LiveIntervals already adds <imp-def> operands for super-registers when a subreg
def defines the whole register. Thus, it is not necessary to do it again when
rewriting.

In fact, the super-register imp-defs caused miscompilations because the late
scheduler couldn't see that the super-register was read.

We still add super-reg <imp-use,kill> operands when rewriting virtuals to
physicals.

llvm-svn: 113299

register must be one of the destination registers for the load. Otherwise,
the tLDM instruction will write-back to the base register, which isn't what's
desired (otherwise, we'd have a t2LDM_UPD instead).

rdar://8394087

llvm-svn: 113297

llvm-svn: 113293

llvm-svn: 113289

llvm-svn: 113287

Factor out some x86 vector shuffle rewriting and add comments about the direction the shuffle lowering is heading to

llvm-svn: 113286

llvm-svn: 113285

the provided cleanup function is never actually called.

llvm-svn: 113284

switch to using a ManagedStatic for the global PassRegistry instead of a
ManagedCleanup, and fix a destruction ordering bug this exposed.

llvm-svn: 113283

llvm-svn: 113281

llvm-svn: 113275

llvm-svn: 113267

llvm-svn: 113264

llvm-svn: 113262

llvm-svn: 113261

turning (fptrunc (sqrt (fpext x))) -> (sqrtf x)  is great, but we have
to delete the original sqrt as well.  Not doing so causes us to do 
two sqrt's when building with -fmath-errno (the default on linux).

llvm-svn: 113260

llvm-svn: 113259

llvm-svn: 113257

llvm-svn: 113255

and exposing internal details.

llvm-svn: 113252

llvm-svn: 113250

llvm-svn: 113249

decouple MMX check from regular splat checks. Some refactoring is coming, and MMX should be left alone to be easily removed after moving to intrinsics

llvm-svn: 113247

llvm-svn: 113242

GetMainExecutable yet.

llvm-svn: 113240

llvm-svn: 113239

llvm-svn: 113237

llvm-svn: 113236

llvm-svn: 113235

Enable palignr intrinsic.
These may need adjustment for a new VT in due course.

llvm-svn: 113233

llvm-svn: 113232

llvm-svn: 113224

llvm-svn: 113213

Add a test case to check handling of dbg-declare  during hybrid mode where we begin using fast-isel but switch back to DAG building at some point.

llvm-svn: 113210

llvm-svn: 113209

llvm-svn: 113208

llvm-svn: 113200

llvm-svn: 113198

llvm-svn: 113197

teach LazyValueInfo to use them.

llvm-svn: 113196