llvm-svn: 107615

llvm-svn: 107612

Infer alignments of fixed frame objects when they are constructed. This ensures remat'ed loads from fixed slots have the right alignments.

llvm-svn: 107591

llvm-svn: 107581

PrologEpilog code, and use it to determine whether
the asm forces stack alignment or not.  gcc consistently
does not do this for GCC-style asms; Apple gcc inconsistently
sometimes does it for asm blocks.  There is no
convenient place to put a bit in either the SDNode or
the MachineInstr form, so I've added an extra operand
to each; unlovely, but it does allow for expansion for
more bits, should we need it.  PR 5125.  Some
existing testcases are affected.
The operand lists of the SDNode and MachineInstr forms
are indexed with awesome mnemonics, like "2"; I may
fix this someday, but not now.  I'm not making it any
worse.  If anyone is inspired I think you can find all
the right places from this patch.

llvm-svn: 107506

llvm-svn: 107490

llvm-svn: 107489

llvm-svn: 107451

llvm-svn: 107393

doing the work manually.

llvm-svn: 107384

from a Value to a Type, because it doesn't actually care about
the Value.

llvm-svn: 107383

available in a register. This is pretty primitive, but it reduces the
number of instructions in common testcases by 4%.

llvm-svn: 107380

llvm-svn: 107377

SelectionDAGBuilder::getValue into a helper function, with fixes to
use DenseMaps safely.

llvm-svn: 107371

value if the search fails.

llvm-svn: 107368

llvm-svn: 107321

this could break something (but doesn't seem to).

llvm-svn: 107295

llvm-svn: 107282

llvm-svn: 107279

of getPhysicalRegisterRegClass with it.

If we want to make a copy (or estimate its cost), it is better to use the
smallest class as more efficient operations might be possible.

llvm-svn: 107140

llvm-svn: 107127

can't be changed arbitrarily by the DAGCombiner without checking if it is
running after legalization.

llvm-svn: 107097

have to be registers, per gcc documentation.  This affects
the logic for determining what "g" should lower to.  PR 7393.
A couple of existing testcases are affected.

llvm-svn: 107079

This produces terrible but correct code.

llvm-svn: 106952

llvm-svn: 106901

for an "i" constraint should get lowered; PR 6309.  While
this argument was passed around a lot, this is the only
place it was used, so it goes away from a lot of other
places.

llvm-svn: 106893

is not used.  Spotted by gcc-4.6.

llvm-svn: 106854

llvm-svn: 106836

llvm-svn: 106828

llvm-svn: 106827

original SDNode. This is badness. Also, this function allows one SDNode to point
multiple flags to another SDNode. Badness as well.

llvm-svn: 106793

which trivially fold away.

llvm-svn: 106765

as many tests as possible.

llvm-svn: 106749

when the condition is constant. This optimization shouldn't be
necessary, because codegen shouldn't be able to find dead control
paths that the IR-level optimizer can't find. And it's undesirable,
because it encourages bugpoint to leave "br i1 false" branches
in its output. And it wasn't updating the CFG.

I updated all the tests I could, but some tests are too reduced
and I wasn't able to meaningfully preserve them.

llvm-svn: 106748

llvm-svn: 106746

case where the bit mask has exactly one bit.

llvm-svn: 106716

into the same node, but with different non-memory operands, we need to replace
the memory operands after it's finished morphing.

llvm-svn: 106643

Revert r106263, "Fold the ShrinkDemandedOps pass into the regular DAGCombiner pass,"... it was causing both 'file' (with clang) and 176.gcc (with llvm-gcc) to be miscompiled.

llvm-svn: 106634

atomic intrinsics, either because the use locking instructions for the
atomics, or because they perform the locking directly. Add support in the
DAG combiner to fold away the fences.

llvm-svn: 106630

SmallVector, and other SmallVector simplifications.

llvm-svn: 106452