llvm-svn: 105061

implementation that is correct for most targets. Tablegen will override where
needed.

Add MachineOperand::subst{Virt,Phys}Reg methods that correctly handle existing
subreg indices when sustituting registers.

llvm-svn: 104985

llvm-svn: 104953

This will help reduce the amount of casting required on 64-bit targets.

llvm-svn: 104911

optimization level.

This only really affects llc for now because both the llvm-gcc and clang front
ends override the default register allocator. I intend to remove that code later.

llvm-svn: 104904

an alloca() or an llvm.stackrestore(). rdar://8031573

llvm-svn: 104900

llvm-svn: 104897

llvm-svn: 104884

llvm-svn: 104869

inlined function's arguments need a label to mark the start point because they are not directly attached to current function.

llvm-svn: 104848

llvm-svn: 104841

llvm-svn: 104785

llvm-svn: 104749

-filetype=obj.

llvm-svn: 104747

There is no need to force an line number entry (using previous location) for a temp label at unknown location.

llvm-svn: 104740

usual suspects that could "return twice".

llvm-svn: 104737

ISD::. No functional change.

llvm-svn: 104734

llvm-svn: 104732

more than just the stack slot coloring algorithm.

llvm-svn: 104722

Identify instructions, that needs a label to mark debug info entity, in advance. This simplifies beginScope().

llvm-svn: 104720

implementing pop with a linear search for a "best" element. The priority
queue was a neat idea, but in practice the comparison functions depend
on dynamic information.

llvm-svn: 104718

llvm-svn: 104716

llvm-svn: 104706

llvm-svn: 104705

llvm-svn: 104664

base class, since all the implementations are the same.

llvm-svn: 104659

llvm-svn: 104657

machine code generation. That's a good idea, so I made it so.

llvm-svn: 104655

This is used to track variable information.

llvm-svn: 104649

llvm-svn: 104646

llvm-svn: 104645

as well.

llvm-svn: 104642

If you have a setjmp/longjmp situation, it's possible for stack slot coloring to
reuse a stack slot before it's really dead. For instance, if we have something
like this:

1:  y = g;
    x = sigsetjmp(env, 0);
    switch (x) {
    case 1:
      /* ... */
      goto run;  
    case 0:
  run:
      do_run(); /* marked as "no return" */
      break;
    case 3:
      if (...) {
        /* ... */
        goto run;
      }
      /* ... */
      break;
    }

2:  g = y;

"y" may be put onto the stack, so the expression "g = y" is relying upon the
fact that the stack slot containing "y" isn't modified between (1) and (2). But
it can be, because of the "no return" calls in there. A longjmp might come back
with 3, modify the stack slot, and then go to case 0. And it's perfectly
acceptable to reuse the stack slot there because there's no CFG flow from case 3
to (2).

The fix is to disable certain optimizations in these situations. Ideally, we'd
disable them for all "returns twice" functions. But we don't support that
attribute. Check for "setjmp" and "sigsetjmp" instead.

llvm-svn: 104640

and testcases accordingly.

llvm-svn: 104635

llvm-svn: 104628

Mon Ping provided; unfortunately bugpoint failed to
reduce it, but I think it's important to have a test for
this in the suite.  8023512.

llvm-svn: 104624

(oye, a word which should be better known to people writing tree
traversals, means grandchild.)

llvm-svn: 104619

llvm-svn: 104574

llvm-svn: 104572

llvm-svn: 104560