FIXME: I don't think it'd be smart.
llvm-svn: 193256

llvm-svn: 193255

llvm-svn: 193254

llvm-svn: 193253

Per Nadav's review comments for r192866.

llvm-svn: 193252

 Major steps include:
 1). introduces a not-addr-taken bit-field in GlobalVariable
 2). GlobalOpt pass sets "not-address-taken" if it proves a global varirable 
    dosen't have its address taken.
 3). AA use this info for disambiguation. 

llvm-svn: 193251

Pacifies GCC's -Wstrict-prototypes.

llvm-svn: 193249

llvm-svn: 193247

llvm-svn: 193243

llvm-svn: 193240

llvm-svn: 193239

llvm-svn: 193238

llvm-svn: 193235

This provides rudimentary testing of the llvm-c api.

The following commands are implemented:

  * --module-dump
    Read bytecode from stdin - print ir

  * --module-list-functions
    Read bytecode from stdin - list summary of functions

  * --module-list-globals
    Read bytecode from stdin - list summary of globals

  * --targets-list
    List available targets

  * --object-list-sections
    Read object file from stdin - list sections

  * --object-list-symbols
    Read object file from stdin - list symbols (like nm)

  * --disassemble
    Read lines of triple, hex ascii machine code from stdin - print disassembly

  * --calc
    Read lines of name, rpn from stdin - print generated module ir

Differential-Revision: http://llvm-reviews.chandlerc.com/D1776
llvm-svn: 193233

ffs(x) broke the mingw buildbot.

llvm-svn: 193225

llvm-svn: 193224

This fixes piglit:
- shaders/glsl-fs-texture2d-masked
- shaders/glsl-fs-texture2d-masked-4

Patch by: Marek Olšák

Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 193222

The SelectionDAGBuilder was promoting vector kernel arguments to legal
types, but this won't work for R600 and SI since kernel arguments are
stored in memory and can't be promoted.  In order to handle vector
arguments correctly we need to look at the original types from the LLVM IR
function.

llvm-svn: 193215

For some targets, it is useful to be able to look at the original
type of an argument without having to dig through the original IR.

This also fixes a bug in SelectionDAGBuilder where InputArg.PartOffset
was not taking into account the offset of structure elements.

Patch by: Justin Holewinski

Tom Stellard:
  - Changed the type of ArgVT to EVT, so it can store non-simple types
    like v3i32.

llvm-svn: 193214

llvm-svn: 193213

llvm-svn: 193212

Code review by Eric Christopher and Rafael Espindola.

llvm-svn: 193209

<rdar://problem/15192473>

llvm-svn: 193199

llvm-svn: 193198

Remove unnecessary creation of LexicalScope in collectDeadVariables.
The created LexicialScope was only used to get isAbstractScope, which
should be false from the creation:
"new LexicalScope(NULL, DIDescriptor(SP), NULL, false);".

We can also remove a DenseMap that holds the created LexicalScopes.

llvm-svn: 193196

llvm-svn: 193194

Since (as of r190716) Clang no longer emits debug info for C++ friend
declarations (and it seems GCC never has/does, which was the motivation
for the Clang change), there's no actual reachable case for implementing
the part of DWARF 4, Section 7.27 part 5 that pertains to friends.

Leave an assert here so that if/when we do have a client producing
friends and using type units, we can fill in the gap and add appropriate
(unit and feature) tests.

llvm-svn: 193193

The loop bounds here are uint32_t variables, so it makes sense for the
loop variables to have the same type.

Patch by Yuchen Wu!

llvm-svn: 193192

Rename Size to EndPos, which makes more sense because the variable
stores the last location of the blocks.

Patch by Yuchen Wu!

llvm-svn: 193189

The set of circumstances where the writeback register is allowed to be in the
list of registers is rather baroque, but I think this implements them all on
the assembly parsing side.

For disassembly, we still warn about an ARM-mode LDM even if the architecture
revision is < v7 (the required architecture information isn't available). It's
a silly instruction anyway, so hopefully no-one will mind.

rdar://problem/15223374

llvm-svn: 193185

llvm-svn: 193183

llvm-svn: 193180

The AMDGPUIndirectAddressing pass was previously responsible for
lowering private loads and stores to indirect addressing instructions.
However, this pass was buggy and way too complicated.  The only
advantage it had over the new simplified code was that it saved one
instruction per direct write to private memory.  This optimization
likely has a minimal impact on performance, and we may be able
to duplicate it using some other transformation.

For the private address space, we now:
1. Lower private loads/store to Register(Load|Store) instructions
2. Reserve part of the register file as 'private memory'
3. After regalloc lower the Register(Load|Store) instructions to
   MOV instructions that use indirect addressing.

llvm-svn: 193179

llvm-svn: 193178

llvm-svn: 193177

Includes a test case/FIXME demonstrating a bug/limitation in pointer to
member hashing. To be honest I'm not sure why we don't just always use
summary hashing for referenced types... but perhaps I'm missing
something.

llvm-svn: 193175

Line counts in llvm-cov are read in as 64-bit integers but were being truncated
to 32-bit in collectLineCounts(), which caused overflow for large counts.
This patch fixes all counts to be uint64_t.

Patch by Yuchen Wu!

llvm-svn: 193172

llvm-svn: 193165

This is a stopgap fix for cast warnings introduced in r192864.

A proper fix should be investigated by the author when possible.

llvm-svn: 193160

These branches have a 16-bit offset (R_MIPS_PC16).

List of conditional branch instructions:
bnz.{b,h,w,d}
bnz.v
bz.{b,h,w,d}
bz.v

llvm-svn: 193157