XMM registers.  There are many known deficiencies and fixmes, which will be
addressed ASAP.  The major benefit of this work is that it will allow the
LLVM register allocator to allocate FP registers across basic blocks.

The x86 backend will still default to x87 style FP.  To enable this work,
you must pass -enable-sse-scalar-fp and either -sse2 or -sse3 to llc.

An example before and after would be for:
double foo(double *P) { double Sum = 0; int i; for (i = 0; i < 1000; ++i)
                        Sum += P[i]; return Sum; }

The inner loop looks like the following:
x87:
.LBB_foo_1:     # no_exit
        fldl (%esp)
        faddl (%eax,%ecx,8)
        fstpl (%esp)
        incl %ecx
        cmpl $1000, %ecx
        #FP_REG_KILL
        jne .LBB_foo_1  # no_exit

SSE2:
        addsd (%eax,%ecx,8), %xmm0
        incl %ecx
        cmpl $1000, %ecx
        #FP_REG_KILL
        jne .LBB_foo_1  # no_exit

llvm-svn: 22340

1. Pass Value*'s into lowering methods so that the proper pointers can be
   added to load/stores from the valist
2. Intrinsics that return void should only return a token chain, not a token
   chain/retval pair.
3. Rename LowerVAArgNext -> LowerVAArg, because VANext is long gone.
4. Now that we have Value*'s available in the lowering methods, pass them
   into any load/stores from the valist that are emitted

llvm-svn: 22339

llvm-svn: 22336

llvm-svn: 22335

llvm-svn: 22332

llvm-svn: 22331

llvm-svn: 22330

llvm-svn: 22327

is at least overloading the right virtual methods.  The implementations
are currently wrong though.  This fixes Ptrdist/bc, but not other programs
(e.g. siod).

llvm-svn: 22326

by Aaron Gray, cleaned up by me.

llvm-svn: 22324

llvm-svn: 22323

llvm-svn: 22322

llvm-svn: 22321

thing as cygwin most of the time, and printing our alignments in log2
rather than number of bytes.

llvm-svn: 22316

llvm-svn: 22315

tracking the instructions causing loads and stores provides more information than just the pointer being loaded or stored

llvm-svn: 22311

llvm-svn: 22309

llvm-svn: 22308

llvm-svn: 22307

llvm-svn: 22306

rather than relying on the assembler.  Only a few more pseudo instructions
left.  Also merge load code paths.

llvm-svn: 22305

llvm-svn: 22303

is to use RA, which assumes the called function uses RA for the register
holding the return address when it issues a ret.

llvm-svn: 22301

working.  The instruction selector changes will hopefully be coming later
this week once they are debugged.  This is necessary to support the darwin
x86 FP model, and is recommended by intel as the replacement for x87.  As
a bonus, the register allocator knows how to deal with these registers
across basic blocks, unliky the FP stackifier.  This leads to significantly
better codegen in several cases.

llvm-svn: 22300

llvm-svn: 22299

llvm-svn: 22298

llvm-svn: 22297

Namely, output the rellocation flags explicitly when loading constants.
Added benifit: save a load when loading from the constant pool.

llvm-svn: 22296

llvm-svn: 22295

llvm-svn: 22294

who said we had to use the return address in the return address register.  Might save a move in many cases

llvm-svn: 22293

currently use: llc t.bc --filetype=obj

This will produce a t.o file which is dumpable with readelf.  Currently
the file produced is empty, but the scaffolding to do more is now in place.

llvm-svn: 22292

llvm-svn: 22289

should be able to handle this case.  The code is there, so let's see
if it works.

llvm-svn: 22288

interface.

llvm-svn: 22282

llvm-svn: 22276

llvm-svn: 22275

llvm-svn: 22274

See last commit for LangRef, this implements it on all targets.

llvm-svn: 22273

of memory instructions in the assembly, to allow later linking
of traces with LLVM Value*s.

llvm-svn: 22271