llvm-svn: 36618

argument space for call sites in the function immediately on entry to the
current function. This eliminates the need for add/sub sp brackets around call
sites. However, this is not always a good idea. If the "call frame" is large and
the target load / store instructions have small immediate field to encode sp
offset, this can cause poor codegen. In the worst case, this can make it
impossible to scavenge a register if the reserved spill slot is pushed too far
apart from sp / fp.

llvm-svn: 36607

llvm-svn: 36586

llvm-svn: 36585

llvm-svn: 36583

around islands and jump tables).

llvm-svn: 36573

llvm-svn: 36572

should be taught to deal with protected symbols.

llvm-svn: 36565

llvm-svn: 36519

llvm-svn: 36518

llvm-svn: 36514

llvm-svn: 36513

llvm-svn: 36512

llvm-svn: 36506

llvm-svn: 36505

llvm-svn: 36504

llvm-svn: 36503

llvm-svn: 36502

llvm-svn: 36495

llvm-svn: 36494

llvm-svn: 36483

llvm-svn: 36456

llvm-svn: 36452

clobbered by a call.

llvm-svn: 36448

when promoted to 64-bits

llvm-svn: 36442

llvm-svn: 36438

will follow.

llvm-svn: 36435

llvm-svn: 36430

llvm-svn: 36425

llvm-svn: 36423

llvm-svn: 36410

llvm-svn: 36407

llvm-svn: 36406

llvm-svn: 36405

llvm-svn: 36404

        vector_shuffle v1, v2, <2, 6, 3, 7>

I.e.

         vector_shuffle v, undef, <2, 2, 3, 3>

MMX only has a shuffle for v4i16 vectors. It needs to use the unpackh for
this type of operation.

llvm-svn: 36403

BBOffsets and BBSizes when adjusting conditional branches.

llvm-svn: 36372

llvm-svn: 36371

llvm-svn: 36370

llvm-svn: 36369