//==- SystemZInstrVector.td - SystemZ Vector instructions ------*- tblgen-*-==//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//===----------------------------------------------------------------------===//
// Move instructions
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
// Register move.
def VLR : UnaryVRRa<"vlr", 0xE756, null_frag, v128any, v128any>;
def VLR32 : UnaryAliasVRR<null_frag, v32sb, v32sb>;
def VLR64 : UnaryAliasVRR<null_frag, v64db, v64db>;
// Load GR from VR element.
def VLGV : BinaryVRScGeneric<"vlgv", 0xE721>;
def VLGVB : BinaryVRSc<"vlgvb", 0xE721, null_frag, v128b, 0>;
def VLGVH : BinaryVRSc<"vlgvh", 0xE721, null_frag, v128h, 1>;
def VLGVF : BinaryVRSc<"vlgvf", 0xE721, null_frag, v128f, 2>;
def VLGVG : BinaryVRSc<"vlgvg", 0xE721, z_vector_extract, v128g, 3>;
// Load VR element from GR.
def VLVG : TernaryVRSbGeneric<"vlvg", 0xE722>;
def VLVGB : TernaryVRSb<"vlvgb", 0xE722, z_vector_insert,
v128b, v128b, GR32, 0>;
def VLVGH : TernaryVRSb<"vlvgh", 0xE722, z_vector_insert,
v128h, v128h, GR32, 1>;
def VLVGF : TernaryVRSb<"vlvgf", 0xE722, z_vector_insert,
v128f, v128f, GR32, 2>;
def VLVGG : TernaryVRSb<"vlvgg", 0xE722, z_vector_insert,
v128g, v128g, GR64, 3>;
// Load VR from GRs disjoint.
def VLVGP : BinaryVRRf<"vlvgp", 0xE762, z_join_dwords, v128g>;
def VLVGP32 : BinaryAliasVRRf<GR32>;
}
// Extractions always assign to the full GR64, even if the element would
// fit in the lower 32 bits. Sub-i64 extracts therefore need to take a
// subreg of the result.
class VectorExtractSubreg<ValueType type, Instruction insn>
: Pat<(i32 (z_vector_extract (type VR128:$vec), shift12only:$index)),
(EXTRACT_SUBREG (insn VR128:$vec, shift12only:$index), subreg_l32)>;
def : VectorExtractSubreg<v16i8, VLGVB>;
def : VectorExtractSubreg<v8i16, VLGVH>;
def : VectorExtractSubreg<v4i32, VLGVF>;
//===----------------------------------------------------------------------===//
// Immediate instructions
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
let isAsCheapAsAMove = 1, isMoveImm = 1, isReMaterializable = 1 in {
// Generate byte mask.
def VZERO : InherentVRIa<"vzero", 0xE744, 0>;
def VONE : InherentVRIa<"vone", 0xE744, 0xffff>;
def VGBM : UnaryVRIa<"vgbm", 0xE744, z_byte_mask, v128b, imm32zx16_timm>;
// Generate mask.
def VGM : BinaryVRIbGeneric<"vgm", 0xE746>;
def VGMB : BinaryVRIb<"vgmb", 0xE746, z_rotate_mask, v128b, 0>;
def VGMH : BinaryVRIb<"vgmh", 0xE746, z_rotate_mask, v128h, 1>;
def VGMF : BinaryVRIb<"vgmf", 0xE746, z_rotate_mask, v128f, 2>;
def VGMG : BinaryVRIb<"vgmg", 0xE746, z_rotate_mask, v128g, 3>;
// Replicate immediate.
def VREPI : UnaryVRIaGeneric<"vrepi", 0xE745, imm32sx16>;
def VREPIB : UnaryVRIa<"vrepib", 0xE745, z_replicate, v128b, imm32sx16_timm, 0>;
def VREPIH : UnaryVRIa<"vrepih", 0xE745, z_replicate, v128h, imm32sx16_timm, 1>;
def VREPIF : UnaryVRIa<"vrepif", 0xE745, z_replicate, v128f, imm32sx16_timm, 2>;
def VREPIG : UnaryVRIa<"vrepig", 0xE745, z_replicate, v128g, imm32sx16_timm, 3>;
}
// Load element immediate.
//
// We want these instructions to be used ahead of VLVG* where possible.
// However, VLVG* takes a variable BD-format index whereas VLEI takes
// a plain immediate index. This means that VLVG* has an extra "base"
// register operand and is 3 units more complex. Bumping the complexity
// of the VLEI* instructions by 4 means that they are strictly better
// than VLVG* in cases where both forms match.
let AddedComplexity = 4 in {
def VLEIB : TernaryVRIa<"vleib", 0xE740, z_vector_insert,
v128b, v128b, imm32sx16trunc, imm32zx4>;
def VLEIH : TernaryVRIa<"vleih", 0xE741, z_vector_insert,
v128h, v128h, imm32sx16trunc, imm32zx3>;
def VLEIF : TernaryVRIa<"vleif", 0xE743, z_vector_insert,
v128f, v128f, imm32sx16, imm32zx2>;
def VLEIG : TernaryVRIa<"vleig", 0xE742, z_vector_insert,
v128g, v128g, imm64sx16, imm32zx1>;
}
}
//===----------------------------------------------------------------------===//
// Loads
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
// Load.
defm VL : UnaryVRXAlign<"vl", 0xE706>;
// Load to block boundary. The number of loaded bytes is only known
// at run time. The instruction is really polymorphic, but v128b matches
// the return type of the associated intrinsic.
def VLBB : BinaryVRX<"vlbb", 0xE707, int_s390_vlbb, v128b, 0>;
// Load count to block boundary.
let Defs = [CC] in
def LCBB : InstRXE<0xE727, (outs GR32:$R1),
(ins bdxaddr12only:$XBD2, imm32zx4:$M3),
"lcbb\t$R1, $XBD2, $M3",
[(set GR32:$R1, (int_s390_lcbb bdxaddr12only:$XBD2,
imm32zx4_timm:$M3))]>;
// Load with length. The number of loaded bytes is only known at run time.
def VLL : BinaryVRSb<"vll", 0xE737, int_s390_vll, 0>;
// Load multiple.
defm VLM : LoadMultipleVRSaAlign<"vlm", 0xE736>;
// Load and replicate
def VLREP : UnaryVRXGeneric<"vlrep", 0xE705>;
def VLREPB : UnaryVRX<"vlrepb", 0xE705, z_replicate_loadi8, v128b, 1, 0>;
def VLREPH : UnaryVRX<"vlreph", 0xE705, z_replicate_loadi16, v128h, 2, 1>;
def VLREPF : UnaryVRX<"vlrepf", 0xE705, z_replicate_loadi32, v128f, 4, 2>;
def VLREPG : UnaryVRX<"vlrepg", 0xE705, z_replicate_loadi64, v128g, 8, 3>;
def : Pat<(v4f32 (z_replicate_loadf32 bdxaddr12only:$addr)),
(VLREPF bdxaddr12only:$addr)>;
def : Pat<(v2f64 (z_replicate_loadf64 bdxaddr12only:$addr)),
(VLREPG bdxaddr12only:$addr)>;
// Use VLREP to load subvectors. These patterns use "12pair" because
// LEY and LDY offer full 20-bit displacement fields. It's often better
// to use those instructions rather than force a 20-bit displacement
// into a GPR temporary.
let mayLoad = 1 in {
def VL32 : UnaryAliasVRX<load, v32sb, bdxaddr12pair>;
def VL64 : UnaryAliasVRX<load, v64db, bdxaddr12pair>;
}
// Load logical element and zero.
def VLLEZ : UnaryVRXGeneric<"vllez", 0xE704>;
def VLLEZB : UnaryVRX<"vllezb", 0xE704, z_vllezi8, v128b, 1, 0>;
def VLLEZH : UnaryVRX<"vllezh", 0xE704, z_vllezi16, v128h, 2, 1>;
def VLLEZF : UnaryVRX<"vllezf", 0xE704, z_vllezi32, v128f, 4, 2>;
def VLLEZG : UnaryVRX<"vllezg", 0xE704, z_vllezi64, v128g, 8, 3>;
def : Pat<(z_vllezf32 bdxaddr12only:$addr),
(VLLEZF bdxaddr12only:$addr)>;
def : Pat<(z_vllezf64 bdxaddr12only:$addr),
(VLLEZG bdxaddr12only:$addr)>;
let Predicates = [FeatureVectorEnhancements1] in {
def VLLEZLF : UnaryVRX<"vllezlf", 0xE704, z_vllezli32, v128f, 4, 6>;
def : Pat<(z_vllezlf32 bdxaddr12only:$addr),
(VLLEZLF bdxaddr12only:$addr)>;
}
// Load element.
def VLEB : TernaryVRX<"vleb", 0xE700, z_vlei8, v128b, v128b, 1, imm32zx4>;
def VLEH : TernaryVRX<"vleh", 0xE701, z_vlei16, v128h, v128h, 2, imm32zx3>;
def VLEF : TernaryVRX<"vlef", 0xE703, z_vlei32, v128f, v128f, 4, imm32zx2>;
def VLEG : TernaryVRX<"vleg", 0xE702, z_vlei64, v128g, v128g, 8, imm32zx1>;
def : Pat<(z_vlef32 (v4f32 VR128:$val), bdxaddr12only:$addr, imm32zx2:$index),
(VLEF VR128:$val, bdxaddr12only:$addr, imm32zx2:$index)>;
def : Pat<(z_vlef64 (v2f64 VR128:$val), bdxaddr12only:$addr, imm32zx1:$index),
(VLEG VR128:$val, bdxaddr12only:$addr, imm32zx1:$index)>;
// Gather element.
def VGEF : TernaryVRV<"vgef", 0xE713, 4, imm32zx2>;
def VGEG : TernaryVRV<"vgeg", 0xE712, 8, imm32zx1>;
}
let Predicates = [FeatureVectorPackedDecimal] in {
// Load rightmost with length. The number of loaded bytes is only known
// at run time. Note that while the instruction will accept immediate
// lengths larger that 15 at runtime, those will always result in a trap,
// so we never emit them here.
def VLRL : BinaryVSI<"vlrl", 0xE635, null_frag, 0>;
def VLRLR : BinaryVRSd<"vlrlr", 0xE637, int_s390_vlrl, 0>;
def : Pat<(int_s390_vlrl imm32zx4:$len, bdaddr12only:$addr),
(VLRL bdaddr12only:$addr, imm32zx4:$len)>;
}
// Use replicating loads if we're inserting a single element into an
// undefined vector. This avoids a false dependency on the previous
// register contents.
multiclass ReplicatePeephole<Instruction vlrep, ValueType vectype,
SDPatternOperator load, ValueType scalartype> {
def : Pat<(vectype (z_vector_insert
(undef), (scalartype (load bdxaddr12only:$addr)), 0)),
(vlrep bdxaddr12only:$addr)>;
def : Pat<(vectype (scalar_to_vector
(scalartype (load bdxaddr12only:$addr)))),
(vlrep bdxaddr12only:$addr)>;
}
defm : ReplicatePeephole<VLREPB, v16i8, anyextloadi8, i32>;
defm : ReplicatePeephole<VLREPH, v8i16, anyextloadi16, i32>;
defm : ReplicatePeephole<VLREPF, v4i32, load, i32>;
defm : ReplicatePeephole<VLREPG, v2i64, load, i64>;
defm : ReplicatePeephole<VLREPF, v4f32, load, f32>;
defm : ReplicatePeephole<VLREPG, v2f64, load, f64>;
//===----------------------------------------------------------------------===//
// Stores
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
// Store.
defm VST : StoreVRXAlign<"vst", 0xE70E>;
// Store with length. The number of stored bytes is only known at run time.
def VSTL : StoreLengthVRSb<"vstl", 0xE73F, int_s390_vstl, 0>;
// Store multiple.
defm VSTM : StoreMultipleVRSaAlign<"vstm", 0xE73E>;
// Store element.
def VSTEB : StoreBinaryVRX<"vsteb", 0xE708, z_vstei8, v128b, 1, imm32zx4>;
def VSTEH : StoreBinaryVRX<"vsteh", 0xE709, z_vstei16, v128h, 2, imm32zx3>;
def VSTEF : StoreBinaryVRX<"vstef", 0xE70B, z_vstei32, v128f, 4, imm32zx2>;
def VSTEG : StoreBinaryVRX<"vsteg", 0xE70A, z_vstei64, v128g, 8, imm32zx1>;
def : Pat<(z_vstef32 (v4f32 VR128:$val), bdxaddr12only:$addr,
imm32zx2:$index),
(VSTEF VR128:$val, bdxaddr12only:$addr, imm32zx2:$index)>;
def : Pat<(z_vstef64 (v2f64 VR128:$val), bdxaddr12only:$addr,
imm32zx1:$index),
(VSTEG VR128:$val, bdxaddr12only:$addr, imm32zx1:$index)>;
// Use VSTE to store subvectors. These patterns use "12pair" because
// STEY and STDY offer full 20-bit displacement fields. It's often better
// to use those instructions rather than force a 20-bit displacement
// into a GPR temporary.
let mayStore = 1 in {
def VST32 : StoreAliasVRX<store, v32sb, bdxaddr12pair>;
def VST64 : StoreAliasVRX<store, v64db, bdxaddr12pair>;
}
// Scatter element.
def VSCEF : StoreBinaryVRV<"vscef", 0xE71B, 4, imm32zx2>;
def VSCEG : StoreBinaryVRV<"vsceg", 0xE71A, 8, imm32zx1>;
}
let Predicates = [FeatureVectorPackedDecimal] in {
// Store rightmost with length. The number of stored bytes is only known
// at run time. Note that while the instruction will accept immediate
// lengths larger that 15 at runtime, those will always result in a trap,
// so we never emit them here.
def VSTRL : StoreLengthVSI<"vstrl", 0xE63D, null_frag, 0>;
def VSTRLR : StoreLengthVRSd<"vstrlr", 0xE63F, int_s390_vstrl, 0>;
def : Pat<(int_s390_vstrl VR128:$val, imm32zx4:$len, bdaddr12only:$addr),
(VSTRL VR128:$val, bdaddr12only:$addr, imm32zx4:$len)>;
}
//===----------------------------------------------------------------------===//
// Byte swaps
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVectorEnhancements2] in {
// Load byte-reversed elements.
def VLBR : UnaryVRXGeneric<"vlbr", 0xE606>;
def VLBRH : UnaryVRX<"vlbrh", 0xE606, z_loadbswap, v128h, 16, 1>;
def VLBRF : UnaryVRX<"vlbrf", 0xE606, z_loadbswap, v128f, 16, 2>;
def VLBRG : UnaryVRX<"vlbrg", 0xE606, z_loadbswap, v128g, 16, 3>;
def VLBRQ : UnaryVRX<"vlbrq", 0xE606, null_frag, v128q, 16, 4>;
// Load elements reversed.
def VLER : UnaryVRXGeneric<"vler", 0xE607>;
def VLERH : UnaryVRX<"vlerh", 0xE607, z_loadeswap, v128h, 16, 1>;
def VLERF : UnaryVRX<"vlerf", 0xE607, z_loadeswap, v128f, 16, 2>;
def VLERG : UnaryVRX<"vlerg", 0xE607, z_loadeswap, v128g, 16, 3>;
def : Pat<(v4f32 (z_loadeswap bdxaddr12only:$addr)),
(VLERF bdxaddr12only:$addr)>;
def : Pat<(v2f64 (z_loadeswap bdxaddr12only:$addr)),
(VLERG bdxaddr12only:$addr)>;
def : Pat<(v16i8 (z_loadeswap bdxaddr12only:$addr)),
(VLBRQ bdxaddr12only:$addr)>;
// Load byte-reversed element.
def VLEBRH : TernaryVRX<"vlebrh", 0xE601, z_vlebri16, v128h, v128h, 2, imm32zx3>;
def VLEBRF : TernaryVRX<"vlebrf", 0xE603, z_vlebri32, v128f, v128f, 4, imm32zx2>;
def VLEBRG : TernaryVRX<"vlebrg", 0xE602, z_vlebri64, v128g, v128g, 8, imm32zx1>;
// Load byte-reversed element and zero.
def VLLEBRZ : UnaryVRXGeneric<"vllebrz", 0xE604>;
def VLLEBRZH : UnaryVRX<"vllebrzh", 0xE604, z_vllebrzi16, v128h, 2, 1>;
def VLLEBRZF : UnaryVRX<"vllebrzf", 0xE604, z_vllebrzi32, v128f, 4, 2>;
def VLLEBRZG : UnaryVRX<"vllebrzg", 0xE604, z_vllebrzi64, v128g, 8, 3>;
def VLLEBRZE : UnaryVRX<"vllebrze", 0xE604, z_vllebrzli32, v128f, 4, 6>;
def : InstAlias<"lerv\t$V1, $XBD2",
(VLLEBRZE VR128:$V1, bdxaddr12only:$XBD2), 0>;
def : InstAlias<"ldrv\t$V1, $XBD2",
(VLLEBRZG VR128:$V1, bdxaddr12only:$XBD2), 0>;
// Load byte-reversed element and replicate.
def VLBRREP : UnaryVRXGeneric<"vlbrrep", 0xE605>;
def VLBRREPH : UnaryVRX<"vlbrreph", 0xE605, z_replicate_loadbswapi16, v128h, 2, 1>;
def VLBRREPF : UnaryVRX<"vlbrrepf", 0xE605, z_replicate_loadbswapi32, v128f, 4, 2>;
def VLBRREPG : UnaryVRX<"vlbrrepg", 0xE605, z_replicate_loadbswapi64, v128g, 8, 3>;
// Store byte-reversed elements.
def VSTBR : StoreVRXGeneric<"vstbr", 0xE60E>;
def VSTBRH : StoreVRX<"vstbrh", 0xE60E, z_storebswap, v128h, 16, 1>;
def VSTBRF : StoreVRX<"vstbrf", 0xE60E, z_storebswap, v128f, 16, 2>;
def VSTBRG : StoreVRX<"vstbrg", 0xE60E, z_storebswap, v128g, 16, 3>;
def VSTBRQ : StoreVRX<"vstbrq", 0xE60E, null_frag, v128q, 16, 4>;
// Store elements reversed.
def VSTER : StoreVRXGeneric<"vster", 0xE60F>;
def VSTERH : StoreVRX<"vsterh", 0xE60F, z_storeeswap, v128h, 16, 1>;
def VSTERF : StoreVRX<"vsterf", 0xE60F, z_storeeswap, v128f, 16, 2>;
def VSTERG : StoreVRX<"vsterg", 0xE60F, z_storeeswap, v128g, 16, 3>;
def : Pat<(z_storeeswap (v4f32 VR128:$val), bdxaddr12only:$addr),
(VSTERF VR128:$val, bdxaddr12only:$addr)>;
def : Pat<(z_storeeswap (v2f64 VR128:$val), bdxaddr12only:$addr),
(VSTERG VR128:$val, bdxaddr12only:$addr)>;
def : Pat<(z_storeeswap (v16i8 VR128:$val), bdxaddr12only:$addr),
(VSTBRQ VR128:$val, bdxaddr12only:$addr)>;
// Store byte-reversed element.
def VSTEBRH : StoreBinaryVRX<"vstebrh", 0xE609, z_vstebri16, v128h, 2, imm32zx3>;
def VSTEBRF : StoreBinaryVRX<"vstebrf", 0xE60B, z_vstebri32, v128f, 4, imm32zx2>;
def VSTEBRG : StoreBinaryVRX<"vstebrg", 0xE60A, z_vstebri64, v128g, 8, imm32zx1>;
def : InstAlias<"sterv\t$V1, $XBD2",
(VSTEBRF VR128:$V1, bdxaddr12only:$XBD2, 0), 0>;
def : InstAlias<"stdrv\t$V1, $XBD2",
(VSTEBRG VR128:$V1, bdxaddr12only:$XBD2, 0), 0>;
}
//===----------------------------------------------------------------------===//
// Selects and permutes
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
// Merge high.
def VMRH: BinaryVRRcGeneric<"vmrh", 0xE761>;
def VMRHB : BinaryVRRc<"vmrhb", 0xE761, z_merge_high, v128b, v128b, 0>;
def VMRHH : BinaryVRRc<"vmrhh", 0xE761, z_merge_high, v128h, v128h, 1>;
def VMRHF : BinaryVRRc<"vmrhf", 0xE761, z_merge_high, v128f, v128f, 2>;
def VMRHG : BinaryVRRc<"vmrhg", 0xE761, z_merge_high, v128g, v128g, 3>;
def : BinaryRRWithType<VMRHF, VR128, z_merge_high, v4f32>;
def : BinaryRRWithType<VMRHG, VR128, z_merge_high, v2f64>;
// Merge low.
def VMRL: BinaryVRRcGeneric<"vmrl", 0xE760>;
def VMRLB : BinaryVRRc<"vmrlb", 0xE760, z_merge_low, v128b, v128b, 0>;
def VMRLH : BinaryVRRc<"vmrlh", 0xE760, z_merge_low, v128h, v128h, 1>;
def VMRLF : BinaryVRRc<"vmrlf", 0xE760, z_merge_low, v128f, v128f, 2>;
def VMRLG : BinaryVRRc<"vmrlg", 0xE760, z_merge_low, v128g, v128g, 3>;
def : BinaryRRWithType<VMRLF, VR128, z_merge_low, v4f32>;
def : BinaryRRWithType<VMRLG, VR128, z_merge_low, v2f64>;
// Permute.
def VPERM : TernaryVRRe<"vperm", 0xE78C, z_permute, v128b, v128b>;
// Permute doubleword immediate.
def VPDI : TernaryVRRc<"vpdi", 0xE784, z_permute_dwords, v128g, v128g>;
// Bit Permute.
let Predicates = [FeatureVectorEnhancements1] in
def VBPERM : BinaryVRRc<"vbperm", 0xE785, int_s390_vbperm, v128g, v128b>;
// Replicate.
def VREP: BinaryVRIcGeneric<"vrep", 0xE74D>;
def VREPB : BinaryVRIc<"vrepb", 0xE74D, z_splat, v128b, v128b, 0>;
def VREPH : BinaryVRIc<"vreph", 0xE74D, z_splat, v128h, v128h, 1>;
def VREPF : BinaryVRIc<"vrepf", 0xE74D, z_splat, v128f, v128f, 2>;
def VREPG : BinaryVRIc<"vrepg", 0xE74D, z_splat, v128g, v128g, 3>;
def : Pat<(v4f32 (z_splat VR128:$vec, imm32zx16_timm:$index)),
(VREPF VR128:$vec, imm32zx16:$index)>;
def : Pat<(v2f64 (z_splat VR128:$vec, imm32zx16_timm:$index)),
(VREPG VR128:$vec, imm32zx16:$index)>;
// Select.
def VSEL : TernaryVRRe<"vsel", 0xE78D, null_frag, v128any, v128any>;
}
//===----------------------------------------------------------------------===//
// Widening and narrowing
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
// Pack
def VPK : BinaryVRRcGeneric<"vpk", 0xE794>;
def VPKH : BinaryVRRc<"vpkh", 0xE794, z_pack, v128b, v128h, 1>;
def VPKF : BinaryVRRc<"vpkf", 0xE794, z_pack, v128h, v128f, 2>;
def VPKG : BinaryVRRc<"vpkg", 0xE794, z_pack, v128f, v128g, 3>;
// Pack saturate.
def VPKS : BinaryVRRbSPairGeneric<"vpks", 0xE797>;
defm VPKSH : BinaryVRRbSPair<"vpksh", 0xE797, int_s390_vpksh, z_packs_cc,
v128b, v128h, 1>;
defm VPKSF : BinaryVRRbSPair<"vpksf", 0xE797, int_s390_vpksf, z_packs_cc,
v128h, v128f, 2>;
defm VPKSG : BinaryVRRbSPair<"vpksg", 0xE797, int_s390_vpksg, z_packs_cc,
v128f, v128g, 3>;
// Pack saturate logical.
def VPKLS : BinaryVRRbSPairGeneric<"vpkls", 0xE795>;
defm VPKLSH : BinaryVRRbSPair<"vpklsh", 0xE795, int_s390_vpklsh, z_packls_cc,
v128b, v128h, 1>;
defm VPKLSF : BinaryVRRbSPair<"vpklsf", 0xE795, int_s390_vpklsf, z_packls_cc,
v128h, v128f, 2>;
defm VPKLSG : BinaryVRRbSPair<"vpklsg", 0xE795, int_s390_vpklsg, z_packls_cc,
v128f, v128g, 3>;
// Sign-extend to doubleword.
def VSEG : UnaryVRRaGeneric<"vseg", 0xE75F>;
def VSEGB : UnaryVRRa<"vsegb", 0xE75F, z_vsei8, v128g, v128g, 0>;
def VSEGH : UnaryVRRa<"vsegh", 0xE75F, z_vsei16, v128g, v128g, 1>;
def VSEGF : UnaryVRRa<"vsegf", 0xE75F, z_vsei32, v128g, v128g, 2>;
def : Pat<(z_vsei8_by_parts (v16i8 VR128:$src)), (VSEGB VR128:$src)>;
def : Pat<(z_vsei16_by_parts (v8i16 VR128:$src)), (VSEGH VR128:$src)>;
def : Pat<(z_vsei32_by_parts (v4i32 VR128:$src)), (VSEGF VR128:$src)>;
// Unpack high.
def VUPH : UnaryVRRaGeneric<"vuph", 0xE7D7>;
def VUPHB : UnaryVRRa<"vuphb", 0xE7D7, z_unpack_high, v128h, v128b, 0>;
def VUPHH : UnaryVRRa<"vuphh", 0xE7D7, z_unpack_high, v128f, v128h, 1>;
def VUPHF : UnaryVRRa<"vuphf", 0xE7D7, z_unpack_high, v128g, v128f, 2>;
// Unpack logical high.
def VUPLH : UnaryVRRaGeneric<"vuplh", 0xE7D5>;
def VUPLHB : UnaryVRRa<"vuplhb", 0xE7D5, z_unpackl_high, v128h, v128b, 0>;
def VUPLHH : UnaryVRRa<"vuplhh", 0xE7D5, z_unpackl_high, v128f, v128h, 1>;
def VUPLHF : UnaryVRRa<"vuplhf", 0xE7D5, z_unpackl_high, v128g, v128f, 2>;
// Unpack low.
def VUPL : UnaryVRRaGeneric<"vupl", 0xE7D6>;
def VUPLB : UnaryVRRa<"vuplb", 0xE7D6, z_unpack_low, v128h, v128b, 0>;
def VUPLHW : UnaryVRRa<"vuplhw", 0xE7D6, z_unpack_low, v128f, v128h, 1>;
def VUPLF : UnaryVRRa<"vuplf", 0xE7D6, z_unpack_low, v128g, v128f, 2>;
// Unpack logical low.
def VUPLL : UnaryVRRaGeneric<"vupll", 0xE7D4>;
def VUPLLB : UnaryVRRa<"vupllb", 0xE7D4, z_unpackl_low, v128h, v128b, 0>;
def VUPLLH : UnaryVRRa<"vupllh", 0xE7D4, z_unpackl_low, v128f, v128h, 1>;
def VUPLLF : UnaryVRRa<"vupllf", 0xE7D4, z_unpackl_low, v128g, v128f, 2>;
}
//===----------------------------------------------------------------------===//
// Instantiating generic operations for specific types.
//===----------------------------------------------------------------------===//
multiclass GenericVectorOps<ValueType type, ValueType inttype> {
let Predicates = [FeatureVector] in {
def : Pat<(type (load bdxaddr12only:$addr)),
(VL bdxaddr12only:$addr)>;
def : Pat<(store (type VR128:$src), bdxaddr12only:$addr),
(VST VR128:$src, bdxaddr12only:$addr)>;
def : Pat<(type (vselect (inttype VR128:$x), VR128:$y, VR128:$z)),
(VSEL VR128:$y, VR128:$z, VR128:$x)>;
def : Pat<(type (vselect (inttype (z_vnot VR128:$x)), VR128:$y, VR128:$z)),
(VSEL VR128:$z, VR128:$y, VR128:$x)>;
}
}
defm : GenericVectorOps<v16i8, v16i8>;
defm : GenericVectorOps<v8i16, v8i16>;
defm : GenericVectorOps<v4i32, v4i32>;
defm : GenericVectorOps<v2i64, v2i64>;
defm : GenericVectorOps<v4f32, v4i32>;
defm : GenericVectorOps<v2f64, v2i64>;
//===----------------------------------------------------------------------===//
// Integer arithmetic
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
let isCommutable = 1 in {
// Add.
def VA : BinaryVRRcGeneric<"va", 0xE7F3>;
def VAB : BinaryVRRc<"vab", 0xE7F3, add, v128b, v128b, 0>;
def VAH : BinaryVRRc<"vah", 0xE7F3, add, v128h, v128h, 1>;
def VAF : BinaryVRRc<"vaf", 0xE7F3, add, v128f, v128f, 2>;
def VAG : BinaryVRRc<"vag", 0xE7F3, add, v128g, v128g, 3>;
def VAQ : BinaryVRRc<"vaq", 0xE7F3, int_s390_vaq, v128q, v128q, 4>;
}
let isCommutable = 1 in {
// Add compute carry.
def VACC : BinaryVRRcGeneric<"vacc", 0xE7F1>;
def VACCB : BinaryVRRc<"vaccb", 0xE7F1, int_s390_vaccb, v128b, v128b, 0>;
def VACCH : BinaryVRRc<"vacch", 0xE7F1, int_s390_vacch, v128h, v128h, 1>;
def VACCF : BinaryVRRc<"vaccf", 0xE7F1, int_s390_vaccf, v128f, v128f, 2>;
def VACCG : BinaryVRRc<"vaccg", 0xE7F1, int_s390_vaccg, v128g, v128g, 3>;
def VACCQ : BinaryVRRc<"vaccq", 0xE7F1, int_s390_vaccq, v128q, v128q, 4>;
// Add with carry.
def VAC : TernaryVRRdGeneric<"vac", 0xE7BB>;
def VACQ : TernaryVRRd<"vacq", 0xE7BB, int_s390_vacq, v128q, v128q, 4>;
// Add with carry compute carry.
def VACCC : TernaryVRRdGeneric<"vaccc", 0xE7B9>;
def VACCCQ : TernaryVRRd<"vacccq", 0xE7B9, int_s390_vacccq, v128q, v128q, 4>;
}
// And.
let isCommutable = 1 in
def VN : BinaryVRRc<"vn", 0xE768, null_frag, v128any, v128any>;
// And with complement.
def VNC : BinaryVRRc<"vnc", 0xE769, null_frag, v128any, v128any>;
let isCommutable = 1 in {
// Average.
def VAVG : BinaryVRRcGeneric<"vavg", 0xE7F2>;
def VAVGB : BinaryVRRc<"vavgb", 0xE7F2, int_s390_vavgb, v128b, v128b, 0>;
def VAVGH : BinaryVRRc<"vavgh", 0xE7F2, int_s390_vavgh, v128h, v128h, 1>;
def VAVGF : BinaryVRRc<"vavgf", 0xE7F2, int_s390_vavgf, v128f, v128f, 2>;
def VAVGG : BinaryVRRc<"vavgg", 0xE7F2, int_s390_vavgg, v128g, v128g, 3>;
// Average logical.
def VAVGL : BinaryVRRcGeneric<"vavgl", 0xE7F0>;
def VAVGLB : BinaryVRRc<"vavglb", 0xE7F0, int_s390_vavglb, v128b, v128b, 0>;
def VAVGLH : BinaryVRRc<"vavglh", 0xE7F0, int_s390_vavglh, v128h, v128h, 1>;
def VAVGLF : BinaryVRRc<"vavglf", 0xE7F0, int_s390_vavglf, v128f, v128f, 2>;
def VAVGLG : BinaryVRRc<"vavglg", 0xE7F0, int_s390_vavglg, v128g, v128g, 3>;
}
// Checksum.
def VCKSM : BinaryVRRc<"vcksm", 0xE766, int_s390_vcksm, v128f, v128f>;
// Count leading zeros.
def VCLZ : UnaryVRRaGeneric<"vclz", 0xE753>;
def VCLZB : UnaryVRRa<"vclzb", 0xE753, ctlz, v128b, v128b, 0>;
def VCLZH : UnaryVRRa<"vclzh", 0xE753, ctlz, v128h, v128h, 1>;
def VCLZF : UnaryVRRa<"vclzf", 0xE753, ctlz, v128f, v128f, 2>;
def VCLZG : UnaryVRRa<"vclzg", 0xE753, ctlz, v128g, v128g, 3>;
// Count trailing zeros.
def VCTZ : UnaryVRRaGeneric<"vctz", 0xE752>;
def VCTZB : UnaryVRRa<"vctzb", 0xE752, cttz, v128b, v128b, 0>;
def VCTZH : UnaryVRRa<"vctzh", 0xE752, cttz, v128h, v128h, 1>;
def VCTZF : UnaryVRRa<"vctzf", 0xE752, cttz, v128f, v128f, 2>;
def VCTZG : UnaryVRRa<"vctzg", 0xE752, cttz, v128g, v128g, 3>;
let isCommutable = 1 in {
// Not exclusive or.
let Predicates = [FeatureVectorEnhancements1] in
def VNX : BinaryVRRc<"vnx", 0xE76C, null_frag, v128any, v128any>;
// Exclusive or.
def VX : BinaryVRRc<"vx", 0xE76D, null_frag, v128any, v128any>;
}
// Galois field multiply sum.
def VGFM : BinaryVRRcGeneric<"vgfm", 0xE7B4>;
def VGFMB : BinaryVRRc<"vgfmb", 0xE7B4, int_s390_vgfmb, v128h, v128b, 0>;
def VGFMH : BinaryVRRc<"vgfmh", 0xE7B4, int_s390_vgfmh, v128f, v128h, 1>;
def VGFMF : BinaryVRRc<"vgfmf", 0xE7B4, int_s390_vgfmf, v128g, v128f, 2>;
def VGFMG : BinaryVRRc<"vgfmg", 0xE7B4, int_s390_vgfmg, v128q, v128g, 3>;
// Galois field multiply sum and accumulate.
def VGFMA : TernaryVRRdGeneric<"vgfma", 0xE7BC>;
def VGFMAB : TernaryVRRd<"vgfmab", 0xE7BC, int_s390_vgfmab, v128h, v128b, 0>;
def VGFMAH : TernaryVRRd<"vgfmah", 0xE7BC, int_s390_vgfmah, v128f, v128h, 1>;
def VGFMAF : TernaryVRRd<"vgfmaf", 0xE7BC, int_s390_vgfmaf, v128g, v128f, 2>;
def VGFMAG : TernaryVRRd<"vgfmag", 0xE7BC, int_s390_vgfmag, v128q, v128g, 3>;
// Load complement.
def VLC : UnaryVRRaGeneric<"vlc", 0xE7DE>;
def VLCB : UnaryVRRa<"vlcb", 0xE7DE, z_vneg, v128b, v128b, 0>;
def VLCH : UnaryVRRa<"vlch", 0xE7DE, z_vneg, v128h, v128h, 1>;
def VLCF : UnaryVRRa<"vlcf", 0xE7DE, z_vneg, v128f, v128f, 2>;
def VLCG : UnaryVRRa<"vlcg", 0xE7DE, z_vneg, v128g, v128g, 3>;
// Load positive.
def VLP : UnaryVRRaGeneric<"vlp", 0xE7DF>;
def VLPB : UnaryVRRa<"vlpb", 0xE7DF, z_viabs8, v128b, v128b, 0>;
def VLPH : UnaryVRRa<"vlph", 0xE7DF, z_viabs16, v128h, v128h, 1>;
def VLPF : UnaryVRRa<"vlpf", 0xE7DF, z_viabs32, v128f, v128f, 2>;
def VLPG : UnaryVRRa<"vlpg", 0xE7DF, z_viabs64, v128g, v128g, 3>;
let isCommutable = 1 in {
// Maximum.
def VMX : BinaryVRRcGeneric<"vmx", 0xE7FF>;
def VMXB : BinaryVRRc<"vmxb", 0xE7FF, null_frag, v128b, v128b, 0>;
def VMXH : BinaryVRRc<"vmxh", 0xE7FF, null_frag, v128h, v128h, 1>;
def VMXF : BinaryVRRc<"vmxf", 0xE7FF, null_frag, v128f, v128f, 2>;
def VMXG : BinaryVRRc<"vmxg", 0xE7FF, null_frag, v128g, v128g, 3>;
// Maximum logical.
def VMXL : BinaryVRRcGeneric<"vmxl", 0xE7FD>;
def VMXLB : BinaryVRRc<"vmxlb", 0xE7FD, null_frag, v128b, v128b, 0>;
def VMXLH : BinaryVRRc<"vmxlh", 0xE7FD, null_frag, v128h, v128h, 1>;
def VMXLF : BinaryVRRc<"vmxlf", 0xE7FD, null_frag, v128f, v128f, 2>;
def VMXLG : BinaryVRRc<"vmxlg", 0xE7FD, null_frag, v128g, v128g, 3>;
}
let isCommutable = 1 in {
// Minimum.
def VMN : BinaryVRRcGeneric<"vmn", 0xE7FE>;
def VMNB : BinaryVRRc<"vmnb", 0xE7FE, null_frag, v128b, v128b, 0>;
def VMNH : BinaryVRRc<"vmnh", 0xE7FE, null_frag, v128h, v128h, 1>;
def VMNF : BinaryVRRc<"vmnf", 0xE7FE, null_frag, v128f, v128f, 2>;
def VMNG : BinaryVRRc<"vmng", 0xE7FE, null_frag, v128g, v128g, 3>;
// Minimum logical.
def VMNL : BinaryVRRcGeneric<"vmnl", 0xE7FC>;
def VMNLB : BinaryVRRc<"vmnlb", 0xE7FC, null_frag, v128b, v128b, 0>;
def VMNLH : BinaryVRRc<"vmnlh", 0xE7FC, null_frag, v128h, v128h, 1>;
def VMNLF : BinaryVRRc<"vmnlf", 0xE7FC, null_frag, v128f, v128f, 2>;
def VMNLG : BinaryVRRc<"vmnlg", 0xE7FC, null_frag, v128g, v128g, 3>;
}
let isCommutable = 1 in {
// Multiply and add low.
def VMAL : TernaryVRRdGeneric<"vmal", 0xE7AA>;
def VMALB : TernaryVRRd<"vmalb", 0xE7AA, z_muladd, v128b, v128b, 0>;
def VMALHW : TernaryVRRd<"vmalhw", 0xE7AA, z_muladd, v128h, v128h, 1>;
def VMALF : TernaryVRRd<"vmalf", 0xE7AA, z_muladd, v128f, v128f, 2>;
// Multiply and add high.
def VMAH : TernaryVRRdGeneric<"vmah", 0xE7AB>;
def VMAHB : TernaryVRRd<"vmahb", 0xE7AB, int_s390_vmahb, v128b, v128b, 0>;
def VMAHH : TernaryVRRd<"vmahh", 0xE7AB, int_s390_vmahh, v128h, v128h, 1>;
def VMAHF : TernaryVRRd<"vmahf", 0xE7AB, int_s390_vmahf, v128f, v128f, 2>;
// Multiply and add logical high.
def VMALH : TernaryVRRdGeneric<"vmalh", 0xE7A9>;
def VMALHB : TernaryVRRd<"vmalhb", 0xE7A9, int_s390_vmalhb, v128b, v128b, 0>;
def VMALHH : TernaryVRRd<"vmalhh", 0xE7A9, int_s390_vmalhh, v128h, v128h, 1>;
def VMALHF : TernaryVRRd<"vmalhf", 0xE7A9, int_s390_vmalhf, v128f, v128f, 2>;
// Multiply and add even.
def VMAE : TernaryVRRdGeneric<"vmae", 0xE7AE>;
def VMAEB : TernaryVRRd<"vmaeb", 0xE7AE, int_s390_vmaeb, v128h, v128b, 0>;
def VMAEH : TernaryVRRd<"vmaeh", 0xE7AE, int_s390_vmaeh, v128f, v128h, 1>;
def VMAEF : TernaryVRRd<"vmaef", 0xE7AE, int_s390_vmaef, v128g, v128f, 2>;
// Multiply and add logical even.
def VMALE : TernaryVRRdGeneric<"vmale", 0xE7AC>;
def VMALEB : TernaryVRRd<"vmaleb", 0xE7AC, int_s390_vmaleb, v128h, v128b, 0>;
def VMALEH : TernaryVRRd<"vmaleh", 0xE7AC, int_s390_vmaleh, v128f, v128h, 1>;
def VMALEF : TernaryVRRd<"vmalef", 0xE7AC, int_s390_vmalef, v128g, v128f, 2>;
// Multiply and add odd.
def VMAO : TernaryVRRdGeneric<"vmao", 0xE7AF>;
def VMAOB : TernaryVRRd<"vmaob", 0xE7AF, int_s390_vmaob, v128h, v128b, 0>;
def VMAOH : TernaryVRRd<"vmaoh", 0xE7AF, int_s390_vmaoh, v128f, v128h, 1>;
def VMAOF : TernaryVRRd<"vmaof", 0xE7AF, int_s390_vmaof, v128g, v128f, 2>;
// Multiply and add logical odd.
def VMALO : TernaryVRRdGeneric<"vmalo", 0xE7AD>;
def VMALOB : TernaryVRRd<"vmalob", 0xE7AD, int_s390_vmalob, v128h, v128b, 0>;
def VMALOH : TernaryVRRd<"vmaloh", 0xE7AD, int_s390_vmaloh, v128f, v128h, 1>;
def VMALOF : TernaryVRRd<"vmalof", 0xE7AD, int_s390_vmalof, v128g, v128f, 2>;
}
let isCommutable = 1 in {
// Multiply high.
def VMH : BinaryVRRcGeneric<"vmh", 0xE7A3>;
def VMHB : BinaryVRRc<"vmhb", 0xE7A3, int_s390_vmhb, v128b, v128b, 0>;
def VMHH : BinaryVRRc<"vmhh", 0xE7A3, int_s390_vmhh, v128h, v128h, 1>;
def VMHF : BinaryVRRc<"vmhf", 0xE7A3, int_s390_vmhf, v128f, v128f, 2>;
// Multiply logical high.
def VMLH : BinaryVRRcGeneric<"vmlh", 0xE7A1>;
def VMLHB : BinaryVRRc<"vmlhb", 0xE7A1, int_s390_vmlhb, v128b, v128b, 0>;
def VMLHH : BinaryVRRc<"vmlhh", 0xE7A1, int_s390_vmlhh, v128h, v128h, 1>;
def VMLHF : BinaryVRRc<"vmlhf", 0xE7A1, int_s390_vmlhf, v128f, v128f, 2>;
// Multiply low.
def VML : BinaryVRRcGeneric<"vml", 0xE7A2>;
def VMLB : BinaryVRRc<"vmlb", 0xE7A2, mul, v128b, v128b, 0>;
def VMLHW : BinaryVRRc<"vmlhw", 0xE7A2, mul, v128h, v128h, 1>;
def VMLF : BinaryVRRc<"vmlf", 0xE7A2, mul, v128f, v128f, 2>;
// Multiply even.
def VME : BinaryVRRcGeneric<"vme", 0xE7A6>;
def VMEB : BinaryVRRc<"vmeb", 0xE7A6, int_s390_vmeb, v128h, v128b, 0>;
def VMEH : BinaryVRRc<"vmeh", 0xE7A6, int_s390_vmeh, v128f, v128h, 1>;
def VMEF : BinaryVRRc<"vmef", 0xE7A6, int_s390_vmef, v128g, v128f, 2>;
// Multiply logical even.
def VMLE : BinaryVRRcGeneric<"vmle", 0xE7A4>;
def VMLEB : BinaryVRRc<"vmleb", 0xE7A4, int_s390_vmleb, v128h, v128b, 0>;
def VMLEH : BinaryVRRc<"vmleh", 0xE7A4, int_s390_vmleh, v128f, v128h, 1>;
def VMLEF : BinaryVRRc<"vmlef", 0xE7A4, int_s390_vmlef, v128g, v128f, 2>;
// Multiply odd.
def VMO : BinaryVRRcGeneric<"vmo", 0xE7A7>;
def VMOB : BinaryVRRc<"vmob", 0xE7A7, int_s390_vmob, v128h, v128b, 0>;
def VMOH : BinaryVRRc<"vmoh", 0xE7A7, int_s390_vmoh, v128f, v128h, 1>;
def VMOF : BinaryVRRc<"vmof", 0xE7A7, int_s390_vmof, v128g, v128f, 2>;
// Multiply logical odd.
def VMLO : BinaryVRRcGeneric<"vmlo", 0xE7A5>;
def VMLOB : BinaryVRRc<"vmlob", 0xE7A5, int_s390_vmlob, v128h, v128b, 0>;
def VMLOH : BinaryVRRc<"vmloh", 0xE7A5, int_s390_vmloh, v128f, v128h, 1>;
def VMLOF : BinaryVRRc<"vmlof", 0xE7A5, int_s390_vmlof, v128g, v128f, 2>;
}
// Multiply sum logical.
let Predicates = [FeatureVectorEnhancements1], isCommutable = 1 in {
def VMSL : QuaternaryVRRdGeneric<"vmsl", 0xE7B8>;
def VMSLG : QuaternaryVRRd<"vmslg", 0xE7B8, int_s390_vmslg,
v128q, v128g, v128g, v128q, 3>;
}
// Nand.
let Predicates = [FeatureVectorEnhancements1], isCommutable = 1 in
def VNN : BinaryVRRc<"vnn", 0xE76E, null_frag, v128any, v128any>;
// Nor.
let isCommutable = 1 in
def VNO : BinaryVRRc<"vno", 0xE76B, null_frag, v128any, v128any>;
def : InstAlias<"vnot\t$V1, $V2", (VNO VR128:$V1, VR128:$V2, VR128:$V2), 0>;
// Or.
let isCommutable = 1 in
def VO : BinaryVRRc<"vo", 0xE76A, null_frag, v128any, v128any>;
// Or with complement.
let Predicates = [FeatureVectorEnhancements1] in
def VOC : BinaryVRRc<"voc", 0xE76F, null_frag, v128any, v128any>;
// Population count.
def VPOPCT : UnaryVRRaGeneric<"vpopct", 0xE750>;
def : Pat<(v16i8 (z_popcnt VR128:$x)), (VPOPCT VR128:$x, 0)>;
let Predicates = [FeatureVectorEnhancements1] in {
def VPOPCTB : UnaryVRRa<"vpopctb", 0xE750, ctpop, v128b, v128b, 0>;
def VPOPCTH : UnaryVRRa<"vpopcth", 0xE750, ctpop, v128h, v128h, 1>;
def VPOPCTF : UnaryVRRa<"vpopctf", 0xE750, ctpop, v128f, v128f, 2>;
def VPOPCTG : UnaryVRRa<"vpopctg", 0xE750, ctpop, v128g, v128g, 3>;
}
// Element rotate left logical (with vector shift amount).
def VERLLV : BinaryVRRcGeneric<"verllv", 0xE773>;
def VERLLVB : BinaryVRRc<"verllvb", 0xE773, int_s390_verllvb,
v128b, v128b, 0>;
def VERLLVH : BinaryVRRc<"verllvh", 0xE773, int_s390_verllvh,
v128h, v128h, 1>;
def VERLLVF : BinaryVRRc<"verllvf", 0xE773, int_s390_verllvf,
v128f, v128f, 2>;
def VERLLVG : BinaryVRRc<"verllvg", 0xE773, int_s390_verllvg,
v128g, v128g, 3>;
// Element rotate left logical (with scalar shift amount).
def VERLL : BinaryVRSaGeneric<"verll", 0xE733>;
def VERLLB : BinaryVRSa<"verllb", 0xE733, int_s390_verllb, v128b, v128b, 0>;
def VERLLH : BinaryVRSa<"verllh", 0xE733, int_s390_verllh, v128h, v128h, 1>;
def VERLLF : BinaryVRSa<"verllf", 0xE733, int_s390_verllf, v128f, v128f, 2>;
def VERLLG : BinaryVRSa<"verllg", 0xE733, int_s390_verllg, v128g, v128g, 3>;
// Element rotate and insert under mask.
def VERIM : QuaternaryVRIdGeneric<"verim", 0xE772>;
def VERIMB : QuaternaryVRId<"verimb", 0xE772, int_s390_verimb, v128b, v128b, 0>;
def VERIMH : QuaternaryVRId<"verimh", 0xE772, int_s390_verimh, v128h, v128h, 1>;
def VERIMF : QuaternaryVRId<"verimf", 0xE772, int_s390_verimf, v128f, v128f, 2>;
def VERIMG : QuaternaryVRId<"verimg", 0xE772, int_s390_verimg, v128g, v128g, 3>;
// Element shift left (with vector shift amount).
def VESLV : BinaryVRRcGeneric<"veslv", 0xE770>;
def VESLVB : BinaryVRRc<"veslvb", 0xE770, z_vshl, v128b, v128b, 0>;
def VESLVH : BinaryVRRc<"veslvh", 0xE770, z_vshl, v128h, v128h, 1>;
def VESLVF : BinaryVRRc<"veslvf", 0xE770, z_vshl, v128f, v128f, 2>;
def VESLVG : BinaryVRRc<"veslvg", 0xE770, z_vshl, v128g, v128g, 3>;
// Element shift left (with scalar shift amount).
def VESL : BinaryVRSaGeneric<"vesl", 0xE730>;
def VESLB : BinaryVRSa<"veslb", 0xE730, z_vshl_by_scalar, v128b, v128b, 0>;
def VESLH : BinaryVRSa<"veslh", 0xE730, z_vshl_by_scalar, v128h, v128h, 1>;
def VESLF : BinaryVRSa<"veslf", 0xE730, z_vshl_by_scalar, v128f, v128f, 2>;
def VESLG : BinaryVRSa<"veslg", 0xE730, z_vshl_by_scalar, v128g, v128g, 3>;
// Element shift right arithmetic (with vector shift amount).
def VESRAV : BinaryVRRcGeneric<"vesrav", 0xE77A>;
def VESRAVB : BinaryVRRc<"vesravb", 0xE77A, z_vsra, v128b, v128b, 0>;
def VESRAVH : BinaryVRRc<"vesravh", 0xE77A, z_vsra, v128h, v128h, 1>;
def VESRAVF : BinaryVRRc<"vesravf", 0xE77A, z_vsra, v128f, v128f, 2>;
def VESRAVG : BinaryVRRc<"vesravg", 0xE77A, z_vsra, v128g, v128g, 3>;
// Element shift right arithmetic (with scalar shift amount).
def VESRA : BinaryVRSaGeneric<"vesra", 0xE73A>;
def VESRAB : BinaryVRSa<"vesrab", 0xE73A, z_vsra_by_scalar, v128b, v128b, 0>;
def VESRAH : BinaryVRSa<"vesrah", 0xE73A, z_vsra_by_scalar, v128h, v128h, 1>;
def VESRAF : BinaryVRSa<"vesraf", 0xE73A, z_vsra_by_scalar, v128f, v128f, 2>;
def VESRAG : BinaryVRSa<"vesrag", 0xE73A, z_vsra_by_scalar, v128g, v128g, 3>;
// Element shift right logical (with vector shift amount).
def VESRLV : BinaryVRRcGeneric<"vesrlv", 0xE778>;
def VESRLVB : BinaryVRRc<"vesrlvb", 0xE778, z_vsrl, v128b, v128b, 0>;
def VESRLVH : BinaryVRRc<"vesrlvh", 0xE778, z_vsrl, v128h, v128h, 1>;
def VESRLVF : BinaryVRRc<"vesrlvf", 0xE778, z_vsrl, v128f, v128f, 2>;
def VESRLVG : BinaryVRRc<"vesrlvg", 0xE778, z_vsrl, v128g, v128g, 3>;
// Element shift right logical (with scalar shift amount).
def VESRL : BinaryVRSaGeneric<"vesrl", 0xE738>;
def VESRLB : BinaryVRSa<"vesrlb", 0xE738, z_vsrl_by_scalar, v128b, v128b, 0>;
def VESRLH : BinaryVRSa<"vesrlh", 0xE738, z_vsrl_by_scalar, v128h, v128h, 1>;
def VESRLF : BinaryVRSa<"vesrlf", 0xE738, z_vsrl_by_scalar, v128f, v128f, 2>;
def VESRLG : BinaryVRSa<"vesrlg", 0xE738, z_vsrl_by_scalar, v128g, v128g, 3>;
// Shift left.
def VSL : BinaryVRRc<"vsl", 0xE774, int_s390_vsl, v128b, v128b>;
// Shift left by byte.
def VSLB : BinaryVRRc<"vslb", 0xE775, int_s390_vslb, v128b, v128b>;
// Shift left double by byte.
def VSLDB : TernaryVRId<"vsldb", 0xE777, z_shl_double, v128b, v128b, 0>;
def : Pat<(int_s390_vsldb VR128:$x, VR128:$y, imm32zx8_timm:$z),
(VSLDB VR128:$x, VR128:$y, imm32zx8:$z)>;
// Shift left double by bit.
let Predicates = [FeatureVectorEnhancements2] in
def VSLD : TernaryVRId<"vsld", 0xE786, int_s390_vsld, v128b, v128b, 0>;
// Shift right arithmetic.
def VSRA : BinaryVRRc<"vsra", 0xE77E, int_s390_vsra, v128b, v128b>;
// Shift right arithmetic by byte.
def VSRAB : BinaryVRRc<"vsrab", 0xE77F, int_s390_vsrab, v128b, v128b>;
// Shift right logical.
def VSRL : BinaryVRRc<"vsrl", 0xE77C, int_s390_vsrl, v128b, v128b>;
// Shift right logical by byte.
def VSRLB : BinaryVRRc<"vsrlb", 0xE77D, int_s390_vsrlb, v128b, v128b>;
// Shift right double by bit.
let Predicates = [FeatureVectorEnhancements2] in
def VSRD : TernaryVRId<"vsrd", 0xE787, int_s390_vsrd, v128b, v128b, 0>;
// Subtract.
def VS : BinaryVRRcGeneric<"vs", 0xE7F7>;
def VSB : BinaryVRRc<"vsb", 0xE7F7, sub, v128b, v128b, 0>;
def VSH : BinaryVRRc<"vsh", 0xE7F7, sub, v128h, v128h, 1>;
def VSF : BinaryVRRc<"vsf", 0xE7F7, sub, v128f, v128f, 2>;
def VSG : BinaryVRRc<"vsg", 0xE7F7, sub, v128g, v128g, 3>;
def VSQ : BinaryVRRc<"vsq", 0xE7F7, int_s390_vsq, v128q, v128q, 4>;
// Subtract compute borrow indication.
def VSCBI : BinaryVRRcGeneric<"vscbi", 0xE7F5>;
def VSCBIB : BinaryVRRc<"vscbib", 0xE7F5, int_s390_vscbib, v128b, v128b, 0>;
def VSCBIH : BinaryVRRc<"vscbih", 0xE7F5, int_s390_vscbih, v128h, v128h, 1>;
def VSCBIF : BinaryVRRc<"vscbif", 0xE7F5, int_s390_vscbif, v128f, v128f, 2>;
def VSCBIG : BinaryVRRc<"vscbig", 0xE7F5, int_s390_vscbig, v128g, v128g, 3>;
def VSCBIQ : BinaryVRRc<"vscbiq", 0xE7F5, int_s390_vscbiq, v128q, v128q, 4>;
// Subtract with borrow indication.
def VSBI : TernaryVRRdGeneric<"vsbi", 0xE7BF>;
def VSBIQ : TernaryVRRd<"vsbiq", 0xE7BF, int_s390_vsbiq, v128q, v128q, 4>;
// Subtract with borrow compute borrow indication.
def VSBCBI : TernaryVRRdGeneric<"vsbcbi", 0xE7BD>;
def VSBCBIQ : TernaryVRRd<"vsbcbiq", 0xE7BD, int_s390_vsbcbiq,
v128q, v128q, 4>;
// Sum across doubleword.
def VSUMG : BinaryVRRcGeneric<"vsumg", 0xE765>;
def VSUMGH : BinaryVRRc<"vsumgh", 0xE765, z_vsum, v128g, v128h, 1>;
def VSUMGF : BinaryVRRc<"vsumgf", 0xE765, z_vsum, v128g, v128f, 2>;
// Sum across quadword.
def VSUMQ : BinaryVRRcGeneric<"vsumq", 0xE767>;
def VSUMQF : BinaryVRRc<"vsumqf", 0xE767, z_vsum, v128q, v128f, 2>;
def VSUMQG : BinaryVRRc<"vsumqg", 0xE767, z_vsum, v128q, v128g, 3>;
// Sum across word.
def VSUM : BinaryVRRcGeneric<"vsum", 0xE764>;
def VSUMB : BinaryVRRc<"vsumb", 0xE764, z_vsum, v128f, v128b, 0>;
def VSUMH : BinaryVRRc<"vsumh", 0xE764, z_vsum, v128f, v128h, 1>;
}
// Instantiate the bitwise ops for type TYPE.
multiclass BitwiseVectorOps<ValueType type> {
let Predicates = [FeatureVector] in {
def : Pat<(type (and VR128:$x, VR128:$y)), (VN VR128:$x, VR128:$y)>;
def : Pat<(type (and VR128:$x, (z_vnot VR128:$y))),
(VNC VR128:$x, VR128:$y)>;
def : Pat<(type (or VR128:$x, VR128:$y)), (VO VR128:$x, VR128:$y)>;
def : Pat<(type (xor VR128:$x, VR128:$y)), (VX VR128:$x, VR128:$y)>;
def : Pat<(type (or (and VR128:$x, VR128:$z),
(and VR128:$y, (z_vnot VR128:$z)))),
(VSEL VR128:$x, VR128:$y, VR128:$z)>;
def : Pat<(type (z_vnot (or VR128:$x, VR128:$y))),
(VNO VR128:$x, VR128:$y)>;
def : Pat<(type (z_vnot VR128:$x)), (VNO VR128:$x, VR128:$x)>;
}
let Predicates = [FeatureVectorEnhancements1] in {
def : Pat<(type (z_vnot (xor VR128:$x, VR128:$y))),
(VNX VR128:$x, VR128:$y)>;
def : Pat<(type (z_vnot (and VR128:$x, VR128:$y))),
(VNN VR128:$x, VR128:$y)>;
def : Pat<(type (or VR128:$x, (z_vnot VR128:$y))),
(VOC VR128:$x, VR128:$y)>;
}
}
defm : BitwiseVectorOps<v16i8>;
defm : BitwiseVectorOps<v8i16>;
defm : BitwiseVectorOps<v4i32>;
defm : BitwiseVectorOps<v2i64>;
// Instantiate additional patterns for absolute-related expressions on
// type TYPE. LC is the negate instruction for TYPE and LP is the absolute
// instruction.
multiclass IntegerAbsoluteVectorOps<ValueType type, Instruction lc,
Instruction lp, int shift> {
let Predicates = [FeatureVector] in {
def : Pat<(type (vselect (type (z_vicmph_zero VR128:$x)),
(z_vneg VR128:$x), VR128:$x)),
(lc (lp VR128:$x))>;
def : Pat<(type (vselect (type (z_vnot (z_vicmph_zero VR128:$x))),
VR128:$x, (z_vneg VR128:$x))),
(lc (lp VR128:$x))>;
def : Pat<(type (vselect (type (z_vicmpl_zero VR128:$x)),
VR128:$x, (z_vneg VR128:$x))),
(lc (lp VR128:$x))>;
def : Pat<(type (vselect (type (z_vnot (z_vicmpl_zero VR128:$x))),
(z_vneg VR128:$x), VR128:$x)),
(lc (lp VR128:$x))>;
def : Pat<(type (or (and (z_vsra_by_scalar VR128:$x, (i32 shift)),
(z_vneg VR128:$x)),
(and (z_vnot (z_vsra_by_scalar VR128:$x, (i32 shift))),
VR128:$x))),
(lp VR128:$x)>;
def : Pat<(type (or (and (z_vsra_by_scalar VR128:$x, (i32 shift)),
VR128:$x),
(and (z_vnot (z_vsra_by_scalar VR128:$x, (i32 shift))),
(z_vneg VR128:$x)))),
(lc (lp VR128:$x))>;
}
}
defm : IntegerAbsoluteVectorOps<v16i8, VLCB, VLPB, 7>;
defm : IntegerAbsoluteVectorOps<v8i16, VLCH, VLPH, 15>;
defm : IntegerAbsoluteVectorOps<v4i32, VLCF, VLPF, 31>;
defm : IntegerAbsoluteVectorOps<v2i64, VLCG, VLPG, 63>;
// Instantiate minimum- and maximum-related patterns for TYPE. CMPH is the
// signed or unsigned "set if greater than" comparison instruction and
// MIN and MAX are the associated minimum and maximum instructions.
multiclass IntegerMinMaxVectorOps<ValueType type, SDPatternOperator cmph,
Instruction min, Instruction max> {
let Predicates = [FeatureVector] in {
def : Pat<(type (vselect (cmph VR128:$x, VR128:$y), VR128:$x, VR128:$y)),
(max VR128:$x, VR128:$y)>;
def : Pat<(type (vselect (cmph VR128:$x, VR128:$y), VR128:$y, VR128:$x)),
(min VR128:$x, VR128:$y)>;
def : Pat<(type (vselect (z_vnot (cmph VR128:$x, VR128:$y)),
VR128:$x, VR128:$y)),
(min VR128:$x, VR128:$y)>;
def : Pat<(type (vselect (z_vnot (cmph VR128:$x, VR128:$y)),
VR128:$y, VR128:$x)),
(max VR128:$x, VR128:$y)>;
}
}
// Signed min/max.
defm : IntegerMinMaxVectorOps<v16i8, z_vicmph, VMNB, VMXB>;
defm : IntegerMinMaxVectorOps<v8i16, z_vicmph, VMNH, VMXH>;
defm : IntegerMinMaxVectorOps<v4i32, z_vicmph, VMNF, VMXF>;
defm : IntegerMinMaxVectorOps<v2i64, z_vicmph, VMNG, VMXG>;
// Unsigned min/max.
defm : IntegerMinMaxVectorOps<v16i8, z_vicmphl, VMNLB, VMXLB>;
defm : IntegerMinMaxVectorOps<v8i16, z_vicmphl, VMNLH, VMXLH>;
defm : IntegerMinMaxVectorOps<v4i32, z_vicmphl, VMNLF, VMXLF>;
defm : IntegerMinMaxVectorOps<v2i64, z_vicmphl, VMNLG, VMXLG>;
//===----------------------------------------------------------------------===//
// Integer comparison
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
// Element compare.
let Defs = [CC] in {
def VEC : CompareVRRaGeneric<"vec", 0xE7DB>;
def VECB : CompareVRRa<"vecb", 0xE7DB, null_frag, v128b, 0>;
def VECH : CompareVRRa<"vech", 0xE7DB, null_frag, v128h, 1>;
def VECF : CompareVRRa<"vecf", 0xE7DB, null_frag, v128f, 2>;
def VECG : CompareVRRa<"vecg", 0xE7DB, null_frag, v128g, 3>;
}
// Element compare logical.
let Defs = [CC] in {
def VECL : CompareVRRaGeneric<"vecl", 0xE7D9>;
def VECLB : CompareVRRa<"veclb", 0xE7D9, null_frag, v128b, 0>;
def VECLH : CompareVRRa<"veclh", 0xE7D9, null_frag, v128h, 1>;
def VECLF : CompareVRRa<"veclf", 0xE7D9, null_frag, v128f, 2>;
def VECLG : CompareVRRa<"veclg", 0xE7D9, null_frag, v128g, 3>;
}
// Compare equal.
def VCEQ : BinaryVRRbSPairGeneric<"vceq", 0xE7F8>;
defm VCEQB : BinaryVRRbSPair<"vceqb", 0xE7F8, z_vicmpe, z_vicmpes,
v128b, v128b, 0>;
defm VCEQH : BinaryVRRbSPair<"vceqh", 0xE7F8, z_vicmpe, z_vicmpes,
v128h, v128h, 1>;
defm VCEQF : BinaryVRRbSPair<"vceqf", 0xE7F8, z_vicmpe, z_vicmpes,
v128f, v128f, 2>;
defm VCEQG : BinaryVRRbSPair<"vceqg", 0xE7F8, z_vicmpe, z_vicmpes,
v128g, v128g, 3>;
// Compare high.
def VCH : BinaryVRRbSPairGeneric<"vch", 0xE7FB>;
defm VCHB : BinaryVRRbSPair<"vchb", 0xE7FB, z_vicmph, z_vicmphs,
v128b, v128b, 0>;
defm VCHH : BinaryVRRbSPair<"vchh", 0xE7FB, z_vicmph, z_vicmphs,
v128h, v128h, 1>;
defm VCHF : BinaryVRRbSPair<"vchf", 0xE7FB, z_vicmph, z_vicmphs,
v128f, v128f, 2>;
defm VCHG : BinaryVRRbSPair<"vchg", 0xE7FB, z_vicmph, z_vicmphs,
v128g, v128g, 3>;
// Compare high logical.
def VCHL : BinaryVRRbSPairGeneric<"vchl", 0xE7F9>;
defm VCHLB : BinaryVRRbSPair<"vchlb", 0xE7F9, z_vicmphl, z_vicmphls,
v128b, v128b, 0>;
defm VCHLH : BinaryVRRbSPair<"vchlh", 0xE7F9, z_vicmphl, z_vicmphls,
v128h, v128h, 1>;
defm VCHLF : BinaryVRRbSPair<"vchlf", 0xE7F9, z_vicmphl, z_vicmphls,
v128f, v128f, 2>;
defm VCHLG : BinaryVRRbSPair<"vchlg", 0xE7F9, z_vicmphl, z_vicmphls,
v128g, v128g, 3>;
// Test under mask.
let Defs = [CC] in
def VTM : CompareVRRa<"vtm", 0xE7D8, z_vtm, v128b, 0>;
}
//===----------------------------------------------------------------------===//
// Floating-point arithmetic
//===----------------------------------------------------------------------===//
// See comments in SystemZInstrFP.td for the suppression flags and
// rounding modes.
multiclass VectorRounding<Instruction insn, TypedReg tr> {
def : FPConversion<insn, any_frint, tr, tr, 0, 0>;
def : FPConversion<insn, any_fnearbyint, tr, tr, 4, 0>;
def : FPConversion<insn, any_ffloor, tr, tr, 4, 7>;
def : FPConversion<insn, any_fceil, tr, tr, 4, 6>;
def : FPConversion<insn, any_ftrunc, tr, tr, 4, 5>;
def : FPConversion<insn, any_fround, tr, tr, 4, 1>;
}
let Predicates = [FeatureVector] in {
// Add.
let Uses = [FPC], mayRaiseFPException = 1, isCommutable = 1 in {
def VFA : BinaryVRRcFloatGeneric<"vfa", 0xE7E3>;
def VFADB : BinaryVRRc<"vfadb", 0xE7E3, any_fadd, v128db, v128db, 3, 0>;
def WFADB : BinaryVRRc<"wfadb", 0xE7E3, any_fadd, v64db, v64db, 3, 8, 0,
"adbr">;
let Predicates = [FeatureVectorEnhancements1] in {
def VFASB : BinaryVRRc<"vfasb", 0xE7E3, any_fadd, v128sb, v128sb, 2, 0>;
def WFASB : BinaryVRRc<"wfasb", 0xE7E3, any_fadd, v32sb, v32sb, 2, 8, 0,
"aebr">;
def WFAXB : BinaryVRRc<"wfaxb", 0xE7E3, any_fadd, v128xb, v128xb, 4, 8>;
}
}
// Convert from fixed.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VCDG : TernaryVRRaFloatGeneric<"vcdg", 0xE7C3>;
def VCDGB : TernaryVRRa<"vcdgb", 0xE7C3, null_frag, v128db, v128g, 3, 0>;
def WCDGB : TernaryVRRa<"wcdgb", 0xE7C3, null_frag, v64db, v64g, 3, 8>;
}
def : FPConversion<VCDGB, any_sint_to_fp, v128db, v128g, 0, 0>;
let Predicates = [FeatureVectorEnhancements2] in {
let Uses = [FPC], mayRaiseFPException = 1 in {
let isAsmParserOnly = 1 in
def VCFPS : TernaryVRRaFloatGeneric<"vcfps", 0xE7C3>;
def VCEFB : TernaryVRRa<"vcefb", 0xE7C3, null_frag, v128sb, v128g, 2, 0>;
def WCEFB : TernaryVRRa<"wcefb", 0xE7C3, null_frag, v32sb, v32f, 2, 8>;
}
def : FPConversion<VCEFB, any_sint_to_fp, v128sb, v128f, 0, 0>;
}
// Convert from logical.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VCDLG : TernaryVRRaFloatGeneric<"vcdlg", 0xE7C1>;
def VCDLGB : TernaryVRRa<"vcdlgb", 0xE7C1, null_frag, v128db, v128g, 3, 0>;
def WCDLGB : TernaryVRRa<"wcdlgb", 0xE7C1, null_frag, v64db, v64g, 3, 8>;
}
def : FPConversion<VCDLGB, any_uint_to_fp, v128db, v128g, 0, 0>;
let Predicates = [FeatureVectorEnhancements2] in {
let Uses = [FPC], mayRaiseFPException = 1 in {
let isAsmParserOnly = 1 in
def VCFPL : TernaryVRRaFloatGeneric<"vcfpl", 0xE7C1>;
def VCELFB : TernaryVRRa<"vcelfb", 0xE7C1, null_frag, v128sb, v128g, 2, 0>;
def WCELFB : TernaryVRRa<"wcelfb", 0xE7C1, null_frag, v32sb, v32f, 2, 8>;
}
def : FPConversion<VCELFB, any_uint_to_fp, v128sb, v128f, 0, 0>;
}
// Convert to fixed.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VCGD : TernaryVRRaFloatGeneric<"vcgd", 0xE7C2>;
def VCGDB : TernaryVRRa<"vcgdb", 0xE7C2, null_frag, v128g, v128db, 3, 0>;
def WCGDB : TernaryVRRa<"wcgdb", 0xE7C2, null_frag, v64g, v64db, 3, 8>;
}
// Rounding mode should agree with SystemZInstrFP.td.
def : FPConversion<VCGDB, any_fp_to_sint, v128g, v128db, 0, 5>;
let Predicates = [FeatureVectorEnhancements2] in {
let Uses = [FPC], mayRaiseFPException = 1 in {
let isAsmParserOnly = 1 in
def VCSFP : TernaryVRRaFloatGeneric<"vcsfp", 0xE7C2>;
def VCFEB : TernaryVRRa<"vcfeb", 0xE7C2, null_frag, v128sb, v128g, 2, 0>;
def WCFEB : TernaryVRRa<"wcfeb", 0xE7C2, null_frag, v32sb, v32f, 2, 8>;
}
// Rounding mode should agree with SystemZInstrFP.td.
def : FPConversion<VCFEB, any_fp_to_sint, v128f, v128sb, 0, 5>;
}
// Convert to logical.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VCLGD : TernaryVRRaFloatGeneric<"vclgd", 0xE7C0>;
def VCLGDB : TernaryVRRa<"vclgdb", 0xE7C0, null_frag, v128g, v128db, 3, 0>;
def WCLGDB : TernaryVRRa<"wclgdb", 0xE7C0, null_frag, v64g, v64db, 3, 8>;
}
// Rounding mode should agree with SystemZInstrFP.td.
def : FPConversion<VCLGDB, any_fp_to_uint, v128g, v128db, 0, 5>;
let Predicates = [FeatureVectorEnhancements2] in {
let Uses = [FPC], mayRaiseFPException = 1 in {
let isAsmParserOnly = 1 in
def VCLFP : TernaryVRRaFloatGeneric<"vclfp", 0xE7C0>;
def VCLFEB : TernaryVRRa<"vclfeb", 0xE7C0, null_frag, v128sb, v128g, 2, 0>;
def WCLFEB : TernaryVRRa<"wclfeb", 0xE7C0, null_frag, v32sb, v32f, 2, 8>;
}
// Rounding mode should agree with SystemZInstrFP.td.
def : FPConversion<VCLFEB, any_fp_to_uint, v128f, v128sb, 0, 5>;
}
// Divide.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VFD : BinaryVRRcFloatGeneric<"vfd", 0xE7E5>;
def VFDDB : BinaryVRRc<"vfddb", 0xE7E5, any_fdiv, v128db, v128db, 3, 0>;
def WFDDB : BinaryVRRc<"wfddb", 0xE7E5, any_fdiv, v64db, v64db, 3, 8, 0,
"ddbr">;
let Predicates = [FeatureVectorEnhancements1] in {
def VFDSB : BinaryVRRc<"vfdsb", 0xE7E5, any_fdiv, v128sb, v128sb, 2, 0>;
def WFDSB : BinaryVRRc<"wfdsb", 0xE7E5, any_fdiv, v32sb, v32sb, 2, 8, 0,
"debr">;
def WFDXB : BinaryVRRc<"wfdxb", 0xE7E5, any_fdiv, v128xb, v128xb, 4, 8>;
}
}
// Load FP integer.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VFI : TernaryVRRaFloatGeneric<"vfi", 0xE7C7>;
def VFIDB : TernaryVRRa<"vfidb", 0xE7C7, int_s390_vfidb, v128db, v128db, 3, 0>;
def WFIDB : TernaryVRRa<"wfidb", 0xE7C7, null_frag, v64db, v64db, 3, 8>;
}
defm : VectorRounding<VFIDB, v128db>;
defm : VectorRounding<WFIDB, v64db>;
let Predicates = [FeatureVectorEnhancements1] in {
let Uses = [FPC], mayRaiseFPException = 1 in {
def VFISB : TernaryVRRa<"vfisb", 0xE7C7, int_s390_vfisb, v128sb, v128sb, 2, 0>;
def WFISB : TernaryVRRa<"wfisb", 0xE7C7, null_frag, v32sb, v32sb, 2, 8>;
def WFIXB : TernaryVRRa<"wfixb", 0xE7C7, null_frag, v128xb, v128xb, 4, 8>;
}
defm : VectorRounding<VFISB, v128sb>;
defm : VectorRounding<WFISB, v32sb>;
defm : VectorRounding<WFIXB, v128xb>;
}
// Load lengthened.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VLDE : UnaryVRRaFloatGeneric<"vlde", 0xE7C4>;
def VLDEB : UnaryVRRa<"vldeb", 0xE7C4, z_any_vextend, v128db, v128sb, 2, 0>;
def WLDEB : UnaryVRRa<"wldeb", 0xE7C4, any_fpextend, v64db, v32sb, 2, 8, 0,
"ldebr">;
}
let Predicates = [FeatureVectorEnhancements1] in {
let Uses = [FPC], mayRaiseFPException = 1 in {
let isAsmParserOnly = 1 in {
def VFLL : UnaryVRRaFloatGeneric<"vfll", 0xE7C4>;
def VFLLS : UnaryVRRa<"vflls", 0xE7C4, null_frag, v128db, v128sb, 2, 0>;
def WFLLS : UnaryVRRa<"wflls", 0xE7C4, null_frag, v64db, v32sb, 2, 8>;
}
def WFLLD : UnaryVRRa<"wflld", 0xE7C4, any_fpextend, v128xb, v64db, 3, 8>;
}
def : Pat<(f128 (any_fpextend (f32 VR32:$src))),
(WFLLD (WLDEB VR32:$src))>;
}
// Load rounded.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VLED : TernaryVRRaFloatGeneric<"vled", 0xE7C5>;
def VLEDB : TernaryVRRa<"vledb", 0xE7C5, null_frag, v128sb, v128db, 3, 0>;
def WLEDB : TernaryVRRa<"wledb", 0xE7C5, null_frag, v32sb, v64db, 3, 8>;
}
def : Pat<(v4f32 (z_any_vround (v2f64 VR128:$src))), (VLEDB VR128:$src, 0, 0)>;
def : FPConversion<WLEDB, any_fpround, v32sb, v64db, 0, 0>;
let Predicates = [FeatureVectorEnhancements1] in {
let Uses = [FPC], mayRaiseFPException = 1 in {
let isAsmParserOnly = 1 in {
def VFLR : TernaryVRRaFloatGeneric<"vflr", 0xE7C5>;
def VFLRD : TernaryVRRa<"vflrd", 0xE7C5, null_frag, v128sb, v128db, 3, 0>;
def WFLRD : TernaryVRRa<"wflrd", 0xE7C5, null_frag, v32sb, v64db, 3, 8>;
}
def WFLRX : TernaryVRRa<"wflrx", 0xE7C5, null_frag, v64db, v128xb, 4, 8>;
}
def : FPConversion<WFLRX, any_fpround, v64db, v128xb, 0, 0>;
def : Pat<(f32 (any_fpround (f128 VR128:$src))),
(WLEDB (WFLRX VR128:$src, 0, 3), 0, 0)>;
}
// Maximum.
multiclass VectorMax<Instruction insn, TypedReg tr> {
def : FPMinMax<insn, any_fmaxnum, tr, 4>;
def : FPMinMax<insn, any_fmaximum, tr, 1>;
}
let Predicates = [FeatureVectorEnhancements1] in {
let Uses = [FPC], mayRaiseFPException = 1, isCommutable = 1 in {
def VFMAX : TernaryVRRcFloatGeneric<"vfmax", 0xE7EF>;
def VFMAXDB : TernaryVRRcFloat<"vfmaxdb", 0xE7EF, int_s390_vfmaxdb,
v128db, v128db, 3, 0>;
def WFMAXDB : TernaryVRRcFloat<"wfmaxdb", 0xE7EF, null_frag,
v64db, v64db, 3, 8>;
def VFMAXSB : TernaryVRRcFloat<"vfmaxsb", 0xE7EF, int_s390_vfmaxsb,
v128sb, v128sb, 2, 0>;
def WFMAXSB : TernaryVRRcFloat<"wfmaxsb", 0xE7EF, null_frag,
v32sb, v32sb, 2, 8>;
def WFMAXXB : TernaryVRRcFloat<"wfmaxxb", 0xE7EF, null_frag,
v128xb, v128xb, 4, 8>;
}
defm : VectorMax<VFMAXDB, v128db>;
defm : VectorMax<WFMAXDB, v64db>;
defm : VectorMax<VFMAXSB, v128sb>;
defm : VectorMax<WFMAXSB, v32sb>;
defm : VectorMax<WFMAXXB, v128xb>;
}
// Minimum.
multiclass VectorMin<Instruction insn, TypedReg tr> {
def : FPMinMax<insn, any_fminnum, tr, 4>;
def : FPMinMax<insn, any_fminimum, tr, 1>;
}
let Predicates = [FeatureVectorEnhancements1] in {
let Uses = [FPC], mayRaiseFPException = 1, isCommutable = 1 in {
def VFMIN : TernaryVRRcFloatGeneric<"vfmin", 0xE7EE>;
def VFMINDB : TernaryVRRcFloat<"vfmindb", 0xE7EE, int_s390_vfmindb,
v128db, v128db, 3, 0>;
def WFMINDB : TernaryVRRcFloat<"wfmindb", 0xE7EE, null_frag,
v64db, v64db, 3, 8>;
def VFMINSB : TernaryVRRcFloat<"vfminsb", 0xE7EE, int_s390_vfminsb,
v128sb, v128sb, 2, 0>;
def WFMINSB : TernaryVRRcFloat<"wfminsb", 0xE7EE, null_frag,
v32sb, v32sb, 2, 8>;
def WFMINXB : TernaryVRRcFloat<"wfminxb", 0xE7EE, null_frag,
v128xb, v128xb, 4, 8>;
}
defm : VectorMin<VFMINDB, v128db>;
defm : VectorMin<WFMINDB, v64db>;
defm : VectorMin<VFMINSB, v128sb>;
defm : VectorMin<WFMINSB, v32sb>;
defm : VectorMin<WFMINXB, v128xb>;
}
// Multiply.
let Uses = [FPC], mayRaiseFPException = 1, isCommutable = 1 in {
def VFM : BinaryVRRcFloatGeneric<"vfm", 0xE7E7>;
def VFMDB : BinaryVRRc<"vfmdb", 0xE7E7, any_fmul, v128db, v128db, 3, 0>;
def WFMDB : BinaryVRRc<"wfmdb", 0xE7E7, any_fmul, v64db, v64db, 3, 8, 0,
"mdbr">;
let Predicates = [FeatureVectorEnhancements1] in {
def VFMSB : BinaryVRRc<"vfmsb", 0xE7E7, any_fmul, v128sb, v128sb, 2, 0>;
def WFMSB : BinaryVRRc<"wfmsb", 0xE7E7, any_fmul, v32sb, v32sb, 2, 8, 0,
"meebr">;
def WFMXB : BinaryVRRc<"wfmxb", 0xE7E7, any_fmul, v128xb, v128xb, 4, 8>;
}
}
// Multiply and add.
let Uses = [FPC], mayRaiseFPException = 1, isCommutable = 1 in {
def VFMA : TernaryVRReFloatGeneric<"vfma", 0xE78F>;
def VFMADB : TernaryVRRe<"vfmadb", 0xE78F, any_fma, v128db, v128db, 0, 3>;
def WFMADB : TernaryVRRe<"wfmadb", 0xE78F, any_fma, v64db, v64db, 8, 3,
"madbr">;
let Predicates = [FeatureVectorEnhancements1] in {
def VFMASB : TernaryVRRe<"vfmasb", 0xE78F, any_fma, v128sb, v128sb, 0, 2>;
def WFMASB : TernaryVRRe<"wfmasb", 0xE78F, any_fma, v32sb, v32sb, 8, 2,
"maebr">;
def WFMAXB : TernaryVRRe<"wfmaxb", 0xE78F, any_fma, v128xb, v128xb, 8, 4>;
}
}
// Multiply and subtract.
let Uses = [FPC], mayRaiseFPException = 1, isCommutable = 1 in {
def VFMS : TernaryVRReFloatGeneric<"vfms", 0xE78E>;
def VFMSDB : TernaryVRRe<"vfmsdb", 0xE78E, any_fms, v128db, v128db, 0, 3>;
def WFMSDB : TernaryVRRe<"wfmsdb", 0xE78E, any_fms, v64db, v64db, 8, 3,
"msdbr">;
let Predicates = [FeatureVectorEnhancements1] in {
def VFMSSB : TernaryVRRe<"vfmssb", 0xE78E, any_fms, v128sb, v128sb, 0, 2>;
def WFMSSB : TernaryVRRe<"wfmssb", 0xE78E, any_fms, v32sb, v32sb, 8, 2,
"msebr">;
def WFMSXB : TernaryVRRe<"wfmsxb", 0xE78E, any_fms, v128xb, v128xb, 8, 4>;
}
}
// Negative multiply and add.
let Uses = [FPC], mayRaiseFPException = 1, isCommutable = 1,
Predicates = [FeatureVectorEnhancements1] in {
def VFNMA : TernaryVRReFloatGeneric<"vfnma", 0xE79F>;
def VFNMADB : TernaryVRRe<"vfnmadb", 0xE79F, any_fnma, v128db, v128db, 0, 3>;
def WFNMADB : TernaryVRRe<"wfnmadb", 0xE79F, any_fnma, v64db, v64db, 8, 3>;
def VFNMASB : TernaryVRRe<"vfnmasb", 0xE79F, any_fnma, v128sb, v128sb, 0, 2>;
def WFNMASB : TernaryVRRe<"wfnmasb", 0xE79F, any_fnma, v32sb, v32sb, 8, 2>;
def WFNMAXB : TernaryVRRe<"wfnmaxb", 0xE79F, any_fnma, v128xb, v128xb, 8, 4>;
}
// Negative multiply and subtract.
let Uses = [FPC], mayRaiseFPException = 1, isCommutable = 1,
Predicates = [FeatureVectorEnhancements1] in {
def VFNMS : TernaryVRReFloatGeneric<"vfnms", 0xE79E>;
def VFNMSDB : TernaryVRRe<"vfnmsdb", 0xE79E, any_fnms, v128db, v128db, 0, 3>;
def WFNMSDB : TernaryVRRe<"wfnmsdb", 0xE79E, any_fnms, v64db, v64db, 8, 3>;
def VFNMSSB : TernaryVRRe<"vfnmssb", 0xE79E, any_fnms, v128sb, v128sb, 0, 2>;
def WFNMSSB : TernaryVRRe<"wfnmssb", 0xE79E, any_fnms, v32sb, v32sb, 8, 2>;
def WFNMSXB : TernaryVRRe<"wfnmsxb", 0xE79E, any_fnms, v128xb, v128xb, 8, 4>;
}
// Perform sign operation.
def VFPSO : BinaryVRRaFloatGeneric<"vfpso", 0xE7CC>;
def VFPSODB : BinaryVRRa<"vfpsodb", 0xE7CC, null_frag, v128db, v128db, 3, 0>;
def WFPSODB : BinaryVRRa<"wfpsodb", 0xE7CC, null_frag, v64db, v64db, 3, 8>;
let Predicates = [FeatureVectorEnhancements1] in {
def VFPSOSB : BinaryVRRa<"vfpsosb", 0xE7CC, null_frag, v128sb, v128sb, 2, 0>;
def WFPSOSB : BinaryVRRa<"wfpsosb", 0xE7CC, null_frag, v32sb, v32sb, 2, 8>;
def WFPSOXB : BinaryVRRa<"wfpsoxb", 0xE7CC, null_frag, v128xb, v128xb, 4, 8>;
}
// Load complement.
def VFLCDB : UnaryVRRa<"vflcdb", 0xE7CC, fneg, v128db, v128db, 3, 0, 0>;
def WFLCDB : UnaryVRRa<"wflcdb", 0xE7CC, fneg, v64db, v64db, 3, 8, 0>;
let Predicates = [FeatureVectorEnhancements1] in {
def VFLCSB : UnaryVRRa<"vflcsb", 0xE7CC, fneg, v128sb, v128sb, 2, 0, 0>;
def WFLCSB : UnaryVRRa<"wflcsb", 0xE7CC, fneg, v32sb, v32sb, 2, 8, 0>;
def WFLCXB : UnaryVRRa<"wflcxb", 0xE7CC, fneg, v128xb, v128xb, 4, 8, 0>;
}
// Load negative.
def VFLNDB : UnaryVRRa<"vflndb", 0xE7CC, fnabs, v128db, v128db, 3, 0, 1>;
def WFLNDB : UnaryVRRa<"wflndb", 0xE7CC, fnabs, v64db, v64db, 3, 8, 1>;
let Predicates = [FeatureVectorEnhancements1] in {
def VFLNSB : UnaryVRRa<"vflnsb", 0xE7CC, fnabs, v128sb, v128sb, 2, 0, 1>;
def WFLNSB : UnaryVRRa<"wflnsb", 0xE7CC, fnabs, v32sb, v32sb, 2, 8, 1>;
def WFLNXB : UnaryVRRa<"wflnxb", 0xE7CC, fnabs, v128xb, v128xb, 4, 8, 1>;
}
// Load positive.
def VFLPDB : UnaryVRRa<"vflpdb", 0xE7CC, fabs, v128db, v128db, 3, 0, 2>;
def WFLPDB : UnaryVRRa<"wflpdb", 0xE7CC, fabs, v64db, v64db, 3, 8, 2>;
let Predicates = [FeatureVectorEnhancements1] in {
def VFLPSB : UnaryVRRa<"vflpsb", 0xE7CC, fabs, v128sb, v128sb, 2, 0, 2>;
def WFLPSB : UnaryVRRa<"wflpsb", 0xE7CC, fabs, v32sb, v32sb, 2, 8, 2>;
def WFLPXB : UnaryVRRa<"wflpxb", 0xE7CC, fabs, v128xb, v128xb, 4, 8, 2>;
}
// Square root.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VFSQ : UnaryVRRaFloatGeneric<"vfsq", 0xE7CE>;
def VFSQDB : UnaryVRRa<"vfsqdb", 0xE7CE, any_fsqrt, v128db, v128db, 3, 0>;
def WFSQDB : UnaryVRRa<"wfsqdb", 0xE7CE, any_fsqrt, v64db, v64db, 3, 8, 0,
"sqdbr">;
let Predicates = [FeatureVectorEnhancements1] in {
def VFSQSB : UnaryVRRa<"vfsqsb", 0xE7CE, any_fsqrt, v128sb, v128sb, 2, 0>;
def WFSQSB : UnaryVRRa<"wfsqsb", 0xE7CE, any_fsqrt, v32sb, v32sb, 2, 8, 0,
"sqebr">;
def WFSQXB : UnaryVRRa<"wfsqxb", 0xE7CE, any_fsqrt, v128xb, v128xb, 4, 8>;
}
}
// Subtract.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VFS : BinaryVRRcFloatGeneric<"vfs", 0xE7E2>;
def VFSDB : BinaryVRRc<"vfsdb", 0xE7E2, any_fsub, v128db, v128db, 3, 0>;
def WFSDB : BinaryVRRc<"wfsdb", 0xE7E2, any_fsub, v64db, v64db, 3, 8, 0,
"sdbr">;
let Predicates = [FeatureVectorEnhancements1] in {
def VFSSB : BinaryVRRc<"vfssb", 0xE7E2, any_fsub, v128sb, v128sb, 2, 0>;
def WFSSB : BinaryVRRc<"wfssb", 0xE7E2, any_fsub, v32sb, v32sb, 2, 8, 0,
"sebr">;
def WFSXB : BinaryVRRc<"wfsxb", 0xE7E2, any_fsub, v128xb, v128xb, 4, 8>;
}
}
// Test data class immediate.
let Defs = [CC] in {
def VFTCI : BinaryVRIeFloatGeneric<"vftci", 0xE74A>;
def VFTCIDB : BinaryVRIe<"vftcidb", 0xE74A, z_vftci, v128g, v128db, 3, 0>;
def WFTCIDB : BinaryVRIe<"wftcidb", 0xE74A, null_frag, v64g, v64db, 3, 8>;
let Predicates = [FeatureVectorEnhancements1] in {
def VFTCISB : BinaryVRIe<"vftcisb", 0xE74A, z_vftci, v128f, v128sb, 2, 0>;
def WFTCISB : BinaryVRIe<"wftcisb", 0xE74A, null_frag, v32f, v32sb, 2, 8>;
def WFTCIXB : BinaryVRIe<"wftcixb", 0xE74A, null_frag, v128q, v128xb, 4, 8>;
}
}
}
//===----------------------------------------------------------------------===//
// Floating-point comparison
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
// Compare scalar.
let Uses = [FPC], mayRaiseFPException = 1, Defs = [CC] in {
def WFC : CompareVRRaFloatGeneric<"wfc", 0xE7CB>;
def WFCDB : CompareVRRa<"wfcdb", 0xE7CB, z_any_fcmp, v64db, 3, "cdbr">;
let Predicates = [FeatureVectorEnhancements1] in {
def WFCSB : CompareVRRa<"wfcsb", 0xE7CB, z_any_fcmp, v32sb, 2, "cebr">;
def WFCXB : CompareVRRa<"wfcxb", 0xE7CB, z_any_fcmp, v128xb, 4>;
}
}
// Compare and signal scalar.
let Uses = [FPC], mayRaiseFPException = 1, Defs = [CC] in {
def WFK : CompareVRRaFloatGeneric<"wfk", 0xE7CA>;
def WFKDB : CompareVRRa<"wfkdb", 0xE7CA, z_strict_fcmps, v64db, 3, "kdbr">;
let Predicates = [FeatureVectorEnhancements1] in {
def WFKSB : CompareVRRa<"wfksb", 0xE7CA, z_strict_fcmps, v32sb, 2, "kebr">;
def WFKXB : CompareVRRa<"wfkxb", 0xE7CA, z_strict_fcmps, v128xb, 4>;
}
}
// Compare equal.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VFCE : BinaryVRRcSPairFloatGeneric<"vfce", 0xE7E8>;
defm VFCEDB : BinaryVRRcSPair<"vfcedb", 0xE7E8, z_any_vfcmpe, z_vfcmpes,
v128g, v128db, 3, 0>;
defm WFCEDB : BinaryVRRcSPair<"wfcedb", 0xE7E8, null_frag, null_frag,
v64g, v64db, 3, 8>;
let Predicates = [FeatureVectorEnhancements1] in {
defm VFCESB : BinaryVRRcSPair<"vfcesb", 0xE7E8, z_any_vfcmpe, z_vfcmpes,
v128f, v128sb, 2, 0>;
defm WFCESB : BinaryVRRcSPair<"wfcesb", 0xE7E8, null_frag, null_frag,
v32f, v32sb, 2, 8>;
defm WFCEXB : BinaryVRRcSPair<"wfcexb", 0xE7E8, null_frag, null_frag,
v128q, v128xb, 4, 8>;
}
}
// Compare and signal equal.
let Uses = [FPC], mayRaiseFPException = 1,
Predicates = [FeatureVectorEnhancements1] in {
defm VFKEDB : BinaryVRRcSPair<"vfkedb", 0xE7E8, z_strict_vfcmpes, null_frag,
v128g, v128db, 3, 4>;
defm WFKEDB : BinaryVRRcSPair<"wfkedb", 0xE7E8, null_frag, null_frag,
v64g, v64db, 3, 12>;
defm VFKESB : BinaryVRRcSPair<"vfkesb", 0xE7E8, z_strict_vfcmpes, null_frag,
v128f, v128sb, 2, 4>;
defm WFKESB : BinaryVRRcSPair<"wfkesb", 0xE7E8, null_frag, null_frag,
v32f, v32sb, 2, 12>;
defm WFKEXB : BinaryVRRcSPair<"wfkexb", 0xE7E8, null_frag, null_frag,
v128q, v128xb, 4, 12>;
}
// Compare high.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VFCH : BinaryVRRcSPairFloatGeneric<"vfch", 0xE7EB>;
defm VFCHDB : BinaryVRRcSPair<"vfchdb", 0xE7EB, z_any_vfcmph, z_vfcmphs,
v128g, v128db, 3, 0>;
defm WFCHDB : BinaryVRRcSPair<"wfchdb", 0xE7EB, null_frag, null_frag,
v64g, v64db, 3, 8>;
let Predicates = [FeatureVectorEnhancements1] in {
defm VFCHSB : BinaryVRRcSPair<"vfchsb", 0xE7EB, z_any_vfcmph, z_vfcmphs,
v128f, v128sb, 2, 0>;
defm WFCHSB : BinaryVRRcSPair<"wfchsb", 0xE7EB, null_frag, null_frag,
v32f, v32sb, 2, 8>;
defm WFCHXB : BinaryVRRcSPair<"wfchxb", 0xE7EB, null_frag, null_frag,
v128q, v128xb, 4, 8>;
}
}
// Compare and signal high.
let Uses = [FPC], mayRaiseFPException = 1,
Predicates = [FeatureVectorEnhancements1] in {
defm VFKHDB : BinaryVRRcSPair<"vfkhdb", 0xE7EB, z_strict_vfcmphs, null_frag,
v128g, v128db, 3, 4>;
defm WFKHDB : BinaryVRRcSPair<"wfkhdb", 0xE7EB, null_frag, null_frag,
v64g, v64db, 3, 12>;
defm VFKHSB : BinaryVRRcSPair<"vfkhsb", 0xE7EB, z_strict_vfcmphs, null_frag,
v128f, v128sb, 2, 4>;
defm WFKHSB : BinaryVRRcSPair<"wfkhsb", 0xE7EB, null_frag, null_frag,
v32f, v32sb, 2, 12>;
defm WFKHXB : BinaryVRRcSPair<"wfkhxb", 0xE7EB, null_frag, null_frag,
v128q, v128xb, 4, 12>;
}
// Compare high or equal.
let Uses = [FPC], mayRaiseFPException = 1 in {
def VFCHE : BinaryVRRcSPairFloatGeneric<"vfche", 0xE7EA>;
defm VFCHEDB : BinaryVRRcSPair<"vfchedb", 0xE7EA, z_any_vfcmphe, z_vfcmphes,
v128g, v128db, 3, 0>;
defm WFCHEDB : BinaryVRRcSPair<"wfchedb", 0xE7EA, null_frag, null_frag,
v64g, v64db, 3, 8>;
let Predicates = [FeatureVectorEnhancements1] in {
defm VFCHESB : BinaryVRRcSPair<"vfchesb", 0xE7EA, z_any_vfcmphe, z_vfcmphes,
v128f, v128sb, 2, 0>;
defm WFCHESB : BinaryVRRcSPair<"wfchesb", 0xE7EA, null_frag, null_frag,
v32f, v32sb, 2, 8>;
defm WFCHEXB : BinaryVRRcSPair<"wfchexb", 0xE7EA, null_frag, null_frag,
v128q, v128xb, 4, 8>;
}
}
// Compare and signal high or equal.
let Uses = [FPC], mayRaiseFPException = 1,
Predicates = [FeatureVectorEnhancements1] in {
defm VFKHEDB : BinaryVRRcSPair<"vfkhedb", 0xE7EA, z_strict_vfcmphes, null_frag,
v128g, v128db, 3, 4>;
defm WFKHEDB : BinaryVRRcSPair<"wfkhedb", 0xE7EA, null_frag, null_frag,
v64g, v64db, 3, 12>;
defm VFKHESB : BinaryVRRcSPair<"vfkhesb", 0xE7EA, z_strict_vfcmphes, null_frag,
v128f, v128sb, 2, 4>;
defm WFKHESB : BinaryVRRcSPair<"wfkhesb", 0xE7EA, null_frag, null_frag,
v32f, v32sb, 2, 12>;
defm WFKHEXB : BinaryVRRcSPair<"wfkhexb", 0xE7EA, null_frag, null_frag,
v128q, v128xb, 4, 12>;
}
}
//===----------------------------------------------------------------------===//
// Conversions
//===----------------------------------------------------------------------===//
def : Pat<(v16i8 (bitconvert (v8i16 VR128:$src))), (v16i8 VR128:$src)>;
def : Pat<(v16i8 (bitconvert (v4i32 VR128:$src))), (v16i8 VR128:$src)>;
def : Pat<(v16i8 (bitconvert (v2i64 VR128:$src))), (v16i8 VR128:$src)>;
def : Pat<(v16i8 (bitconvert (v4f32 VR128:$src))), (v16i8 VR128:$src)>;
def : Pat<(v16i8 (bitconvert (v2f64 VR128:$src))), (v16i8 VR128:$src)>;
def : Pat<(v16i8 (bitconvert (f128 VR128:$src))), (v16i8 VR128:$src)>;
def : Pat<(v8i16 (bitconvert (v16i8 VR128:$src))), (v8i16 VR128:$src)>;
def : Pat<(v8i16 (bitconvert (v4i32 VR128:$src))), (v8i16 VR128:$src)>;
def : Pat<(v8i16 (bitconvert (v2i64 VR128:$src))), (v8i16 VR128:$src)>;
def : Pat<(v8i16 (bitconvert (v4f32 VR128:$src))), (v8i16 VR128:$src)>;
def : Pat<(v8i16 (bitconvert (v2f64 VR128:$src))), (v8i16 VR128:$src)>;
def : Pat<(v8i16 (bitconvert (f128 VR128:$src))), (v8i16 VR128:$src)>;
def : Pat<(v4i32 (bitconvert (v16i8 VR128:$src))), (v4i32 VR128:$src)>;
def : Pat<(v4i32 (bitconvert (v8i16 VR128:$src))), (v4i32 VR128:$src)>;
def : Pat<(v4i32 (bitconvert (v2i64 VR128:$src))), (v4i32 VR128:$src)>;
def : Pat<(v4i32 (bitconvert (v4f32 VR128:$src))), (v4i32 VR128:$src)>;
def : Pat<(v4i32 (bitconvert (v2f64 VR128:$src))), (v4i32 VR128:$src)>;
def : Pat<(v4i32 (bitconvert (f128 VR128:$src))), (v4i32 VR128:$src)>;
def : Pat<(v2i64 (bitconvert (v16i8 VR128:$src))), (v2i64 VR128:$src)>;
def : Pat<(v2i64 (bitconvert (v8i16 VR128:$src))), (v2i64 VR128:$src)>;
def : Pat<(v2i64 (bitconvert (v4i32 VR128:$src))), (v2i64 VR128:$src)>;
def : Pat<(v2i64 (bitconvert (v4f32 VR128:$src))), (v2i64 VR128:$src)>;
def : Pat<(v2i64 (bitconvert (v2f64 VR128:$src))), (v2i64 VR128:$src)>;
def : Pat<(v2i64 (bitconvert (f128 VR128:$src))), (v2i64 VR128:$src)>;
def : Pat<(v4f32 (bitconvert (v16i8 VR128:$src))), (v4f32 VR128:$src)>;
def : Pat<(v4f32 (bitconvert (v8i16 VR128:$src))), (v4f32 VR128:$src)>;
def : Pat<(v4f32 (bitconvert (v4i32 VR128:$src))), (v4f32 VR128:$src)>;
def : Pat<(v4f32 (bitconvert (v2i64 VR128:$src))), (v4f32 VR128:$src)>;
def : Pat<(v4f32 (bitconvert (v2f64 VR128:$src))), (v4f32 VR128:$src)>;
def : Pat<(v4f32 (bitconvert (f128 VR128:$src))), (v4f32 VR128:$src)>;
def : Pat<(v2f64 (bitconvert (v16i8 VR128:$src))), (v2f64 VR128:$src)>;
def : Pat<(v2f64 (bitconvert (v8i16 VR128:$src))), (v2f64 VR128:$src)>;
def : Pat<(v2f64 (bitconvert (v4i32 VR128:$src))), (v2f64 VR128:$src)>;
def : Pat<(v2f64 (bitconvert (v2i64 VR128:$src))), (v2f64 VR128:$src)>;
def : Pat<(v2f64 (bitconvert (v4f32 VR128:$src))), (v2f64 VR128:$src)>;
def : Pat<(v2f64 (bitconvert (f128 VR128:$src))), (v2f64 VR128:$src)>;
def : Pat<(f128 (bitconvert (v16i8 VR128:$src))), (f128 VR128:$src)>;
def : Pat<(f128 (bitconvert (v8i16 VR128:$src))), (f128 VR128:$src)>;
def : Pat<(f128 (bitconvert (v4i32 VR128:$src))), (f128 VR128:$src)>;
def : Pat<(f128 (bitconvert (v2i64 VR128:$src))), (f128 VR128:$src)>;
def : Pat<(f128 (bitconvert (v4f32 VR128:$src))), (f128 VR128:$src)>;
def : Pat<(f128 (bitconvert (v2f64 VR128:$src))), (f128 VR128:$src)>;
//===----------------------------------------------------------------------===//
// Replicating scalars
//===----------------------------------------------------------------------===//
// Define patterns for replicating a scalar GR32 into a vector of type TYPE.
// INDEX is 8 minus the element size in bytes.
class VectorReplicateScalar<ValueType type, Instruction insn, bits<16> index>
: Pat<(type (z_replicate GR32:$scalar)),
(insn (VLVGP32 GR32:$scalar, GR32:$scalar), index)>;
def : VectorReplicateScalar<v16i8, VREPB, 7>;
def : VectorReplicateScalar<v8i16, VREPH, 3>;
def : VectorReplicateScalar<v4i32, VREPF, 1>;
// i64 replications are just a single instruction.
def : Pat<(v2i64 (z_replicate GR64:$scalar)),
(VLVGP GR64:$scalar, GR64:$scalar)>;
//===----------------------------------------------------------------------===//
// Floating-point insertion and extraction
//===----------------------------------------------------------------------===//
// Moving 32-bit values between GPRs and FPRs can be done using VLVGF
// and VLGVF.
let Predicates = [FeatureVector] in {
def LEFR : UnaryAliasVRS<VR32, GR32>;
def LFER : UnaryAliasVRS<GR64, VR32>;
def : Pat<(f32 (bitconvert (i32 GR32:$src))), (LEFR GR32:$src)>;
def : Pat<(i32 (bitconvert (f32 VR32:$src))),
(EXTRACT_SUBREG (LFER VR32:$src), subreg_l32)>;
}
// Floating-point values are stored in element 0 of the corresponding
// vector register. Scalar to vector conversion is just a subreg and
// scalar replication can just replicate element 0 of the vector register.
multiclass ScalarToVectorFP<Instruction vrep, ValueType vt, RegisterOperand cls,
SubRegIndex subreg> {
def : Pat<(vt (scalar_to_vector cls:$scalar)),
(INSERT_SUBREG (vt (IMPLICIT_DEF)), cls:$scalar, subreg)>;
def : Pat<(vt (z_replicate cls:$scalar)),
(vrep (INSERT_SUBREG (vt (IMPLICIT_DEF)), cls:$scalar,
subreg), 0)>;
}
defm : ScalarToVectorFP<VREPF, v4f32, FP32, subreg_h32>;
defm : ScalarToVectorFP<VREPG, v2f64, FP64, subreg_h64>;
// Match v2f64 insertions. The AddedComplexity counters the 3 added by
// TableGen for the base register operand in VLVG-based integer insertions
// and ensures that this version is strictly better.
let AddedComplexity = 4 in {
def : Pat<(z_vector_insert (v2f64 VR128:$vec), FP64:$elt, 0),
(VPDI (INSERT_SUBREG (v2f64 (IMPLICIT_DEF)), FP64:$elt,
subreg_h64), VR128:$vec, 1)>;
def : Pat<(z_vector_insert (v2f64 VR128:$vec), FP64:$elt, 1),
(VPDI VR128:$vec, (INSERT_SUBREG (v2f64 (IMPLICIT_DEF)), FP64:$elt,
subreg_h64), 0)>;
}
// We extract floating-point element X by replicating (for elements other
// than 0) and then taking a high subreg. The AddedComplexity counters the
// 3 added by TableGen for the base register operand in VLGV-based integer
// extractions and ensures that this version is strictly better.
let AddedComplexity = 4 in {
def : Pat<(f32 (z_vector_extract (v4f32 VR128:$vec), 0)),
(EXTRACT_SUBREG VR128:$vec, subreg_h32)>;
def : Pat<(f32 (z_vector_extract (v4f32 VR128:$vec), imm32zx2:$index)),
(EXTRACT_SUBREG (VREPF VR128:$vec, imm32zx2:$index), subreg_h32)>;
def : Pat<(f64 (z_vector_extract (v2f64 VR128:$vec), 0)),
(EXTRACT_SUBREG VR128:$vec, subreg_h64)>;
def : Pat<(f64 (z_vector_extract (v2f64 VR128:$vec), imm32zx1:$index)),
(EXTRACT_SUBREG (VREPG VR128:$vec, imm32zx1:$index), subreg_h64)>;
}
//===----------------------------------------------------------------------===//
// Support for 128-bit floating-point values in vector registers
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVectorEnhancements1] in {
def : Pat<(f128 (load bdxaddr12only:$addr)),
(VL bdxaddr12only:$addr)>;
def : Pat<(store (f128 VR128:$src), bdxaddr12only:$addr),
(VST VR128:$src, bdxaddr12only:$addr)>;
def : Pat<(f128 fpimm0), (VZERO)>;
def : Pat<(f128 fpimmneg0), (WFLNXB (VZERO))>;
}
//===----------------------------------------------------------------------===//
// String instructions
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVector] in {
defm VFAE : TernaryOptVRRbSPairGeneric<"vfae", 0xE782>;
defm VFAEB : TernaryOptVRRbSPair<"vfaeb", 0xE782, int_s390_vfaeb,
z_vfae_cc, v128b, v128b, 0>;
defm VFAEH : TernaryOptVRRbSPair<"vfaeh", 0xE782, int_s390_vfaeh,
z_vfae_cc, v128h, v128h, 1>;
defm VFAEF : TernaryOptVRRbSPair<"vfaef", 0xE782, int_s390_vfaef,
z_vfae_cc, v128f, v128f, 2>;
defm VFAEZB : TernaryOptVRRbSPair<"vfaezb", 0xE782, int_s390_vfaezb,
z_vfaez_cc, v128b, v128b, 0, 2>;
defm VFAEZH : TernaryOptVRRbSPair<"vfaezh", 0xE782, int_s390_vfaezh,
z_vfaez_cc, v128h, v128h, 1, 2>;
defm VFAEZF : TernaryOptVRRbSPair<"vfaezf", 0xE782, int_s390_vfaezf,
z_vfaez_cc, v128f, v128f, 2, 2>;
defm VFEE : BinaryExtraVRRbSPairGeneric<"vfee", 0xE780>;
defm VFEEB : BinaryExtraVRRbSPair<"vfeeb", 0xE780, int_s390_vfeeb,
z_vfee_cc, v128b, v128b, 0>;
defm VFEEH : BinaryExtraVRRbSPair<"vfeeh", 0xE780, int_s390_vfeeh,
z_vfee_cc, v128h, v128h, 1>;
defm VFEEF : BinaryExtraVRRbSPair<"vfeef", 0xE780, int_s390_vfeef,
z_vfee_cc, v128f, v128f, 2>;
defm VFEEZB : BinaryVRRbSPair<"vfeezb", 0xE780, int_s390_vfeezb,
z_vfeez_cc, v128b, v128b, 0, 2>;
defm VFEEZH : BinaryVRRbSPair<"vfeezh", 0xE780, int_s390_vfeezh,
z_vfeez_cc, v128h, v128h, 1, 2>;
defm VFEEZF : BinaryVRRbSPair<"vfeezf", 0xE780, int_s390_vfeezf,
z_vfeez_cc, v128f, v128f, 2, 2>;
defm VFENE : BinaryExtraVRRbSPairGeneric<"vfene", 0xE781>;
defm VFENEB : BinaryExtraVRRbSPair<"vfeneb", 0xE781, int_s390_vfeneb,
z_vfene_cc, v128b, v128b, 0>;
defm VFENEH : BinaryExtraVRRbSPair<"vfeneh", 0xE781, int_s390_vfeneh,
z_vfene_cc, v128h, v128h, 1>;
defm VFENEF : BinaryExtraVRRbSPair<"vfenef", 0xE781, int_s390_vfenef,
z_vfene_cc, v128f, v128f, 2>;
defm VFENEZB : BinaryVRRbSPair<"vfenezb", 0xE781, int_s390_vfenezb,
z_vfenez_cc, v128b, v128b, 0, 2>;
defm VFENEZH : BinaryVRRbSPair<"vfenezh", 0xE781, int_s390_vfenezh,
z_vfenez_cc, v128h, v128h, 1, 2>;
defm VFENEZF : BinaryVRRbSPair<"vfenezf", 0xE781, int_s390_vfenezf,
z_vfenez_cc, v128f, v128f, 2, 2>;
defm VISTR : UnaryExtraVRRaSPairGeneric<"vistr", 0xE75C>;
defm VISTRB : UnaryExtraVRRaSPair<"vistrb", 0xE75C, int_s390_vistrb,
z_vistr_cc, v128b, v128b, 0>;
defm VISTRH : UnaryExtraVRRaSPair<"vistrh", 0xE75C, int_s390_vistrh,
z_vistr_cc, v128h, v128h, 1>;
defm VISTRF : UnaryExtraVRRaSPair<"vistrf", 0xE75C, int_s390_vistrf,
z_vistr_cc, v128f, v128f, 2>;
defm VSTRC : QuaternaryOptVRRdSPairGeneric<"vstrc", 0xE78A>;
defm VSTRCB : QuaternaryOptVRRdSPair<"vstrcb", 0xE78A, int_s390_vstrcb,
z_vstrc_cc, v128b, v128b, 0>;
defm VSTRCH : QuaternaryOptVRRdSPair<"vstrch", 0xE78A, int_s390_vstrch,
z_vstrc_cc, v128h, v128h, 1>;
defm VSTRCF : QuaternaryOptVRRdSPair<"vstrcf", 0xE78A, int_s390_vstrcf,
z_vstrc_cc, v128f, v128f, 2>;
defm VSTRCZB : QuaternaryOptVRRdSPair<"vstrczb", 0xE78A, int_s390_vstrczb,
z_vstrcz_cc, v128b, v128b, 0, 2>;
defm VSTRCZH : QuaternaryOptVRRdSPair<"vstrczh", 0xE78A, int_s390_vstrczh,
z_vstrcz_cc, v128h, v128h, 1, 2>;
defm VSTRCZF : QuaternaryOptVRRdSPair<"vstrczf", 0xE78A, int_s390_vstrczf,
z_vstrcz_cc, v128f, v128f, 2, 2>;
}
let Predicates = [FeatureVectorEnhancements2] in {
defm VSTRS : TernaryExtraVRRdGeneric<"vstrs", 0xE78B>;
defm VSTRSB : TernaryExtraVRRd<"vstrsb", 0xE78B,
z_vstrs_cc, v128b, v128b, 0>;
defm VSTRSH : TernaryExtraVRRd<"vstrsh", 0xE78B,
z_vstrs_cc, v128b, v128h, 1>;
defm VSTRSF : TernaryExtraVRRd<"vstrsf", 0xE78B,
z_vstrs_cc, v128b, v128f, 2>;
let Defs = [CC] in {
def VSTRSZB : TernaryVRRd<"vstrszb", 0xE78B,
z_vstrsz_cc, v128b, v128b, 0, 2>;
def VSTRSZH : TernaryVRRd<"vstrszh", 0xE78B,
z_vstrsz_cc, v128b, v128h, 1, 2>;
def VSTRSZF : TernaryVRRd<"vstrszf", 0xE78B,
z_vstrsz_cc, v128b, v128f, 2, 2>;
}
}
//===----------------------------------------------------------------------===//
// Packed-decimal instructions
//===----------------------------------------------------------------------===//
let Predicates = [FeatureVectorPackedDecimal] in {
def VLIP : BinaryVRIh<"vlip", 0xE649>;
def VPKZ : BinaryVSI<"vpkz", 0xE634, null_frag, 0>;
def VUPKZ : StoreLengthVSI<"vupkz", 0xE63C, null_frag, 0>;
let Defs = [CC] in {
let Predicates = [FeatureVectorPackedDecimalEnhancement] in {
def VCVBOpt : TernaryVRRi<"vcvb", 0xE650, GR32>;
def VCVBGOpt : TernaryVRRi<"vcvbg", 0xE652, GR64>;
}
def VCVB : BinaryVRRi<"vcvb", 0xE650, GR32>;
def VCVBG : BinaryVRRi<"vcvbg", 0xE652, GR64>;
def VCVD : TernaryVRIi<"vcvd", 0xE658, GR32>;
def VCVDG : TernaryVRIi<"vcvdg", 0xE65A, GR64>;
def VAP : QuaternaryVRIf<"vap", 0xE671>;
def VSP : QuaternaryVRIf<"vsp", 0xE673>;
def VMP : QuaternaryVRIf<"vmp", 0xE678>;
def VMSP : QuaternaryVRIf<"vmsp", 0xE679>;
def VDP : QuaternaryVRIf<"vdp", 0xE67A>;
def VRP : QuaternaryVRIf<"vrp", 0xE67B>;
def VSDP : QuaternaryVRIf<"vsdp", 0xE67E>;
def VSRP : QuaternaryVRIg<"vsrp", 0xE659>;
def VPSOP : QuaternaryVRIg<"vpsop", 0xE65B>;
def VTP : TestVRRg<"vtp", 0xE65F>;
def VCP : CompareVRRh<"vcp", 0xE677>;
}
}