llvm-svn: 134820

llvm-svn: 134819

llvm-svn: 134818

With Lit (not bash) in a test, multiple redirects >%t might open(%t, "w") multiple. It can be avoided if latter redirect is >>%t.

It might work even if ">/dev/null" were used.

llvm-svn: 134814

llvm-svn: 134778

Try to move spills as early as possible in their basic block. This can
help eliminate interferences by shortening the live range being
spilled.

This fixes PR10221.

llvm-svn: 134776

llvm-svn: 134760

llvm-svn: 134756

Default 64-bit target features and SSE2 on when a triple specifies x86-64.  Clean up all the other hacks which are now unnecessary.

llvm-svn: 134753

The normal tBX instruction is predicable, so there's no reason the
pseudos for using it as a return shouldn't be. Gives us some nice code-gen
improvements as can be seen by the test changes. In particular, several
tests now have to disable if-conversion because it works too well and defeats
the test.

llvm-svn: 134746

http://llvm.org/bugs/show_bug.cgi?id=10305

llvm-svn: 134744

is to use this for architectures that have a native FMA instruction.

llvm-svn: 134742

RAGreedy::tryAssign will now evict interference from the preferred
register even when another register is free.

To support this, add the EvictionCost struct that counts how many hints
are broken by an eviction. We don't want to break one hint just to
satisfy another.

Rename canEvict to shouldEvict, and add the first bit of eviction policy
that doesn't depend on spill weights: Always make room in the preferred
register as long as the evictees can be split and aren't already
assigned to their preferred register.

Also make the CSR avoidance more accurate. When looking for a cheaper
register it is OK to use a new volatile register. Only CSR aliases that
have never been used before should be avoided.

llvm-svn: 134735

llvm-svn: 134719

We have to do this in DAGBuilder instead of DAGCombiner, because the exact bit is lost after building.

  struct foo { char x[24]; };
  long bar(struct foo *a, struct foo *b) { return a-b; }
is now compiled into
  movl	4(%esp), %eax
  subl	8(%esp), %eax
  sarl	$3, %eax
  imull	$-1431655765, %eax, %eax
instead of
  movl	4(%esp), %eax
  subl	8(%esp), %eax
  movl	$715827883, %ecx
  imull	%ecx
  movl	%edx, %eax
  shrl	$31, %eax
  sarl	$2, %edx
  addl	%eax, %edx
  movl	%edx, %eax

llvm-svn: 134695

Make GVN look through extractvalues for recognised intrinsics. GVN can then CSE ops that match values produced by the intrinsics.

llvm-svn: 134677

llvm-svn: 134667

It was testing a linear scan feature:

  Test if linearscan is unfavoring registers for allocation to allow
  more reuse of reloads from stack slots.

The greedy register allocator doesn't access any stack slots in this
function, so the linear scan feature was not being tested.

llvm-svn: 134666

Fixes PR9602!

llvm-svn: 134665

processor supports it just fine.

Fixes PR9675 and rdar://9740801

llvm-svn: 134664

Fixes PR10149 and rdar://9738585

llvm-svn: 134648

llvm-svn: 134641

Change some ARM subtarget features to be single bit yes/no in order to sink them down to MC layer. Also fix tests.

llvm-svn: 134590

llvm-svn: 134573

llvm-svn: 134530

The promotion code lost any alignment information, when hoisting loads and
stores out of the loop. This lead to incorrect aligned memory accesses. We now
use the largest alignment we can prove to be correct.

llvm-svn: 134520

llvm-svn: 134516

llvm-svn: 134511

push with a small constant produces a 2-byte push.

llvm-svn: 134501

llvm-svn: 134447

Add assembler/disassembler support for non-AVX pclmulqdq. While I'm here, use proper aliases for the pclmullqlqdq and friends. PR10269.

llvm-svn: 134424

Remat during spilling triggers dead code elimination. If a phi-def
becomes unused, that may also cause live ranges to split into separate
connected components.

This type of splitting is different from normal live range splitting. In
particular, there may not be a common original interval.

When the split range is its own original, make sure that the new
siblings are also their own originals. The range being split cannot be
used as an original since it doesn't cover the new siblings.

llvm-svn: 134413

PR10267: Don't combine an equality compare with an AND into an inequality compare when the AND has more than one use.

This can pessimize code, inequalities are generally more expensive.

llvm-svn: 134379

llvm-svn: 134366

makes one of the tests actually mean something (as the string 'add' will
always appear in the output of this file).

llvm-svn: 134358

a file descriptor.

llvm-svn: 134355

desired result based on the comments in the file.

llvm-svn: 134354

llvm-svn: 134353

llvm-svn: 134352

llvm-svn: 134351