llvm-svn: 200349

llvm-svn: 200348

Needed to fix PR18303 to correctly re-encode the instruction if it
is relaxed.

We keep a copy of the MCSubtargetInfo to make sure that we are not
effected by future changes to the subtarget info coming from the
assembler (e.g. when parsing .code 16 directived).

llvm-svn: 200347

Add MCSubtargetInfo parameter
virtual void EmitInstToFragment(const MCInst &Inst, const MCSubtargetInfo &);
virtual void EmitInstToData(const MCInst &Inst, const MCSubtargetInfo &);

llvm-svn: 200346

llvm-svn: 200345

llvm-svn: 200341

Reviewed at http://llvm-reviews.chandlerc.com/D2232

llvm-svn: 200340

llvm-svn: 200335

llvm-svn: 200333

As opposed to GCC/GAS the default ABI for Mips64 is n64.
Compatibility bit should be set if o32 ABI is used when targeting Mips64.

llvm-svn: 200332

Makes it easy to use BumpPtrAllocator to make a copy of StringRef strings.

llvm-svn: 200331

The code was missing the case for aggregate parameters and
hence was emitting them as .b0 type. Also fixed a couple
of comments.

llvm-svn: 200325

This improves the fix committed at revision 199683 adding the
following new target specific combine rules:

1) fold (v4i32: vselect <0,0,-1,-1>, A, B) ->
        (v4i32 (bitcast (movsd (v2i64 (bitcast A)), (v2i64 (bitcast B))) ))

2) fold (v4f32: vselect <0,0,-1,-1>, A, B) ->
        (v4f32 (bitcast (movsd (v2f64 (bitcast A)), (v2f64 (bitcast B))) ))

3) fold (v4i32: vselect <-1,-1,0,0>, A, B) ->
        (v4i32 (bitcast (movsd (v2i64 (bitcast B)), (v2i64 (bitcast A))) ))

4) fold (v4f32: vselect <-1,-1,0,0>, A, B) ->
        (v4f32 (bitcast (movsd (v2i64 (bitcast B)), (v2i64 (bitcast A))) ))

llvm-svn: 200324

llvm-svn: 200323

When simplifycfg moves an instruction, it must drop metadata it doesn't know
is still valid with the preconditions changes. In particular, it must drop
the range and tbaa metadata.

The patch implements this with an utility function to drop all metadata not
in a white list.

llvm-svn: 200322

The llvm_headers_do_not_build project needs to be excluded from the default build, otherwise it gets built (at least in Visual Studio 2013).

Thanks to chapuni200000 for help with this in IRC!

llvm-svn: 200321

Make sure that we don't introduce illegal build_vector dag nodes
when trying to fold a sign_extend of a build_vector.

This fixes a regression introduced by r200234.
Added test CodeGen/X86/fold-vector-sext-crash.ll
to verify that llc no longer crashes with an assertion failure
due to an illegal build_vector of type MVT::v4i64.

Thanks to Ilia Filippov for spotting this regression and for
providing a reproducible test case.

llvm-svn: 200313

[CMake] llvm_update_compile_flags(name) doesn't require source files. TARGET PROPERTY SOURCES has them.

llvm-svn: 200311

At present, this handles .tc (error) and needs to be expanded to deal properly with .machine

llvm-svn: 200309

It is the final step to deprecate contextual CMAKE_CXX_FLAGS.

llvm-svn: 200303

[CMake] Enhance llvm_update_compile_flags(name sources) to handle LLVM_REQUIRES_EH and LLVM_REQUIRES_RTTI.

LLVM_REQUIRES_EH implies LLVM_REQUIRES_RTTI. It is as same behavior as Makefile.rule's.
llvm/examples/ExceptionDemo is affected. (It was built with -fno-rtti.)

For MSVC, Remove flags like "/EHsc /GR" in HandleLLVMOptions, or CL.EXE complains with flags like "/GR /GR-".

llvm_update_compile_flags() updates source file property if the target contains *.c.
COMPILE_FLAGS in target properties affects both C++ and C!

LLVM_NO_RTTI is deprecated. It was introduced by me and was my mistake.

llvm-svn: 200301

llvm-svn: 200300

With this tweaks, also unittests are compiled with -ffunction-sections.

It's hard to control contextual CMAKE_CXX_FLAGS. We should get rid of twiddling it as possible.

llvm-svn: 200299

llvm-svn: 200298

llvm-svn: 200297

vectorizer, placing it behind an off-by-default flag.

It turns out that block frequency isn't what we want at all, here or
elsewhere. This has been I think a nagging feeling for several of us
working with it, but Arnold has given some really nice simple examples
where the results are so comprehensively wrong that they aren't useful.

I'm planning to email the dev list with a summary of why its not really
useful and a couple of ideas about how to better structure these types
of heuristics.

llvm-svn: 200294

GPRC_NOR0 is not a subclass of GPRC (because it also contains the ZERO pseudo
register). As a result, we also need to check for it in the spilling code.

llvm-svn: 200288

Improve handling of EnforceSmallerThan. Remove all types that are smaller from the larger set not just the smallest type from the smaller set. Ensure 'smaller' vectors have the same or fewer total bits. Similar for 'larger' vectors.

llvm-svn: 200287

llvm-svn: 200285

Fixes half a dozen piglit tests with radeonsi.

Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
llvm-svn: 200283

Also emit the stubs that were generated for references to typeinfo
symbols.

llvm-svn: 200282

Summary:
I searched Transforms/ and Analysis/ for 'ByVal' and updated those call
sites to check for inalloca if appropriate.

I added tests for any change that would allow an optimization to fire on
inalloca.

Reviewers: nlewycky

Differential Revision: http://llvm-reviews.chandlerc.com/D2449

llvm-svn: 200281

This avoids miscompiling MS inline asm in LLVM where we have to infer
clobbers.  Test case forthcoming in Clang.

llvm-svn: 200279

LCSSA from it caused a crasher with the LoopUnroll pass.

This crasher is really nasty. We destroy LCSSA form in a suprising way.
When unrolling a loop into an outer loop, we not only need to restore
LCSSA form for the outer loop, but for all children of the outer loop.
This is somewhat obvious in retrospect, but hey!

While this seems pretty heavy-handed, it's not that bad. Fundamentally,
we only do this when we unroll a loop, which is already a heavyweight
operation. We're unrolling all of these hypothetical inner loops as
well, so their size and complexity is already on the critical path. This
is just adding another pass over them to re-canonicalize.

I have a test case from PR18616 that is great for reproducing this, but
pretty useless to check in as it relies on many 10s of nested empty
loops that get unrolled and deleted in just the right order. =/ What's
worse is that investigating this has exposed another source of failure
that is likely to be even harder to test. I'll try to come up with test
cases for these fixes, but I want to get the fixes into the tree first
as they're causing crashes in the wild.

llvm-svn: 200273

[TLI] Add a new hook to TargetLowering to query the target if a load of a constant should be converted to simply the constant itself.

Before this patch we used getIntImmCost from TargetTransformInfo to determine if
a load of a constant should be converted to just a constant, but the threshold
for this was set to an arbitrary value. This value works well for the two
targets (X86 and ARM) that implement this target-hook, but it isn't
target-independent at all.

Now targets have the possibility to decide directly if this optimization should
be performed. The default value is set to false to preserve the current
behavior. The target hook has been moved to TargetLowering, which removed the
last use and need of TargetTransformInfo in SelectionDAG.

llvm-svn: 200271

The vectorizer takes a loop like this and widens all instructions except for the
store. The stores are scalarized/unrolled and hidden behind an "if" block.

  for (i = 0; i < 128; ++i) {
    if (a[i] < 10)
      a[i] += val;
  }

  for (i = 0; i < 128; i+=2) {
    v = a[i:i+1];
    v0 = (extract v, 0) + 10;
    v1 = (extract v, 1) + 10;
    if (v0 < 10)
      a[i] = v0;
    if (v1 < 10)
      a[i] = v1;
  }

The vectorizer relies on subsequent optimizations to sink instructions into the
conditional block where they are anticipated.

The flag "vectorize-num-stores-pred" controls whether and how many stores to
handle this way. Vectorization of conditional stores is disabled per default for
now.

This patch also adds a change to the heuristic when the flag
"enable-loadstore-runtime-unroll" is enabled (off by default). It unrolls small
loops until load/store ports are saturated. This heuristic uses TTI's
getMaxUnrollFactor as a measure for load/store ports.

I also added a second flag -enable-cond-stores-vec. It will enable vectorization
of conditional stores. But there is no cost model for vectorization of
conditional stores in place yet so this will not do good at the moment.

rdar://15892953

Results for x86-64 -O3 -mavx +/- -mllvm -enable-loadstore-runtime-unroll
-vectorize-num-stores-pred=1 (before the BFI change):

 Performance Regressions:
   Benchmarks/Ptrdist/yacr2/yacr2 7.35% (maze3() is identical but 10% slower)
   Applications/siod/siod         2.18%
 Performance improvements:
   mesa                          -4.42%
   libquantum                    -4.15%

 With a patch that slightly changes the register heuristics (by subtracting the
 induction variable on both sides of the register pressure equation, as the
 induction variable is probably not really unrolled):

 Performance Regressions:
   Benchmarks/Ptrdist/yacr2/yacr2  7.73%
   Applications/siod/siod          1.97%

 Performance Improvements:
   libquantum                    -13.05% (we now also unroll quantum_toffoli)
   mesa                           -4.27%

llvm-svn: 200270

code to see if we're emitting a function into a non-default
text section. This is still a less-than-ideal solution, but more
contained than r199871 to determine whether or not we're emitting
code into an array of comdat sections.

llvm-svn: 200269

llvm-svn: 200264

uint32.

When folding branches to common destination, the updated branch weights
can exceed uint32 by more than factor of 2. We should keep halving the
weights until they can fit into uint32.

llvm-svn: 200262

Without this fix, WaitResult is not defined.

llvm-svn: 200259