- Used to mark fake instructions which don't correspond to an actual machine
   instruction (or are duplicates of a real instruction). This is to be used for
   "special cases" in the .td files, which should be ignored by things like the
   assembler and disassembler. We still need a good solution to handle pervasive
   duplication, like with the Int_ instructions.

 - Set the bit on fake "mov 0" style instructions, which allows turning an
   assembler matcher warning into a hard error.

 - -2 FIXMEs.

llvm-svn: 78731

No functional change.

llvm-svn: 78608

llvm-svn: 77978

llvm-svn: 77956

llvm-svn: 77944

Fixes PR4669

llvm-svn: 77940

llvm-svn: 76713

llvm-svn: 75142

instructions, which implies that there is an explicit memory operand.  There is
(however) no explicit memory operand; although this is a store, the only memory
operand is implicit, indicated by DS:EDI.  This causes the table-generation code
for the disassembler to report an error."

Patch by Sean Callanan!

llvm-svn: 73989

nodes for vectors with an i16 element type.  Add an optimization for 
building a vector which is all zeros/undef except for the bottom 
element, where the bottom element is an i8 or i16.

llvm-svn: 72988

llvm-svn: 72985

llvm-svn: 72817

llvm-svn: 72808

the Intel manual (screenshot) says it should be 0b11110110 (f6).  The existing
encoding causes a disassembly conflict with MMX_PAVGBrm, which really should be
0f e0."

Patch by Sean Callanan!

llvm-svn: 72508

PR2957

ISD::VECTOR_SHUFFLE now stores an array of integers representing the shuffle
mask internal to the node, rather than taking a BUILD_VECTOR of ConstantSDNodes
as the shuffle mask.  A value of -1 represents UNDEF.

In addition to eliminating the creation of illegal BUILD_VECTORS just to 
represent shuffle masks, we are better about canonicalizing the shuffle mask,
resulting in substantially better code for some classes of shuffles.

llvm-svn: 70225

llvm-svn: 69967

ISD::VECTOR_SHUFFLE now stores an array of integers representing the shuffle
mask internal to the node, rather than taking a BUILD_VECTOR of ConstantSDNodes
as the shuffle mask.  A value of -1 represents UNDEF.

In addition to eliminating the creation of illegal BUILD_VECTORS just to 
represent shuffle masks, we are better about canonicalizing the shuffle mask,
resulting in substantially better code for some classes of shuffles.

A clean up of x86 shuffle code, and some canonicalizing in DAGCombiner is next.

llvm-svn: 69952

llvm-svn: 65313

Added support for TRUNC v8i16 to v8i8 for X86 (MMX)

llvm-svn: 60916

llvm-svn: 60499

llvm-svn: 60487

llvm-svn: 58752

SSE2 registers as well as the MMX registers.

llvm-svn: 55436

llvm-svn: 55318

instructions on having SSE2.

llvm-svn: 55317

Is there way to avoid explicit target check?

llvm-svn: 55238

llvm-svn: 54026

load,store,call,return,bitcast.  This is enough to
make call and return work.

llvm-svn: 52691

llvm-svn: 51667

Note, some of the code will be moved into target independent part of DAG combiner in a subsequent patch.

llvm-svn: 50918

Handle vector move / load which zero the destination register top bits (i.e. movd, movq, movss (addr), movsd (addr)) with X86 specific dag combine.

llvm-svn: 50838

Add separate intrinsics for MMX / SSE shifts with i32 integer operands. This allow us to simplify the horribly complicated matching code.

llvm-svn: 50601

llvm-svn: 50291

llvm-svn: 50289

llvm-svn: 50278

llvm-svn: 50053

to 64-bit GPR registers on x86-64.

llvm-svn: 49757

llvm-svn: 48627

llvm-svn: 48569

Replace all target specific implicit def instructions with a target independent one: TargetInstrInfo::IMPLICIT_DEF.

llvm-svn: 48380