llvm-svn: 45416

llvm-svn: 45415

llvm-svn: 45414

llvm-svn: 45413

llvm-svn: 45412

discussion of this change.  Boy are my fingers tired. ;-)

llvm-svn: 45411

discussion of this change.

llvm-svn: 45410

discussion of this change.

llvm-svn: 45409

llvm-svn: 45408

eliminating the llvm.x86.sse2.loadl.pd intrinsic?), one shuffle optzn
may be done (if shufps is better than pinsw, Evan, please review), and
we already know about LICM of simple instructions.

llvm-svn: 45407

llvm-svn: 45406

comparisons with a constant.  This allows us to compile isnan to:

_foo:
	fcmpu cr7, f1, f1
	mfcr r2
	rlwinm r3, r2, 0, 31, 31
	blr 

instead of:

LCPI1_0:					;  float
	.space	4
_foo:
	lis r2, ha16(LCPI1_0)
	lfs f0, lo16(LCPI1_0)(r2)
	fcmpu cr7, f1, f0
	mfcr r2
	rlwinm r3, r2, 0, 31, 31
	blr 

llvm-svn: 45405

llvm-svn: 45404

Disable null pointer folding transforms for non-generic address spaces. This should probably be a target-specific predicate based on address space. That way for targets where this isn't applicable the predicate can be optimized away.

llvm-svn: 45403

llvm-svn: 45402

as:

_bar:
	pushl	%esi
	subl	$8, %esp
	movl	16(%esp), %esi
	call	L_foo$stub
	fstps	(%esi)
	addl	$8, %esp
	popl	%esi
	#FP_REG_KILL
	ret

instead of:

_bar:
	pushl	%esi
	subl	$8, %esp
	movl	16(%esp), %esi
	call	L_foo$stub
	fstpl	(%esi)
	cvtsd2ss	(%esi), %xmm0
	movss	%xmm0, (%esi)
	addl	$8, %esp
	popl	%esi
	#FP_REG_KILL
	ret

llvm-svn: 45401

llvm-svn: 45400

if we are just going to store it back anyway.  This improves things 
like:
double foo();
void bar(double *P) { *P = foo(); }

llvm-svn: 45399

  x = load p
  store x -> p

llvm-svn: 45398

llvm-svn: 45397

llvm-svn: 45396

Enable CodeGen for member expressions based on call expressions returning aggregate types. This enables expressions like 'foo().member.submember'.

llvm-svn: 45395

llvm-svn: 45394

llvm-svn: 45393

have potential side-effects.

llvm-svn: 45392

llvm-svn: 45391

Fix an error in the base/idx accessors for ArraySubscriptExpr's that crops up with vector element access.

llvm-svn: 45390

Test -simplifycfg only.

llvm-svn: 45389

llvm-svn: 45388

llvm-svn: 45387

llvm-svn: 45386

Repair a transform that Chris noticed a bug in.  Thanks to Nicholas for pointing out my stupid mistakes when writing this patch. :-)

llvm-svn: 45384

define i32 @main() {
entry:
	%z = alloca i32		; <i32*> [#uses=2]
	store i32 0, i32* %z
	%tmp = load i32* %z		; <i32> [#uses=1]
	%sub = sub i32 %tmp, 1		; <i32> [#uses=1]
	%cmp = icmp ult i32 %sub, 0		; <i1> [#uses=1]
	%retval = select i1 %cmp, i32 1, i32 0		; <i32> [#uses=1]
	ret i32 %retval
}

into ret 1, instead of ret 0.

Christopher, please investigate.

llvm-svn: 45383

llvm-svn: 45382

llvm-svn: 45381

and OwningPtr instead of constructing only after all of sema is done.  This
has a couple of effects:
1. it fixes memory leaks from all the error cases in sema
2. it simplifies the code significantly.

The cost of this is that the error case now new's and delete's an expr where
it did not before, but we don't care about the perf of the error case.

llvm-svn: 45380

llvm-svn: 45379

llvm-svn: 45378

llvm-svn: 45377

"Here's a tiny patch that lets the clang Xcode project build in any 
location, so llvm doesn't have to be checked out in your home folder."

llvm-svn: 45376