Vendor import of llvm trunk r242221:vendor/llvm/llvm-trunk-r242221

https://llvm.org/svn/llvm-project/llvm/trunk@242221

860 files changed, 32133 insertions, 11025 deletions

diff --git a/.gitignore b/.gitignore
index f3424d231e2d..e3d191dcbcce 100644
--- a/.gitignore
+++ b/.gitignore
@@ -43,7 +43,9 @@ autoconf/autom4te.cache
 # Directories to ignore (do not add trailing '/'s, they skip symlinks).
 #==============================================================================#
 # External projects that are tracked independently.
-projects/*/
+projects/*
+!projects/*.*
+!projects/Makefile
 # Clang, which is tracked independently.
 tools/clang
 # LLDB, which is tracked independently.
diff --git a/autoconf/config.guess b/autoconf/config.guess
index 73d671bc2c6b..b209a1a06c6f 100755
--- a/autoconf/config.guess
+++ b/autoconf/config.guess
@@ -810,6 +810,9 @@ EOF
     *:MINGW*:*)
 	echo ${UNAME_MACHINE}-pc-mingw32
 	exit ;;
+    *:MSYS*:*)
+	echo ${UNAME_MACHINE}-pc-msys
+	exit ;;
     i*:windows32*:*)
 	# uname -m includes "-pc" on this system.
 	echo ${UNAME_MACHINE}-mingw32
diff --git a/bindings/python/llvm/object.py b/bindings/python/llvm/object.py
index 4e912ed5da9d..b427113e9cea 100644
--- a/bindings/python/llvm/object.py
+++ b/bindings/python/llvm/object.py
@@ -372,14 +372,6 @@ class Relocation(LLVMObject):
         self.expired = False
 
     @CachedProperty
-    def address(self):
-        """The address of this relocation, in long bytes."""
-        if self.expired:
-            raise Exception('Relocation instance has expired.')
-
-        return lib.LLVMGetRelocationAddress(self)
-
-    @CachedProperty
     def offset(self):
         """The offset of this relocation, in long bytes."""
         if self.expired:
@@ -498,9 +490,6 @@ def register_library(library):
     library.LLVMGetSymbolSize.argtypes = [Symbol]
     library.LLVMGetSymbolSize.restype = c_uint64
 
-    library.LLVMGetRelocationAddress.argtypes = [c_object_p]
-    library.LLVMGetRelocationAddress.restype = c_uint64
-
     library.LLVMGetRelocationOffset.argtypes = [c_object_p]
     library.LLVMGetRelocationOffset.restype = c_uint64
 
diff --git a/cmake/modules/AddLLVM.cmake b/cmake/modules/AddLLVM.cmake
index 4f60d9e6e7d6..45f6746948d2 100755
--- a/cmake/modules/AddLLVM.cmake
+++ b/cmake/modules/AddLLVM.cmake
@@ -93,20 +93,9 @@ function(add_llvm_symbol_exports target_name export_file)
   else()
     set(native_export_file "${target_name}.def")
 
-    set(CAT "cat")
-    set(export_file_nativeslashes ${export_file})
-    if(WIN32 AND NOT CYGWIN AND NOT MSYS)
-      set(CAT "type")
-      # Convert ${export_file} to native format (backslashes) for "type"
-      # Does not use file(TO_NATIVE_PATH) as it doesn't create a native
-      # path but a build-system specific format (see CMake bug
-      # http://public.kitware.com/Bug/print_bug_page.php?bug_id=5939 )
-      string(REPLACE / \\ export_file_nativeslashes ${export_file})
-    endif()
-
     add_custom_command(OUTPUT ${native_export_file}
-      COMMAND ${CMAKE_COMMAND} -E echo "EXPORTS" > ${native_export_file}
-      COMMAND ${CAT} ${export_file_nativeslashes} >> ${native_export_file}
+      COMMAND ${PYTHON_EXECUTABLE} -c "import sys;print(''.join(['EXPORTS\\n']+sys.stdin.readlines(),))"
+        < ${export_file} > ${native_export_file}
       DEPENDS ${export_file}
       VERBATIM
       COMMENT "Creating export file for ${target_name}")
@@ -700,10 +689,18 @@ macro(add_llvm_external_project name)
   list(APPEND LLVM_IMPLICIT_PROJECT_IGNORE "${CMAKE_CURRENT_SOURCE_DIR}/${add_llvm_external_dir}")
   string(REPLACE "-" "_" nameUNDERSCORE ${name})
   string(TOUPPER ${nameUNDERSCORE} nameUPPER)
-  set(LLVM_EXTERNAL_${nameUPPER}_SOURCE_DIR "${CMAKE_CURRENT_SOURCE_DIR}/${add_llvm_external_dir}"
-      CACHE PATH "Path to ${name} source directory")
-  if (NOT ${LLVM_EXTERNAL_${nameUPPER}_SOURCE_DIR} STREQUAL ""
-      AND EXISTS ${LLVM_EXTERNAL_${nameUPPER}_SOURCE_DIR}/CMakeLists.txt)
+  #TODO: Remove this check in a few days once it has circulated through
+  # buildbots and people's checkouts (cbieneman - July 14, 2015)
+  if("${LLVM_EXTERNAL_${nameUPPER}_SOURCE_DIR}" STREQUAL "${CMAKE_CURRENT_SOURCE_DIR}/${add_llvm_external_dir}")
+    unset(LLVM_EXTERNAL_${nameUPPER}_SOURCE_DIR CACHE)
+  endif()
+  if(NOT LLVM_EXTERNAL_${nameUPPER}_SOURCE_DIR)
+    set(LLVM_EXTERNAL_${nameUPPER}_SOURCE_DIR "${CMAKE_CURRENT_SOURCE_DIR}/${add_llvm_external_dir}")
+  else()
+    set(LLVM_EXTERNAL_${nameUPPER}_SOURCE_DIR
+        CACHE PATH "Path to ${name} source directory")
+  endif()
+  if (EXISTS ${LLVM_EXTERNAL_${nameUPPER}_SOURCE_DIR}/CMakeLists.txt)
     option(LLVM_EXTERNAL_${nameUPPER}_BUILD
            "Whether to build ${name} as part of LLVM" ON)
     if (LLVM_EXTERNAL_${nameUPPER}_BUILD)
diff --git a/docs/ExceptionHandling.rst b/docs/ExceptionHandling.rst
index 72ed78a3c99a..55ffdb45efe9 100644
--- a/docs/ExceptionHandling.rst
+++ b/docs/ExceptionHandling.rst
@@ -339,11 +339,11 @@ original context before code generation.
 
 Catch handlers are called with a pointer to the handler itself as the first
 argument and a pointer to the parent function's stack frame as the second
-argument.  The catch handler uses the `llvm.recoverframe
-<LangRef.html#llvm-frameallocate-and-llvm-framerecover-intrinsics>`_ to get a
+argument.  The catch handler uses the `llvm.localrecover
+<LangRef.html#llvm-localescape-and-llvm-localrecover-intrinsics>`_ to get a
 pointer to a frame allocation block that is created in the parent frame using
-the `llvm.allocateframe 
-<LangRef.html#llvm-frameallocate-and-llvm-framerecover-intrinsics>`_ intrinsic.
+the `llvm.localescape
+<LangRef.html#llvm-localescape-and-llvm-localrecover-intrinsics>`_ intrinsic.
 The ``WinEHPrepare`` pass will have created a structure definition for the
 contents of this block.  The first two members of the structure will always be
 (1) a 32-bit integer that the runtime uses to track the exception state of the
@@ -520,12 +520,12 @@ action.
 A code of ``i32 1`` indicates a catch action, which expects three additional
 arguments. Different EH schemes give different meanings to the three arguments,
 but the first argument indicates whether the catch should fire, the second is
-the frameescape index of the exception object, and the third is the code to run
+the localescape index of the exception object, and the third is the code to run
 to catch the exception.
 
 For Windows C++ exception handling, the first argument for a catch handler is a
 pointer to the RTTI type descriptor for the object to catch. The second
-argument is an index into the argument list of the ``llvm.frameescape`` call in
+argument is an index into the argument list of the ``llvm.localescape`` call in
 the main function. The exception object will be copied into the provided stack
 object. If the exception object is not required, this argument should be -1.
 The third argument is a pointer to a function implementing the catch.  This
diff --git a/docs/LangRef.rst b/docs/LangRef.rst
index 167280f75f7f..e7d6f67c9399 100644
--- a/docs/LangRef.rst
+++ b/docs/LangRef.rst
@@ -1326,6 +1326,14 @@ example:
     On an argument, this attribute indicates that the function does not write
     through this pointer argument, even though it may write to the memory that
     the pointer points to.
+``argmemonly``
+    This attribute indicates that the only memory accesses inside function are
+    loads and stores from objects pointed to by its pointer-typed arguments,
+    with arbitrary offsets. Or in other words, all memory operations in the
+    function can refer to memory only using pointers based on its function
+    arguments.
+    Note that ``argmemonly`` can be used together with ``readonly`` attribute
+    in order to specify that function reads only from its arguments.
 ``returns_twice``
     This attribute indicates that this function can return twice. The C
     ``setjmp`` is an example of such a function. The compiler disables
@@ -1446,8 +1454,8 @@ The strings can contain any character by escaping non-printable
 characters. The escape sequence used is simply "\\xx" where "xx" is the
 two digit hex code for the number.
 
-The inline asm code is simply printed to the machine code .s file when
-assembly code is generated.
+Note that the assembly string *must* be parseable by LLVM's integrated assembler
+(unless it is disabled), even when emitting a ``.s`` file.
 
 .. _langref_datalayout:
 
@@ -1837,8 +1845,8 @@ Fast-Math Flags
 
 LLVM IR floating-point binary ops (:ref:`fadd <i_fadd>`,
 :ref:`fsub <i_fsub>`, :ref:`fmul <i_fmul>`, :ref:`fdiv <i_fdiv>`,
-:ref:`frem <i_frem>`) have the following flags that can be set to enable
-otherwise unsafe floating point operations
+:ref:`frem <i_frem>`, :ref:`fcmp <i_fcmp>`) have the following flags that can
+be set to enable otherwise unsafe floating point operations
 
 ``nnan``
    No NaNs - Allow optimizations to assume the arguments and result are not
@@ -2800,13 +2808,36 @@ Inline Assembler Expressions
 ----------------------------
 
 LLVM supports inline assembler expressions (as opposed to :ref:`Module-Level
-Inline Assembly <moduleasm>`) through the use of a special value. This
-value represents the inline assembler as a string (containing the
-instructions to emit), a list of operand constraints (stored as a
-string), a flag that indicates whether or not the inline asm expression
-has side effects, and a flag indicating whether the function containing
-the asm needs to align its stack conservatively. An example inline
-assembler expression is:
+Inline Assembly <moduleasm>`) through the use of a special value. This value
+represents the inline assembler as a template string (containing the
+instructions to emit), a list of operand constraints (stored as a string), a
+flag that indicates whether or not the inline asm expression has side effects,
+and a flag indicating whether the function containing the asm needs to align its
+stack conservatively.
+
+The template string supports argument substitution of the operands using "``$``"
+followed by a number, to indicate substitution of the given register/memory
+location, as specified by the constraint string. "``${NUM:MODIFIER}``" may also
+be used, where ``MODIFIER`` is a target-specific annotation for how to print the
+operand (See :ref:`inline-asm-modifiers`).
+
+A literal "``$``" may be included by using "``$$``" in the template. To include
+other special characters into the output, the usual "``\XX``" escapes may be
+used, just as in other strings. Note that after template substitution, the
+resulting assembly string is parsed by LLVM's integrated assembler unless it is
+disabled -- even when emitting a ``.s`` file -- and thus must contain assembly
+syntax known to LLVM.
+
+LLVM's support for inline asm is modeled closely on the requirements of Clang's
+GCC-compatible inline-asm support. Thus, the feature-set and the constraint and
+modifier codes listed here are similar or identical to those in GCC's inline asm
+support. However, to be clear, the syntax of the template and constraint strings
+described here is *not* the same as the syntax accepted by GCC and Clang, and,
+while most constraint letters are passed through as-is by Clang, some get
+translated to other codes when converting from the C source to the LLVM
+assembly.
+
+An example inline assembler expression is:
 
 .. code-block:: llvm
 
@@ -2852,6 +2883,596 @@ If multiple keywords appear the '``sideeffect``' keyword must come
 first, the '``alignstack``' keyword second and the '``inteldialect``'
 keyword last.
 
+Inline Asm Constraint String
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+The constraint list is a comma-separated string, each element containing one or
+more constraint codes.
+
+For each element in the constraint list an appropriate register or memory
+operand will be chosen, and it will be made available to assembly template
+string expansion as ``$0`` for the first constraint in the list, ``$1`` for the
+second, etc.
+
+There are three different types of constraints, which are distinguished by a
+prefix symbol in front of the constraint code: Output, Input, and Clobber. The
+constraints must always be given in that order: outputs first, then inputs, then
+clobbers. They cannot be intermingled.
+
+There are also three different categories of constraint codes:
+
+- Register constraint. This is either a register class, or a fixed physical
+  register. This kind of constraint will allocate a register, and if necessary,
+  bitcast the argument or result to the appropriate type.
+- Memory constraint. This kind of constraint is for use with an instruction
+  taking a memory operand. Different constraints allow for different addressing
+  modes used by the target.
+- Immediate value constraint. This kind of constraint is for an integer or other
+  immediate value which can be rendered directly into an instruction. The
+  various target-specific constraints allow the selection of a value in the
+  proper range for the instruction you wish to use it with.
+
+Output constraints
+""""""""""""""""""
+
+Output constraints are specified by an "``=``" prefix (e.g. "``=r``"). This
+indicates that the assembly will write to this operand, and the operand will
+then be made available as a return value of the ``asm`` expression. Output
+constraints do not consume an argument from the call instruction. (Except, see
+below about indirect outputs).
+
+Normally, it is expected that no output locations are written to by the assembly
+expression until *all* of the inputs have been read. As such, LLVM may assign
+the same register to an output and an input. If this is not safe (e.g. if the
+assembly contains two instructions, where the first writes to one output, and
+the second reads an input and writes to a second output), then the "``&``"
+modifier must be used (e.g. "``=&r``") to specify that the output is an
+"early-clobber" output. Marking an ouput as "early-clobber" ensures that LLVM
+will not use the same register for any inputs (other than an input tied to this
+output).
+
+Input constraints
+"""""""""""""""""
+
+Input constraints do not have a prefix -- just the constraint codes. Each input
+constraint will consume one argument from the call instruction. It is not
+permitted for the asm to write to any input register or memory location (unless
+that input is tied to an output). Note also that multiple inputs may all be
+assigned to the same register, if LLVM can determine that they necessarily all
+contain the same value.
+
+Instead of providing a Constraint Code, input constraints may also "tie"
+themselves to an output constraint, by providing an integer as the constraint
+string. Tied inputs still consume an argument from the call instruction, and
+take up a position in the asm template numbering as is usual -- they will simply
+be constrained to always use the same register as the output they've been tied
+to. For example, a constraint string of "``=r,0``" says to assign a register for
+output, and use that register as an input as well (it being the 0'th
+constraint).
+
+It is permitted to tie an input to an "early-clobber" output. In that case, no
+*other* input may share the same register as the input tied to the early-clobber
+(even when the other input has the same value).
+
+You may only tie an input to an output which has a register constraint, not a
+memory constraint. Only a single input may be tied to an output.
+
+There is also an "interesting" feature which deserves a bit of explanation: if a
+register class constraint allocates a register which is too small for the value
+type operand provided as input, the input value will be split into multiple
+registers, and all of them passed to the inline asm.
+
+However, this feature is often not as useful as you might think.
+
+Firstly, the registers are *not* guaranteed to be consecutive. So, on those
+architectures that have instructions which operate on multiple consecutive
+instructions, this is not an appropriate way to support them. (e.g. the 32-bit
+SparcV8 has a 64-bit load, which instruction takes a single 32-bit register. The
+hardware then loads into both the named register, and the next register. This
+feature of inline asm would not be useful to support that.)
+
+A few of the targets provide a template string modifier allowing explicit access
+to the second register of a two-register operand (e.g. MIPS ``L``, ``M``, and
+``D``). On such an architecture, you can actually access the second allocated
+register (yet, still, not any subsequent ones). But, in that case, you're still
+probably better off simply splitting the value into two separate operands, for
+clarity. (e.g. see the description of the ``A`` constraint on X86, which,
+despite existing only for use with this feature, is not really a good idea to
+use)
+
+Indirect inputs and outputs
+"""""""""""""""""""""""""""
+
+Indirect output or input constraints can be specified by the "``*``" modifier
+(which goes after the "``=``" in case of an output). This indicates that the asm
+will write to or read from the contents of an *address* provided as an input
+argument. (Note that in this way, indirect outputs act more like an *input* than
+an output: just like an input, they consume an argument of the call expression,
+rather than producing a return value. An indirect output constraint is an
+"output" only in that the asm is expected to write to the contents of the input
+memory location, instead of just read from it).
+
+This is most typically used for memory constraint, e.g. "``=*m``", to pass the
+address of a variable as a value.
+
+It is also possible to use an indirect *register* constraint, but only on output
+(e.g. "``=*r``"). This will cause LLVM to allocate a register for an output
+value normally, and then, separately emit a store to the address provided as
+input, after the provided inline asm. (It's not clear what value this
+functionality provides, compared to writing the store explicitly after the asm
+statement, and it can only produce worse code, since it bypasses many
+optimization passes. I would recommend not using it.)
+
+
+Clobber constraints
+"""""""""""""""""""
+
+A clobber constraint is indicated by a "``~``" prefix. A clobber does not
+consume an input operand, nor generate an output. Clobbers cannot use any of the
+general constraint code letters -- they may use only explicit register
+constraints, e.g. "``~{eax}``". The one exception is that a clobber string of
+"``~{memory}``" indicates that the assembly writes to arbitrary undeclared
+memory locations -- not only the memory pointed to by a declared indirect
+output.
+
+
+Constraint Codes
+""""""""""""""""
+After a potential prefix comes constraint code, or codes.
+
+A Constraint Code is either a single letter (e.g. "``r``"), a "``^``" character
+followed by two letters (e.g. "``^wc``"), or "``{``" register-name "``}``"
+(e.g. "``{eax}``").
+
+The one and two letter constraint codes are typically chosen to be the same as
+GCC's constraint codes.
+
+A single constraint may include one or more than constraint code in it, leaving
+it up to LLVM to choose which one to use. This is included mainly for
+compatibility with the translation of GCC inline asm coming from clang.
+
+There are two ways to specify alternatives, and either or both may be used in an
+inline asm constraint list:
+
+1) Append the codes to each other, making a constraint code set. E.g. "``im``"
+   or "``{eax}m``". This means "choose any of the options in the set". The
+   choice of constraint is made independently for each constraint in the
+   constraint list.
+
+2) Use "``|``" between constraint code sets, creating alternatives. Every
+   constraint in the constraint list must have the same number of alternative
+   sets. With this syntax, the same alternative in *all* of the items in the
+   constraint list will be chosen together.
+
+Putting those together, you might have a two operand constraint string like
+``"rm|r,ri|rm"``. This indicates that if operand 0 is ``r`` or ``m``, then
+operand 1 may be one of ``r`` or ``i``. If operand 0 is ``r``, then operand 1
+may be one of ``r`` or ``m``. But, operand 0 and 1 cannot both be of type m.
+
+However, the use of either of the alternatives features is *NOT* recommended, as
+LLVM is not able to make an intelligent choice about which one to use. (At the
+point it currently needs to choose, not enough information is available to do so
+in a smart way.) Thus, it simply tries to make a choice that's most likely to
+compile, not one that will be optimal performance. (e.g., given "``rm``", it'll
+always choose to use memory, not registers). And, if given multiple registers,
+or multiple register classes, it will simply choose the first one. (In fact, it
+doesn't currently even ensure explicitly specified physical registers are
+unique, so specifying multiple physical registers as alternatives, like
+``{r11}{r12},{r11}{r12}``, will assign r11 to both operands, not at all what was
+intended.)
+
+Supported Constraint Code List
+""""""""""""""""""""""""""""""
+
+The constraint codes are, in general, expected to behave the same way they do in
+GCC. LLVM's support is often implemented on an 'as-needed' basis, to support C
+inline asm code which was supported by GCC. A mismatch in behavior between LLVM
+and GCC likely indicates a bug in LLVM.
+
+Some constraint codes are typically supported by all targets:
+
+- ``r``: A register in the target's general purpose register class.
+- ``m``: A memory address operand. It is target-specific what addressing modes
+  are supported, typical examples are register, or register + register offset,
+  or register + immediate offset (of some target-specific size).
+- ``i``: An integer constant (of target-specific width). Allows either a simple
+  immediate, or a relocatable value.
+- ``n``: An integer constant -- *not* including relocatable values.
+- ``s``: An integer constant, but allowing *only* relocatable values.
+- ``X``: Allows an operand of any kind, no constraint whatsoever. Typically
+  useful to pass a label for an asm branch or call.
+
+  .. FIXME: but that surely isn't actually okay to jump out of an asm
+     block without telling llvm about the control transfer???)
+
+- ``{register-name}``: Requires exactly the named physical register.
+
+Other constraints are target-specific:
+
+AArch64:
+
+- ``z``: An immediate integer 0. Outputs ``WZR`` or ``XZR``, as appropriate.
+- ``I``: An immediate integer valid for an ``ADD`` or ``SUB`` instruction,
+  i.e. 0 to 4095 with optional shift by 12.
+- ``J``: An immediate integer that, when negated, is valid for an ``ADD`` or
+  ``SUB`` instruction, i.e. -1 to -4095 with optional left shift by 12.
+- ``K``: An immediate integer that is valid for the 'bitmask immediate 32' of a
+  logical instruction like ``AND``, ``EOR``, or ``ORR`` with a 32-bit register.
+- ``L``: An immediate integer that is valid for the 'bitmask immediate 64' of a
+  logical instruction like ``AND``, ``EOR``, or ``ORR`` with a 64-bit register.
+- ``M``: An immediate integer for use with the ``MOV`` assembly alias on a
+  32-bit register. This is a superset of ``K``: in addition to the bitmask
+  immediate, also allows immediate integers which can be loaded with a single
+  ``MOVZ`` or ``MOVL`` instruction.
+- ``N``: An immediate integer for use with the ``MOV`` assembly alias on a
+  64-bit register. This is a superset of ``L``.
+- ``Q``: Memory address operand must be in a single register (no
+  offsets). (However, LLVM currently does this for the ``m`` constraint as
+  well.)
+- ``r``: A 32 or 64-bit integer register (W* or X*).
+- ``w``: A 32, 64, or 128-bit floating-point/SIMD register.
+- ``x``: A lower 128-bit floating-point/SIMD register (``V0`` to ``V15``).
+
+AMDGPU:
+
+- ``r``: A 32 or 64-bit integer register.
+- ``[0-9]v``: The 32-bit VGPR register, number 0-9.
+- ``[0-9]s``: The 32-bit SGPR register, number 0-9.
+
+
+All ARM modes:
+
+- ``Q``, ``Um``, ``Un``, ``Uq``, ``Us``, ``Ut``, ``Uv``, ``Uy``: Memory address
+  operand. Treated the same as operand ``m``, at the moment.
+
+ARM and ARM's Thumb2 mode:
+
+- ``j``: An immediate integer between 0 and 65535 (valid for ``MOVW``)
+- ``I``: An immediate integer valid for a data-processing instruction.
+- ``J``: An immediate integer between -4095 and 4095.
+- ``K``: An immediate integer whose bitwise inverse is valid for a
+  data-processing instruction. (Can be used with template modifier "``B``" to
+  print the inverted value).
+- ``L``: An immediate integer whose negation is valid for a data-processing
+  instruction. (Can be used with template modifier "``n``" to print the negated
+  value).
+- ``M``: A power of two or a integer between 0 and 32.
+- ``N``: Invalid immediate constraint.
+- ``O``: Invalid immediate constraint.
+- ``r``: A general-purpose 32-bit integer register (``r0-r15``).
+- ``l``: In Thumb2 mode, low 32-bit GPR registers (``r0-r7``). In ARM mode, same
+  as ``r``.
+- ``h``: In Thumb2 mode, a high 32-bit GPR register (``r8-r15``). In ARM mode,
+  invalid.
+- ``w``: A 32, 64, or 128-bit floating-point/SIMD register: ``s0-s31``,
+  ``d0-d31``, or ``q0-q15``.
+- ``x``: A 32, 64, or 128-bit floating-point/SIMD register: ``s0-s15``,
+  ``d0-d7``, or ``q0-q3``.
+- ``t``: A floating-point/SIMD register, only supports 32-bit values:
+  ``s0-s31``.
+
+ARM's Thumb1 mode:
+
+- ``I``: An immediate integer between 0 and 255.
+- ``J``: An immediate integer between -255 and -1.
+- ``K``: An immediate integer between 0 and 255, with optional left-shift by
+  some amount.
+- ``L``: An immediate integer between -7 and 7.
+- ``M``: An immediate integer which is a multiple of 4 between 0 and 1020.
+- ``N``: An immediate integer between 0 and 31.
+- ``O``: An immediate integer which is a multiple of 4 between -508 and 508.
+- ``r``: A low 32-bit GPR register (``r0-r7``).
+- ``l``: A low 32-bit GPR register (``r0-r7``).
+- ``h``: A high GPR register (``r0-r7``).
+- ``w``: A 32, 64, or 128-bit floating-point/SIMD register: ``s0-s31``,
+  ``d0-d31``, or ``q0-q15``.
+- ``x``: A 32, 64, or 128-bit floating-point/SIMD register: ``s0-s15``,
+  ``d0-d7``, or ``q0-q3``.
+- ``t``: A floating-point/SIMD register, only supports 32-bit values:
+  ``s0-s31``.
+
+
+Hexagon:
+
+- ``o``, ``v``: A memory address operand, treated the same as constraint ``m``,
+  at the moment.
+- ``r``: A 32 or 64-bit register.
+
+MSP430:
+
+- ``r``: An 8 or 16-bit register.
+
+MIPS:
+
+- ``I``: An immediate signed 16-bit integer.
+- ``J``: An immediate integer zero.
+- ``K``: An immediate unsigned 16-bit integer.
+- ``L``: An immediate 32-bit integer, where the lower 16 bits are 0.
+- ``N``: An immediate integer between -65535 and -1.
+- ``O``: An immediate signed 15-bit integer.
+- ``P``: An immediate integer between 1 and 65535.
+- ``m``: A memory address operand. In MIPS-SE mode, allows a base address
+  register plus 16-bit immediate offset. In MIPS mode, just a base register.
+- ``R``: A memory address operand. In MIPS-SE mode, allows a base address
+  register plus a 9-bit signed offset. In MIPS mode, the same as constraint
+  ``m``.
+- ``ZC``: A memory address operand, suitable for use in a ``pref``, ``ll``, or
+  ``sc`` instruction on the given subtarget (details vary).
+- ``r``, ``d``,  ``y``: A 32 or 64-bit GPR register.
+- ``f``: A 32 or 64-bit FPU register (``F0-F31``), or a 128-bit MSA register
+  (``W0-W31``). In the case of MSA registers, it is recommended to use the ``w``
+  argument modifier for compatibility with GCC.
+- ``c``: A 32-bit or 64-bit GPR register suitable for indirect jump (always
+  ``25``).
+- ``l``: The ``lo`` register, 32 or 64-bit.
+- ``x``: Invalid.
+
+NVPTX:
+
+- ``b``: A 1-bit integer register.
+- ``c`` or ``h``: A 16-bit integer register.
+- ``r``: A 32-bit integer register.
+- ``l`` or ``N``: A 64-bit integer register.
+- ``f``: A 32-bit float register.
+- ``d``: A 64-bit float register.
+
+
+PowerPC:
+
+- ``I``: An immediate signed 16-bit integer.
+- ``J``: An immediate unsigned 16-bit integer, shifted left 16 bits.
+- ``K``: An immediate unsigned 16-bit integer.
+- ``L``: An immediate signed 16-bit integer, shifted left 16 bits.
+- ``M``: An immediate integer greater than 31.
+- ``N``: An immediate integer that is an exact power of 2.
+- ``O``: The immediate integer constant 0.
+- ``P``: An immediate integer constant whose negation is a signed 16-bit
+  constant.
+- ``es``, ``o``, ``Q``, ``Z``, ``Zy``: A memory address operand, currently
+  treated the same as ``m``.
+- ``r``: A 32 or 64-bit integer register.
+- ``b``: A 32 or 64-bit integer register, excluding ``R0`` (that is:
+  ``R1-R31``).
+- ``f``: A 32 or 64-bit float register (``F0-F31``), or when QPX is enabled, a
+  128 or 256-bit QPX register (``Q0-Q31``; aliases the ``F`` registers).
+- ``v``: For ``4 x f32`` or ``4 x f64`` types, when QPX is enabled, a
+  128 or 256-bit QPX register (``Q0-Q31``), otherwise a 128-bit
+  altivec vector register (``V0-V31``).
+
+  .. FIXME: is this a bug that v accepts QPX registers? I think this
+     is supposed to only use the altivec vector registers?
+
+- ``y``: Condition register (``CR0-CR7``).
+- ``wc``: An individual CR bit in a CR register.
+- ``wa``, ``wd``, ``wf``: Any 128-bit VSX vector register, from the full VSX
+  register set (overlapping both the floating-point and vector register files).
+- ``ws``: A 32 or 64-bit floating point register, from the full VSX register
+  set.
+
+Sparc:
+
+- ``I``: An immediate 13-bit signed integer.
+- ``r``: A 32-bit integer register.
+
+SystemZ:
+
+- ``I``: An immediate unsigned 8-bit integer.
+- ``J``: An immediate unsigned 12-bit integer.
+- ``K``: An immediate signed 16-bit integer.
+- ``L``: An immediate signed 20-bit integer.
+- ``M``: An immediate integer 0x7fffffff.
+- ``Q``, ``R``, ``S``, ``T``: A memory address operand, treated the same as
+  ``m``, at the moment.
+- ``r`` or ``d``: A 32, 64, or 128-bit integer register.
+- ``a``: A 32, 64, or 128-bit integer address register (excludes R0, which in an
+  address context evaluates as zero).
+- ``h``: A 32-bit value in the high part of a 64bit data register
+  (LLVM-specific)
+- ``f``: A 32, 64, or 128-bit floating point register.
+
+X86:
+
+- ``I``: An immediate integer between 0 and 31.
+- ``J``: An immediate integer between 0 and 64.
+- ``K``: An immediate signed 8-bit integer.
+- ``L``: An immediate integer, 0xff or 0xffff or (in 64-bit mode only)
+  0xffffffff.
+- ``M``: An immediate integer between 0 and 3.
+- ``N``: An immediate unsigned 8-bit integer.
+- ``O``: An immediate integer between 0 and 127.
+- ``e``: An immediate 32-bit signed integer.
+- ``Z``: An immediate 32-bit unsigned integer.
+- ``o``, ``v``: Treated the same as ``m``, at the moment.
+- ``q``: An 8, 16, 32, or 64-bit register which can be accessed as an 8-bit
+  ``l`` integer register. On X86-32, this is the ``a``, ``b``, ``c``, and ``d``
+  registers, and on X86-64, it is all of the integer registers.
+- ``Q``: An 8, 16, 32, or 64-bit register which can be accessed as an 8-bit
+  ``h`` integer register. This is the ``a``, ``b``, ``c``, and ``d`` registers.
+- ``r`` or ``l``: An 8, 16, 32, or 64-bit integer register.
+- ``R``: An 8, 16, 32, or 64-bit "legacy" integer register -- one which has
+  existed since i386, and can be accessed without the REX prefix.
+- ``f``: A 32, 64, or 80-bit '387 FPU stack pseudo-register.
+- ``y``: A 64-bit MMX register, if MMX is enabled.
+- ``x``: If SSE is enabled: a 32 or 64-bit scalar operand, or 128-bit vector
+  operand in a SSE register. If AVX is also enabled, can also be a 256-bit
+  vector operand in an AVX register. If AVX-512 is also enabled, can also be a
+  512-bit vector operand in an AVX512 register, Otherwise, an error.
+- ``Y``: The same as ``x``, if *SSE2* is enabled, otherwise an error.
+- ``A``: Special case: allocates EAX first, then EDX, for a single operand (in
+  32-bit mode, a 64-bit integer operand will get split into two registers). It
+  is not recommended to use this constraint, as in 64-bit mode, the 64-bit
+  operand will get allocated only to RAX -- if two 32-bit operands are needed,
+  you're better off splitting it yourself, before passing it to the asm
+  statement.
+
+XCore:
+
+- ``r``: A 32-bit integer register.
+
+
+.. _inline-asm-modifiers:
+
+Asm template argument modifiers
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+In the asm template string, modifiers can be used on the operand reference, like
+"``${0:n}``".
+
+The modifiers are, in general, expected to behave the same way they do in
+GCC. LLVM's support is often implemented on an 'as-needed' basis, to support C
+inline asm code which was supported by GCC. A mismatch in behavior between LLVM
+and GCC likely indicates a bug in LLVM.
+
+Target-independent:
+
+- ``c``: Print an immediate integer constant  unadorned, without
+  the target-specific immediate punctuation (e.g. no ``$`` prefix).
+- ``n``: Negate and print immediate integer constant unadorned, without the
+  target-specific immediate punctuation (e.g. no ``$`` prefix).
+- ``l``: Print as an unadorned label, without the target-specific label
+  punctuation (e.g. no ``$`` prefix).
+
+AArch64:
+
+- ``w``: Print a GPR register with a ``w*`` name instead of ``x*`` name. E.g.,
+  instead of ``x30``, print ``w30``.
+- ``x``: Print a GPR register with a ``x*`` name. (this is the default, anyhow).
+- ``b``, ``h``, ``s``, ``d``, ``q``: Print a floating-point/SIMD register with a
+  ``b*``, ``h*``, ``s*``, ``d*``, or ``q*`` name, rather than the default of
+  ``v*``.
+
+AMDGPU:
+
+- ``r``: No effect.
+
+ARM:
+
+- ``a``: Print an operand as an address (with ``[`` and ``]`` surrounding a
+  register).
+- ``P``: No effect.
+- ``q``: No effect.
+- ``y``: Print a VFP single-precision register as an indexed double (e.g. print
+  as ``d4[1]`` instead of ``s9``)
+- ``B``: Bitwise invert and print an immediate integer constant without ``#``
+  prefix.
+- ``L``: Print the low 16-bits of an immediate integer constant.
+- ``M``: Print as a register set suitable for ldm/stm. Also prints *all*
+  register operands subsequent to the specified one (!), so use carefully.
+- ``Q``: Print the low-order register of a register-pair, or the low-order
+  register of a two-register operand.
+- ``R``: Print the high-order register of a register-pair, or the high-order
+  register of a two-register operand.
+- ``H``: Print the second register of a register-pair. (On a big-endian system,
+  ``H`` is equivalent to ``Q``, and on little-endian system, ``H`` is equivalent
+  to ``R``.)
+
+  .. FIXME: H doesn't currently support printing the second register
+     of a two-register operand.
+
+- ``e``: Print the low doubleword register of a NEON quad register.
+- ``f``: Print the high doubleword register of a NEON quad register.
+- ``m``: Print the base register of a memory operand without the ``[`` and ``]``
+  adornment.
+
+Hexagon:
+
+- ``L``: Print the second register of a two-register operand. Requires that it
+  has been allocated consecutively to the first.
+
+  .. FIXME: why is it restricted to consecutive ones? And there's
+     nothing that ensures that happens, is there?
+
+- ``I``: Print the letter 'i' if the operand is an integer constant, otherwise
+  nothing. Used to print 'addi' vs 'add' instructions.
+
+MSP430:
+
+No additional modifiers.
+
+MIPS:
+
+- ``X``: Print an immediate integer as hexadecimal
+- ``x``: Print the low 16 bits of an immediate integer as hexadecimal.
+- ``d``: Print an immediate integer as decimal.
+- ``m``: Subtract one and print an immediate integer as decimal.
+- ``z``: Print $0 if an immediate zero, otherwise print normally.
+- ``L``: Print the low-order register of a two-register operand, or prints the
+  address of the low-order word of a double-word memory operand.
+
+  .. FIXME: L seems to be missing memory operand support.
+
+- ``M``: Print the high-order register of a two-register operand, or prints the
+  address of the high-order word of a double-word memory operand.
+
+  .. FIXME: M seems to be missing memory operand support.
+
+- ``D``: Print the second register of a two-register operand, or prints the
+  second word of a double-word memory operand. (On a big-endian system, ``D`` is
+  equivalent to ``L``, and on little-endian system, ``D`` is equivalent to
+  ``M``.)
+- ``w``: No effect. Provided for compatibility with GCC which requires this
+  modifier in order to print MSA registers (``W0-W31``) with the ``f``
+  constraint.
+
+NVPTX:
+
+- ``r``: No effect.
+
+PowerPC:
+
+- ``L``: Print the second register of a two-register operand. Requires that it
+  has been allocated consecutively to the first.
+
+  .. FIXME: why is it restricted to consecutive ones? And there's
+     nothing that ensures that happens, is there?
+
+- ``I``: Print the letter 'i' if the operand is an integer constant, otherwise
+  nothing. Used to print 'addi' vs 'add' instructions.
+- ``y``: For a memory operand, prints formatter for a two-register X-form
+  instruction. (Currently always prints ``r0,OPERAND``).
+- ``U``: Prints 'u' if the memory operand is an update form, and nothing
+  otherwise. (NOTE: LLVM does not support update form, so this will currently
+  always print nothing)
+- ``X``: Prints 'x' if the memory operand is an indexed form. (NOTE: LLVM does
+  not support indexed form, so this will currently always print nothing)
+
+Sparc:
+
+- ``r``: No effect.
+
+SystemZ:
+
+SystemZ implements only ``n``, and does *not* support any of the other
+target-independent modifiers.
+
+X86:
+
+- ``c``: Print an unadorned integer or symbol name. (The latter is
+  target-specific behavior for this typically target-independent modifier).
+- ``A``: Print a register name with a '``*``' before it.
+- ``b``: Print an 8-bit register name (e.g. ``al``); do nothing on a memory
+  operand.
+- ``h``: Print the upper 8-bit register name (e.g. ``ah``); do nothing on a
+  memory operand.
+- ``w``: Print the 16-bit register name (e.g. ``ax``); do nothing on a memory
+  operand.
+- ``k``: Print the 32-bit register name (e.g. ``eax``); do nothing on a memory
+  operand.
+- ``q``: Print the 64-bit register name (e.g. ``rax``), if 64-bit registers are
+  available, otherwise the 32-bit register name; do nothing on a memory operand.
+- ``n``: Negate and print an unadorned integer, or, for operands other than an
+  immediate integer (e.g. a relocatable symbol expression), print a '-' before
+  the operand. (The behavior for relocatable symbol expressions is a
+  target-specific behavior for this typically target-independent modifier)
+- ``H``: Print a memory reference with additional offset +8.
+- ``P``: Print a memory reference or operand for use as the argument of a call
+  instruction. (E.g. omit ``(rip)``, even though it's PC-relative.)
+
+XCore:
+
+No additional modifiers.
+
+
 Inline Asm Metadata
 ^^^^^^^^^^^^^^^^^^^
 
@@ -6108,7 +6729,8 @@ Overview:
 
 The '``getelementptr``' instruction is used to get the address of a
 subelement of an :ref:`aggregate <t_aggregate>` data structure. It performs
-address calculation only and does not access memory.
+address calculation only and does not access memory. The instruction can also
+be used to calculate a vector of such addresses.
 
 Arguments:
 """"""""""
@@ -6234,12 +6856,61 @@ Example:
         ; yields i32*:iptr
         %iptr = getelementptr [10 x i32], [10 x i32]* @arr, i16 0, i16 0
 
-In cases where the pointer argument is a vector of pointers, each index
-must be a vector with the same number of elements. For example:
+Vector of pointers:
+"""""""""""""""""""
+
+The ``getelementptr`` returns a vector of pointers, instead of a single address,
+when one or more of its arguments is a vector. In such cases, all vector
+arguments should have the same number of elements, and every scalar argument
+will be effectively broadcast into a vector during address calculation.
+
+.. code-block:: llvm
+
+     ; All arguments are vectors:
+     ;   A[i] = ptrs[i] + offsets[i]*sizeof(i8)
+     %A = getelementptr i8, <4 x i8*> %ptrs, <4 x i64> %offsets
+     
+     ; Add the same scalar offset to each pointer of a vector:
+     ;   A[i] = ptrs[i] + offset*sizeof(i8)
+     %A = getelementptr i8, <4 x i8*> %ptrs, i64 %offset
+     
+     ; Add distinct offsets to the same pointer:
+     ;   A[i] = ptr + offsets[i]*sizeof(i8)
+     %A = getelementptr i8, i8* %ptr, <4 x i64> %offsets
+     
+     ; In all cases described above the type of the result is <4 x i8*>
+
+The two following instructions are equivalent:
 
 .. code-block:: llvm
 
-     %A = getelementptr i8, <4 x i8*> %ptrs, <4 x i64> %offsets,
+     getelementptr  %struct.ST, <4 x %struct.ST*> %s, <4 x i64> %ind1,
+       <4 x i32> <i32 2, i32 2, i32 2, i32 2>,
+       <4 x i32> <i32 1, i32 1, i32 1, i32 1>,
+       <4 x i32> %ind4,
+       <4 x i64> <i64 13, i64 13, i64 13, i64 13>
+     
+     getelementptr  %struct.ST, <4 x %struct.ST*> %s, <4 x i64> %ind1,
+       i32 2, i32 1, <4 x i32> %ind4, i64 13
+
+Let's look at the C code, where the vector version of ``getelementptr``
+makes sense:
+
+.. code-block:: c
+
+    // Let's assume that we vectorize the following loop:
+    double *A, B; int *C;
+    for (int i = 0; i < size; ++i) {
+      A[i] = B[C[i]];
+    }
+
+.. code-block:: llvm
+
+    ; get pointers for 8 elements from array B
+    %ptrs = getelementptr double, double* %B, <8 x i32> %C
+    ; load 8 elements from array B into A
+    %A = call <8 x double> @llvm.masked.gather.v8f64(<8 x double*> %ptrs,
+         i32 8, <8 x i1> %mask, <8 x double> %passthru)
 
 Conversion Operations
 ---------------------
@@ -6913,7 +7584,7 @@ Syntax:
 
 ::
 
-      <result> = fcmp <cond> <ty> <op1>, <op2>     ; yields i1 or <N x i1>:result
+      <result> = fcmp [fast-math flags]* <cond> <ty> <op1>, <op2>     ; yields i1 or <N x i1>:result
 
 Overview:
 """""""""
@@ -6996,6 +7667,15 @@ always yields an :ref:`i1 <t_integer>` result, as follows:
 #. ``uno``: yields ``true`` if either operand is a QNAN.
 #. ``true``: always yields ``true``, regardless of operands.
 
+The ``fcmp`` instruction can also optionally take any number of
+:ref:`fast-math flags <fastmath>`, which are optimization hints to enable
+otherwise unsafe floating point optimizations.
+
+Any set of fast-math flags are legal on an ``fcmp`` instruction, but the
+only flags that have any effect on its semantics are those that allow
+assumptions to be made about the values of input arguments; namely
+``nnan``, ``ninf``, and ``nsz``. See :ref:`fastmath` for more information.
+
 Example:
 """"""""
 
@@ -7780,7 +8460,7 @@ Note that calling this intrinsic does not prevent function inlining or
 other aggressive transformations, so the value returned may not be that
 of the obvious source-language caller.
 
-'``llvm.frameescape``' and '``llvm.framerecover``' Intrinsics
+'``llvm.localescape``' and '``llvm.localrecover``' Intrinsics
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 Syntax:
@@ -7788,49 +8468,47 @@ Syntax:
 
 ::
 
-      declare void @llvm.frameescape(...)
-      declare i8* @llvm.framerecover(i8* %func, i8* %fp, i32 %idx)
+      declare void @llvm.localescape(...)
+      declare i8* @llvm.localrecover(i8* %func, i8* %fp, i32 %idx)
 
 Overview:
 """""""""
 
-The '``llvm.frameescape``' intrinsic escapes offsets of a collection of static
-allocas, and the '``llvm.framerecover``' intrinsic applies those offsets to a
+The '``llvm.localescape``' intrinsic escapes offsets of a collection of static
+allocas, and the '``llvm.localrecover``' intrinsic applies those offsets to a
 live frame pointer to recover the address of the allocation. The offset is
-computed during frame layout of the caller of ``llvm.frameescape``.
+computed during frame layout of the caller of ``llvm.localescape``.
 
 Arguments:
 """"""""""
 
-All arguments to '``llvm.frameescape``' must be pointers to static allocas or
-casts of static allocas. Each function can only call '``llvm.frameescape``'
+All arguments to '``llvm.localescape``' must be pointers to static allocas or
+casts of static allocas. Each function can only call '``llvm.localescape``'
 once, and it can only do so from the entry block.
 
-The ``func`` argument to '``llvm.framerecover``' must be a constant
+The ``func`` argument to '``llvm.localrecover``' must be a constant
 bitcasted pointer to a function defined in the current module. The code
 generator cannot determine the frame allocation offset of functions defined in
 other modules.
 
-The ``fp`` argument to '``llvm.framerecover``' must be a frame
-pointer of a call frame that is currently live. The return value of
-'``llvm.frameaddress``' is one way to produce such a value, but most platforms
-also expose the frame pointer through stack unwinding mechanisms.
+The ``fp`` argument to '``llvm.localrecover``' must be a frame pointer of a
+call frame that is currently live. The return value of '``llvm.localaddress``'
+is one way to produce such a value, but various runtimes also expose a suitable
+pointer in platform-specific ways.
 
-The ``idx`` argument to '``llvm.framerecover``' indicates which alloca passed to
-'``llvm.frameescape``' to recover. It is zero-indexed.
+The ``idx`` argument to '``llvm.localrecover``' indicates which alloca passed to
+'``llvm.localescape``' to recover. It is zero-indexed.
 
 Semantics:
 """"""""""
 
-These intrinsics allow a group of functions to access one stack memory
-allocation in an ancestor stack frame. The memory returned from
-'``llvm.frameallocate``' may be allocated prior to stack realignment, so the
-memory is only aligned to the ABI-required stack alignment.  Each function may
-only call '``llvm.frameallocate``' one or zero times from the function entry
-block.  The frame allocation intrinsic inhibits inlining, as any frame
-allocations in the inlined function frame are likely to be at a different
-offset from the one used by '``llvm.framerecover``' called with the
-uninlined function.
+These intrinsics allow a group of functions to share access to a set of local
+stack allocations of a one parent function. The parent function may call the
+'``llvm.localescape``' intrinsic once from the function entry block, and the
+child functions can use '``llvm.localrecover``' to access the escaped allocas.
+The '``llvm.localescape``' intrinsic blocks inlining, as inlining changes where
+the escaped allocas are allocated, which would break attempts to use
+'``llvm.localrecover``'.
 
 .. _int_read_register:
 .. _int_write_register:
@@ -9532,6 +10210,75 @@ Examples:
 Specialised Arithmetic Intrinsics
 ---------------------------------
 
+'``llvm.canonicalize.*``' Intrinsic
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Syntax:
+"""""""
+
+::
+
+      declare float @llvm.canonicalize.f32(float %a)
+      declare double @llvm.canonicalize.f64(double %b)
+
+Overview:
+"""""""""
+
+The '``llvm.canonicalize.*``' intrinsic returns the platform specific canonical
+encoding of a floating point number.  This canonicalization is useful for
+implementing certain numeric primitives such as frexp. The canonical encoding is
+defined by IEEE-754-2008 to be:
+
+::
+
+      2.1.8 canonical encoding: The preferred encoding of a floating-point
+      representation in a format.  Applied to declets, significands of finite
+      numbers, infinities, and NaNs, especially in decimal formats.
+
+This operation can also be considered equivalent to the IEEE-754-2008
+conversion of a floating-point value to the same format.  NaNs are handled
+according to section 6.2.
+
+Examples of non-canonical encodings:
+
+- x87 pseudo denormals, pseudo NaNs, pseudo Infinity, Unnormals.  These are
+  converted to a canonical representation per hardware-specific protocol.
+- Many normal decimal floating point numbers have non-canonical alternative
+  encodings.
+- Some machines, like GPUs or ARMv7 NEON, do not support subnormal values.
+  These are treated as non-canonical encodings of zero and with be flushed to
+  a zero of the same sign by this operation.
+
+Note that per IEEE-754-2008 6.2, systems that support signaling NaNs with
+default exception handling must signal an invalid exception, and produce a
+quiet NaN result.
+
+This function should always be implementable as multiplication by 1.0, provided
+that the compiler does not constant fold the operation.  Likewise, division by
+1.0 and ``llvm.minnum(x, x)`` are possible implementations.  Addition with
+-0.0 is also sufficient provided that the rounding mode is not -Infinity.
+
+``@llvm.canonicalize`` must preserve the equality relation.  That is:
+
+- ``(@llvm.canonicalize(x) == x)`` is equivalent to ``(x == x)``
+- ``(@llvm.canonicalize(x) == @llvm.canonicalize(y))`` is equivalent to
+  to ``(x == y)``
+
+Additionally, the sign of zero must be conserved:
+``@llvm.canonicalize(-0.0) = -0.0`` and ``@llvm.canonicalize(+0.0) = +0.0``
+
+The payload bits of a NaN must be conserved, with two exceptions.
+First, environments which use only a single canonical representation of NaN
+must perform said canonicalization.  Second, SNaNs must be quieted per the
+usual methods.
+
+The canonicalization operation may be optimized away if:
+
+- The input is known to be canonical.  For example, it was produced by a
+  floating-point operation that is required by the standard to be canonical.
+- The result is consumed only by (or fused with) other floating-point
+  operations.  That is, the bits of the floating point value are not examined.
+
 '``llvm.fmuladd.*``' Intrinsic
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
diff --git a/docs/ProgrammersManual.rst b/docs/ProgrammersManual.rst
index ceb39e18efdf..08cc61a187b5 100644
--- a/docs/ProgrammersManual.rst
+++ b/docs/ProgrammersManual.rst
@@ -1868,7 +1868,7 @@ Iterating over predecessors & successors of blocks
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 Iterating over the predecessors and successors of a block is quite easy with the
-routines defined in ``"llvm/Support/CFG.h"``.  Just use code like this to
+routines defined in ``"llvm/IR/CFG.h"``.  Just use code like this to
 iterate over all predecessors of BB:
 
 .. code-block:: c++
diff --git a/docs/StackMaps.rst b/docs/StackMaps.rst
index 43c60c9e7850..dbdf78f992ca 100644
--- a/docs/StackMaps.rst
+++ b/docs/StackMaps.rst
@@ -221,12 +221,10 @@ lowered according to the calling convention specified at the
 intrinsic's callsite. Variants of the intrinsic with non-void return
 type also return a value according to calling convention.
 
-On PowerPC, note that ``<target>`` must be the actual intended target of
-the indirect call. Specifically, even when compiling for the ELF V1 ABI,
-``<target>`` is not the function-descriptor address normally used as the C/C++
-function-pointer representation. As a result, the call target must be local
-because no adjustment or restoration of the TOC pointer (in register r2) will
-be performed.
+On PowerPC, note that ``<target>`` must be the ABI function pointer for the
+intended target of the indirect call. Specifically, when compiling for the
+ELF V1 ABI, ``<target>`` is the function-descriptor address normally used as
+the C/C++ function-pointer representation.
 
 Requesting zero patch point arguments is valid. In this case, all
 variable operands are handled just like
diff --git a/include/llvm-c/Core.h b/include/llvm-c/Core.h
index 713894f57639..15290072abe8 100644
--- a/include/llvm-c/Core.h
+++ b/include/llvm-c/Core.h
@@ -1888,6 +1888,20 @@ LLVMValueRef LLVMAddAlias(LLVMModuleRef M, LLVMTypeRef Ty, LLVMValueRef Aliasee,
 void LLVMDeleteFunction(LLVMValueRef Fn);
 
 /**
+ * Obtain the personality function attached to the function.
+ *
+ * @see llvm::Function::getPersonalityFn()
+ */
+LLVMValueRef LLVMGetPersonalityFn(LLVMValueRef Fn);
+
+/**
+ * Set the personality function attached to the function.
+ *
+ * @see llvm::Function::setPersonalityFn()
+ */
+void LLVMSetPersonalityFn(LLVMValueRef Fn, LLVMValueRef PersonalityFn);
+
+/**
  * Obtain the ID number from a function instance.
  *
  * @see llvm::Function::getIntrinsicID()
diff --git a/include/llvm-c/Object.h b/include/llvm-c/Object.h
index 447fcea7bc24..9cab5c426c45 100644
--- a/include/llvm-c/Object.h
+++ b/include/llvm-c/Object.h
@@ -81,7 +81,6 @@ uint64_t LLVMGetSymbolAddress(LLVMSymbolIteratorRef SI);
 uint64_t LLVMGetSymbolSize(LLVMSymbolIteratorRef SI);
 
 // RelocationRef accessors
-uint64_t LLVMGetRelocationAddress(LLVMRelocationIteratorRef RI);
 uint64_t LLVMGetRelocationOffset(LLVMRelocationIteratorRef RI);
 LLVMSymbolIteratorRef LLVMGetRelocationSymbol(LLVMRelocationIteratorRef RI);
 uint64_t LLVMGetRelocationType(LLVMRelocationIteratorRef RI);
diff --git a/include/llvm-c/lto.h b/include/llvm-c/lto.h
index 42c05a2fc59e..cb3a69160454 100644
--- a/include/llvm-c/lto.h
+++ b/include/llvm-c/lto.h
@@ -40,7 +40,7 @@ typedef bool lto_bool_t;
  * @{
  */
 
-#define LTO_API_VERSION 16
+#define LTO_API_VERSION 17
 
 /**
  * \since prior to LTO_API_VERSION=3
@@ -63,7 +63,8 @@ typedef enum {
     LTO_SYMBOL_SCOPE_PROTECTED             = 0x00002000,
     LTO_SYMBOL_SCOPE_DEFAULT               = 0x00001800,
     LTO_SYMBOL_SCOPE_DEFAULT_CAN_BE_HIDDEN = 0x00002800,
-    LTO_SYMBOL_COMDAT                      = 0x00004000
+    LTO_SYMBOL_COMDAT                      = 0x00004000,
+    LTO_SYMBOL_ALIAS                       = 0x00008000
 } lto_symbol_attributes;
 
 /**
diff --git a/include/llvm/ADT/APFloat.h b/include/llvm/ADT/APFloat.h
index 958e3fdaea14..76615affb253 100644
--- a/include/llvm/ADT/APFloat.h
+++ b/include/llvm/ADT/APFloat.h
@@ -276,6 +276,10 @@ public:
   /// \param isIEEE   - If 128 bit number, select between PPC and IEEE
   static APFloat getAllOnesValue(unsigned BitWidth, bool isIEEE = false);
 
+  /// Returns the size of the floating point number (in bits) in the given
+  /// semantics.
+  static unsigned getSizeInBits(const fltSemantics &Sem);
+
   /// @}
 
   /// Used to insert APFloat objects, or objects that contain APFloat objects,
diff --git a/include/llvm/ADT/Triple.h b/include/llvm/ADT/Triple.h
index 06f5870119c8..947812d94ecb 100644
--- a/include/llvm/ADT/Triple.h
+++ b/include/llvm/ADT/Triple.h
@@ -569,6 +569,22 @@ public:
   ///          architecture if no such variant can be found.
   llvm::Triple get64BitArchVariant() const;
 
+  /// Form a triple with a big endian variant of the current architecture.
+  ///
+  /// This can be used to move across "families" of architectures where useful.
+  ///
+  /// \returns A new triple with a big endian architecture or an unknown
+  ///          architecture if no such variant can be found.
+  llvm::Triple getBigEndianArchVariant() const;
+
+  /// Form a triple with a little endian variant of the current architecture.
+  ///
+  /// This can be used to move across "families" of architectures where useful.
+  ///
+  /// \returns A new triple with a little endian architecture or an unknown
+  ///          architecture if no such variant can be found.
+  llvm::Triple getLittleEndianArchVariant() const;
+
   /// Get the (LLVM) name of the minimum ARM CPU for the arch we are targeting.
   ///
   /// \param Arch the architecture name (e.g., "armv7s"). If it is an empty
diff --git a/include/llvm/ADT/edit_distance.h b/include/llvm/ADT/edit_distance.h
index c2b2041242aa..06a01b18a9fb 100644
--- a/include/llvm/ADT/edit_distance.h
+++ b/include/llvm/ADT/edit_distance.h
@@ -50,50 +50,51 @@ unsigned ComputeEditDistance(ArrayRef<T> FromArray, ArrayRef<T> ToArray,
   //   http://en.wikipedia.org/wiki/Levenshtein_distance
   //
   // Although the algorithm is typically described using an m x n
-  // array, only two rows are used at a time, so this implementation
-  // just keeps two separate vectors for those two rows.
+  // array, only one row plus one element are used at a time, so this
+  // implementation just keeps one vector for the row.  To update one entry,
+  // only the entries to the left, top, and top-left are needed.  The left
+  // entry is in Row[x-1], the top entry is what's in Row[x] from the last
+  // iteration, and the top-left entry is stored in Previous.
   typename ArrayRef<T>::size_type m = FromArray.size();
   typename ArrayRef<T>::size_type n = ToArray.size();
 
   const unsigned SmallBufferSize = 64;
   unsigned SmallBuffer[SmallBufferSize];
   std::unique_ptr<unsigned[]> Allocated;
-  unsigned *Previous = SmallBuffer;
-  if (2*(n + 1) > SmallBufferSize) {
-    Previous = new unsigned [2*(n+1)];
-    Allocated.reset(Previous);
+  unsigned *Row = SmallBuffer;
+  if (n + 1 > SmallBufferSize) {
+    Row = new unsigned[n + 1];
+    Allocated.reset(Row);
   }
-  unsigned *Current = Previous + (n + 1);
 
-  for (unsigned i = 0; i <= n; ++i)
-    Previous[i] = i;
+  for (unsigned i = 1; i <= n; ++i)
+    Row[i] = i;
 
   for (typename ArrayRef<T>::size_type y = 1; y <= m; ++y) {
-    Current[0] = y;
-    unsigned BestThisRow = Current[0];
+    Row[0] = y;
+    unsigned BestThisRow = Row[0];
 
+    unsigned Previous = y - 1;
     for (typename ArrayRef<T>::size_type x = 1; x <= n; ++x) {
+      int OldRow = Row[x];
       if (AllowReplacements) {
-        Current[x] = std::min(
-            Previous[x-1] + (FromArray[y-1] == ToArray[x-1] ? 0u : 1u),
-            std::min(Current[x-1], Previous[x])+1);
+        Row[x] = std::min(
+            Previous + (FromArray[y-1] == ToArray[x-1] ? 0u : 1u),
+            std::min(Row[x-1], Row[x])+1);
       }
       else {
-        if (FromArray[y-1] == ToArray[x-1]) Current[x] = Previous[x-1];
-        else Current[x] = std::min(Current[x-1], Previous[x]) + 1;
+        if (FromArray[y-1] == ToArray[x-1]) Row[x] = Previous;
+        else Row[x] = std::min(Row[x-1], Row[x]) + 1;
       }
-      BestThisRow = std::min(BestThisRow, Current[x]);
+      Previous = OldRow;
+      BestThisRow = std::min(BestThisRow, Row[x]);
     }
 
     if (MaxEditDistance && BestThisRow > MaxEditDistance)
       return MaxEditDistance + 1;
-
-    unsigned *tmp = Current;
-    Current = Previous;
-    Previous = tmp;
   }
 
-  unsigned Result = Previous[n];
+  unsigned Result = Row[n];
   return Result;
 }
 
diff --git a/include/llvm/Analysis/AliasAnalysis.h b/include/llvm/Analysis/AliasAnalysis.h
index f4c1167314a1..36f8199a0322 100644
--- a/include/llvm/Analysis/AliasAnalysis.h
+++ b/include/llvm/Analysis/AliasAnalysis.h
@@ -211,6 +211,8 @@ public:
     /// (if it has any) are non-volatile loads from objects pointed to by its
     /// pointer-typed arguments, with arbitrary offsets.
     ///
+    /// This property corresponds to the LLVM IR 'argmemonly' attribute combined
+    /// with 'readonly' attribute.
     /// This property corresponds to the IntrReadArgMem LLVM intrinsic flag.
     OnlyReadsArgumentPointees = ArgumentPointees | Ref,
 
@@ -218,6 +220,7 @@ public:
     /// function (if it has any) are non-volatile loads and stores from objects
     /// pointed to by its pointer-typed arguments, with arbitrary offsets.
     ///
+    /// This property corresponds to the LLVM IR 'argmemonly' attribute.
     /// This property corresponds to the IntrReadWriteArgMem LLVM intrinsic flag.
     OnlyAccessesArgumentPointees = ArgumentPointees | ModRef,
 
@@ -518,14 +521,6 @@ public:
   ///
   virtual void deleteValue(Value *V);
 
-  /// copyValue - This method should be used whenever a preexisting value in the
-  /// program is copied or cloned, introducing a new value.  Note that analysis
-  /// implementations should tolerate clients that use this method to introduce
-  /// the same value multiple times: if the analysis already knows about a
-  /// value, it should ignore the request.
-  ///
-  virtual void copyValue(Value *From, Value *To);
-
   /// addEscapingUse - This method should be used whenever an escaping use is
   /// added to a pointer value.  Analysis implementations may either return
   /// conservative responses for that value in the future, or may recompute
@@ -541,7 +536,6 @@ public:
   /// above, and it provided as a helper to simplify client code.
   ///
   void replaceWithNewValue(Value *Old, Value *New) {
-    copyValue(Old, New);
     deleteValue(Old);
   }
 };
diff --git a/include/llvm/Analysis/ConstantFolding.h b/include/llvm/Analysis/ConstantFolding.h
index 541a2109af6c..e8185b3b6307 100644
--- a/include/llvm/Analysis/ConstantFolding.h
+++ b/include/llvm/Analysis/ConstantFolding.h
@@ -72,6 +72,17 @@ namespace llvm {
 Constant *ConstantFoldInsertValueInstruction(Constant *Agg, Constant *Val,
                                              ArrayRef<unsigned> Idxs);
 
+/// \brief Attempt to constant fold an extractvalue instruction with the
+/// specified operands and indices.  The constant result is returned if
+/// successful; if not, null is returned.
+Constant *ConstantFoldExtractValueInstruction(Constant *Agg,
+                                              ArrayRef<unsigned> Idxs);
+
+/// \brief Attempt to constant fold an extractelement instruction with the
+/// specified operands and indices.  The constant result is returned if
+/// successful; if not, null is returned.
+Constant *ConstantFoldExtractElementInstruction(Constant *Val, Constant *Idx);
+
 /// ConstantFoldLoadFromConstPtr - Return the value that a load from C would
 /// produce if it is constant and determinable.  If this is not determinable,
 /// return null.
diff --git a/include/llvm/Analysis/DominanceFrontier.h b/include/llvm/Analysis/DominanceFrontier.h
index 996700efdb60..fb730054a8e5 100644
--- a/include/llvm/Analysis/DominanceFrontier.h
+++ b/include/llvm/Analysis/DominanceFrontier.h
@@ -202,8 +202,8 @@ public:
   void dump() const;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class DominanceFrontierBase<BasicBlock>);
-EXTERN_TEMPLATE_INSTANTIATION(class ForwardDominanceFrontierBase<BasicBlock>);
+extern template class DominanceFrontierBase<BasicBlock>;
+extern template class ForwardDominanceFrontierBase<BasicBlock>;
 
 } // End llvm namespace
 
diff --git a/include/llvm/Analysis/IVUsers.h b/include/llvm/Analysis/IVUsers.h
index ae9c1f5bd9ac..00dbcbdd7806 100644
--- a/include/llvm/Analysis/IVUsers.h
+++ b/include/llvm/Analysis/IVUsers.h
@@ -21,6 +21,7 @@
 
 namespace llvm {
 
+class AssumptionCache;
 class DominatorTree;
 class Instruction;
 class Value;
@@ -119,15 +120,19 @@ private:
 class IVUsers : public LoopPass {
   friend class IVStrideUse;
   Loop *L;
+  AssumptionCache *AC;
   LoopInfo *LI;
   DominatorTree *DT;
   ScalarEvolution *SE;
-  SmallPtrSet<Instruction*,16> Processed;
+  SmallPtrSet<Instruction*, 16> Processed;
 
   /// IVUses - A list of all tracked IV uses of induction variable expressions
   /// we are interested in.
   ilist<IVStrideUse> IVUses;
 
+  // Ephemeral values used by @llvm.assume in this function.
+  SmallPtrSet<const Value *, 32> EphValues;
+
   void getAnalysisUsage(AnalysisUsage &AU) const override;
 
   bool runOnLoop(Loop *L, LPPassManager &LPM) override;
diff --git a/include/llvm/Analysis/InstructionSimplify.h b/include/llvm/Analysis/InstructionSimplify.h
index 706bd8000d3a..d44c5ff4078d 100644
--- a/include/llvm/Analysis/InstructionSimplify.h
+++ b/include/llvm/Analysis/InstructionSimplify.h
@@ -212,7 +212,7 @@ namespace llvm {
   /// SimplifyFCmpInst - Given operands for an FCmpInst, see if we can
   /// fold the result.  If not, this returns null.
   Value *SimplifyFCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
-                          const DataLayout &DL,
+                          FastMathFlags FMF, const DataLayout &DL,
                           const TargetLibraryInfo *TLI = nullptr,
                           const DominatorTree *DT = nullptr,
                           AssumptionCache *AC = nullptr,
@@ -244,6 +244,24 @@ namespace llvm {
                                  AssumptionCache *AC = nullptr,
                                  const Instruction *CxtI = nullptr);
 
+  /// \brief Given operands for an ExtractValueInst, see if we can fold the
+  /// result.  If not, this returns null.
+  Value *SimplifyExtractValueInst(Value *Agg, ArrayRef<unsigned> Idxs,
+                                  const DataLayout &DL,
+                                  const TargetLibraryInfo *TLI = nullptr,
+                                  const DominatorTree *DT = nullptr,
+                                  AssumptionCache *AC = nullptr,
+                                  const Instruction *CxtI = nullptr);
+
+  /// \brief Given operands for an ExtractElementInst, see if we can fold the
+  /// result.  If not, this returns null.
+  Value *SimplifyExtractElementInst(Value *Vec, Value *Idx,
+                                    const DataLayout &DL,
+                                    const TargetLibraryInfo *TLI = nullptr,
+                                    const DominatorTree *DT = nullptr,
+                                    AssumptionCache *AC = nullptr,
+                                    const Instruction *CxtI = nullptr);
+
   /// SimplifyTruncInst - Given operands for an TruncInst, see if we can fold
   /// the result.  If not, this returns null.
   Value *SimplifyTruncInst(Value *Op, Type *Ty, const DataLayout &DL,
diff --git a/include/llvm/Analysis/JumpInstrTableInfo.h b/include/llvm/Analysis/JumpInstrTableInfo.h
deleted file mode 100644
index b6dad478cdf2..000000000000
--- a/include/llvm/Analysis/JumpInstrTableInfo.h
+++ /dev/null
@@ -1,71 +0,0 @@
-//===-- JumpInstrTableInfo.h: Info for Jump-Instruction Tables --*- C++ -*-===//
-//
-// This file is distributed under the University of Illinois Open Source
-// License. See LICENSE.TXT for details.
-//
-//===----------------------------------------------------------------------===//
-///
-/// \file
-/// \brief Information about jump-instruction tables that have been created by
-/// JumpInstrTables pass.
-///
-//===----------------------------------------------------------------------===//
-
-#ifndef LLVM_ANALYSIS_JUMPINSTRTABLEINFO_H
-#define LLVM_ANALYSIS_JUMPINSTRTABLEINFO_H
-
-#include "llvm/ADT/DenseMap.h"
-#include "llvm/Pass.h"
-#include <vector>
-
-namespace llvm {
-class Function;
-class FunctionType;
-
-/// This class stores information about jump-instruction tables created by the
-/// JumpInstrTables pass (in lib/CodeGen/JumpInstrTables.cpp). Each table is a
-/// map from a function type to a vector of pairs. The first element of each
-/// pair is the function that has the jumptable annotation. The second element
-/// is a function that was declared by JumpInstrTables and used to replace all
-/// address-taking sites for the original function.
-///
-/// The information in this pass is used in AsmPrinter
-/// (lib/CodeGen/AsmPrinter/AsmPrinter.cpp) to generate the required assembly
-/// for the jump-instruction tables.
-class JumpInstrTableInfo : public ImmutablePass {
-public:
-  static char ID;
-
-  /// The default byte alignment for jump tables is 16, which is large but
-  /// usually safe.
-  JumpInstrTableInfo(uint64_t ByteAlign = 16);
-  ~JumpInstrTableInfo() override;
-  const char *getPassName() const override {
-    return "Jump-Instruction Table Info";
-  }
-
-  typedef std::pair<Function *, Function *> JumpPair;
-  typedef DenseMap<FunctionType *, std::vector<JumpPair> > JumpTables;
-
-  /// Inserts an entry in a table, adding the table if it doesn't exist.
-  void insertEntry(FunctionType *TableFunTy, Function *Target, Function *Jump);
-
-  /// Gets the tables.
-  const JumpTables &getTables() const { return Tables; }
-
-  /// Gets the alignment in bytes of a jumptable entry.
-  uint64_t entryByteAlignment() const { return ByteAlignment; }
-private:
-  JumpTables Tables;
-
-  /// A power-of-two alignment of a jumptable entry.
-  uint64_t ByteAlignment;
-};
-
-/// Creates a JumpInstrTableInfo pass with the given bound on entry size. This
-/// bound specifies the maximum number of bytes needed to represent an
-/// unconditional jump or a trap instruction in the back end currently in use.
-ModulePass *createJumpInstrTableInfoPass(unsigned Bound);
-}
-
-#endif /* LLVM_ANALYSIS_JUMPINSTRTABLEINFO_H */
diff --git a/include/llvm/Analysis/LibCallSemantics.h b/include/llvm/Analysis/LibCallSemantics.h
index 170e2a49a8ea..b4bef310e590 100644
--- a/include/llvm/Analysis/LibCallSemantics.h
+++ b/include/llvm/Analysis/LibCallSemantics.h
@@ -206,6 +206,18 @@ class InvokeInst;
     llvm_unreachable("invalid enum");
   }
 
+  /// \brief Return true if this personality may be safely removed if there
+  /// are no invoke instructions remaining in the current function.
+  inline bool isNoOpWithoutInvoke(EHPersonality Pers) {
+    switch (Pers) {
+    case EHPersonality::Unknown:
+      return false;
+    // All known personalities currently have this behavior
+    default: return true;
+    }
+    llvm_unreachable("invalid enum");
+  }
+
   bool canSimplifyInvokeNoUnwind(const Function *F);
 
 } // end namespace llvm
diff --git a/include/llvm/Analysis/LoopAccessAnalysis.h b/include/llvm/Analysis/LoopAccessAnalysis.h
index 7b635a8b4960..476e4b6686bb 100644
--- a/include/llvm/Analysis/LoopAccessAnalysis.h
+++ b/include/llvm/Analysis/LoopAccessAnalysis.h
@@ -292,6 +292,133 @@ private:
   bool couldPreventStoreLoadForward(unsigned Distance, unsigned TypeByteSize);
 };
 
+/// \brief Holds information about the memory runtime legality checks to verify
+/// that a group of pointers do not overlap.
+class RuntimePointerChecking {
+public:
+  struct PointerInfo {
+    /// Holds the pointer value that we need to check.
+    TrackingVH<Value> PointerValue;
+    /// Holds the pointer value at the beginning of the loop.
+    const SCEV *Start;
+    /// Holds the pointer value at the end of the loop.
+    const SCEV *End;
+    /// Holds the information if this pointer is used for writing to memory.
+    bool IsWritePtr;
+    /// Holds the id of the set of pointers that could be dependent because of a
+    /// shared underlying object.
+    unsigned DependencySetId;
+    /// Holds the id of the disjoint alias set to which this pointer belongs.
+    unsigned AliasSetId;
+    /// SCEV for the access.
+    const SCEV *Expr;
+
+    PointerInfo(Value *PointerValue, const SCEV *Start, const SCEV *End,
+                bool IsWritePtr, unsigned DependencySetId, unsigned AliasSetId,
+                const SCEV *Expr)
+        : PointerValue(PointerValue), Start(Start), End(End),
+          IsWritePtr(IsWritePtr), DependencySetId(DependencySetId),
+          AliasSetId(AliasSetId), Expr(Expr) {}
+  };
+
+  RuntimePointerChecking(ScalarEvolution *SE) : Need(false), SE(SE) {}
+
+  /// Reset the state of the pointer runtime information.
+  void reset() {
+    Need = false;
+    Pointers.clear();
+  }
+
+  /// Insert a pointer and calculate the start and end SCEVs.
+  void insert(Loop *Lp, Value *Ptr, bool WritePtr, unsigned DepSetId,
+              unsigned ASId, const ValueToValueMap &Strides);
+
+  /// \brief No run-time memory checking is necessary.
+  bool empty() const { return Pointers.empty(); }
+
+  /// A grouping of pointers. A single memcheck is required between
+  /// two groups.
+  struct CheckingPtrGroup {
+    /// \brief Create a new pointer checking group containing a single
+    /// pointer, with index \p Index in RtCheck.
+    CheckingPtrGroup(unsigned Index, RuntimePointerChecking &RtCheck)
+        : RtCheck(RtCheck), High(RtCheck.Pointers[Index].End),
+          Low(RtCheck.Pointers[Index].Start) {
+      Members.push_back(Index);
+    }
+
+    /// \brief Tries to add the pointer recorded in RtCheck at index
+    /// \p Index to this pointer checking group. We can only add a pointer
+    /// to a checking group if we will still be able to get
+    /// the upper and lower bounds of the check. Returns true in case
+    /// of success, false otherwise.
+    bool addPointer(unsigned Index);
+
+    /// Constitutes the context of this pointer checking group. For each
+    /// pointer that is a member of this group we will retain the index
+    /// at which it appears in RtCheck.
+    RuntimePointerChecking &RtCheck;
+    /// The SCEV expression which represents the upper bound of all the
+    /// pointers in this group.
+    const SCEV *High;
+    /// The SCEV expression which represents the lower bound of all the
+    /// pointers in this group.
+    const SCEV *Low;
+    /// Indices of all the pointers that constitute this grouping.
+    SmallVector<unsigned, 2> Members;
+  };
+
+  /// \brief Groups pointers such that a single memcheck is required
+  /// between two different groups. This will clear the CheckingGroups vector
+  /// and re-compute it. We will only group dependecies if \p UseDependencies
+  /// is true, otherwise we will create a separate group for each pointer.
+  void groupChecks(MemoryDepChecker::DepCandidates &DepCands,
+                   bool UseDependencies);
+
+  /// \brief Decide if we need to add a check between two groups of pointers,
+  /// according to needsChecking.
+  bool needsChecking(const CheckingPtrGroup &M, const CheckingPtrGroup &N,
+                     const SmallVectorImpl<int> *PtrPartition) const;
+
+  /// \brief Return true if any pointer requires run-time checking according
+  /// to needsChecking.
+  bool needsAnyChecking(const SmallVectorImpl<int> *PtrPartition) const;
+
+  /// \brief Returns the number of run-time checks required according to
+  /// needsChecking.
+  unsigned getNumberOfChecks(const SmallVectorImpl<int> *PtrPartition) const;
+
+  /// \brief Print the list run-time memory checks necessary.
+  ///
+  /// If \p PtrPartition is set, it contains the partition number for
+  /// pointers (-1 if the pointer belongs to multiple partitions).  In this
+  /// case omit checks between pointers belonging to the same partition.
+  void print(raw_ostream &OS, unsigned Depth = 0,
+             const SmallVectorImpl<int> *PtrPartition = nullptr) const;
+
+  /// This flag indicates if we need to add the runtime check.
+  bool Need;
+
+  /// Information about the pointers that may require checking.
+  SmallVector<PointerInfo, 2> Pointers;
+
+  /// Holds a partitioning of pointers into "check groups".
+  SmallVector<CheckingPtrGroup, 2> CheckingGroups;
+
+private:
+  /// \brief Decide whether we need to issue a run-time check for pointer at
+  /// index \p I and \p J to prove their independence.
+  ///
+  /// If \p PtrPartition is set, it contains the partition number for
+  /// pointers (-1 if the pointer belongs to multiple partitions).  In this
+  /// case omit checks between pointers belonging to the same partition.
+  bool needsChecking(unsigned I, unsigned J,
+                     const SmallVectorImpl<int> *PtrPartition) const;
+
+  /// Holds a pointer to the ScalarEvolution analysis.
+  ScalarEvolution *SE;
+};
+
 /// \brief Drive the analysis of memory accesses in the loop
 ///
 /// This class is responsible for analyzing the memory accesses of a loop.  It
@@ -308,72 +435,6 @@ private:
 /// RuntimePointerCheck class.
 class LoopAccessInfo {
 public:
-  /// This struct holds information about the memory runtime legality check that
-  /// a group of pointers do not overlap.
-  struct RuntimePointerCheck {
-    RuntimePointerCheck() : Need(false) {}
-
-    /// Reset the state of the pointer runtime information.
-    void reset() {
-      Need = false;
-      Pointers.clear();
-      Starts.clear();
-      Ends.clear();
-      IsWritePtr.clear();
-      DependencySetId.clear();
-      AliasSetId.clear();
-    }
-
-    /// Insert a pointer and calculate the start and end SCEVs.
-    void insert(ScalarEvolution *SE, Loop *Lp, Value *Ptr, bool WritePtr,
-                unsigned DepSetId, unsigned ASId,
-                const ValueToValueMap &Strides);
-
-    /// \brief No run-time memory checking is necessary.
-    bool empty() const { return Pointers.empty(); }
-
-    /// \brief Decide whether we need to issue a run-time check for pointer at
-    /// index \p I and \p J to prove their independence.
-    ///
-    /// If \p PtrPartition is set, it contains the partition number for
-    /// pointers (-1 if the pointer belongs to multiple partitions).  In this
-    /// case omit checks between pointers belonging to the same partition.
-    bool needsChecking(unsigned I, unsigned J,
-                       const SmallVectorImpl<int> *PtrPartition) const;
-
-    /// \brief Return true if any pointer requires run-time checking according
-    /// to needsChecking.
-    bool needsAnyChecking(const SmallVectorImpl<int> *PtrPartition) const;
-
-    /// \brief Returns the number of run-time checks required according to
-    /// needsChecking.
-    unsigned getNumberOfChecks(const SmallVectorImpl<int> *PtrPartition) const;
-
-    /// \brief Print the list run-time memory checks necessary.
-    ///
-    /// If \p PtrPartition is set, it contains the partition number for
-    /// pointers (-1 if the pointer belongs to multiple partitions).  In this
-    /// case omit checks between pointers belonging to the same partition.
-    void print(raw_ostream &OS, unsigned Depth = 0,
-               const SmallVectorImpl<int> *PtrPartition = nullptr) const;
-
-    /// This flag indicates if we need to add the runtime check.
-    bool Need;
-    /// Holds the pointers that we need to check.
-    SmallVector<TrackingVH<Value>, 2> Pointers;
-    /// Holds the pointer value at the beginning of the loop.
-    SmallVector<const SCEV*, 2> Starts;
-    /// Holds the pointer value at the end of the loop.
-    SmallVector<const SCEV*, 2> Ends;
-    /// Holds the information if this pointer is used for writing to memory.
-    SmallVector<bool, 2> IsWritePtr;
-    /// Holds the id of the set of pointers that could be dependent because of a
-    /// shared underlying object.
-    SmallVector<unsigned, 2> DependencySetId;
-    /// Holds the id of the disjoint alias set to which this pointer belongs.
-    SmallVector<unsigned, 2> AliasSetId;
-  };
-
   LoopAccessInfo(Loop *L, ScalarEvolution *SE, const DataLayout &DL,
                  const TargetLibraryInfo *TLI, AliasAnalysis *AA,
                  DominatorTree *DT, LoopInfo *LI,
@@ -383,15 +444,15 @@ public:
   /// no memory dependence cycles.
   bool canVectorizeMemory() const { return CanVecMem; }
 
-  const RuntimePointerCheck *getRuntimePointerCheck() const {
-    return &PtrRtCheck;
+  const RuntimePointerChecking *getRuntimePointerChecking() const {
+    return &PtrRtChecking;
   }
 
   /// \brief Number of memchecks required to prove independence of otherwise
   /// may-alias pointers.
   unsigned getNumRuntimePointerChecks(
     const SmallVectorImpl<int> *PtrPartition = nullptr) const {
-    return PtrRtCheck.getNumberOfChecks(PtrPartition);
+    return PtrRtChecking.getNumberOfChecks(PtrPartition);
   }
 
   /// Return true if the block BB needs to be predicated in order for the loop
@@ -461,7 +522,7 @@ private:
 
   /// We need to check that all of the pointers in this list are disjoint
   /// at runtime.
-  RuntimePointerCheck PtrRtCheck;
+  RuntimePointerChecking PtrRtChecking;
 
   /// \brief the Memory Dependence Checker which can determine the
   /// loop-independent and loop-carried dependences between memory accesses.
diff --git a/include/llvm/Analysis/LoopInfo.h b/include/llvm/Analysis/LoopInfo.h
index bbcde8d9721a..3ec83f2c21fd 100644
--- a/include/llvm/Analysis/LoopInfo.h
+++ b/include/llvm/Analysis/LoopInfo.h
@@ -347,9 +347,7 @@ raw_ostream& operator<<(raw_ostream &OS, const LoopBase<BlockT, LoopT> &Loop) {
 }
 
 // Implementation in LoopInfoImpl.h
-#ifdef __GNUC__
-__extension__ extern template class LoopBase<BasicBlock, Loop>;
-#endif
+extern template class LoopBase<BasicBlock, Loop>;
 
 class Loop : public LoopBase<BasicBlock, Loop> {
 public:
@@ -633,9 +631,7 @@ public:
 };
 
 // Implementation in LoopInfoImpl.h
-#ifdef __GNUC__
-__extension__ extern template class LoopInfoBase<BasicBlock, Loop>;
-#endif
+extern template class LoopInfoBase<BasicBlock, Loop>;
 
 class LoopInfo : public LoopInfoBase<BasicBlock, Loop> {
   typedef LoopInfoBase<BasicBlock, Loop> BaseT;
diff --git a/include/llvm/Analysis/RegionInfo.h b/include/llvm/Analysis/RegionInfo.h
index 7ceb086ee0a1..8560f1f67160 100644
--- a/include/llvm/Analysis/RegionInfo.h
+++ b/include/llvm/Analysis/RegionInfo.h
@@ -902,9 +902,9 @@ inline raw_ostream &operator<<(raw_ostream &OS,
     return OS << Node.template getNodeAs<BlockT>()->getName();
 }
 
-EXTERN_TEMPLATE_INSTANTIATION(class RegionBase<RegionTraits<Function>>);
-EXTERN_TEMPLATE_INSTANTIATION(class RegionNodeBase<RegionTraits<Function>>);
-EXTERN_TEMPLATE_INSTANTIATION(class RegionInfoBase<RegionTraits<Function>>);
+extern template class RegionBase<RegionTraits<Function>>;
+extern template class RegionNodeBase<RegionTraits<Function>>;
+extern template class RegionInfoBase<RegionTraits<Function>>;
 
 } // End llvm namespace
 #endif
diff --git a/include/llvm/Analysis/TargetTransformInfo.h b/include/llvm/Analysis/TargetTransformInfo.h
index bb6e266b1f5b..01f00896410e 100644
--- a/include/llvm/Analysis/TargetTransformInfo.h
+++ b/include/llvm/Analysis/TargetTransformInfo.h
@@ -69,7 +69,7 @@ public:
   ///
   /// The TTI implementation will reflect the information in the DataLayout
   /// provided if non-null.
-  explicit TargetTransformInfo(const DataLayout *DL);
+  explicit TargetTransformInfo(const DataLayout &DL);
 
   // Provide move semantics.
   TargetTransformInfo(TargetTransformInfo &&Arg);
@@ -541,7 +541,7 @@ private:
 class TargetTransformInfo::Concept {
 public:
   virtual ~Concept() = 0;
-
+  virtual const DataLayout &getDataLayout() const = 0;
   virtual unsigned getOperationCost(unsigned Opcode, Type *Ty, Type *OpTy) = 0;
   virtual unsigned getGEPCost(const Value *Ptr,
                               ArrayRef<const Value *> Operands) = 0;
@@ -636,6 +636,10 @@ public:
   Model(T Impl) : Impl(std::move(Impl)) {}
   ~Model() override {}
 
+  const DataLayout &getDataLayout() const override {
+    return Impl.getDataLayout();
+  }
+
   unsigned getOperationCost(unsigned Opcode, Type *Ty, Type *OpTy) override {
     return Impl.getOperationCost(Opcode, Ty, OpTy);
   }
diff --git a/include/llvm/Analysis/TargetTransformInfoImpl.h b/include/llvm/Analysis/TargetTransformInfoImpl.h
index 403175acae02..035cb04870a1 100644
--- a/include/llvm/Analysis/TargetTransformInfoImpl.h
+++ b/include/llvm/Analysis/TargetTransformInfoImpl.h
@@ -30,26 +30,17 @@ class TargetTransformInfoImplBase {
 protected:
   typedef TargetTransformInfo TTI;
 
-  const DataLayout *DL;
+  const DataLayout &DL;
 
-  explicit TargetTransformInfoImplBase(const DataLayout *DL)
-      : DL(DL) {}
+  explicit TargetTransformInfoImplBase(const DataLayout &DL) : DL(DL) {}
 
 public:
   // Provide value semantics. MSVC requires that we spell all of these out.
   TargetTransformInfoImplBase(const TargetTransformInfoImplBase &Arg)
       : DL(Arg.DL) {}
-  TargetTransformInfoImplBase(TargetTransformInfoImplBase &&Arg)
-      : DL(std::move(Arg.DL)) {}
-  TargetTransformInfoImplBase &
-  operator=(const TargetTransformInfoImplBase &RHS) {
-    DL = RHS.DL;
-    return *this;
-  }
-  TargetTransformInfoImplBase &operator=(TargetTransformInfoImplBase &&RHS) {
-    DL = std::move(RHS.DL);
-    return *this;
-  }
+  TargetTransformInfoImplBase(TargetTransformInfoImplBase &&Arg) : DL(Arg.DL) {}
+
+  const DataLayout &getDataLayout() const { return DL; }
 
   unsigned getOperationCost(unsigned Opcode, Type *Ty, Type *OpTy) {
     switch (Opcode) {
@@ -70,28 +61,22 @@ public:
       return TTI::TCC_Basic;
 
     case Instruction::IntToPtr: {
-      if (!DL)
-        return TTI::TCC_Basic;
-
       // An inttoptr cast is free so long as the input is a legal integer type
       // which doesn't contain values outside the range of a pointer.
       unsigned OpSize = OpTy->getScalarSizeInBits();
-      if (DL->isLegalInteger(OpSize) &&
-          OpSize <= DL->getPointerTypeSizeInBits(Ty))
+      if (DL.isLegalInteger(OpSize) &&
+          OpSize <= DL.getPointerTypeSizeInBits(Ty))
         return TTI::TCC_Free;
 
       // Otherwise it's not a no-op.
       return TTI::TCC_Basic;
     }
     case Instruction::PtrToInt: {
-      if (!DL)
-        return TTI::TCC_Basic;
-
       // A ptrtoint cast is free so long as the result is large enough to store
       // the pointer, and a legal integer type.
       unsigned DestSize = Ty->getScalarSizeInBits();
-      if (DL->isLegalInteger(DestSize) &&
-          DestSize >= DL->getPointerTypeSizeInBits(OpTy))
+      if (DL.isLegalInteger(DestSize) &&
+          DestSize >= DL.getPointerTypeSizeInBits(OpTy))
         return TTI::TCC_Free;
 
       // Otherwise it's not a no-op.
@@ -100,7 +85,7 @@ public:
     case Instruction::Trunc:
       // trunc to a native type is free (assuming the target has compare and
       // shift-right of the same width).
-      if (DL && DL->isLegalInteger(DL->getTypeSizeInBits(Ty)))
+      if (DL.isLegalInteger(DL.getTypeSizeInBits(Ty)))
         return TTI::TCC_Free;
 
       return TTI::TCC_Basic;
@@ -353,8 +338,7 @@ private:
   typedef TargetTransformInfoImplBase BaseT;
 
 protected:
-  explicit TargetTransformInfoImplCRTPBase(const DataLayout *DL)
-      : BaseT(DL) {}
+  explicit TargetTransformInfoImplCRTPBase(const DataLayout &DL) : BaseT(DL) {}
 
 public:
   // Provide value semantics. MSVC requires that we spell all of these out.
@@ -362,16 +346,6 @@ public:
       : BaseT(static_cast<const BaseT &>(Arg)) {}
   TargetTransformInfoImplCRTPBase(TargetTransformInfoImplCRTPBase &&Arg)
       : BaseT(std::move(static_cast<BaseT &>(Arg))) {}
-  TargetTransformInfoImplCRTPBase &
-  operator=(const TargetTransformInfoImplCRTPBase &RHS) {
-    BaseT::operator=(static_cast<const BaseT &>(RHS));
-    return *this;
-  }
-  TargetTransformInfoImplCRTPBase &
-  operator=(TargetTransformInfoImplCRTPBase &&RHS) {
-    BaseT::operator=(std::move(static_cast<BaseT &>(RHS)));
-    return *this;
-  }
 
   using BaseT::getCallCost;
 
diff --git a/include/llvm/Analysis/VectorUtils.h b/include/llvm/Analysis/VectorUtils.h
index aa538ecc0137..d8e9ca42e623 100644
--- a/include/llvm/Analysis/VectorUtils.h
+++ b/include/llvm/Analysis/VectorUtils.h
@@ -20,6 +20,12 @@
 
 namespace llvm {
 
+class GetElementPtrInst;
+class Loop;
+class ScalarEvolution;
+class Type;
+class Value;
+
 /// \brief Identify if the intrinsic is trivially vectorizable.
 /// This method returns true if the intrinsic's argument types are all
 /// scalars for the scalar form of the intrinsic and all vectors for
@@ -51,6 +57,28 @@ Intrinsic::ID checkBinaryFloatSignature(const CallInst &I,
 /// its intrinsic ID, in case it does not found it return not_intrinsic.
 Intrinsic::ID getIntrinsicIDForCall(CallInst *CI, const TargetLibraryInfo *TLI);
 
+/// \brief Find the operand of the GEP that should be checked for consecutive
+/// stores. This ignores trailing indices that have no effect on the final
+/// pointer.
+unsigned getGEPInductionOperand(const GetElementPtrInst *Gep);
+
+/// \brief If the argument is a GEP, then returns the operand identified by 
+/// getGEPInductionOperand. However, if there is some other non-loop-invariant 
+/// operand, it returns that instead.
+Value *stripGetElementPtr(Value *Ptr, ScalarEvolution *SE, Loop *Lp);
+
+/// \brief If a value has only one user that is a CastInst, return it.
+Value *getUniqueCastUse(Value *Ptr, Loop *Lp, Type *Ty);
+
+/// \brief Get the stride of a pointer access in a loop. Looks for symbolic
+/// strides "a[i*stride]". Returns the symbolic stride, or null otherwise.
+Value *getStrideFromPointer(Value *Ptr, ScalarEvolution *SE, Loop *Lp);
+
+/// \brief Given a vector and an element number, see if the scalar value is
+/// already around as a register, for example if it were inserted then extracted
+/// from the vector.
+Value *findScalarElement(Value *V, unsigned EltNo);
+
 } // llvm namespace
 
 #endif
diff --git a/include/llvm/Bitcode/LLVMBitCodes.h b/include/llvm/Bitcode/LLVMBitCodes.h
index 605c4172dd87..7130ee755237 100644
--- a/include/llvm/Bitcode/LLVMBitCodes.h
+++ b/include/llvm/Bitcode/LLVMBitCodes.h
@@ -407,6 +407,7 @@ namespace bitc {
     ATTR_KIND_DEREFERENCEABLE_OR_NULL = 42,
     ATTR_KIND_CONVERGENT = 43,
     ATTR_KIND_SAFESTACK = 44,
+    ATTR_KIND_ARGMEMONLY = 45
   };
 
   enum ComdatSelectionKindCodes {
diff --git a/include/llvm/Bitcode/ReaderWriter.h b/include/llvm/Bitcode/ReaderWriter.h
index 6797aa133c42..452ec3bd0187 100644
--- a/include/llvm/Bitcode/ReaderWriter.h
+++ b/include/llvm/Bitcode/ReaderWriter.h
@@ -146,7 +146,7 @@ namespace llvm {
   }
 
   const std::error_category &BitcodeErrorCategory();
-  enum class BitcodeError { InvalidBitcodeSignature, CorruptedBitcode };
+  enum class BitcodeError { InvalidBitcodeSignature = 1, CorruptedBitcode };
   inline std::error_code make_error_code(BitcodeError E) {
     return std::error_code(static_cast<int>(E), BitcodeErrorCategory());
   }
diff --git a/include/llvm/CodeGen/Analysis.h b/include/llvm/CodeGen/Analysis.h
index c4b94ede4f55..82d1e8ada17d 100644
--- a/include/llvm/CodeGen/Analysis.h
+++ b/include/llvm/CodeGen/Analysis.h
@@ -64,7 +64,7 @@ inline unsigned ComputeLinearIndex(Type *Ty,
 /// If Offsets is non-null, it points to a vector to be filled in
 /// with the in-memory offsets of each of the individual values.
 ///
-void ComputeValueVTs(const TargetLowering &TLI, Type *Ty,
+void ComputeValueVTs(const TargetLowering &TLI, const DataLayout &DL, Type *Ty,
                      SmallVectorImpl<EVT> &ValueVTs,
                      SmallVectorImpl<uint64_t> *Offsets = nullptr,
                      uint64_t StartingOffset = 0);
diff --git a/include/llvm/CodeGen/BasicTTIImpl.h b/include/llvm/CodeGen/BasicTTIImpl.h
index 3e464f4f1e5a..9ba25169fda6 100644
--- a/include/llvm/CodeGen/BasicTTIImpl.h
+++ b/include/llvm/CodeGen/BasicTTIImpl.h
@@ -91,8 +91,10 @@ private:
   }
 
 protected:
-  explicit BasicTTIImplBase(const TargetMachine *TM)
-      : BaseT(TM->getDataLayout()) {}
+  explicit BasicTTIImplBase(const TargetMachine *TM, const DataLayout &DL)
+      : BaseT(DL) {}
+
+  using TargetTransformInfoImplBase::DL;
 
 public:
   // Provide value semantics. MSVC requires that we spell all of these out.
@@ -100,14 +102,6 @@ public:
       : BaseT(static_cast<const BaseT &>(Arg)) {}
   BasicTTIImplBase(BasicTTIImplBase &&Arg)
       : BaseT(std::move(static_cast<BaseT &>(Arg))) {}
-  BasicTTIImplBase &operator=(const BasicTTIImplBase &RHS) {
-    BaseT::operator=(static_cast<const BaseT &>(RHS));
-    return *this;
-  }
-  BasicTTIImplBase &operator=(BasicTTIImplBase &&RHS) {
-    BaseT::operator=(std::move(static_cast<BaseT &>(RHS)));
-    return *this;
-  }
 
   /// \name Scalar TTI Implementations
   /// @{
@@ -132,7 +126,7 @@ public:
     AM.BaseOffs = BaseOffset;
     AM.HasBaseReg = HasBaseReg;
     AM.Scale = Scale;
-    return getTLI()->isLegalAddressingMode(AM, Ty, AddrSpace);
+    return getTLI()->isLegalAddressingMode(DL, AM, Ty, AddrSpace);
   }
 
   int getScalingFactorCost(Type *Ty, GlobalValue *BaseGV, int64_t BaseOffset,
@@ -142,7 +136,7 @@ public:
     AM.BaseOffs = BaseOffset;
     AM.HasBaseReg = HasBaseReg;
     AM.Scale = Scale;
-    return getTLI()->getScalingFactorCost(AM, Ty, AddrSpace);
+    return getTLI()->getScalingFactorCost(DL, AM, Ty, AddrSpace);
   }
 
   bool isTruncateFree(Type *Ty1, Type *Ty2) {
@@ -154,7 +148,7 @@ public:
   }
 
   bool isTypeLegal(Type *Ty) {
-    EVT VT = getTLI()->getValueType(Ty);
+    EVT VT = getTLI()->getValueType(DL, Ty);
     return getTLI()->isTypeLegal(VT);
   }
 
@@ -192,7 +186,7 @@ public:
 
   bool haveFastSqrt(Type *Ty) {
     const TargetLoweringBase *TLI = getTLI();
-    EVT VT = TLI->getValueType(Ty);
+    EVT VT = TLI->getValueType(DL, Ty);
     return TLI->isTypeLegal(VT) &&
            TLI->isOperationLegalOrCustom(ISD::FSQRT, VT);
   }
@@ -282,7 +276,7 @@ public:
   /// \name Vector TTI Implementations
   /// @{
 
-  unsigned getNumberOfRegisters(bool Vector) { return 1; }
+  unsigned getNumberOfRegisters(bool Vector) { return Vector ? 0 : 1; }
 
   unsigned getRegisterBitWidth(bool Vector) { return 32; }
 
@@ -299,7 +293,7 @@ public:
     int ISD = TLI->InstructionOpcodeToISD(Opcode);
     assert(ISD && "Invalid opcode");
 
-    std::pair<unsigned, MVT> LT = TLI->getTypeLegalizationCost(Ty);
+    std::pair<unsigned, MVT> LT = TLI->getTypeLegalizationCost(DL, Ty);
 
     bool IsFloat = Ty->getScalarType()->isFloatingPointTy();
     // Assume that floating point arithmetic operations cost twice as much as
@@ -349,9 +343,8 @@ public:
     const TargetLoweringBase *TLI = getTLI();
     int ISD = TLI->InstructionOpcodeToISD(Opcode);
     assert(ISD && "Invalid opcode");
-
-    std::pair<unsigned, MVT> SrcLT = TLI->getTypeLegalizationCost(Src);
-    std::pair<unsigned, MVT> DstLT = TLI->getTypeLegalizationCost(Dst);
+    std::pair<unsigned, MVT> SrcLT = TLI->getTypeLegalizationCost(DL, Src);
+    std::pair<unsigned, MVT> DstLT = TLI->getTypeLegalizationCost(DL, Dst);
 
     // Check for NOOP conversions.
     if (SrcLT.first == DstLT.first &&
@@ -455,8 +448,7 @@ public:
       if (CondTy->isVectorTy())
         ISD = ISD::VSELECT;
     }
-
-    std::pair<unsigned, MVT> LT = TLI->getTypeLegalizationCost(ValTy);
+    std::pair<unsigned, MVT> LT = TLI->getTypeLegalizationCost(DL, ValTy);
 
     if (!(ValTy->isVectorTy() && !LT.second.isVector()) &&
         !TLI->isOperationExpand(ISD, LT.second)) {
@@ -485,7 +477,7 @@ public:
 
   unsigned getVectorInstrCost(unsigned Opcode, Type *Val, unsigned Index) {
     std::pair<unsigned, MVT> LT =
-        getTLI()->getTypeLegalizationCost(Val->getScalarType());
+        getTLI()->getTypeLegalizationCost(DL, Val->getScalarType());
 
     return LT.first;
   }
@@ -493,7 +485,7 @@ public:
   unsigned getMemoryOpCost(unsigned Opcode, Type *Src, unsigned Alignment,
                            unsigned AddressSpace) {
     assert(!Src->isVoidTy() && "Invalid type");
-    std::pair<unsigned, MVT> LT = getTLI()->getTypeLegalizationCost(Src);
+    std::pair<unsigned, MVT> LT = getTLI()->getTypeLegalizationCost(DL, Src);
 
     // Assuming that all loads of legal types cost 1.
     unsigned Cost = LT.first;
@@ -504,7 +496,7 @@ public:
       // itself. Unless the corresponding extending load or truncating store is
       // legal, then this will scalarize.
       TargetLowering::LegalizeAction LA = TargetLowering::Expand;
-      EVT MemVT = getTLI()->getValueType(Src, true);
+      EVT MemVT = getTLI()->getValueType(DL, Src, true);
       if (MemVT.isSimple() && MemVT != MVT::Other) {
         if (Opcode == Instruction::Store)
           LA = getTLI()->getTruncStoreAction(LT.second, MemVT.getSimpleVT());
@@ -700,7 +692,7 @@ public:
     }
 
     const TargetLoweringBase *TLI = getTLI();
-    std::pair<unsigned, MVT> LT = TLI->getTypeLegalizationCost(RetTy);
+    std::pair<unsigned, MVT> LT = TLI->getTypeLegalizationCost(DL, RetTy);
 
     if (TLI->isOperationLegalOrPromote(ISD, LT.second)) {
       // The operation is legal. Assume it costs 1.
@@ -771,7 +763,7 @@ public:
   }
 
   unsigned getNumberOfParts(Type *Tp) {
-    std::pair<unsigned, MVT> LT = getTLI()->getTypeLegalizationCost(Tp);
+    std::pair<unsigned, MVT> LT = getTLI()->getTypeLegalizationCost(DL, Tp);
     return LT.first;
   }
 
@@ -816,18 +808,6 @@ public:
   BasicTTIImpl(BasicTTIImpl &&Arg)
       : BaseT(std::move(static_cast<BaseT &>(Arg))), ST(std::move(Arg.ST)),
         TLI(std::move(Arg.TLI)) {}
-  BasicTTIImpl &operator=(const BasicTTIImpl &RHS) {
-    BaseT::operator=(static_cast<const BaseT &>(RHS));
-    ST = RHS.ST;
-    TLI = RHS.TLI;
-    return *this;
-  }
-  BasicTTIImpl &operator=(BasicTTIImpl &&RHS) {
-    BaseT::operator=(std::move(static_cast<BaseT &>(RHS)));
-    ST = std::move(RHS.ST);
-    TLI = std::move(RHS.TLI);
-    return *this;
-  }
 };
 
 }
diff --git a/include/llvm/CodeGen/CommandFlags.h b/include/llvm/CodeGen/CommandFlags.h
index 554511d6f4ab..4b2e0b06584c 100644
--- a/include/llvm/CodeGen/CommandFlags.h
+++ b/include/llvm/CodeGen/CommandFlags.h
@@ -206,6 +206,10 @@ cl::opt<std::string> StartAfter("start-after",
                           cl::value_desc("pass-name"),
                           cl::init(""));
 
+cl::opt<std::string>
+    RunPass("run-pass", cl::desc("Run compiler only for one specific pass"),
+            cl::value_desc("pass-name"), cl::init(""));
+
 cl::opt<bool> DataSections("data-sections",
                            cl::desc("Emit data into separate sections"),
                            cl::init(false));
diff --git a/include/llvm/CodeGen/ISDOpcodes.h b/include/llvm/CodeGen/ISDOpcodes.h
index c7237fd55b27..fa44301a2d4a 100644
--- a/include/llvm/CodeGen/ISDOpcodes.h
+++ b/include/llvm/CodeGen/ISDOpcodes.h
@@ -72,10 +72,13 @@ namespace ISD {
     /// the parent's frame or return address, and so on.
     FRAMEADDR, RETURNADDR,
 
-    /// FRAME_ALLOC_RECOVER - Represents the llvm.framerecover
-    /// intrinsic. Materializes the offset from the frame pointer of another
-    /// function to the result of llvm.frameallocate.
-    FRAME_ALLOC_RECOVER,
+    /// LOCAL_RECOVER - Represents the llvm.localrecover intrinsic.
+    /// Materializes the offset from the local object pointer of another
+    /// function to a particular local object passed to llvm.localescape. The
+    /// operand is the MCSymbol label used to represent this offset, since
+    /// typically the offset is not known until after code generation of the
+    /// parent.
+    LOCAL_RECOVER,
 
     /// READ_REGISTER, WRITE_REGISTER - This node represents llvm.register on
     /// the DAG, which implements the named register global variables extension.
@@ -725,7 +728,7 @@ namespace ISD {
   /// which do not reference a specific memory location should be less than
   /// this value. Those that do must not be less than this value, and can
   /// be used with SelectionDAG::getMemIntrinsicNode.
-  static const int FIRST_TARGET_MEMORY_OPCODE = BUILTIN_OP_END+200;
+  static const int FIRST_TARGET_MEMORY_OPCODE = BUILTIN_OP_END+300;
 
   //===--------------------------------------------------------------------===//
   /// MemIndexedMode enum - This enum defines the load / store indexed
diff --git a/include/llvm/CodeGen/LiveIntervalUnion.h b/include/llvm/CodeGen/LiveIntervalUnion.h
index 967f0cbac719..f0f1637dc92d 100644
--- a/include/llvm/CodeGen/LiveIntervalUnion.h
+++ b/include/llvm/CodeGen/LiveIntervalUnion.h
@@ -203,6 +203,11 @@ public:
       assert(idx <  Size && "idx out of bounds");
       return LIUs[idx];
     }
+
+    const LiveIntervalUnion& operator[](unsigned Idx) const {
+      assert(Idx < Size && "Idx out of bounds");
+      return LIUs[Idx];
+    }
   };
 };
 
diff --git a/include/llvm/CodeGen/LiveRegMatrix.h b/include/llvm/CodeGen/LiveRegMatrix.h
index 878b4d9836f2..e169058ca563 100644
--- a/include/llvm/CodeGen/LiveRegMatrix.h
+++ b/include/llvm/CodeGen/LiveRegMatrix.h
@@ -32,13 +32,11 @@ namespace llvm {
 
 class LiveInterval;
 class LiveIntervalAnalysis;
-class MachineRegisterInfo;
 class TargetRegisterInfo;
 class VirtRegMap;
 
 class LiveRegMatrix : public MachineFunctionPass {
   const TargetRegisterInfo *TRI;
-  MachineRegisterInfo *MRI;
   LiveIntervals *LIS;
   VirtRegMap *VRM;
 
@@ -114,6 +112,9 @@ public:
   /// the assignment and updates VirtRegMap accordingly.
   void unassign(LiveInterval &VirtReg);
 
+  /// Returns true if the given \p PhysReg has any live intervals assigned.
+  bool isPhysRegUsed(unsigned PhysReg) const;
+
   //===--------------------------------------------------------------------===//
   // Low-level interface.
   //===--------------------------------------------------------------------===//
diff --git a/include/llvm/CodeGen/MIRYamlMapping.h b/include/llvm/CodeGen/MIRYamlMapping.h
index a6ffeb382978..9798e5cef645 100644
--- a/include/llvm/CodeGen/MIRYamlMapping.h
+++ b/include/llvm/CodeGen/MIRYamlMapping.h
@@ -81,15 +81,30 @@ LLVM_YAML_IS_FLOW_SEQUENCE_VECTOR(llvm::yaml::FlowStringValue)
 namespace llvm {
 namespace yaml {
 
+struct VirtualRegisterDefinition {
+  unsigned ID;
+  StringValue Class;
+  // TODO: Serialize the virtual register hints.
+};
+
+template <> struct MappingTraits<VirtualRegisterDefinition> {
+  static void mapping(IO &YamlIO, VirtualRegisterDefinition &Reg) {
+    YamlIO.mapRequired("id", Reg.ID);
+    YamlIO.mapRequired("class", Reg.Class);
+  }
+
+  static const bool flow = true;
+};
+
 struct MachineBasicBlock {
   unsigned ID;
-  std::string Name;
+  StringValue Name;
   unsigned Alignment = 0;
   bool IsLandingPad = false;
   bool AddressTaken = false;
-  // TODO: Serialize the successor weights and liveins.
+  // TODO: Serialize the successor weights.
   std::vector<FlowStringValue> Successors;
-
+  std::vector<FlowStringValue> LiveIns;
   std::vector<StringValue> Instructions;
 };
 
@@ -97,23 +112,153 @@ template <> struct MappingTraits<MachineBasicBlock> {
   static void mapping(IO &YamlIO, MachineBasicBlock &MBB) {
     YamlIO.mapRequired("id", MBB.ID);
     YamlIO.mapOptional("name", MBB.Name,
-                       std::string()); // Don't print out an empty name.
+                       StringValue()); // Don't print out an empty name.
     YamlIO.mapOptional("alignment", MBB.Alignment);
     YamlIO.mapOptional("isLandingPad", MBB.IsLandingPad);
     YamlIO.mapOptional("addressTaken", MBB.AddressTaken);
     YamlIO.mapOptional("successors", MBB.Successors);
+    YamlIO.mapOptional("liveins", MBB.LiveIns);
     YamlIO.mapOptional("instructions", MBB.Instructions);
   }
 };
 
+/// Serializable representation of stack object from the MachineFrameInfo class.
+///
+/// The flags 'isImmutable' and 'isAliased' aren't serialized, as they are
+/// determined by the object's type and frame information flags.
+/// Dead stack objects aren't serialized.
+///
+/// TODO: Determine isPreallocated flag by mapping between objects and local
+/// objects (Serialize local objects).
+struct MachineStackObject {
+  enum ObjectType { DefaultType, SpillSlot, VariableSized };
+  // TODO: Serialize LLVM alloca reference.
+  unsigned ID;
+  ObjectType Type = DefaultType;
+  int64_t Offset = 0;
+  uint64_t Size = 0;
+  unsigned Alignment = 0;
+};
+
+template <> struct ScalarEnumerationTraits<MachineStackObject::ObjectType> {
+  static void enumeration(yaml::IO &IO, MachineStackObject::ObjectType &Type) {
+    IO.enumCase(Type, "default", MachineStackObject::DefaultType);
+    IO.enumCase(Type, "spill-slot", MachineStackObject::SpillSlot);
+    IO.enumCase(Type, "variable-sized", MachineStackObject::VariableSized);
+  }
+};
+
+template <> struct MappingTraits<MachineStackObject> {
+  static void mapping(yaml::IO &YamlIO, MachineStackObject &Object) {
+    YamlIO.mapRequired("id", Object.ID);
+    YamlIO.mapOptional(
+        "type", Object.Type,
+        MachineStackObject::DefaultType); // Don't print the default type.
+    YamlIO.mapOptional("offset", Object.Offset);
+    if (Object.Type != MachineStackObject::VariableSized)
+      YamlIO.mapRequired("size", Object.Size);
+    YamlIO.mapOptional("alignment", Object.Alignment);
+  }
+
+  static const bool flow = true;
+};
+
+/// Serializable representation of the fixed stack object from the
+/// MachineFrameInfo class.
+struct FixedMachineStackObject {
+  enum ObjectType { DefaultType, SpillSlot };
+  unsigned ID;
+  ObjectType Type = DefaultType;
+  int64_t Offset = 0;
+  uint64_t Size = 0;
+  unsigned Alignment = 0;
+  bool IsImmutable = false;
+  bool IsAliased = false;
+};
+
+template <>
+struct ScalarEnumerationTraits<FixedMachineStackObject::ObjectType> {
+  static void enumeration(yaml::IO &IO,
+                          FixedMachineStackObject::ObjectType &Type) {
+    IO.enumCase(Type, "default", FixedMachineStackObject::DefaultType);
+    IO.enumCase(Type, "spill-slot", FixedMachineStackObject::SpillSlot);
+  }
+};
+
+template <> struct MappingTraits<FixedMachineStackObject> {
+  static void mapping(yaml::IO &YamlIO, FixedMachineStackObject &Object) {
+    YamlIO.mapRequired("id", Object.ID);
+    YamlIO.mapOptional(
+        "type", Object.Type,
+        FixedMachineStackObject::DefaultType); // Don't print the default type.
+    YamlIO.mapOptional("offset", Object.Offset);
+    YamlIO.mapOptional("size", Object.Size);
+    YamlIO.mapOptional("alignment", Object.Alignment);
+    if (Object.Type != FixedMachineStackObject::SpillSlot) {
+      YamlIO.mapOptional("isImmutable", Object.IsImmutable);
+      YamlIO.mapOptional("isAliased", Object.IsAliased);
+    }
+  }
+
+  static const bool flow = true;
+};
+
 } // end namespace yaml
 } // end namespace llvm
 
+LLVM_YAML_IS_SEQUENCE_VECTOR(llvm::yaml::VirtualRegisterDefinition)
 LLVM_YAML_IS_SEQUENCE_VECTOR(llvm::yaml::MachineBasicBlock)
+LLVM_YAML_IS_SEQUENCE_VECTOR(llvm::yaml::MachineStackObject)
+LLVM_YAML_IS_SEQUENCE_VECTOR(llvm::yaml::FixedMachineStackObject)
 
 namespace llvm {
 namespace yaml {
 
+/// Serializable representation of MachineFrameInfo.
+///
+/// Doesn't serialize attributes like 'StackAlignment', 'IsStackRealignable' and
+/// 'RealignOption' as they are determined by the target and LLVM function
+/// attributes.
+/// It also doesn't serialize attributes like 'NumFixedObject' and
+/// 'HasVarSizedObjects' as they are determined by the frame objects themselves.
+struct MachineFrameInfo {
+  bool IsFrameAddressTaken = false;
+  bool IsReturnAddressTaken = false;
+  bool HasStackMap = false;
+  bool HasPatchPoint = false;
+  uint64_t StackSize = 0;
+  int OffsetAdjustment = 0;
+  unsigned MaxAlignment = 0;
+  bool AdjustsStack = false;
+  bool HasCalls = false;
+  // TODO: Serialize StackProtectorIdx and FunctionContextIdx
+  unsigned MaxCallFrameSize = 0;
+  // TODO: Serialize callee saved info.
+  // TODO: Serialize local frame objects.
+  bool HasOpaqueSPAdjustment = false;
+  bool HasVAStart = false;
+  bool HasMustTailInVarArgFunc = false;
+  // TODO: Serialize save and restore MBB references.
+};
+
+template <> struct MappingTraits<MachineFrameInfo> {
+  static void mapping(IO &YamlIO, MachineFrameInfo &MFI) {
+    YamlIO.mapOptional("isFrameAddressTaken", MFI.IsFrameAddressTaken);
+    YamlIO.mapOptional("isReturnAddressTaken", MFI.IsReturnAddressTaken);
+    YamlIO.mapOptional("hasStackMap", MFI.HasStackMap);
+    YamlIO.mapOptional("hasPatchPoint", MFI.HasPatchPoint);
+    YamlIO.mapOptional("stackSize", MFI.StackSize);
+    YamlIO.mapOptional("offsetAdjustment", MFI.OffsetAdjustment);
+    YamlIO.mapOptional("maxAlignment", MFI.MaxAlignment);
+    YamlIO.mapOptional("adjustsStack", MFI.AdjustsStack);
+    YamlIO.mapOptional("hasCalls", MFI.HasCalls);
+    YamlIO.mapOptional("maxCallFrameSize", MFI.MaxCallFrameSize);
+    YamlIO.mapOptional("hasOpaqueSPAdjustment", MFI.HasOpaqueSPAdjustment);
+    YamlIO.mapOptional("hasVAStart", MFI.HasVAStart);
+    YamlIO.mapOptional("hasMustTailInVarArgFunc", MFI.HasMustTailInVarArgFunc);
+  }
+};
+
 struct MachineFunction {
   StringRef Name;
   unsigned Alignment = 0;
@@ -123,9 +268,13 @@ struct MachineFunction {
   bool IsSSA = false;
   bool TracksRegLiveness = false;
   bool TracksSubRegLiveness = false;
-  // TODO: Serialize virtual register definitions.
+  std::vector<VirtualRegisterDefinition> VirtualRegisters;
   // TODO: Serialize the various register masks.
   // TODO: Serialize live in registers.
+  // Frame information
+  MachineFrameInfo FrameInfo;
+  std::vector<FixedMachineStackObject> FixedStackObjects;
+  std::vector<MachineStackObject> StackObjects;
 
   std::vector<MachineBasicBlock> BasicBlocks;
 };
@@ -139,6 +288,10 @@ template <> struct MappingTraits<MachineFunction> {
     YamlIO.mapOptional("isSSA", MF.IsSSA);
     YamlIO.mapOptional("tracksRegLiveness", MF.TracksRegLiveness);
     YamlIO.mapOptional("tracksSubRegLiveness", MF.TracksSubRegLiveness);
+    YamlIO.mapOptional("registers", MF.VirtualRegisters);
+    YamlIO.mapOptional("frameInfo", MF.FrameInfo);
+    YamlIO.mapOptional("fixedStack", MF.FixedStackObjects);
+    YamlIO.mapOptional("stack", MF.StackObjects);
     YamlIO.mapOptional("body", MF.BasicBlocks);
   }
 };
diff --git a/include/llvm/CodeGen/MachineConstantPool.h b/include/llvm/CodeGen/MachineConstantPool.h
index c619afb83333..628400322f60 100644
--- a/include/llvm/CodeGen/MachineConstantPool.h
+++ b/include/llvm/CodeGen/MachineConstantPool.h
@@ -135,17 +135,18 @@ public:
 /// address of the function constant pool values.
 /// @brief The machine constant pool.
 class MachineConstantPool {
-  const TargetMachine &TM;      ///< The target machine.
   unsigned PoolAlignment;       ///< The alignment for the pool.
   std::vector<MachineConstantPoolEntry> Constants; ///< The pool of constants.
   /// MachineConstantPoolValues that use an existing MachineConstantPoolEntry.
   DenseSet<MachineConstantPoolValue*> MachineCPVsSharingEntries;
+  const DataLayout &DL;
+
+  const DataLayout &getDataLayout() const { return DL; }
 
-  const DataLayout *getDataLayout() const;
 public:
   /// @brief The only constructor.
-  explicit MachineConstantPool(const TargetMachine &TM)
-    : TM(TM), PoolAlignment(1) {}
+  explicit MachineConstantPool(const DataLayout &DL)
+      : PoolAlignment(1), DL(DL) {}
   ~MachineConstantPool();
     
   /// getConstantPoolAlignment - Return the alignment required by
diff --git a/include/llvm/CodeGen/MachineDominators.h b/include/llvm/CodeGen/MachineDominators.h
index 4428fa618fb0..735dd069cf7f 100644
--- a/include/llvm/CodeGen/MachineDominators.h
+++ b/include/llvm/CodeGen/MachineDominators.h
@@ -29,8 +29,8 @@ inline void DominatorTreeBase<MachineBasicBlock>::addRoot(MachineBasicBlock* MBB
   this->Roots.push_back(MBB);
 }
 
-EXTERN_TEMPLATE_INSTANTIATION(class DomTreeNodeBase<MachineBasicBlock>);
-EXTERN_TEMPLATE_INSTANTIATION(class DominatorTreeBase<MachineBasicBlock>);
+extern template class DomTreeNodeBase<MachineBasicBlock>;
+extern template class DominatorTreeBase<MachineBasicBlock>;
 
 typedef DomTreeNodeBase<MachineBasicBlock> MachineDomTreeNode;
 
diff --git a/include/llvm/CodeGen/MachineFrameInfo.h b/include/llvm/CodeGen/MachineFrameInfo.h
index 0f5a4b1b09ec..cbc4e66ccc46 100644
--- a/include/llvm/CodeGen/MachineFrameInfo.h
+++ b/include/llvm/CodeGen/MachineFrameInfo.h
@@ -229,9 +229,9 @@ class MachineFrameInfo {
   /// Whether the "realign-stack" option is on.
   bool RealignOption;
 
-  /// True if the function includes inline assembly that adjusts the stack
-  /// pointer.
-  bool HasInlineAsmWithSPAdjust;
+  /// True if the function dynamically adjusts the stack pointer through some
+  /// opaque mechanism like inline assembly or Win32 EH.
+  bool HasOpaqueSPAdjustment;
 
   /// True if the function contains a call to the llvm.vastart intrinsic.
   bool HasVAStart;
@@ -269,7 +269,7 @@ public:
     LocalFrameSize = 0;
     LocalFrameMaxAlign = 0;
     UseLocalStackAllocationBlock = false;
-    HasInlineAsmWithSPAdjust = false;
+    HasOpaqueSPAdjustment = false;
     HasVAStart = false;
     HasMustTailInVarArgFunc = false;
     Save = nullptr;
@@ -468,9 +468,9 @@ public:
   bool hasCalls() const { return HasCalls; }
   void setHasCalls(bool V) { HasCalls = V; }
 
-  /// Returns true if the function contains any stack-adjusting inline assembly.
-  bool hasInlineAsmWithSPAdjust() const { return HasInlineAsmWithSPAdjust; }
-  void setHasInlineAsmWithSPAdjust(bool B) { HasInlineAsmWithSPAdjust = B; }
+  /// Returns true if the function contains opaque dynamic stack adjustments.
+  bool hasOpaqueSPAdjustment() const { return HasOpaqueSPAdjustment; }
+  void setHasOpaqueSPAdjustment(bool B) { HasOpaqueSPAdjustment = B; }
 
   /// Returns true if the function calls the llvm.va_start intrinsic.
   bool hasVAStart() const { return HasVAStart; }
@@ -541,6 +541,14 @@ public:
     return Objects[ObjectIdx+NumFixedObjects].Size == ~0ULL;
   }
 
+  /// Returns true if the specified index corresponds to a variable sized
+  /// object.
+  bool isVariableSizedObjectIndex(int ObjectIdx) const {
+    assert(unsigned(ObjectIdx + NumFixedObjects) < Objects.size() &&
+           "Invalid Object Idx!");
+    return Objects[ObjectIdx + NumFixedObjects].Size == 0;
+  }
+
   /// Create a new statically sized stack object, returning
   /// a nonnegative identifier to represent it.
   int CreateStackObject(uint64_t Size, unsigned Alignment, bool isSS,
diff --git a/include/llvm/CodeGen/MachineFunction.h b/include/llvm/CodeGen/MachineFunction.h
index 94610cabf566..c15ee1c006cd 100644
--- a/include/llvm/CodeGen/MachineFunction.h
+++ b/include/llvm/CodeGen/MachineFunction.h
@@ -155,6 +155,9 @@ public:
   MachineModuleInfo &getMMI() const { return MMI; }
   MCContext &getContext() const { return Ctx; }
 
+  /// Return the DataLayout attached to the Module associated to this MF.
+  const DataLayout &getDataLayout() const;
+
   /// getFunction - Return the LLVM function that this machine code represents
   ///
   const Function *getFunction() const { return Fn; }
diff --git a/include/llvm/CodeGen/MachineLoopInfo.h b/include/llvm/CodeGen/MachineLoopInfo.h
index 438ef2e37255..4868b7363f82 100644
--- a/include/llvm/CodeGen/MachineLoopInfo.h
+++ b/include/llvm/CodeGen/MachineLoopInfo.h
@@ -37,10 +37,8 @@
 namespace llvm {
 
 // Implementation in LoopInfoImpl.h
-#ifdef __GNUC__
 class MachineLoop;
-__extension__ extern template class LoopBase<MachineBasicBlock, MachineLoop>;
-#endif
+extern template class LoopBase<MachineBasicBlock, MachineLoop>;
 
 class MachineLoop : public LoopBase<MachineBasicBlock, MachineLoop> {
 public:
@@ -65,10 +63,7 @@ private:
 };
 
 // Implementation in LoopInfoImpl.h
-#ifdef __GNUC__
-__extension__ extern template
-class LoopInfoBase<MachineBasicBlock, MachineLoop>;
-#endif
+extern template class LoopInfoBase<MachineBasicBlock, MachineLoop>;
 
 class MachineLoopInfo : public MachineFunctionPass {
   LoopInfoBase<MachineBasicBlock, MachineLoop> LI;
diff --git a/include/llvm/CodeGen/MachineModuleInfo.h b/include/llvm/CodeGen/MachineModuleInfo.h
index ccaa83a238a6..4cdfe2463c99 100644
--- a/include/llvm/CodeGen/MachineModuleInfo.h
+++ b/include/llvm/CodeGen/MachineModuleInfo.h
@@ -320,6 +320,7 @@ public:
   /// information.
   void addPersonality(MachineBasicBlock *LandingPad,
                       const Function *Personality);
+  void addPersonality(const Function *Personality);
 
   void addWinEHState(MachineBasicBlock *LandingPad, int State);
 
diff --git a/include/llvm/CodeGen/MachineRegionInfo.h b/include/llvm/CodeGen/MachineRegionInfo.h
index cf49c297c288..df9823f741dc 100644
--- a/include/llvm/CodeGen/MachineRegionInfo.h
+++ b/include/llvm/CodeGen/MachineRegionInfo.h
@@ -172,10 +172,9 @@ template <> struct GraphTraits<MachineRegionInfoPass*>
   }
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class RegionBase<RegionTraits<MachineFunction>>);
-EXTERN_TEMPLATE_INSTANTIATION(class RegionNodeBase<RegionTraits<MachineFunction>>);
-EXTERN_TEMPLATE_INSTANTIATION(class RegionInfoBase<RegionTraits<MachineFunction>>);
-
+extern template class RegionBase<RegionTraits<MachineFunction>>;
+extern template class RegionNodeBase<RegionTraits<MachineFunction>>;
+extern template class RegionInfoBase<RegionTraits<MachineFunction>>;
 }
 
 #endif
diff --git a/include/llvm/CodeGen/MachineRegisterInfo.h b/include/llvm/CodeGen/MachineRegisterInfo.h
index e5b837aeea28..67583be616c3 100644
--- a/include/llvm/CodeGen/MachineRegisterInfo.h
+++ b/include/llvm/CodeGen/MachineRegisterInfo.h
@@ -95,20 +95,8 @@ private:
     return MO->Contents.Reg.Next;
   }
 
-  /// UsedRegUnits - This is a bit vector that is computed and set by the
-  /// register allocator, and must be kept up to date by passes that run after
-  /// register allocation (though most don't modify this).  This is used
-  /// so that the code generator knows which callee save registers to save and
-  /// for other target specific uses.
-  /// This vector has bits set for register units that are modified in the
-  /// current function. It doesn't include registers clobbered by function
-  /// calls with register mask operands.
-  BitVector UsedRegUnits;
-
   /// UsedPhysRegMask - Additional used physregs including aliases.
   /// This bit vector represents all the registers clobbered by function calls.
-  /// It can model things that UsedRegUnits can't, such as function calls that
-  /// clobber ymm7 but preserve the low half in xmm7.
   BitVector UsedPhysRegMask;
 
   /// ReservedRegs - This is a bit vector of reserved registers.  The target
@@ -647,40 +635,11 @@ public:
   /// deleted during LiveDebugVariables analysis.
   void markUsesInDebugValueAsUndef(unsigned Reg) const;
 
-  //===--------------------------------------------------------------------===//
-  // Physical Register Use Info
-  //===--------------------------------------------------------------------===//
-
-  /// isPhysRegUsed - Return true if the specified register is used in this
-  /// function. Also check for clobbered aliases and registers clobbered by
-  /// function calls with register mask operands.
-  ///
-  /// This only works after register allocation. It is primarily used by
-  /// PrologEpilogInserter to determine which callee-saved registers need
-  /// spilling.
-  bool isPhysRegUsed(unsigned Reg) const {
-    if (UsedPhysRegMask.test(Reg))
-      return true;
-    for (MCRegUnitIterator Units(Reg, getTargetRegisterInfo());
-         Units.isValid(); ++Units)
-      if (UsedRegUnits.test(*Units))
-        return true;
-    return false;
-  }
-
-  /// Mark the specified register unit as used in this function.
-  /// This should only be called during and after register allocation.
-  void setRegUnitUsed(unsigned RegUnit) {
-    UsedRegUnits.set(RegUnit);
-  }
-
-  /// setPhysRegUsed - Mark the specified register used in this function.
-  /// This should only be called during and after register allocation.
-  void setPhysRegUsed(unsigned Reg) {
-    for (MCRegUnitIterator Units(Reg, getTargetRegisterInfo());
-         Units.isValid(); ++Units)
-      UsedRegUnits.set(*Units);
-  }
+  /// Return true if the specified register is modified in this function.
+  /// This checks that no defining machine operands exist for the register or
+  /// any of its aliases. Definitions found on functions marked noreturn are
+  /// ignored.
+  bool isPhysRegModified(unsigned PhysReg) const;
 
   /// addPhysRegsUsedFromRegMask - Mark any registers not in RegMask as used.
   /// This corresponds to the bit mask attached to register mask operands.
@@ -688,16 +647,6 @@ public:
     UsedPhysRegMask.setBitsNotInMask(RegMask);
   }
 
-  /// setPhysRegUnused - Mark the specified register unused in this function.
-  /// This should only be called during and after register allocation.
-  void setPhysRegUnused(unsigned Reg) {
-    UsedPhysRegMask.reset(Reg);
-    for (MCRegUnitIterator Units(Reg, getTargetRegisterInfo());
-         Units.isValid(); ++Units)
-      UsedRegUnits.reset(*Units);
-  }
-
-
   //===--------------------------------------------------------------------===//
   // Reserved Register Info
   //===--------------------------------------------------------------------===//
diff --git a/include/llvm/CodeGen/Passes.h b/include/llvm/CodeGen/Passes.h
index 538c995a7b44..5d8292174476 100644
--- a/include/llvm/CodeGen/Passes.h
+++ b/include/llvm/CodeGen/Passes.h
@@ -101,7 +101,7 @@ public:
 
 private:
   PassManagerBase *PM;
-  AnalysisID StartAfter;
+  AnalysisID StartBefore, StartAfter;
   AnalysisID StopAfter;
   bool Started;
   bool Stopped;
@@ -142,16 +142,24 @@ public:
 
   CodeGenOpt::Level getOptLevel() const { return TM->getOptLevel(); }
 
-  /// setStartStopPasses - Set the StartAfter and StopAfter passes to allow
-  /// running only a portion of the normal code-gen pass sequence.  If the
-  /// Start pass ID is zero, then compilation will begin at the normal point;
-  /// otherwise, clear the Started flag to indicate that passes should not be
-  /// added until the starting pass is seen.  If the Stop pass ID is zero,
-  /// then compilation will continue to the end.
-  void setStartStopPasses(AnalysisID Start, AnalysisID Stop) {
-    StartAfter = Start;
-    StopAfter = Stop;
-    Started = (StartAfter == nullptr);
+  /// Set the StartAfter, StartBefore and StopAfter passes to allow running only
+  /// a portion of the normal code-gen pass sequence.
+  ///
+  /// If the StartAfter and StartBefore pass ID is zero, then compilation will
+  /// begin at the normal point; otherwise, clear the Started flag to indicate
+  /// that passes should not be added until the starting pass is seen.  If the
+  /// Stop pass ID is zero, then compilation will continue to the end.
+  ///
+  /// This function expects that at least one of the StartAfter or the
+  /// StartBefore pass IDs is null.
+  void setStartStopPasses(AnalysisID StartBefore, AnalysisID StartAfter,
+                          AnalysisID StopAfter) {
+    if (StartAfter)
+      assert(!StartBefore && "Start after and start before passes are given");
+    this->StartBefore = StartBefore;
+    this->StartAfter = StartAfter;
+    this->StopAfter = StopAfter;
+    Started = (StartAfter == nullptr) && (StartBefore == nullptr);
   }
 
   void setDisableVerify(bool Disable) { setOpt(DisableVerify, Disable); }
@@ -597,7 +605,7 @@ namespace llvm {
   /// createSjLjEHPreparePass - This pass adapts exception handling code to use
   /// the GCC-style builtin setjmp/longjmp (sjlj) to handling EH control flow.
   ///
-  FunctionPass *createSjLjEHPreparePass(const TargetMachine *TM);
+  FunctionPass *createSjLjEHPreparePass();
 
   /// LocalStackSlotAllocation - This pass assigns local frame indices to stack
   /// slots relative to one another and allocates base registers to access them
diff --git a/include/llvm/CodeGen/RegisterPressure.h b/include/llvm/CodeGen/RegisterPressure.h
index fcb6feed68ca..9d8843d1d74a 100644
--- a/include/llvm/CodeGen/RegisterPressure.h
+++ b/include/llvm/CodeGen/RegisterPressure.h
@@ -135,6 +135,8 @@ public:
 
   void addPressureChange(unsigned RegUnit, bool IsDec,
                          const MachineRegisterInfo *MRI);
+
+  LLVM_DUMP_METHOD void dump(const TargetRegisterInfo &TRI) const;
 };
 
 /// Array of PressureDiffs.
diff --git a/include/llvm/CodeGen/SelectionDAG.h b/include/llvm/CodeGen/SelectionDAG.h
index c2b1243ee26e..1ee92380e690 100644
--- a/include/llvm/CodeGen/SelectionDAG.h
+++ b/include/llvm/CodeGen/SelectionDAG.h
@@ -281,6 +281,7 @@ public:
   void clear();
 
   MachineFunction &getMachineFunction() const { return *MF; }
+  const DataLayout &getDataLayout() const { return MF->getDataLayout(); }
   const TargetMachine &getTarget() const { return TM; }
   const TargetSubtargetInfo &getSubtarget() const { return MF->getSubtarget(); }
   const TargetLowering &getTargetLoweringInfo() const { return *TLI; }
@@ -322,6 +323,14 @@ public:
     return AllNodes.size();
   }
 
+  iterator_range<allnodes_iterator> allnodes() {
+    return iterator_range<allnodes_iterator>(allnodes_begin(), allnodes_end());
+  }
+  iterator_range<allnodes_const_iterator> allnodes() const {
+    return iterator_range<allnodes_const_iterator>(allnodes_begin(),
+                                                   allnodes_end());
+  }
+
   /// Return the root tag of the SelectionDAG.
   const SDValue &getRoot() const { return Root; }
 
diff --git a/include/llvm/CodeGen/SelectionDAGNodes.h b/include/llvm/CodeGen/SelectionDAGNodes.h
index 619119096d20..4821d1aae9e5 100644
--- a/include/llvm/CodeGen/SelectionDAGNodes.h
+++ b/include/llvm/CodeGen/SelectionDAGNodes.h
@@ -140,7 +140,7 @@ public:
   }
 
   // Return true if this node is an operand of N.
-  bool isOperandOf(SDNode *N) const;
+  bool isOperandOf(const SDNode *N) const;
 
   /// Return the ValueType of the referenced return value.
   inline EVT getValueType() const;
@@ -357,9 +357,6 @@ private:
   /// The number of entries in the Operand/Value list.
   unsigned short NumOperands, NumValues;
 
-  /// Source line information.
-  DebugLoc debugLoc;
-
   // The ordering of the SDNodes. It roughly corresponds to the ordering of the
   // original LLVM instructions.
   // This is used for turning off scheduling, because we'll forgo
@@ -367,6 +364,9 @@ private:
   // this ordering.
   unsigned IROrder;
 
+  /// Source line information.
+  DebugLoc debugLoc;
+
   /// Return a pointer to the specified value type.
   static const EVT *getValueTypeList(EVT VT);
 
@@ -532,10 +532,10 @@ public:
   bool hasAnyUseOfValue(unsigned Value) const;
 
   /// Return true if this node is the only use of N.
-  bool isOnlyUserOf(SDNode *N) const;
+  bool isOnlyUserOf(const SDNode *N) const;
 
   /// Return true if this node is an operand of N.
-  bool isOperandOf(SDNode *N) const;
+  bool isOperandOf(const SDNode *N) const;
 
   /// Return true if this node is a predecessor of N.
   /// NOTE: Implemented on top of hasPredecessor and every bit as
@@ -732,7 +732,7 @@ protected:
         SubclassData(0), NodeId(-1),
         OperandList(Ops.size() ? new SDUse[Ops.size()] : nullptr),
         ValueList(VTs.VTs), UseList(nullptr), NumOperands(Ops.size()),
-        NumValues(VTs.NumVTs), debugLoc(std::move(dl)), IROrder(Order) {
+        NumValues(VTs.NumVTs), IROrder(Order), debugLoc(std::move(dl)) {
     assert(debugLoc.hasTrivialDestructor() && "Expected trivial destructor");
     assert(NumOperands == Ops.size() &&
            "NumOperands wasn't wide enough for its operands!");
@@ -752,7 +752,7 @@ protected:
       : NodeType(Opc), OperandsNeedDelete(false), HasDebugValue(false),
         SubclassData(0), NodeId(-1), OperandList(nullptr), ValueList(VTs.VTs),
         UseList(nullptr), NumOperands(0), NumValues(VTs.NumVTs),
-        debugLoc(std::move(dl)), IROrder(Order) {
+        IROrder(Order), debugLoc(std::move(dl)) {
     assert(debugLoc.hasTrivialDestructor() && "Expected trivial destructor");
     assert(NumValues == VTs.NumVTs &&
            "NumValues wasn't wide enough for its operands!");
diff --git a/include/llvm/CodeGen/StackMaps.h b/include/llvm/CodeGen/StackMaps.h
index 46a773f74aac..fdc1a9143ed2 100644
--- a/include/llvm/CodeGen/StackMaps.h
+++ b/include/llvm/CodeGen/StackMaps.h
@@ -1,5 +1,4 @@
 //===------------------- StackMaps.h - StackMaps ----------------*- C++ -*-===//
-
 //
 //                     The LLVM Compiler Infrastructure
 //
@@ -42,10 +41,12 @@ class PatchPointOpers {
 public:
   /// Enumerate the meta operands.
   enum { IDPos, NBytesPos, TargetPos, NArgPos, CCPos, MetaEnd };
+
 private:
   const MachineInstr *MI;
   bool HasDef;
   bool IsAnyReg;
+
 public:
   explicit PatchPointOpers(const MachineInstr *MI);
 
@@ -66,8 +67,8 @@ public:
   /// Get the operand index of the variable list of non-argument operands.
   /// These hold the "live state".
   unsigned getVarIdx() const {
-    return getMetaIdx() + MetaEnd
-      + MI->getOperand(getMetaIdx(NArgPos)).getImm();
+    return getMetaIdx() + MetaEnd +
+           MI->getOperand(getMetaIdx(NArgPos)).getImm();
   }
 
   /// Get the index at which stack map locations will be recorded.
@@ -98,15 +99,10 @@ private:
 
   // These values are relative offests from the start of the statepoint meta
   // arguments (i.e. the end of the call arguments).
-  enum {
-    CCOffset = 1,
-    FlagsOffset = 3,
-    NumVMSArgsOffset = 5
-  };
+  enum { CCOffset = 1, FlagsOffset = 3, NumVMSArgsOffset = 5 };
 
 public:
-  explicit StatepointOpers(const MachineInstr *MI):
-    MI(MI) { }
+  explicit StatepointOpers(const MachineInstr *MI) : MI(MI) {}
 
   /// Get starting index of non call related arguments
   /// (calling convention, statepoint flags, vm state and gc state).
@@ -134,31 +130,32 @@ private:
 class StackMaps {
 public:
   struct Location {
-    enum LocationType { Unprocessed, Register, Direct, Indirect, Constant,
-                        ConstantIndex };
-    LocationType LocType;
+    enum LocationType {
+      Unprocessed,
+      Register,
+      Direct,
+      Indirect,
+      Constant,
+      ConstantIndex
+    };
+    LocationType Type;
     unsigned Size;
     unsigned Reg;
     int64_t Offset;
-    Location() : LocType(Unprocessed), Size(0), Reg(0), Offset(0) {}
-    Location(LocationType LocType, unsigned Size, unsigned Reg, int64_t Offset)
-      : LocType(LocType), Size(Size), Reg(Reg), Offset(Offset) {}
+    Location() : Type(Unprocessed), Size(0), Reg(0), Offset(0) {}
+    Location(LocationType Type, unsigned Size, unsigned Reg, int64_t Offset)
+        : Type(Type), Size(Size), Reg(Reg), Offset(Offset) {}
   };
 
   struct LiveOutReg {
     unsigned short Reg;
-    unsigned short RegNo;
+    unsigned short DwarfRegNum;
     unsigned short Size;
 
-    LiveOutReg() : Reg(0), RegNo(0), Size(0) {}
-    LiveOutReg(unsigned short Reg, unsigned short RegNo, unsigned short Size)
-      : Reg(Reg), RegNo(RegNo), Size(Size) {}
-
-    void MarkInvalid() { Reg = 0; }
-
-    // Only sort by the dwarf register number.
-    bool operator< (const LiveOutReg &LO) const { return RegNo < LO.RegNo; }
-    static bool IsInvalid(const LiveOutReg &LO) { return LO.Reg == 0; }
+    LiveOutReg() : Reg(0), DwarfRegNum(0), Size(0) {}
+    LiveOutReg(unsigned short Reg, unsigned short DwarfRegNum,
+               unsigned short Size)
+        : Reg(Reg), DwarfRegNum(DwarfRegNum), Size(Size) {}
   };
 
   // OpTypes are used to encode information about the following logical
@@ -205,8 +202,8 @@ private:
     CallsiteInfo() : CSOffsetExpr(nullptr), ID(0) {}
     CallsiteInfo(const MCExpr *CSOffsetExpr, uint64_t ID,
                  LocationVec &&Locations, LiveOutVec &&LiveOuts)
-      : CSOffsetExpr(CSOffsetExpr), ID(ID), Locations(std::move(Locations)),
-        LiveOuts(std::move(LiveOuts)) {}
+        : CSOffsetExpr(CSOffsetExpr), ID(ID), Locations(std::move(Locations)),
+          LiveOuts(std::move(LiveOuts)) {}
   };
 
   typedef std::vector<CallsiteInfo> CallsiteInfoList;
@@ -218,8 +215,8 @@ private:
 
   MachineInstr::const_mop_iterator
   parseOperand(MachineInstr::const_mop_iterator MOI,
-               MachineInstr::const_mop_iterator MOE,
-               LocationVec &Locs, LiveOutVec &LiveOuts) const;
+               MachineInstr::const_mop_iterator MOE, LocationVec &Locs,
+               LiveOutVec &LiveOuts) const;
 
   /// \brief Create a live-out register record for the given register @p Reg.
   LiveOutReg createLiveOutReg(unsigned Reg,
@@ -254,7 +251,6 @@ private:
   void print(raw_ostream &OS);
   void debug() { print(dbgs()); }
 };
-
 }
 
 #endif
diff --git a/include/llvm/CodeGen/WinEHFuncInfo.h b/include/llvm/CodeGen/WinEHFuncInfo.h
index 291f3905512c..75638a058a30 100644
--- a/include/llvm/CodeGen/WinEHFuncInfo.h
+++ b/include/llvm/CodeGen/WinEHFuncInfo.h
@@ -91,7 +91,7 @@ private:
   // When the parseEHActions function is called to populate a vector of
   // instances of this class, the ExceptionObjectVar field will be nullptr
   // and the ExceptionObjectIndex will be the index of the exception object in
-  // the parent function's frameescape block.
+  // the parent function's localescape block.
   const Value *ExceptionObjectVar;
   int ExceptionObjectIndex;
   TinyPtrVector<BasicBlock *> ReturnTargets;
@@ -148,7 +148,7 @@ struct WinEHFuncInfo {
   int UnwindHelpFrameOffset = -1;
   unsigned NumIPToStateFuncsVisited = 0;
 
-  /// frameescape index of the 32-bit EH registration node. Set by
+  /// localescape index of the 32-bit EH registration node. Set by
   /// WinEHStatePass and used indirectly by SEH filter functions of the parent.
   int EHRegNodeEscapeIndex = INT_MAX;
 
diff --git a/include/llvm/ExecutionEngine/ExecutionEngine.h b/include/llvm/ExecutionEngine/ExecutionEngine.h
index e8af601d83b6..821c0181ce83 100644
--- a/include/llvm/ExecutionEngine/ExecutionEngine.h
+++ b/include/llvm/ExecutionEngine/ExecutionEngine.h
@@ -31,6 +31,7 @@
 #include <map>
 #include <string>
 #include <vector>
+#include <functional>
 
 namespace llvm {
 
@@ -89,6 +90,8 @@ public:
   uint64_t RemoveMapping(StringRef Name);
 };
 
+using FunctionCreator = std::function<void *(const std::string &)>;
+
 /// \brief Abstract interface for implementation execution of LLVM modules,
 /// designed to support both interpreter and just-in-time (JIT) compiler
 /// implementations.
@@ -147,7 +150,7 @@ protected:
   /// LazyFunctionCreator - If an unknown function is needed, this function
   /// pointer is invoked to create it.  If this returns null, the JIT will
   /// abort.
-  void *(*LazyFunctionCreator)(const std::string &);
+  FunctionCreator LazyFunctionCreator;
 
   /// getMangledName - Get mangled name.
   std::string getMangledName(const GlobalValue *GV);
@@ -470,8 +473,8 @@ public:
   /// InstallLazyFunctionCreator - If an unknown function is needed, the
   /// specified function pointer is invoked to create it.  If it returns null,
   /// the JIT will abort.
-  void InstallLazyFunctionCreator(void* (*P)(const std::string &)) {
-    LazyFunctionCreator = P;
+  void InstallLazyFunctionCreator(FunctionCreator C) {
+    LazyFunctionCreator = C;
   }
 
 protected:
diff --git a/include/llvm/ExecutionEngine/RuntimeDyld.h b/include/llvm/ExecutionEngine/RuntimeDyld.h
index 94c4038e8183..a808d9231167 100644
--- a/include/llvm/ExecutionEngine/RuntimeDyld.h
+++ b/include/llvm/ExecutionEngine/RuntimeDyld.h
@@ -153,6 +153,10 @@ public:
 
     /// This method returns the address of the specified function or variable.
     /// It is used to resolve symbols during module linking.
+    ///
+    /// If the returned symbol's address is equal to ~0ULL then RuntimeDyld will
+    /// skip all relocations for that symbol, and the client will be responsible
+    /// for handling them manually.
     virtual SymbolInfo findSymbol(const std::string &Name) = 0;
 
     /// This method returns the address of the specified symbol if it exists
diff --git a/include/llvm/IR/Attributes.h b/include/llvm/IR/Attributes.h
index 366bf709ab16..4d6d7da1fa5b 100644
--- a/include/llvm/IR/Attributes.h
+++ b/include/llvm/IR/Attributes.h
@@ -98,6 +98,8 @@ public:
     OptimizeNone,          ///< Function must not be optimized.
     ReadNone,              ///< Function does not access memory
     ReadOnly,              ///< Function only reads from memory
+    ArgMemOnly,            ///< Funciton can access memory only using pointers
+                           ///< based on its arguments.
     Returned,              ///< Return value is always equal to this argument
     ReturnsTwice,          ///< Function can return twice
     SExt,                  ///< Sign extended before/after call
diff --git a/include/llvm/IR/CallSite.h b/include/llvm/IR/CallSite.h
index dd2903e807e1..2841781e8a9e 100644
--- a/include/llvm/IR/CallSite.h
+++ b/include/llvm/IR/CallSite.h
@@ -290,6 +290,15 @@ public:
     CALLSITE_DELEGATE_SETTER(setOnlyReadsMemory());
   }
 
+  /// @brief Determine if the call can access memmory only using pointers based
+  /// on its arguments.
+  bool onlyAccessesArgMemory() const {
+    CALLSITE_DELEGATE_GETTER(onlyAccessesArgMemory());
+  }
+  void setOnlyAccessesArgMemory() {
+    CALLSITE_DELEGATE_SETTER(setOnlyAccessesArgMemory());
+  }
+
   /// @brief Determine if the call cannot return.
   bool doesNotReturn() const {
     CALLSITE_DELEGATE_GETTER(doesNotReturn());
diff --git a/include/llvm/IR/DIBuilder.h b/include/llvm/IR/DIBuilder.h
index d6296b622aab..aa43c02d5cd8 100644
--- a/include/llvm/IR/DIBuilder.h
+++ b/include/llvm/IR/DIBuilder.h
@@ -47,7 +47,7 @@ namespace llvm {
     SmallVector<Metadata *, 4> AllGVs;
     SmallVector<TrackingMDNodeRef, 4> AllImportedModules;
 
-    /// \brief Track nodes that may be unresolved.
+    /// Track nodes that may be unresolved.
     SmallVector<TrackingMDNodeRef, 4> UnresolvedNodes;
     bool AllowUnresolvedNodes;
 
@@ -57,49 +57,52 @@ namespace llvm {
     DIBuilder(const DIBuilder &) = delete;
     void operator=(const DIBuilder &) = delete;
 
-    /// \brief Create a temporary.
+    /// Create a temporary.
     ///
     /// Create an \a temporary node and track it in \a UnresolvedNodes.
     void trackIfUnresolved(MDNode *N);
 
   public:
-    /// \brief Construct a builder for a module.
+    /// Construct a builder for a module.
     ///
     /// If \c AllowUnresolved, collect unresolved nodes attached to the module
     /// in order to resolve cycles during \a finalize().
     explicit DIBuilder(Module &M, bool AllowUnresolved = true);
     enum DebugEmissionKind { FullDebug=1, LineTablesOnly };
 
-    /// finalize - Construct any deferred debug info descriptors.
+    /// Construct any deferred debug info descriptors.
     void finalize();
 
-    /// createCompileUnit - A CompileUnit provides an anchor for all debugging
+    /// A CompileUnit provides an anchor for all debugging
     /// information generated during this instance of compilation.
-    /// @param Lang     Source programming language, eg. dwarf::DW_LANG_C99
-    /// @param File     File name
-    /// @param Dir      Directory
-    /// @param Producer Identify the producer of debugging information and code.
-    ///                 Usually this is a compiler version string.
-    /// @param isOptimized A boolean flag which indicates whether optimization
-    ///                    is ON or not.
-    /// @param Flags    This string lists command line options. This string is
-    ///                 directly embedded in debug info output which may be used
-    ///                 by a tool analyzing generated debugging information.
-    /// @param RV       This indicates runtime version for languages like
-    ///                 Objective-C.
-    /// @param SplitName The name of the file that we'll split debug info out
-    ///                  into.
-    /// @param Kind     The kind of debug information to generate.
-    /// @param DWOId    The DWOId if this is a split skeleton compile unit.
-    /// @param EmitDebugInfo   A boolean flag which indicates whether debug
-    ///                        information should be written to the final
-    ///                        output or not. When this is false, debug
-    ///                        information annotations will be present in
-    ///                        the IL but they are not written to the final
-    ///                        assembly or object file. This supports tracking
-    ///                        source location information in the back end
-    ///                        without actually changing the output (e.g.,
-    ///                        when using optimization remarks).
+    /// \param Lang          Source programming language, eg. dwarf::DW_LANG_C99
+    /// \param File          File name
+    /// \param Dir           Directory
+    /// \param Producer      Identify the producer of debugging information
+    ///                      and code.  Usually this is a compiler
+    ///                      version string.
+    /// \param isOptimized   A boolean flag which indicates whether optimization
+    ///                      is enabled or not.
+    /// \param Flags         This string lists command line options. This
+    ///                      string is directly embedded in debug info
+    ///                      output which may be used by a tool
+    ///                      analyzing generated debugging information.
+    /// \param RV            This indicates runtime version for languages like
+    ///                      Objective-C.
+    /// \param SplitName     The name of the file that we'll split debug info
+    ///                      out into.
+    /// \param Kind          The kind of debug information to generate.
+    /// \param DWOId         The DWOId if this is a split skeleton compile unit.
+    /// \param EmitDebugInfo A boolean flag which indicates whether
+    ///                      debug information should be written to
+    ///                      the final output or not. When this is
+    ///                      false, debug information annotations will
+    ///                      be present in the IL but they are not
+    ///                      written to the final assembly or object
+    ///                      file. This supports tracking source
+    ///                      location information in the back end
+    ///                      without actually changing the output
+    ///                      (e.g., when using optimization remarks).
     DICompileUnit *
     createCompileUnit(unsigned Lang, StringRef File, StringRef Dir,
                       StringRef Producer, bool isOptimized, StringRef Flags,
@@ -107,155 +110,155 @@ namespace llvm {
                       DebugEmissionKind Kind = FullDebug, uint64_t DWOId = 0,
                       bool EmitDebugInfo = true);
 
-    /// createFile - Create a file descriptor to hold debugging information
+    /// Create a file descriptor to hold debugging information
     /// for a file.
     DIFile *createFile(StringRef Filename, StringRef Directory);
 
-    /// createEnumerator - Create a single enumerator value.
+    /// Create a single enumerator value.
     DIEnumerator *createEnumerator(StringRef Name, int64_t Val);
 
-    /// \brief Create a DWARF unspecified type.
+    /// Create a DWARF unspecified type.
     DIBasicType *createUnspecifiedType(StringRef Name);
 
-    /// \brief Create C++11 nullptr type.
+    /// Create C++11 nullptr type.
     DIBasicType *createNullPtrType();
 
-    /// createBasicType - Create debugging information entry for a basic
+    /// Create debugging information entry for a basic
     /// type.
-    /// @param Name        Type name.
-    /// @param SizeInBits  Size of the type.
-    /// @param AlignInBits Type alignment.
-    /// @param Encoding    DWARF encoding code, e.g. dwarf::DW_ATE_float.
+    /// \param Name        Type name.
+    /// \param SizeInBits  Size of the type.
+    /// \param AlignInBits Type alignment.
+    /// \param Encoding    DWARF encoding code, e.g. dwarf::DW_ATE_float.
     DIBasicType *createBasicType(StringRef Name, uint64_t SizeInBits,
                                  uint64_t AlignInBits, unsigned Encoding);
 
-    /// createQualifiedType - Create debugging information entry for a qualified
+    /// Create debugging information entry for a qualified
     /// type, e.g. 'const int'.
-    /// @param Tag         Tag identifing type, e.g. dwarf::TAG_volatile_type
-    /// @param FromTy      Base Type.
+    /// \param Tag         Tag identifing type, e.g. dwarf::TAG_volatile_type
+    /// \param FromTy      Base Type.
     DIDerivedType *createQualifiedType(unsigned Tag, DIType *FromTy);
 
-    /// createPointerType - Create debugging information entry for a pointer.
-    /// @param PointeeTy   Type pointed by this pointer.
-    /// @param SizeInBits  Size.
-    /// @param AlignInBits Alignment. (optional)
-    /// @param Name        Pointer type name. (optional)
+    /// Create debugging information entry for a pointer.
+    /// \param PointeeTy   Type pointed by this pointer.
+    /// \param SizeInBits  Size.
+    /// \param AlignInBits Alignment. (optional)
+    /// \param Name        Pointer type name. (optional)
     DIDerivedType *createPointerType(DIType *PointeeTy, uint64_t SizeInBits,
                                      uint64_t AlignInBits = 0,
                                      StringRef Name = "");
 
-    /// \brief Create debugging information entry for a pointer to member.
-    /// @param PointeeTy Type pointed to by this pointer.
-    /// @param SizeInBits  Size.
-    /// @param AlignInBits Alignment. (optional)
-    /// @param Class Type for which this pointer points to members of.
+    /// Create debugging information entry for a pointer to member.
+    /// \param PointeeTy Type pointed to by this pointer.
+    /// \param SizeInBits  Size.
+    /// \param AlignInBits Alignment. (optional)
+    /// \param Class Type for which this pointer points to members of.
     DIDerivedType *createMemberPointerType(DIType *PointeeTy, DIType *Class,
                                            uint64_t SizeInBits,
                                            uint64_t AlignInBits = 0);
 
-    /// createReferenceType - Create debugging information entry for a c++
+    /// Create debugging information entry for a c++
     /// style reference or rvalue reference type.
     DIDerivedType *createReferenceType(unsigned Tag, DIType *RTy);
 
-    /// createTypedef - Create debugging information entry for a typedef.
-    /// @param Ty          Original type.
-    /// @param Name        Typedef name.
-    /// @param File        File where this type is defined.
-    /// @param LineNo      Line number.
-    /// @param Context     The surrounding context for the typedef.
+    /// Create debugging information entry for a typedef.
+    /// \param Ty          Original type.
+    /// \param Name        Typedef name.
+    /// \param File        File where this type is defined.
+    /// \param LineNo      Line number.
+    /// \param Context     The surrounding context for the typedef.
     DIDerivedType *createTypedef(DIType *Ty, StringRef Name, DIFile *File,
                                  unsigned LineNo, DIScope *Context);
 
-    /// createFriend - Create debugging information entry for a 'friend'.
+    /// Create debugging information entry for a 'friend'.
     DIDerivedType *createFriend(DIType *Ty, DIType *FriendTy);
 
-    /// createInheritance - Create debugging information entry to establish
+    /// Create debugging information entry to establish
     /// inheritance relationship between two types.
-    /// @param Ty           Original type.
-    /// @param BaseTy       Base type. Ty is inherits from base.
-    /// @param BaseOffset   Base offset.
-    /// @param Flags        Flags to describe inheritance attribute,
+    /// \param Ty           Original type.
+    /// \param BaseTy       Base type. Ty is inherits from base.
+    /// \param BaseOffset   Base offset.
+    /// \param Flags        Flags to describe inheritance attribute,
     ///                     e.g. private
     DIDerivedType *createInheritance(DIType *Ty, DIType *BaseTy,
                                      uint64_t BaseOffset, unsigned Flags);
 
-    /// createMemberType - Create debugging information entry for a member.
-    /// @param Scope        Member scope.
-    /// @param Name         Member name.
-    /// @param File         File where this member is defined.
-    /// @param LineNo       Line number.
-    /// @param SizeInBits   Member size.
-    /// @param AlignInBits  Member alignment.
-    /// @param OffsetInBits Member offset.
-    /// @param Flags        Flags to encode member attribute, e.g. private
-    /// @param Ty           Parent type.
+    /// Create debugging information entry for a member.
+    /// \param Scope        Member scope.
+    /// \param Name         Member name.
+    /// \param File         File where this member is defined.
+    /// \param LineNo       Line number.
+    /// \param SizeInBits   Member size.
+    /// \param AlignInBits  Member alignment.
+    /// \param OffsetInBits Member offset.
+    /// \param Flags        Flags to encode member attribute, e.g. private
+    /// \param Ty           Parent type.
     DIDerivedType *createMemberType(DIScope *Scope, StringRef Name,
                                     DIFile *File, unsigned LineNo,
                                     uint64_t SizeInBits, uint64_t AlignInBits,
                                     uint64_t OffsetInBits, unsigned Flags,
                                     DIType *Ty);
 
-    /// createStaticMemberType - Create debugging information entry for a
+    /// Create debugging information entry for a
     /// C++ static data member.
-    /// @param Scope      Member scope.
-    /// @param Name       Member name.
-    /// @param File       File where this member is declared.
-    /// @param LineNo     Line number.
-    /// @param Ty         Type of the static member.
-    /// @param Flags      Flags to encode member attribute, e.g. private.
-    /// @param Val        Const initializer of the member.
+    /// \param Scope      Member scope.
+    /// \param Name       Member name.
+    /// \param File       File where this member is declared.
+    /// \param LineNo     Line number.
+    /// \param Ty         Type of the static member.
+    /// \param Flags      Flags to encode member attribute, e.g. private.
+    /// \param Val        Const initializer of the member.
     DIDerivedType *createStaticMemberType(DIScope *Scope, StringRef Name,
                                           DIFile *File, unsigned LineNo,
                                           DIType *Ty, unsigned Flags,
                                           llvm::Constant *Val);
 
-    /// createObjCIVar - Create debugging information entry for Objective-C
+    /// Create debugging information entry for Objective-C
     /// instance variable.
-    /// @param Name         Member name.
-    /// @param File         File where this member is defined.
-    /// @param LineNo       Line number.
-    /// @param SizeInBits   Member size.
-    /// @param AlignInBits  Member alignment.
-    /// @param OffsetInBits Member offset.
-    /// @param Flags        Flags to encode member attribute, e.g. private
-    /// @param Ty           Parent type.
-    /// @param PropertyNode Property associated with this ivar.
+    /// \param Name         Member name.
+    /// \param File         File where this member is defined.
+    /// \param LineNo       Line number.
+    /// \param SizeInBits   Member size.
+    /// \param AlignInBits  Member alignment.
+    /// \param OffsetInBits Member offset.
+    /// \param Flags        Flags to encode member attribute, e.g. private
+    /// \param Ty           Parent type.
+    /// \param PropertyNode Property associated with this ivar.
     DIDerivedType *createObjCIVar(StringRef Name, DIFile *File, unsigned LineNo,
                                   uint64_t SizeInBits, uint64_t AlignInBits,
                                   uint64_t OffsetInBits, unsigned Flags,
                                   DIType *Ty, MDNode *PropertyNode);
 
-    /// createObjCProperty - Create debugging information entry for Objective-C
+    /// Create debugging information entry for Objective-C
     /// property.
-    /// @param Name         Property name.
-    /// @param File         File where this property is defined.
-    /// @param LineNumber   Line number.
-    /// @param GetterName   Name of the Objective C property getter selector.
-    /// @param SetterName   Name of the Objective C property setter selector.
-    /// @param PropertyAttributes Objective C property attributes.
-    /// @param Ty           Type.
+    /// \param Name         Property name.
+    /// \param File         File where this property is defined.
+    /// \param LineNumber   Line number.
+    /// \param GetterName   Name of the Objective C property getter selector.
+    /// \param SetterName   Name of the Objective C property setter selector.
+    /// \param PropertyAttributes Objective C property attributes.
+    /// \param Ty           Type.
     DIObjCProperty *createObjCProperty(StringRef Name, DIFile *File,
                                        unsigned LineNumber,
                                        StringRef GetterName,
                                        StringRef SetterName,
                                        unsigned PropertyAttributes, DIType *Ty);
 
-    /// createClassType - Create debugging information entry for a class.
-    /// @param Scope        Scope in which this class is defined.
-    /// @param Name         class name.
-    /// @param File         File where this member is defined.
-    /// @param LineNumber   Line number.
-    /// @param SizeInBits   Member size.
-    /// @param AlignInBits  Member alignment.
-    /// @param OffsetInBits Member offset.
-    /// @param Flags        Flags to encode member attribute, e.g. private
-    /// @param Elements     class members.
-    /// @param VTableHolder Debug info of the base class that contains vtable
+    /// Create debugging information entry for a class.
+    /// \param Scope        Scope in which this class is defined.
+    /// \param Name         class name.
+    /// \param File         File where this member is defined.
+    /// \param LineNumber   Line number.
+    /// \param SizeInBits   Member size.
+    /// \param AlignInBits  Member alignment.
+    /// \param OffsetInBits Member offset.
+    /// \param Flags        Flags to encode member attribute, e.g. private
+    /// \param Elements     class members.
+    /// \param VTableHolder Debug info of the base class that contains vtable
     ///                     for this type. This is used in
     ///                     DW_AT_containing_type. See DWARF documentation
     ///                     for more info.
-    /// @param TemplateParms Template type parameters.
-    /// @param UniqueIdentifier A unique identifier for the class.
+    /// \param TemplateParms Template type parameters.
+    /// \param UniqueIdentifier A unique identifier for the class.
     DICompositeType *createClassType(DIScope *Scope, StringRef Name,
                                      DIFile *File, unsigned LineNumber,
                                      uint64_t SizeInBits, uint64_t AlignInBits,
@@ -265,34 +268,34 @@ namespace llvm {
                                      MDNode *TemplateParms = nullptr,
                                      StringRef UniqueIdentifier = "");
 
-    /// createStructType - Create debugging information entry for a struct.
-    /// @param Scope        Scope in which this struct is defined.
-    /// @param Name         Struct name.
-    /// @param File         File where this member is defined.
-    /// @param LineNumber   Line number.
-    /// @param SizeInBits   Member size.
-    /// @param AlignInBits  Member alignment.
-    /// @param Flags        Flags to encode member attribute, e.g. private
-    /// @param Elements     Struct elements.
-    /// @param RunTimeLang  Optional parameter, Objective-C runtime version.
-    /// @param UniqueIdentifier A unique identifier for the struct.
+    /// Create debugging information entry for a struct.
+    /// \param Scope        Scope in which this struct is defined.
+    /// \param Name         Struct name.
+    /// \param File         File where this member is defined.
+    /// \param LineNumber   Line number.
+    /// \param SizeInBits   Member size.
+    /// \param AlignInBits  Member alignment.
+    /// \param Flags        Flags to encode member attribute, e.g. private
+    /// \param Elements     Struct elements.
+    /// \param RunTimeLang  Optional parameter, Objective-C runtime version.
+    /// \param UniqueIdentifier A unique identifier for the struct.
     DICompositeType *createStructType(
         DIScope *Scope, StringRef Name, DIFile *File, unsigned LineNumber,
         uint64_t SizeInBits, uint64_t AlignInBits, unsigned Flags,
         DIType *DerivedFrom, DINodeArray Elements, unsigned RunTimeLang = 0,
         DIType *VTableHolder = nullptr, StringRef UniqueIdentifier = "");
 
-    /// createUnionType - Create debugging information entry for an union.
-    /// @param Scope        Scope in which this union is defined.
-    /// @param Name         Union name.
-    /// @param File         File where this member is defined.
-    /// @param LineNumber   Line number.
-    /// @param SizeInBits   Member size.
-    /// @param AlignInBits  Member alignment.
-    /// @param Flags        Flags to encode member attribute, e.g. private
-    /// @param Elements     Union elements.
-    /// @param RunTimeLang  Optional parameter, Objective-C runtime version.
-    /// @param UniqueIdentifier A unique identifier for the union.
+    /// Create debugging information entry for an union.
+    /// \param Scope        Scope in which this union is defined.
+    /// \param Name         Union name.
+    /// \param File         File where this member is defined.
+    /// \param LineNumber   Line number.
+    /// \param SizeInBits   Member size.
+    /// \param AlignInBits  Member alignment.
+    /// \param Flags        Flags to encode member attribute, e.g. private
+    /// \param Elements     Union elements.
+    /// \param RunTimeLang  Optional parameter, Objective-C runtime version.
+    /// \param UniqueIdentifier A unique identifier for the union.
     DICompositeType *createUnionType(DIScope *Scope, StringRef Name,
                                      DIFile *File, unsigned LineNumber,
                                      uint64_t SizeInBits, uint64_t AlignInBits,
@@ -300,95 +303,95 @@ namespace llvm {
                                      unsigned RunTimeLang = 0,
                                      StringRef UniqueIdentifier = "");
 
-    /// createTemplateTypeParameter - Create debugging information for template
+    /// Create debugging information for template
     /// type parameter.
-    /// @param Scope        Scope in which this type is defined.
-    /// @param Name         Type parameter name.
-    /// @param Ty           Parameter type.
+    /// \param Scope        Scope in which this type is defined.
+    /// \param Name         Type parameter name.
+    /// \param Ty           Parameter type.
     DITemplateTypeParameter *
     createTemplateTypeParameter(DIScope *Scope, StringRef Name, DIType *Ty);
 
-    /// createTemplateValueParameter - Create debugging information for template
+    /// Create debugging information for template
     /// value parameter.
-    /// @param Scope        Scope in which this type is defined.
-    /// @param Name         Value parameter name.
-    /// @param Ty           Parameter type.
-    /// @param Val          Constant parameter value.
+    /// \param Scope        Scope in which this type is defined.
+    /// \param Name         Value parameter name.
+    /// \param Ty           Parameter type.
+    /// \param Val          Constant parameter value.
     DITemplateValueParameter *createTemplateValueParameter(DIScope *Scope,
                                                            StringRef Name,
                                                            DIType *Ty,
                                                            Constant *Val);
 
-    /// \brief Create debugging information for a template template parameter.
-    /// @param Scope        Scope in which this type is defined.
-    /// @param Name         Value parameter name.
-    /// @param Ty           Parameter type.
-    /// @param Val          The fully qualified name of the template.
+    /// Create debugging information for a template template parameter.
+    /// \param Scope        Scope in which this type is defined.
+    /// \param Name         Value parameter name.
+    /// \param Ty           Parameter type.
+    /// \param Val          The fully qualified name of the template.
     DITemplateValueParameter *createTemplateTemplateParameter(DIScope *Scope,
                                                               StringRef Name,
                                                               DIType *Ty,
                                                               StringRef Val);
 
-    /// \brief Create debugging information for a template parameter pack.
-    /// @param Scope        Scope in which this type is defined.
-    /// @param Name         Value parameter name.
-    /// @param Ty           Parameter type.
-    /// @param Val          An array of types in the pack.
+    /// Create debugging information for a template parameter pack.
+    /// \param Scope        Scope in which this type is defined.
+    /// \param Name         Value parameter name.
+    /// \param Ty           Parameter type.
+    /// \param Val          An array of types in the pack.
     DITemplateValueParameter *createTemplateParameterPack(DIScope *Scope,
                                                           StringRef Name,
                                                           DIType *Ty,
                                                           DINodeArray Val);
 
-    /// createArrayType - Create debugging information entry for an array.
-    /// @param Size         Array size.
-    /// @param AlignInBits  Alignment.
-    /// @param Ty           Element type.
-    /// @param Subscripts   Subscripts.
+    /// Create debugging information entry for an array.
+    /// \param Size         Array size.
+    /// \param AlignInBits  Alignment.
+    /// \param Ty           Element type.
+    /// \param Subscripts   Subscripts.
     DICompositeType *createArrayType(uint64_t Size, uint64_t AlignInBits,
                                      DIType *Ty, DINodeArray Subscripts);
 
-    /// createVectorType - Create debugging information entry for a vector type.
-    /// @param Size         Array size.
-    /// @param AlignInBits  Alignment.
-    /// @param Ty           Element type.
-    /// @param Subscripts   Subscripts.
+    /// Create debugging information entry for a vector type.
+    /// \param Size         Array size.
+    /// \param AlignInBits  Alignment.
+    /// \param Ty           Element type.
+    /// \param Subscripts   Subscripts.
     DICompositeType *createVectorType(uint64_t Size, uint64_t AlignInBits,
                                       DIType *Ty, DINodeArray Subscripts);
 
-    /// createEnumerationType - Create debugging information entry for an
+    /// Create debugging information entry for an
     /// enumeration.
-    /// @param Scope          Scope in which this enumeration is defined.
-    /// @param Name           Union name.
-    /// @param File           File where this member is defined.
-    /// @param LineNumber     Line number.
-    /// @param SizeInBits     Member size.
-    /// @param AlignInBits    Member alignment.
-    /// @param Elements       Enumeration elements.
-    /// @param UnderlyingType Underlying type of a C++11/ObjC fixed enum.
-    /// @param UniqueIdentifier A unique identifier for the enum.
+    /// \param Scope          Scope in which this enumeration is defined.
+    /// \param Name           Union name.
+    /// \param File           File where this member is defined.
+    /// \param LineNumber     Line number.
+    /// \param SizeInBits     Member size.
+    /// \param AlignInBits    Member alignment.
+    /// \param Elements       Enumeration elements.
+    /// \param UnderlyingType Underlying type of a C++11/ObjC fixed enum.
+    /// \param UniqueIdentifier A unique identifier for the enum.
     DICompositeType *createEnumerationType(
         DIScope *Scope, StringRef Name, DIFile *File, unsigned LineNumber,
         uint64_t SizeInBits, uint64_t AlignInBits, DINodeArray Elements,
         DIType *UnderlyingType, StringRef UniqueIdentifier = "");
 
-    /// createSubroutineType - Create subroutine type.
-    /// @param File            File in which this subroutine is defined.
-    /// @param ParameterTypes  An array of subroutine parameter types. This
+    /// Create subroutine type.
+    /// \param File            File in which this subroutine is defined.
+    /// \param ParameterTypes  An array of subroutine parameter types. This
     ///                        includes return type at 0th index.
-    /// @param Flags           E.g.: LValueReference.
+    /// \param Flags           E.g.: LValueReference.
     ///                        These flags are used to emit dwarf attributes.
     DISubroutineType *createSubroutineType(DIFile *File,
                                            DITypeRefArray ParameterTypes,
                                            unsigned Flags = 0);
 
-    /// createArtificialType - Create a new DIType* with "artificial" flag set.
+    /// Create a new DIType* with "artificial" flag set.
     DIType *createArtificialType(DIType *Ty);
 
-    /// createObjectPointerType - Create a new DIType* with the "object pointer"
+    /// Create a new DIType* with the "object pointer"
     /// flag set.
     DIType *createObjectPointerType(DIType *Ty);
 
-    /// \brief Create a permanent forward-declared type.
+    /// Create a permanent forward-declared type.
     DICompositeType *createForwardDecl(unsigned Tag, StringRef Name,
                                        DIScope *Scope, DIFile *F, unsigned Line,
                                        unsigned RuntimeLang = 0,
@@ -396,43 +399,43 @@ namespace llvm {
                                        uint64_t AlignInBits = 0,
                                        StringRef UniqueIdentifier = "");
 
-    /// \brief Create a temporary forward-declared type.
+    /// Create a temporary forward-declared type.
     DICompositeType *createReplaceableCompositeType(
         unsigned Tag, StringRef Name, DIScope *Scope, DIFile *F, unsigned Line,
         unsigned RuntimeLang = 0, uint64_t SizeInBits = 0,
         uint64_t AlignInBits = 0, unsigned Flags = DINode::FlagFwdDecl,
         StringRef UniqueIdentifier = "");
 
-    /// retainType - Retain DIType* in a module even if it is not referenced
+    /// Retain DIType* in a module even if it is not referenced
     /// through debug info anchors.
     void retainType(DIType *T);
 
-    /// createUnspecifiedParameter - Create unspecified parameter type
+    /// Create unspecified parameter type
     /// for a subroutine type.
     DIBasicType *createUnspecifiedParameter();
 
-    /// getOrCreateArray - Get a DINodeArray, create one if required.
+    /// Get a DINodeArray, create one if required.
     DINodeArray getOrCreateArray(ArrayRef<Metadata *> Elements);
 
-    /// getOrCreateTypeArray - Get a DITypeRefArray, create one if required.
+    /// Get a DITypeRefArray, create one if required.
     DITypeRefArray getOrCreateTypeArray(ArrayRef<Metadata *> Elements);
 
-    /// getOrCreateSubrange - Create a descriptor for a value range.  This
+    /// Create a descriptor for a value range.  This
     /// implicitly uniques the values returned.
     DISubrange *getOrCreateSubrange(int64_t Lo, int64_t Count);
 
-    /// createGlobalVariable - Create a new descriptor for the specified
+    /// Create a new descriptor for the specified
     /// variable.
-    /// @param Context     Variable scope.
-    /// @param Name        Name of the variable.
-    /// @param LinkageName Mangled  name of the variable.
-    /// @param File        File where this variable is defined.
-    /// @param LineNo      Line number.
-    /// @param Ty          Variable Type.
-    /// @param isLocalToUnit Boolean flag indicate whether this variable is
+    /// \param Context     Variable scope.
+    /// \param Name        Name of the variable.
+    /// \param LinkageName Mangled  name of the variable.
+    /// \param File        File where this variable is defined.
+    /// \param LineNo      Line number.
+    /// \param Ty          Variable Type.
+    /// \param isLocalToUnit Boolean flag indicate whether this variable is
     ///                      externally visible or not.
-    /// @param Val         llvm::Value of the variable.
-    /// @param Decl        Reference to the corresponding declaration.
+    /// \param Val         llvm::Value of the variable.
+    /// \param Decl        Reference to the corresponding declaration.
     DIGlobalVariable *createGlobalVariable(DIScope *Context, StringRef Name,
                                            StringRef LinkageName, DIFile *File,
                                            unsigned LineNo, DIType *Ty,
@@ -440,26 +443,26 @@ namespace llvm {
                                            llvm::Constant *Val,
                                            MDNode *Decl = nullptr);
 
-    /// createTempGlobalVariableFwdDecl - Identical to createGlobalVariable
+    /// Identical to createGlobalVariable
     /// except that the resulting DbgNode is temporary and meant to be RAUWed.
     DIGlobalVariable *createTempGlobalVariableFwdDecl(
         DIScope *Context, StringRef Name, StringRef LinkageName, DIFile *File,
         unsigned LineNo, DIType *Ty, bool isLocalToUnit, llvm::Constant *Val,
         MDNode *Decl = nullptr);
 
-    /// createLocalVariable - Create a new descriptor for the specified
+    /// Create a new descriptor for the specified
     /// local variable.
-    /// @param Tag         Dwarf TAG. Usually DW_TAG_auto_variable or
+    /// \param Tag         Dwarf TAG. Usually DW_TAG_auto_variable or
     ///                    DW_TAG_arg_variable.
-    /// @param Scope       Variable scope.
-    /// @param Name        Variable name.
-    /// @param File        File where this variable is defined.
-    /// @param LineNo      Line number.
-    /// @param Ty          Variable Type
-    /// @param AlwaysPreserve Boolean. Set to true if debug info for this
+    /// \param Scope       Variable scope.
+    /// \param Name        Variable name.
+    /// \param File        File where this variable is defined.
+    /// \param LineNo      Line number.
+    /// \param Ty          Variable Type
+    /// \param AlwaysPreserve Boolean. Set to true if debug info for this
     ///                       variable should be preserved in optimized build.
-    /// @param Flags       Flags, e.g. artificial variable.
-    /// @param ArgNo       If this variable is an argument then this argument's
+    /// \param Flags       Flags, e.g. artificial variable.
+    /// \param ArgNo       If this variable is an argument then this argument's
     ///                    number. 1 indicates 1st argument.
     DILocalVariable *createLocalVariable(unsigned Tag, DIScope *Scope,
                                          StringRef Name, DIFile *File,
@@ -468,36 +471,36 @@ namespace llvm {
                                          unsigned Flags = 0,
                                          unsigned ArgNo = 0);
 
-    /// createExpression - Create a new descriptor for the specified
+    /// Create a new descriptor for the specified
     /// variable which has a complex address expression for its address.
-    /// @param Addr        An array of complex address operations.
+    /// \param Addr        An array of complex address operations.
     DIExpression *createExpression(ArrayRef<uint64_t> Addr = None);
     DIExpression *createExpression(ArrayRef<int64_t> Addr);
 
-    /// createBitPieceExpression - Create a descriptor to describe one part
+    /// Create a descriptor to describe one part
     /// of aggregate variable that is fragmented across multiple Values.
     ///
-    /// @param OffsetInBits Offset of the piece in bits.
-    /// @param SizeInBits   Size of the piece in bits.
+    /// \param OffsetInBits Offset of the piece in bits.
+    /// \param SizeInBits   Size of the piece in bits.
     DIExpression *createBitPieceExpression(unsigned OffsetInBits,
                                            unsigned SizeInBits);
 
-    /// createFunction - Create a new descriptor for the specified subprogram.
+    /// Create a new descriptor for the specified subprogram.
     /// See comments in DISubprogram* for descriptions of these fields.
-    /// @param Scope         Function scope.
-    /// @param Name          Function name.
-    /// @param LinkageName   Mangled function name.
-    /// @param File          File where this variable is defined.
-    /// @param LineNo        Line number.
-    /// @param Ty            Function type.
-    /// @param isLocalToUnit True if this function is not externally visible.
-    /// @param isDefinition  True if this is a function definition.
-    /// @param ScopeLine     Set to the beginning of the scope this starts
-    /// @param Flags         e.g. is this function prototyped or not.
+    /// \param Scope         Function scope.
+    /// \param Name          Function name.
+    /// \param LinkageName   Mangled function name.
+    /// \param File          File where this variable is defined.
+    /// \param LineNo        Line number.
+    /// \param Ty            Function type.
+    /// \param isLocalToUnit True if this function is not externally visible.
+    /// \param isDefinition  True if this is a function definition.
+    /// \param ScopeLine     Set to the beginning of the scope this starts
+    /// \param Flags         e.g. is this function prototyped or not.
     ///                      These flags are used to emit dwarf attributes.
-    /// @param isOptimized   True if optimization is ON.
-    /// @param Fn            llvm::Function pointer.
-    /// @param TParam        Function template parameters.
+    /// \param isOptimized   True if optimization is ON.
+    /// \param Fn            llvm::Function pointer.
+    /// \param TParam        Function template parameters.
     DISubprogram *
     createFunction(DIScope *Scope, StringRef Name, StringRef LinkageName,
                    DIFile *File, unsigned LineNo, DISubroutineType *Ty,
@@ -506,7 +509,7 @@ namespace llvm {
                    Function *Fn = nullptr, MDNode *TParam = nullptr,
                    MDNode *Decl = nullptr);
 
-    /// createTempFunctionFwdDecl - Identical to createFunction,
+    /// Identical to createFunction,
     /// except that the resulting DbgNode is meant to be RAUWed.
     DISubprogram *createTempFunctionFwdDecl(
         DIScope *Scope, StringRef Name, StringRef LinkageName, DIFile *File,
@@ -525,25 +528,25 @@ namespace llvm {
                    Function *Fn = nullptr, MDNode *TParam = nullptr,
                    MDNode *Decl = nullptr);
 
-    /// createMethod - Create a new descriptor for the specified C++ method.
-    /// See comments in DISubprogram* for descriptions of these fields.
-    /// @param Scope         Function scope.
-    /// @param Name          Function name.
-    /// @param LinkageName   Mangled function name.
-    /// @param File          File where this variable is defined.
-    /// @param LineNo        Line number.
-    /// @param Ty            Function type.
-    /// @param isLocalToUnit True if this function is not externally visible..
-    /// @param isDefinition  True if this is a function definition.
-    /// @param Virtuality    Attributes describing virtualness. e.g. pure
+    /// Create a new descriptor for the specified C++ method.
+    /// See comments in \a DISubprogram* for descriptions of these fields.
+    /// \param Scope         Function scope.
+    /// \param Name          Function name.
+    /// \param LinkageName   Mangled function name.
+    /// \param File          File where this variable is defined.
+    /// \param LineNo        Line number.
+    /// \param Ty            Function type.
+    /// \param isLocalToUnit True if this function is not externally visible..
+    /// \param isDefinition  True if this is a function definition.
+    /// \param Virtuality    Attributes describing virtualness. e.g. pure
     ///                      virtual function.
-    /// @param VTableIndex   Index no of this method in virtual table.
-    /// @param VTableHolder  Type that holds vtable.
-    /// @param Flags         e.g. is this function prototyped or not.
+    /// \param VTableIndex   Index no of this method in virtual table.
+    /// \param VTableHolder  Type that holds vtable.
+    /// \param Flags         e.g. is this function prototyped or not.
     ///                      This flags are used to emit dwarf attributes.
-    /// @param isOptimized   True if optimization is ON.
-    /// @param Fn            llvm::Function pointer.
-    /// @param TParam        Function template parameters.
+    /// \param isOptimized   True if optimization is ON.
+    /// \param Fn            llvm::Function pointer.
+    /// \param TParam        Function template parameters.
     DISubprogram *
     createMethod(DIScope *Scope, StringRef Name, StringRef LinkageName,
                  DIFile *File, unsigned LineNo, DISubroutineType *Ty,
@@ -552,131 +555,131 @@ namespace llvm {
                  unsigned Flags = 0, bool isOptimized = false,
                  Function *Fn = nullptr, MDNode *TParam = nullptr);
 
-    /// createNameSpace - This creates new descriptor for a namespace
-    /// with the specified parent scope.
-    /// @param Scope       Namespace scope
-    /// @param Name        Name of this namespace
-    /// @param File        Source file
-    /// @param LineNo      Line number
+    /// This creates new descriptor for a namespace with the specified
+    /// parent scope.
+    /// \param Scope       Namespace scope
+    /// \param Name        Name of this namespace
+    /// \param File        Source file
+    /// \param LineNo      Line number
     DINamespace *createNameSpace(DIScope *Scope, StringRef Name, DIFile *File,
                                  unsigned LineNo);
 
-    /// createModule - This creates new descriptor for a module
-    /// with the specified parent scope.
-    /// @param Scope       Parent scope
-    /// @param Name        Name of this module
-    /// @param ConfigurationMacros
+    /// This creates new descriptor for a module with the specified
+    /// parent scope.
+    /// \param Scope       Parent scope
+    /// \param Name        Name of this module
+    /// \param ConfigurationMacros
     ///                    A space-separated shell-quoted list of -D macro
     ///                    definitions as they would appear on a command line.
-    /// @param IncludePath The path to the module map file.
-    /// @param ISysRoot    The clang system root (value of -isysroot).
+    /// \param IncludePath The path to the module map file.
+    /// \param ISysRoot    The clang system root (value of -isysroot).
     DIModule *createModule(DIScope *Scope, StringRef Name,
                            StringRef ConfigurationMacros,
                            StringRef IncludePath,
                            StringRef ISysRoot);
 
-    /// createLexicalBlockFile - This creates a descriptor for a lexical
-    /// block with a new file attached. This merely extends the existing
+    /// This creates a descriptor for a lexical block with a new file
+    /// attached. This merely extends the existing
     /// lexical block as it crosses a file.
-    /// @param Scope       Lexical block.
-    /// @param File        Source file.
-    /// @param Discriminator DWARF path discriminator value.
+    /// \param Scope       Lexical block.
+    /// \param File        Source file.
+    /// \param Discriminator DWARF path discriminator value.
     DILexicalBlockFile *createLexicalBlockFile(DIScope *Scope, DIFile *File,
                                                unsigned Discriminator = 0);
 
-    /// createLexicalBlock - This creates a descriptor for a lexical block
-    /// with the specified parent context.
-    /// @param Scope         Parent lexical scope.
-    /// @param File          Source file.
-    /// @param Line          Line number.
-    /// @param Col           Column number.
+    /// This creates a descriptor for a lexical block with the
+    /// specified parent context.
+    /// \param Scope         Parent lexical scope.
+    /// \param File          Source file.
+    /// \param Line          Line number.
+    /// \param Col           Column number.
     DILexicalBlock *createLexicalBlock(DIScope *Scope, DIFile *File,
                                        unsigned Line, unsigned Col);
 
-    /// \brief Create a descriptor for an imported module.
-    /// @param Context The scope this module is imported into
-    /// @param NS The namespace being imported here
-    /// @param Line Line number
+    /// Create a descriptor for an imported module.
+    /// \param Context The scope this module is imported into
+    /// \param NS The namespace being imported here
+    /// \param Line Line number
     DIImportedEntity *createImportedModule(DIScope *Context, DINamespace *NS,
                                            unsigned Line);
 
-    /// \brief Create a descriptor for an imported module.
-    /// @param Context The scope this module is imported into
-    /// @param NS An aliased namespace
-    /// @param Line Line number
+    /// Create a descriptor for an imported module.
+    /// \param Context The scope this module is imported into
+    /// \param NS An aliased namespace
+    /// \param Line Line number
     DIImportedEntity *createImportedModule(DIScope *Context,
                                            DIImportedEntity *NS, unsigned Line);
 
-    /// \brief Create a descriptor for an imported module.
-    /// @param Context The scope this module is imported into
-    /// @param M The module being imported here
-    /// @param Line Line number
+    /// Create a descriptor for an imported module.
+    /// \param Context The scope this module is imported into
+    /// \param M The module being imported here
+    /// \param Line Line number
     DIImportedEntity *createImportedModule(DIScope *Context, DIModule *M,
                                            unsigned Line);
 
-    /// \brief Create a descriptor for an imported function.
-    /// @param Context The scope this module is imported into
-    /// @param Decl The declaration (or definition) of a function, type, or
+    /// Create a descriptor for an imported function.
+    /// \param Context The scope this module is imported into
+    /// \param Decl The declaration (or definition) of a function, type, or
     ///             variable
-    /// @param Line Line number
+    /// \param Line Line number
     DIImportedEntity *createImportedDeclaration(DIScope *Context, DINode *Decl,
                                                 unsigned Line,
                                                 StringRef Name = "");
 
-    /// insertDeclare - Insert a new llvm.dbg.declare intrinsic call.
-    /// @param Storage     llvm::Value of the variable
-    /// @param VarInfo     Variable's debug info descriptor.
-    /// @param Expr         A complex location expression.
-    /// @param DL           Debug info location.
-    /// @param InsertAtEnd Location for the new intrinsic.
+    /// Insert a new llvm.dbg.declare intrinsic call.
+    /// \param Storage     llvm::Value of the variable
+    /// \param VarInfo     Variable's debug info descriptor.
+    /// \param Expr        A complex location expression.
+    /// \param DL          Debug info location.
+    /// \param InsertAtEnd Location for the new intrinsic.
     Instruction *insertDeclare(llvm::Value *Storage, DILocalVariable *VarInfo,
                                DIExpression *Expr, const DILocation *DL,
                                BasicBlock *InsertAtEnd);
 
-    /// insertDeclare - Insert a new llvm.dbg.declare intrinsic call.
-    /// @param Storage      llvm::Value of the variable
-    /// @param VarInfo      Variable's debug info descriptor.
-    /// @param Expr         A complex location expression.
-    /// @param DL           Debug info location.
-    /// @param InsertBefore Location for the new intrinsic.
+    /// Insert a new llvm.dbg.declare intrinsic call.
+    /// \param Storage      llvm::Value of the variable
+    /// \param VarInfo      Variable's debug info descriptor.
+    /// \param Expr         A complex location expression.
+    /// \param DL           Debug info location.
+    /// \param InsertBefore Location for the new intrinsic.
     Instruction *insertDeclare(llvm::Value *Storage, DILocalVariable *VarInfo,
                                DIExpression *Expr, const DILocation *DL,
                                Instruction *InsertBefore);
 
-    /// insertDbgValueIntrinsic - Insert a new llvm.dbg.value intrinsic call.
-    /// @param Val          llvm::Value of the variable
-    /// @param Offset       Offset
-    /// @param VarInfo      Variable's debug info descriptor.
-    /// @param Expr         A complex location expression.
-    /// @param DL           Debug info location.
-    /// @param InsertAtEnd Location for the new intrinsic.
+    /// Insert a new llvm.dbg.value intrinsic call.
+    /// \param Val          llvm::Value of the variable
+    /// \param Offset       Offset
+    /// \param VarInfo      Variable's debug info descriptor.
+    /// \param Expr         A complex location expression.
+    /// \param DL           Debug info location.
+    /// \param InsertAtEnd Location for the new intrinsic.
     Instruction *insertDbgValueIntrinsic(llvm::Value *Val, uint64_t Offset,
                                          DILocalVariable *VarInfo,
                                          DIExpression *Expr,
                                          const DILocation *DL,
                                          BasicBlock *InsertAtEnd);
 
-    /// insertDbgValueIntrinsic - Insert a new llvm.dbg.value intrinsic call.
-    /// @param Val          llvm::Value of the variable
-    /// @param Offset       Offset
-    /// @param VarInfo      Variable's debug info descriptor.
-    /// @param Expr         A complex location expression.
-    /// @param DL           Debug info location.
-    /// @param InsertBefore Location for the new intrinsic.
+    /// Insert a new llvm.dbg.value intrinsic call.
+    /// \param Val          llvm::Value of the variable
+    /// \param Offset       Offset
+    /// \param VarInfo      Variable's debug info descriptor.
+    /// \param Expr         A complex location expression.
+    /// \param DL           Debug info location.
+    /// \param InsertBefore Location for the new intrinsic.
     Instruction *insertDbgValueIntrinsic(llvm::Value *Val, uint64_t Offset,
                                          DILocalVariable *VarInfo,
                                          DIExpression *Expr,
                                          const DILocation *DL,
                                          Instruction *InsertBefore);
 
-    /// \brief Replace the vtable holder in the given composite type.
+    /// Replace the vtable holder in the given composite type.
     ///
     /// If this creates a self reference, it may orphan some unresolved cycles
     /// in the operands of \c T, so \a DIBuilder needs to track that.
     void replaceVTableHolder(DICompositeType *&T,
                              DICompositeType *VTableHolder);
 
-    /// \brief Replace arrays on a composite type.
+    /// Replace arrays on a composite type.
     ///
     /// If \c T is resolved, but the arrays aren't -- which can happen if \c T
     /// has a self-reference -- \a DIBuilder needs to track the array to
@@ -684,7 +687,7 @@ namespace llvm {
     void replaceArrays(DICompositeType *&T, DINodeArray Elements,
                        DINodeArray TParems = DINodeArray());
 
-    /// \brief Replace a temporary node.
+    /// Replace a temporary node.
     ///
     /// Call \a MDNode::replaceAllUsesWith() on \c N, replacing it with \c
     /// Replacement.
diff --git a/include/llvm/IR/DebugInfoMetadata.h b/include/llvm/IR/DebugInfoMetadata.h
index 5c99300c35c7..9c5a95721d79 100644
--- a/include/llvm/IR/DebugInfoMetadata.h
+++ b/include/llvm/IR/DebugInfoMetadata.h
@@ -1085,10 +1085,10 @@ public:
   /// deleted on a uniquing collision.  In practice, uniquing collisions on \a
   /// DICompileUnit should be fairly rare.
   /// @{
-  void replaceEnumTypes(DISubprogramArray N) {
+  void replaceEnumTypes(DICompositeTypeArray N) {
     replaceOperandWith(4, N.get());
   }
-  void replaceRetainedTypes(DISubprogramArray N) {
+  void replaceRetainedTypes(DITypeArray N) {
     replaceOperandWith(5, N.get());
   }
   void replaceSubprograms(DISubprogramArray N) {
@@ -1097,7 +1097,7 @@ public:
   void replaceGlobalVariables(DIGlobalVariableArray N) {
     replaceOperandWith(7, N.get());
   }
-  void replaceImportedEntities(DIGlobalVariableArray N) {
+  void replaceImportedEntities(DIImportedEntityArray N) {
     replaceOperandWith(8, N.get());
   }
   /// @}
@@ -1650,14 +1650,14 @@ class DIModule : public DIScope {
                            StorageType Storage, bool ShouldCreate = true) {
     return getImpl(Context, Scope, getCanonicalMDString(Context, Name),
                    getCanonicalMDString(Context, ConfigurationMacros),
-		   getCanonicalMDString(Context, IncludePath),
-		   getCanonicalMDString(Context, ISysRoot),
+                   getCanonicalMDString(Context, IncludePath),
+                   getCanonicalMDString(Context, ISysRoot),
                    Storage, ShouldCreate);
   }
   static DIModule *getImpl(LLVMContext &Context, Metadata *Scope,
                            MDString *Name, MDString *ConfigurationMacros,
-			   MDString *IncludePath, MDString *ISysRoot,
-			   StorageType Storage, bool ShouldCreate = true);
+                           MDString *IncludePath, MDString *ISysRoot,
+                           StorageType Storage, bool ShouldCreate = true);
 
   TempDIModule cloneImpl() const {
     return getTemporary(getContext(), getScope(), getName(),
@@ -1667,12 +1667,12 @@ class DIModule : public DIScope {
 
 public:
   DEFINE_MDNODE_GET(DIModule, (DIScope *Scope, StringRef Name,
-			       StringRef ConfigurationMacros, StringRef IncludePath,
-			       StringRef ISysRoot),
+                               StringRef ConfigurationMacros, StringRef IncludePath,
+                               StringRef ISysRoot),
                     (Scope, Name, ConfigurationMacros, IncludePath, ISysRoot))
   DEFINE_MDNODE_GET(DIModule,
                     (Metadata *Scope, MDString *Name, MDString *ConfigurationMacros,
-		     MDString *IncludePath, MDString *ISysRoot),
+                     MDString *IncludePath, MDString *ISysRoot),
                     (Scope, Name, ConfigurationMacros, IncludePath, ISysRoot))
 
   TempDIModule clone() const { return cloneImpl(); }
diff --git a/include/llvm/IR/Dominators.h b/include/llvm/IR/Dominators.h
index c1f208e3d72f..27d989b0344c 100644
--- a/include/llvm/IR/Dominators.h
+++ b/include/llvm/IR/Dominators.h
@@ -36,18 +36,14 @@ namespace llvm {
 template <typename IRUnitT> class AnalysisManager;
 class PreservedAnalyses;
 
-EXTERN_TEMPLATE_INSTANTIATION(class DomTreeNodeBase<BasicBlock>);
-EXTERN_TEMPLATE_INSTANTIATION(class DominatorTreeBase<BasicBlock>);
-
-#define LLVM_COMMA ,
-EXTERN_TEMPLATE_INSTANTIATION(void Calculate<Function LLVM_COMMA BasicBlock *>(
-    DominatorTreeBase<GraphTraits<BasicBlock *>::NodeType> &DT LLVM_COMMA
-        Function &F));
-EXTERN_TEMPLATE_INSTANTIATION(
-    void Calculate<Function LLVM_COMMA Inverse<BasicBlock *> >(
-        DominatorTreeBase<GraphTraits<Inverse<BasicBlock *> >::NodeType> &DT
-            LLVM_COMMA Function &F));
-#undef LLVM_COMMA
+extern template class DomTreeNodeBase<BasicBlock>;
+extern template class DominatorTreeBase<BasicBlock>;
+
+extern template void Calculate<Function, BasicBlock *>(
+    DominatorTreeBase<GraphTraits<BasicBlock *>::NodeType> &DT, Function &F);
+extern template void Calculate<Function, Inverse<BasicBlock *>>(
+    DominatorTreeBase<GraphTraits<Inverse<BasicBlock *>>::NodeType> &DT,
+    Function &F);
 
 typedef DomTreeNodeBase<BasicBlock> DomTreeNode;
 
diff --git a/include/llvm/IR/Function.h b/include/llvm/IR/Function.h
index 02ea056de39b..ec9f4cad094a 100644
--- a/include/llvm/IR/Function.h
+++ b/include/llvm/IR/Function.h
@@ -293,6 +293,16 @@ public:
     addFnAttr(Attribute::ReadOnly);
   }
 
+  /// @brief Determine if the call can access memmory only using pointers based
+  /// on its arguments.
+  bool onlyAccessesArgMemory() const {
+    return AttributeSets.hasAttribute(AttributeSet::FunctionIndex,
+                                      Attribute::ArgMemOnly);
+  }
+  void setOnlyAccessesArgMemory() {
+    addFnAttr(Attribute::ArgMemOnly);
+  }
+  
   /// @brief Determine if the function cannot return.
   bool doesNotReturn() const {
     return AttributeSets.hasAttribute(AttributeSet::FunctionIndex,
diff --git a/include/llvm/IR/GlobalValue.h b/include/llvm/IR/GlobalValue.h
index f2379705d460..2961369a7327 100644
--- a/include/llvm/IR/GlobalValue.h
+++ b/include/llvm/IR/GlobalValue.h
@@ -252,10 +252,9 @@ public:
   /// mistake: when working at the IR level use mayBeOverridden instead as it
   /// knows about ODR semantics.
   static bool isWeakForLinker(LinkageTypes Linkage)  {
-    return Linkage == AvailableExternallyLinkage || Linkage == WeakAnyLinkage ||
-           Linkage == WeakODRLinkage || Linkage == LinkOnceAnyLinkage ||
-           Linkage == LinkOnceODRLinkage || Linkage == CommonLinkage ||
-           Linkage == ExternalWeakLinkage;
+    return Linkage == WeakAnyLinkage || Linkage == WeakODRLinkage ||
+           Linkage == LinkOnceAnyLinkage || Linkage == LinkOnceODRLinkage ||
+           Linkage == CommonLinkage || Linkage == ExternalWeakLinkage;
   }
 
   bool hasExternalLinkage() const { return isExternalLinkage(Linkage); }
@@ -349,6 +348,12 @@ public:
     return isDeclaration();
   }
 
+  /// Returns true if this global's definition will be the one chosen by the
+  /// linker.
+  bool isStrongDefinitionForLinker() const {
+    return !(isDeclarationForLinker() || isWeakForLinker());
+  }
+
   /// This method unlinks 'this' from the containing module, but does not delete
   /// it.
   virtual void removeFromParent() = 0;
diff --git a/include/llvm/IR/IRBuilder.h b/include/llvm/IR/IRBuilder.h
index e6b5393c3397..6c67c79b6c0e 100644
--- a/include/llvm/IR/IRBuilder.h
+++ b/include/llvm/IR/IRBuilder.h
@@ -1382,47 +1382,61 @@ public:
     return CreateICmp(ICmpInst::ICMP_SLE, LHS, RHS, Name);
   }
 
-  Value *CreateFCmpOEQ(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_OEQ, LHS, RHS, Name);
+  Value *CreateFCmpOEQ(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_OEQ, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpOGT(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_OGT, LHS, RHS, Name);
+  Value *CreateFCmpOGT(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_OGT, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpOGE(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_OGE, LHS, RHS, Name);
+  Value *CreateFCmpOGE(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_OGE, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpOLT(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_OLT, LHS, RHS, Name);
+  Value *CreateFCmpOLT(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_OLT, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpOLE(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_OLE, LHS, RHS, Name);
+  Value *CreateFCmpOLE(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_OLE, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpONE(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_ONE, LHS, RHS, Name);
+  Value *CreateFCmpONE(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_ONE, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpORD(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_ORD, LHS, RHS, Name);
+  Value *CreateFCmpORD(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_ORD, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpUNO(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_UNO, LHS, RHS, Name);
+  Value *CreateFCmpUNO(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_UNO, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpUEQ(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_UEQ, LHS, RHS, Name);
+  Value *CreateFCmpUEQ(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_UEQ, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpUGT(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_UGT, LHS, RHS, Name);
+  Value *CreateFCmpUGT(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_UGT, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpUGE(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_UGE, LHS, RHS, Name);
+  Value *CreateFCmpUGE(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_UGE, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpULT(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_ULT, LHS, RHS, Name);
+  Value *CreateFCmpULT(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_ULT, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpULE(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_ULE, LHS, RHS, Name);
+  Value *CreateFCmpULE(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_ULE, LHS, RHS, Name, FPMathTag);
   }
-  Value *CreateFCmpUNE(Value *LHS, Value *RHS, const Twine &Name = "") {
-    return CreateFCmp(FCmpInst::FCMP_UNE, LHS, RHS, Name);
+  Value *CreateFCmpUNE(Value *LHS, Value *RHS, const Twine &Name = "",
+                       MDNode *FPMathTag = nullptr) {
+    return CreateFCmp(FCmpInst::FCMP_UNE, LHS, RHS, Name, FPMathTag);
   }
 
   Value *CreateICmp(CmpInst::Predicate P, Value *LHS, Value *RHS,
@@ -1433,11 +1447,12 @@ public:
     return Insert(new ICmpInst(P, LHS, RHS), Name);
   }
   Value *CreateFCmp(CmpInst::Predicate P, Value *LHS, Value *RHS,
-                    const Twine &Name = "") {
+                    const Twine &Name = "", MDNode *FPMathTag = nullptr) {
     if (Constant *LC = dyn_cast<Constant>(LHS))
       if (Constant *RC = dyn_cast<Constant>(RHS))
         return Insert(Folder.CreateFCmp(P, LC, RC), Name);
-    return Insert(new FCmpInst(P, LHS, RHS), Name);
+    return Insert(AddFPMathAttributes(new FCmpInst(P, LHS, RHS),
+                                      FPMathTag, FMF), Name);
   }
 
   //===--------------------------------------------------------------------===//
@@ -1449,7 +1464,7 @@ public:
     return Insert(PHINode::Create(Ty, NumReservedValues), Name);
   }
 
-  CallInst *CreateCall(Value *Callee, ArrayRef<Value *> Args,
+  CallInst *CreateCall(Value *Callee, ArrayRef<Value *> Args = None,
                        const Twine &Name = "") {
     return Insert(CallInst::Create(Callee, Args), Name);
   }
diff --git a/include/llvm/IR/Instruction.h b/include/llvm/IR/Instruction.h
index 6e3de1f13545..31f363f70a5b 100644
--- a/include/llvm/IR/Instruction.h
+++ b/include/llvm/IR/Instruction.h
@@ -382,7 +382,7 @@ public:
   ///
   /// Note that this does not consider malloc and alloca to have side
   /// effects because the newly allocated memory is completely invisible to
-  /// instructions which don't used the returned value.  For cases where this
+  /// instructions which don't use the returned value.  For cases where this
   /// matters, isSafeToSpeculativelyExecute may be more appropriate.
   bool mayHaveSideEffects() const {
     return mayWriteToMemory() || mayThrow() || !mayReturn();
diff --git a/include/llvm/IR/Instructions.h b/include/llvm/IR/Instructions.h
index c5890f01ea70..07d5f111b9e1 100644
--- a/include/llvm/IR/Instructions.h
+++ b/include/llvm/IR/Instructions.h
@@ -990,10 +990,14 @@ public:
                                    Ptr->getType()->getPointerAddressSpace());
     // Vector GEP
     if (Ptr->getType()->isVectorTy()) {
-      unsigned NumElem = cast<VectorType>(Ptr->getType())->getNumElements();
+      unsigned NumElem = Ptr->getType()->getVectorNumElements();
       return VectorType::get(PtrTy, NumElem);
     }
-
+    for (Value *Index : IdxList)
+      if (Index->getType()->isVectorTy()) {
+        unsigned NumElem = Index->getType()->getVectorNumElements();
+        return VectorType::get(PtrTy, NumElem);
+      }
     // Scalar GEP
     return PtrTy;
   }
@@ -1591,6 +1595,15 @@ public:
     addAttribute(AttributeSet::FunctionIndex, Attribute::ReadOnly);
   }
 
+  /// @brief Determine if the call can access memmory only using pointers based
+  /// on its arguments.
+  bool onlyAccessesArgMemory() const {
+    return hasFnAttr(Attribute::ArgMemOnly);
+  }
+  void setOnlyAccessesArgMemory() {
+    addAttribute(AttributeSet::FunctionIndex, Attribute::ArgMemOnly);
+  }
+
   /// \brief Determine if the call cannot return.
   bool doesNotReturn() const { return hasFnAttr(Attribute::NoReturn); }
   void setDoesNotReturn() {
@@ -3360,6 +3373,15 @@ public:
     addAttribute(AttributeSet::FunctionIndex, Attribute::ReadOnly);
   }
 
+  /// @brief Determine if the call access memmory only using it's pointer
+  /// arguments.
+  bool onlyAccessesArgMemory() const {
+    return hasFnAttr(Attribute::ArgMemOnly);
+  }
+  void setOnlyAccessesArgMemory() {
+    addAttribute(AttributeSet::FunctionIndex, Attribute::ArgMemOnly);
+  }
+
   /// \brief Determine if the call cannot return.
   bool doesNotReturn() const { return hasFnAttr(Attribute::NoReturn); }
   void setDoesNotReturn() {
diff --git a/include/llvm/IR/Intrinsics.td b/include/llvm/IR/Intrinsics.td
index e6f6d0ffe8b6..bbae720b4e12 100644
--- a/include/llvm/IR/Intrinsics.td
+++ b/include/llvm/IR/Intrinsics.td
@@ -268,15 +268,23 @@ def int_gcwrite : Intrinsic<[],
 //
 def int_returnaddress : Intrinsic<[llvm_ptr_ty], [llvm_i32_ty], [IntrNoMem]>;
 def int_frameaddress  : Intrinsic<[llvm_ptr_ty], [llvm_i32_ty], [IntrNoMem]>;
-def int_frameescape : Intrinsic<[], [llvm_vararg_ty]>;
-def int_framerecover : Intrinsic<[llvm_ptr_ty],
-                                 [llvm_ptr_ty, llvm_ptr_ty, llvm_i32_ty],
-                                 [IntrNoMem]>;
 def int_read_register  : Intrinsic<[llvm_anyint_ty], [llvm_metadata_ty],
                                    [IntrReadMem], "llvm.read_register">;
 def int_write_register : Intrinsic<[], [llvm_metadata_ty, llvm_anyint_ty],
                                    [], "llvm.write_register">;
 
+// Gets the address of the local variable area. This is typically a copy of the
+// stack, frame, or base pointer depending on the type of prologue.
+def int_localaddress : Intrinsic<[llvm_ptr_ty], [], [IntrNoMem]>;
+
+// Escapes local variables to allow access from other functions.
+def int_localescape : Intrinsic<[], [llvm_vararg_ty]>;
+
+// Given a function and the localaddress of a parent frame, returns a pointer
+// to an escaped allocation indicated by the index.
+def int_localrecover : Intrinsic<[llvm_ptr_ty],
+                                 [llvm_ptr_ty, llvm_ptr_ty, llvm_i32_ty],
+                                 [IntrNoMem]>;
 // Note: we treat stacksave/stackrestore as writemem because we don't otherwise
 // model their dependencies on allocas.
 def int_stacksave     : Intrinsic<[llvm_ptr_ty]>,
@@ -362,6 +370,8 @@ let Properties = [IntrNoMem] in {
   def int_rint  : Intrinsic<[llvm_anyfloat_ty], [LLVMMatchType<0>]>;
   def int_nearbyint : Intrinsic<[llvm_anyfloat_ty], [LLVMMatchType<0>]>;
   def int_round : Intrinsic<[llvm_anyfloat_ty], [LLVMMatchType<0>]>;
+  def int_canonicalize : Intrinsic<[llvm_anyfloat_ty], [LLVMMatchType<0>],
+                                   [IntrNoMem]>;
 }
 
 // NOTE: these are internal interfaces.
@@ -638,3 +648,4 @@ include "llvm/IR/IntrinsicsMips.td"
 include "llvm/IR/IntrinsicsAMDGPU.td"
 include "llvm/IR/IntrinsicsBPF.td"
 include "llvm/IR/IntrinsicsSystemZ.td"
+include "llvm/IR/IntrinsicsWebAssembly.td"
diff --git a/include/llvm/IR/IntrinsicsPowerPC.td b/include/llvm/IR/IntrinsicsPowerPC.td
index 05adc5a757be..eb8f1e6cd079 100644
--- a/include/llvm/IR/IntrinsicsPowerPC.td
+++ b/include/llvm/IR/IntrinsicsPowerPC.td
@@ -694,6 +694,18 @@ def int_ppc_vsx_xvrspip :
 def int_ppc_vsx_xvrdpip :
       Intrinsic<[llvm_v2f64_ty], [llvm_v2f64_ty], [IntrNoMem]>;
 
+// Vector reciprocal estimate
+def int_ppc_vsx_xvresp : GCCBuiltin<"__builtin_vsx_xvresp">,
+      Intrinsic<[llvm_v4f32_ty], [llvm_v4f32_ty], [IntrNoMem]>;
+def int_ppc_vsx_xvredp : GCCBuiltin<"__builtin_vsx_xvredp">,
+      Intrinsic<[llvm_v2f64_ty], [llvm_v2f64_ty], [IntrNoMem]>;
+
+// Vector rsqrte
+def int_ppc_vsx_xvrsqrtesp : GCCBuiltin<"__builtin_vsx_xvrsqrtesp">,
+      Intrinsic<[llvm_v4f32_ty], [llvm_v4f32_ty], [IntrNoMem]>;
+def int_ppc_vsx_xvrsqrtedp : GCCBuiltin<"__builtin_vsx_xvrsqrtedp">,
+      Intrinsic<[llvm_v2f64_ty], [llvm_v2f64_ty], [IntrNoMem]>;
+
 // Vector compare
 def int_ppc_vsx_xvcmpeqdp :
       PowerPC_VSX_Intrinsic<"xvcmpeqdp", [llvm_v2i64_ty],
@@ -713,6 +725,9 @@ def int_ppc_vsx_xvcmpgtdp :
 def int_ppc_vsx_xvcmpgtsp :
       PowerPC_VSX_Intrinsic<"xvcmpgtsp", [llvm_v4i32_ty],
                             [llvm_v4f32_ty, llvm_v4f32_ty], [IntrNoMem]>;
+def int_ppc_vsx_xxleqv :
+      PowerPC_VSX_Intrinsic<"xxleqv", [llvm_v4i32_ty],
+                            [llvm_v4i32_ty, llvm_v4i32_ty], [IntrNoMem]>;
 }
 
 //===----------------------------------------------------------------------===//
diff --git a/include/llvm/IR/IntrinsicsWebAssembly.td b/include/llvm/IR/IntrinsicsWebAssembly.td
new file mode 100644
index 000000000000..3ccde4742384
--- /dev/null
+++ b/include/llvm/IR/IntrinsicsWebAssembly.td
@@ -0,0 +1,16 @@
+//===- IntrinsicsWebAssembly.td - Defines wasm intrinsics --*- tablegen -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// \brief This file defines all of the WebAssembly-specific intrinsics.
+///
+//===----------------------------------------------------------------------===//
+
+let TargetPrefix = "wasm" in {  // All intrinsics start with "llvm.wasm.".
+}
diff --git a/include/llvm/IR/IntrinsicsX86.td b/include/llvm/IR/IntrinsicsX86.td
index b90825db93cd..a3bc4af84308 100644
--- a/include/llvm/IR/IntrinsicsX86.td
+++ b/include/llvm/IR/IntrinsicsX86.td
@@ -28,7 +28,7 @@ let TargetPrefix = "x86" in {
   def int_x86_seh_restoreframe : Intrinsic<[], [], []>;
 
   // Given a pointer to the end of an EH registration object, returns the true
-  // parent frame address that can be used with llvm.framerecover.
+  // parent frame address that can be used with llvm.localrecover.
   def int_x86_seh_recoverfp : Intrinsic<[llvm_ptr_ty],
                                         [llvm_ptr_ty, llvm_ptr_ty],
                                         [IntrNoMem]>;
@@ -2107,6 +2107,15 @@ let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
   def int_x86_avx2_pmul_hr_sw : GCCBuiltin<"__builtin_ia32_pmulhrsw256">,
               Intrinsic<[llvm_v16i16_ty], [llvm_v16i16_ty,
                          llvm_v16i16_ty], [IntrNoMem, Commutative]>;
+  def int_x86_avx512_mask_pmul_hr_sw_128 : GCCBuiltin<"__builtin_ia32_pmulhrsw128_mask">,
+              Intrinsic<[llvm_v8i16_ty], [llvm_v8i16_ty, llvm_v8i16_ty, 
+                     llvm_v8i16_ty, llvm_i8_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_pmul_hr_sw_256 : GCCBuiltin<"__builtin_ia32_pmulhrsw256_mask">,
+              Intrinsic<[llvm_v16i16_ty], [llvm_v16i16_ty, llvm_v16i16_ty, 
+                     llvm_v16i16_ty, llvm_i16_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_pmul_hr_sw_512 : GCCBuiltin<"__builtin_ia32_pmulhrsw512_mask">,
+              Intrinsic<[llvm_v32i16_ty], [llvm_v32i16_ty, llvm_v32i16_ty, 
+                     llvm_v32i16_ty, llvm_i32_ty], [IntrNoMem]>;
 }
 
 // Vector sign and zero extend
@@ -4466,6 +4475,24 @@ let TargetPrefix = "x86" in {
   def int_x86_avx512_mask_pmull_q_512 : GCCBuiltin<"__builtin_ia32_pmullq512_mask">,
               Intrinsic<[llvm_v8i64_ty], [llvm_v8i64_ty, llvm_v8i64_ty,
                      llvm_v8i64_ty, llvm_i8_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_pmulhu_w_512 : GCCBuiltin<"__builtin_ia32_pmulhuw512_mask">,
+              Intrinsic<[llvm_v32i16_ty], [llvm_v32i16_ty, llvm_v32i16_ty,
+                     llvm_v32i16_ty, llvm_i32_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_pmulh_w_512 : GCCBuiltin<"__builtin_ia32_pmulhw512_mask">,
+              Intrinsic<[llvm_v32i16_ty], [llvm_v32i16_ty, llvm_v32i16_ty,
+                     llvm_v32i16_ty, llvm_i32_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_pmulhu_w_128 : GCCBuiltin<"__builtin_ia32_pmulhuw128_mask">,
+              Intrinsic<[llvm_v8i16_ty], [llvm_v8i16_ty, llvm_v8i16_ty, 
+                     llvm_v8i16_ty, llvm_i8_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_pmulhu_w_256 : GCCBuiltin<"__builtin_ia32_pmulhuw256_mask">,
+              Intrinsic<[llvm_v16i16_ty], [llvm_v16i16_ty, llvm_v16i16_ty, 
+                     llvm_v16i16_ty, llvm_i16_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_pmulh_w_128 : GCCBuiltin<"__builtin_ia32_pmulhw128_mask">,
+              Intrinsic<[llvm_v8i16_ty], [llvm_v8i16_ty, llvm_v8i16_ty, 
+                     llvm_v8i16_ty, llvm_i8_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_pmulh_w_256 : GCCBuiltin<"__builtin_ia32_pmulhw256_mask">,
+              Intrinsic<[llvm_v16i16_ty], [llvm_v16i16_ty, llvm_v16i16_ty, 
+                     llvm_v16i16_ty, llvm_i16_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_pavg_b_512 : GCCBuiltin<"__builtin_ia32_pavgb512_mask">,
           Intrinsic<[llvm_v64i8_ty], [llvm_v64i8_ty, llvm_v64i8_ty, 
                     llvm_v64i8_ty, llvm_i64_ty], [IntrNoMem]>;
diff --git a/include/llvm/IR/Operator.h b/include/llvm/IR/Operator.h
index 1b9102ecc7e4..372b254ab183 100644
--- a/include/llvm/IR/Operator.h
+++ b/include/llvm/IR/Operator.h
@@ -305,7 +305,8 @@ public:
   float getFPAccuracy() const;
 
   static inline bool classof(const Instruction *I) {
-    return I->getType()->isFPOrFPVectorTy();
+    return I->getType()->isFPOrFPVectorTy() ||
+      I->getOpcode() == Instruction::FCmp;
   }
   static inline bool classof(const Value *V) {
     return isa<Instruction>(V) && classof(cast<Instruction>(V));
diff --git a/include/llvm/IR/Value.h b/include/llvm/IR/Value.h
index 484afc6d232c..17a80c82d1bc 100644
--- a/include/llvm/IR/Value.h
+++ b/include/llvm/IR/Value.h
@@ -104,8 +104,8 @@ protected:
   ///
   /// Note, this should *NOT* be used directly by any class other than User.
   /// User uses this value to find the Use list.
-  static const unsigned NumUserOperandsBits = 29;
-  unsigned NumUserOperands : 29;
+  enum : unsigned { NumUserOperandsBits = 29 };
+  unsigned NumUserOperands : NumUserOperandsBits;
 
   bool IsUsedByMD : 1;
   bool HasName : 1;
diff --git a/include/llvm/InitializePasses.h b/include/llvm/InitializePasses.h
index 74fbc0f94b03..e3b9a95f0a3d 100644
--- a/include/llvm/InitializePasses.h
+++ b/include/llvm/InitializePasses.h
@@ -130,6 +130,7 @@ void initializeSanitizerCoverageModulePass(PassRegistry&);
 void initializeDataFlowSanitizerPass(PassRegistry&);
 void initializeScalarizerPass(PassRegistry&);
 void initializeEarlyCSELegacyPassPass(PassRegistry &);
+void initializeEliminateAvailableExternallyPass(PassRegistry&);
 void initializeExpandISelPseudosPass(PassRegistry&);
 void initializeFunctionAttrsPass(PassRegistry&);
 void initializeGCMachineCodeAnalysisPass(PassRegistry&);
@@ -302,6 +303,7 @@ void initializePlaceSafepointsPass(PassRegistry&);
 void initializeDwarfEHPreparePass(PassRegistry&);
 void initializeFloat2IntPass(PassRegistry&);
 void initializeLoopDistributePass(PassRegistry&);
+void initializeSjLjEHPreparePass(PassRegistry&);
 }
 
 #endif
diff --git a/include/llvm/LinkAllPasses.h b/include/llvm/LinkAllPasses.h
index 8ac1b212ae5f..cea5530db3b8 100644
--- a/include/llvm/LinkAllPasses.h
+++ b/include/llvm/LinkAllPasses.h
@@ -176,6 +176,7 @@ namespace {
       (void) llvm::createStraightLineStrengthReducePass();
       (void) llvm::createMemDerefPrinter();
       (void) llvm::createFloat2IntPass();
+      (void) llvm::createEliminateAvailableExternallyPass();
 
       (void)new llvm::IntervalPartition();
       (void)new llvm::ScalarEvolution();
diff --git a/include/llvm/MC/MCContext.h b/include/llvm/MC/MCContext.h
index 52017fda189b..41169e9a12a0 100644
--- a/include/llvm/MC/MCContext.h
+++ b/include/llvm/MC/MCContext.h
@@ -273,7 +273,7 @@ namespace llvm {
     /// Gets a symbol that will be defined to the final stack offset of a local
     /// variable after codegen.
     ///
-    /// \param Idx - The index of a local variable passed to @llvm.frameescape.
+    /// \param Idx - The index of a local variable passed to @llvm.localescape.
     MCSymbol *getOrCreateFrameAllocSymbol(StringRef FuncName, unsigned Idx);
 
     MCSymbol *getOrCreateParentFrameOffsetSymbol(StringRef FuncName);
diff --git a/include/llvm/MC/MCDwarf.h b/include/llvm/MC/MCDwarf.h
index c7bed8eccda9..1e72dfee4ad1 100644
--- a/include/llvm/MC/MCDwarf.h
+++ b/include/llvm/MC/MCDwarf.h
@@ -54,13 +54,13 @@ struct MCDwarfFile {
 /// \brief Instances of this class represent the information from a
 /// dwarf .loc directive.
 class MCDwarfLoc {
-  unsigned FileNum;
-  unsigned Line;
-  unsigned Column;
+  uint32_t FileNum;
+  uint32_t Line;
+  uint16_t Column;
   // Flags (see #define's below)
-  unsigned Flags;
-  unsigned Isa;
-  unsigned Discriminator;
+  uint8_t Flags;
+  uint8_t Isa;
+  uint32_t Discriminator;
 
 // Flag that indicates the initial value of the is_stmt_start flag.
 #define DWARF2_LINE_DEFAULT_IS_STMT 1
@@ -107,13 +107,22 @@ public:
   void setLine(unsigned line) { Line = line; }
 
   /// \brief Set the Column of this MCDwarfLoc.
-  void setColumn(unsigned column) { Column = column; }
+  void setColumn(unsigned column) {
+    assert(column <= UINT16_MAX);
+    Column = column;
+  }
 
   /// \brief Set the Flags of this MCDwarfLoc.
-  void setFlags(unsigned flags) { Flags = flags; }
+  void setFlags(unsigned flags) {
+    assert(flags <= UINT8_MAX);
+    Flags = flags;
+  }
 
   /// \brief Set the Isa of this MCDwarfLoc.
-  void setIsa(unsigned isa) { Isa = isa; }
+  void setIsa(unsigned isa) {
+    assert(isa <= UINT8_MAX);
+    Isa = isa;
+  }
 
   /// \brief Set the Discriminator of this MCDwarfLoc.
   void setDiscriminator(unsigned discriminator) {
diff --git a/include/llvm/MC/MCInstrDesc.h b/include/llvm/MC/MCInstrDesc.h
index 3209a2ce0408..6a582e82d00e 100644
--- a/include/llvm/MC/MCInstrDesc.h
+++ b/include/llvm/MC/MCInstrDesc.h
@@ -154,7 +154,8 @@ public:
 
   // A complex method to determine is a certain is deprecated or not, and return
   // the reason for deprecation.
-  bool (*ComplexDeprecationInfo)(MCInst &, MCSubtargetInfo &, std::string &);
+  bool (*ComplexDeprecationInfo)(MCInst &, const MCSubtargetInfo &,
+                                 std::string &);
 
   /// \brief Returns the value of the specific constraint if
   /// it is set. Returns -1 if it is not set.
@@ -170,7 +171,7 @@ public:
 
   /// \brief Returns true if a certain instruction is deprecated and if so
   /// returns the reason in \p Info.
-  bool getDeprecatedInfo(MCInst &MI, MCSubtargetInfo &STI,
+  bool getDeprecatedInfo(MCInst &MI, const MCSubtargetInfo &STI,
                          std::string &Info) const;
 
   /// \brief Return the opcode number for this descriptor.
diff --git a/include/llvm/MC/MCSchedule.h b/include/llvm/MC/MCSchedule.h
index 1adfedd2638a..c09791631056 100644
--- a/include/llvm/MC/MCSchedule.h
+++ b/include/llvm/MC/MCSchedule.h
@@ -224,25 +224,9 @@ struct MCSchedModel {
     return &SchedClassTable[SchedClassIdx];
   }
 
-  // /\brief Returns a default initialized model. Used for unknown processors.
-  static MCSchedModel GetDefaultSchedModel() {
-    MCSchedModel Ret = { DefaultIssueWidth,
-                         DefaultMicroOpBufferSize,
-                         DefaultLoopMicroOpBufferSize,
-                         DefaultLoadLatency,
-                         DefaultHighLatency,
-                         DefaultMispredictPenalty,
-                         false,
-                         true,
-                         0,
-                         nullptr,
-                         nullptr,
-                         0,
-                         0,
-                         nullptr
-                       };
-    return Ret;
-  }
+  /// Returns the default initialized model.
+  static const MCSchedModel &GetDefaultSchedModel() { return Default; }
+  static const MCSchedModel Default;
 };
 
 } // End llvm namespace
diff --git a/include/llvm/MC/MCSubtargetInfo.h b/include/llvm/MC/MCSubtargetInfo.h
index b8ad02fbe696..d5ad4eebf9ef 100644
--- a/include/llvm/MC/MCSubtargetInfo.h
+++ b/include/llvm/MC/MCSubtargetInfo.h
@@ -37,22 +37,26 @@ class MCSubtargetInfo {
   const MCWriteProcResEntry *WriteProcResTable;
   const MCWriteLatencyEntry *WriteLatencyTable;
   const MCReadAdvanceEntry *ReadAdvanceTable;
-  MCSchedModel CPUSchedModel;
+  const MCSchedModel *CPUSchedModel;
 
   const InstrStage *Stages;            // Instruction itinerary stages
   const unsigned *OperandCycles;       // Itinerary operand cycles
   const unsigned *ForwardingPaths;     // Forwarding paths
   FeatureBitset FeatureBits;           // Feature bits for current CPU + FS
 
+  MCSubtargetInfo() = delete;
+  MCSubtargetInfo &operator=(MCSubtargetInfo &&) = delete;
+  MCSubtargetInfo &operator=(const MCSubtargetInfo &) = delete;
+
 public:
-  void InitMCSubtargetInfo(const Triple &TT, StringRef CPU, StringRef FS,
-                           ArrayRef<SubtargetFeatureKV> PF,
-                           ArrayRef<SubtargetFeatureKV> PD,
-                           const SubtargetInfoKV *ProcSched,
-                           const MCWriteProcResEntry *WPR,
-                           const MCWriteLatencyEntry *WL,
-                           const MCReadAdvanceEntry *RA, const InstrStage *IS,
-                           const unsigned *OC, const unsigned *FP);
+  MCSubtargetInfo(const MCSubtargetInfo &) = default;
+  MCSubtargetInfo(const Triple &TT, StringRef CPU, StringRef FS,
+                  ArrayRef<SubtargetFeatureKV> PF,
+                  ArrayRef<SubtargetFeatureKV> PD,
+                  const SubtargetInfoKV *ProcSched,
+                  const MCWriteProcResEntry *WPR, const MCWriteLatencyEntry *WL,
+                  const MCReadAdvanceEntry *RA, const InstrStage *IS,
+                  const unsigned *OC, const unsigned *FP);
 
   /// getTargetTriple - Return the target triple string.
   const Triple &getTargetTriple() const { return TargetTriple; }
@@ -74,12 +78,16 @@ public:
     FeatureBits = FeatureBits_;
   }
 
-  /// InitMCProcessorInfo - Set or change the CPU (optionally supplemented with
-  /// feature string). Recompute feature bits and scheduling model.
+protected:
+  /// Initialize the scheduling model and feature bits.
+  ///
+  /// FIXME: Find a way to stick this in the constructor, since it should only
+  /// be called during initialization.
   void InitMCProcessorInfo(StringRef CPU, StringRef FS);
 
-  /// InitCPUSchedModel - Recompute scheduling model based on CPU.
-  void InitCPUSchedModel(StringRef CPU);
+public:
+  /// Set the features to the default for the given CPU.
+  void setDefaultFeatures(StringRef CPU);
 
   /// ToggleFeature - Toggle a feature and returns the re-computed feature
   /// bits. This version does not change the implied bits.
@@ -99,11 +107,10 @@ public:
 
   /// getSchedModelForCPU - Get the machine model of a CPU.
   ///
-  MCSchedModel getSchedModelForCPU(StringRef CPU) const;
+  const MCSchedModel &getSchedModelForCPU(StringRef CPU) const;
 
-  /// getSchedModel - Get the machine model for this subtarget's CPU.
-  ///
-  const MCSchedModel &getSchedModel() const { return CPUSchedModel; }
+  /// Get the machine model for this subtarget's CPU.
+  const MCSchedModel &getSchedModel() const { return *CPUSchedModel; }
 
   /// Return an iterator at the first process resource consumed by the given
   /// scheduling class.
@@ -151,7 +158,7 @@ public:
   void initInstrItins(InstrItineraryData &InstrItins) const;
 
   /// Check whether the CPU string is valid.
-  bool isCPUStringValid(StringRef CPU) {
+  bool isCPUStringValid(StringRef CPU) const {
     auto Found = std::find_if(ProcDesc.begin(), ProcDesc.end(),
                               [=](const SubtargetFeatureKV &KV) {
                                 return CPU == KV.Key; 
diff --git a/include/llvm/MC/MCSymbol.h b/include/llvm/MC/MCSymbol.h
index 17e6b857cf20..b2910dfccd63 100644
--- a/include/llvm/MC/MCSymbol.h
+++ b/include/llvm/MC/MCSymbol.h
@@ -114,12 +114,12 @@ protected:
   /// The alignment is stored as log2(align) + 1.  This allows all values from
   /// 0 to 2^31 to be stored which is every power of 2 representable by an
   /// unsigned.
-  static const unsigned NumCommonAlignmentBits = 5;
+  enum : unsigned { NumCommonAlignmentBits = 5 };
   unsigned CommonAlignLog2 : NumCommonAlignmentBits;
 
   /// The Flags field is used by object file implementations to store
   /// additional per symbol information which is not easily classified.
-  static const unsigned NumFlagsBits = 16;
+  enum : unsigned { NumFlagsBits = 16 };
   mutable uint32_t Flags : NumFlagsBits;
 
   /// Index field, for use by the object file implementation.
diff --git a/include/llvm/MC/MCSymbolMachO.h b/include/llvm/MC/MCSymbolMachO.h
index 166ae9e755a1..5b0321fe9f73 100644
--- a/include/llvm/MC/MCSymbolMachO.h
+++ b/include/llvm/MC/MCSymbolMachO.h
@@ -7,7 +7,7 @@
 //
 //===----------------------------------------------------------------------===//
 #ifndef LLVM_MC_MCSYMBOLMACHO_H
-#define setIsWeakExternal
+#define LLVM_MC_MCSYMBOLMACHO_H
 
 #include "llvm/MC/MCSymbol.h"
 
diff --git a/include/llvm/MC/MCTargetOptions.h b/include/llvm/MC/MCTargetOptions.h
index ce28a196e974..7f4f23eda27f 100644
--- a/include/llvm/MC/MCTargetOptions.h
+++ b/include/llvm/MC/MCTargetOptions.h
@@ -55,7 +55,7 @@ inline bool operator==(const MCTargetOptions &LHS, const MCTargetOptions &RHS) {
           ARE_EQUAL(ShowMCInst) &&
           ARE_EQUAL(AsmVerbose) &&
           ARE_EQUAL(DwarfVersion) &&
-	  ARE_EQUAL(ABIName));
+          ARE_EQUAL(ABIName));
 #undef ARE_EQUAL
 }
 
diff --git a/include/llvm/Object/Archive.h b/include/llvm/Object/Archive.h
index 8da6919a4655..597f0d48c118 100644
--- a/include/llvm/Object/Archive.h
+++ b/include/llvm/Object/Archive.h
@@ -94,9 +94,7 @@ public:
     /// \return the size in the archive header for this member.
     uint64_t getRawSize() const;
 
-    StringRef getBuffer() const {
-      return StringRef(Data.data() + StartOfFile, getSize());
-    }
+    ErrorOr<StringRef> getBuffer() const;
     uint64_t getChildOffset() const;
 
     ErrorOr<MemoryBufferRef> getMemoryBufferRef() const;
@@ -183,6 +181,7 @@ public:
   };
 
   Kind kind() const { return (Kind)Format; }
+  bool isThin() const { return IsThin; }
 
   child_iterator child_begin(bool SkipInternal = true) const;
   child_iterator child_end() const;
@@ -207,6 +206,11 @@ public:
 
   bool hasSymbolTable() const;
   child_iterator getSymbolTableChild() const { return SymbolTable; }
+  StringRef getSymbolTable() const {
+    // We know that the symbol table is not an external file,
+    // so we just assert there is no error.
+    return *SymbolTable->getBuffer();
+  }
   uint32_t getNumberOfSymbols() const;
 
 private:
@@ -215,6 +219,7 @@ private:
   child_iterator FirstRegular;
   unsigned Format : 2;
   unsigned IsThin : 1;
+  mutable std::vector<std::unique_ptr<MemoryBuffer>> ThinBuffers;
 };
 
 }
diff --git a/include/llvm/Object/ArchiveWriter.h b/include/llvm/Object/ArchiveWriter.h
index 1616e46d3e6f..3648d0c77fb5 100644
--- a/include/llvm/Object/ArchiveWriter.h
+++ b/include/llvm/Object/ArchiveWriter.h
@@ -31,7 +31,6 @@ class NewArchiveIterator {
 public:
   NewArchiveIterator(object::Archive::child_iterator I, StringRef Name);
   NewArchiveIterator(StringRef I, StringRef Name);
-  NewArchiveIterator();
   bool isNewMember() const;
   StringRef getName() const;
 
@@ -44,8 +43,7 @@ public:
 
 std::pair<StringRef, std::error_code>
 writeArchive(StringRef ArcName, std::vector<NewArchiveIterator> &NewMembers,
-             bool WriteSymtab);
-
+             bool WriteSymtab, object::Archive::Kind Kind, bool Deterministic);
 }
 
 #endif
diff --git a/include/llvm/Object/COFF.h b/include/llvm/Object/COFF.h
index fc605826a8b0..025a9dbc6bc0 100644
--- a/include/llvm/Object/COFF.h
+++ b/include/llvm/Object/COFF.h
@@ -474,7 +474,7 @@ struct coff_import_header {
   support::ulittle16_t OrdinalHint;
   support::ulittle16_t TypeInfo;
   int getType() const { return TypeInfo & 0x3; }
-  int getNameType() const { return (TypeInfo & 0x7) >> 2; }
+  int getNameType() const { return (TypeInfo >> 2) & 0x7; }
 };
 
 struct coff_import_directory_table_entry {
@@ -648,9 +648,8 @@ public:
 protected:
   void moveSymbolNext(DataRefImpl &Symb) const override;
   ErrorOr<StringRef> getSymbolName(DataRefImpl Symb) const override;
-  std::error_code getSymbolAddress(DataRefImpl Symb,
-                                   uint64_t &Res) const override;
-  uint64_t getSymbolValue(DataRefImpl Symb) const override;
+  ErrorOr<uint64_t> getSymbolAddress(DataRefImpl Symb) const override;
+  uint64_t getSymbolValueImpl(DataRefImpl Symb) const override;
   uint64_t getCommonSymbolSizeImpl(DataRefImpl Symb) const override;
   uint32_t getSymbolFlags(DataRefImpl Symb) const override;
   SymbolRef::Type getSymbolType(DataRefImpl Symb) const override;
@@ -672,7 +671,6 @@ protected:
   relocation_iterator section_rel_end(DataRefImpl Sec) const override;
 
   void moveRelocationNext(DataRefImpl &Rel) const override;
-  ErrorOr<uint64_t> getRelocationAddress(DataRefImpl Rel) const override;
   uint64_t getRelocationOffset(DataRefImpl Rel) const override;
   symbol_iterator getRelocationSymbol(DataRefImpl Rel) const override;
   uint64_t getRelocationType(DataRefImpl Rel) const override;
diff --git a/include/llvm/Object/ELF.h b/include/llvm/Object/ELF.h
index 3b0c548ffe15..cc271851e6b0 100644
--- a/include/llvm/Object/ELF.h
+++ b/include/llvm/Object/ELF.h
@@ -16,6 +16,7 @@
 
 #include "llvm/ADT/ArrayRef.h"
 #include "llvm/ADT/DenseMap.h"
+#include "llvm/ADT/IntervalMap.h"
 #include "llvm/ADT/PointerIntPair.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/ADT/StringSwitch.h"
@@ -139,6 +140,7 @@ public:
   typedef Elf_Verneed_Impl<ELFT> Elf_Verneed;
   typedef Elf_Vernaux_Impl<ELFT> Elf_Vernaux;
   typedef Elf_Versym_Impl<ELFT> Elf_Versym;
+  typedef Elf_Hash_Impl<ELFT> Elf_Hash;
   typedef ELFEntityIterator<const Elf_Dyn> Elf_Dyn_Iter;
   typedef iterator_range<Elf_Dyn_Iter> Elf_Dyn_Range;
   typedef ELFEntityIterator<const Elf_Rela> Elf_Rela_Iter;
@@ -174,8 +176,8 @@ private:
   StringRef DotShstrtab;                    // Section header string table.
   StringRef DotStrtab;                      // Symbol header string table.
   const Elf_Shdr *dot_symtab_sec = nullptr; // Symbol table section.
-  StringRef DynSymStrTab;                   // Dynnamic symbol string table.
   const Elf_Shdr *DotDynSymSec = nullptr;   // Dynamic symbol table section.
+  const Elf_Hash *HashTable = nullptr;
 
   const Elf_Shdr *SymbolTableSectionHeaderIndex = nullptr;
   DenseMap<const Elf_Sym *, ELF::Elf64_Word> ExtendedSymbolTable;
@@ -197,6 +199,7 @@ private:
 
   DynRegionInfo DynamicRegion;
   DynRegionInfo DynHashRegion;
+  DynRegionInfo DynStrRegion;
   DynRegionInfo DynRelaRegion;
 
   // Pointer to SONAME entry in dynamic string table
@@ -229,6 +232,8 @@ private:
   void LoadVersionNeeds(const Elf_Shdr *ec) const;
   void LoadVersionMap() const;
 
+  void scanDynamicTable();
+
 public:
   template<typename T>
   const T        *getEntry(uint32_t Section, uint32_t Entry) const;
@@ -237,6 +242,7 @@ public:
 
   const Elf_Shdr *getDotSymtabSec() const { return dot_symtab_sec; }
   const Elf_Shdr *getDotDynSymSec() const { return DotDynSymSec; }
+  const Elf_Hash *getHashTable() const { return HashTable; }
 
   ErrorOr<StringRef> getStringTable(const Elf_Shdr *Section) const;
   const char *getDynamicString(uintX_t Offset) const;
@@ -578,8 +584,10 @@ ELFFile<ELFT>::ELFFile(StringRef Object, std::error_code &EC)
 
   Header = reinterpret_cast<const Elf_Ehdr *>(base());
 
-  if (Header->e_shoff == 0)
+  if (Header->e_shoff == 0) {
+    scanDynamicTable();
     return;
+  }
 
   const uint64_t SectionTableOffset = Header->e_shoff;
 
@@ -604,6 +612,13 @@ ELFFile<ELFT>::ELFFile(StringRef Object, std::error_code &EC)
 
   for (const Elf_Shdr &Sec : sections()) {
     switch (Sec.sh_type) {
+    case ELF::SHT_HASH:
+      if (HashTable) {
+        EC = object_error::parse_failed;
+        return;
+      }
+      HashTable = reinterpret_cast<const Elf_Hash *>(base() + Sec.sh_offset);
+      break;
     case ELF::SHT_SYMTAB_SHNDX:
       if (SymbolTableSectionHeaderIndex) {
         // More than one .symtab_shndx!
@@ -640,7 +655,9 @@ ELFFile<ELFT>::ELFFile(StringRef Object, std::error_code &EC)
       ErrorOr<StringRef> SymtabOrErr = getStringTable(*SectionOrErr);
       if ((EC = SymtabOrErr.getError()))
         return;
-      DynSymStrTab = *SymtabOrErr;
+      DynStrRegion.Addr = SymtabOrErr->data();
+      DynStrRegion.Size = SymtabOrErr->size();
+      DynStrRegion.EntSize = 1;
       break;
     }
     case ELF::SHT_DYNAMIC:
@@ -701,7 +718,23 @@ ELFFile<ELFT>::ELFFile(StringRef Object, std::error_code &EC)
     }
   }
 
-  // Scan program headers.
+  scanDynamicTable();
+
+  EC = std::error_code();
+}
+
+template <class ELFT>
+void ELFFile<ELFT>::scanDynamicTable() {
+  // Build load-address to file-offset map.
+  typedef IntervalMap<
+      uintX_t, uintptr_t,
+      IntervalMapImpl::NodeSizer<uintX_t, uintptr_t>::LeafSize,
+      IntervalMapHalfOpenInfo<uintX_t>> LoadMapT;
+  typename LoadMapT::Allocator Alloc;
+  // Allocate the IntervalMap on the heap to work around MSVC bug where the
+  // stack doesn't get realigned despite LoadMap having alignment 8 (PR24113).
+  std::unique_ptr<LoadMapT> LoadMap(new LoadMapT(Alloc));
+
   for (Elf_Phdr_Iter PhdrI = program_header_begin(),
                      PhdrE = program_header_end();
        PhdrI != PhdrE; ++PhdrI) {
@@ -709,34 +742,44 @@ ELFFile<ELFT>::ELFFile(StringRef Object, std::error_code &EC)
       DynamicRegion.Addr = base() + PhdrI->p_offset;
       DynamicRegion.Size = PhdrI->p_filesz;
       DynamicRegion.EntSize = sizeof(Elf_Dyn);
-      break;
+      continue;
     }
+    if (PhdrI->p_type != ELF::PT_LOAD)
+      continue;
+    if (PhdrI->p_filesz == 0)
+      continue;
+    LoadMap->insert(PhdrI->p_vaddr, PhdrI->p_vaddr + PhdrI->p_filesz,
+                    PhdrI->p_offset);
   }
 
-  // Scan dynamic table.
+  auto toMappedAddr = [&](uint64_t VAddr) -> const uint8_t * {
+    auto I = LoadMap->find(VAddr);
+    if (I == LoadMap->end())
+      return nullptr;
+    return this->base() + I.value() + (VAddr - I.start());
+  };
+
   for (Elf_Dyn_Iter DynI = dynamic_table_begin(), DynE = dynamic_table_end();
        DynI != DynE; ++DynI) {
     switch (DynI->d_tag) {
-    case ELF::DT_RELA: {
-      uint64_t VBase = 0;
-      const uint8_t *FBase = nullptr;
-      for (Elf_Phdr_Iter PhdrI = program_header_begin(),
-                         PhdrE = program_header_end();
-           PhdrI != PhdrE; ++PhdrI) {
-        if (PhdrI->p_type != ELF::PT_LOAD)
-          continue;
-        if (DynI->getPtr() >= PhdrI->p_vaddr &&
-            DynI->getPtr() < PhdrI->p_vaddr + PhdrI->p_memsz) {
-          VBase = PhdrI->p_vaddr;
-          FBase = base() + PhdrI->p_offset;
-          break;
-        }
-      }
-      if (!VBase)
-        return;
-      DynRelaRegion.Addr = FBase + DynI->getPtr() - VBase;
+    case ELF::DT_HASH:
+      if (HashTable)
+        continue;
+      HashTable =
+          reinterpret_cast<const Elf_Hash *>(toMappedAddr(DynI->getPtr()));
+      break;
+    case ELF::DT_STRTAB:
+      if (!DynStrRegion.Addr)
+        DynStrRegion.Addr = toMappedAddr(DynI->getPtr());
+      break;
+    case ELF::DT_STRSZ:
+      if (!DynStrRegion.Size)
+        DynStrRegion.Size = DynI->getVal();
+      break;
+    case ELF::DT_RELA:
+      if (!DynRelaRegion.Addr)
+        DynRelaRegion.Addr = toMappedAddr(DynI->getPtr());
       break;
-    }
     case ELF::DT_RELASZ:
       DynRelaRegion.Size = DynI->getVal();
       break;
@@ -744,8 +787,6 @@ ELFFile<ELFT>::ELFFile(StringRef Object, std::error_code &EC)
       DynRelaRegion.EntSize = DynI->getVal();
     }
   }
-
-  EC = std::error_code();
 }
 
 template <class ELFT>
@@ -868,9 +909,9 @@ ELFFile<ELFT>::getStringTable(const Elf_Shdr *Section) const {
 
 template <class ELFT>
 const char *ELFFile<ELFT>::getDynamicString(uintX_t Offset) const {
-  if (!DotDynSymSec || Offset >= DynSymStrTab.size())
+  if (Offset >= DynStrRegion.Size)
     return nullptr;
-  return (const char *)DynSymStrTab.begin() + Offset;
+  return (const char *)DynStrRegion.Addr + Offset;
 }
 
 template <class ELFT>
@@ -983,7 +1024,7 @@ ErrorOr<StringRef> ELFFile<ELFT>::getSymbolVersion(const Elf_Shdr *section,
     IsDefault = false;
   }
 
-  if (name_offset >= DynSymStrTab.size())
+  if (name_offset >= DynStrRegion.Size)
     return object_error::parse_failed;
   return StringRef(getDynamicString(name_offset));
 }
diff --git a/include/llvm/Object/ELFObjectFile.h b/include/llvm/Object/ELFObjectFile.h
index 5b9b113a2f0b..6e8ace427a20 100644
--- a/include/llvm/Object/ELFObjectFile.h
+++ b/include/llvm/Object/ELFObjectFile.h
@@ -196,9 +196,8 @@ protected:
 
   void moveSymbolNext(DataRefImpl &Symb) const override;
   ErrorOr<StringRef> getSymbolName(DataRefImpl Symb) const override;
-  std::error_code getSymbolAddress(DataRefImpl Symb,
-                                   uint64_t &Res) const override;
-  uint64_t getSymbolValue(DataRefImpl Symb) const override;
+  ErrorOr<uint64_t> getSymbolAddress(DataRefImpl Symb) const override;
+  uint64_t getSymbolValueImpl(DataRefImpl Symb) const override;
   uint32_t getSymbolAlignment(DataRefImpl Symb) const override;
   uint64_t getCommonSymbolSizeImpl(DataRefImpl Symb) const override;
   uint32_t getSymbolFlags(DataRefImpl Symb) const override;
@@ -226,7 +225,6 @@ protected:
   section_iterator getRelocatedSection(DataRefImpl Sec) const override;
 
   void moveRelocationNext(DataRefImpl &Rel) const override;
-  ErrorOr<uint64_t> getRelocationAddress(DataRefImpl Rel) const override;
   uint64_t getRelocationOffset(DataRefImpl Rel) const override;
   symbol_iterator getRelocationSymbol(DataRefImpl Rel) const override;
   uint64_t getRelocationType(DataRefImpl Rel) const override;
@@ -235,7 +233,6 @@ protected:
 
   uint32_t getSectionType(DataRefImpl Sec) const override;
   uint64_t getSectionFlags(DataRefImpl Sec) const override;
-  uint64_t getROffset(DataRefImpl Rel) const;
   StringRef getRelocationTypeName(uint32_t Type) const;
 
   /// \brief Get the relocation section that contains \a Rel.
@@ -276,11 +273,6 @@ protected:
     return DRI;
   }
 
-  Elf_Dyn_Iter toELFDynIter(DataRefImpl Dyn) const {
-    return Elf_Dyn_Iter(EF.begin_dynamic_table().getEntSize(),
-                        reinterpret_cast<const char *>(Dyn.p));
-  }
-
   DataRefImpl toDRI(Elf_Dyn_Iter Dyn) const {
     DataRefImpl DRI;
     DRI.p = reinterpret_cast<uintptr_t>(Dyn.get());
@@ -378,19 +370,13 @@ uint32_t ELFObjectFile<ELFT>::getSectionType(DataRefImpl Sec) const {
 }
 
 template <class ELFT>
-uint64_t ELFObjectFile<ELFT>::getSymbolValue(DataRefImpl Symb) const {
+uint64_t ELFObjectFile<ELFT>::getSymbolValueImpl(DataRefImpl Symb) const {
   const Elf_Sym *ESym = getSymbol(Symb);
-  switch (ESym->st_shndx) {
-  case ELF::SHN_COMMON:
-  case ELF::SHN_UNDEF:
-    return UnknownAddress;
-  case ELF::SHN_ABS:
-    return ESym->st_value;
-  }
-
-  const Elf_Ehdr *Header = EF.getHeader();
   uint64_t Ret = ESym->st_value;
+  if (ESym->st_shndx == ELF::SHN_ABS)
+    return Ret;
 
+  const Elf_Ehdr *Header = EF.getHeader();
   // Clear the ARM/Thumb or microMIPS indicator flag.
   if ((Header->e_machine == ELF::EM_ARM || Header->e_machine == ELF::EM_MIPS) &&
       ESym->getType() == ELF::STT_FUNC)
@@ -400,15 +386,15 @@ uint64_t ELFObjectFile<ELFT>::getSymbolValue(DataRefImpl Symb) const {
 }
 
 template <class ELFT>
-std::error_code ELFObjectFile<ELFT>::getSymbolAddress(DataRefImpl Symb,
-                                                      uint64_t &Result) const {
-  Result = getSymbolValue(Symb);
+ErrorOr<uint64_t>
+ELFObjectFile<ELFT>::getSymbolAddress(DataRefImpl Symb) const {
+  uint64_t Result = getSymbolValue(Symb);
   const Elf_Sym *ESym = getSymbol(Symb);
   switch (ESym->st_shndx) {
   case ELF::SHN_COMMON:
   case ELF::SHN_UNDEF:
   case ELF::SHN_ABS:
-    return std::error_code();
+    return Result;
   }
 
   const Elf_Ehdr *Header = EF.getHeader();
@@ -422,7 +408,7 @@ std::error_code ELFObjectFile<ELFT>::getSymbolAddress(DataRefImpl Symb,
       Result += Section->sh_addr;
   }
 
-  return std::error_code();
+  return Result;
 }
 
 template <class ELFT>
@@ -689,31 +675,9 @@ ELFObjectFile<ELFT>::getRelocationSymbol(DataRefImpl Rel) const {
 }
 
 template <class ELFT>
-ErrorOr<uint64_t>
-ELFObjectFile<ELFT>::getRelocationAddress(DataRefImpl Rel) const {
-  uint64_t ROffset = getROffset(Rel);
-  const Elf_Ehdr *Header = EF.getHeader();
-
-  if (Header->e_type == ELF::ET_REL) {
-    const Elf_Shdr *RelocationSec = getRelSection(Rel);
-    ErrorOr<const Elf_Shdr *> RelocatedSec =
-        EF.getSection(RelocationSec->sh_info);
-    if (std::error_code EC = RelocatedSec.getError())
-      return EC;
-    return ROffset + (*RelocatedSec)->sh_addr;
-  }
-  return ROffset;
-}
-
-template <class ELFT>
 uint64_t ELFObjectFile<ELFT>::getRelocationOffset(DataRefImpl Rel) const {
   assert(EF.getHeader()->e_type == ELF::ET_REL &&
          "Only relocatable object files have relocation offsets");
-  return getROffset(Rel);
-}
-
-template <class ELFT>
-uint64_t ELFObjectFile<ELFT>::getROffset(DataRefImpl Rel) const {
   const Elf_Shdr *sec = getRelSection(Rel);
   if (sec->sh_type == ELF::SHT_REL)
     return getRel(Rel)->r_offset;
diff --git a/include/llvm/Object/ELFTypes.h b/include/llvm/Object/ELFTypes.h
index 63e13909ae5c..27e987ba2852 100644
--- a/include/llvm/Object/ELFTypes.h
+++ b/include/llvm/Object/ELFTypes.h
@@ -10,6 +10,7 @@
 #ifndef LLVM_OBJECT_ELFTYPES_H
 #define LLVM_OBJECT_ELFTYPES_H
 
+#include "llvm/ADT/ArrayRef.h"
 #include "llvm/Object/Error.h"
 #include "llvm/Support/DataTypes.h"
 #include "llvm/Support/ELF.h"
@@ -463,6 +464,23 @@ struct Elf_Phdr_Impl<ELFType<TargetEndianness, true>> {
   Elf_Xword p_align;  // Segment alignment constraint
 };
 
+// ELFT needed for endianess.
+template <class ELFT>
+struct Elf_Hash_Impl {
+  LLVM_ELF_IMPORT_TYPES_ELFT(ELFT)
+  Elf_Word nbucket;
+  Elf_Word nchain;
+
+  ArrayRef<Elf_Word> buckets() const {
+    return ArrayRef<Elf_Word>(&nbucket + 2, &nbucket + 2 + nbucket);
+  }
+
+  ArrayRef<Elf_Word> chains() const {
+    return ArrayRef<Elf_Word>(&nbucket + 2 + nbucket,
+                              &nbucket + 2 + nbucket + nchain);
+  }
+};
+
 // MIPS .reginfo section
 template <class ELFT>
 struct Elf_Mips_RegInfo;
diff --git a/include/llvm/Object/ELFYAML.h b/include/llvm/Object/ELFYAML.h
index b45507920a9c..df0aa500c8a2 100644
--- a/include/llvm/Object/ELFYAML.h
+++ b/include/llvm/Object/ELFYAML.h
@@ -85,7 +85,13 @@ struct SectionOrType {
 };
 
 struct Section {
-  enum class SectionKind { Group, RawContent, Relocation, MipsABIFlags };
+  enum class SectionKind {
+    Group,
+    RawContent,
+    Relocation,
+    NoBits,
+    MipsABIFlags
+  };
   SectionKind Kind;
   StringRef Name;
   ELF_SHT Type;
@@ -106,6 +112,14 @@ struct RawContentSection : Section {
   }
 };
 
+struct NoBitsSection : Section {
+  llvm::yaml::Hex64 Size;
+  NoBitsSection() : Section(SectionKind::NoBits) {}
+  static bool classof(const Section *S) {
+    return S->Kind == SectionKind::NoBits;
+  }
+};
+
 struct Group : Section {
   // Members of a group contain a flag and a list of section indices
   // that are part of the group.
diff --git a/include/llvm/Object/MachO.h b/include/llvm/Object/MachO.h
index f4edfd057303..489ecef5c996 100644
--- a/include/llvm/Object/MachO.h
+++ b/include/llvm/Object/MachO.h
@@ -205,9 +205,7 @@ public:
   std::error_code getIndirectName(DataRefImpl Symb, StringRef &Res) const;
   unsigned getSectionType(SectionRef Sec) const;
 
-  std::error_code getSymbolAddress(DataRefImpl Symb,
-                                   uint64_t &Res) const override;
-  uint64_t getSymbolValue(DataRefImpl Symb) const override;
+  ErrorOr<uint64_t> getSymbolAddress(DataRefImpl Symb) const override;
   uint32_t getSymbolAlignment(DataRefImpl Symb) const override;
   uint64_t getCommonSymbolSizeImpl(DataRefImpl Symb) const override;
   SymbolRef::Type getSymbolType(DataRefImpl Symb) const override;
@@ -233,7 +231,6 @@ public:
   relocation_iterator section_rel_end(DataRefImpl Sec) const override;
 
   void moveRelocationNext(DataRefImpl &Rel) const override;
-  ErrorOr<uint64_t> getRelocationAddress(DataRefImpl Rel) const override;
   uint64_t getRelocationOffset(DataRefImpl Rel) const override;
   symbol_iterator getRelocationSymbol(DataRefImpl Rel) const override;
   section_iterator getRelocationSection(DataRefImpl Rel) const;
@@ -245,6 +242,8 @@ public:
   // MachO specific.
   std::error_code getLibraryShortNameByIndex(unsigned Index, StringRef &) const;
 
+  section_iterator getRelocationRelocatedSection(relocation_iterator Rel) const;
+
   // TODO: Would be useful to have an iterator based version
   // of the load command interface too.
 
@@ -425,6 +424,8 @@ public:
   }
 
 private:
+  uint64_t getSymbolValueImpl(DataRefImpl Symb) const override;
+
   union {
     MachO::mach_header_64 Header64;
     MachO::mach_header Header;
diff --git a/include/llvm/Object/ObjectFile.h b/include/llvm/Object/ObjectFile.h
index 62eab1066be5..8dd525626218 100644
--- a/include/llvm/Object/ObjectFile.h
+++ b/include/llvm/Object/ObjectFile.h
@@ -50,7 +50,6 @@ public:
 
   void moveNext();
 
-  ErrorOr<uint64_t> getAddress() const;
   uint64_t getOffset() const;
   symbol_iterator getSymbol() const;
   uint64_t getType() const;
@@ -135,7 +134,7 @@ public:
   ErrorOr<StringRef> getName() const;
   /// Returns the symbol virtual address (i.e. address at which it will be
   /// mapped).
-  std::error_code getAddress(uint64_t &Result) const;
+  ErrorOr<uint64_t> getAddress() const;
 
   /// Return the value of the symbol depending on the object this can be an
   /// offset or a virtual address.
@@ -198,9 +197,8 @@ protected:
   virtual ErrorOr<StringRef> getSymbolName(DataRefImpl Symb) const = 0;
   std::error_code printSymbolName(raw_ostream &OS,
                                   DataRefImpl Symb) const override;
-  virtual std::error_code getSymbolAddress(DataRefImpl Symb,
-                                           uint64_t &Res) const = 0;
-  virtual uint64_t getSymbolValue(DataRefImpl Symb) const = 0;
+  virtual ErrorOr<uint64_t> getSymbolAddress(DataRefImpl Symb) const = 0;
+  virtual uint64_t getSymbolValueImpl(DataRefImpl Symb) const = 0;
   virtual uint32_t getSymbolAlignment(DataRefImpl Symb) const;
   virtual uint64_t getCommonSymbolSizeImpl(DataRefImpl Symb) const = 0;
   virtual SymbolRef::Type getSymbolType(DataRefImpl Symb) const = 0;
@@ -229,13 +227,14 @@ protected:
   // Same as above for RelocationRef.
   friend class RelocationRef;
   virtual void moveRelocationNext(DataRefImpl &Rel) const = 0;
-  virtual ErrorOr<uint64_t> getRelocationAddress(DataRefImpl Rel) const = 0;
   virtual uint64_t getRelocationOffset(DataRefImpl Rel) const = 0;
   virtual symbol_iterator getRelocationSymbol(DataRefImpl Rel) const = 0;
   virtual uint64_t getRelocationType(DataRefImpl Rel) const = 0;
   virtual void getRelocationTypeName(DataRefImpl Rel,
                                      SmallVectorImpl<char> &Result) const = 0;
 
+  uint64_t getSymbolValue(DataRefImpl Symb) const;
+
 public:
   uint64_t getCommonSymbolSize(DataRefImpl Symb) const {
     assert(getSymbolFlags(Symb) & SymbolRef::SF_Common);
@@ -308,8 +307,8 @@ inline ErrorOr<StringRef> SymbolRef::getName() const {
   return getObject()->getSymbolName(getRawDataRefImpl());
 }
 
-inline std::error_code SymbolRef::getAddress(uint64_t &Result) const {
-  return getObject()->getSymbolAddress(getRawDataRefImpl(), Result);
+inline ErrorOr<uint64_t> SymbolRef::getAddress() const {
+  return getObject()->getSymbolAddress(getRawDataRefImpl());
 }
 
 inline uint64_t SymbolRef::getValue() const {
@@ -430,10 +429,6 @@ inline void RelocationRef::moveNext() {
   return OwningObject->moveRelocationNext(RelocationPimpl);
 }
 
-inline ErrorOr<uint64_t> RelocationRef::getAddress() const {
-  return OwningObject->getRelocationAddress(RelocationPimpl);
-}
-
 inline uint64_t RelocationRef::getOffset() const {
   return OwningObject->getRelocationOffset(RelocationPimpl);
 }
diff --git a/include/llvm/Object/RelocVisitor.h b/include/llvm/Object/RelocVisitor.h
index 950e2ed0e338..d5e4258cb0a7 100644
--- a/include/llvm/Object/RelocVisitor.h
+++ b/include/llvm/Object/RelocVisitor.h
@@ -100,9 +100,9 @@ private:
       case Triple::mips64:
         switch (RelocType) {
         case llvm::ELF::R_MIPS_32:
-          return visitELF_MIPS_32(R, Value);
+          return visitELF_MIPS64_32(R, Value);
         case llvm::ELF::R_MIPS_64:
-          return visitELF_MIPS_64(R, Value);
+          return visitELF_MIPS64_64(R, Value);
         default:
           HasError = true;
           return RelocToApply();
@@ -313,11 +313,18 @@ private:
 
   /// MIPS ELF
   RelocToApply visitELF_MIPS_32(RelocationRef R, uint64_t Value) {
-    uint32_t Res = (Value)&0xFFFFFFFF;
+    uint32_t Res = Value & 0xFFFFFFFF;
+    return RelocToApply(Res, 4);
+  }
+
+  /// MIPS64 ELF
+  RelocToApply visitELF_MIPS64_32(RelocationRef R, uint64_t Value) {
+    int64_t Addend = getELFAddend(R);
+    uint32_t Res = (Value + Addend) & 0xFFFFFFFF;
     return RelocToApply(Res, 4);
   }
 
-  RelocToApply visitELF_MIPS_64(RelocationRef R, uint64_t Value) {
+  RelocToApply visitELF_MIPS64_64(RelocationRef R, uint64_t Value) {
     int64_t Addend = getELFAddend(R);
     uint64_t Res = (Value + Addend);
     return RelocToApply(Res, 8);
diff --git a/include/llvm/Object/SymbolicFile.h b/include/llvm/Object/SymbolicFile.h
index 3a3823159c92..537997ac6318 100644
--- a/include/llvm/Object/SymbolicFile.h
+++ b/include/llvm/Object/SymbolicFile.h
@@ -115,8 +115,6 @@ public:
 
 typedef content_iterator<BasicSymbolRef> basic_symbol_iterator;
 
-const uint64_t UnknownAddress = ~0ULL;
-
 class SymbolicFile : public Binary {
 public:
   ~SymbolicFile() override;
diff --git a/include/llvm/Support/COFF.h b/include/llvm/Support/COFF.h
index b26af61a7c70..3c5ee06969d0 100644
--- a/include/llvm/Support/COFF.h
+++ b/include/llvm/Support/COFF.h
@@ -655,6 +655,7 @@ namespace COFF {
   };
 
   enum CodeViewIdentifiers {
+    DEBUG_LINE_TABLES_HAVE_COLUMN_RECORDS = 0x1,
     DEBUG_SECTION_MAGIC = 0x4,
     DEBUG_SYMBOL_SUBSECTION = 0xF1,
     DEBUG_LINE_TABLE_SUBSECTION = 0xF2,
diff --git a/include/llvm/Support/CommandLine.h b/include/llvm/Support/CommandLine.h
index ed809211ea97..379d06a65741 100644
--- a/include/llvm/Support/CommandLine.h
+++ b/include/llvm/Support/CommandLine.h
@@ -790,7 +790,7 @@ public:
   void anchor() override;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class basic_parser<bool>);
+extern template class basic_parser<bool>;
 
 //--------------------------------------------------
 // parser<boolOrDefault>
@@ -816,7 +816,7 @@ public:
   void anchor() override;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class basic_parser<boolOrDefault>);
+extern template class basic_parser<boolOrDefault>;
 
 //--------------------------------------------------
 // parser<int>
@@ -838,7 +838,7 @@ public:
   void anchor() override;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class basic_parser<int>);
+extern template class basic_parser<int>;
 
 //--------------------------------------------------
 // parser<unsigned>
@@ -860,7 +860,7 @@ public:
   void anchor() override;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class basic_parser<unsigned>);
+extern template class basic_parser<unsigned>;
 
 //--------------------------------------------------
 // parser<unsigned long long>
@@ -885,7 +885,7 @@ public:
   void anchor() override;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class basic_parser<unsigned long long>);
+extern template class basic_parser<unsigned long long>;
 
 //--------------------------------------------------
 // parser<double>
@@ -907,7 +907,7 @@ public:
   void anchor() override;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class basic_parser<double>);
+extern template class basic_parser<double>;
 
 //--------------------------------------------------
 // parser<float>
@@ -929,7 +929,7 @@ public:
   void anchor() override;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class basic_parser<float>);
+extern template class basic_parser<float>;
 
 //--------------------------------------------------
 // parser<std::string>
@@ -954,7 +954,7 @@ public:
   void anchor() override;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class basic_parser<std::string>);
+extern template class basic_parser<std::string>;
 
 //--------------------------------------------------
 // parser<char>
@@ -979,7 +979,7 @@ public:
   void anchor() override;
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class basic_parser<char>);
+extern template class basic_parser<char>;
 
 //--------------------------------------------------
 // PrintOptionDiff
@@ -1254,11 +1254,11 @@ public:
   }
 };
 
-EXTERN_TEMPLATE_INSTANTIATION(class opt<unsigned>);
-EXTERN_TEMPLATE_INSTANTIATION(class opt<int>);
-EXTERN_TEMPLATE_INSTANTIATION(class opt<std::string>);
-EXTERN_TEMPLATE_INSTANTIATION(class opt<char>);
-EXTERN_TEMPLATE_INSTANTIATION(class opt<bool>);
+extern template class opt<unsigned>;
+extern template class opt<int>;
+extern template class opt<std::string>;
+extern template class opt<char>;
+extern template class opt<bool>;
 
 //===----------------------------------------------------------------------===//
 // list_storage class
diff --git a/include/llvm/Support/Compiler.h b/include/llvm/Support/Compiler.h
index 67ef23d43c99..141639839cc2 100644
--- a/include/llvm/Support/Compiler.h
+++ b/include/llvm/Support/Compiler.h
@@ -174,19 +174,6 @@
 #define LLVM_UNLIKELY(EXPR) (EXPR)
 #endif
 
-// C++ doesn't support 'extern template' of template specializations.  GCC does,
-// but requires __extension__ before it.  In the header, use this:
-//   EXTERN_TEMPLATE_INSTANTIATION(class foo<bar>);
-// in the .cpp file, use this:
-//   TEMPLATE_INSTANTIATION(class foo<bar>);
-#ifdef __GNUC__
-#define EXTERN_TEMPLATE_INSTANTIATION(X) __extension__ extern template X
-#define TEMPLATE_INSTANTIATION(X) template X
-#else
-#define EXTERN_TEMPLATE_INSTANTIATION(X)
-#define TEMPLATE_INSTANTIATION(X)
-#endif
-
 /// LLVM_ATTRIBUTE_NOINLINE - On compilers where we have a directive to do so,
 /// mark a method "not for inlining".
 #if __has_attribute(noinline) || LLVM_GNUC_PREREQ(3, 4, 0)
diff --git a/include/llvm/Support/OnDiskHashTable.h b/include/llvm/Support/OnDiskHashTable.h
index 0f097f287286..08e277ad5ce1 100644
--- a/include/llvm/Support/OnDiskHashTable.h
+++ b/include/llvm/Support/OnDiskHashTable.h
@@ -280,13 +280,19 @@ public:
   };
 
   /// \brief Look up the stored data for a particular key.
-  iterator find(const external_key_type &EKey, Info *InfoPtr = 0) {
-    if (!InfoPtr)
-      InfoPtr = &InfoObj;
-
-    using namespace llvm::support;
+  iterator find(const external_key_type &EKey, Info *InfoPtr = nullptr) {
     const internal_key_type &IKey = InfoObj.GetInternalKey(EKey);
     hash_value_type KeyHash = InfoObj.ComputeHash(IKey);
+    return find_hashed(IKey, KeyHash, InfoPtr);
+  }
+
+  /// \brief Look up the stored data for a particular key with a known hash.
+  iterator find_hashed(const internal_key_type &IKey, hash_value_type KeyHash,
+                       Info *InfoPtr = nullptr) {
+    using namespace llvm::support;
+
+    if (!InfoPtr)
+      InfoPtr = &InfoObj;
 
     // Each bucket is just an offset into the hash table file.
     offset_type Idx = KeyHash & (NumBuckets - 1);
diff --git a/include/llvm/Support/TargetRegistry.h b/include/llvm/Support/TargetRegistry.h
index d2e8b95d74f3..40bf6fb20c9f 100644
--- a/include/llvm/Support/TargetRegistry.h
+++ b/include/llvm/Support/TargetRegistry.h
@@ -71,7 +71,7 @@ MCStreamer *createMachOStreamer(MCContext &Ctx, MCAsmBackend &TAB,
 
 MCRelocationInfo *createMCRelocationInfo(const Triple &TT, MCContext &Ctx);
 
-MCSymbolizer *createMCSymbolizer(StringRef TT, LLVMOpInfoCallback GetOpInfo,
+MCSymbolizer *createMCSymbolizer(const Triple &TT, LLVMOpInfoCallback GetOpInfo,
                                  LLVMSymbolLookupCallback SymbolLookUp,
                                  void *DisInfo, MCContext *Ctx,
                                  std::unique_ptr<MCRelocationInfo> &&RelInfo);
@@ -92,17 +92,18 @@ public:
 
   typedef MCAsmInfo *(*MCAsmInfoCtorFnTy)(const MCRegisterInfo &MRI,
                                           const Triple &TT);
-  typedef MCCodeGenInfo *(*MCCodeGenInfoCtorFnTy)(StringRef TT, Reloc::Model RM,
+  typedef MCCodeGenInfo *(*MCCodeGenInfoCtorFnTy)(const Triple &TT,
+                                                  Reloc::Model RM,
                                                   CodeModel::Model CM,
                                                   CodeGenOpt::Level OL);
   typedef MCInstrInfo *(*MCInstrInfoCtorFnTy)(void);
   typedef MCInstrAnalysis *(*MCInstrAnalysisCtorFnTy)(const MCInstrInfo *Info);
-  typedef MCRegisterInfo *(*MCRegInfoCtorFnTy)(StringRef TT);
+  typedef MCRegisterInfo *(*MCRegInfoCtorFnTy)(const Triple &TT);
   typedef MCSubtargetInfo *(*MCSubtargetInfoCtorFnTy)(const Triple &TT,
                                                       StringRef CPU,
                                                       StringRef Features);
   typedef TargetMachine *(*TargetMachineCtorTy)(
-      const Target &T, StringRef TT, StringRef CPU, StringRef Features,
+      const Target &T, const Triple &TT, StringRef CPU, StringRef Features,
       const TargetOptions &Options, Reloc::Model RM, CodeModel::Model CM,
       CodeGenOpt::Level OL);
   // If it weren't for layering issues (this header is in llvm/Support, but
@@ -150,7 +151,7 @@ public:
   typedef MCRelocationInfo *(*MCRelocationInfoCtorTy)(const Triple &TT,
                                                       MCContext &Ctx);
   typedef MCSymbolizer *(*MCSymbolizerCtorTy)(
-      StringRef TT, LLVMOpInfoCallback GetOpInfo,
+      const Triple &TT, LLVMOpInfoCallback GetOpInfo,
       LLVMSymbolLookupCallback SymbolLookUp, void *DisInfo, MCContext *Ctx,
       std::unique_ptr<MCRelocationInfo> &&RelInfo);
 
@@ -300,12 +301,12 @@ public:
 
   /// createMCCodeGenInfo - Create a MCCodeGenInfo implementation.
   ///
-  MCCodeGenInfo *createMCCodeGenInfo(StringRef Triple, Reloc::Model RM,
+  MCCodeGenInfo *createMCCodeGenInfo(StringRef TT, Reloc::Model RM,
                                      CodeModel::Model CM,
                                      CodeGenOpt::Level OL) const {
     if (!MCCodeGenInfoCtorFn)
       return nullptr;
-    return MCCodeGenInfoCtorFn(Triple, RM, CM, OL);
+    return MCCodeGenInfoCtorFn(Triple(TT), RM, CM, OL);
   }
 
   /// createMCInstrInfo - Create a MCInstrInfo implementation.
@@ -326,10 +327,10 @@ public:
 
   /// createMCRegInfo - Create a MCRegisterInfo implementation.
   ///
-  MCRegisterInfo *createMCRegInfo(StringRef Triple) const {
+  MCRegisterInfo *createMCRegInfo(StringRef TT) const {
     if (!MCRegInfoCtorFn)
       return nullptr;
-    return MCRegInfoCtorFn(Triple);
+    return MCRegInfoCtorFn(Triple(TT));
   }
 
   /// createMCSubtargetInfo - Create a MCSubtargetInfo implementation.
@@ -351,20 +352,20 @@ public:
   /// createTargetMachine - Create a target specific machine implementation
   /// for the specified \p Triple.
   ///
-  /// \param Triple This argument is used to determine the target machine
+  /// \param TT This argument is used to determine the target machine
   /// feature set; it should always be provided. Generally this should be
   /// either the target triple from the module, or the target triple of the
   /// host if that does not exist.
   TargetMachine *
-  createTargetMachine(StringRef Triple, StringRef CPU, StringRef Features,
+  createTargetMachine(StringRef TT, StringRef CPU, StringRef Features,
                       const TargetOptions &Options,
                       Reloc::Model RM = Reloc::Default,
                       CodeModel::Model CM = CodeModel::Default,
                       CodeGenOpt::Level OL = CodeGenOpt::Default) const {
     if (!TargetMachineCtorFn)
       return nullptr;
-    return TargetMachineCtorFn(*this, Triple, CPU, Features, Options, RM, CM,
-                               OL);
+    return TargetMachineCtorFn(*this, Triple(TT), CPU, Features, Options, RM,
+                               CM, OL);
   }
 
   /// createMCAsmBackend - Create a target specific assembly parser.
@@ -529,7 +530,8 @@ public:
                      std::unique_ptr<MCRelocationInfo> &&RelInfo) const {
     MCSymbolizerCtorTy Fn =
         MCSymbolizerCtorFn ? MCSymbolizerCtorFn : llvm::createMCSymbolizer;
-    return Fn(TT, GetOpInfo, SymbolLookUp, DisInfo, Ctx, std::move(RelInfo));
+    return Fn(Triple(TT), GetOpInfo, SymbolLookUp, DisInfo, Ctx,
+              std::move(RelInfo));
   }
 
   /// @}
@@ -924,7 +926,7 @@ template <class MCCodeGenInfoImpl> struct RegisterMCCodeGenInfo {
   }
 
 private:
-  static MCCodeGenInfo *Allocator(StringRef /*TT*/, Reloc::Model /*RM*/,
+  static MCCodeGenInfo *Allocator(const Triple & /*TT*/, Reloc::Model /*RM*/,
                                   CodeModel::Model /*CM*/,
                                   CodeGenOpt::Level /*OL*/) {
     return new MCCodeGenInfoImpl();
@@ -1023,7 +1025,7 @@ template <class MCRegisterInfoImpl> struct RegisterMCRegInfo {
   }
 
 private:
-  static MCRegisterInfo *Allocator(StringRef /*TT*/) {
+  static MCRegisterInfo *Allocator(const Triple & /*TT*/) {
     return new MCRegisterInfoImpl();
   }
 };
@@ -1090,11 +1092,11 @@ template <class TargetMachineImpl> struct RegisterTargetMachine {
   }
 
 private:
-  static TargetMachine *Allocator(const Target &T, StringRef TT, StringRef CPU,
-                                  StringRef FS, const TargetOptions &Options,
-                                  Reloc::Model RM, CodeModel::Model CM,
-                                  CodeGenOpt::Level OL) {
-    return new TargetMachineImpl(T, Triple(TT), CPU, FS, Options, RM, CM, OL);
+  static TargetMachine *Allocator(const Target &T, const Triple &TT,
+                                  StringRef CPU, StringRef FS,
+                                  const TargetOptions &Options, Reloc::Model RM,
+                                  CodeModel::Model CM, CodeGenOpt::Level OL) {
+    return new TargetMachineImpl(T, TT, CPU, FS, Options, RM, CM, OL);
   }
 };
 
diff --git a/include/llvm/Support/raw_ostream.h b/include/llvm/Support/raw_ostream.h
index b59317112c44..28e512c86941 100644
--- a/include/llvm/Support/raw_ostream.h
+++ b/include/llvm/Support/raw_ostream.h
@@ -165,8 +165,10 @@ public:
     if (Size > (size_t)(OutBufEnd - OutBufCur))
       return write(Str.data(), Size);
 
-    memcpy(OutBufCur, Str.data(), Size);
-    OutBufCur += Size;
+    if (Size) {
+      memcpy(OutBufCur, Str.data(), Size);
+      OutBufCur += Size;
+    }
     return *this;
   }
 
diff --git a/include/llvm/TableGen/Record.h b/include/llvm/TableGen/Record.h
index 717a2a4ba62a..b4642c991192 100644
--- a/include/llvm/TableGen/Record.h
+++ b/include/llvm/TableGen/Record.h
@@ -1222,11 +1222,11 @@ public:
   /// get the corresponding DefInit.
   DefInit *getDefInit();
 
-  const std::vector<Init *> &getTemplateArgs() const {
+  ArrayRef<Init *> getTemplateArgs() const {
     return TemplateArgs;
   }
-  const std::vector<RecordVal> &getValues() const { return Values; }
-  const std::vector<Record*>   &getSuperClasses() const { return SuperClasses; }
+  ArrayRef<RecordVal> getValues() const { return Values; }
+  ArrayRef<Record *>  getSuperClasses() const { return SuperClasses; }
   ArrayRef<SMRange> getSuperClassRanges() const { return SuperClassRanges; }
 
   bool isTemplateArg(Init *Name) const {
diff --git a/include/llvm/Target/Target.td b/include/llvm/Target/Target.td
index 61234991be44..e0aea181a639 100644
--- a/include/llvm/Target/Target.td
+++ b/include/llvm/Target/Target.td
@@ -872,7 +872,7 @@ def LOAD_STACK_GUARD : Instruction {
   let hasSideEffects = 0;
   bit isPseudo = 1;
 }
-def FRAME_ALLOC : Instruction {
+def LOCAL_ESCAPE : Instruction {
   // This instruction is really just a label. It has to be part of the chain so
   // that it doesn't get dropped from the DAG, but it produces nothing and has
   // no side effects.
@@ -1014,7 +1014,7 @@ class InstAlias<string Asm, dag Result, int Emit = 1> {
   // Predicates - Predicates that must be true for this to match.
   list<Predicate> Predicates = [];
 
-	// If the instruction specified in Result has defined an AsmMatchConverter
+  // If the instruction specified in Result has defined an AsmMatchConverter
   // then setting this to 1 will cause the alias to use the AsmMatchConverter
   // function when converting the OperandVector into an MCInst instead of the
   // function that is generated by the dag Result.
diff --git a/include/llvm/Target/TargetFrameLowering.h b/include/llvm/Target/TargetFrameLowering.h
index 0e317247a59f..3af2227410f7 100644
--- a/include/llvm/Target/TargetFrameLowering.h
+++ b/include/llvm/Target/TargetFrameLowering.h
@@ -19,6 +19,7 @@
 #include <vector>
 
 namespace llvm {
+  class BitVector;
   class CalleeSavedInfo;
   class MachineFunction;
   class RegScavenger;
@@ -226,13 +227,15 @@ public:
     return 0;
   }
 
-  /// processFunctionBeforeCalleeSavedScan - This method is called immediately
-  /// before PrologEpilogInserter scans the physical registers used to determine
-  /// what callee saved registers should be spilled. This method is optional.
-  virtual void processFunctionBeforeCalleeSavedScan(MachineFunction &MF,
-                                             RegScavenger *RS = nullptr) const {
-
-  }
+  /// This method determines which of the registers reported by
+  /// TargetRegisterInfo::getCalleeSavedRegs() should actually get saved.
+  /// The default implementation checks populates the \p SavedRegs bitset with
+  /// all registers which are modified in the function, targets may override
+  /// this function to save additional registers.
+  /// This method also sets up the register scavenger ensuring there is a free
+  /// register or a frameindex available.
+  virtual void determineCalleeSaves(MachineFunction &MF, BitVector &SavedRegs,
+                                    RegScavenger *RS = nullptr) const;
 
   /// processFunctionBeforeFrameFinalized - This method is called immediately
   /// before the specified function's frame layout (MF.getFrameInfo()) is
diff --git a/include/llvm/Target/TargetLowering.h b/include/llvm/Target/TargetLowering.h
index 277487fee6bc..4412d9b3c68e 100644
--- a/include/llvm/Target/TargetLowering.h
+++ b/include/llvm/Target/TargetLowering.h
@@ -161,27 +161,27 @@ protected:
 
 public:
   const TargetMachine &getTargetMachine() const { return TM; }
-  const DataLayout *getDataLayout() const { return TM.getDataLayout(); }
 
-  bool isBigEndian() const { return !IsLittleEndian; }
-  bool isLittleEndian() const { return IsLittleEndian; }
   virtual bool useSoftFloat() const { return false; }
 
   /// Return the pointer type for the given address space, defaults to
   /// the pointer type from the data layout.
   /// FIXME: The default needs to be removed once all the code is updated.
-  virtual MVT getPointerTy(uint32_t /*AS*/ = 0) const;
-  unsigned getPointerSizeInBits(uint32_t AS = 0) const;
-  unsigned getPointerTypeSizeInBits(Type *Ty) const;
-  virtual MVT getScalarShiftAmountTy(EVT LHSTy) const;
+  MVT getPointerTy(const DataLayout &DL, uint32_t AS = 0) const {
+    return MVT::getIntegerVT(DL.getPointerSizeInBits(AS));
+  }
+
+  /// EVT is not used in-tree, but is used by out-of-tree target.
+  /// A documentation for this function would be nice...
+  virtual MVT getScalarShiftAmountTy(const DataLayout &, EVT) const;
 
-  EVT getShiftAmountTy(EVT LHSTy) const;
+  EVT getShiftAmountTy(EVT LHSTy, const DataLayout &DL) const;
 
   /// Returns the type to be used for the index operand of:
   /// ISD::INSERT_VECTOR_ELT, ISD::EXTRACT_VECTOR_ELT,
   /// ISD::INSERT_SUBVECTOR, and ISD::EXTRACT_SUBVECTOR
-  virtual MVT getVectorIdxTy() const {
-    return getPointerTy();
+  virtual MVT getVectorIdxTy(const DataLayout &DL) const {
+    return getPointerTy(DL);
   }
 
   /// Return true if the select operation is expensive for this target.
@@ -327,7 +327,8 @@ public:
   }
 
   /// Return the ValueType of the result of SETCC operations.
-  virtual EVT getSetCCResultType(LLVMContext &Context, EVT VT) const;
+  virtual EVT getSetCCResultType(const DataLayout &DL, LLVMContext &Context,
+                                 EVT VT) const;
 
   /// Return the ValueType for comparison libcalls. Comparions libcalls include
   /// floating point comparion calls, and Ordered/Unordered check calls on
@@ -715,17 +716,18 @@ public:
   /// operations except for the pointer size.  If AllowUnknown is true, this
   /// will return MVT::Other for types with no EVT counterpart (e.g. structs),
   /// otherwise it will assert.
-  EVT getValueType(Type *Ty, bool AllowUnknown = false) const {
+  EVT getValueType(const DataLayout &DL, Type *Ty,
+                   bool AllowUnknown = false) const {
     // Lower scalar pointers to native pointer types.
     if (PointerType *PTy = dyn_cast<PointerType>(Ty))
-      return getPointerTy(PTy->getAddressSpace());
+      return getPointerTy(DL, PTy->getAddressSpace());
 
     if (Ty->isVectorTy()) {
       VectorType *VTy = cast<VectorType>(Ty);
       Type *Elm = VTy->getElementType();
       // Lower vectors of pointers to native pointer types.
       if (PointerType *PT = dyn_cast<PointerType>(Elm)) {
-        EVT PointerTy(getPointerTy(PT->getAddressSpace()));
+        EVT PointerTy(getPointerTy(DL, PT->getAddressSpace()));
         Elm = PointerTy.getTypeForEVT(Ty->getContext());
       }
 
@@ -736,14 +738,15 @@ public:
   }
 
   /// Return the MVT corresponding to this LLVM type. See getValueType.
-  MVT getSimpleValueType(Type *Ty, bool AllowUnknown = false) const {
-    return getValueType(Ty, AllowUnknown).getSimpleVT();
+  MVT getSimpleValueType(const DataLayout &DL, Type *Ty,
+                         bool AllowUnknown = false) const {
+    return getValueType(DL, Ty, AllowUnknown).getSimpleVT();
   }
 
   /// Return the desired alignment for ByVal or InAlloca aggregate function
   /// arguments in the caller parameter area.  This is the actual alignment, not
   /// its logarithm.
-  virtual unsigned getByValTypeAlignment(Type *Ty) const;
+  virtual unsigned getByValTypeAlignment(Type *Ty, const DataLayout &DL) const;
 
   /// Return the type of registers that this ValueType will eventually require.
   MVT getRegisterType(MVT VT) const {
@@ -818,8 +821,8 @@ public:
   /// When splitting a value of the specified type into parts, does the Lo
   /// or Hi part come first?  This usually follows the endianness, except
   /// for ppcf128, where the Hi part always comes first.
-  bool hasBigEndianPartOrdering(EVT VT) const {
-    return isBigEndian() || VT == MVT::ppcf128;
+  bool hasBigEndianPartOrdering(EVT VT, const DataLayout &DL) const {
+    return DL.isBigEndian() || VT == MVT::ppcf128;
   }
 
   /// If true, the target has custom DAG combine transformations that it can
@@ -1006,7 +1009,8 @@ public:
   int InstructionOpcodeToISD(unsigned Opcode) const;
 
   /// Estimate the cost of type-legalization and the legalized type.
-  std::pair<unsigned, MVT> getTypeLegalizationCost(Type *Ty) const;
+  std::pair<unsigned, MVT> getTypeLegalizationCost(const DataLayout &DL,
+                                                   Type *Ty) const;
 
   /// @}
 
@@ -1460,8 +1464,8 @@ public:
   /// If the address space cannot be determined, it will be -1.
   ///
   /// TODO: Remove default argument
-  virtual bool isLegalAddressingMode(const AddrMode &AM, Type *Ty,
-                                     unsigned AddrSpace) const;
+  virtual bool isLegalAddressingMode(const DataLayout &DL, const AddrMode &AM,
+                                     Type *Ty, unsigned AddrSpace) const;
 
   /// \brief Return the cost of the scaling factor used in the addressing mode
   /// represented by AM for this target, for a load/store of the specified type.
@@ -1470,10 +1474,10 @@ public:
   /// If the AM is not supported, it returns a negative value.
   /// TODO: Handle pre/postinc as well.
   /// TODO: Remove default argument
-  virtual int getScalingFactorCost(const AddrMode &AM, Type *Ty,
-                                   unsigned AS = 0) const {
+  virtual int getScalingFactorCost(const DataLayout &DL, const AddrMode &AM,
+                                   Type *Ty, unsigned AS = 0) const {
     // Default: assume that any scaling factor used in a legal AM is free.
-    if (isLegalAddressingMode(AM, Ty, AS))
+    if (isLegalAddressingMode(DL, AM, Ty, AS))
       return 0;
     return -1;
   }
@@ -1734,9 +1738,6 @@ public:
 private:
   const TargetMachine &TM;
 
-  /// True if this is a little endian target.
-  bool IsLittleEndian;
-
   /// Tells the code generator not to expand operations into sequences that use
   /// the select operations if possible.
   bool SelectIsExpensive;
@@ -2414,6 +2415,7 @@ public:
     ArgListTy &getArgs() {
       return Args;
     }
+
   };
 
   /// This function lowers an abstract call to a function into an actual call.
@@ -2485,7 +2487,8 @@ public:
   /// Return the register ID of the name passed in. Used by named register
   /// global variables extension. There is no target-independent behaviour
   /// so the default action is to bail.
-  virtual unsigned getRegisterByName(const char* RegName, EVT VT) const {
+  virtual unsigned getRegisterByName(const char* RegName, EVT VT,
+                                     SelectionDAG &DAG) const {
     report_fatal_error("Named registers not implemented for this target");
   }
 
@@ -2657,7 +2660,8 @@ public:
   /// specific constraints and their prefixes, and also tie in the associated
   /// operand values.  If this returns an empty vector, and if the constraint
   /// string itself isn't empty, there was an error parsing.
-  virtual AsmOperandInfoVector ParseConstraints(const TargetRegisterInfo *TRI,
+  virtual AsmOperandInfoVector ParseConstraints(const DataLayout &DL,
+                                                const TargetRegisterInfo *TRI,
                                                 ImmutableCallSite CS) const;
 
   /// Examine constraint type and operand type and determine a weight value.
@@ -2679,7 +2683,7 @@ public:
                                       SelectionDAG *DAG = nullptr) const;
 
   /// Given a constraint, return the type of constraint it is for this target.
-  virtual ConstraintType getConstraintType(const std::string &Constraint) const;
+  virtual ConstraintType getConstraintType(StringRef Constraint) const;
 
   /// Given a physical register constraint (e.g.  {edx}), return the register
   /// number and the register class for the register.
@@ -2692,10 +2696,9 @@ public:
   /// returns a register number of 0 and a null register class pointer.
   virtual std::pair<unsigned, const TargetRegisterClass *>
   getRegForInlineAsmConstraint(const TargetRegisterInfo *TRI,
-                               const std::string &Constraint, MVT VT) const;
+                               StringRef Constraint, MVT VT) const;
 
-  virtual unsigned
-  getInlineAsmMemConstraint(const std::string &ConstraintCode) const {
+  virtual unsigned getInlineAsmMemConstraint(StringRef ConstraintCode) const {
     if (ConstraintCode == "i")
       return InlineAsm::Constraint_i;
     else if (ConstraintCode == "m")
@@ -2823,9 +2826,9 @@ public:
 /// Given an LLVM IR type and return type attributes, compute the return value
 /// EVTs and flags, and optionally also the offsets, if the return value is
 /// being lowered to memory.
-void GetReturnInfo(Type* ReturnType, AttributeSet attr,
+void GetReturnInfo(Type *ReturnType, AttributeSet attr,
                    SmallVectorImpl<ISD::OutputArg> &Outs,
-                   const TargetLowering &TLI);
+                   const TargetLowering &TLI, const DataLayout &DL);
 
 } // end llvm namespace
 
diff --git a/include/llvm/Target/TargetMachine.h b/include/llvm/Target/TargetMachine.h
index 64a923b80edf..06a2b13836ed 100644
--- a/include/llvm/Target/TargetMachine.h
+++ b/include/llvm/Target/TargetMachine.h
@@ -212,8 +212,8 @@ public:
   /// supported, or false on success.
   virtual bool addPassesToEmitFile(
       PassManagerBase &, raw_pwrite_stream &, CodeGenFileType,
-      bool /*DisableVerify*/ = true, AnalysisID /*StartAfter*/ = nullptr,
-      AnalysisID /*StopAfter*/ = nullptr,
+      bool /*DisableVerify*/ = true, AnalysisID /*StartBefore*/ = nullptr,
+      AnalysisID /*StartAfter*/ = nullptr, AnalysisID /*StopAfter*/ = nullptr,
       MachineFunctionInitializer * /*MFInitializer*/ = nullptr) {
     return true;
   }
@@ -260,8 +260,8 @@ public:
   /// emitted.  Typically this will involve several steps of code generation.
   bool addPassesToEmitFile(
       PassManagerBase &PM, raw_pwrite_stream &Out, CodeGenFileType FileType,
-      bool DisableVerify = true, AnalysisID StartAfter = nullptr,
-      AnalysisID StopAfter = nullptr,
+      bool DisableVerify = true, AnalysisID StartBefore = nullptr,
+      AnalysisID StartAfter = nullptr, AnalysisID StopAfter = nullptr,
       MachineFunctionInitializer *MFInitializer = nullptr) override;
 
   /// Add passes to the specified pass manager to get machine code emitted with
diff --git a/include/llvm/Target/TargetOpcodes.h b/include/llvm/Target/TargetOpcodes.h
index 1f9a5d4ecaf0..50197191109d 100644
--- a/include/llvm/Target/TargetOpcodes.h
+++ b/include/llvm/Target/TargetOpcodes.h
@@ -118,10 +118,10 @@ enum {
   /// collectors and deoptimizations in either the callee or caller.
   STATEPOINT = 20,
 
-  /// Instruction that records the offset of a function's frame allocation in a
-  /// label. Created by the llvm.frameallocate intrinsic. It has two arguments:
-  /// the symbol for the label and the frame index of the stack allocation.
-  FRAME_ALLOC = 21,
+  /// Instruction that records the offset of a local stack allocation passed to
+  /// llvm.localescape. It has two arguments: the symbol for the label and the
+  /// frame index of the local stack allocation.
+  LOCAL_ESCAPE = 21,
 
   /// Loading instruction that may page fault, bundled with associated
   /// information on how to handle such a page fault.  It is intended to support
diff --git a/include/llvm/Target/TargetSelectionDAGInfo.h b/include/llvm/Target/TargetSelectionDAGInfo.h
index bacdd950705b..53db5aa84292 100644
--- a/include/llvm/Target/TargetSelectionDAGInfo.h
+++ b/include/llvm/Target/TargetSelectionDAGInfo.h
@@ -20,8 +20,6 @@
 
 namespace llvm {
 
-class DataLayout;
-
 //===----------------------------------------------------------------------===//
 /// TargetSelectionDAGInfo - Targets can subclass this to parameterize the
 /// SelectionDAG lowering and instruction selection process.
@@ -30,13 +28,8 @@ class TargetSelectionDAGInfo {
   TargetSelectionDAGInfo(const TargetSelectionDAGInfo &) = delete;
   void operator=(const TargetSelectionDAGInfo &) = delete;
 
-  const DataLayout *DL;
-
-protected:
-  const DataLayout *getDataLayout() const { return DL; }
-
 public:
-  explicit TargetSelectionDAGInfo(const DataLayout *DL);
+  explicit TargetSelectionDAGInfo() = default;
   virtual ~TargetSelectionDAGInfo();
 
   /// EmitTargetCodeForMemcpy - Emit target-specific code that performs a
diff --git a/include/llvm/Target/TargetSubtargetInfo.h b/include/llvm/Target/TargetSubtargetInfo.h
index e42c56add7b1..07c0c66bfa18 100644
--- a/include/llvm/Target/TargetSubtargetInfo.h
+++ b/include/llvm/Target/TargetSubtargetInfo.h
@@ -44,9 +44,17 @@ template <typename T> class SmallVectorImpl;
 class TargetSubtargetInfo : public MCSubtargetInfo {
   TargetSubtargetInfo(const TargetSubtargetInfo &) = delete;
   void operator=(const TargetSubtargetInfo &) = delete;
+  TargetSubtargetInfo() = delete;
 
 protected: // Can only create subclasses...
-  TargetSubtargetInfo();
+  TargetSubtargetInfo(const Triple &TT, StringRef CPU, StringRef FS,
+                      ArrayRef<SubtargetFeatureKV> PF,
+                      ArrayRef<SubtargetFeatureKV> PD,
+                      const SubtargetInfoKV *ProcSched,
+                      const MCWriteProcResEntry *WPR,
+                      const MCWriteLatencyEntry *WL,
+                      const MCReadAdvanceEntry *RA, const InstrStage *IS,
+                      const unsigned *OC, const unsigned *FP);
 
 public:
   // AntiDepBreakMode - Type of anti-dependence breaking that should
diff --git a/include/llvm/Transforms/IPO.h b/include/llvm/Transforms/IPO.h
index fbd999cbc946..2ea47301bb4c 100644
--- a/include/llvm/Transforms/IPO.h
+++ b/include/llvm/Transforms/IPO.h
@@ -71,6 +71,12 @@ ModulePass *createGlobalOptimizerPass();
 ModulePass *createGlobalDCEPass();
 
 //===----------------------------------------------------------------------===//
+/// This transform is designed to eliminate available external globals
+/// (functions or global variables)
+///
+ModulePass *createEliminateAvailableExternallyPass();
+
+//===----------------------------------------------------------------------===//
 /// createGVExtractionPass - If deleteFn is true, this pass deletes
 /// the specified global values. Otherwise, it deletes as much of the module as
 /// possible, except for the global values specified.
diff --git a/include/llvm/Transforms/IPO/PassManagerBuilder.h b/include/llvm/Transforms/IPO/PassManagerBuilder.h
index 5d574ae0bf0f..1334dd0da23c 100644
--- a/include/llvm/Transforms/IPO/PassManagerBuilder.h
+++ b/include/llvm/Transforms/IPO/PassManagerBuilder.h
@@ -121,6 +121,7 @@ public:
   bool VerifyInput;
   bool VerifyOutput;
   bool MergeFunctions;
+  bool PrepareForLTO;
 
 private:
   /// ExtensionList - This is list of all of the extensions that are registered.
diff --git a/include/llvm/Transforms/Utils/Cloning.h b/include/llvm/Transforms/Utils/Cloning.h
index cb187ec103d0..2caa9a2462df 100644
--- a/include/llvm/Transforms/Utils/Cloning.h
+++ b/include/llvm/Transforms/Utils/Cloning.h
@@ -45,6 +45,7 @@ class LoopInfo;
 class AllocaInst;
 class AliasAnalysis;
 class AssumptionCacheTracker;
+class DominatorTree;
 
 /// CloneModule - Return an exact copy of the specified module
 ///
@@ -233,6 +234,21 @@ bool InlineFunction(InvokeInst *II, InlineFunctionInfo &IFI,
 bool InlineFunction(CallSite CS, InlineFunctionInfo &IFI,
                     bool InsertLifetime = true);
 
+/// \brief Clones a loop \p OrigLoop.  Returns the loop and the blocks in \p
+/// Blocks.
+///
+/// Updates LoopInfo and DominatorTree assuming the loop is dominated by block
+/// \p LoopDomBB.  Insert the new blocks before block specified in \p Before.
+Loop *cloneLoopWithPreheader(BasicBlock *Before, BasicBlock *LoopDomBB,
+                             Loop *OrigLoop, ValueToValueMapTy &VMap,
+                             const Twine &NameSuffix, LoopInfo *LI,
+                             DominatorTree *DT,
+                             SmallVectorImpl<BasicBlock *> &Blocks);
+
+/// \brief Remaps instructions in \p Blocks using the mapping in \p VMap.
+void remapInstructionsInBlocks(const SmallVectorImpl<BasicBlock *> &Blocks,
+                               ValueToValueMapTy &VMap);
+
 } // End llvm namespace
 
 #endif
diff --git a/include/llvm/Transforms/Utils/LoopVersioning.h b/include/llvm/Transforms/Utils/LoopVersioning.h
new file mode 100644
index 000000000000..009fba48c6a3
--- /dev/null
+++ b/include/llvm/Transforms/Utils/LoopVersioning.h
@@ -0,0 +1,100 @@
+//===- LoopVersioning.h - Utility to version a loop -------------*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// This file defines a utility class to perform loop versioning.  The versioned
+// loop speculates that otherwise may-aliasing memory accesses don't overlap and
+// emits checks to prove this.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_TRANSFORMS_UTILS_LOOPVERSIONING_H
+#define LLVM_TRANSFORMS_UTILS_LOOPVERSIONING_H
+
+#include "llvm/Transforms/Utils/ValueMapper.h"
+
+namespace llvm {
+
+class Loop;
+class LoopAccessInfo;
+class LoopInfo;
+
+/// \brief This class emits a version of the loop where run-time checks ensure
+/// that may-alias pointers can't overlap.
+///
+/// It currently only supports single-exit loops and assumes that the loop
+/// already has a preheader.
+class LoopVersioning {
+public:
+  LoopVersioning(const LoopAccessInfo &LAI, Loop *L, LoopInfo *LI,
+                 DominatorTree *DT,
+                 const SmallVector<int, 8> *PtrToPartition = nullptr);
+
+  /// \brief Returns true if we need memchecks to disambiguate may-aliasing
+  /// accesses.
+  bool needsRuntimeChecks() const;
+
+  /// \brief Performs the CFG manipulation part of versioning the loop including
+  /// the DominatorTree and LoopInfo updates.
+  ///
+  /// The loop that was used to construct the class will be the "versioned" loop
+  /// i.e. the loop that will receive control if all the memchecks pass.
+  ///
+  /// This allows the loop transform pass to operate on the same loop regardless
+  /// of whether versioning was necessary or not:
+  ///
+  ///    for each loop L:
+  ///        analyze L
+  ///        if versioning is necessary version L
+  ///        transform L
+  void versionLoop(Pass *P);
+
+  /// \brief Adds the necessary PHI nodes for the versioned loops based on the
+  /// loop-defined values used outside of the loop.
+  ///
+  /// This needs to be called after versionLoop if there are defs in the loop
+  /// that are used outside the loop.  FIXME: this should be invoked internally
+  /// by versionLoop and made private.
+  void addPHINodes(const SmallVectorImpl<Instruction *> &DefsUsedOutside);
+
+  /// \brief Returns the versioned loop.  Control flows here if pointers in the
+  /// loop don't alias (i.e. all memchecks passed).  (This loop is actually the
+  /// same as the original loop that we got constructed with.)
+  Loop *getVersionedLoop() { return VersionedLoop; }
+
+  /// \brief Returns the fall-back loop.  Control flows here if pointers in the
+  /// loop may alias (i.e. one of the memchecks failed).
+  Loop *getNonVersionedLoop() { return NonVersionedLoop; }
+
+private:
+  /// \brief The original loop.  This becomes the "versioned" one.  I.e.,
+  /// control flows here if pointers in the loop don't alias.
+  Loop *VersionedLoop;
+  /// \brief The fall-back loop.  I.e. control flows here if pointers in the
+  /// loop may alias (memchecks failed).
+  Loop *NonVersionedLoop;
+
+  /// \brief For each memory pointer it contains the partitionId it is used in.
+  /// If nullptr, no partitioning is used.
+  ///
+  /// The I-th entry corresponds to I-th entry in LAI.getRuntimePointerCheck().
+  /// If the pointer is used in multiple partitions the entry is set to -1.
+  const SmallVector<int, 8> *PtrToPartition;
+
+  /// \brief This maps the instructions from VersionedLoop to their counterpart
+  /// in NonVersionedLoop.
+  ValueToValueMapTy VMap;
+
+  /// \brief Analyses used.
+  const LoopAccessInfo &LAI;
+  LoopInfo *LI;
+  DominatorTree *DT;
+};
+}
+
+#endif
diff --git a/lib/Analysis/AliasAnalysis.cpp b/lib/Analysis/AliasAnalysis.cpp
index ad0727a0e0e5..44d137dffd22 100644
--- a/lib/Analysis/AliasAnalysis.cpp
+++ b/lib/Analysis/AliasAnalysis.cpp
@@ -71,11 +71,6 @@ void AliasAnalysis::deleteValue(Value *V) {
   AA->deleteValue(V);
 }
 
-void AliasAnalysis::copyValue(Value *From, Value *To) {
-  assert(AA && "AA didn't call InitializeAliasAnalysis in its run method!");
-  AA->copyValue(From, To);
-}
-
 void AliasAnalysis::addEscapingUse(Use &U) {
   assert(AA && "AA didn't call InitializeAliasAnalysis in its run method!");
   AA->addEscapingUse(U);
diff --git a/lib/Analysis/AliasDebugger.cpp b/lib/Analysis/AliasDebugger.cpp
index 1ef49fc02fef..e5107b3bc827 100644
--- a/lib/Analysis/AliasDebugger.cpp
+++ b/lib/Analysis/AliasDebugger.cpp
@@ -124,10 +124,6 @@ namespace {
       assert(Vals.find(V) != Vals.end() && "Never seen value in AA before");
       AliasAnalysis::deleteValue(V);
     }
-    void copyValue(Value *From, Value *To) override {
-      Vals.insert(To);
-      AliasAnalysis::copyValue(From, To);
-    }
 
   };
 }
diff --git a/lib/Analysis/AliasSetTracker.cpp b/lib/Analysis/AliasSetTracker.cpp
index bf8cda1ffaec..54d0f4304e1f 100644
--- a/lib/Analysis/AliasSetTracker.cpp
+++ b/lib/Analysis/AliasSetTracker.cpp
@@ -544,9 +544,6 @@ void AliasSetTracker::deleteValue(Value *PtrVal) {
 // the tracker already knows about a value, it will ignore the request.
 //
 void AliasSetTracker::copyValue(Value *From, Value *To) {
-  // Notify the alias analysis implementation that this value is copied.
-  AA.copyValue(From, To);
-
   // First, look up the PointerRec for this pointer.
   PointerMapType::iterator I = PointerMap.find_as(From);
   if (I == PointerMap.end())
diff --git a/lib/Analysis/BasicAliasAnalysis.cpp b/lib/Analysis/BasicAliasAnalysis.cpp
index 8e812252fdfe..68f766edb301 100644
--- a/lib/Analysis/BasicAliasAnalysis.cpp
+++ b/lib/Analysis/BasicAliasAnalysis.cpp
@@ -685,6 +685,9 @@ BasicAliasAnalysis::getModRefBehavior(ImmutableCallSite CS) {
   if (CS.onlyReadsMemory())
     Min = OnlyReadsMemory;
 
+  if (CS.onlyAccessesArgMemory())
+    Min = ModRefBehavior(Min & OnlyAccessesArgumentPointees);
+
   // The AliasAnalysis base class has some smarts, lets use them.
   return ModRefBehavior(AliasAnalysis::getModRefBehavior(CS) & Min);
 }
@@ -710,6 +713,9 @@ BasicAliasAnalysis::getModRefBehavior(const Function *F) {
   if (F->onlyReadsMemory())
     Min = OnlyReadsMemory;
 
+  if (F->onlyAccessesArgMemory())
+    Min = ModRefBehavior(Min & OnlyAccessesArgumentPointees);
+
   const TargetLibraryInfo &TLI =
       getAnalysis<TargetLibraryInfoWrapperPass>().getTLI();
   if (isMemsetPattern16(F, TLI))
diff --git a/lib/Analysis/ConstantFolding.cpp b/lib/Analysis/ConstantFolding.cpp
index 2f4c6a92f9af..02a5aef03223 100644
--- a/lib/Analysis/ConstantFolding.cpp
+++ b/lib/Analysis/ConstantFolding.cpp
@@ -1234,6 +1234,8 @@ bool llvm::canConstantFoldCallTo(const Function *F) {
   case Intrinsic::floor:
   case Intrinsic::ceil:
   case Intrinsic::sqrt:
+  case Intrinsic::sin:
+  case Intrinsic::cos:
   case Intrinsic::pow:
   case Intrinsic::powi:
   case Intrinsic::bswap:
@@ -1450,6 +1452,10 @@ static Constant *ConstantFoldScalarCall(StringRef Name, unsigned IntrinsicID,
           return ConstantFoldFP(floor, V, Ty);
         case Intrinsic::ceil:
           return ConstantFoldFP(ceil, V, Ty);
+        case Intrinsic::sin:
+          return ConstantFoldFP(sin, V, Ty);
+        case Intrinsic::cos:
+          return ConstantFoldFP(cos, V, Ty);
       }
 
       if (!TLI)
diff --git a/lib/Analysis/IPA/GlobalsModRef.cpp b/lib/Analysis/IPA/GlobalsModRef.cpp
index f1ddde252924..18d45dd6a396 100644
--- a/lib/Analysis/IPA/GlobalsModRef.cpp
+++ b/lib/Analysis/IPA/GlobalsModRef.cpp
@@ -42,94 +42,111 @@ STATISTIC(NumReadMemFunctions, "Number of functions that only read memory");
 STATISTIC(NumIndirectGlobalVars, "Number of indirect global objects");
 
 namespace {
-  /// FunctionRecord - One instance of this structure is stored for every
-  /// function in the program.  Later, the entries for these functions are
-  /// removed if the function is found to call an external function (in which
-  /// case we know nothing about it.
-  struct FunctionRecord {
-    /// GlobalInfo - Maintain mod/ref info for all of the globals without
-    /// addresses taken that are read or written (transitively) by this
-    /// function.
-    std::map<const GlobalValue*, unsigned> GlobalInfo;
-
-    /// MayReadAnyGlobal - May read global variables, but it is not known which.
-    bool MayReadAnyGlobal;
-
-    unsigned getInfoForGlobal(const GlobalValue *GV) const {
-      unsigned Effect = MayReadAnyGlobal ? AliasAnalysis::Ref : 0;
-      std::map<const GlobalValue*, unsigned>::const_iterator I =
+/// FunctionRecord - One instance of this structure is stored for every
+/// function in the program.  Later, the entries for these functions are
+/// removed if the function is found to call an external function (in which
+/// case we know nothing about it.
+struct FunctionRecord {
+  /// GlobalInfo - Maintain mod/ref info for all of the globals without
+  /// addresses taken that are read or written (transitively) by this
+  /// function.
+  std::map<const GlobalValue *, unsigned> GlobalInfo;
+
+  /// MayReadAnyGlobal - May read global variables, but it is not known which.
+  bool MayReadAnyGlobal;
+
+  unsigned getInfoForGlobal(const GlobalValue *GV) const {
+    unsigned Effect = MayReadAnyGlobal ? AliasAnalysis::Ref : 0;
+    std::map<const GlobalValue *, unsigned>::const_iterator I =
         GlobalInfo.find(GV);
-      if (I != GlobalInfo.end())
-        Effect |= I->second;
-      return Effect;
-    }
+    if (I != GlobalInfo.end())
+      Effect |= I->second;
+    return Effect;
+  }
 
-    /// FunctionEffect - Capture whether or not this function reads or writes to
-    /// ANY memory.  If not, we can do a lot of aggressive analysis on it.
-    unsigned FunctionEffect;
+  /// FunctionEffect - Capture whether or not this function reads or writes to
+  /// ANY memory.  If not, we can do a lot of aggressive analysis on it.
+  unsigned FunctionEffect;
 
-    FunctionRecord() : MayReadAnyGlobal (false), FunctionEffect(0) {}
-  };
+  FunctionRecord() : MayReadAnyGlobal(false), FunctionEffect(0) {}
+};
 
-  /// GlobalsModRef - The actual analysis pass.
-  class GlobalsModRef : public ModulePass, public AliasAnalysis {
-    /// NonAddressTakenGlobals - The globals that do not have their addresses
-    /// taken.
-    std::set<const GlobalValue*> NonAddressTakenGlobals;
+/// GlobalsModRef - The actual analysis pass.
+class GlobalsModRef : public ModulePass, public AliasAnalysis {
+  /// NonAddressTakenGlobals - The globals that do not have their addresses
+  /// taken.
+  std::set<const GlobalValue *> NonAddressTakenGlobals;
 
-    /// IndirectGlobals - The memory pointed to by this global is known to be
-    /// 'owned' by the global.
-    std::set<const GlobalValue*> IndirectGlobals;
+  /// IndirectGlobals - The memory pointed to by this global is known to be
+  /// 'owned' by the global.
+  std::set<const GlobalValue *> IndirectGlobals;
 
-    /// AllocsForIndirectGlobals - If an instruction allocates memory for an
-    /// indirect global, this map indicates which one.
-    std::map<const Value*, const GlobalValue*> AllocsForIndirectGlobals;
+  /// AllocsForIndirectGlobals - If an instruction allocates memory for an
+  /// indirect global, this map indicates which one.
+  std::map<const Value *, const GlobalValue *> AllocsForIndirectGlobals;
 
-    /// FunctionInfo - For each function, keep track of what globals are
-    /// modified or read.
-    std::map<const Function*, FunctionRecord> FunctionInfo;
+  /// FunctionInfo - For each function, keep track of what globals are
+  /// modified or read.
+  std::map<const Function *, FunctionRecord> FunctionInfo;
 
-  public:
-    static char ID;
-    GlobalsModRef() : ModulePass(ID) {
-      initializeGlobalsModRefPass(*PassRegistry::getPassRegistry());
-    }
+public:
+  static char ID;
+  GlobalsModRef() : ModulePass(ID) {
+    initializeGlobalsModRefPass(*PassRegistry::getPassRegistry());
+  }
 
-    bool runOnModule(Module &M) override {
-      InitializeAliasAnalysis(this, &M.getDataLayout());
+  bool runOnModule(Module &M) override {
+    InitializeAliasAnalysis(this, &M.getDataLayout());
 
-      // Find non-addr taken globals.
-      AnalyzeGlobals(M);
+    // Find non-addr taken globals.
+    AnalyzeGlobals(M);
 
-      // Propagate on CG.
-      AnalyzeCallGraph(getAnalysis<CallGraphWrapperPass>().getCallGraph(), M);
-      return false;
-    }
+    // Propagate on CG.
+    AnalyzeCallGraph(getAnalysis<CallGraphWrapperPass>().getCallGraph(), M);
+    return false;
+  }
 
-    void getAnalysisUsage(AnalysisUsage &AU) const override {
-      AliasAnalysis::getAnalysisUsage(AU);
-      AU.addRequired<CallGraphWrapperPass>();
-      AU.setPreservesAll();                         // Does not transform code
-    }
+  void getAnalysisUsage(AnalysisUsage &AU) const override {
+    AliasAnalysis::getAnalysisUsage(AU);
+    AU.addRequired<CallGraphWrapperPass>();
+    AU.setPreservesAll(); // Does not transform code
+  }
+
+  //------------------------------------------------
+  // Implement the AliasAnalysis API
+  //
+  AliasResult alias(const MemoryLocation &LocA,
+                    const MemoryLocation &LocB) override;
+  ModRefResult getModRefInfo(ImmutableCallSite CS,
+                             const MemoryLocation &Loc) override;
+  ModRefResult getModRefInfo(ImmutableCallSite CS1,
+                             ImmutableCallSite CS2) override {
+    return AliasAnalysis::getModRefInfo(CS1, CS2);
+  }
 
-    //------------------------------------------------
-    // Implement the AliasAnalysis API
-    //
-    AliasResult alias(const MemoryLocation &LocA,
-                      const MemoryLocation &LocB) override;
-    ModRefResult getModRefInfo(ImmutableCallSite CS,
-                               const MemoryLocation &Loc) override;
-    ModRefResult getModRefInfo(ImmutableCallSite CS1,
-                               ImmutableCallSite CS2) override {
-      return AliasAnalysis::getModRefInfo(CS1, CS2);
+  /// getModRefBehavior - Return the behavior of the specified function if
+  /// called from the specified call site.  The call site may be null in which
+  /// case the most generic behavior of this function should be returned.
+  ModRefBehavior getModRefBehavior(const Function *F) override {
+    ModRefBehavior Min = UnknownModRefBehavior;
+
+    if (FunctionRecord *FR = getFunctionInfo(F)) {
+      if (FR->FunctionEffect == 0)
+        Min = DoesNotAccessMemory;
+      else if ((FR->FunctionEffect & Mod) == 0)
+        Min = OnlyReadsMemory;
     }
 
-    /// getModRefBehavior - Return the behavior of the specified function if
-    /// called from the specified call site.  The call site may be null in which
-    /// case the most generic behavior of this function should be returned.
-    ModRefBehavior getModRefBehavior(const Function *F) override {
-      ModRefBehavior Min = UnknownModRefBehavior;
+    return ModRefBehavior(AliasAnalysis::getModRefBehavior(F) & Min);
+  }
+
+  /// getModRefBehavior - Return the behavior of the specified function if
+  /// called from the specified call site.  The call site may be null in which
+  /// case the most generic behavior of this function should be returned.
+  ModRefBehavior getModRefBehavior(ImmutableCallSite CS) override {
+    ModRefBehavior Min = UnknownModRefBehavior;
 
+    if (const Function *F = CS.getCalledFunction())
       if (FunctionRecord *FR = getFunctionInfo(F)) {
         if (FR->FunctionEffect == 0)
           Min = DoesNotAccessMemory;
@@ -137,68 +154,50 @@ namespace {
           Min = OnlyReadsMemory;
       }
 
-      return ModRefBehavior(AliasAnalysis::getModRefBehavior(F) & Min);
-    }
-    
-    /// getModRefBehavior - Return the behavior of the specified function if
-    /// called from the specified call site.  The call site may be null in which
-    /// case the most generic behavior of this function should be returned.
-    ModRefBehavior getModRefBehavior(ImmutableCallSite CS) override {
-      ModRefBehavior Min = UnknownModRefBehavior;
-
-      if (const Function* F = CS.getCalledFunction())
-        if (FunctionRecord *FR = getFunctionInfo(F)) {
-          if (FR->FunctionEffect == 0)
-            Min = DoesNotAccessMemory;
-          else if ((FR->FunctionEffect & Mod) == 0)
-            Min = OnlyReadsMemory;
-        }
+    return ModRefBehavior(AliasAnalysis::getModRefBehavior(CS) & Min);
+  }
 
-      return ModRefBehavior(AliasAnalysis::getModRefBehavior(CS) & Min);
-    }
+  void deleteValue(Value *V) override;
+  void addEscapingUse(Use &U) override;
+
+  /// getAdjustedAnalysisPointer - This method is used when a pass implements
+  /// an analysis interface through multiple inheritance.  If needed, it
+  /// should override this to adjust the this pointer as needed for the
+  /// specified pass info.
+  void *getAdjustedAnalysisPointer(AnalysisID PI) override {
+    if (PI == &AliasAnalysis::ID)
+      return (AliasAnalysis *)this;
+    return this;
+  }
 
-    void deleteValue(Value *V) override;
-    void copyValue(Value *From, Value *To) override;
-    void addEscapingUse(Use &U) override;
-
-    /// getAdjustedAnalysisPointer - This method is used when a pass implements
-    /// an analysis interface through multiple inheritance.  If needed, it
-    /// should override this to adjust the this pointer as needed for the
-    /// specified pass info.
-    void *getAdjustedAnalysisPointer(AnalysisID PI) override {
-      if (PI == &AliasAnalysis::ID)
-        return (AliasAnalysis*)this;
-      return this;
-    }
-    
-  private:
-    /// getFunctionInfo - Return the function info for the function, or null if
-    /// we don't have anything useful to say about it.
-    FunctionRecord *getFunctionInfo(const Function *F) {
-      std::map<const Function*, FunctionRecord>::iterator I =
+private:
+  /// getFunctionInfo - Return the function info for the function, or null if
+  /// we don't have anything useful to say about it.
+  FunctionRecord *getFunctionInfo(const Function *F) {
+    std::map<const Function *, FunctionRecord>::iterator I =
         FunctionInfo.find(F);
-      if (I != FunctionInfo.end())
-        return &I->second;
-      return nullptr;
-    }
+    if (I != FunctionInfo.end())
+      return &I->second;
+    return nullptr;
+  }
 
-    void AnalyzeGlobals(Module &M);
-    void AnalyzeCallGraph(CallGraph &CG, Module &M);
-    bool AnalyzeUsesOfPointer(Value *V, std::vector<Function*> &Readers,
-                              std::vector<Function*> &Writers,
-                              GlobalValue *OkayStoreDest = nullptr);
-    bool AnalyzeIndirectGlobalMemory(GlobalValue *GV);
-  };
+  void AnalyzeGlobals(Module &M);
+  void AnalyzeCallGraph(CallGraph &CG, Module &M);
+  bool AnalyzeUsesOfPointer(Value *V, std::vector<Function *> &Readers,
+                            std::vector<Function *> &Writers,
+                            GlobalValue *OkayStoreDest = nullptr);
+  bool AnalyzeIndirectGlobalMemory(GlobalValue *GV);
+};
 }
 
 char GlobalsModRef::ID = 0;
-INITIALIZE_AG_PASS_BEGIN(GlobalsModRef, AliasAnalysis,
-                "globalsmodref-aa", "Simple mod/ref analysis for globals",    
-                false, true, false)
+INITIALIZE_AG_PASS_BEGIN(GlobalsModRef, AliasAnalysis, "globalsmodref-aa",
+                         "Simple mod/ref analysis for globals", false, true,
+                         false)
 INITIALIZE_PASS_DEPENDENCY(CallGraphWrapperPass)
-INITIALIZE_AG_PASS_END(GlobalsModRef, AliasAnalysis,
-                "globalsmodref-aa", "Simple mod/ref analysis for globals",    
-                false, true, false)
+INITIALIZE_AG_PASS_END(GlobalsModRef, AliasAnalysis, "globalsmodref-aa",
+                       "Simple mod/ref analysis for globals", false, true,
+                       false)
 
 Pass *llvm::createGlobalsModRefPass() { return new GlobalsModRef(); }
 
@@ -207,7 +206,7 @@ Pass *llvm::createGlobalsModRefPass() { return new GlobalsModRef(); }
 /// (really, their address passed to something nontrivial), record this fact,
 /// and record the functions that they are used directly in.
 void GlobalsModRef::AnalyzeGlobals(Module &M) {
-  std::vector<Function*> Readers, Writers;
+  std::vector<Function *> Readers, Writers;
   for (Module::iterator I = M.begin(), E = M.end(); I != E; ++I)
     if (I->hasLocalLinkage()) {
       if (!AnalyzeUsesOfPointer(I, Readers, Writers)) {
@@ -215,11 +214,12 @@ void GlobalsModRef::AnalyzeGlobals(Module &M) {
         NonAddressTakenGlobals.insert(I);
         ++NumNonAddrTakenFunctions;
       }
-      Readers.clear(); Writers.clear();
+      Readers.clear();
+      Writers.clear();
     }
 
-  for (Module::global_iterator I = M.global_begin(), E = M.global_end();
-       I != E; ++I)
+  for (Module::global_iterator I = M.global_begin(), E = M.global_end(); I != E;
+       ++I)
     if (I->hasLocalLinkage()) {
       if (!AnalyzeUsesOfPointer(I, Readers, Writers)) {
         // Remember that we are tracking this global, and the mod/ref fns
@@ -228,7 +228,7 @@ void GlobalsModRef::AnalyzeGlobals(Module &M) {
         for (unsigned i = 0, e = Readers.size(); i != e; ++i)
           FunctionInfo[Readers[i]].GlobalInfo[I] |= Ref;
 
-        if (!I->isConstant())  // No need to keep track of writers to constants
+        if (!I->isConstant()) // No need to keep track of writers to constants
           for (unsigned i = 0, e = Writers.size(); i != e; ++i)
             FunctionInfo[Writers[i]].GlobalInfo[I] |= Mod;
         ++NumNonAddrTakenGlobalVars;
@@ -238,7 +238,8 @@ void GlobalsModRef::AnalyzeGlobals(Module &M) {
             AnalyzeIndirectGlobalMemory(I))
           ++NumIndirectGlobalVars;
       }
-      Readers.clear(); Writers.clear();
+      Readers.clear();
+      Writers.clear();
     }
 }
 
@@ -249,10 +250,11 @@ void GlobalsModRef::AnalyzeGlobals(Module &M) {
 ///
 /// If OkayStoreDest is non-null, stores into this global are allowed.
 bool GlobalsModRef::AnalyzeUsesOfPointer(Value *V,
-                                         std::vector<Function*> &Readers,
-                                         std::vector<Function*> &Writers,
+                                         std::vector<Function *> &Readers,
+                                         std::vector<Function *> &Writers,
                                          GlobalValue *OkayStoreDest) {
-  if (!V->getType()->isPointerTy()) return true;
+  if (!V->getType()->isPointerTy())
+    return true;
 
   for (Use &U : V->uses()) {
     User *I = U.getUser();
@@ -262,7 +264,7 @@ bool GlobalsModRef::AnalyzeUsesOfPointer(Value *V,
       if (V == SI->getOperand(1)) {
         Writers.push_back(SI->getParent()->getParent());
       } else if (SI->getOperand(1) != OkayStoreDest) {
-        return true;  // Storing the pointer
+        return true; // Storing the pointer
       }
     } else if (Operator::getOpcode(I) == Instruction::GetElementPtr) {
       if (AnalyzeUsesOfPointer(I, Readers, Writers))
@@ -282,7 +284,7 @@ bool GlobalsModRef::AnalyzeUsesOfPointer(Value *V,
       }
     } else if (ICmpInst *ICI = dyn_cast<ICmpInst>(I)) {
       if (!isa<ConstantPointerNull>(ICI->getOperand(1)))
-        return true;  // Allow comparison against null.
+        return true; // Allow comparison against null.
     } else {
       return true;
     }
@@ -301,7 +303,7 @@ bool GlobalsModRef::AnalyzeUsesOfPointer(Value *V,
 bool GlobalsModRef::AnalyzeIndirectGlobalMemory(GlobalValue *GV) {
   // Keep track of values related to the allocation of the memory, f.e. the
   // value produced by the malloc call and any casts.
-  std::vector<Value*> AllocRelatedValues;
+  std::vector<Value *> AllocRelatedValues;
 
   // Walk the user list of the global.  If we find anything other than a direct
   // load or store, bail out.
@@ -310,13 +312,14 @@ bool GlobalsModRef::AnalyzeIndirectGlobalMemory(GlobalValue *GV) {
       // The pointer loaded from the global can only be used in simple ways:
       // we allow addressing of it and loading storing to it.  We do *not* allow
       // storing the loaded pointer somewhere else or passing to a function.
-      std::vector<Function*> ReadersWriters;
+      std::vector<Function *> ReadersWriters;
       if (AnalyzeUsesOfPointer(LI, ReadersWriters, ReadersWriters))
-        return false;  // Loaded pointer escapes.
+        return false; // Loaded pointer escapes.
       // TODO: Could try some IP mod/ref of the loaded pointer.
     } else if (StoreInst *SI = dyn_cast<StoreInst>(U)) {
       // Storing the global itself.
-      if (SI->getOperand(0) == GV) return false;
+      if (SI->getOperand(0) == GV)
+        return false;
 
       // If storing the null pointer, ignore it.
       if (isa<ConstantPointerNull>(SI->getOperand(0)))
@@ -327,13 +330,13 @@ bool GlobalsModRef::AnalyzeIndirectGlobalMemory(GlobalValue *GV) {
                                        GV->getParent()->getDataLayout());
 
       if (!isAllocLikeFn(Ptr, TLI))
-        return false;  // Too hard to analyze.
+        return false; // Too hard to analyze.
 
       // Analyze all uses of the allocation.  If any of them are used in a
       // non-simple way (e.g. stored to another global) bail out.
-      std::vector<Function*> ReadersWriters;
+      std::vector<Function *> ReadersWriters;
       if (AnalyzeUsesOfPointer(Ptr, ReadersWriters, ReadersWriters, GV))
-        return false;  // Loaded pointer escapes.
+        return false; // Loaded pointer escapes.
 
       // Remember that this allocation is related to the indirect global.
       AllocRelatedValues.push_back(Ptr);
@@ -360,7 +363,7 @@ bool GlobalsModRef::AnalyzeIndirectGlobalMemory(GlobalValue *GV) {
 void GlobalsModRef::AnalyzeCallGraph(CallGraph &CG, Module &M) {
   // We do a bottom-up SCC traversal of the call graph.  In other words, we
   // visit all callees before callers (leaf-first).
-  for (scc_iterator<CallGraph*> I = scc_begin(&CG); !I.isAtEnd(); ++I) {
+  for (scc_iterator<CallGraph *> I = scc_begin(&CG); !I.isAtEnd(); ++I) {
     const std::vector<CallGraphNode *> &SCC = *I;
     assert(!SCC.empty() && "SCC with no functions?");
 
@@ -437,9 +440,10 @@ void GlobalsModRef::AnalyzeCallGraph(CallGraph &CG, Module &M) {
     }
 
     // Scan the function bodies for explicit loads or stores.
-    for (unsigned i = 0, e = SCC.size(); i != e && FunctionEffect != ModRef;++i)
+    for (unsigned i = 0, e = SCC.size(); i != e && FunctionEffect != ModRef;
+         ++i)
       for (inst_iterator II = inst_begin(SCC[i]->getFunction()),
-             E = inst_end(SCC[i]->getFunction());
+                         E = inst_end(SCC[i]->getFunction());
            II != E && FunctionEffect != ModRef; ++II)
         if (LoadInst *LI = dyn_cast<LoadInst>(&*II)) {
           FunctionEffect |= Ref;
@@ -474,8 +478,6 @@ void GlobalsModRef::AnalyzeCallGraph(CallGraph &CG, Module &M) {
   }
 }
 
-
-
 /// alias - If one of the pointers is to a global that we are tracking, and the
 /// other is some random pointer, we know there cannot be an alias, because the
 /// address of the global isn't taken.
@@ -492,8 +494,10 @@ AliasResult GlobalsModRef::alias(const MemoryLocation &LocA,
   if (GV1 || GV2) {
     // If the global's address is taken, pretend we don't know it's a pointer to
     // the global.
-    if (GV1 && !NonAddressTakenGlobals.count(GV1)) GV1 = nullptr;
-    if (GV2 && !NonAddressTakenGlobals.count(GV2)) GV2 = nullptr;
+    if (GV1 && !NonAddressTakenGlobals.count(GV1))
+      GV1 = nullptr;
+    if (GV2 && !NonAddressTakenGlobals.count(GV2))
+      GV2 = nullptr;
 
     // If the two pointers are derived from two different non-addr-taken
     // globals, or if one is and the other isn't, we know these can't alias.
@@ -554,7 +558,6 @@ GlobalsModRef::getModRefInfo(ImmutableCallSite CS, const MemoryLocation &Loc) {
   return ModRefResult(Known & AliasAnalysis::getModRefInfo(CS, Loc));
 }
 
-
 //===----------------------------------------------------------------------===//
 // Methods to update the analysis as a result of the client transformation.
 //
@@ -565,9 +568,10 @@ void GlobalsModRef::deleteValue(Value *V) {
       // any AllocRelatedValues for it.
       if (IndirectGlobals.erase(GV)) {
         // Remove any entries in AllocsForIndirectGlobals for this global.
-        for (std::map<const Value*, const GlobalValue*>::iterator
-             I = AllocsForIndirectGlobals.begin(),
-             E = AllocsForIndirectGlobals.end(); I != E; ) {
+        for (std::map<const Value *, const GlobalValue *>::iterator
+                 I = AllocsForIndirectGlobals.begin(),
+                 E = AllocsForIndirectGlobals.end();
+             I != E;) {
           if (I->second == GV) {
             AllocsForIndirectGlobals.erase(I++);
           } else {
@@ -585,16 +589,12 @@ void GlobalsModRef::deleteValue(Value *V) {
   AliasAnalysis::deleteValue(V);
 }
 
-void GlobalsModRef::copyValue(Value *From, Value *To) {
-  AliasAnalysis::copyValue(From, To);
-}
-
 void GlobalsModRef::addEscapingUse(Use &U) {
   // For the purposes of this analysis, it is conservatively correct to treat
   // a newly escaping value equivalently to a deleted one.  We could perhaps
   // be more precise by processing the new use and attempting to update our
   // saved analysis results to accommodate it.
   deleteValue(U);
-  
+
   AliasAnalysis::addEscapingUse(U);
 }
diff --git a/lib/Analysis/IPA/InlineCost.cpp b/lib/Analysis/IPA/InlineCost.cpp
index 349b9cac2c2d..c0d2e375cb04 100644
--- a/lib/Analysis/IPA/InlineCost.cpp
+++ b/lib/Analysis/IPA/InlineCost.cpp
@@ -783,7 +783,7 @@ bool CallAnalyzer::visitCallSite(CallSite CS) {
       case Intrinsic::memmove:
         // SROA can usually chew through these intrinsics, but they aren't free.
         return false;
-      case Intrinsic::frameescape:
+      case Intrinsic::localescape:
         HasFrameEscape = true;
         return false;
       }
@@ -1424,11 +1424,11 @@ bool InlineCostAnalysis::isInlineViable(Function &F) {
           cast<CallInst>(CS.getInstruction())->canReturnTwice())
         return false;
 
-      // Disallow inlining functions that call @llvm.frameescape. Doing this
+      // Disallow inlining functions that call @llvm.localescape. Doing this
       // correctly would require major changes to the inliner.
       if (CS.getCalledFunction() &&
           CS.getCalledFunction()->getIntrinsicID() ==
-              llvm::Intrinsic::frameescape)
+              llvm::Intrinsic::localescape)
         return false;
     }
   }
diff --git a/lib/Analysis/IVUsers.cpp b/lib/Analysis/IVUsers.cpp
index b88b2496b875..926787d3be91 100644
--- a/lib/Analysis/IVUsers.cpp
+++ b/lib/Analysis/IVUsers.cpp
@@ -12,8 +12,10 @@
 //
 //===----------------------------------------------------------------------===//
 
-#include "llvm/Analysis/IVUsers.h"
 #include "llvm/ADT/STLExtras.h"
+#include "llvm/Analysis/AssumptionCache.h"
+#include "llvm/Analysis/CodeMetrics.h"
+#include "llvm/Analysis/IVUsers.h"
 #include "llvm/Analysis/LoopPass.h"
 #include "llvm/Analysis/ScalarEvolutionExpressions.h"
 #include "llvm/Analysis/ValueTracking.h"
@@ -34,6 +36,7 @@ using namespace llvm;
 char IVUsers::ID = 0;
 INITIALIZE_PASS_BEGIN(IVUsers, "iv-users",
                       "Induction Variable Users", false, true)
+INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker)
 INITIALIZE_PASS_DEPENDENCY(LoopInfoWrapperPass)
 INITIALIZE_PASS_DEPENDENCY(DominatorTreeWrapperPass)
 INITIALIZE_PASS_DEPENDENCY(ScalarEvolution)
@@ -137,6 +140,11 @@ bool IVUsers::AddUsersImpl(Instruction *I,
   if (Width > 64 || !DL.isLegalInteger(Width))
     return false;
 
+  // Don't attempt to promote ephemeral values to indvars. They will be removed
+  // later anyway.
+  if (EphValues.count(I))
+    return false;
+
   // Get the symbolic expression for this instruction.
   const SCEV *ISE = SE->getSCEV(I);
 
@@ -244,6 +252,7 @@ IVUsers::IVUsers()
 }
 
 void IVUsers::getAnalysisUsage(AnalysisUsage &AU) const {
+  AU.addRequired<AssumptionCacheTracker>();
   AU.addRequired<LoopInfoWrapperPass>();
   AU.addRequired<DominatorTreeWrapperPass>();
   AU.addRequired<ScalarEvolution>();
@@ -253,10 +262,16 @@ void IVUsers::getAnalysisUsage(AnalysisUsage &AU) const {
 bool IVUsers::runOnLoop(Loop *l, LPPassManager &LPM) {
 
   L = l;
+  AC = &getAnalysis<AssumptionCacheTracker>().getAssumptionCache(
+      *L->getHeader()->getParent());
   LI = &getAnalysis<LoopInfoWrapperPass>().getLoopInfo();
   DT = &getAnalysis<DominatorTreeWrapperPass>().getDomTree();
   SE = &getAnalysis<ScalarEvolution>();
 
+  // Collect ephemeral values so that AddUsersIfInteresting skips them.
+  EphValues.clear();
+  CodeMetrics::collectEphemeralValues(L, AC, EphValues);
+
   // Find all uses of induction variables in this loop, and categorize
   // them by stride.  Start by finding all of the PHI nodes in the header for
   // this loop.  If they are induction variables, inspect their uses.
diff --git a/lib/Analysis/InstructionSimplify.cpp b/lib/Analysis/InstructionSimplify.cpp
index 12e406bb1a2d..fa42b48b6cdb 100644
--- a/lib/Analysis/InstructionSimplify.cpp
+++ b/lib/Analysis/InstructionSimplify.cpp
@@ -24,6 +24,7 @@
 #include "llvm/Analysis/ConstantFolding.h"
 #include "llvm/Analysis/MemoryBuiltins.h"
 #include "llvm/Analysis/ValueTracking.h"
+#include "llvm/Analysis/VectorUtils.h"
 #include "llvm/IR/ConstantRange.h"
 #include "llvm/IR/DataLayout.h"
 #include "llvm/IR/Dominators.h"
@@ -3046,7 +3047,8 @@ Value *llvm::SimplifyICmpInst(unsigned Predicate, Value *LHS, Value *RHS,
 /// SimplifyFCmpInst - Given operands for an FCmpInst, see if we can
 /// fold the result.  If not, this returns null.
 static Value *SimplifyFCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
-                               const Query &Q, unsigned MaxRecurse) {
+                               FastMathFlags FMF, const Query &Q,
+                               unsigned MaxRecurse) {
   CmpInst::Predicate Pred = (CmpInst::Predicate)Predicate;
   assert(CmpInst::isFPPredicate(Pred) && "Not an FP compare!");
 
@@ -3065,6 +3067,14 @@ static Value *SimplifyFCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
   if (Pred == FCmpInst::FCMP_TRUE)
     return ConstantInt::get(GetCompareTy(LHS), 1);
 
+  // UNO/ORD predicates can be trivially folded if NaNs are ignored.
+  if (FMF.noNaNs()) {
+    if (Pred == FCmpInst::FCMP_UNO)
+      return ConstantInt::get(GetCompareTy(LHS), 0);
+    if (Pred == FCmpInst::FCMP_ORD)
+      return ConstantInt::get(GetCompareTy(LHS), 1);
+  }
+
   // fcmp pred x, undef  and  fcmp pred undef, x
   // fold to true if unordered, false if ordered
   if (isa<UndefValue>(LHS) || isa<UndefValue>(RHS)) {
@@ -3151,12 +3161,12 @@ static Value *SimplifyFCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
 }
 
 Value *llvm::SimplifyFCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
-                              const DataLayout &DL,
+                              FastMathFlags FMF, const DataLayout &DL,
                               const TargetLibraryInfo *TLI,
                               const DominatorTree *DT, AssumptionCache *AC,
                               const Instruction *CxtI) {
-  return ::SimplifyFCmpInst(Predicate, LHS, RHS, Query(DL, TLI, DT, AC, CxtI),
-                            RecursionLimit);
+  return ::SimplifyFCmpInst(Predicate, LHS, RHS, FMF,
+                            Query(DL, TLI, DT, AC, CxtI), RecursionLimit);
 }
 
 /// SimplifyWithOpReplaced - See if V simplifies when its operand Op is
@@ -3511,6 +3521,82 @@ Value *llvm::SimplifyInsertValueInst(
                                    RecursionLimit);
 }
 
+/// SimplifyExtractValueInst - Given operands for an ExtractValueInst, see if we
+/// can fold the result.  If not, this returns null.
+static Value *SimplifyExtractValueInst(Value *Agg, ArrayRef<unsigned> Idxs,
+                                       const Query &, unsigned) {
+  if (auto *CAgg = dyn_cast<Constant>(Agg))
+    return ConstantFoldExtractValueInstruction(CAgg, Idxs);
+
+  // extractvalue x, (insertvalue y, elt, n), n -> elt
+  unsigned NumIdxs = Idxs.size();
+  for (auto *IVI = dyn_cast<InsertValueInst>(Agg); IVI != nullptr;
+       IVI = dyn_cast<InsertValueInst>(IVI->getAggregateOperand())) {
+    ArrayRef<unsigned> InsertValueIdxs = IVI->getIndices();
+    unsigned NumInsertValueIdxs = InsertValueIdxs.size();
+    unsigned NumCommonIdxs = std::min(NumInsertValueIdxs, NumIdxs);
+    if (InsertValueIdxs.slice(0, NumCommonIdxs) ==
+        Idxs.slice(0, NumCommonIdxs)) {
+      if (NumIdxs == NumInsertValueIdxs)
+        return IVI->getInsertedValueOperand();
+      break;
+    }
+  }
+
+  return nullptr;
+}
+
+Value *llvm::SimplifyExtractValueInst(Value *Agg, ArrayRef<unsigned> Idxs,
+                                      const DataLayout &DL,
+                                      const TargetLibraryInfo *TLI,
+                                      const DominatorTree *DT,
+                                      AssumptionCache *AC,
+                                      const Instruction *CxtI) {
+  return ::SimplifyExtractValueInst(Agg, Idxs, Query(DL, TLI, DT, AC, CxtI),
+                                    RecursionLimit);
+}
+
+/// SimplifyExtractElementInst - Given operands for an ExtractElementInst, see if we
+/// can fold the result.  If not, this returns null.
+static Value *SimplifyExtractElementInst(Value *Vec, Value *Idx, const Query &,
+                                         unsigned) {
+  if (auto *CVec = dyn_cast<Constant>(Vec)) {
+    if (auto *CIdx = dyn_cast<Constant>(Idx))
+      return ConstantFoldExtractElementInstruction(CVec, CIdx);
+
+    // The index is not relevant if our vector is a splat.
+    if (auto *Splat = CVec->getSplatValue())
+      return Splat;
+
+    if (isa<UndefValue>(Vec))
+      return UndefValue::get(Vec->getType()->getVectorElementType());
+  }
+
+  // If extracting a specified index from the vector, see if we can recursively
+  // find a previously computed scalar that was inserted into the vector.
+  if (auto *IdxC = dyn_cast<ConstantInt>(Idx)) {
+    unsigned IndexVal = IdxC->getZExtValue();
+    unsigned VectorWidth = Vec->getType()->getVectorNumElements();
+
+    // If this is extracting an invalid index, turn this into undef, to avoid
+    // crashing the code below.
+    if (IndexVal >= VectorWidth)
+      return UndefValue::get(Vec->getType()->getVectorElementType());
+
+    if (Value *Elt = findScalarElement(Vec, IndexVal))
+      return Elt;
+  }
+
+  return nullptr;
+}
+
+Value *llvm::SimplifyExtractElementInst(
+    Value *Vec, Value *Idx, const DataLayout &DL, const TargetLibraryInfo *TLI,
+    const DominatorTree *DT, AssumptionCache *AC, const Instruction *CxtI) {
+  return ::SimplifyExtractElementInst(Vec, Idx, Query(DL, TLI, DT, AC, CxtI),
+                                      RecursionLimit);
+}
+
 /// SimplifyPHINode - See if we can fold the given phi.  If not, returns null.
 static Value *SimplifyPHINode(PHINode *PN, const Query &Q) {
   // If all of the PHI's incoming values are the same then replace the PHI node
@@ -3670,7 +3756,7 @@ static Value *SimplifyCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
                               const Query &Q, unsigned MaxRecurse) {
   if (CmpInst::isIntPredicate((CmpInst::Predicate)Predicate))
     return SimplifyICmpInst(Predicate, LHS, RHS, Q, MaxRecurse);
-  return SimplifyFCmpInst(Predicate, LHS, RHS, Q, MaxRecurse);
+  return SimplifyFCmpInst(Predicate, LHS, RHS, FastMathFlags(), Q, MaxRecurse);
 }
 
 Value *llvm::SimplifyCmpInst(unsigned Predicate, Value *LHS, Value *RHS,
@@ -3900,9 +3986,9 @@ Value *llvm::SimplifyInstruction(Instruction *I, const DataLayout &DL,
                          I->getOperand(1), DL, TLI, DT, AC, I);
     break;
   case Instruction::FCmp:
-    Result =
-        SimplifyFCmpInst(cast<FCmpInst>(I)->getPredicate(), I->getOperand(0),
-                         I->getOperand(1), DL, TLI, DT, AC, I);
+    Result = SimplifyFCmpInst(cast<FCmpInst>(I)->getPredicate(),
+                              I->getOperand(0), I->getOperand(1),
+                              I->getFastMathFlags(), DL, TLI, DT, AC, I);
     break;
   case Instruction::Select:
     Result = SimplifySelectInst(I->getOperand(0), I->getOperand(1),
@@ -3920,6 +4006,18 @@ Value *llvm::SimplifyInstruction(Instruction *I, const DataLayout &DL,
                                      IV->getIndices(), DL, TLI, DT, AC, I);
     break;
   }
+  case Instruction::ExtractValue: {
+    auto *EVI = cast<ExtractValueInst>(I);
+    Result = SimplifyExtractValueInst(EVI->getAggregateOperand(),
+                                      EVI->getIndices(), DL, TLI, DT, AC, I);
+    break;
+  }
+  case Instruction::ExtractElement: {
+    auto *EEI = cast<ExtractElementInst>(I);
+    Result = SimplifyExtractElementInst(
+        EEI->getVectorOperand(), EEI->getIndexOperand(), DL, TLI, DT, AC, I);
+    break;
+  }
   case Instruction::PHI:
     Result = SimplifyPHINode(cast<PHINode>(I), Query(DL, TLI, DT, AC, I));
     break;
diff --git a/lib/Analysis/LoopAccessAnalysis.cpp b/lib/Analysis/LoopAccessAnalysis.cpp
index b11cd7e84a6d..becbae4c5b50 100644
--- a/lib/Analysis/LoopAccessAnalysis.cpp
+++ b/lib/Analysis/LoopAccessAnalysis.cpp
@@ -48,6 +48,13 @@ static cl::opt<unsigned, true> RuntimeMemoryCheckThreshold(
     cl::location(VectorizerParams::RuntimeMemoryCheckThreshold), cl::init(8));
 unsigned VectorizerParams::RuntimeMemoryCheckThreshold;
 
+/// \brief The maximum iterations used to merge memory checks
+static cl::opt<unsigned> MemoryCheckMergeThreshold(
+    "memory-check-merge-threshold", cl::Hidden,
+    cl::desc("Maximum number of comparisons done when trying to merge "
+             "runtime memory checks. (default = 100)"),
+    cl::init(100));
+
 /// Maximum SIMD width.
 const unsigned VectorizerParams::MaxVectorWidth = 64;
 
@@ -112,35 +119,182 @@ const SCEV *llvm::replaceSymbolicStrideSCEV(ScalarEvolution *SE,
   return SE->getSCEV(Ptr);
 }
 
-void LoopAccessInfo::RuntimePointerCheck::insert(
-    ScalarEvolution *SE, Loop *Lp, Value *Ptr, bool WritePtr, unsigned DepSetId,
-    unsigned ASId, const ValueToValueMap &Strides) {
+void RuntimePointerChecking::insert(Loop *Lp, Value *Ptr, bool WritePtr,
+                                    unsigned DepSetId, unsigned ASId,
+                                    const ValueToValueMap &Strides) {
   // Get the stride replaced scev.
   const SCEV *Sc = replaceSymbolicStrideSCEV(SE, Strides, Ptr);
   const SCEVAddRecExpr *AR = dyn_cast<SCEVAddRecExpr>(Sc);
   assert(AR && "Invalid addrec expression");
   const SCEV *Ex = SE->getBackedgeTakenCount(Lp);
   const SCEV *ScEnd = AR->evaluateAtIteration(Ex, *SE);
-  Pointers.push_back(Ptr);
-  Starts.push_back(AR->getStart());
-  Ends.push_back(ScEnd);
-  IsWritePtr.push_back(WritePtr);
-  DependencySetId.push_back(DepSetId);
-  AliasSetId.push_back(ASId);
+  Pointers.emplace_back(Ptr, AR->getStart(), ScEnd, WritePtr, DepSetId, ASId,
+                        Sc);
+}
+
+bool RuntimePointerChecking::needsChecking(
+    const CheckingPtrGroup &M, const CheckingPtrGroup &N,
+    const SmallVectorImpl<int> *PtrPartition) const {
+  for (unsigned I = 0, EI = M.Members.size(); EI != I; ++I)
+    for (unsigned J = 0, EJ = N.Members.size(); EJ != J; ++J)
+      if (needsChecking(M.Members[I], N.Members[J], PtrPartition))
+        return true;
+  return false;
+}
+
+/// Compare \p I and \p J and return the minimum.
+/// Return nullptr in case we couldn't find an answer.
+static const SCEV *getMinFromExprs(const SCEV *I, const SCEV *J,
+                                   ScalarEvolution *SE) {
+  const SCEV *Diff = SE->getMinusSCEV(J, I);
+  const SCEVConstant *C = dyn_cast<const SCEVConstant>(Diff);
+
+  if (!C)
+    return nullptr;
+  if (C->getValue()->isNegative())
+    return J;
+  return I;
+}
+
+bool RuntimePointerChecking::CheckingPtrGroup::addPointer(unsigned Index) {
+  const SCEV *Start = RtCheck.Pointers[Index].Start;
+  const SCEV *End = RtCheck.Pointers[Index].End;
+
+  // Compare the starts and ends with the known minimum and maximum
+  // of this set. We need to know how we compare against the min/max
+  // of the set in order to be able to emit memchecks.
+  const SCEV *Min0 = getMinFromExprs(Start, Low, RtCheck.SE);
+  if (!Min0)
+    return false;
+
+  const SCEV *Min1 = getMinFromExprs(End, High, RtCheck.SE);
+  if (!Min1)
+    return false;
+
+  // Update the low bound  expression if we've found a new min value.
+  if (Min0 == Start)
+    Low = Start;
+
+  // Update the high bound expression if we've found a new max value.
+  if (Min1 != End)
+    High = End;
+
+  Members.push_back(Index);
+  return true;
 }
 
-bool LoopAccessInfo::RuntimePointerCheck::needsChecking(
+void RuntimePointerChecking::groupChecks(
+    MemoryDepChecker::DepCandidates &DepCands, bool UseDependencies) {
+  // We build the groups from dependency candidates equivalence classes
+  // because:
+  //    - We know that pointers in the same equivalence class share
+  //      the same underlying object and therefore there is a chance
+  //      that we can compare pointers
+  //    - We wouldn't be able to merge two pointers for which we need
+  //      to emit a memcheck. The classes in DepCands are already
+  //      conveniently built such that no two pointers in the same
+  //      class need checking against each other.
+
+  // We use the following (greedy) algorithm to construct the groups
+  // For every pointer in the equivalence class:
+  //   For each existing group:
+  //   - if the difference between this pointer and the min/max bounds
+  //     of the group is a constant, then make the pointer part of the
+  //     group and update the min/max bounds of that group as required.
+
+  CheckingGroups.clear();
+
+  // If we don't have the dependency partitions, construct a new
+  // checking pointer group for each pointer.
+  if (!UseDependencies) {
+    for (unsigned I = 0; I < Pointers.size(); ++I)
+      CheckingGroups.push_back(CheckingPtrGroup(I, *this));
+    return;
+  }
+
+  unsigned TotalComparisons = 0;
+
+  DenseMap<Value *, unsigned> PositionMap;
+  for (unsigned Index = 0; Index < Pointers.size(); ++Index)
+    PositionMap[Pointers[Index].PointerValue] = Index;
+
+  // We need to keep track of what pointers we've already seen so we
+  // don't process them twice.
+  SmallSet<unsigned, 2> Seen;
+
+  // Go through all equivalence classes, get the the "pointer check groups"
+  // and add them to the overall solution. We use the order in which accesses
+  // appear in 'Pointers' to enforce determinism.
+  for (unsigned I = 0; I < Pointers.size(); ++I) {
+    // We've seen this pointer before, and therefore already processed
+    // its equivalence class.
+    if (Seen.count(I))
+      continue;
+
+    MemoryDepChecker::MemAccessInfo Access(Pointers[I].PointerValue,
+                                           Pointers[I].IsWritePtr);
+
+    SmallVector<CheckingPtrGroup, 2> Groups;
+    auto LeaderI = DepCands.findValue(DepCands.getLeaderValue(Access));
+
+    // Because DepCands is constructed by visiting accesses in the order in
+    // which they appear in alias sets (which is deterministic) and the
+    // iteration order within an equivalence class member is only dependent on
+    // the order in which unions and insertions are performed on the
+    // equivalence class, the iteration order is deterministic.
+    for (auto MI = DepCands.member_begin(LeaderI), ME = DepCands.member_end();
+         MI != ME; ++MI) {
+      unsigned Pointer = PositionMap[MI->getPointer()];
+      bool Merged = false;
+      // Mark this pointer as seen.
+      Seen.insert(Pointer);
+
+      // Go through all the existing sets and see if we can find one
+      // which can include this pointer.
+      for (CheckingPtrGroup &Group : Groups) {
+        // Don't perform more than a certain amount of comparisons.
+        // This should limit the cost of grouping the pointers to something
+        // reasonable.  If we do end up hitting this threshold, the algorithm
+        // will create separate groups for all remaining pointers.
+        if (TotalComparisons > MemoryCheckMergeThreshold)
+          break;
+
+        TotalComparisons++;
+
+        if (Group.addPointer(Pointer)) {
+          Merged = true;
+          break;
+        }
+      }
+
+      if (!Merged)
+        // We couldn't add this pointer to any existing set or the threshold
+        // for the number of comparisons has been reached. Create a new group
+        // to hold the current pointer.
+        Groups.push_back(CheckingPtrGroup(Pointer, *this));
+    }
+
+    // We've computed the grouped checks for this partition.
+    // Save the results and continue with the next one.
+    std::copy(Groups.begin(), Groups.end(), std::back_inserter(CheckingGroups));
+  }
+}
+
+bool RuntimePointerChecking::needsChecking(
     unsigned I, unsigned J, const SmallVectorImpl<int> *PtrPartition) const {
+  const PointerInfo &PointerI = Pointers[I];
+  const PointerInfo &PointerJ = Pointers[J];
+
   // No need to check if two readonly pointers intersect.
-  if (!IsWritePtr[I] && !IsWritePtr[J])
+  if (!PointerI.IsWritePtr && !PointerJ.IsWritePtr)
     return false;
 
   // Only need to check pointers between two different dependency sets.
-  if (DependencySetId[I] == DependencySetId[J])
+  if (PointerI.DependencySetId == PointerJ.DependencySetId)
     return false;
 
   // Only need to check pointers in the same alias set.
-  if (AliasSetId[I] != AliasSetId[J])
+  if (PointerI.AliasSetId != PointerJ.AliasSetId)
     return false;
 
   // If PtrPartition is set omit checks between pointers of the same partition.
@@ -153,45 +307,75 @@ bool LoopAccessInfo::RuntimePointerCheck::needsChecking(
   return true;
 }
 
-void LoopAccessInfo::RuntimePointerCheck::print(
+void RuntimePointerChecking::print(
     raw_ostream &OS, unsigned Depth,
     const SmallVectorImpl<int> *PtrPartition) const {
-  unsigned NumPointers = Pointers.size();
-  if (NumPointers == 0)
-    return;
 
   OS.indent(Depth) << "Run-time memory checks:\n";
+
   unsigned N = 0;
-  for (unsigned I = 0; I < NumPointers; ++I)
-    for (unsigned J = I + 1; J < NumPointers; ++J)
-      if (needsChecking(I, J, PtrPartition)) {
-        OS.indent(Depth) << N++ << ":\n";
-        OS.indent(Depth + 2) << *Pointers[I];
-        if (PtrPartition)
-          OS << " (Partition: " << (*PtrPartition)[I] << ")";
-        OS << "\n";
-        OS.indent(Depth + 2) << *Pointers[J];
-        if (PtrPartition)
-          OS << " (Partition: " << (*PtrPartition)[J] << ")";
-        OS << "\n";
+  for (unsigned I = 0; I < CheckingGroups.size(); ++I)
+    for (unsigned J = I + 1; J < CheckingGroups.size(); ++J)
+      if (needsChecking(CheckingGroups[I], CheckingGroups[J], PtrPartition)) {
+        OS.indent(Depth) << "Check " << N++ << ":\n";
+        OS.indent(Depth + 2) << "Comparing group " << I << ":\n";
+
+        for (unsigned K = 0; K < CheckingGroups[I].Members.size(); ++K) {
+          OS.indent(Depth + 2)
+              << *Pointers[CheckingGroups[I].Members[K]].PointerValue << "\n";
+          if (PtrPartition)
+            OS << " (Partition: "
+               << (*PtrPartition)[CheckingGroups[I].Members[K]] << ")"
+               << "\n";
+        }
+
+        OS.indent(Depth + 2) << "Against group " << J << ":\n";
+
+        for (unsigned K = 0; K < CheckingGroups[J].Members.size(); ++K) {
+          OS.indent(Depth + 2)
+              << *Pointers[CheckingGroups[J].Members[K]].PointerValue << "\n";
+          if (PtrPartition)
+            OS << " (Partition: "
+               << (*PtrPartition)[CheckingGroups[J].Members[K]] << ")"
+               << "\n";
+        }
       }
+
+  OS.indent(Depth) << "Grouped accesses:\n";
+  for (unsigned I = 0; I < CheckingGroups.size(); ++I) {
+    OS.indent(Depth + 2) << "Group " << I << ":\n";
+    OS.indent(Depth + 4) << "(Low: " << *CheckingGroups[I].Low
+                         << " High: " << *CheckingGroups[I].High << ")\n";
+    for (unsigned J = 0; J < CheckingGroups[I].Members.size(); ++J) {
+      OS.indent(Depth + 6) << "Member: "
+                           << *Pointers[CheckingGroups[I].Members[J]].Expr
+                           << "\n";
+    }
+  }
 }
 
-unsigned LoopAccessInfo::RuntimePointerCheck::getNumberOfChecks(
+unsigned RuntimePointerChecking::getNumberOfChecks(
     const SmallVectorImpl<int> *PtrPartition) const {
-  unsigned NumPointers = Pointers.size();
+
+  unsigned NumPartitions = CheckingGroups.size();
   unsigned CheckCount = 0;
 
-  for (unsigned I = 0; I < NumPointers; ++I)
-    for (unsigned J = I + 1; J < NumPointers; ++J)
-      if (needsChecking(I, J, PtrPartition))
+  for (unsigned I = 0; I < NumPartitions; ++I)
+    for (unsigned J = I + 1; J < NumPartitions; ++J)
+      if (needsChecking(CheckingGroups[I], CheckingGroups[J], PtrPartition))
         CheckCount++;
   return CheckCount;
 }
 
-bool LoopAccessInfo::RuntimePointerCheck::needsAnyChecking(
+bool RuntimePointerChecking::needsAnyChecking(
     const SmallVectorImpl<int> *PtrPartition) const {
-  return getNumberOfChecks(PtrPartition) != 0;
+  unsigned NumPointers = Pointers.size();
+
+  for (unsigned I = 0; I < NumPointers; ++I)
+    for (unsigned J = I + 1; J < NumPointers; ++J)
+      if (needsChecking(I, J, PtrPartition))
+        return true;
+  return false;
 }
 
 namespace {
@@ -207,7 +391,8 @@ public:
 
   AccessAnalysis(const DataLayout &Dl, AliasAnalysis *AA, LoopInfo *LI,
                  MemoryDepChecker::DepCandidates &DA)
-      : DL(Dl), AST(*AA), LI(LI), DepCands(DA), IsRTCheckNeeded(false) {}
+      : DL(Dl), AST(*AA), LI(LI), DepCands(DA),
+        IsRTCheckAnalysisNeeded(false) {}
 
   /// \brief Register a load  and whether it is only read from.
   void addLoad(MemoryLocation &Loc, bool IsReadOnly) {
@@ -226,11 +411,12 @@ public:
   }
 
   /// \brief Check whether we can check the pointers at runtime for
-  /// non-intersection. Returns true when we have 0 pointers
-  /// (a check on 0 pointers for non-intersection will always return true).
-  bool canCheckPtrAtRT(LoopAccessInfo::RuntimePointerCheck &RtCheck,
-                       bool &NeedRTCheck, ScalarEvolution *SE, Loop *TheLoop,
-                       const ValueToValueMap &Strides,
+  /// non-intersection.
+  ///
+  /// Returns true if we need no check or if we do and we can generate them
+  /// (i.e. the pointers have computable bounds).
+  bool canCheckPtrAtRT(RuntimePointerChecking &RtCheck, ScalarEvolution *SE,
+                       Loop *TheLoop, const ValueToValueMap &Strides,
                        bool ShouldCheckStride = false);
 
   /// \brief Goes over all memory accesses, checks whether a RT check is needed
@@ -239,8 +425,11 @@ public:
     processMemAccesses();
   }
 
-  bool isRTCheckNeeded() { return IsRTCheckNeeded; }
-
+  /// \brief Initial processing of memory accesses determined that we need to
+  /// perform dependency checking.
+  ///
+  /// Note that this can later be cleared if we retry memcheck analysis without
+  /// dependency checking (i.e. ShouldRetryWithRuntimeCheck).
   bool isDependencyCheckNeeded() { return !CheckDeps.empty(); }
 
   /// We decided that no dependence analysis would be used.  Reset the state.
@@ -255,7 +444,7 @@ private:
   typedef SetVector<MemAccessInfo> PtrAccessSet;
 
   /// \brief Go over all memory access and check whether runtime pointer checks
-  /// are needed /// and build sets of dependency check candidates.
+  /// are needed and build sets of dependency check candidates.
   void processMemAccesses();
 
   /// Set of all accesses.
@@ -280,7 +469,14 @@ private:
   /// dependence check.
   MemoryDepChecker::DepCandidates &DepCands;
 
-  bool IsRTCheckNeeded;
+  /// \brief Initial processing of memory accesses determined that we may need
+  /// to add memchecks.  Perform the analysis to determine the necessary checks.
+  ///
+  /// Note that, this is different from isDependencyCheckNeeded.  When we retry
+  /// memcheck analysis without dependency checking
+  /// (i.e. ShouldRetryWithRuntimeCheck), isDependencyCheckNeeded is cleared
+  /// while this remains set if we have potentially dependent accesses.
+  bool IsRTCheckAnalysisNeeded;
 };
 
 } // end anonymous namespace
@@ -296,16 +492,16 @@ static bool hasComputableBounds(ScalarEvolution *SE,
   return AR->isAffine();
 }
 
-bool AccessAnalysis::canCheckPtrAtRT(
-    LoopAccessInfo::RuntimePointerCheck &RtCheck, bool &NeedRTCheck,
-    ScalarEvolution *SE, Loop *TheLoop, const ValueToValueMap &StridesMap,
-    bool ShouldCheckStride) {
+bool AccessAnalysis::canCheckPtrAtRT(RuntimePointerChecking &RtCheck,
+                                     ScalarEvolution *SE, Loop *TheLoop,
+                                     const ValueToValueMap &StridesMap,
+                                     bool ShouldCheckStride) {
   // Find pointers with computable bounds. We are going to use this information
   // to place a runtime bound check.
   bool CanDoRT = true;
 
-  NeedRTCheck = false;
-  if (!IsRTCheckNeeded) return true;
+  bool NeedRTCheck = false;
+  if (!IsRTCheckAnalysisNeeded) return true;
 
   bool IsDepCheckNeeded = isDependencyCheckNeeded();
 
@@ -313,6 +509,9 @@ bool AccessAnalysis::canCheckPtrAtRT(
   // Accesses between different groups doesn't need to be checked.
   unsigned ASId = 1;
   for (auto &AS : AST) {
+    int NumReadPtrChecks = 0;
+    int NumWritePtrChecks = 0;
+
     // We assign consecutive id to access from different dependence sets.
     // Accesses within the same set don't need a runtime check.
     unsigned RunningDepId = 1;
@@ -323,6 +522,11 @@ bool AccessAnalysis::canCheckPtrAtRT(
       bool IsWrite = Accesses.count(MemAccessInfo(Ptr, true));
       MemAccessInfo Access(Ptr, IsWrite);
 
+      if (IsWrite)
+        ++NumWritePtrChecks;
+      else
+        ++NumReadPtrChecks;
+
       if (hasComputableBounds(SE, StridesMap, Ptr) &&
           // When we run after a failing dependency check we have to make sure
           // we don't have wrapping pointers.
@@ -341,7 +545,7 @@ bool AccessAnalysis::canCheckPtrAtRT(
           // Each access has its own dependence set.
           DepId = RunningDepId++;
 
-        RtCheck.insert(SE, TheLoop, Ptr, IsWrite, DepId, ASId, StridesMap);
+        RtCheck.insert(TheLoop, Ptr, IsWrite, DepId, ASId, StridesMap);
 
         DEBUG(dbgs() << "LAA: Found a runtime check ptr:" << *Ptr << '\n');
       } else {
@@ -350,15 +554,21 @@ bool AccessAnalysis::canCheckPtrAtRT(
       }
     }
 
+    // If we have at least two writes or one write and a read then we need to
+    // check them.  But there is no need to checks if there is only one
+    // dependence set for this alias set.
+    //
+    // Note that this function computes CanDoRT and NeedRTCheck independently.
+    // For example CanDoRT=false, NeedRTCheck=false means that we have a pointer
+    // for which we couldn't find the bounds but we don't actually need to emit
+    // any checks so it does not matter.
+    if (!(IsDepCheckNeeded && CanDoRT && RunningDepId == 2))
+      NeedRTCheck |= (NumWritePtrChecks >= 2 || (NumReadPtrChecks >= 1 &&
+                                                 NumWritePtrChecks >= 1));
+
     ++ASId;
   }
 
-  // We need a runtime check if there are any accesses that need checking.
-  // However, some accesses cannot be checked (for example because we
-  // can't determine their bounds). In these cases we would need a check
-  // but wouldn't be able to add it.
-  NeedRTCheck = !CanDoRT || RtCheck.needsAnyChecking(nullptr);
-
   // If the pointers that we would use for the bounds comparison have different
   // address spaces, assume the values aren't directly comparable, so we can't
   // use them for the runtime check. We also have to assume they could
@@ -368,14 +578,15 @@ bool AccessAnalysis::canCheckPtrAtRT(
   for (unsigned i = 0; i < NumPointers; ++i) {
     for (unsigned j = i + 1; j < NumPointers; ++j) {
       // Only need to check pointers between two different dependency sets.
-      if (RtCheck.DependencySetId[i] == RtCheck.DependencySetId[j])
+      if (RtCheck.Pointers[i].DependencySetId ==
+          RtCheck.Pointers[j].DependencySetId)
        continue;
       // Only need to check pointers in the same alias set.
-      if (RtCheck.AliasSetId[i] != RtCheck.AliasSetId[j])
+      if (RtCheck.Pointers[i].AliasSetId != RtCheck.Pointers[j].AliasSetId)
         continue;
 
-      Value *PtrI = RtCheck.Pointers[i];
-      Value *PtrJ = RtCheck.Pointers[j];
+      Value *PtrI = RtCheck.Pointers[i].PointerValue;
+      Value *PtrJ = RtCheck.Pointers[j].PointerValue;
 
       unsigned ASi = PtrI->getType()->getPointerAddressSpace();
       unsigned ASj = PtrJ->getType()->getPointerAddressSpace();
@@ -387,7 +598,18 @@ bool AccessAnalysis::canCheckPtrAtRT(
     }
   }
 
-  return CanDoRT;
+  if (NeedRTCheck && CanDoRT)
+    RtCheck.groupChecks(DepCands, IsDepCheckNeeded);
+
+  DEBUG(dbgs() << "LAA: We need to do " << RtCheck.getNumberOfChecks(nullptr)
+               << " pointer comparisons.\n");
+
+  RtCheck.Need = NeedRTCheck;
+
+  bool CanDoRTIfNeeded = !NeedRTCheck || CanDoRT;
+  if (!CanDoRTIfNeeded)
+    RtCheck.reset();
+  return CanDoRTIfNeeded;
 }
 
 void AccessAnalysis::processMemAccesses() {
@@ -470,7 +692,7 @@ void AccessAnalysis::processMemAccesses() {
           // catch "a[i] = a[i] + " without having to do a dependence check).
           if ((IsWrite || IsReadOnlyPtr) && SetHasWrite) {
             CheckDeps.insert(Access);
-            IsRTCheckNeeded = true;
+            IsRTCheckAnalysisNeeded = true;
           }
 
           if (IsWrite)
@@ -600,7 +822,7 @@ int llvm::isStridedPtr(ScalarEvolution *SE, Value *Ptr, const Loop *Lp,
   // Check the step is constant.
   const SCEV *Step = AR->getStepRecurrence(*SE);
 
-  // Calculate the pointer stride and check if it is consecutive.
+  // Calculate the pointer stride and check if it is constant.
   const SCEVConstant *C = dyn_cast<SCEVConstant>(Step);
   if (!C) {
     DEBUG(dbgs() << "LAA: Bad stride - Not a constant strided " << *Ptr <<
@@ -805,11 +1027,11 @@ MemoryDepChecker::isDependent(const MemAccessInfo &A, unsigned AIdx,
   DEBUG(dbgs() << "LAA: Distance for " << *InstMap[AIdx] << " to "
         << *InstMap[BIdx] << ": " << *Dist << "\n");
 
-  // Need consecutive accesses. We don't want to vectorize
+  // Need accesses with constant stride. We don't want to vectorize
   // "A[B[i]] += ..." and similar code or pointer arithmetic that could wrap in
   // the address space.
   if (!StrideAPtr || !StrideBPtr || StrideAPtr != StrideBPtr){
-    DEBUG(dbgs() << "Non-consecutive pointer access\n");
+    DEBUG(dbgs() << "Pointer access with non-constant stride\n");
     return Dependence::Unknown;
   }
 
@@ -859,8 +1081,10 @@ MemoryDepChecker::isDependent(const MemAccessInfo &A, unsigned AIdx,
 
   unsigned Stride = std::abs(StrideAPtr);
   if (Stride > 1 &&
-      areStridedAccessesIndependent(Distance, Stride, TypeByteSize))
+      areStridedAccessesIndependent(Distance, Stride, TypeByteSize)) {
+    DEBUG(dbgs() << "LAA: Strided accesses are independent\n");
     return Dependence::NoDep;
+  }
 
   // Bail out early if passed-in parameters make vectorization not feasible.
   unsigned ForcedFactor = (VectorizerParams::VectorizationFactor ?
@@ -1098,8 +1322,8 @@ void LoopAccessInfo::analyzeLoop(const ValueToValueMap &Strides) {
   unsigned NumReads = 0;
   unsigned NumReadWrites = 0;
 
-  PtrRtCheck.Pointers.clear();
-  PtrRtCheck.Need = false;
+  PtrRtChecking.Pointers.clear();
+  PtrRtChecking.Need = false;
 
   const bool IsAnnotatedParallel = TheLoop->isAnnotatedParallel();
 
@@ -1258,28 +1482,17 @@ void LoopAccessInfo::analyzeLoop(const ValueToValueMap &Strides) {
 
   // Find pointers with computable bounds. We are going to use this information
   // to place a runtime bound check.
-  bool NeedRTCheck;
-  bool CanDoRT = Accesses.canCheckPtrAtRT(PtrRtCheck,
-                                          NeedRTCheck, SE,
-                                          TheLoop, Strides);
-
-  DEBUG(dbgs() << "LAA: We need to do "
-               << PtrRtCheck.getNumberOfChecks(nullptr)
-               << " pointer comparisons.\n");
-
-  // Check that we found the bounds for the pointer.
-  if (CanDoRT)
-    DEBUG(dbgs() << "LAA: We can perform a memory runtime check if needed.\n");
-  else if (NeedRTCheck) {
+  bool CanDoRTIfNeeded =
+      Accesses.canCheckPtrAtRT(PtrRtChecking, SE, TheLoop, Strides);
+  if (!CanDoRTIfNeeded) {
     emitAnalysis(LoopAccessReport() << "cannot identify array bounds");
-    DEBUG(dbgs() << "LAA: We can't vectorize because we can't find " <<
-          "the array bounds.\n");
-    PtrRtCheck.reset();
+    DEBUG(dbgs() << "LAA: We can't vectorize because we can't find "
+                 << "the array bounds.\n");
     CanVecMem = false;
     return;
   }
 
-  PtrRtCheck.Need = NeedRTCheck;
+  DEBUG(dbgs() << "LAA: We can perform a memory runtime check if needed.\n");
 
   CanVecMem = true;
   if (Accesses.isDependencyCheckNeeded()) {
@@ -1290,23 +1503,21 @@ void LoopAccessInfo::analyzeLoop(const ValueToValueMap &Strides) {
 
     if (!CanVecMem && DepChecker.shouldRetryWithRuntimeCheck()) {
       DEBUG(dbgs() << "LAA: Retrying with memory checks\n");
-      NeedRTCheck = true;
 
       // Clear the dependency checks. We assume they are not needed.
       Accesses.resetDepChecks(DepChecker);
 
-      PtrRtCheck.reset();
-      PtrRtCheck.Need = true;
+      PtrRtChecking.reset();
+      PtrRtChecking.Need = true;
 
-      CanDoRT = Accesses.canCheckPtrAtRT(PtrRtCheck, NeedRTCheck, SE,
-                                         TheLoop, Strides, true);
+      CanDoRTIfNeeded =
+          Accesses.canCheckPtrAtRT(PtrRtChecking, SE, TheLoop, Strides, true);
 
       // Check that we found the bounds for the pointer.
-      if (NeedRTCheck && !CanDoRT) {
+      if (!CanDoRTIfNeeded) {
         emitAnalysis(LoopAccessReport()
                      << "cannot check memory dependencies at runtime");
         DEBUG(dbgs() << "LAA: Can't vectorize with memory checks\n");
-        PtrRtCheck.reset();
         CanVecMem = false;
         return;
       }
@@ -1317,8 +1528,8 @@ void LoopAccessInfo::analyzeLoop(const ValueToValueMap &Strides) {
 
   if (CanVecMem)
     DEBUG(dbgs() << "LAA: No unsafe dependent memory operations in loop.  We"
-                 << (NeedRTCheck ? "" : " don't")
-                 << " need a runtime memory check.\n");
+                 << (PtrRtChecking.Need ? "" : " don't")
+                 << " need runtime memory checks.\n");
   else {
     emitAnalysis(LoopAccessReport() <<
                  "unsafe dependent memory operations in loop");
@@ -1357,35 +1568,38 @@ static Instruction *getFirstInst(Instruction *FirstInst, Value *V,
 
 std::pair<Instruction *, Instruction *> LoopAccessInfo::addRuntimeCheck(
     Instruction *Loc, const SmallVectorImpl<int> *PtrPartition) const {
-  if (!PtrRtCheck.Need)
+  if (!PtrRtChecking.Need)
     return std::make_pair(nullptr, nullptr);
 
-  unsigned NumPointers = PtrRtCheck.Pointers.size();
-  SmallVector<TrackingVH<Value> , 2> Starts;
-  SmallVector<TrackingVH<Value> , 2> Ends;
+  SmallVector<TrackingVH<Value>, 2> Starts;
+  SmallVector<TrackingVH<Value>, 2> Ends;
 
   LLVMContext &Ctx = Loc->getContext();
   SCEVExpander Exp(*SE, DL, "induction");
   Instruction *FirstInst = nullptr;
 
-  for (unsigned i = 0; i < NumPointers; ++i) {
-    Value *Ptr = PtrRtCheck.Pointers[i];
+  for (unsigned i = 0; i < PtrRtChecking.CheckingGroups.size(); ++i) {
+    const RuntimePointerChecking::CheckingPtrGroup &CG =
+        PtrRtChecking.CheckingGroups[i];
+    Value *Ptr = PtrRtChecking.Pointers[CG.Members[0]].PointerValue;
     const SCEV *Sc = SE->getSCEV(Ptr);
 
     if (SE->isLoopInvariant(Sc, TheLoop)) {
-      DEBUG(dbgs() << "LAA: Adding RT check for a loop invariant ptr:" <<
-            *Ptr <<"\n");
+      DEBUG(dbgs() << "LAA: Adding RT check for a loop invariant ptr:" << *Ptr
+                   << "\n");
       Starts.push_back(Ptr);
       Ends.push_back(Ptr);
     } else {
-      DEBUG(dbgs() << "LAA: Adding RT check for range:" << *Ptr << '\n');
       unsigned AS = Ptr->getType()->getPointerAddressSpace();
 
       // Use this type for pointer arithmetic.
       Type *PtrArithTy = Type::getInt8PtrTy(Ctx, AS);
+      Value *Start = nullptr, *End = nullptr;
 
-      Value *Start = Exp.expandCodeFor(PtrRtCheck.Starts[i], PtrArithTy, Loc);
-      Value *End = Exp.expandCodeFor(PtrRtCheck.Ends[i], PtrArithTy, Loc);
+      DEBUG(dbgs() << "LAA: Adding RT check for range:\n");
+      Start = Exp.expandCodeFor(CG.Low, PtrArithTy, Loc);
+      End = Exp.expandCodeFor(CG.High, PtrArithTy, Loc);
+      DEBUG(dbgs() << "Start: " << *CG.Low << " End: " << *CG.High << "\n");
       Starts.push_back(Start);
       Ends.push_back(End);
     }
@@ -1394,9 +1608,14 @@ std::pair<Instruction *, Instruction *> LoopAccessInfo::addRuntimeCheck(
   IRBuilder<> ChkBuilder(Loc);
   // Our instructions might fold to a constant.
   Value *MemoryRuntimeCheck = nullptr;
-  for (unsigned i = 0; i < NumPointers; ++i) {
-    for (unsigned j = i+1; j < NumPointers; ++j) {
-      if (!PtrRtCheck.needsChecking(i, j, PtrPartition))
+  for (unsigned i = 0; i < PtrRtChecking.CheckingGroups.size(); ++i) {
+    for (unsigned j = i + 1; j < PtrRtChecking.CheckingGroups.size(); ++j) {
+      const RuntimePointerChecking::CheckingPtrGroup &CGI =
+          PtrRtChecking.CheckingGroups[i];
+      const RuntimePointerChecking::CheckingPtrGroup &CGJ =
+          PtrRtChecking.CheckingGroups[j];
+
+      if (!PtrRtChecking.needsChecking(CGI, CGJ, PtrPartition))
         continue;
 
       unsigned AS0 = Starts[i]->getType()->getPointerAddressSpace();
@@ -1447,7 +1666,7 @@ LoopAccessInfo::LoopAccessInfo(Loop *L, ScalarEvolution *SE,
                                const TargetLibraryInfo *TLI, AliasAnalysis *AA,
                                DominatorTree *DT, LoopInfo *LI,
                                const ValueToValueMap &Strides)
-    : DepChecker(SE, L), TheLoop(L), SE(SE), DL(DL),
+    : PtrRtChecking(SE), DepChecker(SE, L), TheLoop(L), SE(SE), DL(DL),
       TLI(TLI), AA(AA), DT(DT), LI(LI), NumLoads(0), NumStores(0),
       MaxSafeDepDistBytes(-1U), CanVecMem(false),
       StoreToLoopInvariantAddress(false) {
@@ -1457,7 +1676,7 @@ LoopAccessInfo::LoopAccessInfo(Loop *L, ScalarEvolution *SE,
 
 void LoopAccessInfo::print(raw_ostream &OS, unsigned Depth) const {
   if (CanVecMem) {
-    if (PtrRtCheck.Need)
+    if (PtrRtChecking.Need)
       OS.indent(Depth) << "Memory dependences are safe with run-time checks\n";
     else
       OS.indent(Depth) << "Memory dependences are safe\n";
@@ -1476,7 +1695,7 @@ void LoopAccessInfo::print(raw_ostream &OS, unsigned Depth) const {
     OS.indent(Depth) << "Too many interesting dependences, not recorded\n";
 
   // List the pair of accesses need run-time checks to prove independence.
-  PtrRtCheck.print(OS, Depth);
+  PtrRtChecking.print(OS, Depth);
   OS << "\n";
 
   OS.indent(Depth) << "Store to invariant address was "
diff --git a/lib/Analysis/NoAliasAnalysis.cpp b/lib/Analysis/NoAliasAnalysis.cpp
index 7617622b9ab6..322a9a80de4c 100644
--- a/lib/Analysis/NoAliasAnalysis.cpp
+++ b/lib/Analysis/NoAliasAnalysis.cpp
@@ -72,7 +72,6 @@ namespace {
     }
 
     void deleteValue(Value *V) override {}
-    void copyValue(Value *From, Value *To) override {}
     void addEscapingUse(Use &U) override {}
 
     /// getAdjustedAnalysisPointer - This method is used when a pass implements
diff --git a/lib/Analysis/TargetTransformInfo.cpp b/lib/Analysis/TargetTransformInfo.cpp
index 520d1e5ef87d..7d1c3fbef68a 100644
--- a/lib/Analysis/TargetTransformInfo.cpp
+++ b/lib/Analysis/TargetTransformInfo.cpp
@@ -28,12 +28,12 @@ namespace {
 ///
 /// This is used when no target specific information is available.
 struct NoTTIImpl : TargetTransformInfoImplCRTPBase<NoTTIImpl> {
-  explicit NoTTIImpl(const DataLayout *DL)
+  explicit NoTTIImpl(const DataLayout &DL)
       : TargetTransformInfoImplCRTPBase<NoTTIImpl>(DL) {}
 };
 }
 
-TargetTransformInfo::TargetTransformInfo(const DataLayout *DL)
+TargetTransformInfo::TargetTransformInfo(const DataLayout &DL)
     : TTIImpl(new Model<NoTTIImpl>(NoTTIImpl(DL))) {}
 
 TargetTransformInfo::~TargetTransformInfo() {}
@@ -304,7 +304,7 @@ TargetIRAnalysis::Result TargetIRAnalysis::run(Function &F) {
 char TargetIRAnalysis::PassID;
 
 TargetIRAnalysis::Result TargetIRAnalysis::getDefaultTTI(Function &F) {
-  return Result(&F.getParent()->getDataLayout());
+  return Result(F.getParent()->getDataLayout());
 }
 
 // Register the basic pass.
diff --git a/lib/Analysis/ValueTracking.cpp b/lib/Analysis/ValueTracking.cpp
index c45005f343d3..fa0d7798cae9 100644
--- a/lib/Analysis/ValueTracking.cpp
+++ b/lib/Analysis/ValueTracking.cpp
@@ -1464,7 +1464,7 @@ void computeKnownBits(Value *V, APInt &KnownZero, APInt &KnownOne,
           // If the object is defined in the current Module, we'll be giving
           // it the preferred alignment. Otherwise, we have to assume that it
           // may only have the minimum ABI alignment.
-          if (!GVar->isDeclaration() && !GVar->isWeakForLinker())
+          if (GVar->isStrongDefinitionForLinker())
             Align = DL.getPreferredAlignment(GVar);
           else
             Align = DL.getABITypeAlignment(ObjectType);
diff --git a/lib/Analysis/VectorUtils.cpp b/lib/Analysis/VectorUtils.cpp
index 96fddd103cc5..67f68dc8391e 100644
--- a/lib/Analysis/VectorUtils.cpp
+++ b/lib/Analysis/VectorUtils.cpp
@@ -11,7 +11,13 @@
 //
 //===----------------------------------------------------------------------===//
 
+#include "llvm/Analysis/LoopInfo.h"
+#include "llvm/Analysis/ScalarEvolutionExpressions.h"
+#include "llvm/Analysis/ScalarEvolution.h"
 #include "llvm/Analysis/VectorUtils.h"
+#include "llvm/IR/GetElementPtrTypeIterator.h"
+#include "llvm/IR/PatternMatch.h"
+#include "llvm/IR/Value.h"
 
 /// \brief Identify if the intrinsic is trivially vectorizable.
 /// This method returns true if the intrinsic's argument types are all
@@ -211,3 +217,195 @@ llvm::Intrinsic::ID llvm::getIntrinsicIDForCall(CallInst *CI,
 
   return Intrinsic::not_intrinsic;
 }
+
+/// \brief Find the operand of the GEP that should be checked for consecutive
+/// stores. This ignores trailing indices that have no effect on the final
+/// pointer.
+unsigned llvm::getGEPInductionOperand(const GetElementPtrInst *Gep) {
+  const DataLayout &DL = Gep->getModule()->getDataLayout();
+  unsigned LastOperand = Gep->getNumOperands() - 1;
+  unsigned GEPAllocSize = DL.getTypeAllocSize(
+      cast<PointerType>(Gep->getType()->getScalarType())->getElementType());
+
+  // Walk backwards and try to peel off zeros.
+  while (LastOperand > 1 &&
+         match(Gep->getOperand(LastOperand), llvm::PatternMatch::m_Zero())) {
+    // Find the type we're currently indexing into.
+    gep_type_iterator GEPTI = gep_type_begin(Gep);
+    std::advance(GEPTI, LastOperand - 1);
+
+    // If it's a type with the same allocation size as the result of the GEP we
+    // can peel off the zero index.
+    if (DL.getTypeAllocSize(*GEPTI) != GEPAllocSize)
+      break;
+    --LastOperand;
+  }
+
+  return LastOperand;
+}
+
+/// \brief If the argument is a GEP, then returns the operand identified by
+/// getGEPInductionOperand. However, if there is some other non-loop-invariant
+/// operand, it returns that instead.
+llvm::Value *llvm::stripGetElementPtr(llvm::Value *Ptr, ScalarEvolution *SE,
+                                      Loop *Lp) {
+  GetElementPtrInst *GEP = dyn_cast<GetElementPtrInst>(Ptr);
+  if (!GEP)
+    return Ptr;
+
+  unsigned InductionOperand = getGEPInductionOperand(GEP);
+
+  // Check that all of the gep indices are uniform except for our induction
+  // operand.
+  for (unsigned i = 0, e = GEP->getNumOperands(); i != e; ++i)
+    if (i != InductionOperand &&
+        !SE->isLoopInvariant(SE->getSCEV(GEP->getOperand(i)), Lp))
+      return Ptr;
+  return GEP->getOperand(InductionOperand);
+}
+
+/// \brief If a value has only one user that is a CastInst, return it.
+llvm::Value *llvm::getUniqueCastUse(llvm::Value *Ptr, Loop *Lp, Type *Ty) {
+  llvm::Value *UniqueCast = nullptr;
+  for (User *U : Ptr->users()) {
+    CastInst *CI = dyn_cast<CastInst>(U);
+    if (CI && CI->getType() == Ty) {
+      if (!UniqueCast)
+        UniqueCast = CI;
+      else
+        return nullptr;
+    }
+  }
+  return UniqueCast;
+}
+
+/// \brief Get the stride of a pointer access in a loop. Looks for symbolic
+/// strides "a[i*stride]". Returns the symbolic stride, or null otherwise.
+llvm::Value *llvm::getStrideFromPointer(llvm::Value *Ptr, ScalarEvolution *SE,
+                                        Loop *Lp) {
+  const PointerType *PtrTy = dyn_cast<PointerType>(Ptr->getType());
+  if (!PtrTy || PtrTy->isAggregateType())
+    return nullptr;
+
+  // Try to remove a gep instruction to make the pointer (actually index at this
+  // point) easier analyzable. If OrigPtr is equal to Ptr we are analzying the
+  // pointer, otherwise, we are analyzing the index.
+  llvm::Value *OrigPtr = Ptr;
+
+  // The size of the pointer access.
+  int64_t PtrAccessSize = 1;
+
+  Ptr = stripGetElementPtr(Ptr, SE, Lp);
+  const SCEV *V = SE->getSCEV(Ptr);
+
+  if (Ptr != OrigPtr)
+    // Strip off casts.
+    while (const SCEVCastExpr *C = dyn_cast<SCEVCastExpr>(V))
+      V = C->getOperand();
+
+  const SCEVAddRecExpr *S = dyn_cast<SCEVAddRecExpr>(V);
+  if (!S)
+    return nullptr;
+
+  V = S->getStepRecurrence(*SE);
+  if (!V)
+    return nullptr;
+
+  // Strip off the size of access multiplication if we are still analyzing the
+  // pointer.
+  if (OrigPtr == Ptr) {
+    const DataLayout &DL = Lp->getHeader()->getModule()->getDataLayout();
+    DL.getTypeAllocSize(PtrTy->getElementType());
+    if (const SCEVMulExpr *M = dyn_cast<SCEVMulExpr>(V)) {
+      if (M->getOperand(0)->getSCEVType() != scConstant)
+        return nullptr;
+
+      const APInt &APStepVal =
+          cast<SCEVConstant>(M->getOperand(0))->getValue()->getValue();
+
+      // Huge step value - give up.
+      if (APStepVal.getBitWidth() > 64)
+        return nullptr;
+
+      int64_t StepVal = APStepVal.getSExtValue();
+      if (PtrAccessSize != StepVal)
+        return nullptr;
+      V = M->getOperand(1);
+    }
+  }
+
+  // Strip off casts.
+  Type *StripedOffRecurrenceCast = nullptr;
+  if (const SCEVCastExpr *C = dyn_cast<SCEVCastExpr>(V)) {
+    StripedOffRecurrenceCast = C->getType();
+    V = C->getOperand();
+  }
+
+  // Look for the loop invariant symbolic value.
+  const SCEVUnknown *U = dyn_cast<SCEVUnknown>(V);
+  if (!U)
+    return nullptr;
+
+  llvm::Value *Stride = U->getValue();
+  if (!Lp->isLoopInvariant(Stride))
+    return nullptr;
+
+  // If we have stripped off the recurrence cast we have to make sure that we
+  // return the value that is used in this loop so that we can replace it later.
+  if (StripedOffRecurrenceCast)
+    Stride = getUniqueCastUse(Stride, Lp, StripedOffRecurrenceCast);
+
+  return Stride;
+}
+
+/// \brief Given a vector and an element number, see if the scalar value is
+/// already around as a register, for example if it were inserted then extracted
+/// from the vector.
+llvm::Value *llvm::findScalarElement(llvm::Value *V, unsigned EltNo) {
+  assert(V->getType()->isVectorTy() && "Not looking at a vector?");
+  VectorType *VTy = cast<VectorType>(V->getType());
+  unsigned Width = VTy->getNumElements();
+  if (EltNo >= Width)  // Out of range access.
+    return UndefValue::get(VTy->getElementType());
+
+  if (Constant *C = dyn_cast<Constant>(V))
+    return C->getAggregateElement(EltNo);
+
+  if (InsertElementInst *III = dyn_cast<InsertElementInst>(V)) {
+    // If this is an insert to a variable element, we don't know what it is.
+    if (!isa<ConstantInt>(III->getOperand(2)))
+      return nullptr;
+    unsigned IIElt = cast<ConstantInt>(III->getOperand(2))->getZExtValue();
+
+    // If this is an insert to the element we are looking for, return the
+    // inserted value.
+    if (EltNo == IIElt)
+      return III->getOperand(1);
+
+    // Otherwise, the insertelement doesn't modify the value, recurse on its
+    // vector input.
+    return findScalarElement(III->getOperand(0), EltNo);
+  }
+
+  if (ShuffleVectorInst *SVI = dyn_cast<ShuffleVectorInst>(V)) {
+    unsigned LHSWidth = SVI->getOperand(0)->getType()->getVectorNumElements();
+    int InEl = SVI->getMaskValue(EltNo);
+    if (InEl < 0)
+      return UndefValue::get(VTy->getElementType());
+    if (InEl < (int)LHSWidth)
+      return findScalarElement(SVI->getOperand(0), InEl);
+    return findScalarElement(SVI->getOperand(1), InEl - LHSWidth);
+  }
+
+  // Extract a value from a vector add operation with a constant zero.
+  Value *Val = nullptr; Constant *Con = nullptr;
+  if (match(V,
+            llvm::PatternMatch::m_Add(llvm::PatternMatch::m_Value(Val),
+                                      llvm::PatternMatch::m_Constant(Con)))) {
+    if (Con->getAggregateElement(EltNo)->isNullValue())
+      return findScalarElement(Val, EltNo);
+  }
+
+  // Otherwise, we don't know.
+  return nullptr;
+}
diff --git a/lib/AsmParser/LLLexer.cpp b/lib/AsmParser/LLLexer.cpp
index 88f359d4fd5c..5c4bab734b2b 100644
--- a/lib/AsmParser/LLLexer.cpp
+++ b/lib/AsmParser/LLLexer.cpp
@@ -593,6 +593,7 @@ lltok::Kind LLLexer::LexIdentifier() {
   KEYWORD(attributes);
 
   KEYWORD(alwaysinline);
+  KEYWORD(argmemonly);
   KEYWORD(builtin);
   KEYWORD(byval);
   KEYWORD(inalloca);
diff --git a/lib/AsmParser/LLParser.cpp b/lib/AsmParser/LLParser.cpp
index b3c7fa087d40..1c6e7bd18d0e 100644
--- a/lib/AsmParser/LLParser.cpp
+++ b/lib/AsmParser/LLParser.cpp
@@ -946,35 +946,42 @@ bool LLParser::ParseFnAttributeValuePairs(AttrBuilder &B,
       B.addStackAlignmentAttr(Alignment);
       continue;
     }
-    case lltok::kw_alwaysinline:      B.addAttribute(Attribute::AlwaysInline); break;
-    case lltok::kw_builtin:           B.addAttribute(Attribute::Builtin); break;
-    case lltok::kw_cold:              B.addAttribute(Attribute::Cold); break;
-    case lltok::kw_convergent:        B.addAttribute(Attribute::Convergent); break;
-    case lltok::kw_inlinehint:        B.addAttribute(Attribute::InlineHint); break;
-    case lltok::kw_jumptable:         B.addAttribute(Attribute::JumpTable); break;
-    case lltok::kw_minsize:           B.addAttribute(Attribute::MinSize); break;
-    case lltok::kw_naked:             B.addAttribute(Attribute::Naked); break;
-    case lltok::kw_nobuiltin:         B.addAttribute(Attribute::NoBuiltin); break;
-    case lltok::kw_noduplicate:       B.addAttribute(Attribute::NoDuplicate); break;
-    case lltok::kw_noimplicitfloat:   B.addAttribute(Attribute::NoImplicitFloat); break;
-    case lltok::kw_noinline:          B.addAttribute(Attribute::NoInline); break;
-    case lltok::kw_nonlazybind:       B.addAttribute(Attribute::NonLazyBind); break;
-    case lltok::kw_noredzone:         B.addAttribute(Attribute::NoRedZone); break;
-    case lltok::kw_noreturn:          B.addAttribute(Attribute::NoReturn); break;
-    case lltok::kw_nounwind:          B.addAttribute(Attribute::NoUnwind); break;
-    case lltok::kw_optnone:           B.addAttribute(Attribute::OptimizeNone); break;
-    case lltok::kw_optsize:           B.addAttribute(Attribute::OptimizeForSize); break;
-    case lltok::kw_readnone:          B.addAttribute(Attribute::ReadNone); break;
-    case lltok::kw_readonly:          B.addAttribute(Attribute::ReadOnly); break;
-    case lltok::kw_returns_twice:     B.addAttribute(Attribute::ReturnsTwice); break;
-    case lltok::kw_ssp:               B.addAttribute(Attribute::StackProtect); break;
-    case lltok::kw_sspreq:            B.addAttribute(Attribute::StackProtectReq); break;
-    case lltok::kw_sspstrong:         B.addAttribute(Attribute::StackProtectStrong); break;
-    case lltok::kw_safestack:         B.addAttribute(Attribute::SafeStack); break;
-    case lltok::kw_sanitize_address:  B.addAttribute(Attribute::SanitizeAddress); break;
-    case lltok::kw_sanitize_thread:   B.addAttribute(Attribute::SanitizeThread); break;
-    case lltok::kw_sanitize_memory:   B.addAttribute(Attribute::SanitizeMemory); break;
-    case lltok::kw_uwtable:           B.addAttribute(Attribute::UWTable); break;
+    case lltok::kw_alwaysinline: B.addAttribute(Attribute::AlwaysInline); break;
+    case lltok::kw_argmemonly: B.addAttribute(Attribute::ArgMemOnly); break;
+    case lltok::kw_builtin: B.addAttribute(Attribute::Builtin); break;
+    case lltok::kw_cold: B.addAttribute(Attribute::Cold); break;
+    case lltok::kw_convergent: B.addAttribute(Attribute::Convergent); break;
+    case lltok::kw_inlinehint: B.addAttribute(Attribute::InlineHint); break;
+    case lltok::kw_jumptable: B.addAttribute(Attribute::JumpTable); break;
+    case lltok::kw_minsize: B.addAttribute(Attribute::MinSize); break;
+    case lltok::kw_naked: B.addAttribute(Attribute::Naked); break;
+    case lltok::kw_nobuiltin: B.addAttribute(Attribute::NoBuiltin); break;
+    case lltok::kw_noduplicate: B.addAttribute(Attribute::NoDuplicate); break;
+    case lltok::kw_noimplicitfloat:
+      B.addAttribute(Attribute::NoImplicitFloat); break;
+    case lltok::kw_noinline: B.addAttribute(Attribute::NoInline); break;
+    case lltok::kw_nonlazybind: B.addAttribute(Attribute::NonLazyBind); break;
+    case lltok::kw_noredzone: B.addAttribute(Attribute::NoRedZone); break;
+    case lltok::kw_noreturn: B.addAttribute(Attribute::NoReturn); break;
+    case lltok::kw_nounwind: B.addAttribute(Attribute::NoUnwind); break;
+    case lltok::kw_optnone: B.addAttribute(Attribute::OptimizeNone); break;
+    case lltok::kw_optsize: B.addAttribute(Attribute::OptimizeForSize); break;
+    case lltok::kw_readnone: B.addAttribute(Attribute::ReadNone); break;
+    case lltok::kw_readonly: B.addAttribute(Attribute::ReadOnly); break;
+    case lltok::kw_returns_twice:
+      B.addAttribute(Attribute::ReturnsTwice); break;
+    case lltok::kw_ssp: B.addAttribute(Attribute::StackProtect); break;
+    case lltok::kw_sspreq: B.addAttribute(Attribute::StackProtectReq); break;
+    case lltok::kw_sspstrong:
+      B.addAttribute(Attribute::StackProtectStrong); break;
+    case lltok::kw_safestack: B.addAttribute(Attribute::SafeStack); break;
+    case lltok::kw_sanitize_address:
+      B.addAttribute(Attribute::SanitizeAddress); break;
+    case lltok::kw_sanitize_thread:
+      B.addAttribute(Attribute::SanitizeThread); break;
+    case lltok::kw_sanitize_memory:
+      B.addAttribute(Attribute::SanitizeMemory); break;
+    case lltok::kw_uwtable: B.addAttribute(Attribute::UWTable); break;
 
     // Error handling.
     case lltok::kw_inreg:
@@ -1258,6 +1265,7 @@ bool LLParser::ParseOptionalParamAttrs(AttrBuilder &B) {
 
     case lltok::kw_alignstack:
     case lltok::kw_alwaysinline:
+    case lltok::kw_argmemonly:
     case lltok::kw_builtin:
     case lltok::kw_inlinehint:
     case lltok::kw_jumptable:
@@ -1334,6 +1342,7 @@ bool LLParser::ParseOptionalReturnAttrs(AttrBuilder &B) {
 
     case lltok::kw_alignstack:
     case lltok::kw_alwaysinline:
+    case lltok::kw_argmemonly:
     case lltok::kw_builtin:
     case lltok::kw_cold:
     case lltok::kw_inlinehint:
@@ -2873,8 +2882,8 @@ bool LLParser::ParseValID(ValID &ID, PerFunctionState *PFS) {
         if (ValTy->isVectorTy() != BaseType->isVectorTy())
           return Error(ID.Loc, "getelementptr index type missmatch");
         if (ValTy->isVectorTy()) {
-          unsigned ValNumEl = cast<VectorType>(ValTy)->getNumElements();
-          unsigned PtrNumEl = cast<VectorType>(BaseType)->getNumElements();
+          unsigned ValNumEl = ValTy->getVectorNumElements();
+          unsigned PtrNumEl = BaseType->getVectorNumElements();
           if (ValNumEl != PtrNumEl)
             return Error(
                 ID.Loc,
@@ -4534,8 +4543,17 @@ int LLParser::ParseInstruction(Instruction *&Inst, BasicBlock *BB,
   case lltok::kw_and:
   case lltok::kw_or:
   case lltok::kw_xor:    return ParseLogical(Inst, PFS, KeywordVal);
-  case lltok::kw_icmp:
-  case lltok::kw_fcmp:   return ParseCompare(Inst, PFS, KeywordVal);
+  case lltok::kw_icmp:   return ParseCompare(Inst, PFS, KeywordVal);
+  case lltok::kw_fcmp: {
+    FastMathFlags FMF = EatFastMathFlagsIfPresent();
+    int Res = ParseCompare(Inst, PFS, KeywordVal);
+    if (Res != 0)
+      return Res;
+    if (FMF.any())
+      Inst->setFastMathFlags(FMF);
+    return 0;
+  }
+
   // Casts.
   case lltok::kw_trunc:
   case lltok::kw_zext:
@@ -5572,6 +5590,11 @@ int LLParser::ParseGetElementPtr(Instruction *&Inst, PerFunctionState &PFS) {
 
   SmallVector<Value*, 16> Indices;
   bool AteExtraComma = false;
+  // GEP returns a vector of pointers if at least one of parameters is a vector.
+  // All vector parameters should have the same vector width.
+  unsigned GEPWidth = BaseType->isVectorTy() ?
+    BaseType->getVectorNumElements() : 0;
+
   while (EatIfPresent(lltok::comma)) {
     if (Lex.getKind() == lltok::MetadataVar) {
       AteExtraComma = true;
@@ -5580,14 +5603,13 @@ int LLParser::ParseGetElementPtr(Instruction *&Inst, PerFunctionState &PFS) {
     if (ParseTypeAndValue(Val, EltLoc, PFS)) return true;
     if (!Val->getType()->getScalarType()->isIntegerTy())
       return Error(EltLoc, "getelementptr index must be an integer");
-    if (Val->getType()->isVectorTy() != Ptr->getType()->isVectorTy())
-      return Error(EltLoc, "getelementptr index type missmatch");
+
     if (Val->getType()->isVectorTy()) {
-      unsigned ValNumEl = cast<VectorType>(Val->getType())->getNumElements();
-      unsigned PtrNumEl = cast<VectorType>(Ptr->getType())->getNumElements();
-      if (ValNumEl != PtrNumEl)
+      unsigned ValNumEl = Val->getType()->getVectorNumElements();
+      if (GEPWidth && GEPWidth != ValNumEl)
         return Error(EltLoc,
           "getelementptr vector index has a wrong number of elements");
+      GEPWidth = ValNumEl;
     }
     Indices.push_back(Val);
   }
diff --git a/lib/AsmParser/LLToken.h b/lib/AsmParser/LLToken.h
index 2487d1208133..691f085f0c9f 100644
--- a/lib/AsmParser/LLToken.h
+++ b/lib/AsmParser/LLToken.h
@@ -100,6 +100,7 @@ namespace lltok {
     // Attributes:
     kw_attributes,
     kw_alwaysinline,
+    kw_argmemonly,
     kw_sanitize_address,
     kw_builtin,
     kw_byval,
diff --git a/lib/Bitcode/Reader/BitcodeReader.cpp b/lib/Bitcode/Reader/BitcodeReader.cpp
index 09f0b689bdc3..c04e8b9f1f37 100644
--- a/lib/Bitcode/Reader/BitcodeReader.cpp
+++ b/lib/Bitcode/Reader/BitcodeReader.cpp
@@ -697,6 +697,21 @@ static Comdat::SelectionKind getDecodedComdatSelectionKind(unsigned Val) {
   }
 }
 
+static FastMathFlags getDecodedFastMathFlags(unsigned Val) {
+  FastMathFlags FMF;
+  if (0 != (Val & FastMathFlags::UnsafeAlgebra))
+    FMF.setUnsafeAlgebra();
+  if (0 != (Val & FastMathFlags::NoNaNs))
+    FMF.setNoNaNs();
+  if (0 != (Val & FastMathFlags::NoInfs))
+    FMF.setNoInfs();
+  if (0 != (Val & FastMathFlags::NoSignedZeros))
+    FMF.setNoSignedZeros();
+  if (0 != (Val & FastMathFlags::AllowReciprocal))
+    FMF.setAllowReciprocal();
+  return FMF;
+}
+
 static void upgradeDLLImportExportLinkage(llvm::GlobalValue *GV, unsigned Val) {
   switch (Val) {
   case 5: GV->setDLLStorageClass(GlobalValue::DLLImportStorageClass); break;
@@ -1075,6 +1090,8 @@ static Attribute::AttrKind getAttrFromCode(uint64_t Code) {
     return Attribute::Alignment;
   case bitc::ATTR_KIND_ALWAYS_INLINE:
     return Attribute::AlwaysInline;
+  case bitc::ATTR_KIND_ARGMEMONLY:
+    return Attribute::ArgMemOnly;
   case bitc::ATTR_KIND_BUILTIN:
     return Attribute::Builtin;
   case bitc::ATTR_KIND_BY_VAL:
@@ -3472,17 +3489,7 @@ std::error_code BitcodeReader::parseFunctionBody(Function *F) {
           if (Record[OpNum] & (1 << bitc::PEO_EXACT))
             cast<BinaryOperator>(I)->setIsExact(true);
         } else if (isa<FPMathOperator>(I)) {
-          FastMathFlags FMF;
-          if (0 != (Record[OpNum] & FastMathFlags::UnsafeAlgebra))
-            FMF.setUnsafeAlgebra();
-          if (0 != (Record[OpNum] & FastMathFlags::NoNaNs))
-            FMF.setNoNaNs();
-          if (0 != (Record[OpNum] & FastMathFlags::NoInfs))
-            FMF.setNoInfs();
-          if (0 != (Record[OpNum] & FastMathFlags::NoSignedZeros))
-            FMF.setNoSignedZeros();
-          if (0 != (Record[OpNum] & FastMathFlags::AllowReciprocal))
-            FMF.setAllowReciprocal();
+          FastMathFlags FMF = getDecodedFastMathFlags(Record[OpNum]);
           if (FMF.any())
             I->setFastMathFlags(FMF);
         }
@@ -3739,14 +3746,25 @@ std::error_code BitcodeReader::parseFunctionBody(Function *F) {
       unsigned OpNum = 0;
       Value *LHS, *RHS;
       if (getValueTypePair(Record, OpNum, NextValueNo, LHS) ||
-          popValue(Record, OpNum, NextValueNo, LHS->getType(), RHS) ||
-          OpNum+1 != Record.size())
+          popValue(Record, OpNum, NextValueNo, LHS->getType(), RHS))
+        return error("Invalid record");
+
+      unsigned PredVal = Record[OpNum];
+      bool IsFP = LHS->getType()->isFPOrFPVectorTy();
+      FastMathFlags FMF;
+      if (IsFP && Record.size() > OpNum+1)
+        FMF = getDecodedFastMathFlags(Record[++OpNum]);
+
+      if (OpNum+1 != Record.size())
         return error("Invalid record");
 
       if (LHS->getType()->isFPOrFPVectorTy())
-        I = new FCmpInst((FCmpInst::Predicate)Record[OpNum], LHS, RHS);
+        I = new FCmpInst((FCmpInst::Predicate)PredVal, LHS, RHS);
       else
-        I = new ICmpInst((ICmpInst::Predicate)Record[OpNum], LHS, RHS);
+        I = new ICmpInst((ICmpInst::Predicate)PredVal, LHS, RHS);
+
+      if (FMF.any())
+        I->setFastMathFlags(FMF);
       InstructionList.push_back(I);
       break;
     }
@@ -4458,14 +4476,11 @@ std::error_code BitcodeReader::materialize(GlobalValue *GV) {
 
   // Upgrade any old intrinsic calls in the function.
   for (auto &I : UpgradedIntrinsics) {
-    if (I.first != I.second) {
-      for (auto UI = I.first->user_begin(), UE = I.first->user_end();
-           UI != UE;) {
-        User *U = *UI;
-        ++UI;
-        if (CallInst *CI = dyn_cast<CallInst>(U))
-          UpgradeIntrinsicCall(CI, I.second);
-      }
+    for (auto UI = I.first->user_begin(), UE = I.first->user_end(); UI != UE;) {
+      User *U = *UI;
+      ++UI;
+      if (CallInst *CI = dyn_cast<CallInst>(U))
+        UpgradeIntrinsicCall(CI, I.second);
     }
   }
 
@@ -4533,15 +4548,13 @@ std::error_code BitcodeReader::materializeModule(Module *M) {
   // module is materialized because there could always be another function body
   // with calls to the old function.
   for (auto &I : UpgradedIntrinsics) {
-    if (I.first != I.second) {
-      for (auto *U : I.first->users()) {
-        if (CallInst *CI = dyn_cast<CallInst>(U))
-          UpgradeIntrinsicCall(CI, I.second);
-      }
-      if (!I.first->use_empty())
-        I.first->replaceAllUsesWith(I.second);
-      I.first->eraseFromParent();
+    for (auto *U : I.first->users()) {
+      if (CallInst *CI = dyn_cast<CallInst>(U))
+        UpgradeIntrinsicCall(CI, I.second);
     }
+    if (!I.first->use_empty())
+      I.first->replaceAllUsesWith(I.second);
+    I.first->eraseFromParent();
   }
   UpgradedIntrinsics.clear();
 
diff --git a/lib/Bitcode/Writer/BitcodeWriter.cpp b/lib/Bitcode/Writer/BitcodeWriter.cpp
index 622f7eaf0784..1a70ba5ac127 100644
--- a/lib/Bitcode/Writer/BitcodeWriter.cpp
+++ b/lib/Bitcode/Writer/BitcodeWriter.cpp
@@ -162,6 +162,8 @@ static uint64_t getAttrKindEncoding(Attribute::AttrKind Kind) {
     return bitc::ATTR_KIND_ALIGNMENT;
   case Attribute::AlwaysInline:
     return bitc::ATTR_KIND_ALWAYS_INLINE;
+  case Attribute::ArgMemOnly:
+    return bitc::ATTR_KIND_ARGMEMONLY;
   case Attribute::Builtin:
     return bitc::ATTR_KIND_BUILTIN;
   case Attribute::ByVal:
@@ -1759,13 +1761,17 @@ static void WriteInstruction(const Instruction &I, unsigned InstID,
     pushValue(I.getOperand(2), InstID, Vals, VE);
     break;
   case Instruction::ICmp:
-  case Instruction::FCmp:
+  case Instruction::FCmp: {
     // compare returning Int1Ty or vector of Int1Ty
     Code = bitc::FUNC_CODE_INST_CMP2;
     PushValueAndType(I.getOperand(0), InstID, Vals, VE);
     pushValue(I.getOperand(1), InstID, Vals, VE);
     Vals.push_back(cast<CmpInst>(I).getPredicate());
+    uint64_t Flags = GetOptimizationFlags(&I);
+    if (Flags != 0)
+      Vals.push_back(Flags);
     break;
+  }
 
   case Instruction::Ret:
     {
diff --git a/lib/CodeGen/Analysis.cpp b/lib/CodeGen/Analysis.cpp
index 3224fac25cb4..98d4c8afc7b9 100644
--- a/lib/CodeGen/Analysis.cpp
+++ b/lib/CodeGen/Analysis.cpp
@@ -81,27 +81,27 @@ unsigned llvm::ComputeLinearIndex(Type *Ty,
 /// If Offsets is non-null, it points to a vector to be filled in
 /// with the in-memory offsets of each of the individual values.
 ///
-void llvm::ComputeValueVTs(const TargetLowering &TLI, Type *Ty,
-                           SmallVectorImpl<EVT> &ValueVTs,
+void llvm::ComputeValueVTs(const TargetLowering &TLI, const DataLayout &DL,
+                           Type *Ty, SmallVectorImpl<EVT> &ValueVTs,
                            SmallVectorImpl<uint64_t> *Offsets,
                            uint64_t StartingOffset) {
   // Given a struct type, recursively traverse the elements.
   if (StructType *STy = dyn_cast<StructType>(Ty)) {
-    const StructLayout *SL = TLI.getDataLayout()->getStructLayout(STy);
+    const StructLayout *SL = DL.getStructLayout(STy);
     for (StructType::element_iterator EB = STy->element_begin(),
                                       EI = EB,
                                       EE = STy->element_end();
          EI != EE; ++EI)
-      ComputeValueVTs(TLI, *EI, ValueVTs, Offsets,
+      ComputeValueVTs(TLI, DL, *EI, ValueVTs, Offsets,
                       StartingOffset + SL->getElementOffset(EI - EB));
     return;
   }
   // Given an array type, recursively traverse the elements.
   if (ArrayType *ATy = dyn_cast<ArrayType>(Ty)) {
     Type *EltTy = ATy->getElementType();
-    uint64_t EltSize = TLI.getDataLayout()->getTypeAllocSize(EltTy);
+    uint64_t EltSize = DL.getTypeAllocSize(EltTy);
     for (unsigned i = 0, e = ATy->getNumElements(); i != e; ++i)
-      ComputeValueVTs(TLI, EltTy, ValueVTs, Offsets,
+      ComputeValueVTs(TLI, DL, EltTy, ValueVTs, Offsets,
                       StartingOffset + i * EltSize);
     return;
   }
@@ -109,7 +109,7 @@ void llvm::ComputeValueVTs(const TargetLowering &TLI, Type *Ty,
   if (Ty->isVoidTy())
     return;
   // Base case: we can get an EVT for this LLVM IR type.
-  ValueVTs.push_back(TLI.getValueType(Ty));
+  ValueVTs.push_back(TLI.getValueType(DL, Ty));
   if (Offsets)
     Offsets->push_back(StartingOffset);
 }
@@ -233,7 +233,8 @@ static bool isNoopBitcast(Type *T1, Type *T2,
 static const Value *getNoopInput(const Value *V,
                                  SmallVectorImpl<unsigned> &ValLoc,
                                  unsigned &DataBits,
-                                 const TargetLoweringBase &TLI) {
+                                 const TargetLoweringBase &TLI,
+                                 const DataLayout &DL) {
   while (true) {
     // Try to look through V1; if V1 is not an instruction, it can't be looked
     // through.
@@ -255,16 +256,16 @@ static const Value *getNoopInput(const Value *V,
       // Make sure this isn't a truncating or extending cast.  We could
       // support this eventually, but don't bother for now.
       if (!isa<VectorType>(I->getType()) &&
-          TLI.getPointerTy().getSizeInBits() ==
-          cast<IntegerType>(Op->getType())->getBitWidth())
+          DL.getPointerSizeInBits() ==
+              cast<IntegerType>(Op->getType())->getBitWidth())
         NoopInput = Op;
     } else if (isa<PtrToIntInst>(I)) {
       // Look through ptrtoint.
       // Make sure this isn't a truncating or extending cast.  We could
       // support this eventually, but don't bother for now.
       if (!isa<VectorType>(I->getType()) &&
-          TLI.getPointerTy().getSizeInBits() ==
-          cast<IntegerType>(I->getType())->getBitWidth())
+          DL.getPointerSizeInBits() ==
+              cast<IntegerType>(I->getType())->getBitWidth())
         NoopInput = Op;
     } else if (isa<TruncInst>(I) &&
                TLI.allowTruncateForTailCall(Op->getType(), I->getType())) {
@@ -331,14 +332,15 @@ static bool slotOnlyDiscardsData(const Value *RetVal, const Value *CallVal,
                                  SmallVectorImpl<unsigned> &RetIndices,
                                  SmallVectorImpl<unsigned> &CallIndices,
                                  bool AllowDifferingSizes,
-                                 const TargetLoweringBase &TLI) {
+                                 const TargetLoweringBase &TLI,
+                                 const DataLayout &DL) {
 
   // Trace the sub-value needed by the return value as far back up the graph as
   // possible, in the hope that it will intersect with the value produced by the
   // call. In the simple case with no "returned" attribute, the hope is actually
   // that we end up back at the tail call instruction itself.
   unsigned BitsRequired = UINT_MAX;
-  RetVal = getNoopInput(RetVal, RetIndices, BitsRequired, TLI);
+  RetVal = getNoopInput(RetVal, RetIndices, BitsRequired, TLI, DL);
 
   // If this slot in the value returned is undef, it doesn't matter what the
   // call puts there, it'll be fine.
@@ -350,7 +352,7 @@ static bool slotOnlyDiscardsData(const Value *RetVal, const Value *CallVal,
   // a "returned" attribute, the search will be blocked immediately and the loop
   // a Noop.
   unsigned BitsProvided = UINT_MAX;
-  CallVal = getNoopInput(CallVal, CallIndices, BitsProvided, TLI);
+  CallVal = getNoopInput(CallVal, CallIndices, BitsProvided, TLI, DL);
 
   // There's no hope if we can't actually trace them to (the same part of!) the
   // same value.
@@ -606,7 +608,8 @@ bool llvm::returnTypeIsEligibleForTailCall(const Function *F,
     // Finally, we can check whether the value produced by the tail call at this
     // index is compatible with the value we return.
     if (!slotOnlyDiscardsData(RetVal, CallVal, TmpRetPath, TmpCallPath,
-                              AllowDifferingSizes, TLI))
+                              AllowDifferingSizes, TLI,
+                              F->getParent()->getDataLayout()))
       return false;
 
     CallEmpty  = !nextRealType(CallSubTypes, CallPath);
diff --git a/lib/CodeGen/AsmPrinter/ARMException.cpp b/lib/CodeGen/AsmPrinter/ARMException.cpp
index 4cb460a7bbfc..0bad7954b980 100644
--- a/lib/CodeGen/AsmPrinter/ARMException.cpp
+++ b/lib/CodeGen/AsmPrinter/ARMException.cpp
@@ -69,24 +69,32 @@ void ARMException::beginFunction(const MachineFunction *MF) {
 ///
 void ARMException::endFunction(const MachineFunction *MF) {
   ARMTargetStreamer &ATS = getTargetStreamer();
+  const Function *F = MF->getFunction();
+  const Function *Per = nullptr;
+  if (F->hasPersonalityFn())
+    Per = dyn_cast<Function>(F->getPersonalityFn()->stripPointerCasts());
+  assert(!MMI->getPersonality() || Per == MMI->getPersonality());
+  bool forceEmitPersonality =
+    F->hasPersonalityFn() && !isNoOpWithoutInvoke(classifyEHPersonality(Per)) &&
+    F->needsUnwindTableEntry();
+  bool shouldEmitPersonality = forceEmitPersonality ||
+    !MMI->getLandingPads().empty();
   if (!Asm->MF->getFunction()->needsUnwindTableEntry() &&
-      MMI->getLandingPads().empty())
+      !shouldEmitPersonality)
     ATS.emitCantUnwind();
-  else {
-    if (!MMI->getLandingPads().empty()) {
-      // Emit references to personality.
-      if (const Function *Personality = MMI->getPersonality()) {
-        MCSymbol *PerSym = Asm->getSymbol(Personality);
-        Asm->OutStreamer->EmitSymbolAttribute(PerSym, MCSA_Global);
-        ATS.emitPersonality(PerSym);
-      }
-
-      // Emit .handlerdata directive.
-      ATS.emitHandlerData();
-
-      // Emit actual exception table
-      emitExceptionTable();
+  else if (shouldEmitPersonality) {
+    // Emit references to personality.
+    if (Per) {
+      MCSymbol *PerSym = Asm->getSymbol(Per);
+      Asm->OutStreamer->EmitSymbolAttribute(PerSym, MCSA_Global);
+      ATS.emitPersonality(PerSym);
     }
+
+    // Emit .handlerdata directive.
+    ATS.emitHandlerData();
+
+    // Emit actual exception table
+    emitExceptionTable();
   }
 
   if (Asm->MAI->getExceptionHandlingType() == ExceptionHandling::ARM)
diff --git a/lib/CodeGen/AsmPrinter/AsmPrinter.cpp b/lib/CodeGen/AsmPrinter/AsmPrinter.cpp
index 8a7e9f991611..125047e7bbb5 100644
--- a/lib/CodeGen/AsmPrinter/AsmPrinter.cpp
+++ b/lib/CodeGen/AsmPrinter/AsmPrinter.cpp
@@ -19,7 +19,6 @@
 #include "llvm/ADT/SmallString.h"
 #include "llvm/ADT/Statistic.h"
 #include "llvm/Analysis/ConstantFolding.h"
-#include "llvm/Analysis/JumpInstrTableInfo.h"
 #include "llvm/CodeGen/Analysis.h"
 #include "llvm/CodeGen/GCMetadataPrinter.h"
 #include "llvm/CodeGen/MachineConstantPool.h"
@@ -820,7 +819,7 @@ void AsmPrinter::EmitFunctionBody() {
         emitCFIInstruction(MI);
         break;
 
-      case TargetOpcode::FRAME_ALLOC:
+      case TargetOpcode::LOCAL_ESCAPE:
         emitFrameAlloc(MI);
         break;
 
@@ -1024,7 +1023,7 @@ bool AsmPrinter::doFinalization(Module &M) {
 
   // Emit visibility info for declarations
   for (const Function &F : M) {
-    if (!F.isDeclaration())
+    if (!F.isDeclarationForLinker())
       continue;
     GlobalValue::VisibilityTypes V = F.getVisibility();
     if (V == GlobalValue::DefaultVisibility)
diff --git a/lib/CodeGen/AsmPrinter/DwarfCFIException.cpp b/lib/CodeGen/AsmPrinter/DwarfCFIException.cpp
index 0bc873e326be..2c212c7ecee1 100644
--- a/lib/CodeGen/AsmPrinter/DwarfCFIException.cpp
+++ b/lib/CodeGen/AsmPrinter/DwarfCFIException.cpp
@@ -89,6 +89,7 @@ void DwarfCFIException::endModule() {
 
 void DwarfCFIException::beginFunction(const MachineFunction *MF) {
   shouldEmitMoves = shouldEmitPersonality = shouldEmitLSDA = false;
+  const Function *F = MF->getFunction();
 
   // If any landing pads survive, we need an EH table.
   bool hasLandingPads = !MMI->getLandingPads().empty();
@@ -104,10 +105,24 @@ void DwarfCFIException::beginFunction(const MachineFunction *MF) {
 
   const TargetLoweringObjectFile &TLOF = Asm->getObjFileLowering();
   unsigned PerEncoding = TLOF.getPersonalityEncoding();
-  const Function *Per = MMI->getPersonality();
-
-  shouldEmitPersonality = hasLandingPads &&
-    PerEncoding != dwarf::DW_EH_PE_omit && Per;
+  const Function *Per = nullptr;
+  if (F->hasPersonalityFn())
+    Per = dyn_cast<Function>(F->getPersonalityFn()->stripPointerCasts());
+  assert(!MMI->getPersonality() || Per == MMI->getPersonality());
+
+  // Emit a personality function even when there are no landing pads
+  bool forceEmitPersonality =
+      // ...if a personality function is explicitly specified
+      F->hasPersonalityFn() &&
+      // ... and it's not known to be a noop in the absence of invokes
+      !isNoOpWithoutInvoke(classifyEHPersonality(Per)) &&
+      // ... and we're not explicitly asked not to emit it
+      F->needsUnwindTableEntry();
+
+  shouldEmitPersonality =
+      (forceEmitPersonality ||
+       (hasLandingPads && PerEncoding != dwarf::DW_EH_PE_omit)) &&
+      Per;
 
   unsigned LSDAEncoding = TLOF.getLSDAEncoding();
   shouldEmitLSDA = shouldEmitPersonality &&
@@ -123,6 +138,11 @@ void DwarfCFIException::beginFunction(const MachineFunction *MF) {
   if (!shouldEmitPersonality)
     return;
 
+  // If we are forced to emit this personality, make sure to record
+  // it because it might not appear in any landingpad
+  if (forceEmitPersonality)
+    MMI->addPersonality(Per);
+
   const MCSymbol *Sym =
       TLOF.getCFIPersonalitySymbol(Per, *Asm->Mang, Asm->TM, MMI);
   Asm->OutStreamer->EmitCFIPersonality(Sym, PerEncoding);
diff --git a/lib/CodeGen/AsmPrinter/DwarfDebug.h b/lib/CodeGen/AsmPrinter/DwarfDebug.h
index 1c3e2aec64ab..01f34c6eb81c 100644
--- a/lib/CodeGen/AsmPrinter/DwarfDebug.h
+++ b/lib/CodeGen/AsmPrinter/DwarfDebug.h
@@ -49,7 +49,7 @@ class DwarfUnit;
 class MachineModuleInfo;
 
 //===----------------------------------------------------------------------===//
-/// \brief This class is used to record source line correspondence.
+/// This class is used to record source line correspondence.
 class SrcLineInfo {
   unsigned Line;     // Source line number.
   unsigned Column;   // Source column.
@@ -161,7 +161,7 @@ public:
 
     return dwarf::DW_TAG_variable;
   }
-  /// \brief Return true if DbgVariable is artificial.
+  /// Return true if DbgVariable is artificial.
   bool isArtificial() const {