packed word integer (v8i16), and 64-bit packed doubleword integer (v2i32).

llvm-svn: 26294

llvm-svn: 26292

them to get ressurected, in which case, deleting the undead nodes is
unfriendly.

llvm-svn: 26291

i64 before this code, so zero_ext doesn't work.

llvm-svn: 26290

llvm-svn: 26289

advantage of fisttpll.

llvm-svn: 26288

llvm-svn: 26287

that the shift may be more easily folded into other operations.

llvm-svn: 26286

llvm-svn: 26285

where we were doing exactly that which was causing failures on x86 and
alpha.

llvm-svn: 26284

llvm-svn: 26283

usesCustomDAGSchedInserter.

llvm-svn: 26282

llvm-svn: 26281

Thanks to nate for pointing this out :)

llvm-svn: 26280

llvm-svn: 26279

llvm-svn: 26278

the right loop.

llvm-svn: 26277

llvm-svn: 26275

llvm-svn: 26274

llvm-svn: 26273

llvm-svn: 26272

llvm-svn: 26271

exactly the API we wanted to call into.  This fixes the crash on crafty last
night.

llvm-svn: 26269

llvm-svn: 26268

llvm-svn: 26267

llvm-svn: 26266

llvm-svn: 26265

llvm-svn: 26264

Make more use of the new mask helpers in valuetypes.h
Combine (sra (srl x, c1), c1) -> sext_inreg if legal

llvm-svn: 26263

llvm-svn: 26262

generate illegal nodes.

llvm-svn: 26261

llvm-svn: 26260

llvm-svn: 26259

llvm-svn: 26258

llvm-svn: 26257

opcodes on ppc.

llvm-svn: 26256

and SUBE nodes that actually expose what's going on and allow for
significant simplifications in the targets.

llvm-svn: 26255

conversions to __floatdidf instead of __floatdisf on targets that support
f32 but not i64 (e.g. sparc).

llvm-svn: 26254

llvm-svn: 26253

llvm-svn: 26252