This pass:

1. Splits TargetMachine into TargetMachine (generic targets, can be implemented
any way, like the CBE) and LLVMTargetMachine (subclass of TM that is used by
things using libcodegen and other support).
2. Instead of having each target fully populate the passmgr for file or JIT
   output, move all this to common code, and give targets hooks they can
   implement.
3. Commonalize the target population stuff between file emission and JIT
   emission.
4. All (native code) codegen stuff now happens in a FunctionPassManager, which
   paves the way for "fast -O0" stuff in the CFE later, and now LLC could
   lazily stream .bc files from disk to use less memory.
5. There are now many fewer #includes and the targets don't depend on the
   scalar xforms or libanalysis anymore (but codegen does).
6. Changing common code generator pass ordering stuff no longer requires
   touching all targets.
7. The JIT now has the option of "-fast" codegen or normal optimized codegen,
   which is now orthogonal to the fact that JIT'ing is being done.

llvm-svn: 30081

llvm-svn: 30080

llvm-svn: 30078

llvm-svn: 30070

llvm-svn: 30067

llvm-svn: 30056

Suppose the TokenFactor can reach the Op:

       [Load chain]
           ^
           |
         [Load]
         ^    ^
         |    |
        /      \-
       /         |
      /          [Op]
     /          ^ ^
     |        ..  |
     |       /    |
   [TokenFactor]  |
       ^          |
       |          |
        \        /
         \      /
         [Store]

If we move the Load below the TokenFactor, we would have created a cycle in
the DAG.

llvm-svn: 30040

llvm-svn: 30039

llvm-svn: 30038

llvm-svn: 29965

llvm-svn: 29962

or vector result. X86 does not have load/mod/store variants of those
instructions.

llvm-svn: 29957

- Also disable isel load folding if -fast.

llvm-svn: 29956

handling.)

llvm-svn: 29954

llvm-svn: 29953

llvm-svn: 29952

there are clearly some encoding bugs lurking in there somewhere.

llvm-svn: 29949

llvm-svn: 29946

Add an optional pass to preprocess the DAG before x86 isel to allow selecting more load/mod/store instructions.

llvm-svn: 29943

Close out this long standing bug by removing the remaining overloaded
virtual functions in LLVM. The -Woverloaded-virtual option is now turned on.

llvm-svn: 29934

llvm-svn: 29911

SDOperand arguments. Use the variants which take an array and number instead.

llvm-svn: 29907

ppc target, not the itanium target, when run on an itanium machine.
This should fix the CodeGen/PowerPC regtest failures on itanium.

llvm-svn: 29903

llvm-svn: 29901

llvm-svn: 29898

llvm-svn: 29895

llvm-svn: 29894

llvm-svn: 29891

llvm-svn: 29889

asm printer will print it as offset*4. e.g. bne cr0, $+8.

The PPC code emitter was expecting the offset to be number of instructions, not
number of bytes. This fixes a whole bunch of JIT failures.

llvm-svn: 29885

llvm-svn: 29880

store LR in an arbitrary stack slot
add support for writing varargs functions

llvm-svn: 29876

_swap_16:
        slwi r2, r3, 24
        rlwimi r2, r3, 8, 8, 15
        srwi r3, r2, 16
        blr

now.

llvm-svn: 29864

llvm-svn: 29863

implement a movcond instruction

llvm-svn: 29857

llvm-svn: 29856

llvm-svn: 29854

does emit linkable .o files in very simple cases.

llvm-svn: 29850

llvm-svn: 29814

llvm-svn: 29802