these alignment amounts to align scalars when we can.  Patch by Scott Michel!

llvm-svn: 33409

that disassembleBuffer will be called even if NDEBUG, but the result will
be ignored.

llvm-svn: 33408

NetBSD also doesn't have pthread_mutexattr_setpshared

llvm-svn: 33406

a small inline function to sign extend a uint64_t value based on its
type's bitwidth. This function is then used in both executeSExtInst and
the various executeICMP_S** functions.

llvm-svn: 33403

llvm-svn: 33401

llvm-svn: 33400

Last check-in was bogus. There is no need to align the stack if the function is a leaf function (and without alloca).

llvm-svn: 33399

llvm-svn: 33398

llvm-svn: 33397

llvm-svn: 33396

This patch fixes test/Integer/2007-01-17-TruncSext.ll

llvm-svn: 33394

llvm-svn: 33390

llvm-svn: 33389

llvm-svn: 33388

We not align the final stack slot but instead let the target do so in emitPrologue(). Each target can make adjustments to the stack frame and re-align the stack as it deem appropriate. Do not align it twice which can end up wasting stack space.

llvm-svn: 33387

GetRegForReload() now keeps track which registers have been considered and rejected during its quest to find a suitable reload register. This avoids an infinite loop in case like this:
  t1 := op t2, t3
  t2 <- assigned r0 for use by the reload but ended up reuse r1
  t3 <- assigned r1 for use by the reload but ended up reuse r0
  t1 <- desires r1
        sees r1 is taken by t2, tries t2's reload register r0
        sees r0 is taken by t3, tries t3's reload register r1
        sees r1 is taken by t2, tries t2's reload register r0 ...

llvm-svn: 33382

llvm-svn: 33375

needed for output to be generated. On Linux these are both global vars and
macro definitions so we have to special case Linux.

llvm-svn: 33374

llvm-svn: 33373

llvm-svn: 33372

don't allow the transform if V and the pointer's element type are different
width integer types.

llvm-svn: 33371

This is the final patch for this PR. It implements some minor cleanup
in the use of IntegerType, to wit:
1. Type::getIntegerTypeMask -> IntegerType::getBitMask
2. Type::Int*Ty changed to IntegerType* from Type*
3. ConstantInt::getType() returns IntegerType* now, not Type*

This also fixes PR1120.

Patch by Sheng Zhou.

llvm-svn: 33370

llvm-svn: 33368

llvm-svn: 33367

llvm-svn: 33366

the .zerofill directive emission to not be darwin-specific.

llvm-svn: 33365

llvm-svn: 33364

llvm-svn: 33363

delimits the boundaries of jump tables.  This lets the linker's dead code
stripping optimization do a better job.

llvm-svn: 33362

(shl (add x, c1), c2) -> (add (shl x, c2), c1<<c2)
Replace it with:
(add (shl (add x, c1), c2), ) -> (add (add (shl x, c2), c1<<c2), )

This fixes test/CodeGen/ARM/smul.ll

llvm-svn: 33361

llvm-svn: 33359

llvm-svn: 33358

llvm-svn: 33353

llvm-svn: 33352

llvm-svn: 33347

Implement reading and writing of the Module's data layout string.

llvm-svn: 33346

- In x86-64 mode, symbols with external linkage (not just symbols which are
  defined externally) requires GOT indirect reference.
- Stylistic code clean up.

llvm-svn: 33345

infinitely loops.

llvm-svn: 33343

transform. Change some variable names so it is clear what is source and
what is dest of the cast. Also, add an assert to ensure that the integer
to integer case is asserting if the bitwidths are different. This prevents
illegal casts from being formed and catches bitwidth bugs sooner.

llvm-svn: 33337

llvm-svn: 33336