in the duplicated block instead of duplicating them.  

Duplicating them into the end of the loop and the preheader 
means that we got a phi node in the header of the loop, 
which prevented LICM from hoisting them.  GVN would
usually come around later and merge the duplicated 
instructions so we'd get reasonable output... except that
anything dependent on the shoulda-been-hoisted value can't
be hoisted.  In PR5319 (which this fixes), a memory value
didn't get promoted.

llvm-svn: 113134

Loop::hasLoopInvariantOperands method. Remove
a useless and confusing Loop::isLoopInvariant(Instruction)
method, which didn't do what you thought it did.

No functionality change.

llvm-svn: 113133

pattern, so there is no need to define a matching function.

llvm-svn: 113122

llvm-svn: 113119

llvm-svn: 113117

llvm-svn: 113116

llvm-svn: 113115

llvm-svn: 113114

llvm-svn: 113113

llvm-svn: 113109

llvm-svn: 113108

llvm-svn: 113106

This reduces malloc traffic (yay!) and removes MergeFunctionsEqualityInfo.

llvm-svn: 113105

strong functions first to make sure they're the canonical definitions and then
do a second pass looking only for weak functions.

llvm-svn: 113104

Since mem2reg isn't run at -O0, we get a ton of reloads from the stack,
for example, before, this code:

int foo(int x, int y, int z) {
  return x+y+z;
}

used to compile into:

_foo:                                   ## @foo
	subq	$12, %rsp
	movl	%edi, 8(%rsp)
	movl	%esi, 4(%rsp)
	movl	%edx, (%rsp)
	movl	8(%rsp), %edx
	movl	4(%rsp), %esi
	addl	%edx, %esi
	movl	(%rsp), %edx
	addl	%esi, %edx
	movl	%edx, %eax
	addq	$12, %rsp
	ret

Now we produce:

_foo:                                   ## @foo
	subq	$12, %rsp
	movl	%edi, 8(%rsp)
	movl	%esi, 4(%rsp)
	movl	%edx, (%rsp)
	movl	8(%rsp), %edx
	addl	4(%rsp), %edx    ## Folded load
	addl	(%rsp), %edx     ## Folded load
	movl	%edx, %eax
	addq	$12, %rsp
	ret

Fewer instructions and less register use = faster compiles.

llvm-svn: 113102

Clobber ranges are no longer used when joining physical registers.
Instead, all aliases are checked for interference.

llvm-svn: 113084

llvm-svn: 113077

llvm-svn: 113073

not SelectAddr

llvm-svn: 113072

llvm-svn: 113059

llvm-svn: 113058

into an inner loop, as the new loop iteration may differ substantially.
This fixes PR8078.

llvm-svn: 113057

llvm-svn: 113056

llvm-svn: 113055

llvm-svn: 113050

llvm-svn: 113048

llvm-svn: 113047

llvm-svn: 113045

llvm-svn: 113044

llvm-svn: 113043

location is being re-stored to the memory location.  We would get
a dangling pointer from the SSAUpdate data structure and miss a 
use.  This fixes PR8068

llvm-svn: 113042

llvm-svn: 113035

llvm-svn: 113034

checking each standalone condition and decide whether emit target
specific nodes or remove the condition if it's already matched before.

llvm-svn: 113031

llvm-svn: 113025

invertible. ScalarEvolution's folding routines don't always succeed
in canonicalizing equal expressions to a single canonical form, and
this can cause these asserts to fail, even though there's no actual
correctness problem. This fixes PR8066.

llvm-svn: 113021

"Use target specific nodes instead of relying in unpckl and
unpckh pattern fragments during isel time. Also place a
depth limit in getShuffleScalarElt.

llvm-svn: 113020

overload UserInInstr. Explicitly check Allocatable. The early exit in the
condition will mean the performance impact of the extra test should be
minimal.

llvm-svn: 113016

Bruno, please review.

llvm-svn: 113014

llvm-svn: 113009