llvm-svn: 48064

#include <xmmintrin.h>
__m128i doload64(short x) {return _mm_set_epi16(0,0,0,0,0,0,0,1);}

into:
	movl	$1, %eax
	movd	%eax, %xmm0
	ret

instead of a constant pool load.

llvm-svn: 48063

llvm-svn: 48055

llvm-svn: 48054

llvm-svn: 47948

llvm-svn: 47939

llvm-svn: 47828

llvm-svn: 47109

llvm-svn: 47051

Add some notes to the README.

llvm-svn: 46949

llvm-svn: 46413

llvm-svn: 46405

eliminating the llvm.x86.sse2.loadl.pd intrinsic?), one shuffle optzn
may be done (if shufps is better than pinsw, Evan, please review), and
we already know about LICM of simple instructions.

llvm-svn: 45407

llvm-svn: 45280

llvm-svn: 43444

llvm-svn: 42549

llvm-svn: 42548

llvm-svn: 42542

llvm-svn: 42345

llvm-svn: 41359

llvm-svn: 41103

llvm-svn: 39772

llvm-svn: 38507

llvm-svn: 35439

llvm-svn: 35266

llvm-svn: 34670

llvm-svn: 32591

llvm-svn: 32584

llvm-svn: 32561

llvm-svn: 32424

llvm-svn: 32229

llvm-svn: 31650

llvm-svn: 31649

llvm-svn: 31038

llvm-svn: 30717

llvm-svn: 30245

llvm-svn: 29091

llvm-svn: 28787

llvm-svn: 28579

llvm-svn: 28400