llvm-svn: 203492

Extend the error message generated by the Verifier when an intrinsic
name does not match the expected mangling to include the expected
name.  Simplifies debugging.

Patch by Philip Reames!

llvm-svn: 203490

MemCpyOpt: When merging memsets also merge the trivial case of two memsets with the same destination.

The testcase is from PR19092, but I think the bug described there is actually a clang issue.

llvm-svn: 203489

optimize a call to a llvm intrinsic to something that invovles a call to a C
library call, make sure it sets the right calling convention on the call.

e.g.
extern double pow(double, double);
double t(double x) {
  return pow(10, x);
}

Compiles to something like this for AAPCS-VFP:
define arm_aapcs_vfpcc double @t(double %x) #0 {
entry:
  %0 = call double @llvm.pow.f64(double 1.000000e+01, double %x)
  ret double %0
}

declare double @llvm.pow.f64(double, double) #1

Simplify libcall (part of instcombine) will turn the above into:
define arm_aapcs_vfpcc double @t(double %x) #0 {
entry:
  %__exp10 = call double @__exp10(double %x) #1
  ret double %__exp10
}

declare double @__exp10(double)

The pre-instcombine code works because calls to LLVM builtins are special.
Instruction selection will chose the right calling convention for the call.
However, the code after instcombine is wrong. The call to __exp10 will use
the C calling convention.

I can think of 3 options to fix this.

1. Make "C" calling convention just work since the target should know what CC
   is being used.

   This doesn't work because each function can use different CC with the "pcs"
   attribute.

2. Have Clang add the right CC keyword on the calls to LLVM builtin.

   This will work but it doesn't match the LLVM IR specification which states
   these are "Standard C Library Intrinsics".

3. Fix simplify libcall so the resulting calls to the C routines will have the
   proper CC keyword. e.g.
   %__exp10 = call arm_aapcs_vfpcc double @__exp10(double %x) #1

   This works and is the solution I implemented here.

Both solutions #2 and #3 would work. After carefully considering the pros and
cons, I decided to implement #3 for the following reasons.

1. It doesn't change the "spec" of the intrinsics.
2. It's a self-contained fix.

There are a couple of potential downsides.
1. There could be other places in the optimizer that is broken in the same way
   that's not addressed by this.
2. There could be other calling conventions that need to be propagated by
   simplify-libcall that's not handled.

But for now, this is the fix that I'm most comfortable with.

llvm-svn: 203488

llvm-svn: 203486

[forgot to 'svn add' before committing r203483]

llvm-svn: 203485

  * Add masking instructions before loads and stores (in MC layer).
  * Add masking instructions after SP changes (in MC layer).
  * Forbid loads, stores and SP changes in delay slots (in MI layer).

Differential Revision: http://llvm-reviews.chandlerc.com/D2904

llvm-svn: 203484

NVPTX, like the other backends, relies on generic symbol name sanitizing done by
MCSymbol. However, the ptxas assembler is more stringent and disallows some
additional characters in symbol names.

See PR19099 for more details.

llvm-svn: 203483

Patch by Manuel Jacob.

llvm-svn: 203482

There's now a normal UI for that, apparently.

Patch by Manuel Jacob.

llvm-svn: 203481

llvm-svn: 203472

This will replace the now badly named CLANG_IS_PRODUCTION.

llvm-svn: 203471

llvm-svn: 203469

llvm-svn: 203468

No functionality change.

llvm-svn: 203465

Summary:
This is a white lie to workaround a widespread bug in the -mfp64
implementation.

The problem is that none of the 32-bit fpu ops mention the fact that they
clobber the upper 32-bits of the 64-bit FPR. This allows MFHC1 to be
scheduled on the wrong side of most 32-bit FPU ops. Fixing that requires a
major overhaul of the FPU implementation which can't be done right now due to
time constraints.

MFHC1 is one of two affected instructions. These instructions are the only
FPU instructions that don't read or write the lower 32-bits. We therefore
pretend that it reads the bottom 32-bits to artificially create a dependency and
prevent the scheduler changing the behaviour of the code.
The other instruction is MTHC1 which will be fixed once I've have found a failing
test case for it. 

The testcase is test-suite/SingleSource/UnitTests/Vector/simple.c when
given TARGET_CFLAGS="-mips32r2 -mfp64 -mmsa".

Reviewers: jacksprat, matheusalmeida

Reviewed By: jacksprat

Differential Revision: http://llvm-reviews.chandlerc.com/D2966

llvm-svn: 203464

Removing llvm::distance and llvm::copy for iterator_range based on post-commit review feedback. Adding an explicit range-based constructor to SmallVector, which supersedes the llvm::copy functionality.

llvm-svn: 203460

llvm-svn: 203459

The function was making too many assumptions about its input:

1. The NEON_VDUP optimisation was far too aggressive, assuming (I
think) that the input would always be BUILD_VECTOR.

2. We were treating most unknown concats as legal (by returning Op
rather than SDValue()). I think only concats of pairs of vectors are
actually legal.

http://llvm.org/PR19094

llvm-svn: 203450

as well. I don't see any particular need but it imposes no cost to
support it and it makes the API cleaner.

llvm-svn: 203448

now that there is essentially no cost to doing so. Yay C++11.

llvm-svn: 203447

llvm-svn: 203444

llvm-svn: 203442

and caught by the MSan bootstrap build bot. This should hopefully get
the bot green at long last.

llvm-svn: 203441

llvm-svn: 203440

llvm-svn: 203439

commit. Sorry for the churn, just trying to keep it out of any
functionality changed.

llvm-svn: 203438

the stack of the analysis group because they are all immutable passes.
This is made clear by Craig's recent work to use override
systematically -- we weren't overriding anything for 'finalizePass'
because there is no such thing.

This is kind of a lame restriction on the API -- we can no longer push
and pop things, we just set up the stack and run. However, I'm not
invested in building some better solution on top of the existing
(terrifying) immutable pass and legacy pass manager.

llvm-svn: 203437

  - Use constructor instead of initializer list.
  - Disable ManyUnusedBits for now.

llvm-svn: 203436

llvm-svn: 203435

lines under 80-columns, etc.

llvm-svn: 203434

llvm-svn: 203433

clang-format, but with some modifications by me where it got things
wrong or got confused.

llvm-svn: 203432

constructors from the classes which only have a single reference member
to many other places. This resulted in them copying their single member
instead of moving. =/ Fix this.

There's really not a useful test to add sadly because these move
constructors are only called when something deep inside some standard
library implementation *needs* to move them. Many of the types aren't
even user-impacting types. Or, the objects are copyable anyways and so
the result was merely a performance problem rather than a correctness
problem.

Anyways, thanks for the review. And this is a great example of why
I wish I colud have the compiler write these for me.

llvm-svn: 203431

This is fallout from r203429.

llvm-svn: 203430

Split by comma once instead of multiple times.  Moving this upfront
makes the rest of the code considerably simpler.

No functional change.

llvm-svn: 203429

class that might (at some point) need them.

llvm-svn: 203428

members as being te workaround MSVC.

llvm-svn: 203427

synthesize a move constructor. Thus, for any types where move semantics
are important (yea, that's essentially every type...) you must
explicitly define the special members. Do so systematically throughout
the pass manager as the core of the design relies heavily on move
semantics.

This will hopefully fix the build with MSVC 2013. We still don't know
why MSVC 2012 accepted this code, but it almost certainly wasn't doing
the right thing.

I've also added explicit to a few single-argument constructors spotted
in passing.

llvm-svn: 203426

llvm-svn: 203424