[𝘀𝗽𝗿] changes introduced through rebaseusers/vitalybuka/spr/main.nfcsanitizerasan-promote-stack_id-into-threadcontextbase

Created using spr 1.3.4

[skip ci]

72 files changed, 1975 insertions, 1045 deletions

diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index e48835d47380..df165b912525 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -172,7 +172,7 @@ C++23 Feature Support
 - Removed the restriction to literal types in constexpr functions in C++23 mode.
 
 - Extend lifetime of temporaries in mem-default-init for P2718R0. Clang now fully
-  supported `P2718R0 Lifetime extension in range-based for loops <https://wg21.link/P2718R0>`_.
+  supports `P2718R0 Lifetime extension in range-based for loops <https://wg21.link/P2718R0>`_.
 
 C++20 Feature Support
 ^^^^^^^^^^^^^^^^^^^^^
diff --git a/clang/include/clang/Basic/Builtins.td b/clang/include/clang/Basic/Builtins.td
index 9ebee81fcb0d..7068473a0e12 100644
--- a/clang/include/clang/Basic/Builtins.td
+++ b/clang/include/clang/Basic/Builtins.td
@@ -4745,6 +4745,12 @@ def HLSLCross: LangBuiltin<"HLSL_LANG"> {
   let Prototype = "void(...)";
 }
 
+def HLSLDegrees : LangBuiltin<"HLSL_LANG"> {
+  let Spellings = ["__builtin_hlsl_elementwise_degrees"];
+  let Attributes = [NoThrow, Const];
+  let Prototype = "void(...)";
+}
+
 def HLSLDotProduct : LangBuiltin<"HLSL_LANG"> {
   let Spellings = ["__builtin_hlsl_dot"];
   let Attributes = [NoThrow, Const];
diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index 06140d6d4ce2..ff678ee04f9c 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -18755,6 +18755,16 @@ Value *CodeGenFunction::EmitHLSLBuiltinExpr(unsigned BuiltinID,
         CGM.getHLSLRuntime().getNormalizeIntrinsic(), ArrayRef<Value *>{X},
         nullptr, "hlsl.normalize");
   }
+  case Builtin::BI__builtin_hlsl_elementwise_degrees: {
+    Value *X = EmitScalarExpr(E->getArg(0));
+
+    assert(E->getArg(0)->getType()->hasFloatingRepresentation() &&
+           "degree operand must have a float representation");
+
+    return Builder.CreateIntrinsic(
+        /*ReturnType=*/X->getType(), CGM.getHLSLRuntime().getDegreesIntrinsic(),
+        ArrayRef<Value *>{X}, nullptr, "hlsl.degrees");
+  }
   case Builtin::BI__builtin_hlsl_elementwise_frac: {
     Value *Op0 = EmitScalarExpr(E->getArg(0));
     if (!E->getArg(0)->getType()->hasFloatingRepresentation())
diff --git a/clang/lib/CodeGen/CGHLSLRuntime.h b/clang/lib/CodeGen/CGHLSLRuntime.h
index 05ff325216f5..282fa44af212 100644
--- a/clang/lib/CodeGen/CGHLSLRuntime.h
+++ b/clang/lib/CodeGen/CGHLSLRuntime.h
@@ -75,6 +75,7 @@ public:
   GENERATE_HLSL_INTRINSIC_FUNCTION(All, all)
   GENERATE_HLSL_INTRINSIC_FUNCTION(Any, any)
   GENERATE_HLSL_INTRINSIC_FUNCTION(Cross, cross)
+  GENERATE_HLSL_INTRINSIC_FUNCTION(Degrees, degrees)
   GENERATE_HLSL_INTRINSIC_FUNCTION(Frac, frac)
   GENERATE_HLSL_INTRINSIC_FUNCTION(Length, length)
   GENERATE_HLSL_INTRINSIC_FUNCTION(Lerp, lerp)
diff --git a/clang/lib/Headers/hlsl/hlsl_intrinsics.h b/clang/lib/Headers/hlsl/hlsl_intrinsics.h
index 813f8a317bf6..137467e5a782 100644
--- a/clang/lib/Headers/hlsl/hlsl_intrinsics.h
+++ b/clang/lib/Headers/hlsl/hlsl_intrinsics.h
@@ -767,6 +767,36 @@ _HLSL_BUILTIN_ALIAS(__builtin_elementwise_popcount)
 uint64_t4 countbits(uint64_t4);
 
 //===----------------------------------------------------------------------===//
+// degrees builtins
+//===----------------------------------------------------------------------===//
+
+/// \fn T degrees(T x)
+/// \brief Converts the specified value from radians to degrees.
+/// \param x The specified input value.
+
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_degrees)
+half degrees(half);
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_degrees)
+half2 degrees(half2);
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_degrees)
+half3 degrees(half3);
+_HLSL_16BIT_AVAILABILITY(shadermodel, 6.2)
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_degrees)
+half4 degrees(half4);
+
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_degrees)
+float degrees(float);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_degrees)
+float2 degrees(float2);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_degrees)
+float3 degrees(float3);
+_HLSL_BUILTIN_ALIAS(__builtin_hlsl_elementwise_degrees)
+float4 degrees(float4);
+
+//===----------------------------------------------------------------------===//
 // dot product builtins
 //===----------------------------------------------------------------------===//
 
diff --git a/clang/lib/Sema/SemaHLSL.cpp b/clang/lib/Sema/SemaHLSL.cpp
index b0acbbbbb2b1..137b15c8fcfe 100644
--- a/clang/lib/Sema/SemaHLSL.cpp
+++ b/clang/lib/Sema/SemaHLSL.cpp
@@ -1896,6 +1896,7 @@ bool SemaHLSL::CheckBuiltinFunctionCall(unsigned BuiltinID, CallExpr *TheCall) {
       return true;
     break;
   }
+  case Builtin::BI__builtin_hlsl_elementwise_degrees:
   case Builtin::BI__builtin_hlsl_elementwise_radians:
   case Builtin::BI__builtin_hlsl_elementwise_rsqrt:
   case Builtin::BI__builtin_hlsl_elementwise_frac: {
diff --git a/clang/test/CodeGenHLSL/builtins/degrees.hlsl b/clang/test/CodeGenHLSL/builtins/degrees.hlsl
new file mode 100644
index 000000000000..9e131f4badc1
--- /dev/null
+++ b/clang/test/CodeGenHLSL/builtins/degrees.hlsl
@@ -0,0 +1,64 @@
+// RUN: %clang_cc1 -finclude-default-header -triple \
+// RUN:   dxil-pc-shadermodel6.3-library %s -fnative-half-type \
+// RUN:   -emit-llvm -disable-llvm-passes -o - | FileCheck %s \
+// RUN:   --check-prefixes=CHECK,NATIVE_HALF \
+// RUN:   -DFNATTRS=noundef -DTARGET=dx
+// RUN: %clang_cc1 -finclude-default-header -triple \
+// RUN:   dxil-pc-shadermodel6.3-library %s -emit-llvm -disable-llvm-passes \
+// RUN:   -o - | FileCheck %s --check-prefixes=CHECK,NO_HALF \
+// RUN:   -DFNATTRS=noundef -DTARGET=dx
+// RUN: %clang_cc1 -finclude-default-header -triple \
+// RUN:   spirv-unknown-vulkan-compute %s -fnative-half-type \
+// RUN:   -emit-llvm -disable-llvm-passes -o - | FileCheck %s \
+// RUN:   --check-prefixes=CHECK,NATIVE_HALF \
+// RUN:   -DFNATTRS="spir_func noundef" -DTARGET=spv
+// RUN: %clang_cc1 -finclude-default-header -triple \
+// RUN:   spirv-unknown-vulkan-compute %s -emit-llvm -disable-llvm-passes \
+// RUN:   -o - | FileCheck %s --check-prefixes=CHECK,NO_HALF \
+// RUN:   -DFNATTRS="spir_func noundef" -DTARGET=spv
+
+// NATIVE_HALF: define [[FNATTRS]] half @
+// NATIVE_HALF: %hlsl.degrees = call half @llvm.[[TARGET]].degrees.f16(
+// NATIVE_HALF: ret half %hlsl.degrees
+// NO_HALF: define [[FNATTRS]] float @
+// NO_HALF: %hlsl.degrees = call float @llvm.[[TARGET]].degrees.f32(
+// NO_HALF: ret float %hlsl.degrees
+half test_degrees_half(half p0) { return degrees(p0); }
+// NATIVE_HALF: define [[FNATTRS]] <2 x half> @
+// NATIVE_HALF: %hlsl.degrees = call <2 x half> @llvm.[[TARGET]].degrees.v2f16
+// NATIVE_HALF: ret <2 x half> %hlsl.degrees
+// NO_HALF: define [[FNATTRS]] <2 x float> @
+// NO_HALF: %hlsl.degrees = call <2 x float> @llvm.[[TARGET]].degrees.v2f32(
+// NO_HALF: ret <2 x float> %hlsl.degrees
+half2 test_degrees_half2(half2 p0) { return degrees(p0); }
+// NATIVE_HALF: define [[FNATTRS]] <3 x half> @
+// NATIVE_HALF: %hlsl.degrees = call <3 x half> @llvm.[[TARGET]].degrees.v3f16
+// NATIVE_HALF: ret <3 x half> %hlsl.degrees
+// NO_HALF: define [[FNATTRS]] <3 x float> @
+// NO_HALF: %hlsl.degrees = call <3 x float> @llvm.[[TARGET]].degrees.v3f32(
+// NO_HALF: ret <3 x float> %hlsl.degrees
+half3 test_degrees_half3(half3 p0) { return degrees(p0); }
+// NATIVE_HALF: define [[FNATTRS]] <4 x half> @
+// NATIVE_HALF: %hlsl.degrees = call <4 x half> @llvm.[[TARGET]].degrees.v4f16
+// NATIVE_HALF: ret <4 x half> %hlsl.degrees
+// NO_HALF: define [[FNATTRS]] <4 x float> @
+// NO_HALF: %hlsl.degrees = call <4 x float> @llvm.[[TARGET]].degrees.v4f32(
+// NO_HALF: ret <4 x float> %hlsl.degrees
+half4 test_degrees_half4(half4 p0) { return degrees(p0); }
+
+// CHECK: define [[FNATTRS]] float @
+// CHECK: %hlsl.degrees = call float @llvm.[[TARGET]].degrees.f32(
+// CHECK: ret float %hlsl.degrees
+float test_degrees_float(float p0) { return degrees(p0); }
+// CHECK: define [[FNATTRS]] <2 x float> @
+// CHECK: %hlsl.degrees = call <2 x float> @llvm.[[TARGET]].degrees.v2f32
+// CHECK: ret <2 x float> %hlsl.degrees
+float2 test_degrees_float2(float2 p0) { return degrees(p0); }
+// CHECK: define [[FNATTRS]] <3 x float> @
+// CHECK: %hlsl.degrees = call <3 x float> @llvm.[[TARGET]].degrees.v3f32
+// CHECK: ret <3 x float> %hlsl.degrees
+float3 test_degrees_float3(float3 p0) { return degrees(p0); }
+// CHECK: define [[FNATTRS]] <4 x float> @
+// CHECK: %hlsl.degrees = call <4 x float> @llvm.[[TARGET]].degrees.v4f32
+// CHECK: ret <4 x float> %hlsl.degrees
+float4 test_degrees_float4(float4 p0) { return degrees(p0); }
diff --git a/clang/test/SemaHLSL/BuiltIns/degrees-errors.hlsl b/clang/test/SemaHLSL/BuiltIns/degrees-errors.hlsl
new file mode 100644
index 000000000000..9e981f697357
--- /dev/null
+++ b/clang/test/SemaHLSL/BuiltIns/degrees-errors.hlsl
@@ -0,0 +1,26 @@
+// RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.6-library %s -emit-llvm-only -disable-llvm-passes -verify
+
+float test_too_few_arg() {
+  return __builtin_hlsl_elementwise_degrees();
+  // expected-error@-1 {{too few arguments to function call, expected 1, have 0}}
+}
+
+float2 test_too_many_arg(float2 p0) {
+  return __builtin_hlsl_elementwise_degrees(p0, p0);
+  // expected-error@-1 {{too many arguments to function call, expected 1, have 2}}
+}
+
+float builtin_bool_to_float_type_promotion(bool p1) {
+  return __builtin_hlsl_elementwise_degrees(p1);
+  // expected-error@-1 {{passing 'bool' to parameter of incompatible type 'float'}}
+}
+
+float builtin_degrees_int_to_float_promotion(int p1) {
+  return __builtin_hlsl_elementwise_degrees(p1);
+  // expected-error@-1 {{passing 'int' to parameter of incompatible type 'float'}}
+}
+
+float2 builtin_degrees_int2_to_float2_promotion(int2 p1) {
+  return __builtin_hlsl_elementwise_degrees(p1);
+  // expected-error@-1 {{passing 'int2' (aka 'vector<int, 2>') to parameter of incompatible type '__attribute__((__vector_size__(2 * sizeof(float)))) float' (vector of 2 'float' values)}}
+}
diff --git a/clang/test/SemaHLSL/BuiltIns/half-float-only-errors.hlsl b/clang/test/SemaHLSL/BuiltIns/half-float-only-errors.hlsl
index 2cecf7aeb00e..cdd130052b6a 100644
--- a/clang/test/SemaHLSL/BuiltIns/half-float-only-errors.hlsl
+++ b/clang/test/SemaHLSL/BuiltIns/half-float-only-errors.hlsl
@@ -17,6 +17,7 @@
 // RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only -disable-llvm-passes -verify -DTEST_FUNC=__builtin_elementwise_tan
 // RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only -disable-llvm-passes -verify -DTEST_FUNC=__builtin_elementwise_tanh
 // RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only -disable-llvm-passes -verify -DTEST_FUNC=__builtin_elementwise_trunc
+// RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only -disable-llvm-passes -verify -DTEST_FUNC=__builtin_hlsl_elementwise_degrees
 // RUN: %clang_cc1 -finclude-default-header -triple dxil-pc-shadermodel6.6-library %s -fnative-half-type -emit-llvm-only -disable-llvm-passes -verify -DTEST_FUNC=__builtin_hlsl_elementwise_radians
 
 double test_double_builtin(double p0) {
diff --git a/compiler-rt/lib/asan/asan_descriptions.cpp b/compiler-rt/lib/asan/asan_descriptions.cpp
index 778e134f1cf2..674fe9c1e90b 100644
--- a/compiler-rt/lib/asan/asan_descriptions.cpp
+++ b/compiler-rt/lib/asan/asan_descriptions.cpp
@@ -48,9 +48,20 @@ void DescribeThread(AsanThreadContext *context) {
     return;
   }
   context->announced = true;
+
+  AsanThreadContext *parent_context =
+      context->parent_tid == kInvalidTid
+          ? nullptr
+          : GetThreadContextByTidLocked(context->parent_tid);
+
+  // `context->parent_tid` may point to reused slot. Check `unique_id` which
+  // is always smaller for the parent, always greater for a new user.
+  if (context->unique_id <= parent_context->unique_id)
+    parent_context = nullptr;
+
   InternalScopedString str;
   str.AppendF("Thread %s", AsanThreadIdAndName(context).c_str());
-  if (context->parent_tid == kInvalidTid) {
+  if (!parent_context) {
     str.Append(" created by unknown thread\n");
     Printf("%s", str.data());
     return;
@@ -60,14 +71,8 @@ void DescribeThread(AsanThreadContext *context) {
   Printf("%s", str.data());
   StackDepotGet(context->stack_id).Print();
   // Recursively described parent thread if needed.
-  if (flags()->print_full_thread_history) {
-    AsanThreadContext *parent_context =
-        GetThreadContextByTidLocked(context->parent_tid);
-    // `context->parent_tid` may point to reused slot, double check, `unique_id`
-    // of new user will always be greater then the child.
-    if (context->unique_id > parent_context->unique_id)
-      DescribeThread(parent_context);
-  }
+  if (flags()->print_full_thread_history)
+    DescribeThread(parent_context);
 }
 
 // Shadow descriptions
diff --git a/lld/ELF/SymbolTable.cpp b/lld/ELF/SymbolTable.cpp
index db8ee8f4d7b3..b9ef28f0436f 100644
--- a/lld/ELF/SymbolTable.cpp
+++ b/lld/ELF/SymbolTable.cpp
@@ -309,13 +309,43 @@ void SymbolTable::scanVersionScript() {
 
   // Then, assign versions to "*". In GNU linkers they have lower priority than
   // other wildcards.
+  bool globalAsteriskFound = false;
+  bool localAsteriskFound = false;
+  bool asteriskReported = false;
+  auto assignAsterisk = [&](SymbolVersion &pat, VersionDefinition *ver,
+                            bool isLocal) {
+    // Avoid issuing a warning if both '--retain-symbol-file' and a version
+    // script with `global: *` are used.
+    //
+    // '--retain-symbol-file' adds a "*" pattern to
+    // 'config->versionDefinitions[VER_NDX_LOCAL].nonLocalPatterns', see
+    // 'readConfigs()' in 'Driver.cpp'. Note that it is not '.localPatterns',
+    // and may seem counterintuitive, but still works as expected. Here we can
+    // exploit that and skip analyzing the pattern added for this option.
+    if (!asteriskReported && (isLocal || ver->id > VER_NDX_LOCAL)) {
+      if ((isLocal && globalAsteriskFound) ||
+          (!isLocal && localAsteriskFound)) {
+        warn("wildcard pattern '*' is used for both 'local' and 'global' "
+             "scopes in version script");
+        asteriskReported = true;
+      } else if (!isLocal && globalAsteriskFound) {
+        warn("wildcard pattern '*' is used for multiple version definitions in "
+             "version script");
+        asteriskReported = true;
+      } else {
+        localAsteriskFound = isLocal;
+        globalAsteriskFound = !isLocal;
+      }
+    }
+    assignWildcard(pat, isLocal ? VER_NDX_LOCAL : ver->id, ver->name);
+  };
   for (VersionDefinition &v : llvm::reverse(ctx.arg.versionDefinitions)) {
     for (SymbolVersion &pat : v.nonLocalPatterns)
       if (pat.hasWildcard && pat.name == "*")
-        assignWildcard(pat, v.id, v.name);
+        assignAsterisk(pat, &v, false);
     for (SymbolVersion &pat : v.localPatterns)
       if (pat.hasWildcard && pat.name == "*")
-        assignWildcard(pat, VER_NDX_LOCAL, v.name);
+        assignAsterisk(pat, &v, true);
   }
 
   // Symbol themselves might know their versions because symbols
diff --git a/lld/test/ELF/version-script-reassign-glob.s b/lld/test/ELF/version-script-reassign-glob.s
index 39d19a26fc44..8de36467bd8e 100644
--- a/lld/test/ELF/version-script-reassign-glob.s
+++ b/lld/test/ELF/version-script-reassign-glob.s
@@ -10,7 +10,8 @@
 # RUN: llvm-readelf --dyn-syms %t.so | FileCheck --check-prefix=BAR %s
 
 # RUN: echo 'bar1 { *; }; bar2 { *; };' > %t2.ver
-# RUN: ld.lld --version-script %t2.ver %t.o -shared -o %t2.so --fatal-warnings
+# RUN: ld.lld --version-script %t2.ver %t.o -shared -o %t2.so 2>&1 | \
+# RUN:   FileCheck --check-prefix=DUPWARN %s
 # RUN: llvm-readelf --dyn-syms %t2.so | FileCheck --check-prefix=BAR2 %s
 
 ## If both a non-* glob and a * match, non-* wins.
@@ -21,6 +22,7 @@
 
 ## When there are multiple * patterns, the last wins.
 # BAR2: GLOBAL DEFAULT 7 foo@@bar2
+# DUPWARN: warning: wildcard pattern '*' is used for multiple version definitions in version script
 
 .globl foo
 foo:
diff --git a/lld/test/ELF/version-script-warn.s b/lld/test/ELF/version-script-warn.s
new file mode 100644
index 000000000000..9aba59616579
--- /dev/null
+++ b/lld/test/ELF/version-script-warn.s
@@ -0,0 +1,35 @@
+# REQUIRES: x86
+# RUN: llvm-mc -filetype=obj -triple=x86_64 %s -o %t.o
+
+# RUN: echo 'foo { *; }; bar { *; };' > %t.ver
+# RUN: ld.lld --version-script %t.ver %t.o -shared -o %t.so 2>&1 | \
+# RUN:   FileCheck --check-prefix=MULTVER %s
+
+# RUN: echo '{ global: *; local: *;};' > %t.ver
+# RUN: ld.lld --version-script %t.ver %t.o -shared -o %t.so 2>&1 | \
+# RUN:   FileCheck --check-prefix=LOCGLOB %s
+
+# RUN: echo 'V1 { global: *; }; V2 { local: *;};' > %t.ver
+# RUN: ld.lld --version-script %t.ver %t.o -shared -o %t.so 2>&1 | \
+# RUN:   FileCheck --check-prefix=LOCGLOB %s
+
+# RUN: echo 'V1 { local: *; }; V2 { global: *;};' > %t.ver
+# RUN: ld.lld --version-script %t.ver %t.o -shared -o %t.so 2>&1 | \
+# RUN:   FileCheck --check-prefix=LOCGLOB %s
+
+# RUN: echo 'V1 { local: *; }; V2 { local: *;};' > %t.ver
+# RUN: ld.lld --version-script %t.ver %t.o -shared -o %t.so --fatal-warnings
+
+## --retain-symbols-file uses the same internal infrastructure as the support
+## for version scripts. Do not show the warings if they both are used.
+# RUN: echo 'foo' > %t_retain.txt
+# RUN: echo '{ local: *; };' > %t_local.ver
+# RUN: echo '{ global: *; };' > %t_global.ver
+# RUN: ld.lld --retain-symbols-file=%t_retain.txt --version-script %t_local.ver %t.o -shared -o %t.so --fatal-warnings
+# RUN: ld.lld --retain-symbols-file=%t_retain.txt --version-script %t_global.ver %t.o -shared -o %t.so --fatal-warnings
+
+# MULTVER: warning: wildcard pattern '*' is used for multiple version definitions in version script
+# LOCGLOB: warning: wildcard pattern '*' is used for both 'local' and 'global' scopes in version script
+
+.globl foo
+foo:
diff --git a/lldb/include/lldb/API/SBProcess.h b/lldb/include/lldb/API/SBProcess.h
index 8b8ed830b54c..1624e02070b1 100644
--- a/lldb/include/lldb/API/SBProcess.h
+++ b/lldb/include/lldb/API/SBProcess.h
@@ -159,7 +159,6 @@ public:
   lldb::SBError Destroy();
 
   lldb::SBError Continue();
-  lldb::SBError Continue(RunDirection direction);
 
   lldb::SBError Stop();
 
diff --git a/lldb/include/lldb/Target/DynamicLoader.h b/lldb/include/lldb/Target/DynamicLoader.h
index 0629e2faae7e..75bb6cb6bb90 100644
--- a/lldb/include/lldb/Target/DynamicLoader.h
+++ b/lldb/include/lldb/Target/DynamicLoader.h
@@ -11,6 +11,7 @@
 
 #include "lldb/Core/Address.h"
 #include "lldb/Core/PluginInterface.h"
+#include "lldb/Target/CoreFileMemoryRanges.h"
 #include "lldb/Utility/FileSpec.h"
 #include "lldb/Utility/Status.h"
 #include "lldb/Utility/UUID.h"
@@ -337,6 +338,17 @@ public:
     return std::nullopt;
   }
 
+  /// Returns a list of memory ranges that should be saved in the core file,
+  /// specific for this dynamic loader.
+  ///
+  /// For example, an implementation of this function can save the thread
+  /// local data of a given thread.
+  virtual void CalculateDynamicSaveCoreRanges(
+      lldb_private::Process &process,
+      std::vector<lldb_private::MemoryRegionInfo> &ranges,
+      llvm::function_ref<bool(const lldb_private::Thread &)>
+          save_thread_predicate) {};
+
 protected:
   // Utility methods for derived classes
 
diff --git a/lldb/include/lldb/Target/Process.h b/lldb/include/lldb/Target/Process.h
index fe7fbc50fd57..b8c53a474ba6 100644
--- a/lldb/include/lldb/Target/Process.h
+++ b/lldb/include/lldb/Target/Process.h
@@ -857,10 +857,10 @@ public:
   /// \see Thread:Resume()
   /// \see Thread:Step()
   /// \see Thread:Suspend()
-  Status Resume(lldb::RunDirection direction = lldb::eRunForward);
+  Status Resume();
 
   /// Resume a process, and wait for it to stop.
-  Status ResumeSynchronous(Stream *stream, lldb::RunDirection direction = lldb::eRunForward);
+  Status ResumeSynchronous(Stream *stream);
 
   /// Halts a running process.
   ///
@@ -1104,14 +1104,9 @@ public:
   /// \see Thread:Resume()
   /// \see Thread:Step()
   /// \see Thread:Suspend()
-  virtual Status DoResume(lldb::RunDirection direction) {
-    if (direction == lldb::RunDirection::eRunForward) {
-      return Status::FromErrorStringWithFormatv(
-          "error: {0} does not support resuming processes", GetPluginName());
-    } else {
-      return Status::FromErrorStringWithFormatv(
-          "error: {0} does not support reverse execution of processes", GetPluginName());
-    }
+  virtual Status DoResume() {
+    return Status::FromErrorStringWithFormatv(
+        "error: {0} does not support resuming processes", GetPluginName());
   }
 
   /// Called after resuming a process.
@@ -2337,8 +2332,6 @@ public:
 
   bool IsRunning() const;
 
-  lldb::RunDirection GetLastRunDirection() { return m_last_run_direction; }
-
   DynamicCheckerFunctions *GetDynamicCheckers() {
     return m_dynamic_checkers_up.get();
   }
@@ -2858,7 +2851,7 @@ protected:
   ///
   /// \return
   ///     An Status object describing the success or failure of the resume.
-  Status PrivateResume(lldb::RunDirection direction = lldb::eRunForward);
+  Status PrivateResume();
 
   // Called internally
   void CompleteAttach();
@@ -3134,8 +3127,6 @@ protected:
                            // m_currently_handling_do_on_removals are true,
                            // Resume will only request a resume, using this
                            // flag to check.
-  // The direction of execution from the last time this process was resumed.
-  lldb::RunDirection m_last_run_direction;
 
   lldb::tid_t m_interrupt_tid; /// The tid of the thread that issued the async
                                /// interrupt, used by thread plan timeout. It
diff --git a/lldb/include/lldb/Target/StopInfo.h b/lldb/include/lldb/Target/StopInfo.h
index 072f71f6b112..fae90364deaf 100644
--- a/lldb/include/lldb/Target/StopInfo.h
+++ b/lldb/include/lldb/Target/StopInfo.h
@@ -142,12 +142,6 @@ public:
   static lldb::StopInfoSP
   CreateStopReasonProcessorTrace(Thread &thread, const char *description);
 
-  // This creates a StopInfo indicating that execution stopped because
-  // it was replaying some recorded execution history, and execution reached
-  // the end of that recorded history.
-  static lldb::StopInfoSP
-  CreateStopReasonHistoryBoundary(Thread &thread, const char *description);
-
   static lldb::StopInfoSP CreateStopReasonFork(Thread &thread,
                                                lldb::pid_t child_pid,
                                                lldb::tid_t child_tid);
diff --git a/lldb/include/lldb/lldb-enumerations.h b/lldb/include/lldb/lldb-enumerations.h
index 232d1dfdb5c9..938f6e3abe8f 100644
--- a/lldb/include/lldb/lldb-enumerations.h
+++ b/lldb/include/lldb/lldb-enumerations.h
@@ -135,9 +135,6 @@ FLAGS_ENUM(LaunchFlags){
 /// Thread Run Modes.
 enum RunMode { eOnlyThisThread, eAllThreads, eOnlyDuringStepping };
 
-/// Execution directions
-enum RunDirection { eRunForward, eRunReverse };
-
 /// Byte ordering definitions.
 enum ByteOrder {
   eByteOrderInvalid = 0,
@@ -257,9 +254,6 @@ enum StopReason {
   eStopReasonVFork,
   eStopReasonVForkDone,
   eStopReasonInterrupt, ///< Thread requested interrupt
-  // Indicates that execution stopped because the debugger backend relies
-  // on recorded data and we reached the end of that data.
-  eStopReasonHistoryBoundary,
 };
 
 /// Command Return Status Types.
diff --git a/lldb/packages/Python/lldbsuite/test/gdbclientutils.py b/lldb/packages/Python/lldbsuite/test/gdbclientutils.py
index 732d61713206..1784487323ad 100644
--- a/lldb/packages/Python/lldbsuite/test/gdbclientutils.py
+++ b/lldb/packages/Python/lldbsuite/test/gdbclientutils.py
@@ -510,9 +510,8 @@ class MockGDBServer:
         self._thread.start()
 
     def stop(self):
-        if self._thread is not None:
-            self._thread.join()
-            self._thread = None
+        self._thread.join()
+        self._thread = None
 
     def get_connect_address(self):
         return self._socket.get_connect_address()
diff --git a/lldb/packages/Python/lldbsuite/test/lldbgdbproxy.py b/lldb/packages/Python/lldbsuite/test/lldbgdbproxy.py
deleted file mode 100644
index 2a9592bf4545..000000000000
--- a/lldb/packages/Python/lldbsuite/test/lldbgdbproxy.py
+++ /dev/null
@@ -1,175 +0,0 @@
-import logging
-import os
-import os.path
-import random
-
-import lldb
-from lldbsuite.test.lldbtest import *
-from lldbsuite.test.gdbclientutils import *
-import lldbgdbserverutils
-from lldbsuite.support import seven
-
-
-class GDBProxyTestBase(TestBase):
-    """
-    Base class for gdbserver proxy tests.
-
-    This class will setup and start a mock GDB server for the test to use.
-    It pases through requests to a regular lldb-server/debugserver and
-    forwards replies back to the LLDB under test.
-    """
-
-    """The gdbserver that we implement."""
-    server = None
-    """The inner lldb-server/debugserver process that we proxy requests into."""
-    monitor_server = None
-    monitor_sock = None
-
-    server_socket_class = TCPServerSocket
-
-    DEFAULT_TIMEOUT = 20 * (10 if ("ASAN_OPTIONS" in os.environ) else 1)
-
-    _verbose_log_handler = None
-    _log_formatter = logging.Formatter(fmt="%(asctime)-15s %(levelname)-8s %(message)s")
-
-    def setUpBaseLogging(self):
-        self.logger = logging.getLogger(__name__)
-
-        if len(self.logger.handlers) > 0:
-            return  # We have set up this handler already
-
-        self.logger.propagate = False
-        self.logger.setLevel(logging.DEBUG)
-
-        # log all warnings to stderr
-        handler = logging.StreamHandler()
-        handler.setLevel(logging.WARNING)
-        handler.setFormatter(self._log_formatter)
-        self.logger.addHandler(handler)
-
-    def setUp(self):
-        TestBase.setUp(self)
-
-        self.setUpBaseLogging()
-
-        if self.isVerboseLoggingRequested():
-            # If requested, full logs go to a log file
-            log_file_name = self.getLogBasenameForCurrentTest() + "-proxy.log"
-            self._verbose_log_handler = logging.FileHandler(
-               log_file_name
-            )
-            self._verbose_log_handler.setFormatter(self._log_formatter)
-            self._verbose_log_handler.setLevel(logging.DEBUG)
-            self.logger.addHandler(self._verbose_log_handler)
-
-        lldb_server_exe = lldbgdbserverutils.get_lldb_server_exe()
-        if lldb_server_exe is None:
-            self.debug_monitor_exe = lldbgdbserverutils.get_debugserver_exe()
-            self.assertTrue(self.debug_monitor_exe is not None)
-            self.debug_monitor_extra_args = []
-        else:
-            self.debug_monitor_exe = lldb_server_exe
-            self.debug_monitor_extra_args = ["gdbserver"]
-
-        self.server = MockGDBServer(self.server_socket_class())
-        self.server.responder = self
-
-    def tearDown(self):
-        # TestBase.tearDown will kill the process, but we need to kill it early
-        # so its client connection closes and we can stop the server before
-        # finally calling the base tearDown.
-        if self.process() is not None:
-            self.process().Kill()
-        self.server.stop()
-
-        self.logger.removeHandler(self._verbose_log_handler)
-        self._verbose_log_handler = None
-
-        TestBase.tearDown(self)
-
-    def isVerboseLoggingRequested(self):
-        # We will report our detailed logs if the user requested that the "gdb-remote" channel is
-        # logged.
-        return any(("gdb-remote" in channel) for channel in lldbtest_config.channels)
-
-    def connect(self, target):
-        """
-        Create a process by connecting to the mock GDB server.
-        """
-        self.prep_debug_monitor_and_inferior()
-        self.server.start()
-
-        listener = self.dbg.GetListener()
-        error = lldb.SBError()
-        process = target.ConnectRemote(
-            listener, self.server.get_connect_url(), "gdb-remote", error
-        )
-        self.assertTrue(error.Success(), error.description)
-        self.assertTrue(process, PROCESS_IS_VALID)
-        return process
-
-    def get_next_port(self):
-        return 12000 + random.randint(0, 3999)
-
-    def prep_debug_monitor_and_inferior(self):
-        inferior_exe_path = self.getBuildArtifact("a.out")
-        self.connect_to_debug_monitor([inferior_exe_path])
-        self.assertIsNotNone(self.monitor_server)
-        self.initial_handshake()
-
-    def initial_handshake(self):
-        self.monitor_server.send_packet(seven.bitcast_to_bytes("+"))
-        reply = seven.bitcast_to_string(self.monitor_server.get_normal_packet())
-        self.assertEqual(reply, "+")
-        self.monitor_server.send_packet(seven.bitcast_to_bytes("QStartNoAckMode"))
-        reply = seven.bitcast_to_string(self.monitor_server.get_normal_packet())
-        self.assertEqual(reply, "+")
-        reply = seven.bitcast_to_string(self.monitor_server.get_normal_packet())
-        self.assertEqual(reply, "OK")
-        self.monitor_server.send_packet(seven.bitcast_to_bytes("+"))
-        reply = seven.bitcast_to_string(self.monitor_server.get_normal_packet())
-        self.assertEqual(reply, "+")
-
-    def get_debug_monitor_command_line_args(self, connect_address, launch_args):
-        return self.debug_monitor_extra_args + ["--reverse-connect", connect_address] + launch_args
-
-    def launch_debug_monitor(self, launch_args):
-        family, type, proto, _, addr = socket.getaddrinfo(
-            "localhost", 0, proto=socket.IPPROTO_TCP
-        )[0]
-        sock = socket.socket(family, type, proto)
-        sock.settimeout(self.DEFAULT_TIMEOUT)
-        sock.bind(addr)
-        sock.listen(1)
-        addr = sock.getsockname()
-        connect_address = "[{}]:{}".format(*addr)
-
-        commandline_args = self.get_debug_monitor_command_line_args(
-            connect_address, launch_args
-        )
-
-        # Start the server.
-        self.logger.info(f"Spawning monitor {commandline_args}")
-        monitor_process = self.spawnSubprocess(
-            self.debug_monitor_exe, commandline_args, install_remote=False
-        )
-        self.assertIsNotNone(monitor_process)
-
-        self.monitor_sock = sock.accept()[0]
-        self.monitor_sock.settimeout(self.DEFAULT_TIMEOUT)
-        return monitor_process
-
-    def connect_to_debug_monitor(self, launch_args):
-        monitor_process = self.launch_debug_monitor(launch_args)
-        self.monitor_server = lldbgdbserverutils.Server(self.monitor_sock, monitor_process)
-
-    def respond(self, packet):
-        """Subclasses can override this to change how packets are handled."""
-        return self.pass_through(packet)
-
-    def pass_through(self, packet):
-        self.logger.info(f"Sending packet {packet}")
-        self.monitor_server.send_packet(seven.bitcast_to_bytes(packet))
-        reply = seven.bitcast_to_string(self.monitor_server.get_normal_packet())
-        self.logger.info(f"Received reply {reply}")
-        return reply
diff --git a/lldb/packages/Python/lldbsuite/test/lldbreverse.py b/lldb/packages/Python/lldbsuite/test/lldbreverse.py
deleted file mode 100644
index 0f02fdffbdea..000000000000
--- a/lldb/packages/Python/lldbsuite/test/lldbreverse.py
+++ /dev/null
@@ -1,418 +0,0 @@
-import os
-import os.path
-import lldb
-from lldbsuite.test.lldbtest import *
-from lldbsuite.test.gdbclientutils import *
-from lldbsuite.test.lldbgdbproxy import *
-import lldbgdbserverutils
-import re
-
-
-class ThreadSnapshot:
-    def __init__(self, thread_id, registers):
-        self.thread_id = thread_id
-        self.registers = registers
-
-
-class MemoryBlockSnapshot:
-    def __init__(self, address, data):
-        self.address = address
-        self.data = data
-
-
-class StateSnapshot:
-    def __init__(self, thread_snapshots, memory):
-        self.thread_snapshots = thread_snapshots
-        self.memory = memory
-        self.thread_id = None
-
-
-class RegisterInfo:
-    def __init__(self, lldb_index, bitsize, little_endian):
-        self.lldb_index = lldb_index
-        self.bitsize = bitsize
-        self.little_endian = little_endian
-
-
-BELOW_STACK_POINTER = 16384
-ABOVE_STACK_POINTER = 4096
-
-BLOCK_SIZE = 1024
-
-SOFTWARE_BREAKPOINTS = 0
-HARDWARE_BREAKPOINTS = 1
-WRITE_WATCHPOINTS = 2
-
-
-class ReverseTestBase(GDBProxyTestBase):
-    """
-    Base class for tests that need reverse execution.
-
-    This class uses a gdbserver proxy to add very limited reverse-
-    execution capability to lldb-server/debugserver for testing
-    purposes only.
-
-    To use this class, run the inferior forward until some stopping point.
-    Then call `start_recording()` and execute forward again until reaching
-    a software breakpoint; this class records the state before each execution executes.
-    At that point, the server will accept "bc" and "bs" packets to step
-    backwards through the state.
-    When executing during recording, we only allow single-step and continue without
-    delivering a signal, and only software breakpoint stops are allowed.
-
-    We assume that while recording is enabled, the only effects of instructions
-    are on general-purpose registers (read/written by the 'g' and 'G' packets)
-    and on memory bytes between [SP - BELOW_STACK_POINTER, SP + ABOVE_STACK_POINTER).
-    """
-
-    """
-    A list of StateSnapshots in time order.
-
-    There is one snapshot per single-stepped instruction,
-    representing the state before that instruction was
-    executed. The last snapshot in the list is the
-    snapshot before the last instruction was executed.
-    This is an undo log; we snapshot a superset of the state that may have
-    been changed by the instruction's execution.
-    """
-    snapshots = None
-    recording_enabled = False
-
-    breakpoints = None
-
-    pid = None
-
-    pc_register_info = None
-    sp_register_info = None
-    general_purpose_register_info = None
-
-    def __init__(self, *args, **kwargs):
-        GDBProxyTestBase.__init__(self, *args, **kwargs)
-        self.breakpoints = [set(), set(), set(), set(), set()]
-
-    def respond(self, packet):
-        if not packet:
-            raise ValueError("Invalid empty packet")
-        if packet == self.server.PACKET_INTERRUPT:
-            # Don't send a response. We'll just run to completion.
-            return []
-        if self.is_command(packet, "qSupported", ":"):
-            reply = self.pass_through(packet)
-            return reply + ";ReverseStep+;ReverseContinue+"
-        if self.is_command(packet, "vCont", ";"):
-            if self.recording_enabled:
-                return self.continue_with_recording(packet)
-            snapshots = []
-        if packet[0] == "c" or packet[0] == "s" or packet[0] == "C" or packet[0] == "S":
-            raise ValueError("LLDB should not be sending old-style continuation packets")
-        if packet == "bc":
-            return self.reverse_continue()
-        if packet == "bs":
-            return self.reverse_step()
-        if packet == 'jThreadsInfo':
-            # Suppress this because it contains thread stop reasons which we might
-            # need to modify, and we don't want to have to implement that.
-            return ""
-        if packet[0] == "z" or packet[0] == "Z":
-            reply = self.pass_through(packet)
-            if reply == "OK":
-                self.update_breakpoints(packet)
-            return reply
-        return GDBProxyTestBase.respond(self, packet)
-
-    def start_recording(self):
-        self.recording_enabled = True
-        self.snapshots = []
-
-    def stop_recording(self):
-        """
-        Don't record when executing foward.
-
-        Reverse execution is still supported until the next forward continue.
-        """
-        self.recording_enabled = False
-
-    def is_command(self, packet, cmd, follow_token):
-        return packet == cmd or packet[0:len(cmd) + 1] == cmd + follow_token
-
-    def update_breakpoints(self, packet):
-        m = re.match("([zZ])([01234]),([0-9a-f]+),([0-9a-f]+)", packet)
-        if m is None:
-            raise ValueError("Invalid breakpoint packet: " + packet)
-        t = int(m.group(2))
-        addr = int(m.group(3), 16)
-        kind = int(m.group(4), 16)
-        if m.group(1) == 'Z':
-            self.breakpoints[t].add((addr, kind))
-        else:
-            self.breakpoints[t].discard((addr, kind))
-
-    def breakpoint_triggered_at(self, pc):
-        if any(addr == pc for addr, kind in self.breakpoints[SOFTWARE_BREAKPOINTS]):
-            return True
-        if any(addr == pc for addr, kind in self.breakpoints[HARDWARE_BREAKPOINTS]):
-            return True
-        return False
-
-    def watchpoint_triggered(self, new_value_block, current_contents):
-        """Returns the address or None."""
-        for watch_addr, kind in breakpoints[WRITE_WATCHPOINTS]:
-            for offset in range(0, kind):
-                addr = watch_addr + offset
-                if (addr >= new_value_block.address and
-                    addr < new_value_block.address + len(new_value_block.data)):
-                    index = addr - new_value_block.address
-                    if new_value_block.data[index*2:(index + 1)*2] != current_contents[index*2:(index + 1)*2]:
-                        return watch_addr
-        return None
-
-    def continue_with_recording(self, packet):
-        self.logger.debug("Continue with recording enabled")
-
-        step_packet = "vCont;s"
-        if packet == "vCont":
-            requested_step = False
-        else:
-            m = re.match("vCont;(c|s)(.*)", packet)
-            if m is None:
-                raise ValueError("Unsupported vCont packet: " + packet)
-            requested_step = m.group(1) == 's'
-            step_packet += m.group(2)
-
-        while True:
-            snapshot = self.capture_snapshot()
-            reply = self.pass_through(step_packet)
-            (stop_signal, stop_pairs) = self.parse_stop(reply)
-            if stop_signal != 5:
-                raise ValueError("Unexpected stop signal: " + reply)
-            is_swbreak = False
-            thread_id = None
-            for key, value in stop_pairs.items():
-                if key == "thread":
-                    thread_id = self.parse_thread_id(value)
-                    continue
-                if re.match('[0-9a-f]+', key):
-                    continue
-                if key == "swbreak" or (key == "reason" and value == "breakpoint"):
-                    is_swbreak = True
-                    continue
-                if key in ["name", "threads", "thread-pcs", "reason"]:
-                    continue
-                raise ValueError(f"Unknown stop key '{key}' in {reply}")
-            if is_swbreak:
-                self.logger.debug("Recording stopped")
-                return reply
-            if thread_id is None:
-                return ValueError("Expected thread ID: " + reply)
-            snapshot.thread_id = thread_id
-            self.snapshots.append(snapshot)
-            if requested_step:
-                self.logger.debug("Recording stopped for step")
-                return reply
-
-    def parse_stop(self, reply):
-        result = {}
-        if not reply:
-            raise ValueError("Invalid empty packet")
-        if reply[0] == "T" and len(reply) >= 3:
-            result = {k:v for k, v in self.parse_pairs(reply[3:])}
-            return (int(reply[1:3], 16), result)
-        raise "Unsupported stop reply: " + reply
-
-    def parse_pairs(self, text):
-        for pair in text.split(";"):
-            if not pair:
-                continue
-            m = re.match("([^:]+):(.*)", pair)
-            if m is None:
-                raise ValueError("Invalid pair text: " + text)
-            yield (m.group(1), m.group(2))
-
-    def capture_snapshot(self):
-        """Snapshot all threads and their stack memories."""
-        self.ensure_register_info()
-        current_thread = self.get_current_thread()
-        thread_snapshots = []
-        memory = []
-        for thread_id in self.get_thread_list():
-            registers = {}
-            for index in sorted(self.general_purpose_register_info.keys()):
-                reply =  self.pass_through(f"p{index:x};thread:{thread_id:x};")
-                if reply == "" or reply[0] == 'E':
-                    raise ValueError("Can't read register")
-                registers[index] = reply
-            thread_snapshot = ThreadSnapshot(thread_id, registers)
-            thread_sp = self.get_register(self.sp_register_info, thread_snapshot.registers)
-            memory += self.read_memory(thread_sp - BELOW_STACK_POINTER, thread_sp + ABOVE_STACK_POINTER)
-            thread_snapshots.append(thread_snapshot)
-        self.set_current_thread(current_thread)
-        return StateSnapshot(thread_snapshots, memory)
-
-    def restore_snapshot(self, snapshot):
-        """
-        Restore the snapshot during reverse execution.
-
-        If this triggers a breakpoint or watchpoint, return the stop reply,
-        otherwise None.
-        """
-        current_thread = self.get_current_thread()
-        stop_reasons = []
-        for thread_snapshot in snapshot.thread_snapshots:
-            thread_id = thread_snapshot.thread_id
-            for lldb_index in sorted(thread_snapshot.registers.keys()):
-                data = thread_snapshot.registers[lldb_index]
-                reply = self.pass_through(f"P{lldb_index:x}={data};thread:{thread_id:x};")
-                if reply != "OK":
-                    raise ValueError("Can't restore thread register")
-            if thread_id == snapshot.thread_id:
-                new_pc = self.get_register(self.pc_register_info, thread_snapshot.registers)
-                if self.breakpoint_triggered_at(new_pc):
-                    stop_reasons.append([("reason", "breakpoint")])
-        self.set_current_thread(current_thread)
-        for block in snapshot.memory:
-            current_memory = self.pass_through(f"m{block.address:x},{(len(block.data)/2):x}")
-            if not current_memory or current_memory[0] == 'E':
-                raise ValueError("Can't read back memory")
-            reply = self.pass_through(f"M{block.address:x},{len(block.data)/2:x}:" + block.data)
-            if reply != "OK":
-                raise ValueError("Can't restore memory")
-            watch_addr = self.watchpoint_triggered(block, current_memory[1:])
-            if watch_addr is not None:
-                stop_reasons.append([("reason", "watchpoint"), ("watch", f"{watch_addr:x}")])
-        if stop_reasons:
-            pairs = ";".join(f"{key}:{value}" for key, value in stop_reasons[0])
-            return f"T05thread:{self.pid:x}.{snapshot.thread_id:x};{pairs};"
-        return None
-
-    def reverse_step(self):
-        if not self.snapshots:
-            self.logger.debug("Reverse-step at history boundary")
-            return self.history_boundary_reply(self.get_current_thread())
-        self.logger.debug("Reverse-step started")
-        snapshot = self.snapshots.pop()
-        stop_reply = self.restore_snapshot(snapshot)
-        self.set_current_thread(snapshot.thread_id)
-        self.logger.debug("Reverse-step stopped")
-        if stop_reply is None:
-            return self.singlestep_stop_reply(snapshot.thread_id)
-        return stop_reply
-
-    def reverse_continue(self):
-        self.logger.debug("Reverse-continue started")
-        thread_id = None
-        while self.snapshots:
-            snapshot = self.snapshots.pop()
-            stop_reply = self.restore_snapshot(snapshot)
-            thread_id = snapshot.thread_id
-            if stop_reply is not None:
-                self.set_current_thread(thread_id)
-                self.logger.debug("Reverse-continue stopped")
-                return stop_reply
-        if thread_id is None:
-            thread_id = self.get_current_thread()
-        else:
-            self.set_current_thread(snapshot.thread_id)
-        self.logger.debug("Reverse-continue stopped at history boundary")
-        return self.history_boundary_reply(thread_id)
-
-    def get_current_thread(self):
-        reply = self.pass_through("qC")
-        return self.parse_thread_id(reply[2:])
-
-    def parse_thread_id(self, thread_id):
-        m = re.match("(p([0-9a-f]+)[.])?([0-9a-f]+)$", thread_id)
-        if m is None:
-            raise ValueError("Invalid thread ID: " + thread_id)
-        if self.pid is None:
-            self.pid = int(m.group(2), 16)
-        return int(m.group(3), 16)
-
-    def history_boundary_reply(self, thread_id):
-        return f"T00thread:{self.pid:x}.{thread_id:x};replaylog:begin;"
-
-    def singlestep_stop_reply(self, thread_id):
-        return f"T05thread:{self.pid:x}.{thread_id:x};"
-
-    def set_current_thread(self, thread_id):
-        """
-        Set current thread in inner gdbserver.
-        """
-        if thread_id >= 0:
-            self.pass_through(f"Hg{self.pid:x}.{thread_id:x}")
-            self.pass_through(f"Hc{self.pid:x}.{thread_id:x}")
-        else:
-            self.pass_through(f"Hc-1.-1")
-            self.pass_through(f"Hg-1.-1")
-
-    def get_register(self, register_info, registers):
-        if register_info.bitsize % 8 != 0:
-            raise ValueError("Register size must be a multiple of 8 bits")
-        if register_info.lldb_index not in registers:
-            raise ValueError("Register value not captured")
-        data = registers[register_info.lldb_index]
-        num_bytes = register_info.bitsize//8
-        bytes = []
-        for i in range(0, num_bytes):
-            bytes.append(int(data[i*2:(i + 1)*2], 16))
-        if register_info.little_endian:
-            bytes.reverse()
-        result = 0
-        for byte in bytes:
-            result = (result << 8) + byte
-        return result
-
-    def read_memory(self, start_addr, end_addr):
-        """
-        Read a region of memory from the target.
-
-        Some of the addresses may extend into invalid virtual memory;
-        skip those areas.
-        Return a list of blocks containing the valid area(s) in the
-        requested range.
-        """
-        regions = []
-        start_addr = start_addr & (BLOCK_SIZE - 1)
-        end_addr = (end_addr + BLOCK_SIZE - 1) & (BLOCK_SIZE - 1)
-        for addr in range(start_addr, end_addr, BLOCK_SIZE):
-            reply = self.pass_through(f"m{addr:x},{(BLOCK_SIZE - 1):x}")
-            if reply and reply[0] != 'E':
-                block = MemoryBlockSnapshot(addr, reply[1:])
-                regions.append(block)
-        return regions
-
-    def ensure_register_info(self):
-        if self.general_purpose_register_info is not None:
-            return
-        reply = self.pass_through("qHostInfo")
-        little_endian = any(kv == ("endian", "little") for kv in self.parse_pairs(reply))
-        self.general_purpose_register_info = {}
-        lldb_index = 0
-        while True:
-            reply = self.pass_through(f"qRegisterInfo{lldb_index:x}")
-            if not reply or reply[0] == 'E':
-                break
-            info = {k:v for k, v in self.parse_pairs(reply)}
-            reg_info = RegisterInfo(lldb_index, int(info["bitsize"]), little_endian)
-            if info["set"] == "General Purpose Registers" and not "container-regs" in info:
-                self.general_purpose_register_info[lldb_index] = reg_info
-            if "generic" in info:
-                if info["generic"] == "pc":
-                    self.pc_register_info = reg_info
-                elif info["generic"] == "sp":
-                    self.sp_register_info = reg_info
-            lldb_index += 1
-        if self.pc_register_info is None or self.sp_register_info is None:
-            raise ValueError("Can't find generic pc or sp register")
-
-    def get_thread_list(self):
-        threads = []
-        reply = self.pass_through("qfThreadInfo")
-        while True:
-            if not reply:
-                raise ValueError("Missing reply packet")
-            if reply[0] == 'm':
-                for id in reply[1:].split(","):
-                    threads.append(self.parse_thread_id(id))
-            elif reply[0] == 'l':
-                return threads
-            reply = self.pass_through("qsThreadInfo")
diff --git a/lldb/packages/Python/lldbsuite/test/lldbtest.py b/lldb/packages/Python/lldbsuite/test/lldbtest.py
index 7cc1ac9749ec..8884ef5933ad 100644
--- a/lldb/packages/Python/lldbsuite/test/lldbtest.py
+++ b/lldb/packages/Python/lldbsuite/test/lldbtest.py
@@ -143,8 +143,6 @@ STOPPED_DUE_TO_STEP_IN = "Process state is stopped due to step in"
 
 STOPPED_DUE_TO_WATCHPOINT = "Process should be stopped due to watchpoint"
 
-STOPPED_DUE_TO_HISTORY_BOUNDARY = "Process should be stopped due to history boundary"
-
 DATA_TYPES_DISPLAYED_CORRECTLY = "Data type(s) displayed correctly"
 
 VALID_BREAKPOINT = "Got a valid breakpoint"
diff --git a/lldb/source/API/SBProcess.cpp b/lldb/source/API/SBProcess.cpp
index 07780f9f9c83..9773144723c3 100644
--- a/lldb/source/API/SBProcess.cpp
+++ b/lldb/source/API/SBProcess.cpp
@@ -564,10 +564,6 @@ uint32_t SBProcess::GetAddressByteSize() const {
 }
 
 SBError SBProcess::Continue() {
-  return Continue(RunDirection::eRunForward);
-}
-
-SBError SBProcess::Continue(RunDirection direction) {
   LLDB_INSTRUMENT_VA(this);
 
   SBError sb_error;
@@ -578,9 +574,9 @@ SBError SBProcess::Continue(RunDirection direction) {
         process_sp->GetTarget().GetAPIMutex());
 
     if (process_sp->GetTarget().GetDebugger().GetAsyncExecution())
-      sb_error.ref() = process_sp->Resume(direction);
+      sb_error.ref() = process_sp->Resume();
     else
-      sb_error.ref() = process_sp->ResumeSynchronous(nullptr, direction);
+      sb_error.ref() = process_sp->ResumeSynchronous(nullptr);
   } else
     sb_error = Status::FromErrorString("SBProcess is invalid");
 
diff --git a/lldb/source/API/SBThread.cpp b/lldb/source/API/SBThread.cpp
index aca8a0399529..a99456e06d03 100644
--- a/lldb/source/API/SBThread.cpp
+++ b/lldb/source/API/SBThread.cpp
@@ -172,7 +172,6 @@ size_t SBThread::GetStopReasonDataCount() {
         case eStopReasonInstrumentation:
         case eStopReasonProcessorTrace:
         case eStopReasonVForkDone:
-        case eStopReasonHistoryBoundary:
           // There is no data for these stop reasons.
           return 0;
 
@@ -234,7 +233,6 @@ uint64_t SBThread::GetStopReasonDataAtIndex(uint32_t idx) {
         case eStopReasonInstrumentation:
         case eStopReasonProcessorTrace:
         case eStopReasonVForkDone:
-        case eStopReasonHistoryBoundary:
           // There is no data for these stop reasons.
           return 0;
 
diff --git a/lldb/source/Core/DynamicLoader.cpp b/lldb/source/Core/DynamicLoader.cpp
index 7758a87403b5..68d6ab085085 100644
--- a/lldb/source/Core/DynamicLoader.cpp
+++ b/lldb/source/Core/DynamicLoader.cpp
@@ -83,7 +83,11 @@ ModuleSP DynamicLoader::GetTargetExecutable() {
       ModuleSpec module_spec(executable->GetFileSpec(),
                              executable->GetArchitecture());
       auto module_sp = std::make_shared<Module>(module_spec);
-
+      // If we're a coredump and we already have a main executable, we don't
+      // need to reload the module list that target already has
+      if (!m_process->IsLiveDebugSession()) {
+        return executable;
+      }
       // Check if the executable has changed and set it to the target
       // executable if they differ.
       if (module_sp && module_sp->GetUUID().IsValid() &&
@@ -369,4 +373,3 @@ void DynamicLoader::LoadOperatingSystemPlugin(bool flush)
     if (m_process)
         m_process->LoadOperatingSystemPlugin(flush);
 }
-
diff --git a/lldb/source/Interpreter/CommandInterpreter.cpp b/lldb/source/Interpreter/CommandInterpreter.cpp
index ea60492ac46a..8d3a82ef6c99 100644
--- a/lldb/source/Interpreter/CommandInterpreter.cpp
+++ b/lldb/source/Interpreter/CommandInterpreter.cpp
@@ -2553,8 +2553,7 @@ bool CommandInterpreter::DidProcessStopAbnormally() const {
     const StopReason reason = stop_info->GetStopReason();
     if (reason == eStopReasonException ||
         reason == eStopReasonInstrumentation ||
-        reason == eStopReasonProcessorTrace || reason == eStopReasonInterrupt ||
-        reason == eStopReasonHistoryBoundary)
+        reason == eStopReasonProcessorTrace || reason == eStopReasonInterrupt)
       return true;
 
     if (reason == eStopReasonSignal) {
diff --git a/lldb/source/Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.cpp b/lldb/source/Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.cpp
index b9c0e174c3be..34aca50df0ac 100644
--- a/lldb/source/Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.cpp
+++ b/lldb/source/Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.cpp
@@ -18,6 +18,7 @@
 #include "lldb/Symbol/ObjectFile.h"
 #include "lldb/Target/MemoryRegionInfo.h"
 #include "lldb/Target/Platform.h"
+#include "lldb/Target/RegisterContext.h"
 #include "lldb/Target/Target.h"
 #include "lldb/Target/Thread.h"
 #include "lldb/Target/ThreadPlanRunToAddress.h"
@@ -866,3 +867,82 @@ bool DynamicLoaderPOSIXDYLD::AlwaysRelyOnEHUnwindInfo(
 bool DynamicLoaderPOSIXDYLD::IsCoreFile() const {
   return !m_process->IsLiveDebugSession();
 }
+
+// For our ELF/POSIX builds save off the fs_base/gs_base regions
+static void AddThreadLocalMemoryRegions(Process &process, ThreadSP &thread_sp,
+                                        std::vector<MemoryRegionInfo> &ranges) {
+  lldb::RegisterContextSP reg_ctx = thread_sp->GetRegisterContext();
+  if (!reg_ctx)
+    return;
+
+  const RegisterInfo *reg_info = reg_ctx->GetRegisterInfo(
+      lldb::RegisterKind::eRegisterKindGeneric, LLDB_REGNUM_GENERIC_TP);
+  if (!reg_info)
+    return;
+
+  lldb_private::RegisterValue thread_local_register_value;
+  bool success = reg_ctx->ReadRegister(reg_info, thread_local_register_value);
+  if (!success)
+    return;
+
+  const uint64_t fail_value = UINT64_MAX;
+  bool readSuccess = false;
+  const lldb::addr_t reg_value_addr =
+      thread_local_register_value.GetAsUInt64(fail_value, &readSuccess);
+  if (!readSuccess || reg_value_addr == fail_value)
+    return;
+
+  MemoryRegionInfo thread_local_region;
+  Status err = process.GetMemoryRegionInfo(reg_value_addr, thread_local_region);
+  if (err.Fail())
+    return;
+
+  ranges.push_back(thread_local_region);
+}
+
+// Save off the link map for core files.
+static void AddLinkMapSections(Process &process,
+                               std::vector<MemoryRegionInfo> &ranges) {
+  ModuleList &module_list = process.GetTarget().GetImages();
+  Target *target = &process.GetTarget();
+  for (size_t idx = 0; idx < module_list.GetSize(); idx++) {
+    ModuleSP module_sp = module_list.GetModuleAtIndex(idx);
+    if (!module_sp)
+      continue;
+
+    ObjectFile *obj = module_sp->GetObjectFile();
+    if (!obj)
+      continue;
+    Address addr = obj->GetImageInfoAddress(target);
+    addr_t load_addr = addr.GetLoadAddress(target);
+    if (load_addr == LLDB_INVALID_ADDRESS)
+      continue;
+
+    MemoryRegionInfo link_map_section;
+    Status err = process.GetMemoryRegionInfo(load_addr, link_map_section);
+    if (err.Fail())
+      continue;
+
+    ranges.push_back(link_map_section);
+  }
+}
+
+void DynamicLoaderPOSIXDYLD::CalculateDynamicSaveCoreRanges(
+    lldb_private::Process &process,
+    std::vector<lldb_private::MemoryRegionInfo> &ranges,
+    llvm::function_ref<bool(const lldb_private::Thread &)>
+        save_thread_predicate) {
+  ThreadList &thread_list = process.GetThreadList();
+  for (size_t idx = 0; idx < thread_list.GetSize(); idx++) {
+    ThreadSP thread_sp = thread_list.GetThreadAtIndex(idx);
+    if (!thread_sp)
+      continue;
+
+    if (!save_thread_predicate(*thread_sp))
+      continue;
+
+    AddThreadLocalMemoryRegions(process, thread_sp, ranges);
+  }
+
+  AddLinkMapSections(process, ranges);
+}
diff --git a/lldb/source/Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.h b/lldb/source/Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.h
index 4c92335602cd..bde334aaca40 100644
--- a/lldb/source/Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.h
+++ b/lldb/source/Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.h
@@ -60,6 +60,12 @@ public:
                                      lldb::addr_t base_addr,
                                      bool base_addr_is_offset) override;
 
+  void CalculateDynamicSaveCoreRanges(
+      lldb_private::Process &process,
+      std::vector<lldb_private::MemoryRegionInfo> &ranges,
+      llvm::function_ref<bool(const lldb_private::Thread &)>
+          save_thread_predicate) override;
+
 protected:
   /// Runtime linker rendezvous structure.
   DYLDRendezvous m_rendezvous;
diff --git a/lldb/source/Plugins/Process/Linux/NativeThreadLinux.cpp b/lldb/source/Plugins/Process/Linux/NativeThreadLinux.cpp
index b0aa664775b4..de047ee214c1 100644
--- a/lldb/source/Plugins/Process/Linux/NativeThreadLinux.cpp
+++ b/lldb/source/Plugins/Process/Linux/NativeThreadLinux.cpp
@@ -82,9 +82,6 @@ void LogThreadStopInfo(Log &log, const ThreadStopInfo &stop_info,
   case eStopReasonProcessorTrace:
     log.Printf("%s: %s processor trace", __FUNCTION__, header);
     return;
-  case eStopReasonHistoryBoundary:
-    log.Printf("%s: %s history boundary", __FUNCTION__, header);
-    return;
   default:
     log.Printf("%s: %s invalid stop reason %" PRIu32, __FUNCTION__, header,
                static_cast<uint32_t>(stop_info.reason));
diff --git a/lldb/source/Plugins/Process/MacOSX-Kernel/ProcessKDP.cpp b/lldb/source/Plugins/Process/MacOSX-Kernel/ProcessKDP.cpp
index 367fce442bb8..9b2907c68099 100644
--- a/lldb/source/Plugins/Process/MacOSX-Kernel/ProcessKDP.cpp
+++ b/lldb/source/Plugins/Process/MacOSX-Kernel/ProcessKDP.cpp
@@ -402,17 +402,9 @@ lldb_private::DynamicLoader *ProcessKDP::GetDynamicLoader() {
 
 Status ProcessKDP::WillResume() { return Status(); }
 
-Status ProcessKDP::DoResume(RunDirection direction) {
+Status ProcessKDP::DoResume() {
   Status error;
   Log *log = GetLog(KDPLog::Process);
-
-  if (direction == RunDirection::eRunReverse) {
-    error.FromErrorStringWithFormatv(
-        "error: {0} does not support reverse execution of processes",
-        GetPluginName());
-    return error;
-  }
-
   // Only start the async thread if we try to do any process control
   if (!m_async_thread.IsJoinable())
     StartAsyncThread();
diff --git a/lldb/source/Plugins/Process/MacOSX-Kernel/ProcessKDP.h b/lldb/source/Plugins/Process/MacOSX-Kernel/ProcessKDP.h
index 1b71d83f70b0..e5ec5914f960 100644
--- a/lldb/source/Plugins/Process/MacOSX-Kernel/ProcessKDP.h
+++ b/lldb/source/Plugins/Process/MacOSX-Kernel/ProcessKDP.h
@@ -90,7 +90,7 @@ public:
   // Process Control
   lldb_private::Status WillResume() override;
 
-  lldb_private::Status DoResume(lldb::RunDirection direction) override;
+  lldb_private::Status DoResume() override;
 
   lldb_private::Status DoHalt(bool &caused_stop) override;
 
diff --git a/lldb/source/Plugins/Process/Windows/Common/ProcessWindows.cpp b/lldb/source/Plugins/Process/Windows/Common/ProcessWindows.cpp
index 76b7095deaa5..703aa082f047 100644
--- a/lldb/source/Plugins/Process/Windows/Common/ProcessWindows.cpp
+++ b/lldb/source/Plugins/Process/Windows/Common/ProcessWindows.cpp
@@ -204,17 +204,11 @@ ProcessWindows::DoAttachToProcessWithID(lldb::pid_t pid,
   return error;
 }
 
-Status ProcessWindows::DoResume(RunDirection direction) {
+Status ProcessWindows::DoResume() {
   Log *log = GetLog(WindowsLog::Process);
   llvm::sys::ScopedLock lock(m_mutex);
   Status error;
 
-  if (direction == RunDirection::eRunReverse) {
-    error.SetErrorStringWithFormatv(
-        "error: {0} does not support reverse execution of processes", GetPluginName());
-    return error;
-  }
-
   StateType private_state = GetPrivateState();
   if (private_state == eStateStopped || private_state == eStateCrashed) {
     LLDB_LOG(log, "process {0} is in state {1}.  Resuming...",
diff --git a/lldb/source/Plugins/Process/Windows/Common/ProcessWindows.h b/lldb/source/Plugins/Process/Windows/Common/ProcessWindows.h
index 97284b7cd143..e97cfb790248 100644
--- a/lldb/source/Plugins/Process/Windows/Common/ProcessWindows.h
+++ b/lldb/source/Plugins/Process/Windows/Common/ProcessWindows.h
@@ -52,7 +52,7 @@ public:
   Status DoAttachToProcessWithID(
       lldb::pid_t pid,
       const lldb_private::ProcessAttachInfo &attach_info) override;
-  Status DoResume(lldb::RunDirection direction) override;
+  Status DoResume() override;
   Status DoDestroy() override;
   Status DoHalt(bool &caused_stop) override;
 
diff --git a/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationClient.cpp b/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationClient.cpp
index fc792a440941..e42526c8fd72 100644
--- a/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationClient.cpp
+++ b/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationClient.cpp
@@ -199,20 +199,6 @@ uint64_t GDBRemoteCommunicationClient::GetRemoteMaxPacketSize() {
   return m_max_packet_size;
 }
 
-bool GDBRemoteCommunicationClient::GetReverseContinueSupported() {
-  if (m_supports_reverse_continue == eLazyBoolCalculate) {
-    GetRemoteQSupported();
-  }
-  return m_supports_reverse_continue == eLazyBoolYes;
-}
-
-bool GDBRemoteCommunicationClient::GetReverseStepSupported() {
-  if (m_supports_reverse_step == eLazyBoolCalculate) {
-    GetRemoteQSupported();
-  }
-  return m_supports_reverse_step == eLazyBoolYes;
-}
-
 bool GDBRemoteCommunicationClient::QueryNoAckModeSupported() {
   if (m_supports_not_sending_acks == eLazyBoolCalculate) {
     m_send_acks = true;
@@ -309,8 +295,6 @@ void GDBRemoteCommunicationClient::ResetDiscoverableSettings(bool did_exec) {
     m_supports_qXfer_siginfo_read = eLazyBoolCalculate;
     m_supports_augmented_libraries_svr4_read = eLazyBoolCalculate;
     m_uses_native_signals = eLazyBoolCalculate;
-    m_supports_reverse_continue = eLazyBoolCalculate;
-    m_supports_reverse_step = eLazyBoolCalculate;
     m_supports_qProcessInfoPID = true;
     m_supports_qfProcessInfo = true;
     m_supports_qUserName = true;
@@ -364,8 +348,6 @@ void GDBRemoteCommunicationClient::GetRemoteQSupported() {
   m_supports_memory_tagging = eLazyBoolNo;
   m_supports_qSaveCore = eLazyBoolNo;
   m_uses_native_signals = eLazyBoolNo;
-  m_supports_reverse_continue = eLazyBoolNo;
-  m_supports_reverse_step = eLazyBoolNo;
 
   m_max_packet_size = UINT64_MAX; // It's supposed to always be there, but if
                                   // not, we assume no limit
@@ -419,10 +401,6 @@ void GDBRemoteCommunicationClient::GetRemoteQSupported() {
         m_supports_qSaveCore = eLazyBoolYes;
       else if (x == "native-signals+")
         m_uses_native_signals = eLazyBoolYes;
-      else if (x == "ReverseContinue+")
-        m_supports_reverse_continue = eLazyBoolYes;
-      else if (x == "ReverseStep+")
-        m_supports_reverse_step = eLazyBoolYes;
       // Look for a list of compressions in the features list e.g.
       // qXfer:features:read+;PacketSize=20000;qEcho+;SupportedCompressions=zlib-
       // deflate,lzma
diff --git a/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationClient.h b/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationClient.h
index 116b47c1edf0..898d176abc34 100644
--- a/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationClient.h
+++ b/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationClient.h
@@ -331,10 +331,6 @@ public:
 
   bool GetMultiprocessSupported();
 
-  bool GetReverseContinueSupported();
-
-  bool GetReverseStepSupported();
-
   LazyBool SupportsAllocDeallocMemory() // const
   {
     // Uncomment this to have lldb pretend the debug server doesn't respond to
@@ -565,8 +561,6 @@ protected:
   LazyBool m_supports_memory_tagging = eLazyBoolCalculate;
   LazyBool m_supports_qSaveCore = eLazyBoolCalculate;
   LazyBool m_uses_native_signals = eLazyBoolCalculate;
-  LazyBool m_supports_reverse_continue = eLazyBoolCalculate;
-  LazyBool m_supports_reverse_step = eLazyBoolCalculate;
 
   bool m_supports_qProcessInfoPID : 1, m_supports_qfProcessInfo : 1,
       m_supports_qUserName : 1, m_supports_qGroupName : 1,
diff --git a/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationServerLLGS.cpp b/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationServerLLGS.cpp
index 4016cde74ebe..35fa93e53bc6 100644
--- a/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationServerLLGS.cpp
+++ b/lldb/source/Plugins/Process/gdb-remote/GDBRemoteCommunicationServerLLGS.cpp
@@ -716,7 +716,6 @@ static const char *GetStopReasonString(StopReason stop_reason) {
     return "vforkdone";
   case eStopReasonInterrupt:
     return "async interrupt";
-  case eStopReasonHistoryBoundary:
   case eStopReasonInstrumentation:
   case eStopReasonInvalid:
   case eStopReasonPlanComplete:
diff --git a/lldb/source/Plugins/Process/gdb-remote/ProcessGDBRemote.cpp b/lldb/source/Plugins/Process/gdb-remote/ProcessGDBRemote.cpp
index 3fc03bd05d5d..3e09c316d74f 100644
--- a/lldb/source/Plugins/Process/gdb-remote/ProcessGDBRemote.cpp
+++ b/lldb/source/Plugins/Process/gdb-remote/ProcessGDBRemote.cpp
@@ -169,10 +169,6 @@ public:
   }
 };
 
-std::chrono::seconds ResumeTimeout() {
-  return std::chrono::seconds(5);
-}
-
 } // namespace
 
 static PluginProperties &GetGlobalPluginProperties() {
@@ -1184,11 +1180,10 @@ Status ProcessGDBRemote::WillResume() {
   return Status();
 }
 
-Status ProcessGDBRemote::DoResume(RunDirection direction) {
+Status ProcessGDBRemote::DoResume() {
   Status error;
   Log *log = GetLog(GDBRLog::Process);
-  LLDB_LOGF(log, "ProcessGDBRemote::Resume(%s)",
-            direction == RunDirection::eRunForward ? "" : "reverse");
+  LLDB_LOGF(log, "ProcessGDBRemote::Resume()");
 
   ListenerSP listener_sp(
       Listener::MakeListener("gdb-remote.resume-packet-sent"));
@@ -1202,21 +1197,12 @@ Status ProcessGDBRemote::DoResume(RunDirection direction) {
 
     StreamString continue_packet;
     bool continue_packet_error = false;
-    // Number of threads continuing with "c", i.e. continuing without a signal to deliver.
-    const size_t num_continue_c_tids = m_continue_c_tids.size();
-    // Number of threads continuing with "C", i.e. continuing with a signal to deliver.
-    const size_t num_continue_C_tids = m_continue_C_tids.size();
-    // Number of threads continuing with "s", i.e. single-stepping.
-    const size_t num_continue_s_tids = m_continue_s_tids.size();
-    // Number of threads continuing with "S", i.e. single-stepping with a signal to deliver.
-    const size_t num_continue_S_tids = m_continue_S_tids.size();
-    if (direction == RunDirection::eRunForward &&
-        m_gdb_comm.HasAnyVContSupport()) {
+    if (m_gdb_comm.HasAnyVContSupport()) {
       std::string pid_prefix;
       if (m_gdb_comm.GetMultiprocessSupported())
         pid_prefix = llvm::formatv("p{0:x-}.", GetID());
 
-      if (num_continue_c_tids == num_threads ||
+      if (m_continue_c_tids.size() == num_threads ||
           (m_continue_c_tids.empty() && m_continue_C_tids.empty() &&
            m_continue_s_tids.empty() && m_continue_S_tids.empty())) {
         // All threads are continuing
@@ -1279,11 +1265,14 @@ Status ProcessGDBRemote::DoResume(RunDirection direction) {
     } else
       continue_packet_error = true;
 
-    if (direction == RunDirection::eRunForward && continue_packet_error) {
+    if (continue_packet_error) {
       // Either no vCont support, or we tried to use part of the vCont packet
-      // that wasn't supported by the remote GDB server, or it's the reverse
-      // direction. We need to try and make a simple packet that can do our
-      // continue.
+      // that wasn't supported by the remote GDB server. We need to try and
+      // make a simple packet that can do our continue
+      const size_t num_continue_c_tids = m_continue_c_tids.size();
+      const size_t num_continue_C_tids = m_continue_C_tids.size();
+      const size_t num_continue_s_tids = m_continue_s_tids.size();
+      const size_t num_continue_S_tids = m_continue_S_tids.size();
       if (num_continue_c_tids > 0) {
         if (num_continue_c_tids == num_threads) {
           // All threads are resuming...
@@ -1374,41 +1363,9 @@ Status ProcessGDBRemote::DoResume(RunDirection direction) {
       }
     }
 
-    if (direction == RunDirection::eRunReverse && continue_packet_error) {
-      if (num_continue_C_tids > 0 || num_continue_S_tids > 0) {
-        LLDB_LOGF(log, "ProcessGDBRemote::DoResumeReverse: Signals not supported");
-        return Status::FromErrorString("can't deliver signals while running in reverse");
-      }
-
-      if (num_continue_s_tids > 0) {
-        if (num_continue_s_tids > 1) {
-          LLDB_LOGF(log, "ProcessGDBRemote::DoResumeReverse: can't step multiple threads");
-          return Status::FromErrorString("can't step multiple threads while reverse-stepping");
-        }
-
-        if (!m_gdb_comm.GetReverseStepSupported()) {
-          LLDB_LOGF(log, "ProcessGDBRemote::DoResumeReverse: target does not support reverse-stepping");
-          return Status::FromErrorString("target does not support reverse-stepping");
-        }
-
-        m_gdb_comm.SetCurrentThreadForRun(m_continue_s_tids.front());
-        continue_packet.PutCString("bs");
-      } else {
-        if (!m_gdb_comm.GetReverseContinueSupported()) {
-          LLDB_LOGF(log, "ProcessGDBRemote::DoResumeReverse: target does not support reverse-continue");
-          return Status::FromErrorString("target does not support reverse-continue");
-        }
-
-        // All threads continue whether requested or not ---
-        // we can't change how threads ran in the past.
-        continue_packet.PutCString("bc");
-      }
-
-      continue_packet_error = false;
-    }
-
     if (continue_packet_error) {
-      return Status::FromErrorString("can't make continue packet for this resume");
+      error =
+          Status::FromErrorString("can't make continue packet for this resume");
     } else {
       EventSP event_sp;
       if (!m_async_thread.IsJoinable()) {
@@ -1423,7 +1380,7 @@ Status ProcessGDBRemote::DoResume(RunDirection direction) {
           std::make_shared<EventDataBytes>(continue_packet.GetString());
       m_async_broadcaster.BroadcastEvent(eBroadcastBitAsyncContinue, data_sp);
 
-      if (!listener_sp->GetEvent(event_sp, ResumeTimeout())) {
+      if (!listener_sp->GetEvent(event_sp, std::chrono::seconds(5))) {
         error = Status::FromErrorString("Resume timed out.");
         LLDB_LOGF(log, "ProcessGDBRemote::DoResume: Resume timed out.");
       } else if (event_sp->BroadcasterIs(&m_async_broadcaster)) {
@@ -1906,10 +1863,6 @@ ThreadSP ProcessGDBRemote::SetThreadStopInfo(
           thread_sp->SetStopInfo(StopInfo::CreateStopReasonWithException(
               *thread_sp, description.c_str()));
           handled = true;
-        } else if (reason == "replaylog") {
-          thread_sp->SetStopInfo(StopInfo::CreateStopReasonHistoryBoundary(
-              *thread_sp, description.c_str()));
-          handled = true;
         } else if (reason == "exec") {
           did_exec = true;
           thread_sp->SetStopInfo(
@@ -2365,8 +2318,6 @@ StateType ProcessGDBRemote::SetThreadStopInfo(StringExtractor &stop_packet) {
         description = std::string(ostr.GetString());
       } else if (key.compare("swbreak") == 0 || key.compare("hwbreak") == 0) {
         reason = "breakpoint";
-      } else if (key.compare("replaylog") == 0) {
-        reason = "replaylog";
       } else if (key.compare("library") == 0) {
         auto error = LoadModules();
         if (error) {
diff --git a/lldb/source/Plugins/Process/gdb-remote/ProcessGDBRemote.h b/lldb/source/Plugins/Process/gdb-remote/ProcessGDBRemote.h
index fa3e1cec76e2..249279585138 100644
--- a/lldb/source/Plugins/Process/gdb-remote/ProcessGDBRemote.h
+++ b/lldb/source/Plugins/Process/gdb-remote/ProcessGDBRemote.h
@@ -111,7 +111,7 @@ public:
   // Process Control
   Status WillResume() override;
 
-  Status DoResume(lldb::RunDirection direction) override;
+  Status DoResume() override;
 
   Status DoHalt(bool &caused_stop) override;
 
diff --git a/lldb/source/Plugins/Process/minidump/ProcessMinidump.cpp b/lldb/source/Plugins/Process/minidump/ProcessMinidump.cpp
index 32ffba763c08..5ea3db23f114 100644
--- a/lldb/source/Plugins/Process/minidump/ProcessMinidump.cpp
+++ b/lldb/source/Plugins/Process/minidump/ProcessMinidump.cpp
@@ -21,11 +21,13 @@
 #include "lldb/Interpreter/CommandReturnObject.h"
 #include "lldb/Interpreter/OptionArgParser.h"
 #include "lldb/Interpreter/OptionGroupBoolean.h"
+#include "lldb/Target/DynamicLoader.h"
 #include "lldb/Target/JITLoaderList.h"
 #include "lldb/Target/MemoryRegionInfo.h"
 #include "lldb/Target/SectionLoadList.h"
 #include "lldb/Target/Target.h"
 #include "lldb/Target/UnixSignals.h"
+#include "lldb/Utility/DataBufferHeap.h"
 #include "lldb/Utility/LLDBAssert.h"
 #include "lldb/Utility/LLDBLog.h"
 #include "lldb/Utility/Log.h"
@@ -34,6 +36,7 @@
 #include "llvm/Support/MemoryBuffer.h"
 #include "llvm/Support/Threading.h"
 
+#include "Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.h"
 #include "Plugins/ObjectFile/Placeholder/ObjectFilePlaceholder.h"
 #include "Plugins/Process/Utility/StopInfoMachException.h"
 
@@ -333,6 +336,16 @@ ArchSpec ProcessMinidump::GetArchitecture() {
   return ArchSpec(triple);
 }
 
+DataExtractor ProcessMinidump::GetAuxvData() {
+  std::optional<llvm::ArrayRef<uint8_t>> auxv =
+      m_minidump_parser->GetStream(StreamType::LinuxAuxv);
+  if (!auxv)
+    return DataExtractor();
+
+  return DataExtractor(auxv->data(), auxv->size(), GetByteOrder(),
+                       GetAddressByteSize(), GetAddressByteSize());
+}
+
 void ProcessMinidump::BuildMemoryRegions() {
   if (m_memory_regions)
     return;
@@ -534,7 +547,12 @@ void ProcessMinidump::ReadModuleList() {
 
       module_sp = Module::CreateModuleFromObjectFile<ObjectFilePlaceholder>(
           module_spec, load_addr, load_size);
-      GetTarget().GetImages().Append(module_sp, true /* notify */);
+      // If we haven't loaded a main executable yet, set the first module to be
+      // main executable
+      if (!GetTarget().GetExecutableModule())
+        GetTarget().SetExecutableModule(module_sp);
+      else
+        GetTarget().GetImages().Append(module_sp, true /* notify */);
     }
 
     bool load_addr_changed = false;
diff --git a/lldb/source/Plugins/Process/minidump/ProcessMinidump.h b/lldb/source/Plugins/Process/minidump/ProcessMinidump.h
index f2ea0a2b61d1..3d235670a33a 100644
--- a/lldb/source/Plugins/Process/minidump/ProcessMinidump.h
+++ b/lldb/source/Plugins/Process/minidump/ProcessMinidump.h
@@ -53,12 +53,11 @@ public:
 
   Status DoLoadCore() override;
 
-  DynamicLoader *GetDynamicLoader() override { return nullptr; }
+  // Returns AUXV structure found in the core file
+  lldb_private::DataExtractor GetAuxvData() override;
 
   llvm::StringRef GetPluginName() override { return GetPluginNameStatic(); }
 
-  SystemRuntime *GetSystemRuntime() override { return nullptr; }
-
   Status DoDestroy() override;
 
   void RefreshStateAfterStop() override;
diff --git a/lldb/source/Plugins/Process/minidump/RegisterContextMinidump_x86_64.cpp b/lldb/source/Plugins/Process/minidump/RegisterContextMinidump_x86_64.cpp
index e879c4931565..f305d1b7031d 100644
--- a/lldb/source/Plugins/Process/minidump/RegisterContextMinidump_x86_64.cpp
+++ b/lldb/source/Plugins/Process/minidump/RegisterContextMinidump_x86_64.cpp
@@ -44,6 +44,17 @@ static void writeRegister(const void *reg_src, uint8_t *context,
   memcpy(reg_dest.data(), reg_src, reg_dest.size());
 }
 
+// TODO: Fix the registers in this file!
+// writeRegister checks x86_64 registers without base registers. This causes
+// an overlap in the register enum values. So we were truncating fs_base.
+// We should standardize to the x86_64_with_base registers.
+static void writeBaseRegister(const void *reg_src, uint8_t *context,
+                              const RegisterInfo &reg) {
+  auto bytes = reg.mutable_data(context);
+  llvm::MutableArrayRef<uint8_t> reg_dest = bytes.take_front(8);
+  memcpy(reg_dest.data(), reg_src, reg_dest.size());
+}
+
 lldb::DataBufferSP lldb_private::minidump::ConvertMinidumpContext_x86_64(
     llvm::ArrayRef<uint8_t> source_data,
     RegisterInfoInterface *target_reg_interface) {
@@ -105,11 +116,12 @@ lldb::DataBufferSP lldb_private::minidump::ConvertMinidumpContext_x86_64(
     writeRegister(&context->r15, result_base, reg_info[lldb_r15_x86_64]);
   }
 
+  // See comment on base regsiter
   if ((context_flags & LLDBSpecificFlag) == LLDBSpecificFlag) {
-    writeRegister(&context->fs_base, result_base,
-                  reg_info[x86_64_with_base::lldb_fs_base]);
-    writeRegister(&context->gs_base, result_base,
-                  reg_info[x86_64_with_base::lldb_gs_base]);
+    writeBaseRegister(&context->fs_base, result_base,
+                      reg_info[x86_64_with_base::lldb_fs_base]);
+    writeBaseRegister(&context->gs_base, result_base,
+                      reg_info[x86_64_with_base::lldb_gs_base]);
   }
 
   // TODO parse the floating point registers
diff --git a/lldb/source/Plugins/Process/scripted/ScriptedProcess.cpp b/lldb/source/Plugins/Process/scripted/ScriptedProcess.cpp
index 304c12173dd3..d2111ce877ce 100644
--- a/lldb/source/Plugins/Process/scripted/ScriptedProcess.cpp
+++ b/lldb/source/Plugins/Process/scripted/ScriptedProcess.cpp
@@ -182,15 +182,10 @@ void ScriptedProcess::DidResume() {
   m_pid = GetInterface().GetProcessID();
 }
 
-Status ScriptedProcess::DoResume(RunDirection direction) {
+Status ScriptedProcess::DoResume() {
   LLDB_LOGF(GetLog(LLDBLog::Process), "ScriptedProcess::%s resuming process", __FUNCTION__);
 
-  if (direction == RunDirection::eRunForward) {
-    return GetInterface().Resume();
-  } else {
-    return Status::FromErrorStringWithFormatv(
-        "error: {0} does not support reverse execution of processes", GetPluginName());
-  }
+  return GetInterface().Resume();
 }
 
 Status ScriptedProcess::DoAttach(const ProcessAttachInfo &attach_info) {
diff --git a/lldb/source/Plugins/Process/scripted/ScriptedProcess.h b/lldb/source/Plugins/Process/scripted/ScriptedProcess.h
index 8ebe4ca5f3d4..0335364b4010 100644
--- a/lldb/source/Plugins/Process/scripted/ScriptedProcess.h
+++ b/lldb/source/Plugins/Process/scripted/ScriptedProcess.h
@@ -52,7 +52,7 @@ public:
 
   void DidResume() override;
 
-  Status DoResume(lldb::RunDirection direction) override;
+  Status DoResume() override;
 
   Status DoAttachToProcessWithID(lldb::pid_t pid,
                                  const ProcessAttachInfo &attach_info) override;
diff --git a/lldb/source/Target/Process.cpp b/lldb/source/Target/Process.cpp
index ff6a2f59eba3..c009d17d3ba5 100644
--- a/lldb/source/Target/Process.cpp
+++ b/lldb/source/Target/Process.cpp
@@ -446,8 +446,7 @@ Process::Process(lldb::TargetSP target_sp, ListenerSP listener_sp,
       m_memory_cache(*this), m_allocated_memory_cache(*this),
       m_should_detach(false), m_next_event_action_up(), m_public_run_lock(),
       m_private_run_lock(), m_currently_handling_do_on_removals(false),
-      m_resume_requested(false), m_last_run_direction(eRunForward),
-      m_interrupt_tid(LLDB_INVALID_THREAD_ID),
+      m_resume_requested(false), m_interrupt_tid(LLDB_INVALID_THREAD_ID),
       m_finalizing(false), m_destructing(false),
       m_clear_thread_plans_on_stop(false), m_force_next_event_delivery(false),
       m_last_broadcast_state(eStateInvalid), m_destroy_in_process(false),
@@ -846,7 +845,6 @@ bool Process::HandleProcessStateChangedEvent(
             switch (thread_stop_reason) {
             case eStopReasonInvalid:
             case eStopReasonNone:
-            case eStopReasonHistoryBoundary:
               break;
 
             case eStopReasonSignal: {
@@ -1354,7 +1352,7 @@ void Process::SetPublicState(StateType new_state, bool restarted) {
   }
 }
 
-Status Process::Resume(RunDirection direction) {
+Status Process::Resume() {
   Log *log(GetLog(LLDBLog::State | LLDBLog::Process));
   LLDB_LOGF(log, "(plugin = %s) -- locking run lock", GetPluginName().data());
   if (!m_public_run_lock.TrySetRunning()) {
@@ -1363,7 +1361,7 @@ Status Process::Resume(RunDirection direction) {
     return Status::FromErrorString(
         "Resume request failed - process still running.");
   }
-  Status error = PrivateResume(direction);
+  Status error = PrivateResume();
   if (!error.Success()) {
     // Undo running state change
     m_public_run_lock.SetStopped();
@@ -1371,7 +1369,7 @@ Status Process::Resume(RunDirection direction) {
   return error;
 }
 
-Status Process::ResumeSynchronous(Stream *stream, RunDirection direction) {
+Status Process::ResumeSynchronous(Stream *stream) {
   Log *log(GetLog(LLDBLog::State | LLDBLog::Process));
   LLDB_LOGF(log, "Process::ResumeSynchronous -- locking run lock");
   if (!m_public_run_lock.TrySetRunning()) {
@@ -1384,7 +1382,7 @@ Status Process::ResumeSynchronous(Stream *stream, RunDirection direction) {
       Listener::MakeListener(ResumeSynchronousHijackListenerName.data()));
   HijackProcessEvents(listener_sp);
 
-  Status error = PrivateResume(direction);
+  Status error = PrivateResume();
   if (error.Success()) {
     StateType state =
         WaitForProcessToStop(std::nullopt, nullptr, true, listener_sp, stream,
@@ -3241,7 +3239,7 @@ Status Process::ConnectRemote(llvm::StringRef remote_url) {
   return error;
 }
 
-Status Process::PrivateResume(RunDirection direction) {
+Status Process::PrivateResume() {
   Log *log(GetLog(LLDBLog::Process | LLDBLog::Step));
   LLDB_LOGF(log,
             "Process::PrivateResume() m_stop_id = %u, public state: %s "
@@ -3257,15 +3255,6 @@ Status Process::PrivateResume(RunDirection direction) {
   if (!GetModID().IsLastResumeForUserExpression())
     ResetExtendedCrashInfoDict();
 
-  if (m_last_run_direction != direction) {
-    // In the future we might want to support mixed-direction plans,
-    // e.g. a forward step-over stops at a breakpoint, the user does
-    // a reverse-step, then disables the breakpoint and continues forward.
-    // This code will need to be changed to support that.
-    m_thread_list.DiscardThreadPlans();
-    m_last_run_direction = direction;
-  }
-
   Status error(WillResume());
   // Tell the process it is about to resume before the thread list
   if (error.Success()) {
@@ -3283,7 +3272,7 @@ Status Process::PrivateResume(RunDirection direction) {
             "Process::PrivateResume PreResumeActions failed, not resuming.");
       } else {
         m_mod_id.BumpResumeID();
-        error = DoResume(direction);
+        error = DoResume();
         if (error.Success()) {
           DidResume();
           m_thread_list.DidResume();
@@ -3746,7 +3735,7 @@ bool Process::ShouldBroadcastEvent(Event *event_ptr) {
                     "from state: %s",
                     static_cast<void *>(event_ptr), StateAsCString(state));
           ProcessEventData::SetRestartedInEvent(event_ptr, true);
-          PrivateResume(m_last_run_direction);
+          PrivateResume();
         }
       } else {
         return_value = true;
@@ -4357,7 +4346,7 @@ void Process::ProcessEventData::DoOnRemoval(Event *event_ptr) {
     SetRestarted(true);
     // Use the private resume method here, since we aren't changing the run
     // lock state.
-    process_sp->PrivateResume(process_sp->m_last_run_direction);
+    process_sp->PrivateResume();
   } else {
     bool hijacked = process_sp->IsHijackedForEvent(eBroadcastBitStateChanged) &&
                     !process_sp->StateChangedIsHijackedForSynchronousResume();
@@ -6539,6 +6528,29 @@ static void AddRegion(const MemoryRegionInfo &region, bool try_dirty_pages,
                 CreateCoreFileMemoryRange(region));
 }
 
+static void SaveDynamicLoaderSections(Process &process,
+                                      const SaveCoreOptions &options,
+                                      CoreFileMemoryRanges &ranges,
+                                      std::set<addr_t> &stack_ends) {
+  DynamicLoader *dyld = process.GetDynamicLoader();
+  if (!dyld)
+    return;
+
+  std::vector<MemoryRegionInfo> dynamic_loader_mem_regions;
+  std::function<bool(const lldb_private::Thread &)> save_thread_predicate =
+      [&](const lldb_private::Thread &t) -> bool {
+    return options.ShouldThreadBeSaved(t.GetID());
+  };
+  dyld->CalculateDynamicSaveCoreRanges(process, dynamic_loader_mem_regions,
+                                       save_thread_predicate);
+  for (const auto &region : dynamic_loader_mem_regions) {
+    // The Dynamic Loader can give us regions that could include a truncated
+    // stack
+    if (stack_ends.count(region.GetRange().GetRangeEnd()) == 0)
+      AddRegion(region, true, ranges);
+  }
+}
+
 static void SaveOffRegionsWithStackPointers(Process &process,
                                             const SaveCoreOptions &core_options,
                                             const MemoryRegionInfos &regions,
@@ -6570,11 +6582,13 @@ static void SaveOffRegionsWithStackPointers(Process &process,
       // off in other calls
       sp_region.GetRange().SetRangeBase(stack_head);
       sp_region.GetRange().SetByteSize(stack_size);
-      stack_ends.insert(sp_region.GetRange().GetRangeEnd());
+      const addr_t range_end = sp_region.GetRange().GetRangeEnd();
+      stack_ends.insert(range_end);
       // This will return true if the threadlist the user specified is empty,
       // or contains the thread id from thread_sp.
-      if (core_options.ShouldThreadBeSaved(thread_sp->GetID()))
+      if (core_options.ShouldThreadBeSaved(thread_sp->GetID())) {
         AddRegion(sp_region, try_dirty_pages, ranges);
+      }
     }
   }
 }
@@ -6683,9 +6697,14 @@ Status Process::CalculateCoreFileSaveRanges(const SaveCoreOptions &options,
   std::set<addr_t> stack_ends;
   // For fully custom set ups, we don't want to even look at threads if there
   // are no threads specified.
-  if (core_style != lldb::eSaveCoreCustomOnly || options.HasSpecifiedThreads())
+  if (core_style != lldb::eSaveCoreCustomOnly ||
+      options.HasSpecifiedThreads()) {
     SaveOffRegionsWithStackPointers(*this, options, regions, ranges,
                                     stack_ends);
+    // Save off the dynamic loader sections, so if we are on an architecture
+    // that supports Thread Locals, that we include those as well.
+    SaveDynamicLoaderSections(*this, options, ranges, stack_ends);
+  }
 
   switch (core_style) {
   case eSaveCoreUnspecified:
diff --git a/lldb/source/Target/StopInfo.cpp b/lldb/source/Target/StopInfo.cpp
index 08e9a7c099ba..bd7032b803df 100644
--- a/lldb/source/Target/StopInfo.cpp
+++ b/lldb/source/Target/StopInfo.cpp
@@ -1212,30 +1212,6 @@ public:
   }
 };
 
-// StopInfoHistoryBoundary
-
-class StopInfoHistoryBoundary : public StopInfo {
-public:
-  StopInfoHistoryBoundary(Thread &thread, const char *description)
-      : StopInfo(thread, LLDB_INVALID_UID) {
-    if (description)
-      SetDescription(description);
-  }
-
-  ~StopInfoHistoryBoundary() override = default;
-
-  StopReason GetStopReason() const override {
-    return eStopReasonHistoryBoundary;
-  }
-
-  const char *GetDescription() override {
-    if (m_description.empty())
-      return "history boundary";
-    else
-      return m_description.c_str();
-  }
-};
-
 // StopInfoThreadPlan
 
 class StopInfoThreadPlan : public StopInfo {
@@ -1463,11 +1439,6 @@ StopInfoSP StopInfo::CreateStopReasonProcessorTrace(Thread &thread,
   return StopInfoSP(new StopInfoProcessorTrace(thread, description));
 }
 
-StopInfoSP StopInfo::CreateStopReasonHistoryBoundary(Thread &thread,
-                                                     const char *description) {
-  return StopInfoSP(new StopInfoHistoryBoundary(thread, description));
-}
-
 StopInfoSP StopInfo::CreateStopReasonWithExec(Thread &thread) {
   return StopInfoSP(new StopInfoExec(thread));
 }
diff --git a/lldb/source/Target/Thread.cpp b/lldb/source/Target/Thread.cpp
index bbb586f033b7..902fbb2b519e 100644
--- a/lldb/source/Target/Thread.cpp
+++ b/lldb/source/Target/Thread.cpp
@@ -624,12 +624,10 @@ void Thread::SetupForResume() {
     // what the current plan is.
 
     lldb::RegisterContextSP reg_ctx_sp(GetRegisterContext());
-    ProcessSP process_sp(GetProcess());
-    if (reg_ctx_sp && process_sp &&
-        process_sp->GetLastRunDirection() == eRunForward) {
+    if (reg_ctx_sp) {
       const addr_t thread_pc = reg_ctx_sp->GetPC();
       BreakpointSiteSP bp_site_sp =
-          process_sp->GetBreakpointSiteList().FindByAddress(thread_pc);
+          GetProcess()->GetBreakpointSiteList().FindByAddress(thread_pc);
       if (bp_site_sp) {
         // Note, don't assume there's a ThreadPlanStepOverBreakpoint, the
         // target may not require anything special to step over a breakpoint.
@@ -1734,8 +1732,6 @@ std::string Thread::StopReasonAsString(lldb::StopReason reason) {
     return "processor trace";
   case eStopReasonInterrupt:
     return "async interrupt";
-  case eStopReasonHistoryBoundary:
-    return "history boundary";
   }
 
   return "StopReason = " + std::to_string(reason);
diff --git a/lldb/test/API/functionalities/process_save_core_minidump/TestProcessSaveCoreMinidump.py b/lldb/test/API/functionalities/process_save_core_minidump/TestProcessSaveCoreMinidump.py
index 03cc415924e0..4818dde4f3b8 100644
--- a/lldb/test/API/functionalities/process_save_core_minidump/TestProcessSaveCoreMinidump.py
+++ b/lldb/test/API/functionalities/process_save_core_minidump/TestProcessSaveCoreMinidump.py
@@ -523,8 +523,10 @@ class ProcessSaveCoreMinidumpTestCase(TestBase):
         finally:
             self.assertTrue(self.dbg.DeleteTarget(target))
 
-    def minidump_deterministic_difference(self):
-        """Test that verifies that two minidumps produced are identical."""
+    @skipUnlessPlatform(["linux"])
+    @skipUnlessArch("x86_64")
+    def minidump_saves_fs_base_region(self):
+        """Test that verifies the minidump file saves region for fs_base"""
 
         self.build()
         exe = self.getBuildArtifact("a.out")
@@ -534,6 +536,45 @@ class ProcessSaveCoreMinidumpTestCase(TestBase):
                 None, None, self.get_process_working_directory()
             )
             self.assertState(process.GetState(), lldb.eStateStopped)
+            thread = process.GetThreadAtIndex(0)
+            custom_file = self.getBuildArtifact("core.reg_region.dmp")
+            options = lldb.SBSaveCoreOptions()
+            options.SetOutputFile(lldb.SBFileSpec(custom_file))
+            options.SetPluginName("minidump")
+            options.SetStyle(lldb.eSaveCoreCustomOnly)
+            options.AddThread(thread)
+            error = process.SaveCore(options)
+            self.assertTrue(error.Success())
+
+            registers = thread.GetFrameAtIndex(0).GetRegisters()
+            fs_base = registers.GetFirstValueByName("fs_base").GetValueAsUnsigned()
+            self.assertTrue(fs_base != 0)
+            core_target = self.dbg.CreateTarget(None)
+            core_proc = core_target.LoadCore(one_region_file)
+            core_region_list = core_proc.GetMemoryRegions()
+            live_region_list = process.GetMemoryRegions()
+            live_region = lldb.SBMemoryRegionInfo()
+            live_region_list.GetMemoryRegionForAddress(fs_base, live_region)
+            core_region = lldb.SBMemoryRegionInfo()
+            error = core_region_list.GetMemoryRegionForAddress(fs_base, core_region)
+            self.assertTrue(error.Success())
+            self.assertEqual(live_region, core_region)
+
+        finally:
+            self.assertTrue(self.dbg.DeleteTarget(target))
+            self.assertTrue(self.dbg.DeleteTarget(core_target))
+            if os.path.isfile(custom_file):
+                os.unlink(custom_file)
+
+    def minidump_deterministic_difference(self):
+        """Test that verifies that two minidumps produced are identical."""
+        self.build()
+        exe = self.getBuildArtifact("a.out")
+        try:
+            target = self.dbg.CreateTarget(exe)
+            process = target.LaunchSimple(
+                None, None, self.get_process_working_directory()
+            )
 
             core_styles = [
                 lldb.eSaveCoreStackOnly,
@@ -562,6 +603,36 @@ class ProcessSaveCoreMinidumpTestCase(TestBase):
                 self.assertEqual(file_one, file_two)
                 self.assertTrue(os.unlink(spec_one.GetFileName()))
                 self.assertTrue(os.unlink(spec_two.GetFileName()))
-
         finally:
             self.assertTrue(self.dbg.DeleteTarget(target))
+
+    @skipUnlessPlatform(["linux"])
+    @skipUnlessArch("x86_64")
+    def minidump_saves_fs_base_region(self):
+        self.build()
+        exe = self.getBuildArtifact("a.out")
+        try:
+            target = self.dbg.CreateTarget(exe)
+            process = target.LaunchSimple(
+                None, None, self.get_process_working_directory()
+            )
+            self.assertState(process.GetState(), lldb.eStateStopped)
+            thread = process.GetThreadAtIndex(0)
+            tls_file = self.getBuildArtifact("core.tls.dmp")
+            options = lldb.SBSaveCoreOptions()
+            options.SetOutputFile(lldb.SBFileSpec(tls_file))
+            options.SetPluginName("minidump")
+            options.SetStyle(lldb.eSaveCoreCustomOnly)
+            options.AddThread(thread)
+            error = process.SaveCore(options)
+            self.assertTrue(error.Success())
+            core_target = self.dbg.CreateTarget(None)
+            core_proc = core_target.LoadCore(tls_file)
+            frame = core_proc.GetThreadAtIndex(0).GetFrameAtIndex(0)
+            tls_val = frame.FindValue("lf")
+            self.assertEqual(tls_val.GetValueAsUnsigned(), 42)
+
+        except:
+            self.assertTrue(self.dbg.DeleteTarget(target))
+            if os.path.isfile(tls_file):
+                os.unlink(tls_file)
diff --git a/lldb/test/API/functionalities/process_save_core_minidump/main.cpp b/lldb/test/API/functionalities/process_save_core_minidump/main.cpp
index fa34a371f206..15daa68e9a64 100644
--- a/lldb/test/API/functionalities/process_save_core_minidump/main.cpp
+++ b/lldb/test/API/functionalities/process_save_core_minidump/main.cpp
@@ -1,6 +1,7 @@
 #include <cassert>
 #include <iostream>
 #include <thread>
+thread_local size_t lf = 42;
 
 void g() { assert(false); }
 
diff --git a/lldb/test/API/functionalities/reverse-execution/Makefile b/lldb/test/API/functionalities/reverse-execution/Makefile
deleted file mode 100644
index 10495940055b..000000000000
--- a/lldb/test/API/functionalities/reverse-execution/Makefile
+++ /dev/null
@@ -1,3 +0,0 @@
-C_SOURCES := main.c
-
-include Makefile.rules
diff --git a/lldb/test/API/functionalities/reverse-execution/TestReverseContinueBreakpoints.py b/lldb/test/API/functionalities/reverse-execution/TestReverseContinueBreakpoints.py
deleted file mode 100644
index b37578fbd824..000000000000
--- a/lldb/test/API/functionalities/reverse-execution/TestReverseContinueBreakpoints.py
+++ /dev/null
@@ -1,115 +0,0 @@
-import lldb
-import time
-import unittest
-from lldbsuite.test.lldbtest import *
-from lldbsuite.test.decorators import *
-from lldbsuite.test.gdbclientutils import *
-from lldbsuite.test.lldbreverse import ReverseTestBase
-from lldbsuite.test import lldbutil
-
-
-class TestReverseContinueBreakpoints(ReverseTestBase):
-    NO_DEBUG_INFO_TESTCASE = True
-
-    def test_reverse_continue(self):
-        self.reverse_continue_internal(async_mode=False)
-
-    def test_reverse_continue_async(self):
-        self.reverse_continue_internal(async_mode=True)
-
-    def reverse_continue_internal(self, async_mode):
-        target, process, initial_threads = self.setup_recording(async_mode)
-
-        # Reverse-continue. We'll stop at the point where we started recording.
-        status = process.Continue(lldb.eRunReverse)
-        self.assertSuccess(status)
-        self.expect_async_state_changes(async_mode, process, [lldb.eStateRunning, lldb.eStateStopped])
-        self.expect(
-            "thread list",
-            STOPPED_DUE_TO_HISTORY_BOUNDARY,
-            substrs=["stopped", "stop reason = history boundary"],
-        )
-
-        # Continue forward normally until the target exits.
-        status = process.Continue()
-        self.expect_async_state_changes(async_mode, process, [lldb.eStateRunning, lldb.eStateExited])
-        self.assertSuccess(status)
-        self.assertState(process.GetState(), lldb.eStateExited)
-        self.assertEqual(process.GetExitStatus(), 0)
-
-    def test_reverse_continue_breakpoint(self):
-        self.reverse_continue_breakpoint_internal(async_mode=False)
-
-    def test_reverse_continue_breakpoint_async(self):
-        self.reverse_continue_breakpoint_internal(async_mode=True)
-
-    def reverse_continue_breakpoint_internal(self, async_mode):
-        target, process, initial_threads = self.setup_recording(async_mode)
-
-        # Reverse-continue to the function "trigger_breakpoint".
-        trigger_bkpt = target.BreakpointCreateByName("trigger_breakpoint", None)
-        status = process.Continue(lldb.eRunReverse)
-        self.assertSuccess(status)
-        self.expect_async_state_changes(async_mode, process, [lldb.eStateRunning, lldb.eStateStopped])
-        threads_now = lldbutil.get_threads_stopped_at_breakpoint(process, trigger_bkpt)
-        self.assertEqual(threads_now, initial_threads)
-
-    def test_reverse_continue_skip_breakpoint(self):
-        self.reverse_continue_skip_breakpoint_internal(async_mode=False)
-
-    def test_reverse_continue_skip_breakpoint_async(self):
-        self.reverse_continue_skip_breakpoint_internal(async_mode=True)
-
-    def reverse_continue_skip_breakpoint_internal(self, async_mode):
-        target, process, initial_threads = self.setup_recording(async_mode)
-
-        # Reverse-continue over a breakpoint at "trigger_breakpoint" whose
-        # condition is false.
-        # This tests that we continue in the correct direction after hitting
-        # the breakpoint.
-        trigger_bkpt = target.BreakpointCreateByName("trigger_breakpoint", None)
-        trigger_bkpt.SetCondition("false_condition")
-        status = process.Continue(lldb.eRunReverse)
-        self.expect_async_state_changes(async_mode, process, [lldb.eStateRunning, lldb.eStateStopped])
-        self.assertSuccess(status)
-        self.expect(
-            "thread list",
-            STOPPED_DUE_TO_HISTORY_BOUNDARY,
-            substrs=["stopped", "stop reason = history boundary"],
-        )
-
-    def setup_recording(self, async_mode):
-        """
-        Record execution of code between "start_recording" and "stop_recording" breakpoints.
-
-        Returns with the target stopped at "stop_recording", with recording disabled,
-        ready to reverse-execute.
-        """
-        self.build()
-        target = self.dbg.CreateTarget("")
-        process = self.connect(target)
-
-        # Record execution from the start of the function "start_recording"
-        # to the start of the function "stop_recording". We want to keep the
-        # interval that we record as small as possible to minimize the run-time
-        # of our single-stepping recorder.
-        start_recording_bkpt = target.BreakpointCreateByName("start_recording", None)
-        initial_threads = lldbutil.continue_to_breakpoint(process, start_recording_bkpt)
-        self.assertEqual(len(initial_threads), 1)
-        target.BreakpointDelete(start_recording_bkpt.GetID())
-        self.start_recording()
-        stop_recording_bkpt = target.BreakpointCreateByName("stop_recording", None)
-        lldbutil.continue_to_breakpoint(process, stop_recording_bkpt)
-        target.BreakpointDelete(stop_recording_bkpt.GetID())
-        self.stop_recording()
-
-        self.dbg.SetAsync(async_mode)
-        self.expect_async_state_changes(async_mode, process, [lldb.eStateStopped])
-
-        return target, process, initial_threads
-
-    def expect_async_state_changes(self, async_mode, process, states):
-        if not async_mode:
-            return
-        listener = self.dbg.GetListener()
-        lldbutil.expect_state_changes(self, listener, process, states)
diff --git a/lldb/test/API/functionalities/reverse-execution/TestReverseContinueNotSupported.py b/lldb/test/API/functionalities/reverse-execution/TestReverseContinueNotSupported.py
deleted file mode 100644
index d610761b8cb0..000000000000
--- a/lldb/test/API/functionalities/reverse-execution/TestReverseContinueNotSupported.py
+++ /dev/null
@@ -1,30 +0,0 @@
-import lldb
-import unittest
-from lldbsuite.test.lldbtest import *
-from lldbsuite.test.decorators import *
-from lldbsuite.test import lldbutil
-
-
-class TestReverseContinueNotSupported(TestBase):
-    NO_DEBUG_INFO_TESTCASE = True
-
-    def test_reverse_continue_not_supported(self):
-        self.build()
-        exe = self.getBuildArtifact("a.out")
-        target = self.dbg.CreateTarget(exe)
-        self.assertTrue(target, VALID_TARGET)
-
-        main_bkpt = target.BreakpointCreateByName("main", None)
-        self.assertTrue(main_bkpt, VALID_BREAKPOINT)
-
-        process = target.LaunchSimple(None, None, self.get_process_working_directory())
-        self.assertTrue(process, PROCESS_IS_VALID)
-
-        # This will fail gracefully.
-        status = process.Continue(lldb.eRunReverse)
-        self.assertFailure(status, "target does not support reverse-continue")
-
-        status = process.Continue()
-        self.assertSuccess(status)
-        self.assertState(process.GetState(), lldb.eStateExited)
-        self.assertEqual(process.GetExitStatus(), 0)
diff --git a/lldb/test/API/functionalities/reverse-execution/main.c b/lldb/test/API/functionalities/reverse-execution/main.c
deleted file mode 100644
index 40e45dc9f5c3..000000000000
--- a/lldb/test/API/functionalities/reverse-execution/main.c
+++ /dev/null
@@ -1,14 +0,0 @@
-volatile int false_condition = 0;
-
-static void start_recording() {}
-
-static void trigger_breakpoint() {}
-
-static void stop_recording() {}
-
-int main() {
-  start_recording();
-  trigger_breakpoint();
-  stop_recording();
-  return 0;
-}
diff --git a/lldb/tools/lldb-dap/JSONUtils.cpp b/lldb/tools/lldb-dap/JSONUtils.cpp
index 211fd34957f4..558f889c4b7f 100644
--- a/lldb/tools/lldb-dap/JSONUtils.cpp
+++ b/lldb/tools/lldb-dap/JSONUtils.cpp
@@ -1045,9 +1045,6 @@ llvm::json::Value CreateThreadStopped(lldb::SBThread &thread,
   case lldb::eStopReasonProcessorTrace:
     body.try_emplace("reason", "processor trace");
     break;
-  case lldb::eStopReasonHistoryBoundary:
-    body.try_emplace("reason", "history boundary");
-    break;
   case lldb::eStopReasonSignal:
   case lldb::eStopReasonException:
     body.try_emplace("reason", "exception");
diff --git a/lldb/tools/lldb-dap/LLDBUtils.cpp b/lldb/tools/lldb-dap/LLDBUtils.cpp
index 1c5e3ac70087..b38833c0fdb6 100644
--- a/lldb/tools/lldb-dap/LLDBUtils.cpp
+++ b/lldb/tools/lldb-dap/LLDBUtils.cpp
@@ -111,7 +111,6 @@ bool ThreadHasStopReason(lldb::SBThread &thread) {
   case lldb::eStopReasonVFork:
   case lldb::eStopReasonVForkDone:
   case lldb::eStopReasonInterrupt:
-  case lldb::eStopReasonHistoryBoundary:
     return true;
   case lldb::eStopReasonThreadExiting:
   case lldb::eStopReasonInvalid:
diff --git a/llvm/include/llvm/ExecutionEngine/Orc/ELFNixPlatform.h b/llvm/include/llvm/ExecutionEngine/Orc/ELFNixPlatform.h
index 40b85e327201..54442c91096b 100644
--- a/llvm/include/llvm/ExecutionEngine/Orc/ELFNixPlatform.h
+++ b/llvm/include/llvm/ExecutionEngine/Orc/ELFNixPlatform.h
@@ -106,14 +106,14 @@ public:
   /// RuntimeAliases function, in which case the client is responsible for
   /// setting up all aliases (including the required ones).
   static Expected<std::unique_ptr<ELFNixPlatform>>
-  Create(ExecutionSession &ES, ObjectLinkingLayer &ObjLinkingLayer,
-         JITDylib &PlatformJD, std::unique_ptr<DefinitionGenerator> OrcRuntime,
+  Create(ObjectLinkingLayer &ObjLinkingLayer, JITDylib &PlatformJD,
+         std::unique_ptr<DefinitionGenerator> OrcRuntime,
          std::optional<SymbolAliasMap> RuntimeAliases = std::nullopt);
 
   /// Construct using a path to the ORC runtime.
   static Expected<std::unique_ptr<ELFNixPlatform>>
-  Create(ExecutionSession &ES, ObjectLinkingLayer &ObjLinkingLayer,
-         JITDylib &PlatformJD, const char *OrcRuntimePath,
+  Create(ObjectLinkingLayer &ObjLinkingLayer, JITDylib &PlatformJD,
+         const char *OrcRuntimePath,
          std::optional<SymbolAliasMap> RuntimeAliases = std::nullopt);
 
   ExecutionSession &getExecutionSession() const { return ES; }
@@ -211,8 +211,7 @@ private:
 
   static bool supportedTarget(const Triple &TT);
 
-  ELFNixPlatform(ExecutionSession &ES, ObjectLinkingLayer &ObjLinkingLayer,
-                 JITDylib &PlatformJD,
+  ELFNixPlatform(ObjectLinkingLayer &ObjLinkingLayer, JITDylib &PlatformJD,
                  std::unique_ptr<DefinitionGenerator> OrcRuntimeGenerator,
                  Error &Err);
 
@@ -308,4 +307,4 @@ using SPSELFNixJITDylibDepInfoMap =
 } // end namespace orc
 } // end namespace llvm
 
-#endif // LLVM_EXECUTIONENGINE_ORC_ELFNIXPLATFORM_H
\ No newline at end of file
+#endif // LLVM_EXECUTIONENGINE_ORC_ELFNIXPLATFORM_H
diff --git a/llvm/include/llvm/IR/IntrinsicsDirectX.td b/llvm/include/llvm/IR/IntrinsicsDirectX.td
index 1cf6acbf1264..45aea1ccdb6d 100644
--- a/llvm/include/llvm/IR/IntrinsicsDirectX.td
+++ b/llvm/include/llvm/IR/IntrinsicsDirectX.td
@@ -71,6 +71,7 @@ def int_dx_udot :
     [IntrNoMem, Commutative] >;
 
 def int_dx_frac  : DefaultAttrsIntrinsic<[llvm_anyfloat_ty], [LLVMMatchType<0>], [IntrNoMem]>;
+def int_dx_degrees : DefaultAttrsIntrinsic<[LLVMMatchType<0>], [llvm_anyfloat_ty], [IntrNoMem]>;
 
 def int_dx_isinf : DefaultAttrsIntrinsic<[LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>],
     [llvm_anyfloat_ty], [IntrNoMem]>;
diff --git a/llvm/include/llvm/IR/IntrinsicsSPIRV.td b/llvm/include/llvm/IR/IntrinsicsSPIRV.td
index 92d2f67399d2..3d61456589ee 100644
--- a/llvm/include/llvm/IR/IntrinsicsSPIRV.td
+++ b/llvm/include/llvm/IR/IntrinsicsSPIRV.td
@@ -62,6 +62,7 @@ let TargetPrefix = "spv" in {
   def int_spv_all : DefaultAttrsIntrinsic<[llvm_i1_ty], [llvm_any_ty], [IntrNoMem]>;
   def int_spv_any : DefaultAttrsIntrinsic<[llvm_i1_ty], [llvm_any_ty], [IntrNoMem]>;
   def int_spv_cross : DefaultAttrsIntrinsic<[llvm_anyfloat_ty], [LLVMMatchType<0>, LLVMMatchType<0>], [IntrNoMem]>;
+  def int_spv_degrees : DefaultAttrsIntrinsic<[LLVMMatchType<0>], [llvm_anyfloat_ty], [IntrNoMem]>;
   def int_spv_frac : DefaultAttrsIntrinsic<[LLVMMatchType<0>], [llvm_anyfloat_ty], [IntrNoMem]>;
   def int_spv_lerp : DefaultAttrsIntrinsic<[LLVMMatchType<0>], [llvm_anyfloat_ty, LLVMMatchType<0>,LLVMMatchType<0>],
     [IntrNoMem] >;
diff --git a/llvm/include/llvm/Transforms/Vectorize/SandboxVectorizer/DependencyGraph.h b/llvm/include/llvm/Transforms/Vectorize/SandboxVectorizer/DependencyGraph.h
index 7d300ea2b60d..5fa57efc1462 100644
--- a/llvm/include/llvm/Transforms/Vectorize/SandboxVectorizer/DependencyGraph.h
+++ b/llvm/include/llvm/Transforms/Vectorize/SandboxVectorizer/DependencyGraph.h
@@ -122,6 +122,10 @@ public:
   iterator preds_end(DependencyGraph &DAG) const {
     return const_cast<DGNode *>(this)->preds_end(DAG);
   }
+  /// \Returns a range of DAG predecessors nodes. If this is a MemDGNode then
+  /// this will also include the memory dependency predecessors.
+  /// Please note that this can include the same node more than once, if for
+  /// example it's both a use-def predecessor and a mem dep predecessor.
   iterator_range<iterator> preds(DependencyGraph &DAG) const {
     return make_range(preds_begin(DAG), preds_end(DAG));
   }
diff --git a/llvm/lib/ExecutionEngine/Orc/ELFNixPlatform.cpp b/llvm/lib/ExecutionEngine/Orc/ELFNixPlatform.cpp
index d92077dbcbd0..610ecbff5c5c 100644
--- a/llvm/lib/ExecutionEngine/Orc/ELFNixPlatform.cpp
+++ b/llvm/lib/ExecutionEngine/Orc/ELFNixPlatform.cpp
@@ -233,10 +233,13 @@ private:
 namespace llvm {
 namespace orc {
 
-Expected<std::unique_ptr<ELFNixPlatform>> ELFNixPlatform::Create(
-    ExecutionSession &ES, ObjectLinkingLayer &ObjLinkingLayer,
-    JITDylib &PlatformJD, std::unique_ptr<DefinitionGenerator> OrcRuntime,
-    std::optional<SymbolAliasMap> RuntimeAliases) {
+Expected<std::unique_ptr<ELFNixPlatform>>
+ELFNixPlatform::Create(ObjectLinkingLayer &ObjLinkingLayer,
+                       JITDylib &PlatformJD,
+                       std::unique_ptr<DefinitionGenerator> OrcRuntime,
+                       std::optional<SymbolAliasMap> RuntimeAliases) {
+
+  auto &ES = ObjLinkingLayer.getExecutionSession();
 
   // If the target is not supported then bail out immediately.
   if (!supportedTarget(ES.getTargetTriple()))
@@ -271,15 +274,14 @@ Expected<std::unique_ptr<ELFNixPlatform>> ELFNixPlatform::Create(
   // Create the instance.
   Error Err = Error::success();
   auto P = std::unique_ptr<ELFNixPlatform>(new ELFNixPlatform(
-      ES, ObjLinkingLayer, PlatformJD, std::move(OrcRuntime), Err));
+      ObjLinkingLayer, PlatformJD, std::move(OrcRuntime), Err));
   if (Err)
     return std::move(Err);
   return std::move(P);
 }
 
 Expected<std::unique_ptr<ELFNixPlatform>>
-ELFNixPlatform::Create(ExecutionSession &ES,
-                       ObjectLinkingLayer &ObjLinkingLayer,
+ELFNixPlatform::Create(ObjectLinkingLayer &ObjLinkingLayer,
                        JITDylib &PlatformJD, const char *OrcRuntimePath,
                        std::optional<SymbolAliasMap> RuntimeAliases) {
 
@@ -289,7 +291,7 @@ ELFNixPlatform::Create(ExecutionSession &ES,
   if (!OrcRuntimeArchiveGenerator)
     return OrcRuntimeArchiveGenerator.takeError();
 
-  return Create(ES, ObjLinkingLayer, PlatformJD,
+  return Create(ObjLinkingLayer, PlatformJD,
                 std::move(*OrcRuntimeArchiveGenerator),
                 std::move(RuntimeAliases));
 }
@@ -392,10 +394,10 @@ bool ELFNixPlatform::supportedTarget(const Triple &TT) {
 }
 
 ELFNixPlatform::ELFNixPlatform(
-    ExecutionSession &ES, ObjectLinkingLayer &ObjLinkingLayer,
-    JITDylib &PlatformJD,
+    ObjectLinkingLayer &ObjLinkingLayer, JITDylib &PlatformJD,
     std::unique_ptr<DefinitionGenerator> OrcRuntimeGenerator, Error &Err)
-    : ES(ES), PlatformJD(PlatformJD), ObjLinkingLayer(ObjLinkingLayer),
+    : ES(ObjLinkingLayer.getExecutionSession()), PlatformJD(PlatformJD),
+      ObjLinkingLayer(ObjLinkingLayer),
       DSOHandleSymbol(ES.intern("__dso_handle")) {
   ErrorAsOutParameter _(&Err);
   ObjLinkingLayer.addPlugin(std::make_unique<ELFNixPlatformPlugin>(*this));
diff --git a/llvm/lib/ExecutionEngine/Orc/LLJIT.cpp b/llvm/lib/ExecutionEngine/Orc/LLJIT.cpp
index a13443ce57ea..d3dd3b6bedfb 100644
--- a/llvm/lib/ExecutionEngine/Orc/LLJIT.cpp
+++ b/llvm/lib/ExecutionEngine/Orc/LLJIT.cpp
@@ -1185,8 +1185,8 @@ Expected<JITDylibSP> ExecutorNativePlatform::operator()(LLJIT &J) {
     if (!G)
       return G.takeError();
 
-    if (auto P = ELFNixPlatform::Create(ES, *ObjLinkingLayer, PlatformJD,
-                                        std::move(*G)))
+    if (auto P =
+            ELFNixPlatform::Create(*ObjLinkingLayer, PlatformJD, std::move(*G)))
       J.getExecutionSession().setPlatform(std::move(*P));
     else
       return P.takeError();
diff --git a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
index c1aefee3793c..8a217cd1ec5c 100644
--- a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
+++ b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp
@@ -775,6 +775,7 @@ AArch64TargetLowering::AArch64TargetLowering(const TargetMachine &TM,
              ISD::FMAXNUM,
              ISD::FMINIMUM,
              ISD::FMAXIMUM,
+             ISD::FCANONICALIZE,
              ISD::STRICT_FADD,
              ISD::STRICT_FSUB,
              ISD::STRICT_FMUL,
@@ -818,6 +819,7 @@ AArch64TargetLowering::AArch64TargetLowering(const TargetMachine &TM,
     setOperationPromotedToType(ISD::FROUNDEVEN, V4Narrow, MVT::v4f32);
     setOperationPromotedToType(ISD::FRINT,      V4Narrow, MVT::v4f32);
     setOperationPromotedToType(ISD::FNEARBYINT, V4Narrow, MVT::v4f32);
+    setOperationPromotedToType(ISD::FCANONICALIZE, V4Narrow, MVT::v4f32);
 
     setOperationAction(ISD::FABS,        V4Narrow, Legal);
     setOperationAction(ISD::FNEG, 	 V4Narrow, Legal);
@@ -851,6 +853,7 @@ AArch64TargetLowering::AArch64TargetLowering(const TargetMachine &TM,
     setOperationAction(ISD::SELECT,      V8Narrow, Expand);
     setOperationAction(ISD::SELECT_CC,   V8Narrow, Expand);
     setOperationAction(ISD::FP_EXTEND,   V8Narrow, Expand);
+    setOperationPromotedToType(ISD::FCANONICALIZE, V8Narrow, MVT::v8f32);
   };
 
   if (!Subtarget->hasFullFP16()) {
diff --git a/llvm/lib/Target/AArch64/AArch64InstrInfo.td b/llvm/lib/Target/AArch64/AArch64InstrInfo.td
index 33d05d6039b0..325508b62a9f 100644
--- a/llvm/lib/Target/AArch64/AArch64InstrInfo.td
+++ b/llvm/lib/Target/AArch64/AArch64InstrInfo.td
@@ -5052,17 +5052,25 @@ def : Pat<(v1f64 (fminnum (v1f64 FPR64:$Rn), (v1f64 FPR64:$Rm))),
 
 def : Pat<(fminnum_ieee (f64 FPR64:$a), (f64 FPR64:$b)),
           (FMINNMDrr FPR64:$a, FPR64:$b)>;
-def : Pat<(fminnum_ieee (f32 FPR32:$a), (f32 FPR32:$b)),
-          (FMINNMSrr FPR32:$a, FPR32:$b)>;
-def : Pat<(fminnum_ieee (f16 FPR16:$a), (f16 FPR16:$b)),
-          (FMINNMHrr FPR16:$a, FPR16:$b)>;
 def : Pat<(fmaxnum_ieee (f64 FPR64:$a), (f64 FPR64:$b)),
           (FMAXNMDrr FPR64:$a, FPR64:$b)>;
+def : Pat<(f64 (fcanonicalize f64:$a)),
+          (FMINNMDrr   f64:$a, f64:$a)>;
+def : Pat<(fminnum_ieee (f32 FPR32:$a), (f32 FPR32:$b)),
+          (FMINNMSrr FPR32:$a, FPR32:$b)>;
 def : Pat<(fmaxnum_ieee (f32 FPR32:$a), (f32 FPR32:$b)),
           (FMAXNMSrr FPR32:$a, FPR32:$b)>;
+def : Pat<(f32 (fcanonicalize f32:$a)),
+          (FMINNMSrr   f32:$a, f32:$a)>;
+
+let Predicates = [HasFullFP16] in {
+def : Pat<(fminnum_ieee (f16 FPR16:$a), (f16 FPR16:$b)),
+          (FMINNMHrr FPR16:$a, FPR16:$b)>;
 def : Pat<(fmaxnum_ieee (f16 FPR16:$a), (f16 FPR16:$b)),
           (FMAXNMHrr FPR16:$a, FPR16:$b)>;
-
+def : Pat<(f16 (fcanonicalize f16:$a)),
+          (FMINNMHrr   f16:$a, f16:$a)>;
+}
 //===----------------------------------------------------------------------===//
 // Floating point three operand instructions.
 //===----------------------------------------------------------------------===//
@@ -5567,26 +5575,41 @@ defm FMINNM  : SIMDThreeSameVectorFP<0,1,0b000,"fminnm", any_fminnum>;
 defm FMINP   : SIMDThreeSameVectorFP<1,1,0b110,"fminp", int_aarch64_neon_fminp>;
 defm FMIN    : SIMDThreeSameVectorFP<0,1,0b110,"fmin", any_fminimum>;
 
+let Predicates = [HasNEON] in {
 def : Pat<(v2f64 (fminnum_ieee (v2f64 V128:$Rn), (v2f64 V128:$Rm))),
           (v2f64 (FMINNMv2f64 (v2f64 V128:$Rn), (v2f64 V128:$Rm)))>;
-def : Pat<(v4f32 (fminnum_ieee (v4f32 V128:$Rn), (v4f32 V128:$Rm))),
-          (v4f32 (FMINNMv4f32 (v4f32 V128:$Rn), (v4f32 V128:$Rm)))>;
-def : Pat<(v8f16 (fminnum_ieee (v8f16 V128:$Rn), (v8f16 V128:$Rm))),
-          (v8f16 (FMINNMv8f16 (v8f16 V128:$Rn), (v8f16 V128:$Rm)))>;
-def : Pat<(v2f32 (fminnum_ieee (v2f32 V64:$Rn), (v2f32 V64:$Rm))),
-          (v2f32 (FMINNMv2f32 (v2f32 V64:$Rn), (v2f32 V64:$Rm)))>;
-def : Pat<(v4f16 (fminnum_ieee (v4f16 V64:$Rn), (v4f16 V64:$Rm))),
-          (v4f16 (FMINNMv4f16 (v4f16 V64:$Rn), (v4f16 V64:$Rm)))>;
 def : Pat<(v2f64 (fmaxnum_ieee (v2f64 V128:$Rn), (v2f64 V128:$Rm))),
           (v2f64 (FMAXNMv2f64 (v2f64 V128:$Rn), (v2f64 V128:$Rm)))>;
+def : Pat<(v2f64 (fcanonicalize (v2f64 V128:$Rn))),
+          (v2f64 (FMINNMv2f64 (v2f64 V128:$Rn), (v2f64 V128:$Rn)))>;
+def : Pat<(v4f32 (fminnum_ieee (v4f32 V128:$Rn), (v4f32 V128:$Rm))),
+          (v4f32 (FMINNMv4f32 (v4f32 V128:$Rn), (v4f32 V128:$Rm)))>;
 def : Pat<(v4f32 (fmaxnum_ieee (v4f32 V128:$Rn), (v4f32 V128:$Rm))),
           (v4f32 (FMAXNMv4f32 (v4f32 V128:$Rn), (v4f32 V128:$Rm)))>;
-def : Pat<(v8f16 (fmaxnum_ieee (v8f16 V128:$Rn), (v8f16 V128:$Rm))),
-          (v8f16 (FMAXNMv8f16 (v8f16 V128:$Rn), (v8f16 V128:$Rm)))>;
+def : Pat<(v4f32 (fcanonicalize (v4f32 V128:$Rn))),
+          (v4f32 (FMINNMv4f32 (v4f32 V128:$Rn), (v4f32 V128:$Rn)))>;
+def : Pat<(v2f32 (fminnum_ieee (v2f32 V64:$Rn), (v2f32 V64:$Rm))),
+          (v2f32 (FMINNMv2f32 (v2f32 V64:$Rn), (v2f32 V64:$Rm)))>;
 def : Pat<(v2f32 (fmaxnum_ieee (v2f32 V64:$Rn), (v2f32 V64:$Rm))),
           (v2f32 (FMAXNMv2f32 (v2f32 V64:$Rn), (v2f32 V64:$Rm)))>;
+def : Pat<(v2f32 (fcanonicalize (v2f32 V64:$Rn))),
+          (v2f32 (FMINNMv2f32 (v2f32 V64:$Rn), (v2f32 V64:$Rn)))>;
+}
+
+let Predicates = [HasNEON, HasFullFP16] in {
+def : Pat<(v8f16 (fminnum_ieee (v8f16 V128:$Rn), (v8f16 V128:$Rm))),
+          (v8f16 (FMINNMv8f16 (v8f16 V128:$Rn), (v8f16 V128:$Rm)))>;
+def : Pat<(v8f16 (fmaxnum_ieee (v8f16 V128:$Rn), (v8f16 V128:$Rm))),
+          (v8f16 (FMAXNMv8f16 (v8f16 V128:$Rn), (v8f16 V128:$Rm)))>;
+def : Pat<(v8f16 (fcanonicalize (v8f16 V128:$Rn))),
+          (v8f16 (FMINNMv8f16 (v8f16 V128:$Rn), (v8f16 V128:$Rn)))>;
+def : Pat<(v4f16 (fminnum_ieee (v4f16 V64:$Rn), (v4f16 V64:$Rm))),
+          (v4f16 (FMINNMv4f16 (v4f16 V64:$Rn), (v4f16 V64:$Rm)))>;
 def : Pat<(v4f16 (fmaxnum_ieee (v4f16 V64:$Rn), (v4f16 V64:$Rm))),
           (v4f16 (FMAXNMv4f16 (v4f16 V64:$Rn), (v4f16 V64:$Rm)))>;
+def : Pat<(v4f16 (fcanonicalize (v4f16 V64:$Rn))),
+          (v4f16 (FMINNMv4f16 (v4f16 V64:$Rn), (v4f16 V64:$Rn)))>;
+}
 
 // NOTE: The operands of the PatFrag are reordered on FMLA/FMLS because the
 // instruction expects the addend first, while the fma intrinsic puts it last.
diff --git a/llvm/lib/Target/DirectX/DXILIntrinsicExpansion.cpp b/llvm/lib/Target/DirectX/DXILIntrinsicExpansion.cpp
index 1e84a7216013..fb5383b3514a 100644
--- a/llvm/lib/Target/DirectX/DXILIntrinsicExpansion.cpp
+++ b/llvm/lib/Target/DirectX/DXILIntrinsicExpansion.cpp
@@ -56,6 +56,7 @@ static bool isIntrinsicExpansion(Function &F) {
   case Intrinsic::dx_clamp:
   case Intrinsic::dx_cross:
   case Intrinsic::dx_uclamp:
+  case Intrinsic::dx_degrees:
   case Intrinsic::dx_lerp:
   case Intrinsic::dx_length:
   case Intrinsic::dx_normalize:
@@ -490,6 +491,14 @@ static Value *expandClampIntrinsic(CallInst *Orig,
                                  {MaxCall, Max}, nullptr, "dx.min");
 }
 
+static Value *expandDegreesIntrinsic(CallInst *Orig) {
+  Value *X = Orig->getOperand(0);
+  Type *Ty = X->getType();
+  IRBuilder<> Builder(Orig);
+  Value *DegreesRatio = ConstantFP::get(Ty, 180.0 * llvm::numbers::inv_pi);
+  return Builder.CreateFMul(X, DegreesRatio);
+}
+
 static Value *expandSignIntrinsic(CallInst *Orig) {
   Value *X = Orig->getOperand(0);
   Type *Ty = X->getType();
@@ -549,6 +558,9 @@ static bool expandIntrinsic(Function &F, CallInst *Orig) {
   case Intrinsic::dx_clamp:
     Result = expandClampIntrinsic(Orig, IntrinsicId);
     break;
+  case Intrinsic::dx_degrees:
+    Result = expandDegreesIntrinsic(Orig);
+    break;
   case Intrinsic::dx_lerp:
     Result = expandLerpIntrinsic(Orig);
     break;
diff --git a/llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp b/llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp
index e8b769b6fd69..fd92346717c4 100644
--- a/llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp
+++ b/llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp
@@ -2513,6 +2513,8 @@ bool SPIRVInstructionSelector::selectIntrinsic(Register ResVReg,
     return selectExtInst(ResVReg, ResType, I, CL::mix, GL::FMix);
   case Intrinsic::spv_length:
     return selectExtInst(ResVReg, ResType, I, CL::length, GL::Length);
+  case Intrinsic::spv_degrees:
+    return selectExtInst(ResVReg, ResType, I, CL::degrees, GL::Degrees);
   case Intrinsic::spv_frac:
     return selectExtInst(ResVReg, ResType, I, CL::fract, GL::Fract);
   case Intrinsic::spv_normalize:
diff --git a/llvm/test/CodeGen/AArch64/fp-fcanonicalize.ll b/llvm/test/CodeGen/AArch64/fp-fcanonicalize.ll
new file mode 100644
index 000000000000..753e2b734339
--- /dev/null
+++ b/llvm/test/CodeGen/AArch64/fp-fcanonicalize.ll
@@ -0,0 +1,587 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc --mtriple=aarch64 --mattr=-fullfp16,-neon < %s | FileCheck %s --check-prefix=CHECK-NOFP16-NONEON
+; RUN: llc --mtriple=aarch64 --mattr=+fullfp16,-neon < %s | FileCheck %s --check-prefix=CHECK-FP16-NONEON
+; RUN: llc --mtriple=aarch64 --mattr=-fullfp16,+neon < %s | FileCheck %s --check-prefix=CHECK-NOFP16-NEON
+; RUN: llc --mtriple=aarch64 --mattr=+fullfp16,+neon < %s | FileCheck %s --check-prefixes=CHECK-FP16-NEON
+
+declare half @llvm.fcanonicalize.f16(half)
+declare float @llvm.fcanonicalize.f32(float)
+declare double @llvm.fcanonicalize.f64(double)
+
+define half @fcanonicalize_f16(half %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_f16:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_f16:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_f16:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_f16:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call half @llvm.canonicalize.f16(half %x)
+  ret half %z
+}
+
+define half @fcanonicalize_f16_nnan(half %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_f16_nnan:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_f16_nnan:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_f16_nnan:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_f16_nnan:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call nnan half @llvm.canonicalize.f16(half %x)
+  ret half %z
+}
+
+define <2 x half> @fcanonicalize_v2f16(<2 x half> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v2f16:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s1, h1
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h1, s1
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v2f16:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NONEON-NEXT:    fminnm h1, h1, h1
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v2f16:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fcvtl v0.4s, v0.4h
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    fcvtn v0.4h, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v2f16:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.4h, v0.4h, v0.4h
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call <2 x half> @llvm.canonicalize.v2f16(<2 x half> %x)
+  ret <2 x half> %z
+}
+
+define <2 x half> @fcanonicalize_v2f16_nnan(<2 x half> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v2f16_nnan:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s1, h1
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h1, s1
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v2f16_nnan:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NONEON-NEXT:    fminnm h1, h1, h1
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v2f16_nnan:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fcvtl v0.4s, v0.4h
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    fcvtn v0.4h, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v2f16_nnan:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.4h, v0.4h, v0.4h
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call nnan <2 x half> @llvm.canonicalize.v2f16(<2 x half> %x)
+  ret <2 x half> %z
+}
+
+define <4 x half> @fcanonicalize_v4f16(<4 x half> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v4f16:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s1, h1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s2, h2
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s3, h3
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s2, s2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s3, s3, s3
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h3, s3
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v4f16:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NONEON-NEXT:    fminnm h1, h1, h1
+; CHECK-FP16-NONEON-NEXT:    fminnm h2, h2, h2
+; CHECK-FP16-NONEON-NEXT:    fminnm h3, h3, h3
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v4f16:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fcvtl v0.4s, v0.4h
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    fcvtn v0.4h, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v4f16:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.4h, v0.4h, v0.4h
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call <4 x half> @llvm.canonicalize.v4f16(<4 x half> %x)
+  ret <4 x half> %z
+}
+
+define <4 x half> @fcanonicalize_v4f16_nnan(<4 x half> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v4f16_nnan:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s1, h1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s2, h2
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s3, h3
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s2, s2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s3, s3, s3
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h3, s3
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v4f16_nnan:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NONEON-NEXT:    fminnm h1, h1, h1
+; CHECK-FP16-NONEON-NEXT:    fminnm h2, h2, h2
+; CHECK-FP16-NONEON-NEXT:    fminnm h3, h3, h3
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v4f16_nnan:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fcvtl v0.4s, v0.4h
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    fcvtn v0.4h, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v4f16_nnan:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.4h, v0.4h, v0.4h
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call nnan <4 x half> @llvm.canonicalize.v4f16(<4 x half> %x)
+  ret <4 x half> %z
+}
+
+define <8 x half> @fcanonicalize_v8f16(<8 x half> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v8f16:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s1, h1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s2, h2
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s3, h3
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s4, h4
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s5, h5
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s6, h6
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s7, h7
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s2, s2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s3, s3, s3
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s4, s4, s4
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s5, s5, s5
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s6, s6, s6
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s7, s7, s7
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h3, s3
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h4, s4
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h5, s5
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h6, s6
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h7, s7
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v8f16:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NONEON-NEXT:    fminnm h1, h1, h1
+; CHECK-FP16-NONEON-NEXT:    fminnm h2, h2, h2
+; CHECK-FP16-NONEON-NEXT:    fminnm h3, h3, h3
+; CHECK-FP16-NONEON-NEXT:    fminnm h4, h4, h4
+; CHECK-FP16-NONEON-NEXT:    fminnm h5, h5, h5
+; CHECK-FP16-NONEON-NEXT:    fminnm h6, h6, h6
+; CHECK-FP16-NONEON-NEXT:    fminnm h7, h7, h7
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v8f16:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fcvtl v1.4s, v0.4h
+; CHECK-NOFP16-NEON-NEXT:    fcvtl2 v2.4s, v0.8h
+; CHECK-NOFP16-NEON-NEXT:    fminnm v1.4s, v1.4s, v1.4s
+; CHECK-NOFP16-NEON-NEXT:    fcvtn v0.4h, v1.4s
+; CHECK-NOFP16-NEON-NEXT:    fminnm v1.4s, v2.4s, v2.4s
+; CHECK-NOFP16-NEON-NEXT:    fcvtn2 v0.8h, v1.4s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v8f16:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.8h, v0.8h, v0.8h
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call <8 x half> @llvm.canonicalize.v8f16(<8 x half> %x)
+  ret <8 x half> %z
+}
+
+define <8 x half> @fcanonicalize_v8f16_nnan(<8 x half> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v8f16_nnan:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s0, h0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s1, h1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s2, h2
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s3, h3
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s4, h4
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s5, h5
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s6, h6
+; CHECK-NOFP16-NONEON-NEXT:    fcvt s7, h7
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s2, s2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s3, s3, s3
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s4, s4, s4
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s5, s5, s5
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s6, s6, s6
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s7, s7, s7
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h3, s3
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h4, s4
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h5, s5
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h6, s6
+; CHECK-NOFP16-NONEON-NEXT:    fcvt h7, s7
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v8f16_nnan:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm h0, h0, h0
+; CHECK-FP16-NONEON-NEXT:    fminnm h1, h1, h1
+; CHECK-FP16-NONEON-NEXT:    fminnm h2, h2, h2
+; CHECK-FP16-NONEON-NEXT:    fminnm h3, h3, h3
+; CHECK-FP16-NONEON-NEXT:    fminnm h4, h4, h4
+; CHECK-FP16-NONEON-NEXT:    fminnm h5, h5, h5
+; CHECK-FP16-NONEON-NEXT:    fminnm h6, h6, h6
+; CHECK-FP16-NONEON-NEXT:    fminnm h7, h7, h7
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v8f16_nnan:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fcvtl v1.4s, v0.4h
+; CHECK-NOFP16-NEON-NEXT:    fcvtl2 v2.4s, v0.8h
+; CHECK-NOFP16-NEON-NEXT:    fminnm v1.4s, v1.4s, v1.4s
+; CHECK-NOFP16-NEON-NEXT:    fcvtn v0.4h, v1.4s
+; CHECK-NOFP16-NEON-NEXT:    fminnm v1.4s, v2.4s, v2.4s
+; CHECK-NOFP16-NEON-NEXT:    fcvtn2 v0.8h, v1.4s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v8f16_nnan:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.8h, v0.8h, v0.8h
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call nnan <8 x half> @llvm.canonicalize.v8f16(<8 x half> %x)
+  ret <8 x half> %z
+}
+
+define float @fcanonicalize_f32(float %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_f32:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_f32:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_f32:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_f32:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm s0, s0, s0
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call float @llvm.canonicalize.f32(float %x)
+  ret float %z
+}
+
+define float @fcanonicalize_f32_nnan(float %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_f32_nnan:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_f32_nnan:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_f32_nnan:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_f32_nnan:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm s0, s0, s0
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call nnan float @llvm.canonicalize.f32(float %x)
+  ret float %z
+}
+
+define <2 x float> @fcanonicalize_v2f32(<2 x float> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v2f32:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v2f32:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-FP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v2f32:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.2s, v0.2s, v0.2s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v2f32:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.2s, v0.2s, v0.2s
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call <2 x float> @llvm.canonicalize.v2f32(<2 x float> %x)
+  ret <2 x float> %z
+}
+
+define <2 x float> @fcanonicalize_v2f32_nnan(<2 x float> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v2f32_nnan:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v2f32_nnan:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-FP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v2f32_nnan:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.2s, v0.2s, v0.2s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v2f32_nnan:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.2s, v0.2s, v0.2s
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call nnan <2 x float> @llvm.canonicalize.v2f32(<2 x float> %x)
+  ret <2 x float> %z
+}
+
+define <4 x float> @fcanonicalize_v4f32(<4 x float> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v4f32:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s2, s2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s3, s3, s3
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v4f32:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-FP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-FP16-NONEON-NEXT:    fminnm s2, s2, s2
+; CHECK-FP16-NONEON-NEXT:    fminnm s3, s3, s3
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v4f32:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v4f32:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call <4 x float> @llvm.canonicalize.v4f32(<4 x float> %x)
+  ret <4 x float> %z
+}
+
+define <4 x float> @fcanonicalize_v4f32_nnan(<4 x float> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v4f32_nnan:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s2, s2, s2
+; CHECK-NOFP16-NONEON-NEXT:    fminnm s3, s3, s3
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v4f32_nnan:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm s0, s0, s0
+; CHECK-FP16-NONEON-NEXT:    fminnm s1, s1, s1
+; CHECK-FP16-NONEON-NEXT:    fminnm s2, s2, s2
+; CHECK-FP16-NONEON-NEXT:    fminnm s3, s3, s3
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v4f32_nnan:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v4f32_nnan:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call nnan <4 x float> @llvm.canonicalize.v4f32(<4 x float> %x)
+  ret <4 x float> %z
+}
+
+define double @fcanonicalize_f64(double %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_f64:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm d0, d0, d0
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_f64:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm d0, d0, d0
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_f64:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm d0, d0, d0
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_f64:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm d0, d0, d0
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call double @llvm.canonicalize.f64(double %x)
+  ret double %z
+}
+
+define double @fcanonicalize_f64_nnan(double %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_f64_nnan:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm d0, d0, d0
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_f64_nnan:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm d0, d0, d0
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_f64_nnan:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm d0, d0, d0
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_f64_nnan:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm d0, d0, d0
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call nnan double @llvm.canonicalize.f64(double %x)
+  ret double %z
+}
+
+define <2 x double> @fcanonicalize_v2f64(<2 x double> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v2f64:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm d0, d0, d0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm d1, d1, d1
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v2f64:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm d0, d0, d0
+; CHECK-FP16-NONEON-NEXT:    fminnm d1, d1, d1
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v2f64:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v2f64:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call <2 x double> @llvm.canonicalize.v2f64(<2 x double> %x)
+  ret <2 x double> %z
+}
+
+define <2 x double> @fcanonicalize_v2f64_nnan(<2 x double> %x) {
+; CHECK-NOFP16-NONEON-LABEL: fcanonicalize_v2f64_nnan:
+; CHECK-NOFP16-NONEON:       // %bb.0:
+; CHECK-NOFP16-NONEON-NEXT:    fminnm d0, d0, d0
+; CHECK-NOFP16-NONEON-NEXT:    fminnm d1, d1, d1
+; CHECK-NOFP16-NONEON-NEXT:    ret
+;
+; CHECK-FP16-NONEON-LABEL: fcanonicalize_v2f64_nnan:
+; CHECK-FP16-NONEON:       // %bb.0:
+; CHECK-FP16-NONEON-NEXT:    fminnm d0, d0, d0
+; CHECK-FP16-NONEON-NEXT:    fminnm d1, d1, d1
+; CHECK-FP16-NONEON-NEXT:    ret
+;
+; CHECK-NOFP16-NEON-LABEL: fcanonicalize_v2f64_nnan:
+; CHECK-NOFP16-NEON:       // %bb.0:
+; CHECK-NOFP16-NEON-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; CHECK-NOFP16-NEON-NEXT:    ret
+;
+; CHECK-FP16-NEON-LABEL: fcanonicalize_v2f64_nnan:
+; CHECK-FP16-NEON:       // %bb.0:
+; CHECK-FP16-NEON-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; CHECK-FP16-NEON-NEXT:    ret
+  %z = call nnan <2 x double> @llvm.canonicalize.v2f64(<2 x double> %x)
+  ret <2 x double> %z
+}
diff --git a/llvm/test/CodeGen/AArch64/fp-maximumnum-minimumnum.ll b/llvm/test/CodeGen/AArch64/fp-maximumnum-minimumnum.ll
index b8406179f3cb..bb3f9a3e52a1 100644
--- a/llvm/test/CodeGen/AArch64/fp-maximumnum-minimumnum.ll
+++ b/llvm/test/CodeGen/AArch64/fp-maximumnum-minimumnum.ll
@@ -472,3 +472,563 @@ entry:
   %c = call nnan <16 x half> @llvm.minimumnum.v16f16(<16 x half> %a, <16 x half> %b)
   ret <16 x half> %c
 }
+
+;;;;;;;;;;;;;;;;  max_f64
+define double @max_f64(double %a, double %b) {
+; AARCH64-LABEL: max_f64:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm d1, d1, d1
+; AARCH64-NEXT:    fminnm d0, d0, d0
+; AARCH64-NEXT:    fmaxnm d0, d0, d1
+; AARCH64-NEXT:    ret
+entry:
+  %c = call double @llvm.maximumnum.f64(double %a, double %b)
+  ret double %c
+}
+
+define <2 x double> @max_v2f64(<2 x double> %a, <2 x double> %b) {
+; AARCH64-LABEL: max_v2f64:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.2d, v1.2d, v1.2d
+; AARCH64-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; AARCH64-NEXT:    fmaxnm v0.2d, v0.2d, v1.2d
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <2 x double> @llvm.maximumnum.v2f64(<2 x double> %a, <2 x double> %b)
+  ret <2 x double> %c
+}
+
+define <3 x double> @max_v3f64(<3 x double> %a, <3 x double> %b) {
+; AARCH64-LABEL: max_v3f64:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    // kill: def $d3 killed $d3 def $q3
+; AARCH64-NEXT:    // kill: def $d0 killed $d0 def $q0
+; AARCH64-NEXT:    // kill: def $d4 killed $d4 def $q4
+; AARCH64-NEXT:    // kill: def $d1 killed $d1 def $q1
+; AARCH64-NEXT:    // kill: def $d2 killed $d2 def $q2
+; AARCH64-NEXT:    // kill: def $d5 killed $d5 def $q5
+; AARCH64-NEXT:    mov v0.d[1], v1.d[0]
+; AARCH64-NEXT:    mov v3.d[1], v4.d[0]
+; AARCH64-NEXT:    fminnm v2.2d, v2.2d, v2.2d
+; AARCH64-NEXT:    fminnm v1.2d, v3.2d, v3.2d
+; AARCH64-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; AARCH64-NEXT:    fmaxnm v0.2d, v0.2d, v1.2d
+; AARCH64-NEXT:    fminnm v1.2d, v5.2d, v5.2d
+; AARCH64-NEXT:    fmaxnm v2.2d, v2.2d, v1.2d
+; AARCH64-NEXT:    ext v1.16b, v0.16b, v0.16b, #8
+; AARCH64-NEXT:    // kill: def $d0 killed $d0 killed $q0
+; AARCH64-NEXT:    // kill: def $d1 killed $d1 killed $q1
+; AARCH64-NEXT:    // kill: def $d2 killed $d2 killed $q2
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <3 x double> @llvm.maximumnum.v3f64(<3 x double> %a, <3 x double> %b)
+  ret <3 x double> %c
+}
+
+define <4 x double> @max_v4f64(<4 x double> %a, <4 x double> %b) {
+; AARCH64-LABEL: max_v4f64:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v2.2d, v2.2d, v2.2d
+; AARCH64-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; AARCH64-NEXT:    fminnm v3.2d, v3.2d, v3.2d
+; AARCH64-NEXT:    fminnm v1.2d, v1.2d, v1.2d
+; AARCH64-NEXT:    fmaxnm v0.2d, v0.2d, v2.2d
+; AARCH64-NEXT:    fmaxnm v1.2d, v1.2d, v3.2d
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <4 x double> @llvm.maximumnum.v4f64(<4 x double> %a, <4 x double> %b)
+  ret <4 x double> %c
+}
+
+;;;;;;;;;;;;;;;;;; max_f32
+define float @max_f32(float %a, float %b) {
+; AARCH64-LABEL: max_f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm s1, s1, s1
+; AARCH64-NEXT:    fminnm s0, s0, s0
+; AARCH64-NEXT:    fmaxnm s0, s0, s1
+; AARCH64-NEXT:    ret
+entry:
+  %c = call float @llvm.maximumnum.f32(float %a, float %b)
+  ret float %c
+}
+
+define <2 x float> @max_v2f32(<2 x float> %a, <2 x float> %b) {
+; AARCH64-LABEL: max_v2f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.2s, v1.2s, v1.2s
+; AARCH64-NEXT:    fminnm v0.2s, v0.2s, v0.2s
+; AARCH64-NEXT:    fmaxnm v0.2s, v0.2s, v1.2s
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <2 x float> @llvm.maximumnum.v2f32(<2 x float> %a, <2 x float> %b)
+  ret <2 x float> %c
+}
+
+define <3 x float> @max_v3f32(<3 x float> %a, <3 x float> %b) {
+; AARCH64-LABEL: max_v3f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.4s, v1.4s, v1.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; AARCH64-NEXT:    fmaxnm v0.4s, v0.4s, v1.4s
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <3 x float> @llvm.maximumnum.v3f32(<3 x float> %a, <3 x float> %b)
+  ret <3 x float> %c
+}
+
+define <4 x float> @max_v4f32(<4 x float> %a, <4 x float> %b) {
+; AARCH64-LABEL: max_v4f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.4s, v1.4s, v1.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; AARCH64-NEXT:    fmaxnm v0.4s, v0.4s, v1.4s
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <4 x float> @llvm.maximumnum.v4f32(<4 x float> %a, <4 x float> %b)
+  ret <4 x float> %c
+}
+
+define <5 x float> @max_v5f32(<5 x float> %a, <5 x float> %b) {
+; AARCH64-LABEL: max_v5f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    // kill: def $s0 killed $s0 def $q0
+; AARCH64-NEXT:    // kill: def $s5 killed $s5 def $q5
+; AARCH64-NEXT:    // kill: def $s1 killed $s1 def $q1
+; AARCH64-NEXT:    // kill: def $s6 killed $s6 def $q6
+; AARCH64-NEXT:    // kill: def $s2 killed $s2 def $q2
+; AARCH64-NEXT:    // kill: def $s7 killed $s7 def $q7
+; AARCH64-NEXT:    // kill: def $s3 killed $s3 def $q3
+; AARCH64-NEXT:    mov x8, sp
+; AARCH64-NEXT:    // kill: def $s4 killed $s4 def $q4
+; AARCH64-NEXT:    mov v0.s[1], v1.s[0]
+; AARCH64-NEXT:    mov v5.s[1], v6.s[0]
+; AARCH64-NEXT:    mov v0.s[2], v2.s[0]
+; AARCH64-NEXT:    mov v5.s[2], v7.s[0]
+; AARCH64-NEXT:    ldr s2, [sp, #8]
+; AARCH64-NEXT:    fminnm v2.4s, v2.4s, v2.4s
+; AARCH64-NEXT:    mov v0.s[3], v3.s[0]
+; AARCH64-NEXT:    ld1 { v5.s }[3], [x8]
+; AARCH64-NEXT:    fminnm v3.4s, v4.4s, v4.4s
+; AARCH64-NEXT:    fminnm v1.4s, v5.4s, v5.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; AARCH64-NEXT:    fmaxnm v4.4s, v3.4s, v2.4s
+; AARCH64-NEXT:    // kill: def $s4 killed $s4 killed $q4
+; AARCH64-NEXT:    fmaxnm v0.4s, v0.4s, v1.4s
+; AARCH64-NEXT:    mov s1, v0.s[1]
+; AARCH64-NEXT:    mov s2, v0.s[2]
+; AARCH64-NEXT:    mov s3, v0.s[3]
+; AARCH64-NEXT:    // kill: def $s0 killed $s0 killed $q0
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <5 x float> @llvm.maximumnum.v5f32(<5 x float> %a, <5 x float> %b)
+  ret <5 x float> %c
+}
+
+define <8 x float> @max_v8f32(<8 x float> %a, <8 x float> %b) {
+; AARCH64-LABEL: max_v8f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v2.4s, v2.4s, v2.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; AARCH64-NEXT:    fminnm v3.4s, v3.4s, v3.4s
+; AARCH64-NEXT:    fminnm v1.4s, v1.4s, v1.4s
+; AARCH64-NEXT:    fmaxnm v0.4s, v0.4s, v2.4s
+; AARCH64-NEXT:    fmaxnm v1.4s, v1.4s, v3.4s
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <8 x float> @llvm.maximumnum.v8f32(<8 x float> %a, <8 x float> %b)
+  ret <8 x float> %c
+}
+
+;;;;;;;;;;;;;;;;;; max_f16
+define half @max_f16(half %a, half %b) {
+; AARCH64-LABEL: max_f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm h1, h1, h1
+; AARCH64-NEXT:    fminnm h0, h0, h0
+; AARCH64-NEXT:    fmaxnm h0, h0, h1
+; AARCH64-NEXT:    ret
+entry:
+  %c = call half @llvm.maximumnum.f16(half %a, half %b)
+  ret half %c
+}
+
+define <2 x half> @max_v2f16(<2 x half> %a, <2 x half> %b) {
+; AARCH64-LABEL: max_v2f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.4h, v1.4h, v1.4h
+; AARCH64-NEXT:    fminnm v0.4h, v0.4h, v0.4h
+; AARCH64-NEXT:    fmaxnm v0.4h, v0.4h, v1.4h
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <2 x half> @llvm.maximumnum.v2f16(<2 x half> %a, <2 x half> %b)
+  ret <2 x half> %c
+}
+
+define <4 x half> @max_v4f16(<4 x half> %a, <4 x half> %b) {
+; AARCH64-LABEL: max_v4f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.4h, v1.4h, v1.4h
+; AARCH64-NEXT:    fminnm v0.4h, v0.4h, v0.4h
+; AARCH64-NEXT:    fmaxnm v0.4h, v0.4h, v1.4h
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <4 x half> @llvm.maximumnum.v4f16(<4 x half> %a, <4 x half> %b)
+  ret <4 x half> %c
+}
+
+define <8 x half> @max_v8f16(<8 x half> %a, <8 x half> %b) {
+; AARCH64-LABEL: max_v8f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.8h, v1.8h, v1.8h
+; AARCH64-NEXT:    fminnm v0.8h, v0.8h, v0.8h
+; AARCH64-NEXT:    fmaxnm v0.8h, v0.8h, v1.8h
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <8 x half> @llvm.maximumnum.v8f16(<8 x half> %a, <8 x half> %b)
+  ret <8 x half> %c
+}
+
+define <9 x half> @max_v9f16(<9 x half> %a, <9 x half> %b) {
+; AARCH64-LABEL: max_v9f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    // kill: def $h0 killed $h0 def $q0
+; AARCH64-NEXT:    // kill: def $h1 killed $h1 def $q1
+; AARCH64-NEXT:    // kill: def $h2 killed $h2 def $q2
+; AARCH64-NEXT:    add x9, sp, #16
+; AARCH64-NEXT:    // kill: def $h3 killed $h3 def $q3
+; AARCH64-NEXT:    // kill: def $h4 killed $h4 def $q4
+; AARCH64-NEXT:    // kill: def $h5 killed $h5 def $q5
+; AARCH64-NEXT:    // kill: def $h6 killed $h6 def $q6
+; AARCH64-NEXT:    // kill: def $h7 killed $h7 def $q7
+; AARCH64-NEXT:    mov v0.h[1], v1.h[0]
+; AARCH64-NEXT:    ldr h1, [sp, #8]
+; AARCH64-NEXT:    ld1 { v1.h }[1], [x9]
+; AARCH64-NEXT:    add x9, sp, #24
+; AARCH64-NEXT:    mov v0.h[2], v2.h[0]
+; AARCH64-NEXT:    ldr h2, [sp]
+; AARCH64-NEXT:    ld1 { v1.h }[2], [x9]
+; AARCH64-NEXT:    add x9, sp, #32
+; AARCH64-NEXT:    fminnm v2.8h, v2.8h, v2.8h
+; AARCH64-NEXT:    mov v0.h[3], v3.h[0]
+; AARCH64-NEXT:    ld1 { v1.h }[3], [x9]
+; AARCH64-NEXT:    add x9, sp, #40
+; AARCH64-NEXT:    ldr h3, [sp, #72]
+; AARCH64-NEXT:    ld1 { v1.h }[4], [x9]
+; AARCH64-NEXT:    add x9, sp, #48
+; AARCH64-NEXT:    fminnm v3.8h, v3.8h, v3.8h
+; AARCH64-NEXT:    mov v0.h[4], v4.h[0]
+; AARCH64-NEXT:    ld1 { v1.h }[5], [x9]
+; AARCH64-NEXT:    add x9, sp, #56
+; AARCH64-NEXT:    fmaxnm v2.8h, v2.8h, v3.8h
+; AARCH64-NEXT:    mov v0.h[5], v5.h[0]
+; AARCH64-NEXT:    ld1 { v1.h }[6], [x9]
+; AARCH64-NEXT:    add x9, sp, #64
+; AARCH64-NEXT:    str h2, [x8, #16]
+; AARCH64-NEXT:    mov v0.h[6], v6.h[0]
+; AARCH64-NEXT:    ld1 { v1.h }[7], [x9]
+; AARCH64-NEXT:    fminnm v1.8h, v1.8h, v1.8h
+; AARCH64-NEXT:    mov v0.h[7], v7.h[0]
+; AARCH64-NEXT:    fminnm v0.8h, v0.8h, v0.8h
+; AARCH64-NEXT:    fmaxnm v0.8h, v0.8h, v1.8h
+; AARCH64-NEXT:    str q0, [x8]
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <9 x half> @llvm.maximumnum.v9f16(<9 x half> %a, <9 x half> %b)
+  ret <9 x half> %c
+}
+
+define <16 x half> @max_v16f16(<16 x half> %a, <16 x half> %b) {
+; AARCH64-LABEL: max_v16f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v2.8h, v2.8h, v2.8h
+; AARCH64-NEXT:    fminnm v0.8h, v0.8h, v0.8h
+; AARCH64-NEXT:    fminnm v3.8h, v3.8h, v3.8h
+; AARCH64-NEXT:    fminnm v1.8h, v1.8h, v1.8h
+; AARCH64-NEXT:    fmaxnm v0.8h, v0.8h, v2.8h
+; AARCH64-NEXT:    fmaxnm v1.8h, v1.8h, v3.8h
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <16 x half> @llvm.maximumnum.v16f16(<16 x half> %a, <16 x half> %b)
+  ret <16 x half> %c
+}
+
+;;;;;;;;;;;;;;;;  min_f64
+define double @min_f64(double %a, double %b) {
+; AARCH64-LABEL: min_f64:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm d1, d1, d1
+; AARCH64-NEXT:    fminnm d0, d0, d0
+; AARCH64-NEXT:    fminnm d0, d0, d1
+; AARCH64-NEXT:    ret
+entry:
+  %c = call double @llvm.minimumnum.f64(double %a, double %b)
+  ret double %c
+}
+
+define <2 x double> @min_v2f64(<2 x double> %a, <2 x double> %b) {
+; AARCH64-LABEL: min_v2f64:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.2d, v1.2d, v1.2d
+; AARCH64-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; AARCH64-NEXT:    fminnm v0.2d, v0.2d, v1.2d
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <2 x double> @llvm.minimumnum.v2f64(<2 x double> %a, <2 x double> %b)
+  ret <2 x double> %c
+}
+
+define <3 x double> @min_v3f64(<3 x double> %a, <3 x double> %b) {
+; AARCH64-LABEL: min_v3f64:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    // kill: def $d3 killed $d3 def $q3
+; AARCH64-NEXT:    // kill: def $d0 killed $d0 def $q0
+; AARCH64-NEXT:    // kill: def $d4 killed $d4 def $q4
+; AARCH64-NEXT:    // kill: def $d1 killed $d1 def $q1
+; AARCH64-NEXT:    // kill: def $d2 killed $d2 def $q2
+; AARCH64-NEXT:    // kill: def $d5 killed $d5 def $q5
+; AARCH64-NEXT:    mov v0.d[1], v1.d[0]
+; AARCH64-NEXT:    mov v3.d[1], v4.d[0]
+; AARCH64-NEXT:    fminnm v2.2d, v2.2d, v2.2d
+; AARCH64-NEXT:    fminnm v1.2d, v3.2d, v3.2d
+; AARCH64-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; AARCH64-NEXT:    fminnm v0.2d, v0.2d, v1.2d
+; AARCH64-NEXT:    fminnm v1.2d, v5.2d, v5.2d
+; AARCH64-NEXT:    fminnm v2.2d, v2.2d, v1.2d
+; AARCH64-NEXT:    ext v1.16b, v0.16b, v0.16b, #8
+; AARCH64-NEXT:    // kill: def $d0 killed $d0 killed $q0
+; AARCH64-NEXT:    // kill: def $d1 killed $d1 killed $q1
+; AARCH64-NEXT:    // kill: def $d2 killed $d2 killed $q2
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <3 x double> @llvm.minimumnum.v3f64(<3 x double> %a, <3 x double> %b)
+  ret <3 x double> %c
+}
+
+define <4 x double> @min_v4f64(<4 x double> %a, <4 x double> %b) {
+; AARCH64-LABEL: min_v4f64:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v2.2d, v2.2d, v2.2d
+; AARCH64-NEXT:    fminnm v0.2d, v0.2d, v0.2d
+; AARCH64-NEXT:    fminnm v3.2d, v3.2d, v3.2d
+; AARCH64-NEXT:    fminnm v1.2d, v1.2d, v1.2d
+; AARCH64-NEXT:    fminnm v0.2d, v0.2d, v2.2d
+; AARCH64-NEXT:    fminnm v1.2d, v1.2d, v3.2d
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <4 x double> @llvm.minimumnum.v4f64(<4 x double> %a, <4 x double> %b)
+  ret <4 x double> %c
+}
+
+;;;;;;;;;;;;;;;;;; min_f32
+define float @min_f32(float %a, float %b) {
+; AARCH64-LABEL: min_f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm s1, s1, s1
+; AARCH64-NEXT:    fminnm s0, s0, s0
+; AARCH64-NEXT:    fminnm s0, s0, s1
+; AARCH64-NEXT:    ret
+entry:
+  %c = call float @llvm.minimumnum.f32(float %a, float %b)
+  ret float %c
+}
+
+define <2 x float> @min_v2f32(<2 x float> %a, <2 x float> %b) {
+; AARCH64-LABEL: min_v2f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.2s, v1.2s, v1.2s
+; AARCH64-NEXT:    fminnm v0.2s, v0.2s, v0.2s
+; AARCH64-NEXT:    fminnm v0.2s, v0.2s, v1.2s
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <2 x float> @llvm.minimumnum.v2f32(<2 x float> %a, <2 x float> %b)
+  ret <2 x float> %c
+}
+
+define <3 x float> @min_v3f32(<3 x float> %a, <3 x float> %b) {
+; AARCH64-LABEL: min_v3f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.4s, v1.4s, v1.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v1.4s
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <3 x float> @llvm.minimumnum.v3f32(<3 x float> %a, <3 x float> %b)
+  ret <3 x float> %c
+}
+
+define <4 x float> @min_v4f32(<4 x float> %a, <4 x float> %b) {
+; AARCH64-LABEL: min_v4f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.4s, v1.4s, v1.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v1.4s
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <4 x float> @llvm.minimumnum.v4f32(<4 x float> %a, <4 x float> %b)
+  ret <4 x float> %c
+}
+
+define <5 x float> @min_v5f32(<5 x float> %a, <5 x float> %b) {
+; AARCH64-LABEL: min_v5f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    // kill: def $s0 killed $s0 def $q0
+; AARCH64-NEXT:    // kill: def $s5 killed $s5 def $q5
+; AARCH64-NEXT:    // kill: def $s1 killed $s1 def $q1
+; AARCH64-NEXT:    // kill: def $s6 killed $s6 def $q6
+; AARCH64-NEXT:    // kill: def $s2 killed $s2 def $q2
+; AARCH64-NEXT:    // kill: def $s7 killed $s7 def $q7
+; AARCH64-NEXT:    // kill: def $s3 killed $s3 def $q3
+; AARCH64-NEXT:    mov x8, sp
+; AARCH64-NEXT:    // kill: def $s4 killed $s4 def $q4
+; AARCH64-NEXT:    mov v0.s[1], v1.s[0]
+; AARCH64-NEXT:    mov v5.s[1], v6.s[0]
+; AARCH64-NEXT:    mov v0.s[2], v2.s[0]
+; AARCH64-NEXT:    mov v5.s[2], v7.s[0]
+; AARCH64-NEXT:    ldr s2, [sp, #8]
+; AARCH64-NEXT:    fminnm v2.4s, v2.4s, v2.4s
+; AARCH64-NEXT:    mov v0.s[3], v3.s[0]
+; AARCH64-NEXT:    ld1 { v5.s }[3], [x8]
+; AARCH64-NEXT:    fminnm v3.4s, v4.4s, v4.4s
+; AARCH64-NEXT:    fminnm v1.4s, v5.4s, v5.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; AARCH64-NEXT:    fminnm v4.4s, v3.4s, v2.4s
+; AARCH64-NEXT:    // kill: def $s4 killed $s4 killed $q4
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v1.4s
+; AARCH64-NEXT:    mov s1, v0.s[1]
+; AARCH64-NEXT:    mov s2, v0.s[2]
+; AARCH64-NEXT:    mov s3, v0.s[3]
+; AARCH64-NEXT:    // kill: def $s0 killed $s0 killed $q0
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <5 x float> @llvm.minimumnum.v5f32(<5 x float> %a, <5 x float> %b)
+  ret <5 x float> %c
+}
+
+define <8 x float> @min_v8f32(<8 x float> %a, <8 x float> %b) {
+; AARCH64-LABEL: min_v8f32:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v2.4s, v2.4s, v2.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v0.4s
+; AARCH64-NEXT:    fminnm v3.4s, v3.4s, v3.4s
+; AARCH64-NEXT:    fminnm v1.4s, v1.4s, v1.4s
+; AARCH64-NEXT:    fminnm v0.4s, v0.4s, v2.4s
+; AARCH64-NEXT:    fminnm v1.4s, v1.4s, v3.4s
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <8 x float> @llvm.minimumnum.v8f32(<8 x float> %a, <8 x float> %b)
+  ret <8 x float> %c
+}
+
+;;;;;;;;;;;;;;;;;; min_f16
+define half @min_f16(half %a, half %b) {
+; AARCH64-LABEL: min_f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm h1, h1, h1
+; AARCH64-NEXT:    fminnm h0, h0, h0
+; AARCH64-NEXT:    fminnm h0, h0, h1
+; AARCH64-NEXT:    ret
+entry:
+  %c = call half @llvm.minimumnum.f16(half %a, half %b)
+  ret half %c
+}
+
+define <2 x half> @min_v2f16(<2 x half> %a, <2 x half> %b) {
+; AARCH64-LABEL: min_v2f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.4h, v1.4h, v1.4h
+; AARCH64-NEXT:    fminnm v0.4h, v0.4h, v0.4h
+; AARCH64-NEXT:    fminnm v0.4h, v0.4h, v1.4h
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <2 x half> @llvm.minimumnum.v2f16(<2 x half> %a, <2 x half> %b)
+  ret <2 x half> %c
+}
+
+define <4 x half> @min_v4f16(<4 x half> %a, <4 x half> %b) {
+; AARCH64-LABEL: min_v4f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.4h, v1.4h, v1.4h
+; AARCH64-NEXT:    fminnm v0.4h, v0.4h, v0.4h
+; AARCH64-NEXT:    fminnm v0.4h, v0.4h, v1.4h
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <4 x half> @llvm.minimumnum.v4f16(<4 x half> %a, <4 x half> %b)
+  ret <4 x half> %c
+}
+
+define <8 x half> @min_v8f16(<8 x half> %a, <8 x half> %b) {
+; AARCH64-LABEL: min_v8f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v1.8h, v1.8h, v1.8h
+; AARCH64-NEXT:    fminnm v0.8h, v0.8h, v0.8h
+; AARCH64-NEXT:    fminnm v0.8h, v0.8h, v1.8h
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <8 x half> @llvm.minimumnum.v8f16(<8 x half> %a, <8 x half> %b)
+  ret <8 x half> %c
+}
+
+define <9 x half> @min_v9f16(<9 x half> %a, <9 x half> %b) {
+; AARCH64-LABEL: min_v9f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    // kill: def $h0 killed $h0 def $q0
+; AARCH64-NEXT:    // kill: def $h1 killed $h1 def $q1
+; AARCH64-NEXT:    // kill: def $h2 killed $h2 def $q2
+; AARCH64-NEXT:    add x9, sp, #16
+; AARCH64-NEXT:    // kill: def $h3 killed $h3 def $q3
+; AARCH64-NEXT:    // kill: def $h4 killed $h4 def $q4
+; AARCH64-NEXT:    // kill: def $h5 killed $h5 def $q5
+; AARCH64-NEXT:    // kill: def $h6 killed $h6 def $q6
+; AARCH64-NEXT:    // kill: def $h7 killed $h7 def $q7
+; AARCH64-NEXT:    mov v0.h[1], v1.h[0]
+; AARCH64-NEXT:    ldr h1, [sp, #8]
+; AARCH64-NEXT:    ld1 { v1.h }[1], [x9]
+; AARCH64-NEXT:    add x9, sp, #24
+; AARCH64-NEXT:    mov v0.h[2], v2.h[0]
+; AARCH64-NEXT:    ldr h2, [sp]
+; AARCH64-NEXT:    ld1 { v1.h }[2], [x9]
+; AARCH64-NEXT:    add x9, sp, #32
+; AARCH64-NEXT:    fminnm v2.8h, v2.8h, v2.8h
+; AARCH64-NEXT:    mov v0.h[3], v3.h[0]
+; AARCH64-NEXT:    ld1 { v1.h }[3], [x9]
+; AARCH64-NEXT:    add x9, sp, #40
+; AARCH64-NEXT:    ldr h3, [sp, #72]
+; AARCH64-NEXT:    ld1 { v1.h }[4], [x9]
+; AARCH64-NEXT:    add x9, sp, #48
+; AARCH64-NEXT:    fminnm v3.8h, v3.8h, v3.8h
+; AARCH64-NEXT:    mov v0.h[4], v4.h[0]
+; AARCH64-NEXT:    ld1 { v1.h }[5], [x9]
+; AARCH64-NEXT:    add x9, sp, #56
+; AARCH64-NEXT:    fminnm v2.8h, v2.8h, v3.8h
+; AARCH64-NEXT:    mov v0.h[5], v5.h[0]
+; AARCH64-NEXT:    ld1 { v1.h }[6], [x9]
+; AARCH64-NEXT:    add x9, sp, #64
+; AARCH64-NEXT:    str h2, [x8, #16]
+; AARCH64-NEXT:    mov v0.h[6], v6.h[0]
+; AARCH64-NEXT:    ld1 { v1.h }[7], [x9]
+; AARCH64-NEXT:    fminnm v1.8h, v1.8h, v1.8h
+; AARCH64-NEXT:    mov v0.h[7], v7.h[0]
+; AARCH64-NEXT:    fminnm v0.8h, v0.8h, v0.8h
+; AARCH64-NEXT:    fminnm v0.8h, v0.8h, v1.8h
+; AARCH64-NEXT:    str q0, [x8]
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <9 x half> @llvm.minimumnum.v9f16(<9 x half> %a, <9 x half> %b)
+  ret <9 x half> %c
+}
+
+define <16 x half> @min_v16f16(<16 x half> %a, <16 x half> %b) {
+; AARCH64-LABEL: min_v16f16:
+; AARCH64:       // %bb.0: // %entry
+; AARCH64-NEXT:    fminnm v2.8h, v2.8h, v2.8h
+; AARCH64-NEXT:    fminnm v0.8h, v0.8h, v0.8h
+; AARCH64-NEXT:    fminnm v3.8h, v3.8h, v3.8h
+; AARCH64-NEXT:    fminnm v1.8h, v1.8h, v1.8h
+; AARCH64-NEXT:    fminnm v0.8h, v0.8h, v2.8h
+; AARCH64-NEXT:    fminnm v1.8h, v1.8h, v3.8h
+; AARCH64-NEXT:    ret
+entry:
+  %c = call <16 x half> @llvm.minimumnum.v16f16(<16 x half> %a, <16 x half> %b)
+  ret <16 x half> %c
+}
diff --git a/llvm/test/CodeGen/DirectX/degrees.ll b/llvm/test/CodeGen/DirectX/degrees.ll
new file mode 100644
index 000000000000..b38ac13d5f24
--- /dev/null
+++ b/llvm/test/CodeGen/DirectX/degrees.ll
@@ -0,0 +1,54 @@
+; RUN: opt -S -dxil-intrinsic-expansion -scalarizer -dxil-op-lower -mtriple=dxil-pc-shadermodel6.3-library %s | FileCheck %s
+
+; Make sure dxil op function calls for degrees are expanded and lowered as fmul for float and half.
+
+define noundef half @degrees_half(half noundef %a) {
+; CHECK-LABEL: define noundef half @degrees_half(
+; CHECK-SAME: half noundef [[A:%.*]]) {
+; CHECK-NEXT:  [[ENTRY:.*:]]
+; CHECK-NEXT:    [[DX_DEGREES1:%.*]] = fmul half [[A]], 0xH5329
+; CHECK-NEXT:    ret half [[DX_DEGREES1]]
+;
+entry:
+  %dx.degrees = call half @llvm.dx.degrees.f16(half %a)
+  ret half %dx.degrees
+}
+
+define noundef float @degrees_float(float noundef %a) #0 {
+; CHECK-LABEL: define noundef float @degrees_float(
+; CHECK-SAME: float noundef [[A:%.*]]) {
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[DEGREES:%.*]] = fmul float [[A]], 0x404CA5DC20000000
+; CHECK-NEXT:    ret float [[DEGREES]]
+;
+entry:
+  %dx.degrees = call float @llvm.dx.degrees.f32(float %a)
+  ret float %dx.degrees
+}
+
+define noundef <4 x float> @degrees_float4(<4 x float> noundef %a) #0 {
+; CHECK-LABEL: define noundef <4 x float> @degrees_float4(
+; CHECK-SAME: <4 x float> noundef [[A:%.*]]) {
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[A0:%.*]] = extractelement <4 x float> [[A]], i64 0
+; CHECK-NEXT:    [[DEGREES_A0:%.*]] = fmul float [[A0]], 0x404CA5DC20000000
+; CHECK-NEXT:    [[A1:%.*]] = extractelement <4 x float> [[A]], i64 1
+; CHECK-NEXT:    [[DEGREES_A1:%.*]] = fmul float [[A1]], 0x404CA5DC20000000
+; CHECK-NEXT:    [[A2:%.*]] = extractelement <4 x float> [[A]], i64 2
+; CHECK-NEXT:    [[DEGREES_A2:%.*]] = fmul float [[A2]], 0x404CA5DC20000000
+; CHECK-NEXT:    [[A3:%.*]] = extractelement <4 x float> [[A]], i64 3
+; CHECK-NEXT:    [[DEGREES_A3:%.*]] = fmul float [[A3]], 0x404CA5DC20000000
+; CHECK-NEXT:    [[INSERT_0:%.*]] = insertelement <4 x float> poison, float [[DEGREES_A0]], i64 0
+; CHECK-NEXT:    [[INSERT_1:%.*]] = insertelement <4 x float> [[INSERT_0]], float [[DEGREES_A1]], i64 1
+; CHECK-NEXT:    [[INSERT_2:%.*]] = insertelement <4 x float> [[INSERT_1]], float [[DEGREES_A2]], i64 2
+; CHECK-NEXT:    [[RES:%.*]] = insertelement <4 x float> [[INSERT_2]], float [[DEGREES_A3]], i64 3
+; CHECK-NEXT:    ret <4 x float> [[RES]]
+;
+entry:
+  %2 = call <4 x float> @llvm.dx.degrees.v4f32(<4 x float> %a)
+  ret <4 x float> %2
+}
+
+declare half  @llvm.dx.degrees.f16(half)
+declare float @llvm.dx.degrees.f32(float)
+declare <4 x float> @llvm.dx.degrees.v4f32(<4 x float>)
diff --git a/llvm/test/CodeGen/SPIRV/hlsl-intrinsics/degrees.ll b/llvm/test/CodeGen/SPIRV/hlsl-intrinsics/degrees.ll
new file mode 100644
index 000000000000..533bcca6f621
--- /dev/null
+++ b/llvm/test/CodeGen/SPIRV/hlsl-intrinsics/degrees.ll
@@ -0,0 +1,52 @@
+; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv-unknown-unknown %s -o - | FileCheck %s
+; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv-unknown-unknown %s -o - -filetype=obj | spirv-val %}
+
+; CHECK-DAG: %[[#op_ext_glsl:]] = OpExtInstImport "GLSL.std.450"
+
+; CHECK-DAG: %[[#float_32:]] = OpTypeFloat 32
+; CHECK-DAG: %[[#float_16:]] = OpTypeFloat 16
+
+; CHECK-DAG: %[[#vec4_float_32:]] = OpTypeVector %[[#float_32]] 4
+; CHECK-DAG: %[[#vec4_float_16:]] = OpTypeVector %[[#float_16]] 4
+
+; CHECK-LABEL: Begin function degrees_float
+define noundef float @degrees_float(float noundef %a) {
+entry:
+; CHECK: %[[#float_32_arg:]] = OpFunctionParameter %[[#float_32]]
+; CHECK: %[[#]] = OpExtInst %[[#float_32]] %[[#op_ext_glsl]] Degrees %[[#float_32_arg]]
+  %elt.degrees = call float @llvm.spv.degrees.f32(float %a)
+  ret float %elt.degrees
+}
+
+; CHECK-LABEL: Begin function degrees_half
+define noundef half @degrees_half(half noundef %a) {
+entry:
+; CHECK: %[[#float_16_arg:]] = OpFunctionParameter %[[#float_16]]
+; CHECK: %[[#]] = OpExtInst %[[#float_16]] %[[#op_ext_glsl]] Degrees %[[#float_16_arg]]
+  %elt.degrees = call half @llvm.spv.degrees.f16(half %a)
+  ret half %elt.degrees
+}
+
+; CHECK-LABEL: Begin function degrees_float_vector
+define noundef <4 x float> @degrees_float_vector(<4 x float> noundef %a) {
+entry:
+; CHECK: %[[#vec4_float_32_arg:]] = OpFunctionParameter %[[#vec4_float_32]]
+; CHECK: %[[#]] = OpExtInst %[[#vec4_float_32]] %[[#op_ext_glsl]] Degrees %[[#vec4_float_32_arg]]
+  %elt.degrees = call <4 x float> @llvm.spv.degrees.v4f32(<4 x float> %a)
+  ret <4 x float> %elt.degrees
+}
+
+; CHECK-LABEL: Begin function degrees_half_vector
+define noundef <4 x half> @degrees_half_vector(<4 x half> noundef %a) {
+entry:
+; CHECK: %[[#vec4_float_16_arg:]] = OpFunctionParameter %[[#vec4_float_16]]
+; CHECK: %[[#]] = OpExtInst %[[#vec4_float_16]] %[[#op_ext_glsl]] Degrees %[[#vec4_float_16_arg]]
+  %elt.degrees = call <4 x half> @llvm.spv.degrees.v4f16(<4 x half> %a)
+  ret <4 x half> %elt.degrees
+}
+
+declare half @llvm.spv.degrees.f16(half)
+declare float @llvm.spv.degrees.f32(float)
+
+declare <4 x float> @llvm.spv.degrees.v4f32(<4 x float>)
+declare <4 x half> @llvm.spv.degrees.v4f16(<4 x half>)
diff --git a/llvm/test/CodeGen/SPIRV/opencl/degrees.ll b/llvm/test/CodeGen/SPIRV/opencl/degrees.ll
new file mode 100644
index 000000000000..88f97835fe71
--- /dev/null
+++ b/llvm/test/CodeGen/SPIRV/opencl/degrees.ll
@@ -0,0 +1,50 @@
+; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv64-unknown-unknown %s -o - | FileCheck %s
+; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv32-unknown-unknown %s -o - | FileCheck %s
+; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv64-unknown-unknown %s -o - -filetype=obj | spirv-val %}
+; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv32-unknown-unknown %s -o - -filetype=obj | spirv-val %}
+
+; CHECK-DAG: %[[#op_ext_glsl:]] = OpExtInstImport "OpenCL.std"
+
+; CHECK-DAG: %[[#float_32:]] = OpTypeFloat 32
+; CHECK-DAG: %[[#float_16:]] = OpTypeFloat 16
+
+; CHECK-DAG: %[[#vec4_float_32:]] = OpTypeVector %[[#float_32]] 4
+; CHECK-DAG: %[[#vec4_float_16:]] = OpTypeVector %[[#float_16]] 4
+
+declare half @llvm.spv.degrees.f16(half)
+declare float @llvm.spv.degrees.f32(float)
+
+declare <4 x float> @llvm.spv.degrees.v4f32(<4 x float>)
+declare <4 x half> @llvm.spv.degrees.v4f16(<4 x half>)
+
+define noundef float @degrees_float(float noundef %a) {
+entry:
+; CHECK: %[[#float_32_arg:]] = OpFunctionParameter %[[#float_32]]
+; CHECK: %[[#]] = OpExtInst %[[#float_32]] %[[#op_ext_glsl]] degrees %[[#float_32_arg]]
+  %elt.degrees = call float @llvm.spv.degrees.f32(float %a)
+  ret float %elt.degrees
+}
+
+define noundef half @degrees_half(half noundef %a) {
+entry:
+; CHECK: %[[#float_16_arg:]] = OpFunctionParameter %[[#float_16]]
+; CHECK: %[[#]] = OpExtInst %[[#float_16]] %[[#op_ext_glsl]] degrees %[[#float_16_arg]]
+  %elt.degrees = call half @llvm.spv.degrees.f16(half %a)
+  ret half %elt.degrees
+}
+
+define noundef <4 x float> @degrees_float_vector(<4 x float> noundef %a) {
+entry:
+; CHECK: %[[#vec4_float_32_arg:]] = OpFunctionParameter %[[#vec4_float_32]]
+; CHECK: %[[#]] = OpExtInst %[[#vec4_float_32]] %[[#op_ext_glsl]] degrees %[[#vec4_float_32_arg]]
+  %elt.degrees = call <4 x float> @llvm.spv.degrees.v4f32(<4 x float> %a)
+  ret <4 x float> %elt.degrees
+}
+
+define noundef <4 x half> @degrees_half_vector(<4 x half> noundef %a) {
+entry:
+; CHECK: %[[#vec4_float_16_arg:]] = OpFunctionParameter %[[#vec4_float_16]]
+; CHECK: %[[#]] = OpExtInst %[[#vec4_float_16]] %[[#op_ext_glsl]] degrees %[[#vec4_float_16_arg]]
+  %elt.degrees = call <4 x half> @llvm.spv.degrees.v4f16(<4 x half> %a)
+  ret <4 x half> %elt.degrees
+}
diff --git a/llvm/tools/llvm-jitlink/llvm-jitlink.cpp b/llvm/tools/llvm-jitlink/llvm-jitlink.cpp
index a2c05deefa6b..108cadd2e016 100644
--- a/llvm/tools/llvm-jitlink/llvm-jitlink.cpp
+++ b/llvm/tools/llvm-jitlink/llvm-jitlink.cpp
@@ -1041,8 +1041,8 @@ Session::Session(std::unique_ptr<ExecutorProcessControl> EPC, Error &Err)
         return;
       }
     } else if (TT.isOSBinFormatELF()) {
-      if (auto P = ELFNixPlatform::Create(ES, ObjLayer, *PlatformJD,
-                                          OrcRuntime.c_str()))
+      if (auto P =
+              ELFNixPlatform::Create(ObjLayer, *PlatformJD, OrcRuntime.c_str()))
         ES.setPlatform(std::move(*P));
       else {
         Err = P.takeError();
diff --git a/utils/bazel/llvm-project-overlay/clang-tools-extra/include-cleaner/BUILD.bazel b/utils/bazel/llvm-project-overlay/clang-tools-extra/include-cleaner/BUILD.bazel
index 28f90efb3ba7..5b210ad80c87 100644
--- a/utils/bazel/llvm-project-overlay/clang-tools-extra/include-cleaner/BUILD.bazel
+++ b/utils/bazel/llvm-project-overlay/clang-tools-extra/include-cleaner/BUILD.bazel
@@ -2,7 +2,9 @@
 # See https://llvm.org/LICENSE.txt for license information.
 # SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 
+load("@bazel_skylib//rules:expand_template.bzl", "expand_template")
 load("@rules_cc//cc:defs.bzl", "cc_binary", "cc_library")
+load("//llvm:lit_test.bzl", "lit_test", "package_path")
 
 package(
     default_visibility = ["//visibility:public"],
@@ -61,3 +63,71 @@ cc_binary(
         "//llvm:Support",
     ],
 )
+
+cc_test(
+    name = "unittests",
+    srcs = glob(["unittests/*.cpp"]),
+    deps = [
+        ":include_cleaner",
+        ":include_cleaner_internal",
+        "//clang:ast",
+        "//clang:basic",
+        "//clang:format",
+        "//clang:frontend",
+        "//clang:lex",
+        "//clang:serialization",
+        "//clang:testing",
+        "//clang:tooling",
+        "//clang:tooling_inclusions",
+        "//llvm:Support",
+        "//llvm:TestingAnnotations",
+        "//third-party/unittest:gmock",
+        "//third-party/unittest:gtest",
+    ],
+)
+
+LLVM_LIT_PATH_FUNCTION = " " + \
+                         "# Allow generated file to be relocatable.\n" + \
+                         "from pathlib import Path\n" + \
+                         "def path(p):\n" + \
+                         "    p = Path(p)\n" + \
+                         "    if p.exists: return str(p.resolve())\n" + \
+                         "    if not p: return ''\n" + \
+                         "    return str((Path(__file__).parent / p).resolve())\n"
+
+LIT_SITE_CFG_IN_HEADER = "# Autogenerated, do not edit." + LLVM_LIT_PATH_FUNCTION
+
+expand_template(
+    name = "lit_site_cfg_py",
+    testonly = True,
+    out = "test/lit.site.cfg.py",
+    substitutions = {
+        "@CMAKE_CURRENT_BINARY_DIR@": package_path("//clang-tools-extra/include-cleaner:BUILD") + "/test",
+        "@CMAKE_CURRENT_SOURCE_DIR@": package_path("//clang-tools-extra/include-cleaner:BUILD") + "/test",
+        "@CURRENT_TOOLS_DIR@": package_path("//clang-tools-extra/include-cleaner:BUILD"),
+        "@LIT_SITE_CFG_IN_HEADER@": LIT_SITE_CFG_IN_HEADER,
+        "@LLVM_LIBS_DIR@": package_path("//llvm:BUILD"),
+        "@LLVM_LIT_TOOLS_DIR@": package_path("//llvm:BUILD"),
+        "@LLVM_TOOLS_DIR@": package_path("//llvm:BUILD"),
+        "@TARGET_TRIPLE@": "",
+        '"@Python3_EXECUTABLE@"': "sys.executable",
+    },
+    template = "test/lit.site.cfg.py.in",
+)
+
+[
+    lit_test(
+        name = "%s.test" % src,
+        srcs = [src],
+        data = glob(["test/Inputs/**/*"]) + [
+            "test/lit.cfg.py",
+            "test/lit.site.cfg.py",
+            ":clang-include-cleaner",
+            "//llvm:FileCheck",
+            "//llvm:count",
+            "//llvm:not",
+        ],
+        args = ["-svv"],
+    )
+    for src in glob(["test/*.cpp"])
+]
diff --git a/utils/bazel/llvm-project-overlay/lldb/source/Plugins/BUILD.bazel b/utils/bazel/llvm-project-overlay/lldb/source/Plugins/BUILD.bazel
index 96202bf47b84..38493411adde 100644
--- a/utils/bazel/llvm-project-overlay/lldb/source/Plugins/BUILD.bazel
+++ b/utils/bazel/llvm-project-overlay/lldb/source/Plugins/BUILD.bazel
@@ -2248,6 +2248,8 @@ cc_library(
     hdrs = glob(["Process/minidump/*.h"]),
     include_prefix = "Plugins",
     deps = [
+        ":PluginDynamicLoaderPosixDYLD",
+        ":PluginDynamicLoaderPosixDYLDHeaders",
         ":PluginObjectFilePlaceholder",
         ":PluginProcessElfCore",
         ":PluginProcessUtility",