diff --git a/clang/lib/Basic/Targets/AArch64.cpp b/clang/lib/Basic/Targets/AArch64.cpp
index bd1a8834c2fa3..336c7491a5cc8 100644
--- a/clang/lib/Basic/Targets/AArch64.cpp
+++ b/clang/lib/Basic/Targets/AArch64.cpp
@@ -486,17 +486,29 @@ bool AArch64TargetInfo::validateAsmConstraint(
     Info.setAllowsRegister();
     return true;
   case 'U':
+    if (Name[1] == 'p' && (Name[2] == 'l' || Name[2] == 'a')) {
+      // SVE predicate registers ("Upa"=P0-15, "Upl"=P0-P7)
+      Info.setAllowsRegister();
+      Name += 2;
+      return true;
+    }
     // Ump: A memory address suitable for ldp/stp in SI, DI, SF and DF modes.
     // Utf: A memory address suitable for ldp/stp in TF mode.
     // Usa: An absolute symbolic address.
     // Ush: The high part (bits 32:12) of a pc-relative symbolic address.
-    llvm_unreachable("FIXME: Unimplemented support for U* constraints.");
+
+    // Better to return an error saying that it's an unrecognised constraint
+    // even if this is a valid constraint in gcc.
+    return false;
   case 'z': // Zero register, wzr or xzr
     Info.setAllowsRegister();
     return true;
   case 'x': // Floating point and SIMD registers (V0-V15)
     Info.setAllowsRegister();
     return true;
+  case 'y': // SVE registers (V0-V7)
+    Info.setAllowsRegister();
+    return true;
   }
   return false;
 }
diff --git a/clang/lib/Basic/Targets/AArch64.h b/clang/lib/Basic/Targets/AArch64.h
index 5e78237743c90..befbf693ad8b7 100644
--- a/clang/lib/Basic/Targets/AArch64.h
+++ b/clang/lib/Basic/Targets/AArch64.h
@@ -87,6 +87,21 @@ class LLVM_LIBRARY_VISIBILITY AArch64TargetInfo : public TargetInfo {
 
   ArrayRef<const char *> getGCCRegNames() const override;
   ArrayRef<TargetInfo::GCCRegAlias> getGCCRegAliases() const override;
+
+  std::string convertConstraint(const char *&Constraint) const override {
+    std::string R;
+    switch (*Constraint) {
+    case 'U': // Three-character constraint; add "@3" hint for later parsing.
+      R = std::string("@3") + std::string(Constraint, 3);
+      Constraint += 2;
+      break;
+    default:
+      R = std::string(1, *Constraint);
+      break;
+    }
+    return R;
+  }
+
   bool validateAsmConstraint(const char *&Name,
                              TargetInfo::ConstraintInfo &Info) const override;
   bool
diff --git a/clang/lib/CodeGen/CGCall.cpp b/clang/lib/CodeGen/CGCall.cpp
index 1188ea39ba2c9..ad8ebd245b935 100644
--- a/clang/lib/CodeGen/CGCall.cpp
+++ b/clang/lib/CodeGen/CGCall.cpp
@@ -4496,8 +4496,9 @@ RValue CodeGenFunction::EmitCall(const CGFunctionInfo &CallInfo,
   // Update the largest vector width if any arguments have vector types.
   for (unsigned i = 0; i < IRCallArgs.size(); ++i) {
     if (auto *VT = dyn_cast<llvm::VectorType>(IRCallArgs[i]->getType()))
-      LargestVectorWidth = std::max((uint64_t)LargestVectorWidth,
-                                   VT->getPrimitiveSizeInBits().getFixedSize());
+      LargestVectorWidth =
+          std::max((uint64_t)LargestVectorWidth,
+                   VT->getPrimitiveSizeInBits().getKnownMinSize());
   }
 
   // Compute the calling convention and attributes.
@@ -4611,8 +4612,9 @@ RValue CodeGenFunction::EmitCall(const CGFunctionInfo &CallInfo,
 
   // Update largest vector width from the return type.
   if (auto *VT = dyn_cast<llvm::VectorType>(CI->getType()))
-    LargestVectorWidth = std::max((uint64_t)LargestVectorWidth,
-                                  VT->getPrimitiveSizeInBits().getFixedSize());
+    LargestVectorWidth =
+        std::max((uint64_t)LargestVectorWidth,
+                 VT->getPrimitiveSizeInBits().getKnownMinSize());
 
   // Insert instrumentation or attach profile metadata at indirect call sites.
   // For more details, see the comment before the definition of
diff --git a/clang/lib/CodeGen/CGStmt.cpp b/clang/lib/CodeGen/CGStmt.cpp
index a334bab06783e..49f1725ed470d 100644
--- a/clang/lib/CodeGen/CGStmt.cpp
+++ b/clang/lib/CodeGen/CGStmt.cpp
@@ -2095,8 +2095,9 @@ void CodeGenFunction::EmitAsmStmt(const AsmStmt &S) {
 
       // Update largest vector width for any vector types.
       if (auto *VT = dyn_cast<llvm::VectorType>(ResultRegTypes.back()))
-        LargestVectorWidth = std::max((uint64_t)LargestVectorWidth,
-                                   VT->getPrimitiveSizeInBits().getFixedSize());
+        LargestVectorWidth =
+            std::max((uint64_t)LargestVectorWidth,
+                     VT->getPrimitiveSizeInBits().getKnownMinSize());
     } else {
       ArgTypes.push_back(Dest.getAddress(*this).getType());
       Args.push_back(Dest.getPointer(*this));
@@ -2120,8 +2121,9 @@ void CodeGenFunction::EmitAsmStmt(const AsmStmt &S) {
 
       // Update largest vector width for any vector types.
       if (auto *VT = dyn_cast<llvm::VectorType>(Arg->getType()))
-        LargestVectorWidth = std::max((uint64_t)LargestVectorWidth,
-                                   VT->getPrimitiveSizeInBits().getFixedSize());
+        LargestVectorWidth =
+            std::max((uint64_t)LargestVectorWidth,
+                     VT->getPrimitiveSizeInBits().getKnownMinSize());
       if (Info.allowsRegister())
         InOutConstraints += llvm::utostr(i);
       else
@@ -2207,8 +2209,9 @@ void CodeGenFunction::EmitAsmStmt(const AsmStmt &S) {
 
     // Update largest vector width for any vector types.
     if (auto *VT = dyn_cast<llvm::VectorType>(Arg->getType()))
-      LargestVectorWidth = std::max((uint64_t)LargestVectorWidth,
-                                   VT->getPrimitiveSizeInBits().getFixedSize());
+      LargestVectorWidth =
+          std::max((uint64_t)LargestVectorWidth,
+                   VT->getPrimitiveSizeInBits().getKnownMinSize());
 
     ArgTypes.push_back(Arg->getType());
     Args.push_back(Arg);
diff --git a/clang/test/CodeGen/aarch64-sve-inline-asm-crash.c b/clang/test/CodeGen/aarch64-sve-inline-asm-crash.c
new file mode 100644
index 0000000000000..11ddb7f6d5800
--- /dev/null
+++ b/clang/test/CodeGen/aarch64-sve-inline-asm-crash.c
@@ -0,0 +1,24 @@
+// REQUIRES: aarch64-registered-target
+
+// RUN: not %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -fallow-half-arguments-and-returns \
+// RUN:   -target-feature +neon -S -O1 -o - %s 2>&1 | FileCheck %s
+
+// Set a vector constraint for an sve predicate register
+// As the wrong constraint is used for an SVBool,
+// the compiler will try to extend the nxv16i1 to an nxv16i8
+// TODO: We don't have patterns for this yet but once they are added this test
+// should be updated to check for an assembler error
+__SVBool_t funcB1(__SVBool_t in)
+{
+  __SVBool_t ret ;
+  asm volatile (
+    "mov %[ret].b, %[in].b \n"
+    : [ret] "=w" (ret)
+    : [in] "w" (in)
+    :);
+
+  return ret ;
+}
+
+// CHECK: funcB1
+// CHECK-ERROR: fatal error: error in backend: Cannot select
diff --git a/clang/test/CodeGen/aarch64-sve-inline-asm-datatypes.c b/clang/test/CodeGen/aarch64-sve-inline-asm-datatypes.c
new file mode 100644
index 0000000000000..c68bb25e36fa5
--- /dev/null
+++ b/clang/test/CodeGen/aarch64-sve-inline-asm-datatypes.c
@@ -0,0 +1,252 @@
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -fallow-half-arguments-and-returns \
+// RUN:   -target-feature +neon -S -O1 -o - -emit-llvm %s | FileCheck %s
+
+// Tests to check that all sve datatypes can be passed in as input operands
+// and passed out as output operands.
+
+#define SVINT_TEST(DT, KIND)\
+DT func_int_##DT##KIND(DT in)\
+{\
+  DT out;\
+  asm volatile (\
+    "ptrue p0.b\n"\
+    "mov %[out]." #KIND ", p0/m, %[in]." #KIND "\n"\
+    : [out] "=w" (out)\
+    : [in] "w" (in)\
+    : "p0"\
+    );\
+  return out;\
+}
+
+SVINT_TEST(__SVUint8_t,b);
+// CHECK: call <vscale x 16 x i8> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 16 x i8> %in)
+SVINT_TEST(__SVUint8_t,h);
+// CHECK: call <vscale x 16 x i8> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 16 x i8> %in)
+SVINT_TEST(__SVUint8_t,s);
+// CHECK: call <vscale x 16 x i8> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 16 x i8> %in)
+SVINT_TEST(__SVUint8_t,d);
+// CHECK: call <vscale x 16 x i8> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 16 x i8> %in)
+
+SVINT_TEST(__SVUint16_t,b);
+// CHECK: call <vscale x 8 x i16> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 8 x i16> %in)
+SVINT_TEST(__SVUint16_t,h);
+// CHECK: call <vscale x 8 x i16> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 8 x i16> %in)
+SVINT_TEST(__SVUint16_t,s);
+// CHECK: call <vscale x 8 x i16> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 8 x i16> %in)
+SVINT_TEST(__SVUint16_t,d);
+// CHECK: call <vscale x 8 x i16> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 8 x i16> %in)
+
+SVINT_TEST(__SVUint32_t,b);
+// CHECK: call <vscale x 4 x i32> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 4 x i32> %in)
+SVINT_TEST(__SVUint32_t,h);
+// CHECK: call <vscale x 4 x i32> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 4 x i32> %in)
+SVINT_TEST(__SVUint32_t,s);
+// CHECK: call <vscale x 4 x i32> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 4 x i32> %in)
+SVINT_TEST(__SVUint32_t,d);
+// CHECK: call <vscale x 4 x i32> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 4 x i32> %in)
+
+SVINT_TEST(__SVUint64_t,b);
+// CHECK: call <vscale x 2 x i64> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 2 x i64> %in)
+SVINT_TEST(__SVUint64_t,h);
+// CHECK: call <vscale x 2 x i64> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 2 x i64> %in)
+SVINT_TEST(__SVUint64_t,s);
+// CHECK: call <vscale x 2 x i64> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 2 x i64> %in)
+SVINT_TEST(__SVUint64_t,d);
+// CHECK: call <vscale x 2 x i64> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 2 x i64> %in)
+
+SVINT_TEST(__SVInt8_t,b);
+// CHECK: call <vscale x 16 x i8> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 16 x i8> %in)
+SVINT_TEST(__SVInt8_t,h);
+// CHECK: call <vscale x 16 x i8> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 16 x i8> %in)
+SVINT_TEST(__SVInt8_t,s);
+// CHECK: call <vscale x 16 x i8> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 16 x i8> %in)
+SVINT_TEST(__SVInt8_t,d);
+// CHECK: call <vscale x 16 x i8> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 16 x i8> %in)
+
+SVINT_TEST(__SVInt16_t,b);
+// CHECK: call <vscale x 8 x i16> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 8 x i16> %in)
+SVINT_TEST(__SVInt16_t,h);
+// CHECK: call <vscale x 8 x i16> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 8 x i16> %in)
+SVINT_TEST(__SVInt16_t,s);
+// CHECK: call <vscale x 8 x i16> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 8 x i16> %in)
+SVINT_TEST(__SVInt16_t,d);
+// CHECK: call <vscale x 8 x i16> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 8 x i16> %in)
+
+SVINT_TEST(__SVInt32_t,b);
+// CHECK: call <vscale x 4 x i32> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 4 x i32> %in)
+SVINT_TEST(__SVInt32_t,h);
+// CHECK: call <vscale x 4 x i32> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 4 x i32> %in)
+SVINT_TEST(__SVInt32_t,s);
+// CHECK: call <vscale x 4 x i32> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 4 x i32> %in)
+SVINT_TEST(__SVInt32_t,d);
+// CHECK: call <vscale x 4 x i32> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 4 x i32> %in)
+
+SVINT_TEST(__SVInt64_t,b);
+// CHECK: call <vscale x 2 x i64> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 2 x i64> %in)
+SVINT_TEST(__SVInt64_t,h);
+// CHECK: call <vscale x 2 x i64> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 2 x i64> %in)
+SVINT_TEST(__SVInt64_t,s);
+// CHECK: call <vscale x 2 x i64> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 2 x i64> %in)
+SVINT_TEST(__SVInt64_t,d);
+// CHECK: call <vscale x 2 x i64> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 2 x i64> %in)
+
+
+//Test that floats can also be used as datatypes for integer instructions
+//and check all the variants which would not be possible with a float
+//instruction
+SVINT_TEST(__SVFloat16_t,b);
+// CHECK: call <vscale x 8 x half> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 8 x half> %in)
+SVINT_TEST(__SVFloat16_t,h);
+// CHECK: call <vscale x 8 x half> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 8 x half> %in)
+SVINT_TEST(__SVFloat16_t,s);
+// CHECK: call <vscale x 8 x half> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 8 x half> %in)
+SVINT_TEST(__SVFloat16_t,d);
+// CHECK: call <vscale x 8 x half> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 8 x half> %in)
+
+SVINT_TEST(__SVFloat32_t,b);
+// CHECK: call <vscale x 4 x float> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 4 x float> %in)
+SVINT_TEST(__SVFloat32_t,h);
+// CHECK: call <vscale x 4 x float> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 4 x float> %in)
+SVINT_TEST(__SVFloat32_t,s);
+// CHECK: call <vscale x 4 x float> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 4 x float> %in)
+SVINT_TEST(__SVFloat32_t,d);
+// CHECK: call <vscale x 4 x float> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 4 x float> %in)
+
+SVINT_TEST(__SVFloat64_t,b);
+// CHECK: call <vscale x 2 x double> asm sideeffect "ptrue p0.b\0Amov $0.b, p0/m, $1.b\0A", "=w,w,~{p0}"(<vscale x 2 x double> %in)
+SVINT_TEST(__SVFloat64_t,h);
+// CHECK: call <vscale x 2 x double> asm sideeffect "ptrue p0.b\0Amov $0.h, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 2 x double> %in)
+SVINT_TEST(__SVFloat64_t,s);
+// CHECK: call <vscale x 2 x double> asm sideeffect "ptrue p0.b\0Amov $0.s, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 2 x double> %in)
+SVINT_TEST(__SVFloat64_t,d);
+// CHECK: call <vscale x 2 x double> asm sideeffect "ptrue p0.b\0Amov $0.d, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 2 x double> %in)
+
+
+#define SVBOOL_TEST(KIND)\
+__SVBool_t func_bool_##KIND(__SVBool_t in1, __SVBool_t in2)\
+{\
+  __SVBool_t out;\
+  asm volatile (\
+    "zip1 %[out]." #KIND ", %[in1]." #KIND ", %[in2]." #KIND "\n"\
+    : [out] "=Upa" (out)\
+    :  [in1] "Upa" (in1),\
+      [in2] "Upa" (in2)\
+    :);\
+  return out;\
+}
+
+SVBOOL_TEST(b) ;
+// CHECK: call <vscale x 16 x i1> asm sideeffect "zip1 $0.b, $1.b, $2.b\0A", "=@3Upa,@3Upa,@3Upa"(<vscale x 16 x i1> %in1, <vscale x 16 x i1> %in2)
+SVBOOL_TEST(h) ;
+// CHECK: call <vscale x 16 x i1> asm sideeffect "zip1 $0.h, $1.h, $2.h\0A", "=@3Upa,@3Upa,@3Upa"(<vscale x 16 x i1> %in1, <vscale x 16 x i1> %in2)
+SVBOOL_TEST(s) ;
+// CHECK: call <vscale x 16 x i1> asm sideeffect "zip1 $0.s, $1.s, $2.s\0A", "=@3Upa,@3Upa,@3Upa"(<vscale x 16 x i1> %in1, <vscale x 16 x i1> %in2)
+SVBOOL_TEST(d) ;
+// CHECK: call <vscale x 16 x i1> asm sideeffect "zip1 $0.d, $1.d, $2.d\0A", "=@3Upa,@3Upa,@3Upa"(<vscale x 16 x i1> %in1, <vscale x 16 x i1> %in2)
+
+
+#define SVBOOL_TEST_UPL(DT, KIND)\
+__SVBool_t func_bool_upl_##KIND(__SVBool_t in1, DT in2, DT in3)\
+{\
+  __SVBool_t out;\
+  asm volatile (\
+    "fadd %[out]." #KIND ", %[in1]." #KIND ", %[in2]." #KIND ", %[in3]." #KIND "\n"\
+    : [out] "=w" (out)\
+    :  [in1] "Upl" (in1),\
+      [in2] "w" (in2),\
+      [in3] "w" (in3)\
+    :);\
+  return out;\
+}
+
+SVBOOL_TEST_UPL(__SVInt8_t, b) ;
+// CHECK: call <vscale x 16 x i1> asm sideeffect "fadd $0.b, $1.b, $2.b, $3.b\0A", "=w,@3Upl,w,w"(<vscale x 16 x i1> %in1, <vscale x 16 x i8> %in2, <vscale x 16 x i8> %in3)
+SVBOOL_TEST_UPL(__SVInt16_t, h) ;
+// CHECK: call <vscale x 16 x i1> asm sideeffect "fadd $0.h, $1.h, $2.h, $3.h\0A", "=w,@3Upl,w,w"(<vscale x 16 x i1> %in1, <vscale x 8 x i16> %in2, <vscale x 8 x i16> %in3)
+SVBOOL_TEST_UPL(__SVInt32_t, s) ;
+// CHECK: call <vscale x 16 x i1> asm sideeffect "fadd $0.s, $1.s, $2.s, $3.s\0A", "=w,@3Upl,w,w"(<vscale x 16 x i1> %in1, <vscale x 4 x i32> %in2, <vscale x 4 x i32> %in3)
+SVBOOL_TEST_UPL(__SVInt64_t, d) ;
+// CHECK: call <vscale x 16 x i1> asm sideeffect "fadd $0.d, $1.d, $2.d, $3.d\0A", "=w,@3Upl,w,w"(<vscale x 16 x i1> %in1, <vscale x 2 x i64> %in2, <vscale x 2 x i64> %in3)
+
+#define SVFLOAT_TEST(DT,KIND)\
+DT func_float_##DT##KIND(DT inout1, DT in2)\
+{\
+  asm volatile (\
+    "ptrue p0." #KIND ", #1 \n"\
+    "fsub %[inout1]." #KIND ", p0/m, %[inout1]." #KIND ", %[in2]." #KIND "\n"\
+    : [inout1] "=w" (inout1)\
+    : "[inout1]" (inout1),\
+      [in2] "w" (in2)\
+    : "p0");\
+  return inout1 ;\
+}\
+
+SVFLOAT_TEST(__SVFloat16_t,s);
+// CHECK: call <vscale x 8 x half> asm sideeffect "ptrue p0.s, #1 \0Afsub $0.s, p0/m, $0.s, $2.s\0A", "=w,0,w,~{p0}"(<vscale x 8 x half> %inout1, <vscale x 8 x half> %in2)
+SVFLOAT_TEST(__SVFloat16_t,d);
+// CHECK: call <vscale x 8 x half> asm sideeffect "ptrue p0.d, #1 \0Afsub $0.d, p0/m, $0.d, $2.d\0A", "=w,0,w,~{p0}"(<vscale x 8 x half> %inout1, <vscale x 8 x half> %in2)
+
+SVFLOAT_TEST(__SVFloat32_t,s);
+// CHECK: call <vscale x 4 x float> asm sideeffect "ptrue p0.s, #1 \0Afsub $0.s, p0/m, $0.s, $2.s\0A", "=w,0,w,~{p0}"(<vscale x 4 x float> %inout1, <vscale x 4 x float> %in2)
+SVFLOAT_TEST(__SVFloat32_t,d);
+// CHECK: call <vscale x 4 x float> asm sideeffect "ptrue p0.d, #1 \0Afsub $0.d, p0/m, $0.d, $2.d\0A", "=w,0,w,~{p0}"(<vscale x 4 x float> %inout1, <vscale x 4 x float> %in2)
+
+SVFLOAT_TEST(__SVFloat64_t,s);
+// CHECK: call <vscale x 2 x double> asm sideeffect "ptrue p0.s, #1 \0Afsub $0.s, p0/m, $0.s, $2.s\0A", "=w,0,w,~{p0}"(<vscale x 2 x double> %inout1, <vscale x 2 x double> %in2)
+SVFLOAT_TEST(__SVFloat64_t,d);
+// CHECK: call <vscale x 2 x double> asm sideeffect "ptrue p0.d, #1 \0Afsub $0.d, p0/m, $0.d, $2.d\0A", "=w,0,w,~{p0}"(<vscale x 2 x double> %inout1, <vscale x 2 x double> %in2)
+
+#define SVFLOAT_TEST_Y(DT, KIND)\
+__SVBool_t func_float_y_##KIND(DT in1, DT in2)\
+{\
+  __SVBool_t out;\
+  asm volatile (\
+    "fmul %[out]." #KIND ", %[in1]." #KIND ", %[in2]." #KIND "\n"\
+    : [out] "=w" (out)\
+    :  [in1] "w" (in1),\
+      [in2] "y" (in2)\
+    :);\
+  return out;\
+}
+
+SVFLOAT_TEST_Y(__SVFloat16_t,h);
+// CHECK: call <vscale x 16 x i1> asm sideeffect "fmul $0.h, $1.h, $2.h\0A", "=w,w,y"(<vscale x 8 x half> %in1, <vscale x 8 x half> %in2)
+SVFLOAT_TEST_Y(__SVFloat32_t,s);
+// CHECK: call <vscale x 16 x i1> asm sideeffect "fmul $0.s, $1.s, $2.s\0A", "=w,w,y"(<vscale x 4 x float> %in1, <vscale x 4 x float> %in2)
+SVFLOAT_TEST_Y(__SVFloat64_t,d);
+// CHECK: call <vscale x 16 x i1> asm sideeffect "fmul $0.d, $1.d, $2.d\0A", "=w,w,y"(<vscale x 2 x double> %in1, <vscale x 2 x double> %in2)
+
+
+// Another test for floats to include h suffix
+
+#define SVFLOAT_CVT_TEST(DT1,KIND1,DT2,KIND2)\
+DT1 func_float_cvt_##DT1##KIND1##DT2##KIND2(DT2 in1)\
+{\
+  DT1 out1 ;\
+  asm volatile (\
+    "ptrue p0." #KIND2 ", #1 \n"\
+    "fcvt %[out1]." #KIND1 ", p0/m, %[in1]." #KIND2 "\n"\
+    : [out1] "=w" (out1)\
+    : [in1] "w" (in1)\
+    : "p0");\
+  return out1 ;\
+}\
+
+SVFLOAT_CVT_TEST(__SVFloat64_t,d,__SVFloat32_t,s);
+// CHECK: call <vscale x 2 x double> asm sideeffect "ptrue p0.s, #1 \0Afcvt $0.d, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 4 x float> %in1)
+SVFLOAT_CVT_TEST(__SVFloat64_t,d,__SVFloat16_t,h);
+// CHECK: call <vscale x 2 x double> asm sideeffect "ptrue p0.h, #1 \0Afcvt $0.d, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 8 x half> %in1)
+SVFLOAT_CVT_TEST(__SVFloat32_t,s,__SVFloat16_t,h);
+// CHECK: call <vscale x 4 x float> asm sideeffect "ptrue p0.h, #1 \0Afcvt $0.s, p0/m, $1.h\0A", "=w,w,~{p0}"(<vscale x 8 x half> %in1)
+SVFLOAT_CVT_TEST(__SVFloat32_t,s,__SVFloat64_t,d);
+// CHECK: call <vscale x 4 x float> asm sideeffect "ptrue p0.d, #1 \0Afcvt $0.s, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 2 x double> %in1)
+SVFLOAT_CVT_TEST(__SVFloat16_t,h,__SVFloat64_t,d);
+// CHECK: call <vscale x 8 x half> asm sideeffect "ptrue p0.d, #1 \0Afcvt $0.h, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 2 x double> %in1)
+SVFLOAT_CVT_TEST(__SVFloat16_t,h,__SVFloat32_t,s);
+// CHECK: call <vscale x 8 x half> asm sideeffect "ptrue p0.s, #1 \0Afcvt $0.h, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 4 x float> %in1)
+
+//Test a mix of float and ints
+SVFLOAT_CVT_TEST(__SVInt16_t,h,__SVFloat32_t,s);
+// CHECK: call <vscale x 8 x i16> asm sideeffect "ptrue p0.s, #1 \0Afcvt $0.h, p0/m, $1.s\0A", "=w,w,~{p0}"(<vscale x 4 x float> %in1)
+SVFLOAT_CVT_TEST(__SVFloat16_t,s,__SVUint32_t,d);
+// CHECK: call <vscale x 8 x half> asm sideeffect "ptrue p0.d, #1 \0Afcvt $0.s, p0/m, $1.d\0A", "=w,w,~{p0}"(<vscale x 4 x i32> %in1)
diff --git a/clang/test/CodeGen/aarch64-sve-inline-asm-negative-test.c b/clang/test/CodeGen/aarch64-sve-inline-asm-negative-test.c
new file mode 100644
index 0000000000000..ffe7a41946941
--- /dev/null
+++ b/clang/test/CodeGen/aarch64-sve-inline-asm-negative-test.c
@@ -0,0 +1,21 @@
+// REQUIRES: aarch64-registered-target
+
+// RUN: not %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -fallow-half-arguments-and-returns \
+// RUN:   -target-feature +neon -S -O1 -o - %s | FileCheck %s
+
+// Assembler error
+// Output constraint : Set a vector constraint on an integer
+__SVFloat32_t funcB2()
+{
+  __SVFloat32_t ret ;
+  asm volatile (
+    "fmov %[ret], wzr \n"
+    : [ret] "=w" (ret)
+    :
+    :);
+
+  return ret ;
+}
+
+// CHECK: funcB2
+// CHECK-ERROR: error: invalid operand for instruction