llvm-svn: 12625

llvm-svn: 12624

llvm-svn: 12623

llvm-svn: 12622

llvm-svn: 12621

This also implements some new features for the indvars pass, including
linear function test replacement, exit value substitution, and it works with
a much more general class of induction variables and loops.

llvm-svn: 12620

llvm-svn: 12619

llvm-svn: 12618

This actually causes us to turn code like:

  return C ? A : B;

into a select instruction.

llvm-svn: 12617

llvm-svn: 12616

llvm-svn: 12615

Move lowerselect pass to come after preselection.  Move machine
code construction and stack slots pass to come right before instruction
selection. This is to help fix perlbmk.

Update comments.

Make the sequence of passes in addPassesToJITCompile look more like
the sequence of passes in addPassesToEmitAssembly, including support
for -print-machineinstrs.

llvm-svn: 12614

llvm-svn: 12613

llvm-svn: 12612

llvm-svn: 12611

llvm-svn: 12610

Patch contributed by Reid Spencer

llvm-svn: 12609

llvm-svn: 12608

llvm-svn: 12607

llvm-svn: 12606

  1. Each time the loop extractor extracted a loop, we would leak a module.
  2. When we extracted a loop, we didn't add the new function to the list of
     miscompiled functions.  Thus if the bug was in a loop nest and we
     extracted it, we could actually *LOSE THE BUG*, which is very bad.

With these patches, bugpoint has successfully found a bug for me in a function
with several nested loops, and cut it down to just one of them. :) :)

llvm-svn: 12605

hacking on programs with two functions that have the same name.

llvm-svn: 12604

around

llvm-svn: 12603

Address PR305: LLVM tools will happily spew bytecode onto your terminal

llvm-svn: 12602

llvm-svn: 12601

llvm-svn: 12600

llvm-svn: 12598

llvm-svn: 12597

llvm-svn: 12596

llvm-svn: 12595

llvm-svn: 12593

Testcase: LoopSimplify/2004-04-01-IncorrectDomUpdate.ll

llvm-svn: 12592

llvm-svn: 12591

llvm-svn: 12590

llvm-svn: 12589

Implement a small optimization.  In test/Regression/CodeGen/X86/select.ll,
we now generate this for foldSel3:

foldSel3:
        mov %AL, BYTE PTR [%ESP + 4]
        fld DWORD PTR [%ESP + 8]
        fld DWORD PTR [%ESP + 12]
        mov %EAX, DWORD PTR [%ESP + 16]
        mov %ECX, DWORD PTR [%ESP + 20]
        cmp %EAX, %ECX
        fxch %ST(1)
        fcmovae %ST(0), %ST(1)
***     fstp %ST(1)
        ret

Instead of:

foldSel3:
        mov %AL, BYTE PTR [%ESP + 4]
        fld DWORD PTR [%ESP + 8]
        fld DWORD PTR [%ESP + 12]
        mov %EAX, DWORD PTR [%ESP + 16]
        mov %ECX, DWORD PTR [%ESP + 20]
        cmp %EAX, %ECX
        fxch %ST(1)
        fcmovae %ST(0), %ST(1)
***     fxch %ST(1)
***     fstp %ST(0)
        ret

In practice, this only effects code size: performance should be basically
unaffected.

llvm-svn: 12588

llvm-svn: 12587

llvm-svn: 12586

llvm-svn: 12585

add note about select

llvm-svn: 12584