Changing the sign when comparing the base pointer would introduce all
sorts of unexpected things like:
  %gep.i = getelementptr inbounds [1 x i8]* %a, i32 0, i32 0
  %gep2.i = getelementptr inbounds [1 x i8]* %b, i32 0, i32 0
  %cmp.i = icmp ult i8* %gep.i, %gep2.i
  %cmp.i1 = icmp ult [1 x i8]* %a, %b
  %cmp = icmp ne i1 %cmp.i, %cmp.i1
  ret i1 %cmp

into:
  %cmp.i = icmp slt [1 x i8]* %a, %b
  %cmp.i1 = icmp ult [1 x i8]* %a, %b
  %cmp = xor i1 %cmp.i, %cmp.i1
  ret i1 %cmp

By preserving the original sign, we now get:
  ret i1 false

This fixes PR16483.

llvm-svn: 185259

llvm-svn: 185258

Real world code sometimes has the denominator of a 'udiv' be a
'select'.  LLVM can handle such cases but only when the 'select'
operands are symmetric in structure (both select operands are a constant
power of two or a left shift, etc.).  This falls apart if we are dealt a
'udiv' where the code is not symetric or if the select operands lead us
to more select instructions.

Instead, we should treat the LHS and each select operand as a distinct
divide operation and try to optimize them independently.  If we can
to simplify each operation, then we can replace the 'udiv' with, say, a
'lshr' that has a new select with a bunch of new operands for the
select.

llvm-svn: 185257

llvm-svn: 185256

The attribute in the file allows git to perform End-of-Line
normalization (CRLF to LF).

llvm-svn: 185255

llvm-svn: 185254

1. Do not allow creating new memory access record in the InstructionToAccess map
   on the fly in function getAccessFor.
2. Do not allow user to modify the memory accesses returned by getAccessFor
   during the code generation process.

llvm-svn: 185253

llvm-svn: 185252

llvm-svn: 185251

llvm-svn: 185250

No functionality change.
It should suffice to check the type of a debug info metadata, instead of
calling Verify. For cases where we know the type of a DI metadata, use
assert.

llvm-svn: 185249

function is not used in release build.

llvm-svn: 185248

llvm-svn: 185247

llvm-svn: 185246

llvm-svn: 185245

llvm-svn: 185244

llvm-svn: 185243

We may, after other optimizations, find ourselves with IR that looks
like:

  %shl = shl i32 1, %y
  %cmp = icmp ult i32 %shl, 32

Instead, we should just compare the shift count:

  %cmp = icmp ult i32 %y, 5

llvm-svn: 185242

llvm-svn: 185241

OS Plugins' __init__ method takes two arguments: (self,process)

I was erroneously passing the session_dict as well as part of my PyCallable changes and that caused plugins to fail to work

llvm-svn: 185240

Thename says it's an i32*, but it was actually creating another i8*

llvm-svn: 185239

llvm-svn: 185238

llvm-svn: 185237

Allow block frequencies to exceed 32 bits by using the new
BlockFrequency division function.

llvm-svn: 185236

This fixes PR16418, which reports that a function calling
__builtin_unwind_init() asserts. The cause is that this generates a
spill/restore for VRSAVE, and we support that only on Darwin (because VRSAVE is
only really used on Darwin).

The test case checks only that we don't crash. We can add correctness checks
once someone verifies what behavior the function is supposed to have.

llvm-svn: 185235

llvm-svn: 185234

llvm-svn: 185233

- should resolve windows buildbot failure

llvm-svn: 185232

llvm-svn: 185231

To support this we have to insert 'extractelement' instructions to pick the right lane.
We had this functionality before but I removed it when we moved to the multi-block design because it was too complicated.

llvm-svn: 185230

across scopes.

When we declare an extern "C" name that is not a redeclaration of an entity in
the same scope, check whether it redeclares some extern "C" entity from another
scope, and if not, check whether it conflicts with a (non-extern-"C") entity in
the translation unit.

When we declare a name in the translation unit that is not a redeclaration,
check whether it conflicts with any extern "C" entities (possibly from other
scopes).

llvm-svn: 185229

Change assert("text") to assert(0 && "text").  The first case is a const char *
to bool conversion, which always evaluates to true, never triggering the
assert.  The second case will always trigger the assert.

llvm-svn: 185227

llvm-svn: 185226

- mistakenly used get_current_dir() linux function
- replaced with getcwd/_getcwd as appropriate for current platform

llvm-svn: 185225

been suitable for preparing a single IR function
for operation in the target.  However, using blocks
and lambdas creates other IR functions that also
need to be processed.

I have audited IRForTarget to make it process
multiple functions.  Where IRForTarget would add
new instructions at the beginning of the main
expression function, it now adds them on-demand
in the function where they are needed.  This is
enabled by a system of FunctionValueCaches, which
invoke a lambda to create or derive the values as
needed, or report the result of that lambda if it
has already been called for the given function.

<rdar://problem/14180236>

llvm-svn: 185224

llvm-svn: 185222

Revising the MCJIT ObjectCache interface to allow subclasses to avoid retaining references to returned objects

llvm-svn: 185221

llvm-svn: 185219

Hopefully, this fixes the PPC64 buildbots.

llvm-svn: 185218

Default parameters are evil and should not be used. Case and point this checkin that fixes implicit conversions that were happening.

llvm-svn: 185217