[ARM] Add a batch of similarly encoded MVE instructions.

[simon_tatham]

Summary

This adds the MVE_qDest_qSrc superclass and all instructions that
inherit from it. It's not the complete class of _everything_ with a
q-register as both destination and source; it's a subset of them that
all have similar encodings (but it would have been hopelessly unwieldy
to call it anything like MVE_111x11100).

This category includes add/sub with carry; long multiplies; halving
multiplies; multiply and accumulate, and some more complex
instructions.

[simon_tatham]

Remastered patch to apply cleanly against current trunk.

[simon_tatham]

Updated to rename instruction classes in line with the general MVE policy. Also, I spotted some instructions lurking in the middle of this section of ARMInstrMVE.td which aren't derived from the MVE_qDest_qSrc base class, so they should have been part of earlier patches. Moved them into D62674 and D62675 where they belong.

[ostannard]

llvm/lib/Target/ARM/AsmParser/ARMAsmParser.cpp

6704 ↗

(On Diff #205782)

I think this needs to check the mnemonic is exactly "vmul", not just startswith, because we emit a literal "vmullt" below.

[simon_tatham]

Addressed review comment about Mnemonic.startswith("vmul"), and also updated this patch to use the existing NEON complex rotation operands (with matching change in the test to expect the new nicer error messages).

[ostannard]

LGTM

llvm/trunk/lib/Target/ARM/ARMInstrMVE.td

 def MVE_VCMPu32r : MVE_VCMPqru<"u32", 0b10>;
 
 def MVE_VCMPs8r  : MVE_VCMPqrs<"s8",  0b00>;
 def MVE_VCMPs16r : MVE_VCMPqrs<"s16", 0b01>;
 def MVE_VCMPs32r : MVE_VCMPqrs<"s32", 0b10>;
 
 // end of MVE compares
 
+// start of MVE_qDest_qSrc
+
+class MVE_qDest_qSrc<string iname, string suffix, dag oops, dag iops,
+                     string ops, vpred_ops vpred, string cstr,
+                     list<dag> pattern=[]>
+  : MVE_p<oops, iops, NoItinerary, iname, suffix,
+          ops, vpred, cstr, pattern> {
+  bits<4> Qd;
+  bits<4> Qm;
+
+  let Inst{25-23} = 0b100;
+  let Inst{22} = Qd{3};
+  let Inst{15-13} = Qd{2-0};
+  let Inst{11-9} = 0b111;
+  let Inst{6} = 0b0;
+  let Inst{5} = Qm{3};
+  let Inst{4} = 0b0;
+  let Inst{3-1} = Qm{2-0};
+}
+
+class MVE_VQxDMLxDH<string iname, bit exch, bit round, bit subtract,
+                    string suffix, bits<2> size, list<dag> pattern=[]>
+  : MVE_qDest_qSrc<iname, suffix, (outs MQPR:$Qd),
+                   (ins MQPR:$Qn, MQPR:$Qm), "$Qd, $Qn, $Qm",
+                   vpred_r, "", pattern> {
+  bits<4> Qn;
+
+  let Inst{28} = subtract;
+  let Inst{21-20} = size;
+  let Inst{19-17} = Qn{2-0};
+  let Inst{16} = 0b0;
+  let Inst{12} = exch;
+  let Inst{8} = 0b0;
+  let Inst{7} = Qn{3};
+  let Inst{0} = round;
+}
+
+multiclass MVE_VQxDMLxDH_multi<string iname, bit exch,
+                               bit round, bit subtract> {
+  def s8  : MVE_VQxDMLxDH<iname, exch, round, subtract, "s8",  0b00>;
+  def s16 : MVE_VQxDMLxDH<iname, exch, round, subtract, "s16", 0b01>;
+  def s32 : MVE_VQxDMLxDH<iname, exch, round, subtract, "s32", 0b10>;
+}
+
+defm MVE_VQDMLADH   : MVE_VQxDMLxDH_multi<"vqdmladh",   0b0, 0b0, 0b0>;
+defm MVE_VQDMLADHX  : MVE_VQxDMLxDH_multi<"vqdmladhx",  0b1, 0b0, 0b0>;
+defm MVE_VQRDMLADH  : MVE_VQxDMLxDH_multi<"vqrdmladh",  0b0, 0b1, 0b0>;
+defm MVE_VQRDMLADHX : MVE_VQxDMLxDH_multi<"vqrdmladhx", 0b1, 0b1, 0b0>;
+defm MVE_VQDMLSDH   : MVE_VQxDMLxDH_multi<"vqdmlsdh",   0b0, 0b0, 0b1>;
+defm MVE_VQDMLSDHX  : MVE_VQxDMLxDH_multi<"vqdmlsdhx",  0b1, 0b0, 0b1>;
+defm MVE_VQRDMLSDH  : MVE_VQxDMLxDH_multi<"vqrdmlsdh",  0b0, 0b1, 0b1>;
+defm MVE_VQRDMLSDHX : MVE_VQxDMLxDH_multi<"vqrdmlsdhx", 0b1, 0b1, 0b1>;
+
+class MVE_VCMUL<string iname, string suffix, bit size, list<dag> pattern=[]>
+  : MVE_qDest_qSrc<iname, suffix, (outs MQPR:$Qd),
+                   (ins MQPR:$Qn, MQPR:$Qm, complexrotateop:$rot),
+                   "$Qd, $Qn, $Qm, $rot", vpred_r, "", pattern> {
+  bits<4> Qn;
+  bits<2> rot;
+
+  let Inst{28} = size;
+  let Inst{21-20} = 0b11;
+  let Inst{19-17} = Qn{2-0};
+  let Inst{16} = 0b0;
+  let Inst{12} = rot{1};
+  let Inst{8} = 0b0;
+  let Inst{7} = Qn{3};
+  let Inst{0} = rot{0};
+
+  let Predicates = [HasMVEFloat];
+}
+
+def MVE_VCMULf16 : MVE_VCMUL<"vcmul", "f16", 0b0>;
+def MVE_VCMULf32 : MVE_VCMUL<"vcmul", "f32", 0b1>;
+
+class MVE_VMULL<string iname, string suffix, bit bit_28, bits<2> bits_21_20,
+                bit T, list<dag> pattern=[]>
+  : MVE_qDest_qSrc<iname, suffix, (outs MQPR:$Qd),
+                   (ins MQPR:$Qn, MQPR:$Qm), "$Qd, $Qn, $Qm",
+                   vpred_r, "", pattern> {
+  bits<4> Qd;
+  bits<4> Qn;
+  bits<4> Qm;
+
+  let Inst{28} = bit_28;
+  let Inst{21-20} = bits_21_20;
+  let Inst{19-17} = Qn{2-0};
+  let Inst{16} = 0b1;
+  let Inst{12} = T;
+  let Inst{8} = 0b0;
+  let Inst{7} = Qn{3};
+  let Inst{0} = 0b0;
+}
+
+multiclass MVE_VMULL_multi<string iname, string suffix,
+                           bit bit_28, bits<2> bits_21_20> {
+  def bh : MVE_VMULL<iname # "b", suffix, bit_28, bits_21_20, 0b0>;
+  def th : MVE_VMULL<iname # "t", suffix, bit_28, bits_21_20, 0b1>;
+}
+
+// For integer multiplies, bits 21:20 encode size, and bit 28 signedness.
+// For polynomial multiplies, bits 21:20 take the unused value 0b11, and
+// bit 28 switches to encoding the size.
+
+defm MVE_VMULLs8  : MVE_VMULL_multi<"vmull", "s8",  0b0, 0b00>;
+defm MVE_VMULLs16 : MVE_VMULL_multi<"vmull", "s16", 0b0, 0b01>;
+defm MVE_VMULLs32 : MVE_VMULL_multi<"vmull", "s32", 0b0, 0b10>;
+defm MVE_VMULLu8  : MVE_VMULL_multi<"vmull", "u8",  0b1, 0b00>;
+defm MVE_VMULLu16 : MVE_VMULL_multi<"vmull", "u16", 0b1, 0b01>;
+defm MVE_VMULLu32 : MVE_VMULL_multi<"vmull", "u32", 0b1, 0b10>;
+defm MVE_VMULLp8  : MVE_VMULL_multi<"vmull", "p8",  0b0, 0b11>;
+defm MVE_VMULLp16 : MVE_VMULL_multi<"vmull", "p16", 0b1, 0b11>;
+
+class MVE_VxMULH<string iname, string suffix, bit U, bits<2> size,
+                 bit round, list<dag> pattern=[]>
+  : MVE_qDest_qSrc<iname, suffix, (outs MQPR:$Qd),
+                   (ins MQPR:$Qn, MQPR:$Qm), "$Qd, $Qn, $Qm",
+                   vpred_r, "", pattern> {
+  bits<4> Qn;
+
+  let Inst{28} = U;
+  let Inst{21-20} = size;
+  let Inst{19-17} = Qn{2-0};
+  let Inst{16} = 0b1;
+  let Inst{12} = round;
+  let Inst{8} = 0b0;
+  let Inst{7} = Qn{3};
+  let Inst{0} = 0b1;
+}
+
+def MVE_VMULHs8   : MVE_VxMULH<"vmulh",  "s8",  0b0, 0b00, 0b0>;
+def MVE_VMULHs16  : MVE_VxMULH<"vmulh",  "s16", 0b0, 0b01, 0b0>;
+def MVE_VMULHs32  : MVE_VxMULH<"vmulh",  "s32", 0b0, 0b10, 0b0>;
+def MVE_VMULHu8   : MVE_VxMULH<"vmulh",  "u8",  0b1, 0b00, 0b0>;
+def MVE_VMULHu16  : MVE_VxMULH<"vmulh",  "u16", 0b1, 0b01, 0b0>;
+def MVE_VMULHu32  : MVE_VxMULH<"vmulh",  "u32", 0b1, 0b10, 0b0>;
+
+def MVE_VRMULHs8  : MVE_VxMULH<"vrmulh", "s8",  0b0, 0b00, 0b1>;
+def MVE_VRMULHs16 : MVE_VxMULH<"vrmulh", "s16", 0b0, 0b01, 0b1>;
+def MVE_VRMULHs32 : MVE_VxMULH<"vrmulh", "s32", 0b0, 0b10, 0b1>;
+def MVE_VRMULHu8  : MVE_VxMULH<"vrmulh", "u8",  0b1, 0b00, 0b1>;
+def MVE_VRMULHu16 : MVE_VxMULH<"vrmulh", "u16", 0b1, 0b01, 0b1>;
+def MVE_VRMULHu32 : MVE_VxMULH<"vrmulh", "u32", 0b1, 0b10, 0b1>;
+
+class MVE_VxMOVxN<string iname, string suffix, bit bit_28, bit bit_17,
+                  bits<2> size, bit T, list<dag> pattern=[]>
+  : MVE_qDest_qSrc<iname, suffix, (outs MQPR:$Qd),
+                   (ins MQPR:$Qd_src, MQPR:$Qm), "$Qd, $Qm",
+                   vpred_n, "$Qd = $Qd_src", pattern> {
+
+  let Inst{28} = bit_28;
+  let Inst{21-20} = 0b11;
+  let Inst{19-18} = size;
+  let Inst{17} = bit_17;
+  let Inst{16} = 0b1;
+  let Inst{12} = T;
+  let Inst{8} = 0b0;
+  let Inst{7} = !if(!eq(bit_17, 0), 1, 0);
+  let Inst{0} = 0b1;
+}
+
+multiclass MVE_VxMOVxN_halves<string iname, string suffix,
+                              bit bit_28, bit bit_17, bits<2> size> {
+  def bh : MVE_VxMOVxN<iname # "b", suffix, bit_28, bit_17, size, 0b0>;
+  def th : MVE_VxMOVxN<iname # "t", suffix, bit_28, bit_17, size, 0b1>;
+}
+
+defm MVE_VMOVNi16   : MVE_VxMOVxN_halves<"vmovn",   "i16", 0b1, 0b0, 0b00>;
+defm MVE_VMOVNi32   : MVE_VxMOVxN_halves<"vmovn",   "i32", 0b1, 0b0, 0b01>;
+defm MVE_VQMOVNs16  : MVE_VxMOVxN_halves<"vqmovn",  "s16", 0b0, 0b1, 0b00>;
+defm MVE_VQMOVNs32  : MVE_VxMOVxN_halves<"vqmovn",  "s32", 0b0, 0b1, 0b01>;
+defm MVE_VQMOVNu16  : MVE_VxMOVxN_halves<"vqmovn",  "u16", 0b1, 0b1, 0b00>;
+defm MVE_VQMOVNu32  : MVE_VxMOVxN_halves<"vqmovn",  "u32", 0b1, 0b1, 0b01>;
+defm MVE_VQMOVUNs16 : MVE_VxMOVxN_halves<"vqmovun", "s16", 0b0, 0b0, 0b00>;
+defm MVE_VQMOVUNs32 : MVE_VxMOVxN_halves<"vqmovun", "s32", 0b0, 0b0, 0b01>;
+
+class MVE_VCVT_ff<string iname, string suffix, bit op, bit T,
+                  list<dag> pattern=[]>
+  : MVE_qDest_qSrc<iname, suffix, (outs MQPR:$Qd), (ins MQPR:$Qd_src, MQPR:$Qm),
+                   "$Qd, $Qm", vpred_n, "$Qd = $Qd_src", pattern> {
+  let Inst{28} = op;
+  let Inst{21-16} = 0b111111;
+  let Inst{12} = T;
+  let Inst{8-7} = 0b00;
+  let Inst{0} = 0b1;
+
+  let Predicates = [HasMVEFloat];
+}
+
+multiclass MVE_VCVT_ff_halves<string suffix, bit op> {
+  def bh : MVE_VCVT_ff<"vcvtb", suffix, op, 0b0>;
+  def th : MVE_VCVT_ff<"vcvtt", suffix, op, 0b1>;
+}
+
+defm MVE_VCVTf16f32 : MVE_VCVT_ff_halves<"f16.f32", 0b0>;
+defm MVE_VCVTf32f16 : MVE_VCVT_ff_halves<"f32.f16", 0b1>;
+
+class MVE_VxCADD<string iname, string suffix, bits<2> size, bit halve,
+                 list<dag> pattern=[]>
+  : MVE_qDest_qSrc<iname, suffix, (outs MQPR:$Qd),
+                   (ins MQPR:$Qn, MQPR:$Qm, complexrotateopodd:$rot),
+                   "$Qd, $Qn, $Qm, $rot", vpred_r, "",
+          pattern> {
+  bits<4> Qn;
+  bit rot;
+
+  let Inst{28} = halve;
+  let Inst{21-20} = size;
+  let Inst{19-17} = Qn{2-0};
+  let Inst{16} = 0b0;
+  let Inst{12} = rot;
+  let Inst{8} = 0b1;
+  let Inst{7} = Qn{3};
+  let Inst{0} = 0b0;
+}
+
+def MVE_VCADDi8   : MVE_VxCADD<"vcadd", "i8", 0b00, 0b1>;
+def MVE_VCADDi16  : MVE_VxCADD<"vcadd", "i16", 0b01, 0b1>;
+def MVE_VCADDi32  : MVE_VxCADD<"vcadd", "i32", 0b10, 0b1>;
+
+def MVE_VHCADDs8  : MVE_VxCADD<"vhcadd", "s8", 0b00, 0b0>;
+def MVE_VHCADDs16 : MVE_VxCADD<"vhcadd", "s16", 0b01, 0b0>;
+def MVE_VHCADDs32 : MVE_VxCADD<"vhcadd", "s32", 0b10, 0b0>;
+
+class MVE_VADCSBC<string iname, bit I, bit subtract,
+                  dag carryin, list<dag> pattern=[]>
+  : MVE_qDest_qSrc<iname, "i32", (outs MQPR:$Qd, cl_FPSCR_NZCV:$carryout),
+                   !con((ins MQPR:$Qn, MQPR:$Qm), carryin),
+                   "$Qd, $Qn, $Qm", vpred_r, "", pattern> {
+  bits<4> Qn;
+
+  let Inst{28} = subtract;
+  let Inst{21-20} = 0b11;
+  let Inst{19-17} = Qn{2-0};
+  let Inst{16} = 0b0;
+  let Inst{12} = I;
+  let Inst{8} = 0b1;
+  let Inst{7} = Qn{3};
+  let Inst{0} = 0b0;
+
+  // Custom decoder method in order to add the FPSCR operand(s), which
+  // Tablegen won't do right
+  let DecoderMethod = "DecodeMVEVADCInstruction";
+}
+
+def MVE_VADC  : MVE_VADCSBC<"vadc",  0b0, 0b0, (ins cl_FPSCR_NZCV:$carryin)>;
+def MVE_VADCI : MVE_VADCSBC<"vadci", 0b1, 0b0, (ins)>;
+
+def MVE_VSBC  : MVE_VADCSBC<"vsbc",  0b0, 0b1, (ins cl_FPSCR_NZCV:$carryin)>;
+def MVE_VSBCI : MVE_VADCSBC<"vsbci", 0b1, 0b1, (ins)>;
+
+class MVE_VQDMULL<string iname, string suffix, bit size, bit T,
+                  list<dag> pattern=[]>
+  : MVE_qDest_qSrc<iname, suffix, (outs MQPR:$Qd),
+                   (ins MQPR:$Qn, MQPR:$Qm), "$Qd, $Qn, $Qm",
+                   vpred_r, "", pattern> {
+  bits<4> Qn;
+
+  let Inst{28} = size;
+  let Inst{21-20} = 0b11;
+  let Inst{19-17} = Qn{2-0};
+  let Inst{16} = 0b0;
+  let Inst{12} = T;
+  let Inst{8} = 0b1;
+  let Inst{7} = Qn{3};
+  let Inst{0} = 0b1;
+}
+
+multiclass MVE_VQDMULL_halves<string suffix, bit size> {
+  def bh : MVE_VQDMULL<"vqdmullb", suffix, size, 0b0>;
+  def th : MVE_VQDMULL<"vqdmullt", suffix, size, 0b1>;
+}
+
+defm MVE_VQDMULLs16 : MVE_VQDMULL_halves<"s16", 0b0>;
+defm MVE_VQDMULLs32 : MVE_VQDMULL_halves<"s32", 0b1>;
+
+// end of mve_qDest_qSrc
+
 class MVE_VPT<string suffix, bits<2> size, dag iops, string asm, list<dag> pattern=[]>
   : MVE_MI<(outs ), iops, NoItinerary, !strconcat("vpt", "${Mk}", ".", suffix), asm, "", pattern> {
   bits<3> fc;
   bits<4> Mk;
   bits<3> Qn;
 
   let Inst{31-23} = 0b111111100;
   let Inst{22} = Mk{3};

llvm/trunk/lib/Target/ARM/AsmParser/ARMAsmParser.cpp

if (Mnemonic != "adcs" && Mnemonic != "bics" && Mnemonic != "movs" &&
       Mnemonic != "sbcs" && Mnemonic != "rscs" &&
       !(hasMVE() &&
         (Mnemonic == "vmine" ||
          Mnemonic == "vshle" || Mnemonic == "vshlt" || Mnemonic == "vshllt" ||
          Mnemonic == "vmvne" || Mnemonic == "vorne" ||
          Mnemonic == "vnege" || Mnemonic == "vnegt" ||
          Mnemonic == "vmule" || Mnemonic == "vmult" ||
          Mnemonic == "vrintne" ||
+         Mnemonic == "vcmult" || Mnemonic == "vcmule" ||
          Mnemonic.startswith("vq")))) {
     unsigned CC = ARMCondCodeFromString(Mnemonic.substr(Mnemonic.size()-2));
     if (CC != ~0U) {
       Mnemonic = Mnemonic.slice(0, Mnemonic.size() - 2);
       PredicationCode = CC;
     }
   }
 
if (IMod != ~0U) {
       Mnemonic = Mnemonic.slice(0, Mnemonic.size()-2);
       ProcessorIMod = IMod;
     }
   }
 
   if (isMnemonicVPTPredicable(Mnemonic, ExtraToken) && Mnemonic != "vmovlt" &&
       Mnemonic != "vshllt" && Mnemonic != "vrshrnt" && Mnemonic != "vshrnt" &&
       Mnemonic != "vqrshrunt" && Mnemonic != "vqshrunt" &&
-      Mnemonic != "vqrshrnt" && Mnemonic != "vqshrnt" && Mnemonic != "vcvt") {
+      Mnemonic != "vqrshrnt" && Mnemonic != "vqshrnt" && Mnemonic != "vmullt" &&
+      Mnemonic != "vqmovnt" && Mnemonic != "vqmovunt" &&
+      Mnemonic != "vqmovnt" && Mnemonic != "vmovnt" && Mnemonic != "vqdmullt" &&
+      Mnemonic != "vcvtt" && Mnemonic != "vcvt") {
     unsigned CC = ARMVectorCondCodeFromString(Mnemonic.substr(Mnemonic.size()-1));
     if (CC != ~0U) {
       Mnemonic = Mnemonic.slice(0, Mnemonic.size()-1);
       VPTPredicationCode = CC;
     }
     return Mnemonic;
   }
 
if (!shouldOmitVectorPredicateOperand(Mnemonic, Operands) &&
       Operands.erase(Operands.begin());
       SMLoc MLoc = SMLoc::getFromPointer(NameLoc.getPointer());
       SMLoc PLoc = SMLoc::getFromPointer(NameLoc.getPointer() +
                                          Mnemonic.size() - 1 + CarrySetting);
       Operands.insert(Operands.begin(),
                       ARMOperand::CreateVPTPred(ARMVCC::Else, PLoc));
       Operands.insert(Operands.begin(),
                       ARMOperand::CreateToken(StringRef("vcvtn"), MLoc));
+    } else if (Mnemonic == "vmul" && PredicationCode == ARMCC::LT &&
+               !shouldOmitVectorPredicateOperand(Mnemonic, Operands)) {
+      // Another hack, this time to distinguish between scalar predicated vmul
+      // with 'lt' predication code and the vector instruction vmullt with
+      // vector predication code "none"
+      Operands.erase(Operands.begin() + 1);
+      Operands.erase(Operands.begin());
+      SMLoc MLoc = SMLoc::getFromPointer(NameLoc.getPointer());
+      Operands.insert(Operands.begin(),
+                      ARMOperand::CreateToken(StringRef("vmullt"), MLoc));
     }
     // For vmov and vcmp, as mentioned earlier, we did not add the vector
     // predication code, since these may contain operands that require
     // special parsing.  So now we have to see if they require vector
     // predication and replace the scalar one with the vector predication
     // operand if that is the case.
     else if (Mnemonic == "vmov" || Mnemonic.startswith("vcmp") ||
              (Mnemonic.startswith("vcvt") && !Mnemonic.startswith("vcvta") &&
bool ARMAsmParser::validateInstruction(MCInst &Inst,
   case ARM::VSTMDIA: {
     ARMOperand &Op = static_cast<ARMOperand&>(*Operands[3]);
     auto &RegList = Op.getRegList();
     if (RegList.size() < 1 || RegList.size() > 16)
       return Error(Operands[3]->getStartLoc(),
                    "list of registers must be at least 1 and at most 16");
     break;
   }
+  case ARM::MVE_VQDMULLs32bh:
+  case ARM::MVE_VQDMULLs32th:
+  case ARM::MVE_VCMULf32:
+  case ARM::MVE_VMULLs32bh:
+  case ARM::MVE_VMULLs32th:
+  case ARM::MVE_VMULLu32bh:
+  case ARM::MVE_VMULLu32th:
+  case ARM::MVE_VQDMLADHs32:
+  case ARM::MVE_VQDMLADHXs32:
+  case ARM::MVE_VQRDMLADHs32:
+  case ARM::MVE_VQRDMLADHXs32:
+  case ARM::MVE_VQDMLSDHs32:
+  case ARM::MVE_VQDMLSDHXs32:
+  case ARM::MVE_VQRDMLSDHs32:
+  case ARM::MVE_VQRDMLSDHXs32: {
+    if (Operands[3]->getReg() == Operands[4]->getReg()) {
+      return Error (Operands[3]->getStartLoc(),
+                    "Qd register and Qn register can't be identical");
+    }
+    if (Operands[3]->getReg() == Operands[5]->getReg()) {
+      return Error (Operands[3]->getStartLoc(),
+                    "Qd register and Qm register can't be identical");
+    }
+    break;
+  }
   }
 
   return false;
 }
 
 static unsigned getRealVSTOpcode(unsigned Opc, unsigned &Spacing) {
   switch(Opc) {
   default: llvm_unreachable("unexpected opcode!");

llvm/trunk/lib/Target/ARM/Disassembler/ARMDisassembler.cpp

 static DecodeStatus DecodeVLD3DupInstruction(MCInst &Inst, unsigned Val,
                                uint64_t Address, const void *Decoder);
 static DecodeStatus DecodeVLD4DupInstruction(MCInst &Inst, unsigned Val,
                                uint64_t Address, const void *Decoder);
 static DecodeStatus DecodeNEONModImmInstruction(MCInst &Inst,unsigned Val,
                                uint64_t Address, const void *Decoder);
 static DecodeStatus DecodeMVEModImmInstruction(MCInst &Inst,unsigned Val,
                                uint64_t Address, const void *Decoder);
+static DecodeStatus DecodeMVEVADCInstruction(MCInst &Inst, unsigned Insn,
+                               uint64_t Address, const void *Decoder);
 static DecodeStatus DecodeVSHLMaxInstruction(MCInst &Inst, unsigned Val,
                                uint64_t Address, const void *Decoder);
 static DecodeStatus DecodeShiftRight8Imm(MCInst &Inst, unsigned Val,
                                uint64_t Address, const void *Decoder);
 static DecodeStatus DecodeShiftRight16Imm(MCInst &Inst, unsigned Val,
                                uint64_t Address, const void *Decoder);
 static DecodeStatus DecodeShiftRight32Imm(MCInst &Inst, unsigned Val,
                                uint64_t Address, const void *Decoder);
DecodeMVEModImmInstruction(MCInst &Inst, unsigned Insn,
 
   Inst.addOperand(MCOperand::createImm(ARMVCC::None));
   Inst.addOperand(MCOperand::createReg(0));
   Inst.addOperand(MCOperand::createImm(0));
 
   return S;
 }
 
+static DecodeStatus DecodeMVEVADCInstruction(MCInst &Inst, unsigned Insn,
+                               uint64_t Address, const void *Decoder) {
+  DecodeStatus S = MCDisassembler::Success;
+
+  unsigned Qd = fieldFromInstruction(Insn, 13, 3);
+  Qd |= fieldFromInstruction(Insn, 22, 1) << 3;
+  if (!Check(S, DecodeMQPRRegisterClass(Inst, Qd, Address, Decoder)))
+    return MCDisassembler::Fail;
+  Inst.addOperand(MCOperand::createReg(ARM::FPSCR_NZCV));
+
+  unsigned Qn = fieldFromInstruction(Insn, 17, 3);
+  Qn |= fieldFromInstruction(Insn, 7, 1) << 3;
+  if (!Check(S, DecodeMQPRRegisterClass(Inst, Qn, Address, Decoder)))
+    return MCDisassembler::Fail;
+  unsigned Qm = fieldFromInstruction(Insn, 1, 3);
+  Qm |= fieldFromInstruction(Insn, 5, 1) << 3;
+  if (!Check(S, DecodeMQPRRegisterClass(Inst, Qm, Address, Decoder)))
+    return MCDisassembler::Fail;
+  if (!fieldFromInstruction(Insn, 12, 1)) // I bit clear => need input FPSCR
+    Inst.addOperand(MCOperand::createReg(ARM::FPSCR_NZCV));
+  Inst.addOperand(MCOperand::createImm(Qd));
+
+  return S;
+}
+
 static DecodeStatus DecodeVSHLMaxInstruction(MCInst &Inst, unsigned Insn,
                                         uint64_t Address, const void *Decoder) {
   DecodeStatus S = MCDisassembler::Success;
 
   unsigned Rd = fieldFromInstruction(Insn, 12, 4);
   Rd |= fieldFromInstruction(Insn, 22, 1) << 4;
   unsigned Rm = fieldFromInstruction(Insn, 0, 4);
   Rm |= fieldFromInstruction(Insn, 5, 1) << 4;

llvm/trunk/test/MC/ARM/mve-qdest-qsrc.s

+# RUN: not llvm-mc -triple=thumbv8.1m.main-none-eabi -mattr=+mve -show-encoding  < %s \
+# RUN:   | FileCheck --check-prefix=CHECK-NOFP %s
+# RUN: not llvm-mc -triple=thumbv8.1m.main-none-eabi -mattr=+mve.fp,+fp64 -show-encoding  < %s 2>%t \
+# RUN:   | FileCheck --check-prefix=CHECK %s
+# RUN:     FileCheck --check-prefix=ERROR < %t %s
+
+# CHECK: vcvtb.f16.f32 q1, q4  @ encoding: [0x3f,0xee,0x09,0x2e]
+# CHECK-NOFP-NOT: vcvtb.f16.f32 q1, q4  @ encoding: [0x3f,0xee,0x09,0x2e]
+vcvtb.f16.f32 q1, q4
+
+# CHECK: vcvtt.f32.f16 q0, q1  @ encoding: [0x3f,0xfe,0x03,0x1e]
+# CHECK-NOFP-NOT: vcvtt.f32.f16 q0, q1  @ encoding: [0x3f,0xfe,0x03,0x1e]
+vcvtt.f32.f16 q0, q1
+
+# CHECK: vcvtt.f64.f16 d0, s0 @ encoding: [0xb2,0xee,0xc0,0x0b]
+# CHECK-NOFP-NOT: vcvtt.f64.f16 d0, s0 @ encoding: [0xb2,0xee,0xc0,0x0b]
+vcvtt.f64.f16 d0, s0
+
+# CHECK: vcvtt.f16.f64 s1, d2 @ encoding: [0xf3,0xee,0xc2,0x0b]
+# CHECK-NOFP-NOT: vcvtt.f16.f64 s1, d2 @ encoding: [0xf3,0xee,0xc2,0x0b]
+vcvtt.f16.f64 s1, d2
+
+# CHECK: vcvtt.f16.f32 q1, q4 @ encoding: [0x3f,0xee,0x09,0x3e]
+# CHECK-NOFP-NOT: vcvtt.f16.f32 q1, q4 @ encoding: [0x3f,0xee,0x09,0x3e]
+vcvtt.f16.f32 q1, q4
+
+# CHECK: vqdmladhx.s8 q1, q6, q6  @ encoding: [0x0c,0xee,0x0c,0x3e]
+# CHECK-NOFP: vqdmladhx.s8 q1, q6, q6  @ encoding: [0x0c,0xee,0x0c,0x3e]
+vqdmladhx.s8 q1, q6, q6
+
+# CHECK: vqdmladhx.s16 q0, q1, q4  @ encoding: [0x12,0xee,0x08,0x1e]
+# CHECK-NOFP: vqdmladhx.s16 q0, q1, q4  @ encoding: [0x12,0xee,0x08,0x1e]
+vqdmladhx.s16 q0, q1, q4
+
+# CHECK: vqdmladhx.s32 q0, q3, q7  @ encoding: [0x26,0xee,0x0e,0x1e]
+# CHECK-NOFP: vqdmladhx.s32 q0, q3, q7  @ encoding: [0x26,0xee,0x0e,0x1e]
+vqdmladhx.s32 q0, q3, q7
+
+# CHECK: vqdmladh.s8 q0, q1, q1  @ encoding: [0x02,0xee,0x02,0x0e]
+# CHECK-NOFP: vqdmladh.s8 q0, q1, q1  @ encoding: [0x02,0xee,0x02,0x0e]
+vqdmladh.s8 q0, q1, q1
+
+# CHECK: vqdmladh.s16 q0, q2, q2  @ encoding: [0x14,0xee,0x04,0x0e]
+# CHECK-NOFP: vqdmladh.s16 q0, q2, q2  @ encoding: [0x14,0xee,0x04,0x0e]
+vqdmladh.s16 q0, q2, q2
+
+# CHECK: vqdmladh.s32 q1, q5, q7  @ encoding: [0x2a,0xee,0x0e,0x2e]
+# CHECK-NOFP: vqdmladh.s32 q1, q5, q7  @ encoding: [0x2a,0xee,0x0e,0x2e]
+vqdmladh.s32 q1, q5, q7
+
+# CHECK: vqrdmladhx.s8 q0, q7, q0  @ encoding: [0x0e,0xee,0x01,0x1e]
+# CHECK-NOFP: vqrdmladhx.s8 q0, q7, q0  @ encoding: [0x0e,0xee,0x01,0x1e]
+vqrdmladhx.s8 q0, q7, q0
+
+# CHECK: vqrdmladhx.s16 q0, q0, q1  @ encoding: [0x10,0xee,0x03,0x1e]
+# CHECK-NOFP: vqrdmladhx.s16 q0, q0, q1  @ encoding: [0x10,0xee,0x03,0x1e]
+vqrdmladhx.s16 q0, q0, q1
+
+# CHECK: vqrdmladhx.s32 q1, q0, q4  @ encoding: [0x20,0xee,0x09,0x3e]
+# CHECK-NOFP: vqrdmladhx.s32 q1, q0, q4  @ encoding: [0x20,0xee,0x09,0x3e]
+vqrdmladhx.s32 q1, q0, q4
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: Qd register and Qn register can't be identical
+vqrdmladhx.s32 q1, q1, q0
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: Qd register and Qm register can't be identical
+vqrdmladhx.s32 q1, q0, q1
+
+# CHECK: vqrdmladh.s8 q0, q6, q2  @ encoding: [0x0c,0xee,0x05,0x0e]
+# CHECK-NOFP: vqrdmladh.s8 q0, q6, q2  @ encoding: [0x0c,0xee,0x05,0x0e]
+vqrdmladh.s8 q0, q6, q2
+
+# CHECK: vqrdmladh.s16 q1, q5, q4  @ encoding: [0x1a,0xee,0x09,0x2e]
+# CHECK-NOFP: vqrdmladh.s16 q1, q5, q4  @ encoding: [0x1a,0xee,0x09,0x2e]
+vqrdmladh.s16 q1, q5, q4
+
+# CHECK: vqrdmladh.s32 q0, q2, q2  @ encoding: [0x24,0xee,0x05,0x0e]
+# CHECK-NOFP: vqrdmladh.s32 q0, q2, q2  @ encoding: [0x24,0xee,0x05,0x0e]
+vqrdmladh.s32 q0, q2, q2
+
+# CHECK: vqdmlsdhx.s8 q1, q4, q7  @ encoding: [0x08,0xfe,0x0e,0x3e]
+# CHECK-NOFP: vqdmlsdhx.s8 q1, q4, q7  @ encoding: [0x08,0xfe,0x0e,0x3e]
+vqdmlsdhx.s8 q1, q4, q7
+
+# CHECK: vqdmlsdhx.s16 q0, q2, q5  @ encoding: [0x14,0xfe,0x0a,0x1e]
+# CHECK-NOFP: vqdmlsdhx.s16 q0, q2, q5  @ encoding: [0x14,0xfe,0x0a,0x1e]
+vqdmlsdhx.s16 q0, q2, q5
+
+# CHECK: vqdmlsdhx.s32 q3, q4, q6  @ encoding: [0x28,0xfe,0x0c,0x7e]
+# CHECK-NOFP: vqdmlsdhx.s32 q3, q4, q6  @ encoding: [0x28,0xfe,0x0c,0x7e]
+vqdmlsdhx.s32 q3, q4, q6
+
+# CHECK: vqdmlsdh.s8 q0, q3, q6  @ encoding: [0x06,0xfe,0x0c,0x0e]
+# CHECK-NOFP: vqdmlsdh.s8 q0, q3, q6  @ encoding: [0x06,0xfe,0x0c,0x0e]
+vqdmlsdh.s8 q0, q3, q6
+
+# CHECK: vqdmlsdh.s16 q0, q4, q1  @ encoding: [0x18,0xfe,0x02,0x0e]
+# CHECK-NOFP: vqdmlsdh.s16 q0, q4, q1  @ encoding: [0x18,0xfe,0x02,0x0e]
+vqdmlsdh.s16 q0, q4, q1
+
+# CHECK: vqdmlsdh.s32 q2, q5, q0  @ encoding: [0x2a,0xfe,0x00,0x4e]
+# CHECK-NOFP: vqdmlsdh.s32 q2, q5, q0  @ encoding: [0x2a,0xfe,0x00,0x4e]
+vqdmlsdh.s32 q2, q5, q0
+
+# CHECK: vqrdmlsdhx.s8 q0, q3, q1  @ encoding: [0x06,0xfe,0x03,0x1e]
+# CHECK-NOFP: vqrdmlsdhx.s8 q0, q3, q1  @ encoding: [0x06,0xfe,0x03,0x1e]
+vqrdmlsdhx.s8 q0, q3, q1
+
+# CHECK: vqrdmlsdhx.s16 q0, q1, q4  @ encoding: [0x12,0xfe,0x09,0x1e]
+# CHECK-NOFP: vqrdmlsdhx.s16 q0, q1, q4  @ encoding: [0x12,0xfe,0x09,0x1e]
+vqrdmlsdhx.s16 q0, q1, q4
+
+# CHECK: vqrdmlsdhx.s32 q1, q6, q3  @ encoding: [0x2c,0xfe,0x07,0x3e]
+# CHECK-NOFP: vqrdmlsdhx.s32 q1, q6, q3  @ encoding: [0x2c,0xfe,0x07,0x3e]
+vqrdmlsdhx.s32 q1, q6, q3
+
+# CHECK: vqrdmlsdh.s8 q3, q3, q0  @ encoding: [0x06,0xfe,0x01,0x6e]
+# CHECK-NOFP: vqrdmlsdh.s8 q3, q3, q0  @ encoding: [0x06,0xfe,0x01,0x6e]
+vqrdmlsdh.s8 q3, q3, q0
+
+# CHECK: vqrdmlsdh.s16 q0, q7, q4  @ encoding: [0x1e,0xfe,0x09,0x0e]
+# CHECK-NOFP: vqrdmlsdh.s16 q0, q7, q4  @ encoding: [0x1e,0xfe,0x09,0x0e]
+vqrdmlsdh.s16 q0, q7, q4
+
+# CHECK: vqrdmlsdh.s32 q0, q6, q7  @ encoding: [0x2c,0xfe,0x0f,0x0e]
+# CHECK-NOFP: vqrdmlsdh.s32 q0, q6, q7  @ encoding: [0x2c,0xfe,0x0f,0x0e]
+vqrdmlsdh.s32 q0, q6, q7
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: Qd register and Qn register can't be identical
+vqrdmlsdh.s32 q0, q0, q7
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: Qd register and Qm register can't be identical
+vqrdmlsdh.s32 q0, q6, q0
+
+# CHECK: vcmul.f16 q0, q1, q2, #90 @ encoding: [0x32,0xee,0x05,0x0e]
+# CHECK-NOFP-NOT: vcmul.f16 q0, q1, q2, #90 @ encoding: [0x32,0xee,0x05,0x0e]
+vcmul.f16 q0, q1, q2, #90
+
+# CHECK: vcmul.f16 q6, q2, q5, #0  @ encoding: [0x34,0xee,0x0a,0xce]
+# CHECK-NOFP-NOT: vcmul.f16 q6, q2, q5, #0  @ encoding: [0x34,0xee,0x0a,0xce]
+vcmul.f16 q6, q2, q5, #0
+
+# CHECK: vcmul.f16 q1, q0, q5, #90  @ encoding: [0x30,0xee,0x0b,0x2e]
+# CHECK-NOFP-NOT: vcmul.f16 q1, q0, q5, #90  @ encoding: [0x30,0xee,0x0b,0x2e]
+vcmul.f16 q1, q0, q5, #90
+
+# CHECK: vcmul.f16 q1, q0, q5, #180  @ encoding: [0x30,0xee,0x0a,0x3e]
+# CHECK-NOFP-NOT: vcmul.f16 q1, q0, q5, #180  @ encoding: [0x30,0xee,0x0a,0x3e]
+vcmul.f16 q1, q0, q5, #180
+
+# CHECK: vcmul.f16 q1, q0, q5, #270  @ encoding: [0x30,0xee,0x0b,0x3e]
+# CHECK-NOFP-NOT: vcmul.f16 q1, q0, q5, #270  @ encoding: [0x30,0xee,0x0b,0x3e]
+vcmul.f16 q1, q0, q5, #270
+
+# CHECK: vcmul.f16 q1, q0, q1, #270  @ encoding: [0x30,0xee,0x03,0x3e]
+# CHECK-NOFP-NOT: vcmul.f16 q1, q0, q1, #270  @ encoding: [0x30,0xee,0x03,0x3e]
+vcmul.f16 q1, q0, q1, #270
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: complex rotation must be 0, 90, 180 or 270
+vcmul.f16 q1, q0, q5, #300
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: Qd register and Qn register can't be identical
+vcmul.f32 q1, q1, q5, #0
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: Qd register and Qm register can't be identical
+vcmul.f32 q1, q5, q1, #0
+
+# CHECK: vcmul.f32 q1, q7, q5, #0  @ encoding: [0x3e,0xfe,0x0a,0x2e]
+# CHECK-NOFP-NOT: vcmul.f32 q1, q7, q5, #0  @ encoding: [0x3e,0xfe,0x0a,0x2e]
+vcmul.f32 q1, q7, q5, #0
+
+# CHECK: vcmul.f32 q3, q4, q2, #90  @ encoding: [0x38,0xfe,0x05,0x6e]
+# CHECK-NOFP-NOT: vcmul.f32 q3, q4, q2, #90  @ encoding: [0x38,0xfe,0x05,0x6e]
+vcmul.f32 q3, q4, q2, #90
+
+# CHECK: vcmul.f32 q5, q1, q3, #180  @ encoding: [0x32,0xfe,0x06,0xbe]
+# CHECK-NOFP-NOT: vcmul.f32 q5, q1, q3, #180  @ encoding: [0x32,0xfe,0x06,0xbe]
+vcmul.f32 q5, q1, q3, #180
+
+# CHECK: vcmul.f32 q0, q7, q4, #270  @ encoding: [0x3e,0xfe,0x09,0x1e]
+# CHECK-NOFP-NOT: vcmul.f32 q0, q7, q4, #270  @ encoding: [0x3e,0xfe,0x09,0x1e]
+vcmul.f32 q0, q7, q4, #270
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: complex rotation must be 0, 90, 180 or 270
+vcmul.f32 q1, q0, q5, #300
+
+# CHECK: vmullb.s8 q2, q6, q0  @ encoding: [0x0d,0xee,0x00,0x4e]
+# CHECK-NOFP: vmullb.s8 q2, q6, q0  @ encoding: [0x0d,0xee,0x00,0x4e]
+vmullb.s8 q2, q6, q0
+
+# CHECK: vmullb.s16 q3, q4, q3  @ encoding: [0x19,0xee,0x06,0x6e]
+# CHECK-NOFP: vmullb.s16 q3, q4, q3  @ encoding: [0x19,0xee,0x06,0x6e]
+vmullb.s16 q3, q4, q3
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: Qd register and Qm register can't be identical
+vmullb.s32 q3, q4, q3
+
+# CHECK: vmullb.s32 q3, q5, q6  @ encoding: [0x2b,0xee,0x0c,0x6e]
+# CHECK-NOFP: vmullb.s32 q3, q5, q6  @ encoding: [0x2b,0xee,0x0c,0x6e]
+vmullb.s32 q3, q5, q6
+
+# CHECK: vmullt.s8 q0, q6, q2  @ encoding: [0x0d,0xee,0x04,0x1e]
+# CHECK-NOFP: vmullt.s8 q0, q6, q2  @ encoding: [0x0d,0xee,0x04,0x1e]
+vmullt.s8 q0, q6, q2
+
+# CHECK: vmullt.s16 q0, q0, q2  @ encoding: [0x11,0xee,0x04,0x1e]
+# CHECK-NOFP: vmullt.s16 q0, q0, q2  @ encoding: [0x11,0xee,0x04,0x1e]
+vmullt.s16 q0, q0, q2
+
+# CHECK: vmullt.s32 q2, q4, q4  @ encoding: [0x29,0xee,0x08,0x5e]
+# CHECK-NOFP: vmullt.s32 q2, q4, q4  @ encoding: [0x29,0xee,0x08,0x5e]
+vmullt.s32 q2, q4, q4
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: Qd register and Qn register can't be identical
+vmullt.s32 q4, q4, q2
+
+# CHECK: vmullb.p8 q2, q3, q7  @ encoding: [0x37,0xee,0x0e,0x4e]
+# CHECK-NOFP: vmullb.p8 q2, q3, q7  @ encoding: [0x37,0xee,0x0e,0x4e]
+vmullb.p8 q2, q3, q7
+
+# CHECK: vmullb.p16 q0, q1, q3  @ encoding: [0x33,0xfe,0x06,0x0e]
+# CHECK-NOFP: vmullb.p16 q0, q1, q3  @ encoding: [0x33,0xfe,0x06,0x0e]
+vmullb.p16 q0, q1, q3
+
+# CHECK: vmullt.p8 q1, q1, q7  @ encoding: [0x33,0xee,0x0e,0x3e]
+# CHECK-NOFP: vmullt.p8 q1, q1, q7  @ encoding: [0x33,0xee,0x0e,0x3e]
+vmullt.p8 q1, q1, q7
+
+# CHECK: vmullt.p16 q0, q7, q7  @ encoding: [0x3f,0xfe,0x0e,0x1e]
+# CHECK-NOFP: vmullt.p16 q0, q7, q7  @ encoding: [0x3f,0xfe,0x0e,0x1e]
+vmullt.p16 q0, q7, q7
+
+# CHECK: vmulh.s8 q0, q4, q5  @ encoding: [0x09,0xee,0x0b,0x0e]
+# CHECK-NOFP: vmulh.s8 q0, q4, q5  @ encoding: [0x09,0xee,0x0b,0x0e]
+vmulh.s8 q0, q4, q5
+
+# CHECK: vmulh.s16 q0, q7, q4  @ encoding: [0x1f,0xee,0x09,0x0e]
+# CHECK-NOFP: vmulh.s16 q0, q7, q4  @ encoding: [0x1f,0xee,0x09,0x0e]
+vmulh.s16 q0, q7, q4
+
+# CHECK: vmulh.s32 q0, q7, q4  @ encoding: [0x2f,0xee,0x09,0x0e]
+# CHECK-NOFP: vmulh.s32 q0, q7, q4  @ encoding: [0x2f,0xee,0x09,0x0e]
+vmulh.s32 q0, q7, q4
+
+# CHECK: vmulh.u8 q3, q5, q2  @ encoding: [0x0b,0xfe,0x05,0x6e]
+# CHECK-NOFP: vmulh.u8 q3, q5, q2  @ encoding: [0x0b,0xfe,0x05,0x6e]
+vmulh.u8 q3, q5, q2
+
+# CHECK: vmulh.u16 q2, q7, q4  @ encoding: [0x1f,0xfe,0x09,0x4e]
+# CHECK-NOFP: vmulh.u16 q2, q7, q4  @ encoding: [0x1f,0xfe,0x09,0x4e]
+vmulh.u16 q2, q7, q4
+
+# CHECK: vmulh.u32 q1, q3, q2  @ encoding: [0x27,0xfe,0x05,0x2e]
+# CHECK-NOFP: vmulh.u32 q1, q3, q2  @ encoding: [0x27,0xfe,0x05,0x2e]
+vmulh.u32 q1, q3, q2
+
+# CHECK: vrmulh.s8 q1, q1, q2  @ encoding: [0x03,0xee,0x05,0x3e]
+# CHECK-NOFP: vrmulh.s8 q1, q1, q2  @ encoding: [0x03,0xee,0x05,0x3e]
+vrmulh.s8 q1, q1, q2
+
+# CHECK: vrmulh.s16 q1, q1, q2  @ encoding: [0x13,0xee,0x05,0x3e]
+# CHECK-NOFP: vrmulh.s16 q1, q1, q2  @ encoding: [0x13,0xee,0x05,0x3e]
+vrmulh.s16 q1, q1, q2
+
+# CHECK: vrmulh.s32 q3, q1, q0  @ encoding: [0x23,0xee,0x01,0x7e]
+# CHECK-NOFP: vrmulh.s32 q3, q1, q0  @ encoding: [0x23,0xee,0x01,0x7e]
+vrmulh.s32 q3, q1, q0
+
+# CHECK: vrmulh.u8 q1, q6, q0  @ encoding: [0x0d,0xfe,0x01,0x3e]
+# CHECK-NOFP: vrmulh.u8 q1, q6, q0  @ encoding: [0x0d,0xfe,0x01,0x3e]
+vrmulh.u8 q1, q6, q0
+
+# CHECK: vrmulh.u16 q4, q3, q6  @ encoding: [0x17,0xfe,0x0d,0x9e]
+# CHECK-NOFP: vrmulh.u16 q4, q3, q6  @ encoding: [0x17,0xfe,0x0d,0x9e]
+vrmulh.u16 q4, q3, q6
+
+# CHECK: vrmulh.u32 q1, q2, q2  @ encoding: [0x25,0xfe,0x05,0x3e]
+# CHECK-NOFP: vrmulh.u32 q1, q2, q2  @ encoding: [0x25,0xfe,0x05,0x3e]
+vrmulh.u32 q1, q2, q2
+
+# CHECK: vqmovnb.s16 q0, q1  @ encoding: [0x33,0xee,0x03,0x0e]
+# CHECK-NOFP: vqmovnb.s16 q0, q1  @ encoding: [0x33,0xee,0x03,0x0e]
+vqmovnb.s16 q0, q1
+
+# CHECK: vqmovnt.s16 q2, q0  @ encoding: [0x33,0xee,0x01,0x5e]
+# CHECK-NOFP: vqmovnt.s16 q2, q0  @ encoding: [0x33,0xee,0x01,0x5e]
+vqmovnt.s16 q2, q0
+
+# CHECK: vqmovnb.s32 q0, q5  @ encoding: [0x37,0xee,0x0b,0x0e]
+# CHECK-NOFP: vqmovnb.s32 q0, q5  @ encoding: [0x37,0xee,0x0b,0x0e]
+vqmovnb.s32 q0, q5
+
+# CHECK: vqmovnt.s32 q0, q1  @ encoding: [0x37,0xee,0x03,0x1e]
+# CHECK-NOFP: vqmovnt.s32 q0, q1  @ encoding: [0x37,0xee,0x03,0x1e]
+vqmovnt.s32 q0, q1
+
+# CHECK: vqmovnb.u16 q0, q4  @ encoding: [0x33,0xfe,0x09,0x0e]
+# CHECK-NOFP: vqmovnb.u16 q0, q4  @ encoding: [0x33,0xfe,0x09,0x0e]
+vqmovnb.u16 q0, q4
+
+# CHECK: vqmovnt.u16 q0, q7  @ encoding: [0x33,0xfe,0x0f,0x1e]
+# CHECK-NOFP: vqmovnt.u16 q0, q7  @ encoding: [0x33,0xfe,0x0f,0x1e]
+vqmovnt.u16 q0, q7
+
+# CHECK: vqmovnb.u32 q0, q4  @ encoding: [0x37,0xfe,0x09,0x0e]
+# CHECK-NOFP: vqmovnb.u32 q0, q4  @ encoding: [0x37,0xfe,0x09,0x0e]
+vqmovnb.u32 q0, q4
+
+# CHECK: vqmovnt.u32 q0, q2  @ encoding: [0x37,0xfe,0x05,0x1e]
+# CHECK-NOFP: vqmovnt.u32 q0, q2  @ encoding: [0x37,0xfe,0x05,0x1e]
+vqmovnt.u32 q0, q2
+
+# CHECK: vcvtb.f16.f32 q1, q4  @ encoding: [0x3f,0xee,0x09,0x2e]
+# CHECK-NOFP-NOT: vcvtb.f16.f32 q1, q4  @ encoding: [0x3f,0xee,0x09,0x2e]
+vcvtb.f16.f32 q1, q4
+
+# CHECK: vcvtt.f16.f32 q1, q4  @ encoding: [0x3f,0xee,0x09,0x3e]
+# CHECK-NOFP-NOT: vcvtt.f16.f32 q1, q4  @ encoding: [0x3f,0xee,0x09,0x3e]
+vcvtt.f16.f32 q1, q4
+
+# CHECK: vcvtb.f32.f16 q0, q3  @ encoding: [0x3f,0xfe,0x07,0x0e]
+# CHECK-NOFP-NOT: vcvtb.f32.f16 q0, q3  @ encoding: [0x3f,0xfe,0x07,0x0e]
+vcvtb.f32.f16 q0, q3
+
+# CHECK: vcvtt.f32.f16 q0, q1  @ encoding: [0x3f,0xfe,0x03,0x1e]
+# CHECK-NOFP-NOT: vcvtt.f32.f16 q0, q1  @ encoding: [0x3f,0xfe,0x03,0x1e]
+vcvtt.f32.f16 q0, q1
+
+# CHECK: vqmovunb.s16 q0, q3  @ encoding: [0x31,0xee,0x87,0x0e]
+# CHECK-NOFP: vqmovunb.s16 q0, q3  @ encoding: [0x31,0xee,0x87,0x0e]
+vqmovunb.s16 q0, q3
+
+# CHECK: vqmovunt.s16 q4, q1  @ encoding: [0x31,0xee,0x83,0x9e]
+# CHECK-NOFP: vqmovunt.s16 q4, q1  @ encoding: [0x31,0xee,0x83,0x9e]
+vqmovunt.s16 q4, q1
+
+# CHECK: vqmovunb.s32 q1, q7  @ encoding: [0x35,0xee,0x8f,0x2e]
+# CHECK-NOFP: vqmovunb.s32 q1, q7  @ encoding: [0x35,0xee,0x8f,0x2e]
+vqmovunb.s32 q1, q7
+
+# CHECK: vqmovunt.s32 q0, q2  @ encoding: [0x35,0xee,0x85,0x1e]
+# CHECK-NOFP: vqmovunt.s32 q0, q2  @ encoding: [0x35,0xee,0x85,0x1e]
+vqmovunt.s32 q0, q2
+
+# CHECK: vmovnb.i16 q1, q5  @ encoding: [0x31,0xfe,0x8b,0x2e]
+# CHECK-NOFP: vmovnb.i16 q1, q5  @ encoding: [0x31,0xfe,0x8b,0x2e]
+vmovnb.i16 q1, q5
+
+# CHECK: vmovnt.i16 q0, q0  @ encoding: [0x31,0xfe,0x81,0x1e]
+# CHECK-NOFP: vmovnt.i16 q0, q0  @ encoding: [0x31,0xfe,0x81,0x1e]
+vmovnt.i16 q0, q0
+
+# CHECK: vmovnb.i32 q1, q0  @ encoding: [0x35,0xfe,0x81,0x2e]
+# CHECK-NOFP: vmovnb.i32 q1, q0  @ encoding: [0x35,0xfe,0x81,0x2e]
+vmovnb.i32 q1, q0
+
+# CHECK: vmovnt.i32 q3, q3  @ encoding: [0x35,0xfe,0x87,0x7e]
+# CHECK-NOFP: vmovnt.i32 q3, q3  @ encoding: [0x35,0xfe,0x87,0x7e]
+vmovnt.i32 q3, q3
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: complex rotation must be 90 or 270
+vhcadd.s8 q3, q7, q5, #0
+
+# CHECK: vhcadd.s8 q3, q7, q5, #90  @ encoding: [0x0e,0xee,0x0a,0x6f]
+# CHECK-NOFP: vhcadd.s8 q3, q7, q5, #90  @ encoding: [0x0e,0xee,0x0a,0x6f]
+vhcadd.s8 q3, q7, q5, #90
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: complex rotation must be 90 or 270
+vhcadd.s8 q3, q7, q5, #0
+
+# CHECK: vhcadd.s16 q0, q0, q6, #90  @ encoding: [0x10,0xee,0x0c,0x0f]
+# CHECK-NOFP: vhcadd.s16 q0, q0, q6, #90  @ encoding: [0x10,0xee,0x0c,0x0f]
+vhcadd.s16 q0, q0, q6, #90
+
+# CHECK: vhcadd.s16 q0, q0, q6, #90  @ encoding: [0x10,0xee,0x0c,0x0f]
+# CHECK-NOFP: vhcadd.s16 q0, q0, q6, #90  @ encoding: [0x10,0xee,0x0c,0x0f]
+vhcadd.s16 q0, q0, q6, #90
+
+# CHECK: vhcadd.s16 q3, q1, q0, #270  @ encoding: [0x12,0xee,0x00,0x7f]
+# CHECK-NOFP: vhcadd.s16 q3, q1, q0, #270  @ encoding: [0x12,0xee,0x00,0x7f]
+vhcadd.s16 q3, q1, q0, #270
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: complex rotation must be 90 or 270
+vhcadd.s32 q3, q4, q5, #0
+
+# CHECK: vhcadd.s32 q3, q4, q5, #90  @ encoding: [0x28,0xee,0x0a,0x6f]
+# CHECK-NOFP: vhcadd.s32 q3, q4, q5, #90  @ encoding: [0x28,0xee,0x0a,0x6f]
+vhcadd.s32 q3, q4, q5, #90
+
+# CHECK: vhcadd.s32 q6, q7, q2, #270  @ encoding: [0x2e,0xee,0x04,0xdf]
+# CHECK-NOFP: vhcadd.s32 q6, q7, q2, #270  @ encoding: [0x2e,0xee,0x04,0xdf]
+vhcadd.s32 q6, q7, q2, #270
+
+# CHECK: vadc.i32 q1, q0, q2  @ encoding: [0x30,0xee,0x04,0x2f]
+# CHECK-NOFP: vadc.i32 q1, q0, q2  @ encoding: [0x30,0xee,0x04,0x2f]
+vadc.i32 q1, q0, q2
+
+# CHECK: vadci.i32 q0, q1, q1  @ encoding: [0x32,0xee,0x02,0x1f]
+# CHECK-NOFP: vadci.i32 q0, q1, q1  @ encoding: [0x32,0xee,0x02,0x1f]
+vadci.i32 q0, q1, q1
+
+# CHECK: vcadd.i8 q1, q0, q2, #90  @ encoding: [0x00,0xfe,0x04,0x2f]
+# CHECK-NOFP: vcadd.i8 q1, q0, q2, #90  @ encoding: [0x00,0xfe,0x04,0x2f]
+vcadd.i8 q1, q0, q2, #90
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: complex rotation must be 90 or 270
+vcadd.i8 q1, q0, q2, #0
+
+# CHECK: vcadd.i16 q0, q2, q3, #90  @ encoding: [0x14,0xfe,0x06,0x0f]
+# CHECK-NOFP: vcadd.i16 q0, q2, q3, #90  @ encoding: [0x14,0xfe,0x06,0x0f]
+vcadd.i16 q0, q2, q3, #90
+
+# CHECK: vcadd.i16 q0, q5, q5, #270  @ encoding: [0x1a,0xfe,0x0a,0x1f]
+# CHECK-NOFP: vcadd.i16 q0, q5, q5, #270  @ encoding: [0x1a,0xfe,0x0a,0x1f]
+vcadd.i16 q0, q5, q5, #270
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: complex rotation must be 90 or 270
+vcadd.i16 q1, q0, q2, #0
+
+# CHECK: vcadd.i32 q4, q2, q5, #90  @ encoding: [0x24,0xfe,0x0a,0x8f]
+# CHECK-NOFP: vcadd.i32 q4, q2, q5, #90  @ encoding: [0x24,0xfe,0x0a,0x8f]
+vcadd.i32 q4, q2, q5, #90
+
+# CHECK: vcadd.i32 q5, q5, q0, #270  @ encoding: [0x2a,0xfe,0x00,0xbf]
+# CHECK-NOFP: vcadd.i32 q5, q5, q0, #270  @ encoding: [0x2a,0xfe,0x00,0xbf]
+vcadd.i32 q5, q5, q0, #270
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: complex rotation must be 90 or 270
+vcadd.i32 q4, q2, q5, #0
+
+# CHECK: vsbc.i32 q3, q1, q1  @ encoding: [0x32,0xfe,0x02,0x6f]
+# CHECK-NOFP: vsbc.i32 q3, q1, q1  @ encoding: [0x32,0xfe,0x02,0x6f]
+vsbc.i32 q3, q1, q1
+
+# CHECK: vsbci.i32 q2, q6, q2  @ encoding: [0x3c,0xfe,0x04,0x5f]
+# CHECK-NOFP: vsbci.i32 q2, q6, q2  @ encoding: [0x3c,0xfe,0x04,0x5f]
+vsbci.i32 q2, q6, q2
+
+# CHECK: vqdmullb.s16 q0, q4, q5  @ encoding: [0x38,0xee,0x0b,0x0f]
+# CHECK-NOFP: vqdmullb.s16 q0, q4, q5  @ encoding: [0x38,0xee,0x0b,0x0f]
+vqdmullb.s16 q0, q4, q5
+
+# CHECK: vqdmullt.s16 q0, q6, q5  @ encoding: [0x3c,0xee,0x0b,0x1f]
+# CHECK-NOFP: vqdmullt.s16 q0, q6, q5  @ encoding: [0x3c,0xee,0x0b,0x1f]
+vqdmullt.s16 q0, q6, q5
+
+# CHECK: vqdmullb.s32 q0, q3, q7  @ encoding: [0x36,0xfe,0x0f,0x0f]
+# CHECK-NOFP: vqdmullb.s32 q0, q3, q7  @ encoding: [0x36,0xfe,0x0f,0x0f]
+vqdmullb.s32 q0, q3, q7
+
+# CHECK: vqdmullt.s32 q0, q7, q5  @ encoding: [0x3e,0xfe,0x0b,0x1f]
+# CHECK-NOFP: vqdmullt.s32 q0, q7, q5  @ encoding: [0x3e,0xfe,0x0b,0x1f]
+vqdmullt.s32 q0, q7, q5
+
+# CHECK: vqdmullb.s16 q0, q1, q0  @ encoding: [0x32,0xee,0x01,0x0f]
+# CHECK-NOFP: vqdmullb.s16 q0, q1, q0  @ encoding: [0x32,0xee,0x01,0x0f]
+vqdmullb.s16 q0, q1, q0
+
+# CHECK: vqdmullt.s16 q0, q0, q5  @ encoding: [0x30,0xee,0x0b,0x1f]
+# CHECK-NOFP: vqdmullt.s16 q0, q0, q5  @ encoding: [0x30,0xee,0x0b,0x1f]
+vqdmullt.s16 q0, q0, q5
+
+# ERROR: [[@LINE+1]]:{{[0-9]+}}: {{error|note}}: Qd register and Qm register can't be identical
+vqdmullb.s32 q0, q1, q0
+
+vqdmullt.s16 q0, q1, q2
+# CHECK: vqdmullt.s16 q0, q1, q2 @ encoding: [0x32,0xee,0x05,0x1f]
+# CHECK-NOFP: vqdmullt.s16 q0, q1, q2 @ encoding: [0x32,0xee,0x05,0x1f]
+
+vpste
+vqdmulltt.s32 q0, q1, q2
+vqdmullbe.s16 q0, q1, q2
+# CHECK: vpste @ encoding: [0x71,0xfe,0x4d,0x8f]
+# CHECK-NOFP: vpste @ encoding: [0x71,0xfe,0x4d,0x8f]
+# CHECK: vqdmulltt.s32 q0, q1, q2 @ encoding: [0x32,0xfe,0x05,0x1f]
+# CHECK-NOFP: vqdmulltt.s32 q0, q1, q2 @ encoding: [0x32,0xfe,0x05,0x1f]
+# CHECK: vqdmullbe.s16 q0, q1, q2 @ encoding: [0x32,0xee,0x05,0x0f]
+# CHECK-NOFP: vqdmullbe.s16 q0, q1, q2 @ encoding: [0x32,0xee,0x05,0x0f]
+
+vpste
+vmulltt.p8 q0, q1, q2
+vmullbe.p16 q0, q1, q2
+# CHECK: vpste @ encoding: [0x71,0xfe,0x4d,0x8f]
+# CHECK-NOFP: vpste @ encoding: [0x71,0xfe,0x4d,0x8f]
+# CHECK: vmulltt.p8 q0, q1, q2 @ encoding: [0x33,0xee,0x04,0x1e]
+# CHECK-NOFP: vmulltt.p8 q0, q1, q2 @ encoding: [0x33,0xee,0x04,0x1e]
+# CHECK: vmullbe.p16 q0, q1, q2 @ encoding: [0x33,0xfe,0x04,0x0e]
+# CHECK-NOFP: vmullbe.p16 q0, q1, q2 @ encoding: [0x33,0xfe,0x04,0x0e]
+
+# ----------------------------------------------------------------------
+# The following tests have to go last because of the NOFP-NOT checks inside the
+# VPT block.
+
+vpste
+vcmult.f16 q0, q1, q2, #180
+vcmule.f16 q0, q1, q2, #180
+# CHECK: vpste @ encoding: [0x71,0xfe,0x4d,0x8f]
+# CHECK: vcmult.f16 q0, q1, q2, #180 @ encoding: [0x32,0xee,0x04,0x1e]
+# CHECK-NOFP-NOT: vcmult.f16 q0, q1, q2, #180 @ encoding: [0x32,0xee,0x04,0x1e]
+# CHECK: vcmule.f16 q0, q1, q2, #180 @ encoding: [0x32,0xee,0x04,0x1e]
+# CHECK-NOFP-NOT: vcmule.f16 q0, q1, q2, #180 @ encoding: [0x32,0xee,0x04,0x1e]
+
+vpstet
+vcvtbt.f16.f32 q0, q1
+vcvtne.s16.f16 q0, q1
+vcvtmt.s16.f16 q0, q1
+# CHECK: vpstet @ encoding: [0x71,0xfe,0x4d,0xcf]
+# CHECK: vcvtbt.f16.f32 q0, q1 @ encoding: [0x3f,0xee,0x03,0x0e]
+# CHECK-NOFP-NOT: vcvtbt.f16.f32 q0, q1 @ encoding: [0x3f,0xee,0x03,0x0e]
+# CHECK: vcvtne.s16.f16 q0, q1 @ encoding: [0xb7,0xff,0x42,0x01]
+# CHECK-NOFP-NOT: vcvtne.s16.f16 q0, q1 @ encoding: [0xb7,0xff,0x42,0x01]
+# CHECK: vcvtmt.s16.f16 q0, q1 @ encoding: [0xb7,0xff,0x42,0x03
+# CHECK-NOFP-NOT: vcvtmt.s16.f16 q0, q1 @ encoding: [0xb7,0xff,0x42,0x03
+
+vpte.f32 lt, q3, r1
+vcvttt.f16.f32 q2, q0
+vcvtte.f32.f16 q1, q0
+# CHECK: vpte.f32 lt, q3, r1      @ encoding: [0x77,0xee,0xc1,0x9f]
+# CHECK-NOFP-NOT: vpte.f32 lt, q3, r1      @ encoding: [0x77,0xee,0xe1,0x8f]
+# CHECK: vcvttt.f16.f32 q2, q0          @ encoding: [0x3f,0xee,0x01,0x5e]
+# CHECK-NOFP-NOT: vcvttt.f16.f32 q2, q0          @ encoding: [0x3f,0xee,0x01,0x5e]
+# CHECK: vcvtte.f32.f16 q1, q0          @ encoding: [0x3f,0xfe,0x01,0x3e]
+
+vpte.f32 lt, q3, r1
+vcvtbt.f16.f32 q2, q0
+vcvtbe.f32.f16 q1, q0
+# CHECK: vpte.f32 lt, q3, r1      @ encoding: [0x77,0xee,0xc1,0x9f]
+# CHECK-NOFP-NOT: vpte.f32 lt, q3, r1      @ encoding: [0x77,0xee,0xe1,0x8f]
+# CHECK: vcvtbt.f16.f32 q2, q0          @ encoding: [0x3f,0xee,0x01,0x4e]
+# CHECK-NOFP-NOT: vcvtbt.f16.f32 q2, q0          @ encoding: [0x3f,0xee,0x01,0x4e]
+# CHECK: vcvtbe.f32.f16 q1, q0          @ encoding: [0x3f,0xfe,0x01,0x2e]
+# CHECK-NOFP-NOT: vcvtbe.f32.f16 q1, q0          @ encoding: [0x3f,0xfe,0x01,0x2e]
+
+ite eq
+vcvtteq.f16.f32 s0, s1
+vcvttne.f16.f32 s0, s1
+# CHECK: ite eq                      @ encoding: [0x0c,0xbf]
+# CHECK: vcvtteq.f16.f32 s0, s1          @ encoding: [0xb3,0xee,0xe0,0x0a]
+# CHECK-NOFP-NOT: vcvtteq.f16.f32 s0, s1          @ encoding: [0xb3,0xee,0xe0,0x0a]
+# CHECK: vcvttne.f16.f32 s0, s1          @ encoding: [0xb3,0xee,0xe0,0x0a]
+# CHECK-NOFP-NOT: vcvttne.f16.f32 s0, s1          @ encoding: [0xb3,0xee,0xe0,0x0a]

llvm/trunk/test/MC/Disassembler/ARM/mve-qdest-qsrc.txt

+# RUN: llvm-mc -disassemble -triple=thumbv8.1m.main-none-eabi -mattr=+mve.fp,+fp64 -show-encoding %s | FileCheck %s
+# RUN: not llvm-mc -disassemble -triple=thumbv8.1m.main-none-eabi -show-encoding %s &> %t
+# RUN: FileCheck --check-prefix=CHECK-NOMVE < %t %s
+
+# CHECK: vqdmladhx.s8 q1, q6, q6  @ encoding: [0x0c,0xee,0x0c,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x0c,0xee,0x0c,0x3e]
+
+# CHECK: vqdmladhx.s16 q0, q1, q4  @ encoding: [0x12,0xee,0x08,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x12,0xee,0x08,0x1e]
+
+# CHECK: vqdmladhx.s32 q0, q3, q7  @ encoding: [0x26,0xee,0x0e,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x26,0xee,0x0e,0x1e]
+
+# CHECK: vqdmladh.s8 q0, q1, q1  @ encoding: [0x02,0xee,0x02,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x02,0xee,0x02,0x0e]
+
+# CHECK: vqdmladh.s16 q0, q2, q2  @ encoding: [0x14,0xee,0x04,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x14,0xee,0x04,0x0e]
+
+# CHECK: vqdmladh.s32 q1, q5, q7  @ encoding: [0x2a,0xee,0x0e,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x2a,0xee,0x0e,0x2e]
+
+# CHECK: vqrdmladhx.s8 q0, q7, q0  @ encoding: [0x0e,0xee,0x01,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x0e,0xee,0x01,0x1e]
+
+# CHECK: vqrdmladhx.s16 q0, q0, q1  @ encoding: [0x10,0xee,0x03,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x10,0xee,0x03,0x1e]
+
+# CHECK: vqrdmladhx.s32 q1, q0, q4  @ encoding: [0x20,0xee,0x09,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x20,0xee,0x09,0x3e]
+
+# CHECK: vqrdmladh.s8 q0, q6, q2  @ encoding: [0x0c,0xee,0x05,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x0c,0xee,0x05,0x0e]
+
+# CHECK: vqrdmladh.s16 q1, q5, q4  @ encoding: [0x1a,0xee,0x09,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x1a,0xee,0x09,0x2e]
+
+# CHECK: vqrdmladh.s32 q0, q2, q2  @ encoding: [0x24,0xee,0x05,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x24,0xee,0x05,0x0e]
+
+# CHECK: vqdmlsdhx.s8 q1, q4, q7  @ encoding: [0x08,0xfe,0x0e,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x08,0xfe,0x0e,0x3e]
+
+# CHECK: vqdmlsdhx.s16 q0, q2, q5  @ encoding: [0x14,0xfe,0x0a,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x14,0xfe,0x0a,0x1e]
+
+# CHECK: vqdmlsdhx.s32 q3, q4, q6  @ encoding: [0x28,0xfe,0x0c,0x7e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x28,0xfe,0x0c,0x7e]
+
+# CHECK: vqdmlsdh.s8 q0, q3, q6  @ encoding: [0x06,0xfe,0x0c,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x06,0xfe,0x0c,0x0e]
+
+# CHECK: vqdmlsdh.s16 q0, q4, q1  @ encoding: [0x18,0xfe,0x02,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x18,0xfe,0x02,0x0e]
+
+# CHECK: vqdmlsdh.s32 q2, q5, q0  @ encoding: [0x2a,0xfe,0x00,0x4e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x2a,0xfe,0x00,0x4e]
+
+# CHECK: vqrdmlsdhx.s8 q0, q3, q1  @ encoding: [0x06,0xfe,0x03,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x06,0xfe,0x03,0x1e]
+
+# CHECK: vqrdmlsdhx.s16 q0, q1, q4  @ encoding: [0x12,0xfe,0x09,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x12,0xfe,0x09,0x1e]
+
+# CHECK: vqrdmlsdhx.s32 q1, q6, q3  @ encoding: [0x2c,0xfe,0x07,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x2c,0xfe,0x07,0x3e]
+
+# CHECK: vqrdmlsdh.s8 q3, q3, q0  @ encoding: [0x06,0xfe,0x01,0x6e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x06,0xfe,0x01,0x6e]
+
+# CHECK: vqrdmlsdh.s16 q0, q7, q4  @ encoding: [0x1e,0xfe,0x09,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x1e,0xfe,0x09,0x0e]
+
+# CHECK: vqrdmlsdh.s32 q0, q6, q7  @ encoding: [0x2c,0xfe,0x0f,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x2c,0xfe,0x0f,0x0e]
+
+# CHECK: vcmul.f16 q6, q2, q5, #0  @ encoding: [0x34,0xee,0x0a,0xce]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x34,0xee,0x0a,0xce]
+
+# CHECK: vcmul.f16 q1, q0, q5, #90  @ encoding: [0x30,0xee,0x0b,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x30,0xee,0x0b,0x2e]
+
+# CHECK: vcmul.f16 q1, q0, q5, #180  @ encoding: [0x30,0xee,0x0a,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x30,0xee,0x0a,0x3e]
+
+# CHECK: vcmul.f16 q1, q0, q5, #270  @ encoding: [0x30,0xee,0x0b,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x30,0xee,0x0b,0x3e]
+
+# CHECK: vcmul.f32 q1, q7, q5, #0  @ encoding: [0x3e,0xfe,0x0a,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3e,0xfe,0x0a,0x2e]
+
+# CHECK: vcmul.f32 q3, q4, q2, #90  @ encoding: [0x38,0xfe,0x05,0x6e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x38,0xfe,0x05,0x6e]
+
+# CHECK: vcmul.f32 q5, q1, q3, #180  @ encoding: [0x32,0xfe,0x06,0xbe]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x32,0xfe,0x06,0xbe]
+
+# CHECK: vcmul.f32 q0, q7, q4, #270  @ encoding: [0x3e,0xfe,0x09,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3e,0xfe,0x09,0x1e]
+
+# CHECK: vmullb.s8 q2, q6, q0  @ encoding: [0x0d,0xee,0x00,0x4e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x0d,0xee,0x00,0x4e]
+
+# CHECK: vmullb.s16 q3, q4, q3  @ encoding: [0x19,0xee,0x06,0x6e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x19,0xee,0x06,0x6e]
+
+# CHECK: vmullb.s32 q3, q5, q6  @ encoding: [0x2b,0xee,0x0c,0x6e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x2b,0xee,0x0c,0x6e]
+
+# CHECK: vmullt.s8 q0, q6, q2  @ encoding: [0x0d,0xee,0x04,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x0d,0xee,0x04,0x1e]
+
+# CHECK: vmullt.s16 q0, q0, q2  @ encoding: [0x11,0xee,0x04,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x11,0xee,0x04,0x1e]
+
+# CHECK: vmullt.s32 q2, q4, q4  @ encoding: [0x29,0xee,0x08,0x5e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x29,0xee,0x08,0x5e]
+
+# CHECK: vmullb.p8 q2, q3, q7  @ encoding: [0x37,0xee,0x0e,0x4e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x37,0xee,0x0e,0x4e]
+
+# CHECK: vmullb.p16 q0, q1, q3  @ encoding: [0x33,0xfe,0x06,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x33,0xfe,0x06,0x0e]
+
+# CHECK: vmullt.p8 q1, q1, q7  @ encoding: [0x33,0xee,0x0e,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x33,0xee,0x0e,0x3e]
+
+# CHECK: vmullt.p16 q0, q7, q7  @ encoding: [0x3f,0xfe,0x0e,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3f,0xfe,0x0e,0x1e]
+
+# CHECK: vmulh.s8 q0, q4, q5  @ encoding: [0x09,0xee,0x0b,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x09,0xee,0x0b,0x0e]
+
+# CHECK: vmulh.s16 q0, q7, q4  @ encoding: [0x1f,0xee,0x09,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x1f,0xee,0x09,0x0e]
+
+# CHECK: vmulh.s32 q0, q7, q4  @ encoding: [0x2f,0xee,0x09,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x2f,0xee,0x09,0x0e]
+
+# CHECK: vmulh.u8 q3, q5, q2  @ encoding: [0x0b,0xfe,0x05,0x6e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x0b,0xfe,0x05,0x6e]
+
+# CHECK: vmulh.u16 q2, q7, q4  @ encoding: [0x1f,0xfe,0x09,0x4e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x1f,0xfe,0x09,0x4e]
+
+# CHECK: vmulh.u32 q1, q3, q2  @ encoding: [0x27,0xfe,0x05,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x27,0xfe,0x05,0x2e]
+
+# CHECK: vrmulh.s8 q1, q1, q2  @ encoding: [0x03,0xee,0x05,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x03,0xee,0x05,0x3e]
+
+# CHECK: vrmulh.s16 q1, q1, q2  @ encoding: [0x13,0xee,0x05,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x13,0xee,0x05,0x3e]
+
+# CHECK: vrmulh.s32 q3, q1, q0  @ encoding: [0x23,0xee,0x01,0x7e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x23,0xee,0x01,0x7e]
+
+# CHECK: vrmulh.u8 q1, q6, q0  @ encoding: [0x0d,0xfe,0x01,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x0d,0xfe,0x01,0x3e]
+
+# CHECK: vrmulh.u16 q4, q3, q6  @ encoding: [0x17,0xfe,0x0d,0x9e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x17,0xfe,0x0d,0x9e]
+
+# CHECK: vrmulh.u32 q1, q2, q2  @ encoding: [0x25,0xfe,0x05,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x25,0xfe,0x05,0x3e]
+
+# CHECK: vqmovnb.s16 q0, q1  @ encoding: [0x33,0xee,0x03,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x33,0xee,0x03,0x0e]
+
+# CHECK: vqmovnt.s16 q2, q0  @ encoding: [0x33,0xee,0x01,0x5e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x33,0xee,0x01,0x5e]
+
+# CHECK: vqmovnb.s32 q0, q5  @ encoding: [0x37,0xee,0x0b,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x37,0xee,0x0b,0x0e]
+
+# CHECK: vqmovnt.s32 q0, q1  @ encoding: [0x37,0xee,0x03,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x37,0xee,0x03,0x1e]
+
+# CHECK: vqmovnb.u16 q0, q4  @ encoding: [0x33,0xfe,0x09,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x33,0xfe,0x09,0x0e]
+
+# CHECK: vqmovnt.u16 q0, q7  @ encoding: [0x33,0xfe,0x0f,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x33,0xfe,0x0f,0x1e]
+
+# CHECK: vqmovnb.u32 q0, q4  @ encoding: [0x37,0xfe,0x09,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x37,0xfe,0x09,0x0e]
+
+# CHECK: vqmovnt.u32 q0, q2  @ encoding: [0x37,0xfe,0x05,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x37,0xfe,0x05,0x1e]
+
+# CHECK: vcvtb.f16.f32 q1, q4  @ encoding: [0x3f,0xee,0x09,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3f,0xee,0x09,0x2e]
+
+# CHECK: vcvtt.f16.f32 q1, q4  @ encoding: [0x3f,0xee,0x09,0x3e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3f,0xee,0x09,0x3e]
+
+# CHECK: vcvtb.f32.f16 q0, q3  @ encoding: [0x3f,0xfe,0x07,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3f,0xfe,0x07,0x0e]
+
+# CHECK: vcvtt.f32.f16 q0, q1  @ encoding: [0x3f,0xfe,0x03,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3f,0xfe,0x03,0x1e]
+
+# CHECK: vcvtb.f16.f32 q1, q4  @ encoding: [0x3f,0xee,0x09,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3f,0xee,0x09,0x2e]
+
+# CHECK: vcvtt.f32.f16 q0, q1  @ encoding: [0x3f,0xfe,0x03,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3f,0xfe,0x03,0x1e]
+
+# CHECK: vcvtt.f64.f16 d0, s0 @ encoding: [0xb2,0xee,0xc0,0x0b]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0xb2,0xee,0xc0,0x0b]
+
+# CHECK: vcvtt.f16.f64 s1, d2 @ encoding: [0xf3,0xee,0xc2,0x0b]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0xf3,0xee,0xc2,0x0b]
+
+# CHECK: vqmovunb.s16 q0, q3  @ encoding: [0x31,0xee,0x87,0x0e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x31,0xee,0x87,0x0e]
+
+# CHECK: vqmovunt.s16 q4, q1  @ encoding: [0x31,0xee,0x83,0x9e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x31,0xee,0x83,0x9e]
+
+# CHECK: vqmovunb.s32 q1, q7  @ encoding: [0x35,0xee,0x8f,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x35,0xee,0x8f,0x2e]
+
+# CHECK: vqmovunt.s32 q0, q2  @ encoding: [0x35,0xee,0x85,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x35,0xee,0x85,0x1e]
+
+# CHECK: vmovnb.i16 q1, q5  @ encoding: [0x31,0xfe,0x8b,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x31,0xfe,0x8b,0x2e]
+
+# CHECK: vmovnt.i16 q0, q0  @ encoding: [0x31,0xfe,0x81,0x1e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x31,0xfe,0x81,0x1e]
+
+# CHECK: vmovnb.i32 q1, q0  @ encoding: [0x35,0xfe,0x81,0x2e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x35,0xfe,0x81,0x2e]
+
+# CHECK: vmovnt.i32 q3, q3  @ encoding: [0x35,0xfe,0x87,0x7e]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x35,0xfe,0x87,0x7e]
+
+# CHECK: vhcadd.s8 q3, q7, q5, #90  @ encoding: [0x0e,0xee,0x0a,0x6f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x0e,0xee,0x0a,0x6f]
+
+# CHECK: vhcadd.s16 q0, q0, q6, #90  @ encoding: [0x10,0xee,0x0c,0x0f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x10,0xee,0x0c,0x0f]
+
+# CHECK: vhcadd.s16 q0, q0, q6, #90  @ encoding: [0x10,0xee,0x0c,0x0f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x10,0xee,0x0c,0x0f]
+
+# CHECK: vhcadd.s16 q3, q1, q0, #270  @ encoding: [0x12,0xee,0x00,0x7f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x12,0xee,0x00,0x7f]
+
+# CHECK: vhcadd.s32 q3, q4, q5, #90  @ encoding: [0x28,0xee,0x0a,0x6f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x28,0xee,0x0a,0x6f]
+
+# CHECK: vhcadd.s32 q6, q7, q2, #270  @ encoding: [0x2e,0xee,0x04,0xdf]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x2e,0xee,0x04,0xdf]
+
+# CHECK: vadc.i32 q1, q0, q2  @ encoding: [0x30,0xee,0x04,0x2f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x30,0xee,0x04,0x2f]
+
+# CHECK: vadci.i32 q0, q1, q1  @ encoding: [0x32,0xee,0x02,0x1f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x32,0xee,0x02,0x1f]
+
+# CHECK: vcadd.i8 q1, q0, q2, #90  @ encoding: [0x00,0xfe,0x04,0x2f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x00,0xfe,0x04,0x2f]
+
+# CHECK: vcadd.i16 q0, q2, q3, #90  @ encoding: [0x14,0xfe,0x06,0x0f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x14,0xfe,0x06,0x0f]
+
+# CHECK: vcadd.i16 q0, q5, q5, #270  @ encoding: [0x1a,0xfe,0x0a,0x1f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x1a,0xfe,0x0a,0x1f]
+
+# CHECK: vcadd.i32 q4, q2, q5, #90  @ encoding: [0x24,0xfe,0x0a,0x8f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x24,0xfe,0x0a,0x8f]
+
+# CHECK: vcadd.i32 q5, q5, q0, #270  @ encoding: [0x2a,0xfe,0x00,0xbf]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x2a,0xfe,0x00,0xbf]
+
+# CHECK: vsbc.i32 q3, q1, q1  @ encoding: [0x32,0xfe,0x02,0x6f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x32,0xfe,0x02,0x6f]
+
+# CHECK: vsbci.i32 q2, q6, q2  @ encoding: [0x3c,0xfe,0x04,0x5f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3c,0xfe,0x04,0x5f]
+
+# CHECK: vqdmullb.s16 q0, q4, q5  @ encoding: [0x38,0xee,0x0b,0x0f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x38,0xee,0x0b,0x0f]
+
+# CHECK: vqdmullt.s16 q0, q6, q5  @ encoding: [0x3c,0xee,0x0b,0x1f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3c,0xee,0x0b,0x1f]
+
+# CHECK: vqdmullb.s32 q0, q3, q7  @ encoding: [0x36,0xfe,0x0f,0x0f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x36,0xfe,0x0f,0x0f]
+
+# CHECK: vqdmullt.s32 q0, q7, q5  @ encoding: [0x3e,0xfe,0x0b,0x1f]
+# CHECK-NOMVE: [[@LINE+1]]:2: warning: invalid instruction encoding
+[0x3e,0xfe,0x0b,0x1f]