DAG combiner can turn a VAND V, <-1, 0, -1, -1>, i.e. vector clear elements,
into a vector shuffle with a zero vector. It only does so when TLI tells it
the xform is profitable.

llvm-svn: 27874

isShuffleMaskLegal, used to determine if it makes sense to turn a
"vector clear" (e.g. pand V, <0, -1, 0, -1> to a shuffle of the vector and
a zero vector.

llvm-svn: 27873

llvm-svn: 27872

llvm-svn: 27871

llvm-svn: 27870

llvm-svn: 27869

CodeGen/PowerPC/2006-04-19-vmaddfp-crash.ll

llvm-svn: 27868

llvm-svn: 27867

llvm-svn: 27865

llvm-svn: 27863

llvm-svn: 27861

llvm-svn: 27860

llvm-svn: 27859

llvm-svn: 27858

llvm-svn: 27857

llvm-svn: 27856

llvm-svn: 27855

llvm-svn: 27854

llvm-svn: 27853

llvm-svn: 27852

llvm-svn: 27851

llvm-svn: 27850

but i64 is not. If possible, change a i64 op to a f64 (e.g. load, constant)
and then cast it back.

llvm-svn: 27849

llvm-svn: 27848

llvm-svn: 27847

llvm-svn: 27846

  instructions.
- Fixed a commute vector_shuff bug.

llvm-svn: 27845

llvm-svn: 27844

llvm-svn: 27843

- Added more movhlps and movlhps patterns.

llvm-svn: 27842

llvm-svn: 27841

llvm-svn: 27840

llvm-svn: 27839

llvm-svn: 27838

command line so that relinked .o files can be built universal.

llvm-svn: 27837

llvm-svn: 27836

- Increase cost (complexity) of patterns which match mov{h|l}ps ops. These
  are preferred over shufps in most cases.

llvm-svn: 27835

llvm-svn: 27834

llvm-svn: 27833

llvm-svn: 27832