llvm-svn: 132936

Cleaning up the Python script interpreter:  Use the
embedded_interpreter.py file rather than keeping it
all in a string and compiling the string (easier to maintain,
easier to read, remove redundancy).

llvm-svn: 132935

The logic for reserving R4 for use as a scratch needs to match that for
actually using it. Also, it's not necessary for immediate <=508, so adjust
the value checked.

llvm-svn: 132934

Aliased flag options should be directed to stdout, not stderr to be consistent. Patch by Julien Lerouge.

llvm-svn: 132931

More prompt-timing cleanups:  Make multi-line expressions
use the asynchronous stream mechanism rather than writing
directly to the Debugger's output & error streams.

llvm-svn: 132930

<rdar://problem/9600408>

llvm-svn: 132929

arch, but actually compile a file with that arch.

llvm-svn: 132928

might overflow.  Re-typing the alloca to a larger type (e.g. double)
hoists a shift into the alloca, potentially exposing overflow in the
expression.  rdar://problem/9265821

llvm-svn: 132926

llvm-svn: 132925

llvm-svn: 132924

only if the number of packed elements is a power of two.
Bug found in Duncan's testcase.

llvm-svn: 132923

llvm-svn: 132921

reason to allow the user to control these semantics through a flag.

llvm-svn: 132919

llvm-svn: 132918

llvm-svn: 132917

llvm-svn: 132916

The backend already knew this trick.

llvm-svn: 132915

llvm-svn: 132914

llvm-svn: 132913

llvm-svn: 132912

llvm-svn: 132911

llvm-svn: 132910

llvm-svn: 132909

llvm-svn: 132908

intrinsics. In fact, we'll optimize a bitcast to that when possible. Detect it
when looking for the lifetime intrinsics.

No test case, noticed by inspection.

llvm-svn: 132906

llvm-svn: 132905

- Move the diagnostic to the case statement instead of at the end of the switch
- Add a fix-it hint as to how to fix the compilation error

llvm-svn: 132903

handle memcpy and memmove. Spotted by Nico.

llvm-svn: 132902

llvm-svn: 132901

In particular, don't spill dirty registers only to satisfy a hint. It is
not worth it.

The attached test case provides an example where the fast allocator
would spill a register when other registers are available.

llvm-svn: 132900

llvm-svn: 132899

having.

llvm-svn: 132898

InstCombine: Shrink ((zext X) & C1) == C2 to fold away the cast if the "zext" and the "and" have one use.

llvm-svn: 132897

llvm-svn: 132896

Only 163 days late!

llvm-svn: 132895

llvm-svn: 132894

types such as i33 were rounded to i32. Originated from Duncan's testcase.

llvm-svn: 132893

Instead of scalarizing, and doing an element-by-element truncat, use vector
truncate.
Add support for scalarization of vectors:  i8 -> <1 x i1> (from Duncan's
testcase).

llvm-svn: 132892

element type is found.
This fix addresses some of the tests in Duncan's testcase (forthcoming).

llvm-svn: 132891

  struct {
    typedef int A = 0;
  };

According to the C++11 standard, this is not ill-formed, but does not have any ascribed meaning. We can't reasonably accept it, so treat it as ill-formed.

Also switch C++ from an incorrect 'fields can only be initialized in constructors' diagnostic for this case to C's 'illegal initializer (only variables can be initialized)'

llvm-svn: 132890