to_be_bytes());
+    enc[4] = m3 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRIb-type instructions.
+///
+///   47      39 35 31 23 15 11  7
+///   opcode1 v1 -  i2 i3 m4 rxb opcode2
+///        40 36 32 24 16 12   8       0
+///
+fn enc_vri_b(opcode: u16, v1: Reg, i2: u8, i3: u8, m4: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), None, None, None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let m4 = m4 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4;
+    enc[2] = i2;
+    enc[3] = i3;
+    enc[4] = m4 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRIc-type instructions.
+///
+///   47      39 35 31 15 11  7
+///   opcode1 v1 v3 i2 m4 rxb opcode2
+///        40 36 32 16 12   8       0
+///
+fn enc_vri_c(opcode: u16, v1: Reg, i2: u16, v3: Reg, m4: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), Some(v3), None, None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let v3 = machreg_to_vr(v3) & 0x0f;
+    let m4 = m4 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4 | v3;
+    enc[2..4].copy_from_slice(&i2.to_be_bytes());
+    enc[4] = m4 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRRa-type instructions.
+///
+///   47      39 35 31 23 19 15 11  7
+///   opcode1 v1 v2 -  m5 m3 m2 rxb opcode2
+///        40 36 32 24 20 16 12   8       0
+///
+fn enc_vrr_a(opcode: u16, v1: Reg, v2: Reg, m3: u8, m4: u8, m5: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), Some(v2), None, None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let v2 = machreg_to_vr(v2) & 0x0f;
+    let m3 = m3 & 0x0f;
+    let m4 = m4 & 0x0f;
+    let m5 = m5 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4 | v2;
+    enc[2] = 0;
+    enc[3] = m5 << 4 | m4;
+    enc[4] = m3 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRRb-type instructions.
+///
+///   47      39 35 31 27 23 19 15 11  7
+///   opcode1 v1 v2 v3 -  m5 -  m4 rxb opcode2
+///        40 36 32 28 24 20 16 12   8       0
+///
+fn enc_vrr_b(opcode: u16, v1: Reg, v2: Reg, v3: Reg, m4: u8, m5: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), Some(v2), Some(v3), None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let v2 = machreg_to_vr(v2) & 0x0f;
+    let v3 = machreg_to_vr(v3) & 0x0f;
+    let m4 = m4 & 0x0f;
+    let m5 = m5 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4 | v2;
+    enc[2] = v3 << 4;
+    enc[3] = m5 << 4;
+    enc[4] = m4 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRRc-type instructions.
+///
+///   47      39 35 31 27 23 19 15 11  7
+///   opcode1 v1 v2 v3 -  m6 m5 m4 rxb opcode2
+///        40 36 32 28 24 20 16 12   8       0
+///
+fn enc_vrr_c(opcode: u16, v1: Reg, v2: Reg, v3: Reg, m4: u8, m5: u8, m6: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), Some(v2), Some(v3), None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let v2 = machreg_to_vr(v2) & 0x0f;
+    let v3 = machreg_to_vr(v3) & 0x0f;
+    let m4 = m4 & 0x0f;
+    let m5 = m5 & 0x0f;
+    let m6 = m6 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4 | v2;
+    enc[2] = v3 << 4;
+    enc[3] = m6 << 4 | m5;
+    enc[4] = m4 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRRe-type instructions.
+///
+///   47      39 35 31 27 23 19 15 11  7
+///   opcode1 v1 v2 v3 m6 -  m5 v4 rxb opcode2
+///        40 36 32 28 24 20 16 12   8       0
+///
+fn enc_vrr_e(opcode: u16, v1: Reg, v2: Reg, v3: Reg, v4: Reg, m5: u8, m6: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), Some(v2), Some(v3), Some(v4));
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let v2 = machreg_to_vr(v2) & 0x0f;
+    let v3 = machreg_to_vr(v3) & 0x0f;
+    let v4 = machreg_to_vr(v4) & 0x0f;
+    let m5 = m5 & 0x0f;
+    let m6 = m6 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4 | v2;
+    enc[2] = v3 << 4 | m6;
+    enc[3] = m5;
+    enc[4] = v4 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRRf-type instructions.
+///
+///   47      39 35 31 27 11  7
+///   opcode1 v1 r2 r3 -  rxb opcode2
+///        40 36 32 28 12   8       0
+///
+fn enc_vrr_f(opcode: u16, v1: Reg, r2: Reg, r3: Reg) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), None, None, None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let r2 = machreg_to_gpr(r2) & 0x0f;
+    let r3 = machreg_to_gpr(r3) & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4 | r2;
+    enc[2] = r3 << 4;
+    enc[4] = rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRSa-type instructions.
+///
+///   47      39 35 31 27 15 11  7
+///   opcode1 v1 v3 b2 d2 m4 rxb opcode2
+///        40 36 32 28 16 12   8       0
+///
+fn enc_vrs_a(opcode: u16, v1: Reg, b2: Reg, d2: u32, v3: Reg, m4: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), Some(v3), None, None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let b2 = machreg_to_gpr(b2) & 0x0f;
+    let v3 = machreg_to_vr(v3) & 0x0f;
+    let d2_lo = (d2 & 0xff) as u8;
+    let d2_hi = ((d2 >> 8) & 0x0f) as u8;
+    let m4 = m4 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4 | v3;
+    enc[2] = b2 << 4 | d2_hi;
+    enc[3] = d2_lo;
+    enc[4] = m4 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRSb-type instructions.
+///
+///   47      39 35 31 27 15 11  7
+///   opcode1 v1 r3 b2 d2 m4 rxb opcode2
+///        40 36 32 28 16 12   8       0
+///
+fn enc_vrs_b(opcode: u16, v1: Reg, b2: Reg, d2: u32, r3: Reg, m4: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), None, None, None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let b2 = machreg_to_gpr(b2) & 0x0f;
+    let r3 = machreg_to_gpr(r3) & 0x0f;
+    let d2_lo = (d2 & 0xff) as u8;
+    let d2_hi = ((d2 >> 8) & 0x0f) as u8;
+    let m4 = m4 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4 | r3;
+    enc[2] = b2 << 4 | d2_hi;
+    enc[3] = d2_lo;
+    enc[4] = m4 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRSc-type instructions.
+///
+///   47      39 35 31 27 15 11  7
+///   opcode1 r1 v3 b2 d2 m4 rxb opcode2
+///        40 36 32 28 16 12   8       0
+///
+fn enc_vrs_c(opcode: u16, r1: Reg, b2: Reg, d2: u32, v3: Reg, m4: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(None, Some(v3), None, None);
+    let r1 = machreg_to_gpr(r1) & 0x0f;
+    let b2 = machreg_to_gpr(b2) & 0x0f;
+    let v3 = machreg_to_vr(v3) & 0x0f;
+    let d2_lo = (d2 & 0xff) as u8;
+    let d2_hi = ((d2 >> 8) & 0x0f) as u8;
+    let m4 = m4 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = r1 << 4 | v3;
+    enc[2] = b2 << 4 | d2_hi;
+    enc[3] = d2_lo;
+    enc[4] = m4 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// VRX-type instructions.
+///
+///   47      39 35 31 27 15 11  7
+///   opcode1 v1 x2 b2 d2 m3 rxb opcode2
+///        40 36 32 28 16 12   8       0
+///
+fn enc_vrx(opcode: u16, v1: Reg, b2: Reg, x2: Reg, d2: u32, m3: u8) -> [u8; 6] {
+    let opcode1 = ((opcode >> 8) & 0xff) as u8;
+    let opcode2 = (opcode & 0xff) as u8;
+    let rxb = rxb(Some(v1), None, None, None);
+    let v1 = machreg_to_vr(v1) & 0x0f;
+    let b2 = machreg_to_gpr(b2) & 0x0f;
+    let x2 = machreg_to_gpr(x2) & 0x0f;
+    let d2_lo = (d2 & 0xff) as u8;
+    let d2_hi = ((d2 >> 8) & 0x0f) as u8;
+    let m3 = m3 & 0x0f;
+
+    let mut enc: [u8; 6] = [0; 6];
+    enc[0] = opcode1;
+    enc[1] = v1 << 4 | x2;
+    enc[2] = b2 << 4 | d2_hi;
+    enc[3] = d2_lo;
+    enc[4] = m3 << 4 | rxb;
+    enc[5] = opcode2;
+    enc
+}
+
+/// Emit encoding to sink.
+fn put(sink: &mut MachBuffer<Inst>, enc: &[u8]) {
+    for byte in enc {
+        sink.put1(*byte);
+    }
+}
+
+/// Emit encoding to sink, adding a trap on the last byte.
+fn put_with_trap(sink: &mut MachBuffer<Inst>, enc: &[u8], trap_code: TrapCode) {
+    let len = enc.len();
+    for i in 0..len - 1 {
+        sink.put1(enc[i]);
+    }
+    sink.add_trap(trap_code);
+    sink.put1(enc[len - 1]);
+}
+
+/// State carried between emissions of a sequence of instructions.
+#[derive(Default, Clone, Debug)]
+pub struct EmitState {
+    pub(crate) initial_sp_offset: i64,
+    pub(crate) virtual_sp_offset: i64,
+    /// Safepoint stack map for upcoming instruction, as provided to `pre_safepoint()`.
+    stack_map: Option<StackMap>,
+    /// Only used during fuzz-testing. Otherwise, it is a zero-sized struct and
+    /// optimized away at compiletime. See [cranelift_control].
+    ctrl_plane: ControlPlane,
+}
+
+impl MachInstEmitState<Inst> for EmitState {
+    fn new(abi: &Callee<S390xMachineDeps>, ctrl_plane: ControlPlane) -> Self {
+        EmitState {
+            virtual_sp_offset: 0,
+            initial_sp_offset: abi.frame_size() as i64,
+            stack_map: None,
+            ctrl_plane,
+        }
+    }
+
+    fn pre_safepoint(&mut self, stack_map: StackMap) {
+        self.stack_map = Some(stack_map);
+    }
+
+    fn ctrl_plane_mut(&mut self) -> &mut ControlPlane {
+        &mut self.ctrl_plane
+    }
+
+    fn take_ctrl_plane(self) -> ControlPlane {
+        self.ctrl_plane
+    }
+}
+
+impl EmitState {
+    fn take_stack_map(&mut self) -> Option<StackMap> {
+        self.stack_map.take()
+    }
+
+    fn clear_post_insn(&mut self) {
+        self.stack_map = None;
+    }
+}
+
+/// Constant state used during function compilation.
+pub struct EmitInfo {
+    isa_flags: s390x_settings::Flags,
+}
+
+impl EmitInfo {
+    pub(crate) fn new(isa_flags: s390x_settings::Flags) -> Self {
+        Self { isa_flags }
+    }
+}
+
+impl MachInstEmit for Inst {
+    type State = EmitState;
+    type Info = EmitInfo;
+
+    fn emit(
+        &self,
+        allocs: &[Allocation],
+        sink: &mut MachBuffer<Inst>,
+        emit_info: &Self::Info,
+        state: &mut EmitState,
+    ) {
+        let mut allocs = AllocationConsumer::new(allocs);
+        self.emit_with_alloc_consumer(&mut allocs, sink, emit_info, state)
+    }
+
+    fn pretty_print_inst(&self, allocs: &[Allocation], state: &mut EmitState) -> String {
+        let mut allocs = AllocationConsumer::new(allocs);
+        self.print_with_state(state, &mut allocs)
+    }
+}
+
+impl Inst {
+    fn emit_with_alloc_consumer(
+        &self,
+        allocs: &mut AllocationConsumer<'_>,
+        sink: &mut MachBuffer<Inst>,
+        emit_info: &EmitInfo,
+        state: &mut EmitState,
+    ) {
+        // Verify that we can emit this Inst in the current ISA
+        let matches_isa_flags = |iset_requirement: &InstructionSet| -> bool {
+            match iset_requirement {
+                // Baseline ISA is z14
+                InstructionSet::Base => true,
+                // Miscellaneous-Instruction-Extensions Facility 2 (z15)
+                InstructionSet::MIE2 => emit_info.isa_flags.has_mie2(),
+                // Vector-Enhancements Facility 2 (z15)
+                InstructionSet::VXRS_EXT2 => emit_info.isa_flags.has_vxrs_ext2(),
+            }
+        };
+        let isa_requirements = self.available_in_isa();
+        if !matches_isa_flags(&isa_requirements) {
+            panic!(
+                "Cannot emit inst '{:?}' for target; failed to match ISA requirements: {:?}",
+                self, isa_requirements
+            )
+        }
+
+        // N.B.: we *must* not exceed the "worst-case size" used to compute
+        // where to insert islands, except when islands are explicitly triggered
+        // (with an `EmitIsland`). We check this in debug builds. This is `mut`
+        // to allow disabling the check for `JTSequence`, which is always
+        // emitted following an `EmitIsland`.
+        let mut start_off = sink.cur_offset();
+
+        match self {
+            &Inst::AluRRR { alu_op, rd, rn, rm } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                let (opcode, have_rr) = match alu_op {
+                    ALUOp::Add32 => (0xb9f8, true),        // ARK
+                    ALUOp::Add64 => (0xb9e8, true),        // AGRK
+                    ALUOp::AddLogical32 => (0xb9fa, true), // ALRK
+                    ALUOp::AddLogical64 => (0xb9ea, true), // ALGRK
+                    ALUOp::Sub32 => (0xb9f9, true),        // SRK
+                    ALUOp::Sub64 => (0xb9e9, true),        // SGRK
+                    ALUOp::SubLogical32 => (0xb9fb, true), // SLRK
+                    ALUOp::SubLogical64 => (0xb9eb, true), // SLGRK
+                    ALUOp::Mul32 => (0xb9fd, true),        // MSRKC
+                    ALUOp::Mul64 => (0xb9ed, true),        // MSGRKC
+                    ALUOp::And32 => (0xb9f4, true),        // NRK
+                    ALUOp::And64 => (0xb9e4, true),        // NGRK
+                    ALUOp::Orr32 => (0xb9f6, true),        // ORK
+                    ALUOp::Orr64 => (0xb9e6, true),        // OGRK
+                    ALUOp::Xor32 => (0xb9f7, true),        // XRK
+                    ALUOp::Xor64 => (0xb9e7, true),        // XGRK
+                    ALUOp::NotAnd32 => (0xb974, false),    // NNRK
+                    ALUOp::NotAnd64 => (0xb964, false),    // NNGRK
+                    ALUOp::NotOrr32 => (0xb976, false),    // NORK
+                    ALUOp::NotOrr64 => (0xb966, false),    // NOGRK
+                    ALUOp::NotXor32 => (0xb977, false),    // NXRK
+                    ALUOp::NotXor64 => (0xb967, false),    // NXGRK
+                    ALUOp::AndNot32 => (0xb9f5, false),    // NCRK
+                    ALUOp::AndNot64 => (0xb9e5, false),    // NCGRK
+                    ALUOp::OrrNot32 => (0xb975, false),    // OCRK
+                    ALUOp::OrrNot64 => (0xb965, false),    // OCGRK
+                    _ => unreachable!(),
+                };
+                if have_rr && rd.to_reg() == rn {
+                    let inst = Inst::AluRR {
+                        alu_op,
+                        rd,
+                        ri: rn,
+                        rm,
+                    };
+                    inst.emit(&[], sink, emit_info, state);
+                } else {
+                    put(sink, &enc_rrf_ab(opcode, rd.to_reg(), rn, rm, 0));
+                }
+            }
+            &Inst::AluRRSImm16 {
+                alu_op,
+                rd,
+                rn,
+                imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                if rd.to_reg() == rn {
+                    let inst = Inst::AluRSImm16 {
+                        alu_op,
+                        rd,
+                        ri: rn,
+                        imm,
+                    };
+                    inst.emit(&[], sink, emit_info, state);
+                } else {
+                    let opcode = match alu_op {
+                        ALUOp::Add32 => 0xecd8, // AHIK
+                        ALUOp::Add64 => 0xecd9, // AGHIK
+                        _ => unreachable!(),
+                    };
+                    put(sink, &enc_rie_d(opcode, rd.to_reg(), rn, imm as u16));
+                }
+            }
+            &Inst::AluRR { alu_op, rd, ri, rm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let rm = allocs.next(rm);
+
+                let (opcode, is_rre) = match alu_op {
+                    ALUOp::Add32 => (0x1a, false),              // AR
+                    ALUOp::Add64 => (0xb908, true),             // AGR
+                    ALUOp::Add64Ext32 => (0xb918, true),        // AGFR
+                    ALUOp::AddLogical32 => (0x1e, false),       // ALR
+                    ALUOp::AddLogical64 => (0xb90a, true),      // ALGR
+                    ALUOp::AddLogical64Ext32 => (0xb91a, true), // ALGFR
+                    ALUOp::Sub32 => (0x1b, false),              // SR
+                    ALUOp::Sub64 => (0xb909, true),             // SGR
+                    ALUOp::Sub64Ext32 => (0xb919, true),        // SGFR
+                    ALUOp::SubLogical32 => (0x1f, false),       // SLR
+                    ALUOp::SubLogical64 => (0xb90b, true),      // SLGR
+                    ALUOp::SubLogical64Ext32 => (0xb91b, true), // SLGFR
+                    ALUOp::Mul32 => (0xb252, true),             // MSR
+                    ALUOp::Mul64 => (0xb90c, true),             // MSGR
+                    ALUOp::Mul64Ext32 => (0xb91c, true),        // MSGFR
+                    ALUOp::And32 => (0x14, false),              // NR
+                    ALUOp::And64 => (0xb980, true),             // NGR
+                    ALUOp::Orr32 => (0x16, false),              // OR
+                    ALUOp::Orr64 => (0xb981, true),             // OGR
+                    ALUOp::Xor32 => (0x17, false),              // XR
+                    ALUOp::Xor64 => (0xb982, true),             // XGR
+                    _ => unreachable!(),
+                };
+                if is_rre {
+                    put(sink, &enc_rre(opcode, rd.to_reg(), rm));
+                } else {
+                    put(sink, &enc_rr(opcode, rd.to_reg(), rm));
+                }
+            }
+            &Inst::AluRX {
+                alu_op,
+                rd,
+                ri,
+                ref mem,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode_rx, opcode_rxy) = match alu_op {
+                    ALUOp::Add32 => (Some(0x5a), Some(0xe35a)),        // A(Y)
+                    ALUOp::Add32Ext16 => (Some(0x4a), Some(0xe37a)),   // AH(Y)
+                    ALUOp::Add64 => (None, Some(0xe308)),              // AG
+                    ALUOp::Add64Ext16 => (None, Some(0xe338)),         // AGH
+                    ALUOp::Add64Ext32 => (None, Some(0xe318)),         // AGF
+                    ALUOp::AddLogical32 => (Some(0x5e), Some(0xe35e)), // AL(Y)
+                    ALUOp::AddLogical64 => (None, Some(0xe30a)),       // ALG
+                    ALUOp::AddLogical64Ext32 => (None, Some(0xe31a)),  // ALGF
+                    ALUOp::Sub32 => (Some(0x5b), Some(0xe35b)),        // S(Y)
+                    ALUOp::Sub32Ext16 => (Some(0x4b), Some(0xe37b)),   // SH(Y)
+                    ALUOp::Sub64 => (None, Some(0xe309)),              // SG
+                    ALUOp::Sub64Ext16 => (None, Some(0xe339)),         // SGH
+                    ALUOp::Sub64Ext32 => (None, Some(0xe319)),         // SGF
+                    ALUOp::SubLogical32 => (Some(0x5f), Some(0xe35f)), // SL(Y)
+                    ALUOp::SubLogical64 => (None, Some(0xe30b)),       // SLG
+                    ALUOp::SubLogical64Ext32 => (None, Some(0xe31b)),  // SLGF
+                    ALUOp::Mul32 => (Some(0x71), Some(0xe351)),        // MS(Y)
+                    ALUOp::Mul32Ext16 => (Some(0x4c), Some(0xe37c)),   // MH(Y)
+                    ALUOp::Mul64 => (None, Some(0xe30c)),              // MSG
+                    ALUOp::Mul64Ext16 => (None, Some(0xe33c)),         // MSH
+                    ALUOp::Mul64Ext32 => (None, Some(0xe31c)),         // MSGF
+                    ALUOp::And32 => (Some(0x54), Some(0xe354)),        // N(Y)
+                    ALUOp::And64 => (None, Some(0xe380)),              // NG
+                    ALUOp::Orr32 => (Some(0x56), Some(0xe356)),        // O(Y)
+                    ALUOp::Orr64 => (None, Some(0xe381)),              // OG
+                    ALUOp::Xor32 => (Some(0x57), Some(0xe357)),        // X(Y)
+                    ALUOp::Xor64 => (None, Some(0xe382)),              // XG
+                    _ => unreachable!(),
+                };
+                let rd = rd.to_reg();
+                mem_emit(
+                    rd, &mem, opcode_rx, opcode_rxy, None, true, sink, emit_info, state,
+                );
+            }
+            &Inst::AluRSImm16 {
+                alu_op,
+                rd,
+                ri,
+                imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = match alu_op {
+                    ALUOp::Add32 => 0xa7a, // AHI
+                    ALUOp::Add64 => 0xa7b, // AGHI
+                    ALUOp::Mul32 => 0xa7c, // MHI
+                    ALUOp::Mul64 => 0xa7d, // MGHI
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ri_a(opcode, rd.to_reg(), imm as u16));
+            }
+            &Inst::AluRSImm32 {
+                alu_op,
+                rd,
+                ri,
+                imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = match alu_op {
+                    ALUOp::Add32 => 0xc29, // AFI
+                    ALUOp::Add64 => 0xc28, // AGFI
+                    ALUOp::Mul32 => 0xc21, // MSFI
+                    ALUOp::Mul64 => 0xc20, // MSGFI
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ril_a(opcode, rd.to_reg(), imm as u32));
+            }
+            &Inst::AluRUImm32 {
+                alu_op,
+                rd,
+                ri,
+                imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = match alu_op {
+                    ALUOp::AddLogical32 => 0xc2b, // ALFI
+                    ALUOp::AddLogical64 => 0xc2a, // ALGFI
+                    ALUOp::SubLogical32 => 0xc25, // SLFI
+                    ALUOp::SubLogical64 => 0xc24, // SLGFI
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ril_a(opcode, rd.to_reg(), imm));
+            }
+            &Inst::AluRUImm16Shifted {
+                alu_op,
+                rd,
+                ri,
+                imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = match (alu_op, imm.shift) {
+                    (ALUOp::And32, 0) => 0xa57, // NILL
+                    (ALUOp::And32, 1) => 0xa56, // NILH
+                    (ALUOp::And64, 0) => 0xa57, // NILL
+                    (ALUOp::And64, 1) => 0xa56, // NILH
+                    (ALUOp::And64, 2) => 0xa55, // NIHL
+                    (ALUOp::And64, 3) => 0xa54, // NIHL
+                    (ALUOp::Orr32, 0) => 0xa5b, // OILL
+                    (ALUOp::Orr32, 1) => 0xa5a, // OILH
+                    (ALUOp::Orr64, 0) => 0xa5b, // OILL
+                    (ALUOp::Orr64, 1) => 0xa5a, // OILH
+                    (ALUOp::Orr64, 2) => 0xa59, // OIHL
+                    (ALUOp::Orr64, 3) => 0xa58, // OIHH
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ri_a(opcode, rd.to_reg(), imm.bits));
+            }
+            &Inst::AluRUImm32Shifted {
+                alu_op,
+                rd,
+                ri,
+                imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = match (alu_op, imm.shift) {
+                    (ALUOp::And32, 0) => 0xc0b, // NILF
+                    (ALUOp::And64, 0) => 0xc0b, // NILF
+                    (ALUOp::And64, 1) => 0xc0a, // NIHF
+                    (ALUOp::Orr32, 0) => 0xc0d, // OILF
+                    (ALUOp::Orr64, 0) => 0xc0d, // OILF
+                    (ALUOp::Orr64, 1) => 0xc0c, // OILF
+                    (ALUOp::Xor32, 0) => 0xc07, // XILF
+                    (ALUOp::Xor64, 0) => 0xc07, // XILF
+                    (ALUOp::Xor64, 1) => 0xc06, // XILH
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ril_a(opcode, rd.to_reg(), imm.bits));
+            }
+
+            &Inst::SMulWide { rd, rn, rm } => {
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+                let rd1 = allocs.next_writable(rd.hi);
+                let rd2 = allocs.next_writable(rd.lo);
+                debug_assert_valid_regpair!(rd1.to_reg(), rd2.to_reg());
+
+                let opcode = 0xb9ec; // MGRK
+                put(sink, &enc_rrf_ab(opcode, rd1.to_reg(), rn, rm, 0));
+            }
+            &Inst::UMulWide { rd, ri, rn } => {
+                let rn = allocs.next(rn);
+                let rd1 = allocs.next_writable(rd.hi);
+                let rd2 = allocs.next_writable(rd.lo);
+                debug_assert_valid_regpair!(rd1.to_reg(), rd2.to_reg());
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd2.to_reg(), ri);
+
+                let opcode = 0xb986; // MLGR
+                put(sink, &enc_rre(opcode, rd1.to_reg(), rn));
+            }
+            &Inst::SDivMod32 { rd, ri, rn } => {
+                let rn = allocs.next(rn);
+                let rd1 = allocs.next_writable(rd.hi);
+                let rd2 = allocs.next_writable(rd.lo);
+                debug_assert_valid_regpair!(rd1.to_reg(), rd2.to_reg());
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd2.to_reg(), ri);
+
+                let opcode = 0xb91d; // DSGFR
+                let trap_code = TrapCode::IntegerDivisionByZero;
+                put_with_trap(sink, &enc_rre(opcode, rd1.to_reg(), rn), trap_code);
+            }
+            &Inst::SDivMod64 { rd, ri, rn } => {
+                let rn = allocs.next(rn);
+                let rd1 = allocs.next_writable(rd.hi);
+                let rd2 = allocs.next_writable(rd.lo);
+                debug_assert_valid_regpair!(rd1.to_reg(), rd2.to_reg());
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd2.to_reg(), ri);
+
+                let opcode = 0xb90d; // DSGR
+                let trap_code = TrapCode::IntegerDivisionByZero;
+                put_with_trap(sink, &enc_rre(opcode, rd1.to_reg(), rn), trap_code);
+            }
+            &Inst::UDivMod32 { rd, ri, rn } => {
+                let rn = allocs.next(rn);
+                let rd1 = allocs.next_writable(rd.hi);
+                let rd2 = allocs.next_writable(rd.lo);
+                debug_assert_valid_regpair!(rd1.to_reg(), rd2.to_reg());
+                let ri1 = allocs.next(ri.hi);
+                let ri2 = allocs.next(ri.lo);
+                debug_assert_eq!(rd1.to_reg(), ri1);
+                debug_assert_eq!(rd2.to_reg(), ri2);
+
+                let opcode = 0xb997; // DLR
+                let trap_code = TrapCode::IntegerDivisionByZero;
+                put_with_trap(sink, &enc_rre(opcode, rd1.to_reg(), rn), trap_code);
+            }
+            &Inst::UDivMod64 { rd, ri, rn } => {
+                let rn = allocs.next(rn);
+                let rd1 = allocs.next_writable(rd.hi);
+                let rd2 = allocs.next_writable(rd.lo);
+                debug_assert_valid_regpair!(rd1.to_reg(), rd2.to_reg());
+                let ri1 = allocs.next(ri.hi);
+                let ri2 = allocs.next(ri.lo);
+                debug_assert_eq!(rd1.to_reg(), ri1);
+                debug_assert_eq!(rd2.to_reg(), ri2);
+
+                let opcode = 0xb987; // DLGR
+                let trap_code = TrapCode::IntegerDivisionByZero;
+                put_with_trap(sink, &enc_rre(opcode, rd1.to_reg(), rn), trap_code);
+            }
+            &Inst::Flogr { rd, rn } => {
+                let rn = allocs.next(rn);
+                let rd1 = allocs.next_writable(rd.hi);
+                let rd2 = allocs.next_writable(rd.lo);
+                debug_assert_valid_regpair!(rd1.to_reg(), rd2.to_reg());
+
+                let opcode = 0xb983; // FLOGR
+                put(sink, &enc_rre(opcode, rd1.to_reg(), rn));
+            }
+
+            &Inst::ShiftRR {
+                shift_op,
+                rd,
+                rn,
+                shift_imm,
+                shift_reg,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let shift_reg = allocs.next(shift_reg);
+
+                let opcode = match shift_op {
+                    ShiftOp::RotL32 => 0xeb1d, // RLL
+                    ShiftOp::RotL64 => 0xeb1c, // RLLG
+                    ShiftOp::LShL32 => 0xebdf, // SLLK  (SLL ?)
+                    ShiftOp::LShL64 => 0xeb0d, // SLLG
+                    ShiftOp::LShR32 => 0xebde, // SRLK  (SRL ?)
+                    ShiftOp::LShR64 => 0xeb0c, // SRLG
+                    ShiftOp::AShR32 => 0xebdc, // SRAK  (SRA ?)
+                    ShiftOp::AShR64 => 0xeb0a, // SRAG
+                };
+                put(
+                    sink,
+                    &enc_rsy(opcode, rd.to_reg(), rn, shift_reg, shift_imm.into()),
+                );
+            }
+
+            &Inst::RxSBG {
+                op,
+                rd,
+                ri,
+                rn,
+                start_bit,
+                end_bit,
+                rotate_amt,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let rn = allocs.next(rn);
+
+                let opcode = match op {
+                    RxSBGOp::Insert => 0xec59, // RISBGN
+                    RxSBGOp::And => 0xec54,    // RNSBG
+                    RxSBGOp::Or => 0xec56,     // ROSBG
+                    RxSBGOp::Xor => 0xec57,    // RXSBG
+                };
+                put(
+                    sink,
+                    &enc_rie_f(
+                        opcode,
+                        rd.to_reg(),
+                        rn,
+                        start_bit,
+                        end_bit,
+                        (rotate_amt as u8) & 63,
+                    ),
+                );
+            }
+
+            &Inst::RxSBGTest {
+                op,
+                rd,
+                rn,
+                start_bit,
+                end_bit,
+                rotate_amt,
+            } => {
+                let rd = allocs.next(rd);
+                let rn = allocs.next(rn);
+
+                let opcode = match op {
+                    RxSBGOp::And => 0xec54, // RNSBG
+                    RxSBGOp::Or => 0xec56,  // ROSBG
+                    RxSBGOp::Xor => 0xec57, // RXSBG
+                    _ => unreachable!(),
+                };
+                put(
+                    sink,
+                    &enc_rie_f(
+                        opcode,
+                        rd,
+                        rn,
+                        start_bit | 0x80,
+                        end_bit,
+                        (rotate_amt as u8) & 63,
+                    ),
+                );
+            }
+
+            &Inst::UnaryRR { op, rd, rn } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                match op {
+                    UnaryOp::Abs32 => {
+                        let opcode = 0x10; // LPR
+                        put(sink, &enc_rr(opcode, rd.to_reg(), rn));
+                    }
+                    UnaryOp::Abs64 => {
+                        let opcode = 0xb900; // LPGR
+                        put(sink, &enc_rre(opcode, rd.to_reg(), rn));
+                    }
+                    UnaryOp::Abs64Ext32 => {
+                        let opcode = 0xb910; // LPGFR
+                        put(sink, &enc_rre(opcode, rd.to_reg(), rn));
+                    }
+                    UnaryOp::Neg32 => {
+                        let opcode = 0x13; // LCR
+                        put(sink, &enc_rr(opcode, rd.to_reg(), rn));
+                    }
+                    UnaryOp::Neg64 => {
+                        let opcode = 0xb903; // LCGR
+                        put(sink, &enc_rre(opcode, rd.to_reg(), rn));
+                    }
+                    UnaryOp::Neg64Ext32 => {
+                        let opcode = 0xb913; // LCGFR
+                        put(sink, &enc_rre(opcode, rd.to_reg(), rn));
+                    }
+                    UnaryOp::PopcntByte => {
+                        let opcode = 0xb9e1; // POPCNT
+                        put(sink, &enc_rrf_cde(opcode, rd.to_reg(), rn, 0, 0));
+                    }
+                    UnaryOp::PopcntReg => {
+                        let opcode = 0xb9e1; // POPCNT
+                        put(sink, &enc_rrf_cde(opcode, rd.to_reg(), rn, 8, 0));
+                    }
+                    UnaryOp::BSwap32 => {
+                        let opcode = 0xb91f; // LRVR
+                        put(sink, &enc_rre(opcode, rd.to_reg(), rn));
+                    }
+                    UnaryOp::BSwap64 => {
+                        let opcode = 0xb90f; // LRVRG
+                        put(sink, &enc_rre(opcode, rd.to_reg(), rn));
+                    }
+                }
+            }
+
+            &Inst::Extend {
+                rd,
+                rn,
+                signed,
+                from_bits,
+                to_bits,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                let opcode = match (signed, from_bits, to_bits) {
+                    (_, 1, 32) => 0xb926,      // LBR
+                    (_, 1, 64) => 0xb906,      // LGBR
+                    (false, 8, 32) => 0xb994,  // LLCR
+                    (false, 8, 64) => 0xb984,  // LLGCR
+                    (true, 8, 32) => 0xb926,   // LBR
+                    (true, 8, 64) => 0xb906,   // LGBR
+                    (false, 16, 32) => 0xb995, // LLHR
+                    (false, 16, 64) => 0xb985, // LLGHR
+                    (true, 16, 32) => 0xb927,  // LHR
+                    (true, 16, 64) => 0xb907,  // LGHR
+                    (false, 32, 64) => 0xb916, // LLGFR
+                    (true, 32, 64) => 0xb914,  // LGFR
+                    _ => panic!(
+                        "Unsupported extend combination: signed = {}, from_bits = {}, to_bits = {}",
+                        signed, from_bits, to_bits
+                    ),
+                };
+                put(sink, &enc_rre(opcode, rd.to_reg(), rn));
+            }
+
+            &Inst::CmpRR { op, rn, rm } => {
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                let (opcode, is_rre) = match op {
+                    CmpOp::CmpS32 => (0x19, false),       // CR
+                    CmpOp::CmpS64 => (0xb920, true),      // CGR
+                    CmpOp::CmpS64Ext32 => (0xb930, true), // CGFR
+                    CmpOp::CmpL32 => (0x15, false),       // CLR
+                    CmpOp::CmpL64 => (0xb921, true),      // CLGR
+                    CmpOp::CmpL64Ext32 => (0xb931, true), // CLGFR
+                    _ => unreachable!(),
+                };
+                if is_rre {
+                    put(sink, &enc_rre(opcode, rn, rm));
+                } else {
+                    put(sink, &enc_rr(opcode, rn, rm));
+                }
+            }
+            &Inst::CmpRX { op, rn, ref mem } => {
+                let rn = allocs.next(rn);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode_rx, opcode_rxy, opcode_ril) = match op {
+                    CmpOp::CmpS32 => (Some(0x59), Some(0xe359), Some(0xc6d)), // C(Y), CRL
+                    CmpOp::CmpS32Ext16 => (Some(0x49), Some(0xe379), Some(0xc65)), // CH(Y), CHRL
+                    CmpOp::CmpS64 => (None, Some(0xe320), Some(0xc68)),       // CG, CGRL
+                    CmpOp::CmpS64Ext16 => (None, Some(0xe334), Some(0xc64)),  // CGH, CGHRL
+                    CmpOp::CmpS64Ext32 => (None, Some(0xe330), Some(0xc6c)),  // CGF, CGFRL
+                    CmpOp::CmpL32 => (Some(0x55), Some(0xe355), Some(0xc6f)), // CL(Y), CLRL
+                    CmpOp::CmpL32Ext16 => (None, None, Some(0xc67)),          // CLHRL
+                    CmpOp::CmpL64 => (None, Some(0xe321), Some(0xc6a)),       // CLG, CLGRL
+                    CmpOp::CmpL64Ext16 => (None, None, Some(0xc66)),          // CLGHRL
+                    CmpOp::CmpL64Ext32 => (None, Some(0xe331), Some(0xc6e)),  // CLGF, CLGFRL
+                };
+                mem_emit(
+                    rn, &mem, opcode_rx, opcode_rxy, opcode_ril, true, sink, emit_info, state,
+                );
+            }
+            &Inst::CmpRSImm16 { op, rn, imm } => {
+                let rn = allocs.next(rn);
+
+                let opcode = match op {
+                    CmpOp::CmpS32 => 0xa7e, // CHI
+                    CmpOp::CmpS64 => 0xa7f, // CGHI
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ri_a(opcode, rn, imm as u16));
+            }
+            &Inst::CmpRSImm32 { op, rn, imm } => {
+                let rn = allocs.next(rn);
+
+                let opcode = match op {
+                    CmpOp::CmpS32 => 0xc2d, // CFI
+                    CmpOp::CmpS64 => 0xc2c, // CGFI
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ril_a(opcode, rn, imm as u32));
+            }
+            &Inst::CmpRUImm32 { op, rn, imm } => {
+                let rn = allocs.next(rn);
+
+                let opcode = match op {
+                    CmpOp::CmpL32 => 0xc2f, // CLFI
+                    CmpOp::CmpL64 => 0xc2e, // CLGFI
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ril_a(opcode, rn, imm));
+            }
+            &Inst::CmpTrapRR {
+                op,
+                rn,
+                rm,
+                cond,
+                trap_code,
+            } => {
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                let opcode = match op {
+                    CmpOp::CmpS32 => 0xb972, // CRT
+                    CmpOp::CmpS64 => 0xb960, // CGRT
+                    CmpOp::CmpL32 => 0xb973, // CLRT
+                    CmpOp::CmpL64 => 0xb961, // CLGRT
+                    _ => unreachable!(),
+                };
+                put_with_trap(
+                    sink,
+                    &enc_rrf_cde(opcode, rn, rm, cond.bits(), 0),
+                    trap_code,
+                );
+            }
+            &Inst::CmpTrapRSImm16 {
+                op,
+                rn,
+                imm,
+                cond,
+                trap_code,
+            } => {
+                let rn = allocs.next(rn);
+
+                let opcode = match op {
+                    CmpOp::CmpS32 => 0xec72, // CIT
+                    CmpOp::CmpS64 => 0xec70, // CGIT
+                    _ => unreachable!(),
+                };
+                put_with_trap(
+                    sink,
+                    &enc_rie_a(opcode, rn, imm as u16, cond.bits()),
+                    trap_code,
+                );
+            }
+            &Inst::CmpTrapRUImm16 {
+                op,
+                rn,
+                imm,
+                cond,
+                trap_code,
+            } => {
+                let rn = allocs.next(rn);
+
+                let opcode = match op {
+                    CmpOp::CmpL32 => 0xec73, // CLFIT
+                    CmpOp::CmpL64 => 0xec71, // CLGIT
+                    _ => unreachable!(),
+                };
+                put_with_trap(sink, &enc_rie_a(opcode, rn, imm, cond.bits()), trap_code);
+            }
+
+            &Inst::AtomicRmw {
+                alu_op,
+                rd,
+                rn,
+                ref mem,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let mem = mem.with_allocs(allocs);
+
+                let opcode = match alu_op {
+                    ALUOp::Add32 => 0xebf8,        // LAA
+                    ALUOp::Add64 => 0xebe8,        // LAAG
+                    ALUOp::AddLogical32 => 0xebfa, // LAAL
+                    ALUOp::AddLogical64 => 0xebea, // LAALG
+                    ALUOp::And32 => 0xebf4,        // LAN
+                    ALUOp::And64 => 0xebe4,        // LANG
+                    ALUOp::Orr32 => 0xebf6,        // LAO
+                    ALUOp::Orr64 => 0xebe6,        // LAOG
+                    ALUOp::Xor32 => 0xebf7,        // LAX
+                    ALUOp::Xor64 => 0xebe7,        // LAXG
+                    _ => unreachable!(),
+                };
+
+                let rd = rd.to_reg();
+                mem_rs_emit(
+                    rd,
+                    rn,
+                    &mem,
+                    None,
+                    Some(opcode),
+                    true,
+                    sink,
+                    emit_info,
+                    state,
+                );
+            }
+            &Inst::Loop { ref body, cond } => {
+                // This sequence is *one* instruction in the vcode, and is expanded only here at
+                // emission time, because it requires branching to internal labels.
+                let loop_label = sink.get_label();
+                let done_label = sink.get_label();
+
+                // Emit label at the start of the loop.
+                sink.bind_label(loop_label, &mut state.ctrl_plane);
+
+                for inst in (&body).into_iter() {
+                    match &inst {
+                        // Replace a CondBreak with a branch to done_label.
+                        &Inst::CondBreak { cond } => {
+                            let inst = Inst::OneWayCondBr {
+                                target: done_label,
+                                cond: *cond,
+                            };
+                            inst.emit_with_alloc_consumer(allocs, sink, emit_info, state);
+                        }
+                        _ => inst.emit_with_alloc_consumer(allocs, sink, emit_info, state),
+                    };
+                }
+
+                let inst = Inst::OneWayCondBr {
+                    target: loop_label,
+                    cond,
+                };
+                inst.emit(&[], sink, emit_info, state);
+
+                // Emit label at the end of the loop.
+                sink.bind_label(done_label, &mut state.ctrl_plane);
+            }
+            &Inst::CondBreak { .. } => unreachable!(), // Only valid inside a Loop.
+            &Inst::AtomicCas32 {
+                rd,
+                ri,
+                rn,
+                ref mem,
+            }
+            | &Inst::AtomicCas64 {
+                rd,
+                ri,
+                rn,
+                ref mem,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let rn = allocs.next(rn);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode_rs, opcode_rsy) = match self {
+                    &Inst::AtomicCas32 { .. } => (Some(0xba), Some(0xeb14)), // CS(Y)
+                    &Inst::AtomicCas64 { .. } => (None, Some(0xeb30)),       // CSG
+                    _ => unreachable!(),
+                };
+
+                let rd = rd.to_reg();
+                mem_rs_emit(
+                    rd, rn, &mem, opcode_rs, opcode_rsy, true, sink, emit_info, state,
+                );
+            }
+            &Inst::Fence => {
+                put(sink, &enc_e(0x07e0));
+            }
+
+            &Inst::Load32 { rd, ref mem }
+            | &Inst::Load32ZExt8 { rd, ref mem }
+            | &Inst::Load32SExt8 { rd, ref mem }
+            | &Inst::Load32ZExt16 { rd, ref mem }
+            | &Inst::Load32SExt16 { rd, ref mem }
+            | &Inst::Load64 { rd, ref mem }
+            | &Inst::Load64ZExt8 { rd, ref mem }
+            | &Inst::Load64SExt8 { rd, ref mem }
+            | &Inst::Load64ZExt16 { rd, ref mem }
+            | &Inst::Load64SExt16 { rd, ref mem }
+            | &Inst::Load64ZExt32 { rd, ref mem }
+            | &Inst::Load64SExt32 { rd, ref mem }
+            | &Inst::LoadRev16 { rd, ref mem }
+            | &Inst::LoadRev32 { rd, ref mem }
+            | &Inst::LoadRev64 { rd, ref mem } => {
+                let rd = allocs.next_writable(rd);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode_rx, opcode_rxy, opcode_ril) = match self {
+                    &Inst::Load32 { .. } => (Some(0x58), Some(0xe358), Some(0xc4d)), // L(Y), LRL
+                    &Inst::Load32ZExt8 { .. } => (None, Some(0xe394), None),         // LLC
+                    &Inst::Load32SExt8 { .. } => (None, Some(0xe376), None),         // LB
+                    &Inst::Load32ZExt16 { .. } => (None, Some(0xe395), Some(0xc42)), // LLH, LLHRL
+                    &Inst::Load32SExt16 { .. } => (Some(0x48), Some(0xe378), Some(0xc45)), // LH(Y), LHRL
+                    &Inst::Load64 { .. } => (None, Some(0xe304), Some(0xc48)), // LG, LGRL
+                    &Inst::Load64ZExt8 { .. } => (None, Some(0xe390), None),   // LLGC
+                    &Inst::Load64SExt8 { .. } => (None, Some(0xe377), None),   // LGB
+                    &Inst::Load64ZExt16 { .. } => (None, Some(0xe391), Some(0xc46)), // LLGH, LLGHRL
+                    &Inst::Load64SExt16 { .. } => (None, Some(0xe315), Some(0xc44)), // LGH, LGHRL
+                    &Inst::Load64ZExt32 { .. } => (None, Some(0xe316), Some(0xc4e)), // LLGF, LLGFRL
+                    &Inst::Load64SExt32 { .. } => (None, Some(0xe314), Some(0xc4c)), // LGF, LGFRL
+                    &Inst::LoadRev16 { .. } => (None, Some(0xe31f), None),     // LRVH
+                    &Inst::LoadRev32 { .. } => (None, Some(0xe31e), None),     // LRV
+                    &Inst::LoadRev64 { .. } => (None, Some(0xe30f), None),     // LRVG
+                    _ => unreachable!(),
+                };
+                let rd = rd.to_reg();
+                mem_emit(
+                    rd, &mem, opcode_rx, opcode_rxy, opcode_ril, true, sink, emit_info, state,
+                );
+            }
+
+            &Inst::Store8 { rd, ref mem }
+            | &Inst::Store16 { rd, ref mem }
+            | &Inst::Store32 { rd, ref mem }
+            | &Inst::Store64 { rd, ref mem }
+            | &Inst::StoreRev16 { rd, ref mem }
+            | &Inst::StoreRev32 { rd, ref mem }
+            | &Inst::StoreRev64 { rd, ref mem } => {
+                let rd = allocs.next(rd);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode_rx, opcode_rxy, opcode_ril) = match self {
+                    &Inst::Store8 { .. } => (Some(0x42), Some(0xe372), None), // STC(Y)
+                    &Inst::Store16 { .. } => (Some(0x40), Some(0xe370), Some(0xc47)), // STH(Y), STHRL
+                    &Inst::Store32 { .. } => (Some(0x50), Some(0xe350), Some(0xc4f)), // ST(Y), STRL
+                    &Inst::Store64 { .. } => (None, Some(0xe324), Some(0xc4b)),       // STG, STGRL
+                    &Inst::StoreRev16 { .. } => (None, Some(0xe33f), None),           // STRVH
+                    &Inst::StoreRev32 { .. } => (None, Some(0xe33e), None),           // STRV
+                    &Inst::StoreRev64 { .. } => (None, Some(0xe32f), None),           // STRVG
+                    _ => unreachable!(),
+                };
+                mem_emit(
+                    rd, &mem, opcode_rx, opcode_rxy, opcode_ril, true, sink, emit_info, state,
+                );
+            }
+            &Inst::StoreImm8 { imm, ref mem } => {
+                let mem = mem.with_allocs(allocs);
+
+                let opcode_si = 0x92; // MVI
+                let opcode_siy = 0xeb52; // MVIY
+                mem_imm8_emit(
+                    imm, &mem, opcode_si, opcode_siy, true, sink, emit_info, state,
+                );
+            }
+            &Inst::StoreImm16 { imm, ref mem }
+            | &Inst::StoreImm32SExt16 { imm, ref mem }
+            | &Inst::StoreImm64SExt16 { imm, ref mem } => {
+                let mem = mem.with_allocs(allocs);
+
+                let opcode = match self {
+                    &Inst::StoreImm16 { .. } => 0xe544,       // MVHHI
+                    &Inst::StoreImm32SExt16 { .. } => 0xe54c, // MVHI
+                    &Inst::StoreImm64SExt16 { .. } => 0xe548, // MVGHI
+                    _ => unreachable!(),
+                };
+                mem_imm16_emit(imm, &mem, opcode, true, sink, emit_info, state);
+            }
+            &Inst::Mvc {
+                ref dst,
+                ref src,
+                len_minus_one,
+            } => {
+                let dst = dst.with_allocs(allocs);
+                let src = src.with_allocs(allocs);
+                let opcode = 0xd2; // MVC
+                mem_mem_emit(&dst, &src, len_minus_one, opcode, true, sink, state);
+            }
+
+            &Inst::LoadMultiple64 { rt, rt2, ref mem } => {
+                let mem = mem.with_allocs(allocs);
+
+                let opcode = 0xeb04; // LMG
+                let rt = rt.to_reg();
+                let rt2 = rt2.to_reg();
+                mem_rs_emit(
+                    rt,
+                    rt2,
+                    &mem,
+                    None,
+                    Some(opcode),
+                    true,
+                    sink,
+                    emit_info,
+                    state,
+                );
+            }
+            &Inst::StoreMultiple64 { rt, rt2, ref mem } => {
+                let mem = mem.with_allocs(allocs);
+
+                let opcode = 0xeb24; // STMG
+                mem_rs_emit(
+                    rt,
+                    rt2,
+                    &mem,
+                    None,
+                    Some(opcode),
+                    true,
+                    sink,
+                    emit_info,
+                    state,
+                );
+            }
+
+            &Inst::LoadAddr { rd, ref mem } => {
+                let rd = allocs.next_writable(rd);
+                let mem = mem.with_allocs(allocs);
+
+                let opcode_rx = Some(0x41); // LA
+                let opcode_rxy = Some(0xe371); // LAY
+                let opcode_ril = Some(0xc00); // LARL
+                let rd = rd.to_reg();
+                mem_emit(
+                    rd, &mem, opcode_rx, opcode_rxy, opcode_ril, false, sink, emit_info, state,
+                );
+            }
+
+            &Inst::Mov64 { rd, rm } => {
+                let rd = allocs.next_writable(rd);
+                let rm = allocs.next(rm);
+
+                let opcode = 0xb904; // LGR
+                put(sink, &enc_rre(opcode, rd.to_reg(), rm));
+            }
+            &Inst::MovPReg { rd, rm } => {
+                let rm: Reg = rm.into();
+                debug_assert!([regs::gpr(0), regs::gpr(14), regs::gpr(15)].contains(&rm));
+                let rd = allocs.next_writable(rd);
+                Inst::Mov64 { rd, rm }.emit(&[], sink, emit_info, state);
+            }
+            &Inst::Mov32 { rd, rm } => {
+                let rd = allocs.next_writable(rd);
+                let rm = allocs.next(rm);
+
+                let opcode = 0x18; // LR
+                put(sink, &enc_rr(opcode, rd.to_reg(), rm));
+            }
+            &Inst::Mov32Imm { rd, imm } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = 0xc09; // IILF
+                put(sink, &enc_ril_a(opcode, rd.to_reg(), imm));
+            }
+            &Inst::Mov32SImm16 { rd, imm } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = 0xa78; // LHI
+                put(sink, &enc_ri_a(opcode, rd.to_reg(), imm as u16));
+            }
+            &Inst::Mov64SImm16 { rd, imm } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = 0xa79; // LGHI
+                put(sink, &enc_ri_a(opcode, rd.to_reg(), imm as u16));
+            }
+            &Inst::Mov64SImm32 { rd, imm } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = 0xc01; // LGFI
+                put(sink, &enc_ril_a(opcode, rd.to_reg(), imm as u32));
+            }
+            &Inst::CMov32 { rd, cond, ri, rm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let rm = allocs.next(rm);
+
+                let opcode = 0xb9f2; // LOCR
+                put(sink, &enc_rrf_cde(opcode, rd.to_reg(), rm, cond.bits(), 0));
+            }
+            &Inst::CMov64 { rd, cond, ri, rm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let rm = allocs.next(rm);
+
+                let opcode = 0xb9e2; // LOCGR
+                put(sink, &enc_rrf_cde(opcode, rd.to_reg(), rm, cond.bits(), 0));
+            }
+            &Inst::CMov32SImm16 { rd, cond, ri, imm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = 0xec42; // LOCHI
+                put(
+                    sink,
+                    &enc_rie_g(opcode, rd.to_reg(), imm as u16, cond.bits()),
+                );
+            }
+            &Inst::CMov64SImm16 { rd, cond, ri, imm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = 0xec46; // LOCGHI
+                put(
+                    sink,
+                    &enc_rie_g(opcode, rd.to_reg(), imm as u16, cond.bits()),
+                );
+            }
+            &Inst::Mov64UImm16Shifted { rd, imm } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = match imm.shift {
+                    0 => 0xa5f, // LLILL
+                    1 => 0xa5e, // LLILH
+                    2 => 0xa5d, // LLIHL
+                    3 => 0xa5c, // LLIHH
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ri_a(opcode, rd.to_reg(), imm.bits));
+            }
+            &Inst::Mov64UImm32Shifted { rd, imm } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = match imm.shift {
+                    0 => 0xc0f, // LLILF
+                    1 => 0xc0e, // LLIHF
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ril_a(opcode, rd.to_reg(), imm.bits));
+            }
+            &Inst::Insert64UImm16Shifted { rd, ri, imm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = match imm.shift {
+                    0 => 0xa53, // IILL
+                    1 => 0xa52, // IILH
+                    2 => 0xa51, // IIHL
+                    3 => 0xa50, // IIHH
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ri_a(opcode, rd.to_reg(), imm.bits));
+            }
+            &Inst::Insert64UImm32Shifted { rd, ri, imm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = match imm.shift {
+                    0 => 0xc09, // IILF
+                    1 => 0xc08, // IIHF
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_ril_a(opcode, rd.to_reg(), imm.bits));
+            }
+            &Inst::LoadAR { rd, ar } => {
+                let rd = allocs.next_writable(rd);
+                let opcode = 0xb24f; // EAR
+                put(sink, &enc_rre(opcode, rd.to_reg(), gpr(ar)));
+            }
+
+            &Inst::InsertAR { rd, ri, ar } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = 0xb24f; // EAR
+                put(sink, &enc_rre(opcode, rd.to_reg(), gpr(ar)));
+            }
+            &Inst::LoadSymbolReloc {
+                rd,
+                ref symbol_reloc,
+            } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = 0xa75; // BRAS
+                let reg = writable_spilltmp_reg().to_reg();
+                put(sink, &enc_ri_b(opcode, reg, 12));
+                let (reloc, name, offset) = match &**symbol_reloc {
+                    SymbolReloc::Absolute { name, offset } => (Reloc::Abs8, name, *offset),
+                    SymbolReloc::TlsGd { name } => (Reloc::S390xTlsGd64, name, 0),
+                };
+                sink.add_reloc(reloc, name, offset);
+                sink.put8(0);
+                let inst = Inst::Load64 {
+                    rd,
+                    mem: MemArg::reg(reg, MemFlags::trusted()),
+                };
+                inst.emit(&[], sink, emit_info, state);
+            }
+
+            &Inst::FpuMove32 { rd, rn } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                if is_fpr(rd.to_reg()) && is_fpr(rn) {
+                    let opcode = 0x38; // LER
+                    put(sink, &enc_rr(opcode, rd.to_reg(), rn));
+                } else {
+                    let opcode = 0xe756; // VLR
+                    put(sink, &enc_vrr_a(opcode, rd.to_reg(), rn, 0, 0, 0));
+                }
+            }
+            &Inst::FpuMove64 { rd, rn } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                if is_fpr(rd.to_reg()) && is_fpr(rn) {
+                    let opcode = 0x28; // LDR
+                    put(sink, &enc_rr(opcode, rd.to_reg(), rn));
+                } else {
+                    let opcode = 0xe756; // VLR
+                    put(sink, &enc_vrr_a(opcode, rd.to_reg(), rn, 0, 0, 0));
+                }
+            }
+            &Inst::FpuCMov32 { rd, cond, ri, rm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let rm = allocs.next(rm);
+
+                if is_fpr(rd.to_reg()) && is_fpr(rm) {
+                    let opcode = 0xa74; // BCR
+                    put(sink, &enc_ri_c(opcode, cond.invert().bits(), 4 + 2));
+                    let opcode = 0x38; // LER
+                    put(sink, &enc_rr(opcode, rd.to_reg(), rm));
+                } else {
+                    let opcode = 0xa74; // BCR
+                    put(sink, &enc_ri_c(opcode, cond.invert().bits(), 4 + 6));
+                    let opcode = 0xe756; // VLR
+                    put(sink, &enc_vrr_a(opcode, rd.to_reg(), rm, 0, 0, 0));
+                }
+            }
+            &Inst::FpuCMov64 { rd, cond, ri, rm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let rm = allocs.next(rm);
+
+                if is_fpr(rd.to_reg()) && is_fpr(rm) {
+                    let opcode = 0xa74; // BCR
+                    put(sink, &enc_ri_c(opcode, cond.invert().bits(), 4 + 2));
+                    let opcode = 0x28; // LDR
+                    put(sink, &enc_rr(opcode, rd.to_reg(), rm));
+                } else {
+                    let opcode = 0xa74; // BCR
+                    put(sink, &enc_ri_c(opcode, cond.invert().bits(), 4 + 6));
+                    let opcode = 0xe756; // VLR
+                    put(sink, &enc_vrr_a(opcode, rd.to_reg(), rm, 0, 0, 0));
+                }
+            }
+            &Inst::LoadFpuConst32 { rd, const_data } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = 0xa75; // BRAS
+                let reg = writable_spilltmp_reg().to_reg();
+                put(sink, &enc_ri_b(opcode, reg, 8));
+                sink.put4(const_data.swap_bytes());
+                let inst = Inst::VecLoadLaneUndef {
+                    size: 32,
+                    rd,
+                    mem: MemArg::reg(reg, MemFlags::trusted()),
+                    lane_imm: 0,
+                };
+                inst.emit(&[], sink, emit_info, state);
+            }
+            &Inst::LoadFpuConst64 { rd, const_data } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = 0xa75; // BRAS
+                let reg = writable_spilltmp_reg().to_reg();
+                put(sink, &enc_ri_b(opcode, reg, 12));
+                sink.put8(const_data.swap_bytes());
+                let inst = Inst::VecLoadLaneUndef {
+                    size: 64,
+                    rd,
+                    mem: MemArg::reg(reg, MemFlags::trusted()),
+                    lane_imm: 0,
+                };
+                inst.emit(&[], sink, emit_info, state);
+            }
+            &Inst::FpuRR { fpu_op, rd, rn } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                let (opcode, m3, m4, m5, opcode_fpr) = match fpu_op {
+                    FPUOp1::Abs32 => (0xe7cc, 2, 8, 2, Some(0xb300)), // WFPSO, LPEBR
+                    FPUOp1::Abs64 => (0xe7cc, 3, 8, 2, Some(0xb310)), // WFPSO, LPDBR
+                    FPUOp1::Abs32x4 => (0xe7cc, 2, 0, 2, None),       // VFPSO
+                    FPUOp1::Abs64x2 => (0xe7cc, 3, 0, 2, None),       // VFPSO
+                    FPUOp1::Neg32 => (0xe7cc, 2, 8, 0, Some(0xb303)), // WFPSO, LCEBR
+                    FPUOp1::Neg64 => (0xe7cc, 3, 8, 0, Some(0xb313)), // WFPSO, LCDBR
+                    FPUOp1::Neg32x4 => (0xe7cc, 2, 0, 0, None),       // VFPSO
+                    FPUOp1::Neg64x2 => (0xe7cc, 3, 0, 0, None),       // VFPSO
+                    FPUOp1::NegAbs32 => (0xe7cc, 2, 8, 1, Some(0xb301)), // WFPSO, LNEBR
+                    FPUOp1::NegAbs64 => (0xe7cc, 3, 8, 1, Some(0xb311)), // WFPSO, LNDBR
+                    FPUOp1::NegAbs32x4 => (0xe7cc, 2, 0, 1, None),    // VFPSO
+                    FPUOp1::NegAbs64x2 => (0xe7cc, 3, 0, 1, None),    // VFPSO
+                    FPUOp1::Sqrt32 => (0xe7ce, 2, 8, 0, Some(0xb314)), // WFSQ, SQEBR
+                    FPUOp1::Sqrt64 => (0xe7ce, 3, 8, 0, Some(0xb315)), // WFSQ, SQDBR
+                    FPUOp1::Sqrt32x4 => (0xe7ce, 2, 0, 0, None),      // VFSQ
+                    FPUOp1::Sqrt64x2 => (0xe7ce, 3, 0, 0, None),      // VFSQ
+                    FPUOp1::Cvt32To64 => (0xe7c4, 2, 8, 0, Some(0xb304)), // WFLL, LDEBR
+                    FPUOp1::Cvt32x4To64x2 => (0xe7c4, 2, 0, 0, None), // VFLL
+                };
+                if m4 == 8 && is_fpr(rd.to_reg()) && is_fpr(rn) {
+                    put(sink, &enc_rre(opcode_fpr.unwrap(), rd.to_reg(), rn));
+                } else {
+                    put(sink, &enc_vrr_a(opcode, rd.to_reg(), rn, m3, m4, m5));
+                }
+            }
+            &Inst::FpuRRR { fpu_op, rd, rn, rm } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                let (opcode, m4, m5, m6, opcode_fpr) = match fpu_op {
+                    FPUOp2::Add32 => (0xe7e3, 2, 8, 0, Some(0xb30a)), // WFA, AEBR
+                    FPUOp2::Add64 => (0xe7e3, 3, 8, 0, Some(0xb31a)), // WFA, ADBR
+                    FPUOp2::Add32x4 => (0xe7e3, 2, 0, 0, None),       // VFA
+                    FPUOp2::Add64x2 => (0xe7e3, 3, 0, 0, None),       // VFA
+                    FPUOp2::Sub32 => (0xe7e2, 2, 8, 0, Some(0xb30b)), // WFS, SEBR
+                    FPUOp2::Sub64 => (0xe7e2, 3, 8, 0, Some(0xb31b)), // WFS, SDBR
+                    FPUOp2::Sub32x4 => (0xe7e2, 2, 0, 0, None),       // VFS
+                    FPUOp2::Sub64x2 => (0xe7e2, 3, 0, 0, None),       // VFS
+                    FPUOp2::Mul32 => (0xe7e7, 2, 8, 0, Some(0xb317)), // WFM, MEEBR
+                    FPUOp2::Mul64 => (0xe7e7, 3, 8, 0, Some(0xb31c)), // WFM, MDBR
+                    FPUOp2::Mul32x4 => (0xe7e7, 2, 0, 0, None),       // VFM
+                    FPUOp2::Mul64x2 => (0xe7e7, 3, 0, 0, None),       // VFM
+                    FPUOp2::Div32 => (0xe7e5, 2, 8, 0, Some(0xb30d)), // WFD, DEBR
+                    FPUOp2::Div64 => (0xe7e5, 3, 8, 0, Some(0xb31d)), // WFD, DDBR
+                    FPUOp2::Div32x4 => (0xe7e5, 2, 0, 0, None),       // VFD
+                    FPUOp2::Div64x2 => (0xe7e5, 3, 0, 0, None),       // VFD
+                    FPUOp2::Max32 => (0xe7ef, 2, 8, 1, None),         // WFMAX
+                    FPUOp2::Max64 => (0xe7ef, 3, 8, 1, None),         // WFMAX
+                    FPUOp2::Max32x4 => (0xe7ef, 2, 0, 1, None),       // VFMAX
+                    FPUOp2::Max64x2 => (0xe7ef, 3, 0, 1, None),       // VFMAX
+                    FPUOp2::Min32 => (0xe7ee, 2, 8, 1, None),         // WFMIN
+                    FPUOp2::Min64 => (0xe7ee, 3, 8, 1, None),         // WFMIN
+                    FPUOp2::Min32x4 => (0xe7ee, 2, 0, 1, None),       // VFMIN
+                    FPUOp2::Min64x2 => (0xe7ee, 3, 0, 1, None),       // VFMIN
+                    FPUOp2::MaxPseudo32 => (0xe7ef, 2, 8, 3, None),   // WFMAX
+                    FPUOp2::MaxPseudo64 => (0xe7ef, 3, 8, 3, None),   // WFMAX
+                    FPUOp2::MaxPseudo32x4 => (0xe7ef, 2, 0, 3, None), // VFMAX
+                    FPUOp2::MaxPseudo64x2 => (0xe7ef, 3, 0, 3, None), // VFMAX
+                    FPUOp2::MinPseudo32 => (0xe7ee, 2, 8, 3, None),   // WFMIN
+                    FPUOp2::MinPseudo64 => (0xe7ee, 3, 8, 3, None),   // WFMIN
+                    FPUOp2::MinPseudo32x4 => (0xe7ee, 2, 0, 3, None), // VFMIN
+                    FPUOp2::MinPseudo64x2 => (0xe7ee, 3, 0, 3, None), // VFMIN
+                };
+                if m5 == 8 && opcode_fpr.is_some() && rd.to_reg() == rn && is_fpr(rn) && is_fpr(rm)
+                {
+                    put(sink, &enc_rre(opcode_fpr.unwrap(), rd.to_reg(), rm));
+                } else {
+                    put(sink, &enc_vrr_c(opcode, rd.to_reg(), rn, rm, m4, m5, m6));
+                }
+            }
+            &Inst::FpuRRRR {
+                fpu_op,
+                rd,
+                rn,
+                rm,
+                ra,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+                let ra = allocs.next(ra);
+
+                let (opcode, m5, m6, opcode_fpr) = match fpu_op {
+                    FPUOp3::MAdd32 => (0xe78f, 8, 2, Some(0xb30e)), // WFMA, MAEBR
+                    FPUOp3::MAdd64 => (0xe78f, 8, 3, Some(0xb31e)), // WFMA, MADBR
+                    FPUOp3::MAdd32x4 => (0xe78f, 0, 2, None),       // VFMA
+                    FPUOp3::MAdd64x2 => (0xe78f, 0, 3, None),       // VFMA
+                    FPUOp3::MSub32 => (0xe78e, 8, 2, Some(0xb30f)), // WFMS, MSEBR
+                    FPUOp3::MSub64 => (0xe78e, 8, 3, Some(0xb31f)), // WFMS, MSDBR
+                    FPUOp3::MSub32x4 => (0xe78e, 0, 2, None),       // VFMS
+                    FPUOp3::MSub64x2 => (0xe78e, 0, 3, None),       // VFMS
+                };
+                if m5 == 8 && rd.to_reg() == ra && is_fpr(rn) && is_fpr(rm) && is_fpr(ra) {
+                    put(sink, &enc_rrd(opcode_fpr.unwrap(), rd.to_reg(), rm, rn));
+                } else {
+                    put(sink, &enc_vrr_e(opcode, rd.to_reg(), rn, rm, ra, m5, m6));
+                }
+            }
+            &Inst::FpuRound { op, mode, rd, rn } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                let mode = match mode {
+                    FpuRoundMode::Current => 0,
+                    FpuRoundMode::ToNearest => 1,
+                    FpuRoundMode::ShorterPrecision => 3,
+                    FpuRoundMode::ToNearestTiesToEven => 4,
+                    FpuRoundMode::ToZero => 5,
+                    FpuRoundMode::ToPosInfinity => 6,
+                    FpuRoundMode::ToNegInfinity => 7,
+                };
+                let (opcode, m3, m4, opcode_fpr) = match op {
+                    FpuRoundOp::Cvt64To32 => (0xe7c5, 3, 8, Some(0xb344)), // WFLR, LEDBR(A)
+                    FpuRoundOp::Cvt64x2To32x4 => (0xe7c5, 3, 0, None),     // VFLR
+                    FpuRoundOp::Round32 => (0xe7c7, 2, 8, Some(0xb357)),   // WFI, FIEBR
+                    FpuRoundOp::Round64 => (0xe7c7, 3, 8, Some(0xb35f)),   // WFI, FIDBR
+                    FpuRoundOp::Round32x4 => (0xe7c7, 2, 0, None),         // VFI
+                    FpuRoundOp::Round64x2 => (0xe7c7, 3, 0, None),         // VFI
+                    FpuRoundOp::ToSInt32 => (0xe7c2, 2, 8, None),          // WCSFP
+                    FpuRoundOp::ToSInt64 => (0xe7c2, 3, 8, None),          // WCSFP
+                    FpuRoundOp::ToUInt32 => (0xe7c0, 2, 8, None),          // WCLFP
+                    FpuRoundOp::ToUInt64 => (0xe7c0, 3, 8, None),          // WCLFP
+                    FpuRoundOp::ToSInt32x4 => (0xe7c2, 2, 0, None),        // VCSFP
+                    FpuRoundOp::ToSInt64x2 => (0xe7c2, 3, 0, None),        // VCSFP
+                    FpuRoundOp::ToUInt32x4 => (0xe7c0, 2, 0, None),        // VCLFP
+                    FpuRoundOp::ToUInt64x2 => (0xe7c0, 3, 0, None),        // VCLFP
+                    FpuRoundOp::FromSInt32 => (0xe7c3, 2, 8, None),        // WCFPS
+                    FpuRoundOp::FromSInt64 => (0xe7c3, 3, 8, None),        // WCFPS
+                    FpuRoundOp::FromUInt32 => (0xe7c1, 2, 8, None),        // WCFPL
+                    FpuRoundOp::FromUInt64 => (0xe7c1, 3, 8, None),        // WCFPL
+                    FpuRoundOp::FromSInt32x4 => (0xe7c3, 2, 0, None),      // VCFPS
+                    FpuRoundOp::FromSInt64x2 => (0xe7c3, 3, 0, None),      // VCFPS
+                    FpuRoundOp::FromUInt32x4 => (0xe7c1, 2, 0, None),      // VCFPL
+                    FpuRoundOp::FromUInt64x2 => (0xe7c1, 3, 0, None),      // VCFPL
+                };
+                if m4 == 8 && opcode_fpr.is_some() && is_fpr(rd.to_reg()) && is_fpr(rn) {
+                    put(
+                        sink,
+                        &enc_rrf_cde(opcode_fpr.unwrap(), rd.to_reg(), rn, mode, 0),
+                    );
+                } else {
+                    put(sink, &enc_vrr_a(opcode, rd.to_reg(), rn, m3, m4, mode));
+                }
+            }
+            &Inst::FpuCmp32 { rn, rm } => {
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                if is_fpr(rn) && is_fpr(rm) {
+                    let opcode = 0xb309; // CEBR
+                    put(sink, &enc_rre(opcode, rn, rm));
+                } else {
+                    let opcode = 0xe7cb; // WFC
+                    put(sink, &enc_vrr_a(opcode, rn, rm, 2, 0, 0));
+                }
+            }
+            &Inst::FpuCmp64 { rn, rm } => {
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                if is_fpr(rn) && is_fpr(rm) {
+                    let opcode = 0xb319; // CDBR
+                    put(sink, &enc_rre(opcode, rn, rm));
+                } else {
+                    let opcode = 0xe7cb; // WFC
+                    put(sink, &enc_vrr_a(opcode, rn, rm, 3, 0, 0));
+                }
+            }
+
+            &Inst::VecRRR { op, rd, rn, rm } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                let (opcode, m4) = match op {
+                    VecBinaryOp::Add8x16 => (0xe7f3, 0),       // VAB
+                    VecBinaryOp::Add16x8 => (0xe7f3, 1),       // VAH
+                    VecBinaryOp::Add32x4 => (0xe7f3, 2),       // VAF
+                    VecBinaryOp::Add64x2 => (0xe7f3, 3),       // VAG
+                    VecBinaryOp::Add128 => (0xe7f3, 4),        // VAQ
+                    VecBinaryOp::Sub8x16 => (0xe7f7, 0),       // VSB
+                    VecBinaryOp::Sub16x8 => (0xe7f7, 1),       // VSH
+                    VecBinaryOp::Sub32x4 => (0xe7f7, 2),       // VSF
+                    VecBinaryOp::Sub64x2 => (0xe7f7, 3),       // VSG
+                    VecBinaryOp::Sub128 => (0xe7f7, 4),        // VSQ
+                    VecBinaryOp::Mul8x16 => (0xe7a2, 0),       // VMLB
+                    VecBinaryOp::Mul16x8 => (0xe7a2, 1),       // VMLHW
+                    VecBinaryOp::Mul32x4 => (0xe7a2, 2),       // VMLF
+                    VecBinaryOp::UMulHi8x16 => (0xe7a1, 0),    // VMLHB
+                    VecBinaryOp::UMulHi16x8 => (0xe7a1, 1),    // VMLHH
+                    VecBinaryOp::UMulHi32x4 => (0xe7a1, 2),    // VMLHF
+                    VecBinaryOp::SMulHi8x16 => (0xe7a3, 0),    // VMHB
+                    VecBinaryOp::SMulHi16x8 => (0xe7a3, 1),    // VMHH
+                    VecBinaryOp::SMulHi32x4 => (0xe7a3, 2),    // VMHF
+                    VecBinaryOp::UMulEven8x16 => (0xe7a4, 0),  // VMLEB
+                    VecBinaryOp::UMulEven16x8 => (0xe7a4, 1),  // VMLEH
+                    VecBinaryOp::UMulEven32x4 => (0xe7a4, 2),  // VMLEF
+                    VecBinaryOp::SMulEven8x16 => (0xe7a6, 0),  // VMEB
+                    VecBinaryOp::SMulEven16x8 => (0xe7a6, 1),  // VMEH
+                    VecBinaryOp::SMulEven32x4 => (0xe7a6, 2),  // VMEF
+                    VecBinaryOp::UMulOdd8x16 => (0xe7a5, 0),   // VMLOB
+                    VecBinaryOp::UMulOdd16x8 => (0xe7a5, 1),   // VMLOH
+                    VecBinaryOp::UMulOdd32x4 => (0xe7a5, 2),   // VMLOF
+                    VecBinaryOp::SMulOdd8x16 => (0xe7a7, 0),   // VMOB
+                    VecBinaryOp::SMulOdd16x8 => (0xe7a7, 1),   // VMOH
+                    VecBinaryOp::SMulOdd32x4 => (0xe7a7, 2),   // VMOF
+                    VecBinaryOp::UMax8x16 => (0xe7fd, 0),      // VMXLB
+                    VecBinaryOp::UMax16x8 => (0xe7fd, 1),      // VMXLH
+                    VecBinaryOp::UMax32x4 => (0xe7fd, 2),      // VMXLF
+                    VecBinaryOp::UMax64x2 => (0xe7fd, 3),      // VMXLG
+                    VecBinaryOp::SMax8x16 => (0xe7ff, 0),      // VMXB
+                    VecBinaryOp::SMax16x8 => (0xe7ff, 1),      // VMXH
+                    VecBinaryOp::SMax32x4 => (0xe7ff, 2),      // VMXF
+                    VecBinaryOp::SMax64x2 => (0xe7ff, 3),      // VMXG
+                    VecBinaryOp::UMin8x16 => (0xe7fc, 0),      // VMNLB
+                    VecBinaryOp::UMin16x8 => (0xe7fc, 1),      // VMNLH
+                    VecBinaryOp::UMin32x4 => (0xe7fc, 2),      // VMNLF
+                    VecBinaryOp::UMin64x2 => (0xe7fc, 3),      // VMNLG
+                    VecBinaryOp::SMin8x16 => (0xe7fe, 0),      // VMNB
+                    VecBinaryOp::SMin16x8 => (0xe7fe, 1),      // VMNH
+                    VecBinaryOp::SMin32x4 => (0xe7fe, 2),      // VMNF
+                    VecBinaryOp::SMin64x2 => (0xe7fe, 3),      // VMNG
+                    VecBinaryOp::UAvg8x16 => (0xe7f0, 0),      // VAVGLB
+                    VecBinaryOp::UAvg16x8 => (0xe7f0, 1),      // VAVGLH
+                    VecBinaryOp::UAvg32x4 => (0xe7f0, 2),      // VAVGLF
+                    VecBinaryOp::UAvg64x2 => (0xe7f0, 3),      // VAVGLG
+                    VecBinaryOp::SAvg8x16 => (0xe7f2, 0),      // VAVGB
+                    VecBinaryOp::SAvg16x8 => (0xe7f2, 1),      // VAVGH
+                    VecBinaryOp::SAvg32x4 => (0xe7f2, 2),      // VAVGF
+                    VecBinaryOp::SAvg64x2 => (0xe7f2, 3),      // VAVGG
+                    VecBinaryOp::And128 => (0xe768, 0),        // VN
+                    VecBinaryOp::Orr128 => (0xe76a, 0),        // VO
+                    VecBinaryOp::Xor128 => (0xe76d, 0),        // VX
+                    VecBinaryOp::NotAnd128 => (0xe76e, 0),     // VNN
+                    VecBinaryOp::NotOrr128 => (0xe76b, 0),     // VNO
+                    VecBinaryOp::NotXor128 => (0xe76c, 0),     // VNX
+                    VecBinaryOp::AndNot128 => (0xe769, 0),     // VNC
+                    VecBinaryOp::OrrNot128 => (0xe76f, 0),     // VOC
+                    VecBinaryOp::BitPermute128 => (0xe785, 0), // VBPERM
+                    VecBinaryOp::LShLByByte128 => (0xe775, 0), // VSLB
+                    VecBinaryOp::LShRByByte128 => (0xe77d, 0), // VSRLB
+                    VecBinaryOp::AShRByByte128 => (0xe77f, 0), // VSRAB
+                    VecBinaryOp::LShLByBit128 => (0xe774, 0),  // VSL
+                    VecBinaryOp::LShRByBit128 => (0xe77c, 0),  // VSRL
+                    VecBinaryOp::AShRByBit128 => (0xe77e, 0),  // VSRA
+                    VecBinaryOp::Pack16x8 => (0xe794, 1),      // VPKH
+                    VecBinaryOp::Pack32x4 => (0xe794, 2),      // VPKF
+                    VecBinaryOp::Pack64x2 => (0xe794, 3),      // VPKG
+                    VecBinaryOp::PackUSat16x8 => (0xe795, 1),  // VPKLSH
+                    VecBinaryOp::PackUSat32x4 => (0xe795, 2),  // VPKLSF
+                    VecBinaryOp::PackUSat64x2 => (0xe795, 3),  // VPKLSG
+                    VecBinaryOp::PackSSat16x8 => (0xe797, 1),  // VPKSH
+                    VecBinaryOp::PackSSat32x4 => (0xe797, 2),  // VPKSF
+                    VecBinaryOp::PackSSat64x2 => (0xe797, 3),  // VPKSG
+                    VecBinaryOp::MergeLow8x16 => (0xe760, 0),  // VMRLB
+                    VecBinaryOp::MergeLow16x8 => (0xe760, 1),  // VMRLH
+                    VecBinaryOp::MergeLow32x4 => (0xe760, 2),  // VMRLF
+                    VecBinaryOp::MergeLow64x2 => (0xe760, 3),  // VMRLG
+                    VecBinaryOp::MergeHigh8x16 => (0xe761, 0), // VMRHB
+                    VecBinaryOp::MergeHigh16x8 => (0xe761, 1), // VMRHH
+                    VecBinaryOp::MergeHigh32x4 => (0xe761, 2), // VMRHF
+                    VecBinaryOp::MergeHigh64x2 => (0xe761, 3), // VMRHG
+                };
+
+                put(sink, &enc_vrr_c(opcode, rd.to_reg(), rn, rm, m4, 0, 0));
+            }
+            &Inst::VecRR { op, rd, rn } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                let (opcode, m3) = match op {
+                    VecUnaryOp::Abs8x16 => (0xe7df, 0),         // VLPB
+                    VecUnaryOp::Abs16x8 => (0xe7df, 1),         // VLPH
+                    VecUnaryOp::Abs32x4 => (0xe7df, 2),         // VLPF
+                    VecUnaryOp::Abs64x2 => (0xe7df, 3),         // VLPG
+                    VecUnaryOp::Neg8x16 => (0xe7de, 0),         // VLCB
+                    VecUnaryOp::Neg16x8 => (0xe7de, 1),         // VLCH
+                    VecUnaryOp::Neg32x4 => (0xe7de, 2),         // VLCF
+                    VecUnaryOp::Neg64x2 => (0xe7de, 3),         // VLCG
+                    VecUnaryOp::Popcnt8x16 => (0xe750, 0),      // VPOPCTB
+                    VecUnaryOp::Popcnt16x8 => (0xe750, 1),      // VPOPCTH
+                    VecUnaryOp::Popcnt32x4 => (0xe750, 2),      // VPOPCTF
+                    VecUnaryOp::Popcnt64x2 => (0xe750, 3),      // VPOPCTG
+                    VecUnaryOp::Clz8x16 => (0xe753, 0),         // VCLZB
+                    VecUnaryOp::Clz16x8 => (0xe753, 1),         // VCLZH
+                    VecUnaryOp::Clz32x4 => (0xe753, 2),         // VCLZF
+                    VecUnaryOp::Clz64x2 => (0xe753, 3),         // VCLZG
+                    VecUnaryOp::Ctz8x16 => (0xe752, 0),         // VCTZB
+                    VecUnaryOp::Ctz16x8 => (0xe752, 1),         // VCTZH
+                    VecUnaryOp::Ctz32x4 => (0xe752, 2),         // VCTZF
+                    VecUnaryOp::Ctz64x2 => (0xe752, 3),         // VCTZG
+                    VecUnaryOp::UnpackULow8x16 => (0xe7d4, 0),  // VUPLLB
+                    VecUnaryOp::UnpackULow16x8 => (0xe7d4, 1),  // VUPLLH
+                    VecUnaryOp::UnpackULow32x4 => (0xe7d4, 2),  // VUPLLF
+                    VecUnaryOp::UnpackUHigh8x16 => (0xe7d5, 0), // VUPLHB
+                    VecUnaryOp::UnpackUHigh16x8 => (0xe7d5, 1), // VUPLHH
+                    VecUnaryOp::UnpackUHigh32x4 => (0xe7d5, 2), // VUPLHF
+                    VecUnaryOp::UnpackSLow8x16 => (0xe7d6, 0),  // VUPLB
+                    VecUnaryOp::UnpackSLow16x8 => (0xe7d6, 1),  // VUPLH
+                    VecUnaryOp::UnpackSLow32x4 => (0xe7d6, 2),  // VUPLF
+                    VecUnaryOp::UnpackSHigh8x16 => (0xe7d7, 0), // VUPHB
+                    VecUnaryOp::UnpackSHigh16x8 => (0xe7d7, 1), // VUPHH
+                    VecUnaryOp::UnpackSHigh32x4 => (0xe7d7, 2), // VUPHF
+                };
+
+                put(sink, &enc_vrr_a(opcode, rd.to_reg(), rn, m3, 0, 0));
+            }
+            &Inst::VecShiftRR {
+                shift_op,
+                rd,
+                rn,
+                shift_imm,
+                shift_reg,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let shift_reg = allocs.next(shift_reg);
+
+                let (opcode, m4) = match shift_op {
+                    VecShiftOp::RotL8x16 => (0xe733, 0), // VERLLB
+                    VecShiftOp::RotL16x8 => (0xe733, 1), // VERLLH
+                    VecShiftOp::RotL32x4 => (0xe733, 2), // VERLLF
+                    VecShiftOp::RotL64x2 => (0xe733, 3), // VERLLG
+                    VecShiftOp::LShL8x16 => (0xe730, 0), // VESLB
+                    VecShiftOp::LShL16x8 => (0xe730, 1), // VESLH
+                    VecShiftOp::LShL32x4 => (0xe730, 2), // VESLF
+                    VecShiftOp::LShL64x2 => (0xe730, 3), // VESLG
+                    VecShiftOp::LShR8x16 => (0xe738, 0), // VESRLB
+                    VecShiftOp::LShR16x8 => (0xe738, 1), // VESRLH
+                    VecShiftOp::LShR32x4 => (0xe738, 2), // VESRLF
+                    VecShiftOp::LShR64x2 => (0xe738, 3), // VESRLG
+                    VecShiftOp::AShR8x16 => (0xe73a, 0), // VESRAB
+                    VecShiftOp::AShR16x8 => (0xe73a, 1), // VESRAH
+                    VecShiftOp::AShR32x4 => (0xe73a, 2), // VESRAF
+                    VecShiftOp::AShR64x2 => (0xe73a, 3), // VESRAG
+                };
+                put(
+                    sink,
+                    &enc_vrs_a(opcode, rd.to_reg(), shift_reg, shift_imm.into(), rn, m4),
+                );
+            }
+            &Inst::VecSelect { rd, rn, rm, ra } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+                let ra = allocs.next(ra);
+
+                let opcode = 0xe78d; // VSEL
+                put(sink, &enc_vrr_e(opcode, rd.to_reg(), rn, rm, ra, 0, 0));
+            }
+            &Inst::VecPermute { rd, rn, rm, ra } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+                let ra = allocs.next(ra);
+
+                let opcode = 0xe78c; // VPERM
+                put(sink, &enc_vrr_e(opcode, rd.to_reg(), rn, rm, ra, 0, 0));
+            }
+            &Inst::VecPermuteDWImm {
+                rd,
+                rn,
+                rm,
+                idx1,
+                idx2,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+                let m4 = (idx1 & 1) * 4 + (idx2 & 1);
+
+                let opcode = 0xe784; // VPDI
+                put(sink, &enc_vrr_c(opcode, rd.to_reg(), rn, rm, m4, 0, 0));
+            }
+            &Inst::VecIntCmp { op, rd, rn, rm } | &Inst::VecIntCmpS { op, rd, rn, rm } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                let (opcode, m4) = match op {
+                    VecIntCmpOp::CmpEq8x16 => (0xe7f8, 0),  // VCEQB
+                    VecIntCmpOp::CmpEq16x8 => (0xe7f8, 1),  // VCEQH
+                    VecIntCmpOp::CmpEq32x4 => (0xe7f8, 2),  // VCEQF
+                    VecIntCmpOp::CmpEq64x2 => (0xe7f8, 3),  // VCEQG
+                    VecIntCmpOp::SCmpHi8x16 => (0xe7fb, 0), // VCHB
+                    VecIntCmpOp::SCmpHi16x8 => (0xe7fb, 1), // VCHH
+                    VecIntCmpOp::SCmpHi32x4 => (0xe7fb, 2), // VCHG
+                    VecIntCmpOp::SCmpHi64x2 => (0xe7fb, 3), // VCHG
+                    VecIntCmpOp::UCmpHi8x16 => (0xe7f9, 0), // VCHLB
+                    VecIntCmpOp::UCmpHi16x8 => (0xe7f9, 1), // VCHLH
+                    VecIntCmpOp::UCmpHi32x4 => (0xe7f9, 2), // VCHLG
+                    VecIntCmpOp::UCmpHi64x2 => (0xe7f9, 3), // VCHLG
+                };
+                let m5 = match self {
+                    &Inst::VecIntCmp { .. } => 0,
+                    &Inst::VecIntCmpS { .. } => 1,
+                    _ => unreachable!(),
+                };
+
+                put(sink, &enc_vrr_b(opcode, rd.to_reg(), rn, rm, m4, m5));
+            }
+            &Inst::VecFloatCmp { op, rd, rn, rm } | &Inst::VecFloatCmpS { op, rd, rn, rm } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                let (opcode, m4) = match op {
+                    VecFloatCmpOp::CmpEq32x4 => (0xe7e8, 2),   // VFCESB
+                    VecFloatCmpOp::CmpEq64x2 => (0xe7e8, 3),   // VFCEDB
+                    VecFloatCmpOp::CmpHi32x4 => (0xe7eb, 2),   // VFCHSB
+                    VecFloatCmpOp::CmpHi64x2 => (0xe7eb, 3),   // VFCHDB
+                    VecFloatCmpOp::CmpHiEq32x4 => (0xe7ea, 2), // VFCHESB
+                    VecFloatCmpOp::CmpHiEq64x2 => (0xe7ea, 3), // VFCHEDB
+                };
+                let m6 = match self {
+                    &Inst::VecFloatCmp { .. } => 0,
+                    &Inst::VecFloatCmpS { .. } => 1,
+                    _ => unreachable!(),
+                };
+
+                put(sink, &enc_vrr_c(opcode, rd.to_reg(), rn, rm, m4, 0, m6));
+            }
+            &Inst::VecInt128SCmpHi { tmp, rn, rm } | &Inst::VecInt128UCmpHi { tmp, rn, rm } => {
+                // Synthetic instruction to compare 128-bit values.
+                // Sets CC 1 if rn > rm, sets a different CC otherwise.
+                let tmp = allocs.next_writable(tmp);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                // Use VECTOR ELEMENT COMPARE to compare the high parts.
+                // Swap the inputs to get:
+                //    CC 1 if high(rn) > high(rm)
+                //    CC 2 if high(rn) < high(rm)
+                //    CC 0 if high(rn) == high(rm)
+                let (opcode, m3) = match self {
+                    &Inst::VecInt128SCmpHi { .. } => (0xe7db, 3), // VECG
+                    &Inst::VecInt128UCmpHi { .. } => (0xe7d9, 3), // VECLG
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_vrr_a(opcode, rm, rn, m3, 0, 0));
+
+                // If CC != 0, we'd done, so jump over the next instruction.
+                let opcode = 0xa74; // BCR
+                put(sink, &enc_ri_c(opcode, 7, 4 + 6));
+
+                // Otherwise, use VECTOR COMPARE HIGH LOGICAL.
+                // Since we already know the high parts are equal, the CC
+                // result will only depend on the low parts:
+                //     CC 1 if low(rn) > low(rm)
+                //     CC 3 if low(rn) <= low(rm)
+                let inst = Inst::VecIntCmpS {
+                    op: VecIntCmpOp::UCmpHi64x2,
+                    // N.B.: This is the first write to tmp, and it happens
+                    // after all uses of rn and rm.  If this were to ever
+                    // change, tmp would have to become an early-def.
+                    rd: tmp,
+                    rn,
+                    rm,
+                };
+                inst.emit(&[], sink, emit_info, state);
+            }
+
+            &Inst::VecLoad { rd, ref mem }
+            | &Inst::VecLoadRev { rd, ref mem }
+            | &Inst::VecLoadByte16Rev { rd, ref mem }
+            | &Inst::VecLoadByte32Rev { rd, ref mem }
+            | &Inst::VecLoadByte64Rev { rd, ref mem }
+            | &Inst::VecLoadElt16Rev { rd, ref mem }
+            | &Inst::VecLoadElt32Rev { rd, ref mem }
+            | &Inst::VecLoadElt64Rev { rd, ref mem } => {
+                let rd = allocs.next_writable(rd);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode, m3) = match self {
+                    &Inst::VecLoad { .. } => (0xe706, 0),          // VL
+                    &Inst::VecLoadRev { .. } => (0xe606, 4),       // VLBRQ
+                    &Inst::VecLoadByte16Rev { .. } => (0xe606, 1), // VLBRH
+                    &Inst::VecLoadByte32Rev { .. } => (0xe606, 2), // VLBRF
+                    &Inst::VecLoadByte64Rev { .. } => (0xe606, 3), // VLBRG
+                    &Inst::VecLoadElt16Rev { .. } => (0xe607, 1),  // VLERH
+                    &Inst::VecLoadElt32Rev { .. } => (0xe607, 2),  // VLERF
+                    &Inst::VecLoadElt64Rev { .. } => (0xe607, 3),  // VLERG
+                    _ => unreachable!(),
+                };
+                mem_vrx_emit(rd.to_reg(), &mem, opcode, m3, true, sink, emit_info, state);
+            }
+            &Inst::VecStore { rd, ref mem }
+            | &Inst::VecStoreRev { rd, ref mem }
+            | &Inst::VecStoreByte16Rev { rd, ref mem }
+            | &Inst::VecStoreByte32Rev { rd, ref mem }
+            | &Inst::VecStoreByte64Rev { rd, ref mem }
+            | &Inst::VecStoreElt16Rev { rd, ref mem }
+            | &Inst::VecStoreElt32Rev { rd, ref mem }
+            | &Inst::VecStoreElt64Rev { rd, ref mem } => {
+                let rd = allocs.next(rd);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode, m3) = match self {
+                    &Inst::VecStore { .. } => (0xe70e, 0),          // VST
+                    &Inst::VecStoreRev { .. } => (0xe60e, 4),       // VSTBRQ
+                    &Inst::VecStoreByte16Rev { .. } => (0xe60e, 1), // VSTBRH
+                    &Inst::VecStoreByte32Rev { .. } => (0xe60e, 2), // VSTBRF
+                    &Inst::VecStoreByte64Rev { .. } => (0xe60e, 3), // VSTBRG
+                    &Inst::VecStoreElt16Rev { .. } => (0xe60f, 1),  // VSTERH
+                    &Inst::VecStoreElt32Rev { .. } => (0xe60f, 2),  // VSTERF
+                    &Inst::VecStoreElt64Rev { .. } => (0xe60f, 3),  // VSTERG
+                    _ => unreachable!(),
+                };
+                mem_vrx_emit(rd, &mem, opcode, m3, true, sink, emit_info, state);
+            }
+            &Inst::VecLoadReplicate { size, rd, ref mem }
+            | &Inst::VecLoadReplicateRev { size, rd, ref mem } => {
+                let rd = allocs.next_writable(rd);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode, m3) = match (self, size) {
+                    (&Inst::VecLoadReplicate { .. }, 8) => (0xe705, 0), // VLREPB
+                    (&Inst::VecLoadReplicate { .. }, 16) => (0xe705, 1), // VLREPH
+                    (&Inst::VecLoadReplicate { .. }, 32) => (0xe705, 2), // VLREPF
+                    (&Inst::VecLoadReplicate { .. }, 64) => (0xe705, 3), // VLREPG
+                    (&Inst::VecLoadReplicateRev { .. }, 16) => (0xe605, 1), // VLREPBRH
+                    (&Inst::VecLoadReplicateRev { .. }, 32) => (0xe605, 2), // VLREPBRF
+                    (&Inst::VecLoadReplicateRev { .. }, 64) => (0xe605, 3), // VLREPBRG
+                    _ => unreachable!(),
+                };
+                mem_vrx_emit(rd.to_reg(), &mem, opcode, m3, true, sink, emit_info, state);
+            }
+
+            &Inst::VecMov { rd, rn } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                let opcode = 0xe756; // VLR
+                put(sink, &enc_vrr_a(opcode, rd.to_reg(), rn, 0, 0, 0));
+            }
+            &Inst::VecCMov { rd, cond, ri, rm } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let rm = allocs.next(rm);
+
+                let opcode = 0xa74; // BCR
+                put(sink, &enc_ri_c(opcode, cond.invert().bits(), 4 + 6));
+                let opcode = 0xe756; // VLR
+                put(sink, &enc_vrr_a(opcode, rd.to_reg(), rm, 0, 0, 0));
+            }
+            &Inst::MovToVec128 { rd, rn, rm } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let rm = allocs.next(rm);
+
+                let opcode = 0xe762; // VLVGP
+                put(sink, &enc_vrr_f(opcode, rd.to_reg(), rn, rm));
+            }
+            &Inst::VecLoadConst { rd, const_data } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = 0xa75; // BRAS
+                let reg = writable_spilltmp_reg().to_reg();
+                put(sink, &enc_ri_b(opcode, reg, 20));
+                for i in const_data.to_be_bytes().iter() {
+                    sink.put1(*i);
+                }
+                let inst = Inst::VecLoad {
+                    rd,
+                    mem: MemArg::reg(reg, MemFlags::trusted()),
+                };
+                inst.emit(&[], sink, emit_info, state);
+            }
+            &Inst::VecLoadConstReplicate {
+                size,
+                rd,
+                const_data,
+            } => {
+                let rd = allocs.next_writable(rd);
+
+                let opcode = 0xa75; // BRAS
+                let reg = writable_spilltmp_reg().to_reg();
+                put(sink, &enc_ri_b(opcode, reg, (4 + size / 8) as i32));
+                for i in 0..size / 8 {
+                    sink.put1((const_data >> (size - 8 - 8 * i)) as u8);
+                }
+                let inst = Inst::VecLoadReplicate {
+                    size,
+                    rd,
+                    mem: MemArg::reg(reg, MemFlags::trusted()),
+                };
+                inst.emit(&[], sink, emit_info, state);
+            }
+            &Inst::VecImmByteMask { rd, mask } => {
+                let rd = allocs.next_writable(rd);
+                let opcode = 0xe744; // VGBM
+                put(sink, &enc_vri_a(opcode, rd.to_reg(), mask, 0));
+            }
+            &Inst::VecImmBitMask {
+                size,
+                rd,
+                start_bit,
+                end_bit,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let (opcode, m4) = match size {
+                    8 => (0xe746, 0),  // VGMB
+                    16 => (0xe746, 1), // VGMH
+                    32 => (0xe746, 2), // VGMF
+                    64 => (0xe746, 3), // VGMG
+                    _ => unreachable!(),
+                };
+                put(
+                    sink,
+                    &enc_vri_b(opcode, rd.to_reg(), start_bit, end_bit, m4),
+                );
+            }
+            &Inst::VecImmReplicate { size, rd, imm } => {
+                let rd = allocs.next_writable(rd);
+                let (opcode, m3) = match size {
+                    8 => (0xe745, 0),  // VREPIB
+                    16 => (0xe745, 1), // VREPIH
+                    32 => (0xe745, 2), // VREPIF
+                    64 => (0xe745, 3), // VREPIG
+                    _ => unreachable!(),
+                };
+                put(sink, &enc_vri_a(opcode, rd.to_reg(), imm as u16, m3));
+            }
+            &Inst::VecLoadLane {
+                size,
+                rd,
+                ri,
+                ref mem,
+                lane_imm,
+            }
+            | &Inst::VecLoadLaneRev {
+                size,
+                rd,
+                ri,
+                ref mem,
+                lane_imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let mem = mem.with_allocs(allocs);
+
+                let opcode_vrx = match (self, size) {
+                    (&Inst::VecLoadLane { .. }, 8) => 0xe700,     // VLEB
+                    (&Inst::VecLoadLane { .. }, 16) => 0xe701,    // VLEH
+                    (&Inst::VecLoadLane { .. }, 32) => 0xe703,    // VLEF
+                    (&Inst::VecLoadLane { .. }, 64) => 0xe702,    // VLEG
+                    (&Inst::VecLoadLaneRev { .. }, 16) => 0xe601, // VLEBRH
+                    (&Inst::VecLoadLaneRev { .. }, 32) => 0xe603, // VLEBRF
+                    (&Inst::VecLoadLaneRev { .. }, 64) => 0xe602, // VLEBRG
+                    _ => unreachable!(),
+                };
+
+                let rd = rd.to_reg();
+                mem_vrx_emit(
+                    rd,
+                    &mem,
+                    opcode_vrx,
+                    lane_imm.into(),
+                    true,
+                    sink,
+                    emit_info,
+                    state,
+                );
+            }
+            &Inst::VecLoadLaneUndef {
+                size,
+                rd,
+                ref mem,
+                lane_imm,
+            }
+            | &Inst::VecLoadLaneRevUndef {
+                size,
+                rd,
+                ref mem,
+                lane_imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode_vrx, opcode_rx, opcode_rxy) = match (self, size) {
+                    (&Inst::VecLoadLaneUndef { .. }, 8) => (0xe700, None, None), // VLEB
+                    (&Inst::VecLoadLaneUndef { .. }, 16) => (0xe701, None, None), // VLEH
+                    (&Inst::VecLoadLaneUndef { .. }, 32) => (0xe703, Some(0x78), Some(0xed64)), // VLEF, LE(Y)
+                    (&Inst::VecLoadLaneUndef { .. }, 64) => (0xe702, Some(0x68), Some(0xed65)), // VLEG, LD(Y)
+                    (&Inst::VecLoadLaneRevUndef { .. }, 16) => (0xe601, None, None), // VLEBRH
+                    (&Inst::VecLoadLaneRevUndef { .. }, 32) => (0xe603, None, None), // VLEBRF
+                    (&Inst::VecLoadLaneRevUndef { .. }, 64) => (0xe602, None, None), // VLEBRG
+                    _ => unreachable!(),
+                };
+
+                let rd = rd.to_reg();
+                if lane_imm == 0 && is_fpr(rd) && opcode_rx.is_some() {
+                    mem_emit(
+                        rd, &mem, opcode_rx, opcode_rxy, None, true, sink, emit_info, state,
+                    );
+                } else {
+                    mem_vrx_emit(
+                        rd,
+                        &mem,
+                        opcode_vrx,
+                        lane_imm.into(),
+                        true,
+                        sink,
+                        emit_info,
+                        state,
+                    );
+                }
+            }
+            &Inst::VecStoreLane {
+                size,
+                rd,
+                ref mem,
+                lane_imm,
+            }
+            | &Inst::VecStoreLaneRev {
+                size,
+                rd,
+                ref mem,
+                lane_imm,
+            } => {
+                let rd = allocs.next(rd);
+                let mem = mem.with_allocs(allocs);
+
+                let (opcode_vrx, opcode_rx, opcode_rxy) = match (self, size) {
+                    (&Inst::VecStoreLane { .. }, 8) => (0xe708, None, None), // VSTEB
+                    (&Inst::VecStoreLane { .. }, 16) => (0xe709, None, None), // VSTEH
+                    (&Inst::VecStoreLane { .. }, 32) => (0xe70b, Some(0x70), Some(0xed66)), // VSTEF, STE(Y)
+                    (&Inst::VecStoreLane { .. }, 64) => (0xe70a, Some(0x60), Some(0xed67)), // VSTEG, STD(Y)
+                    (&Inst::VecStoreLaneRev { .. }, 16) => (0xe609, None, None), // VSTEBRH
+                    (&Inst::VecStoreLaneRev { .. }, 32) => (0xe60b, None, None), // VSTEBRF
+                    (&Inst::VecStoreLaneRev { .. }, 64) => (0xe60a, None, None), // VSTEBRG
+                    _ => unreachable!(),
+                };
+
+                if lane_imm == 0 && is_fpr(rd) && opcode_rx.is_some() {
+                    mem_emit(
+                        rd, &mem, opcode_rx, opcode_rxy, None, true, sink, emit_info, state,
+                    );
+                } else {
+                    mem_vrx_emit(
+                        rd,
+                        &mem,
+                        opcode_vrx,
+                        lane_imm.into(),
+                        true,
+                        sink,
+                        emit_info,
+                        state,
+                    );
+                }
+            }
+            &Inst::VecInsertLane {
+                size,
+                rd,
+                ri,
+                rn,
+                lane_imm,
+                lane_reg,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+                let rn = allocs.next(rn);
+                let lane_reg = allocs.next(lane_reg);
+
+                let (opcode_vrs, m4) = match size {
+                    8 => (0xe722, 0),  // VLVGB
+                    16 => (0xe722, 1), // VLVGH
+                    32 => (0xe722, 2), // VLVGF
+                    64 => (0xe722, 3), // VLVGG
+                    _ => unreachable!(),
+                };
+                put(
+                    sink,
+                    &enc_vrs_b(opcode_vrs, rd.to_reg(), lane_reg, lane_imm.into(), rn, m4),
+                );
+            }
+            &Inst::VecInsertLaneUndef {
+                size,
+                rd,
+                rn,
+                lane_imm,
+                lane_reg,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let lane_reg = allocs.next(lane_reg);
+
+                let (opcode_vrs, m4, opcode_rre) = match size {
+                    8 => (0xe722, 0, None),          // VLVGB
+                    16 => (0xe722, 1, None),         // VLVGH
+                    32 => (0xe722, 2, None),         // VLVGF
+                    64 => (0xe722, 3, Some(0xb3c1)), // VLVGG, LDGR
+                    _ => unreachable!(),
+                };
+                if opcode_rre.is_some()
+                    && lane_imm == 0
+                    && lane_reg == zero_reg()
+                    && is_fpr(rd.to_reg())
+                {
+                    put(sink, &enc_rre(opcode_rre.unwrap(), rd.to_reg(), rn));
+                } else {
+                    put(
+                        sink,
+                        &enc_vrs_b(opcode_vrs, rd.to_reg(), lane_reg, lane_imm.into(), rn, m4),
+                    );
+                }
+            }
+            &Inst::VecExtractLane {
+                size,
+                rd,
+                rn,
+                lane_imm,
+                lane_reg,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+                let lane_reg = allocs.next(lane_reg);
+
+                let (opcode_vrs, m4, opcode_rre) = match size {
+                    8 => (0xe721, 0, None),          // VLGVB
+                    16 => (0xe721, 1, None),         // VLGVH
+                    32 => (0xe721, 2, None),         // VLGVF
+                    64 => (0xe721, 3, Some(0xb3cd)), // VLGVG, LGDR
+                    _ => unreachable!(),
+                };
+                if opcode_rre.is_some() && lane_imm == 0 && lane_reg == zero_reg() && is_fpr(rn) {
+                    put(sink, &enc_rre(opcode_rre.unwrap(), rd.to_reg(), rn));
+                } else {
+                    put(
+                        sink,
+                        &enc_vrs_c(opcode_vrs, rd.to_reg(), lane_reg, lane_imm.into(), rn, m4),
+                    );
+                }
+            }
+            &Inst::VecInsertLaneImm {
+                size,
+                rd,
+                ri,
+                imm,
+                lane_imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let ri = allocs.next(ri);
+                debug_assert_eq!(rd.to_reg(), ri);
+
+                let opcode = match size {
+                    8 => 0xe740,  // VLEIB
+                    16 => 0xe741, // LEIVH
+                    32 => 0xe743, // VLEIF
+                    64 => 0xe742, // VLEIG
+                    _ => unreachable!(),
+                };
+                put(
+                    sink,
+                    &enc_vri_a(opcode, rd.to_reg(), imm as u16, lane_imm.into()),
+                );
+            }
+            &Inst::VecReplicateLane {
+                size,
+                rd,
+                rn,
+                lane_imm,
+            } => {
+                let rd = allocs.next_writable(rd);
+                let rn = allocs.next(rn);
+
+                let (opcode, m4) = match size {
+                    8 => (0xe74d, 0),  // VREPB
+                    16 => (0xe74d, 1), // VREPH
+                    32 => (0xe74d, 2), // VREPF
+                    64 => (0xe74d, 3), // VREPG
+                    _ => unreachable!(),
+                };
+                put(
+                    sink,
+                    &enc_vri_c(opcode, rd.to_reg(), lane_imm.into(), rn, m4),
+                );
+            }
+
+            &Inst::Call { link, ref info } => {
+                debug_assert_eq!(link.to_reg(), gpr(14));
+
+                let opcode = 0xc05; // BRASL
+
+                // Add relocation for target function.  This has to be done *before*
+                // the S390xTlsGdCall relocation if any, to ensure linker relaxation
+                // works correctly.
+                sink.add_reloc_at_offset(2, Reloc::S390xPLTRel32Dbl, &info.dest, 2);
+
+                // Add relocation for TLS libcalls to enable linker optimizations.
+                match &info.tls_symbol {
+                    None => {}
+                    Some(SymbolReloc::TlsGd { name }) => {
+                        sink.add_reloc(Reloc::S390xTlsGdCall, name, 0)
+                    }
+                    _ => unreachable!(),
+                }
+
+                if let Some(s) = state.take_stack_map() {
+                    sink.add_stack_map(StackMapExtent::UpcomingBytes(6), s);
+                }
+                put(sink, &enc_ril_b(opcode, link.to_reg(), 0));
+                if info.opcode.is_call() {
+                    sink.add_call_site(info.opcode);
+                }
+            }
+            &Inst::CallInd { link, ref info } => {
+                debug_assert_eq!(link.to_reg(), gpr(14));
+                let rn = allocs.next(info.rn);
+
+                let opcode = 0x0d; // BASR
+                if let Some(s) = state.take_stack_map() {
+                    sink.add_stack_map(StackMapExtent::UpcomingBytes(2), s);
+                }
+                put(sink, &enc_rr(opcode, link.to_reg(), rn));
+                if info.opcode.is_call() {
+                    sink.add_call_site(info.opcode);
+                }
+            }
+            &Inst::Args { .. } => {}
+            &Inst::Rets { .. } => {}
+            &Inst::Ret { link } => {
+                debug_assert_eq!(link, gpr(14));
+                let opcode = 0x07; // BCR
+                put(sink, &enc_rr(opcode, gpr(15), link));
+            }
+            &Inst::Jump { dest } => {
+                let off = sink.cur_offset();
+                // Indicate that the jump uses a label, if so, so that a fixup can occur later.
+                sink.use_label_at_offset(off, dest, LabelUse::BranchRIL);
+                sink.add_uncond_branch(off, off + 6, dest);
+                // Emit the jump itself.
+                let opcode = 0xc04; // BCRL
+                put(sink, &enc_ril_c(opcode, 15, 0));
+            }
+            &Inst::IndirectBr { rn, .. } => {
+                let rn = allocs.next(rn);
+
+                let opcode = 0x07; // BCR
+                put(sink, &enc_rr(opcode, gpr(15), rn));
+            }
+            &Inst::CondBr {
+                taken,
+                not_taken,
+                cond,
+            } => {
+                let opcode = 0xc04; // BCRL
+
+                // Conditional part first.
+                let cond_off = sink.cur_offset();
+                sink.use_label_at_offset(cond_off, taken, LabelUse::BranchRIL);
+                let inverted = &enc_ril_c(opcode, cond.invert().bits(), 0);
+                sink.add_cond_branch(cond_off, cond_off + 6, taken, inverted);
+                put(sink, &enc_ril_c(opcode, cond.bits(), 0));
+
+                // Unconditional part next.
+                let uncond_off = sink.cur_offset();
+                sink.use_label_at_offset(uncond_off, not_taken, LabelUse::BranchRIL);
+                sink.add_uncond_branch(uncond_off, uncond_off + 6, not_taken);
+                put(sink, &enc_ril_c(opcode, 15, 0));
+            }
+            &Inst::OneWayCondBr { target, cond } => {
+                let opcode = 0xc04; // BCRL
+                sink.use_label_at_offset(sink.cur_offset(), target, LabelUse::BranchRIL);
+                put(sink, &enc_ril_c(opcode, cond.bits(), 0));
+            }
+            &Inst::Nop0 => {}
+            &Inst::Nop2 => {
+                put(sink, &enc_e(0x0707));
+            }
+            &Inst::Debugtrap => {
+                put(sink, &enc_e(0x0001));
+            }
+            &Inst::Trap { trap_code } => {
+                if let Some(s) = state.take_stack_map() {
+                    sink.add_stack_map(StackMapExtent::UpcomingBytes(2), s);
+                }
+                put_with_trap(sink, &enc_e(0x0000), trap_code);
+            }
+            &Inst::TrapIf { cond, trap_code } => {
+                if let Some(s) = state.take_stack_map() {
+                    sink.add_stack_map(StackMapExtent::UpcomingBytes(6), s);
+                }
+                // We implement a TrapIf as a conditional branch into the middle
+                // of the branch (BRCL) instruction itself - those middle two bytes
+                // are zero, which matches the trap instruction itself.
+                let opcode = 0xc04; // BCRL
+                let enc = &enc_ril_c(opcode, cond.bits(), 2);
+                debug_assert!(enc.len() == 6 && enc[2] == 0 && enc[3] == 0);
+                // The trap must be placed on the last byte of the embedded trap
+                // instruction, so we need to emit the encoding in two parts.
+                put_with_trap(sink, &enc[0..4], trap_code);
+                put(sink, &enc[4..6]);
+            }
+            &Inst::JTSequence { ridx, ref targets } => {
+                let ridx = allocs.next(ridx);
+
+                let table_label = sink.get_label();
+
+                // This sequence is *one* instruction in the vcode, and is expanded only here at
+                // emission time, because we cannot allow the regalloc to insert spills/reloads in
+                // the middle; we depend on hardcoded PC-rel addressing below.
+
+                // Set temp register to address of jump table.
+                let rtmp = writable_spilltmp_reg();
+                let inst = Inst::LoadAddr {
+                    rd: rtmp,
+                    mem: MemArg::Label {
+                        target: table_label,
+                    },
+                };
+                inst.emit(&[], sink, emit_info, state);
+
+                // Set temp to target address by adding the value of the jump table entry.
+                let inst = Inst::AluRX {
+                    alu_op: ALUOp::Add64Ext32,
+                    rd: rtmp,
+                    ri: rtmp.to_reg(),
+                    mem: MemArg::reg_plus_reg(rtmp.to_reg(), ridx, MemFlags::trusted()),
+                };
+                inst.emit(&[], sink, emit_info, state);
+
+                // Branch to computed address. (`targets` here is only used for successor queries
+                // and is not needed for emission.)
+                let inst = Inst::IndirectBr {
+                    rn: rtmp.to_reg(),
+                    targets: vec![],
+                };
+                inst.emit(&[], sink, emit_info, state);
+
+                // Emit jump table (table of 32-bit offsets).
+                sink.bind_label(table_label, &mut state.ctrl_plane);
+                let jt_off = sink.cur_offset();
+                for &target in targets.iter() {
+                    let word_off = sink.cur_offset();
+                    let off_into_table = word_off - jt_off;
+                    sink.use_label_at_offset(word_off, target, LabelUse::PCRel32);
+                    sink.put4(off_into_table.swap_bytes());
+                }
+
+                // Lowering produces an EmitIsland before using a JTSequence, so we can safely
+                // disable the worst-case-size check in this case.
+                start_off = sink.cur_offset();
+            }
+
+            &Inst::VirtualSPOffsetAdj { offset } => {
+                trace!(
+                    "virtual sp offset adjusted by {} -> {}",
+                    offset,
+                    state.virtual_sp_offset + offset
+                );
+                state.virtual_sp_offset += offset;
+            }
+
+            &Inst::Unwind { ref inst } => {
+                sink.add_unwind(inst.clone());
+            }
+
+            &Inst::DummyUse { .. } => {}
+        }
+
+        let end_off = sink.cur_offset();
+        debug_assert!((end_off - start_off) <= Inst::worst_case_size());
+
+        state.clear_post_insn();
+    }
+}