If "optimize for size" attribute is set then block non-trivial loop unswitches but allow trivial loop unswitches.

llvm-svn: 63670

llvm-svn: 63659

renaming it to ConvertScalar_ExtractValue

llvm-svn: 63658

no functionality change.

llvm-svn: 63652

functionality change.

llvm-svn: 63651

aggregate values.  loads are not yet handled (coming
soon to an sroa near you).

llvm-svn: 63649

accessed at least once as a vector.  This prevents it from
compiling the example in not-a-vector into:

define double @test(double %A, double %B) {
	%tmp4 = insertelement <7 x double> undef, double %A, i32 0
	%tmp = insertelement <7 x double> %tmp4, double %B, i32 4
	%tmp2 = extractelement <7 x double> %tmp, i32 4
	ret double %tmp2
}

instead, producing the integer code.  Producing vectors when they
aren't otherwise in the program is dangerous because a lot of other
code treats them carefully and doesn't want to break them down.
OTOH, many things want to break down tasty i448's.

llvm-svn: 63638

llvm-svn: 63631

llvm-svn: 63620

crazy cases like:

struct f {  int A, B, C, D, E, F; };
short test4() {
  struct f A;
  A.A = 1;
  memset(&A.B, 2, 12);
  return A.C;
}

llvm-svn: 63596

With the new world order, it can handle cases where the first
store into the alloca is an element of the vector, instead of
requiring the first analyzed store to have the vector type 
itself.  This allows us to un-xfail 
test/CodeGen/X86/vec_ins_extract.ll.

llvm-svn: 63590

llvm-svn: 63544

tests.  Thanks for the beautiful reduced testcase Duncan!

llvm-svn: 63529

and remove trailing whitespace.  No functionality
change.

llvm-svn: 63511

llvm-svn: 63510

llvm-svn: 63500

turn icmp eq a+x, b+x into icmp eq a, b if a+x or b+x has other uses. This
may have been increasing register pressure leading to the bzip2 slowdown.

llvm-svn: 63487

improvements to the EvaluateInDifferentType code.  This code works 
by just inserted a bunch of new code and then seeing if it is 
useful.  Instcombine is not allowed to do this: it can only insert
new code if it is useful, and only when it is converging to a more
canonical fixed point.  Now that we iterate when DCE makes progress,
this causes an infinite loop when the code ends up not being used.

llvm-svn: 63483

simplifydemandedbits to simplify instructions with *multiple
uses* in contexts where it can get away with it.  This allows
it to simplify the code in multi-use-or.ll into a single 'add 
double'.

This change is particularly interesting because it will cover
up for some common codegen bugs with large integers created due
to the recent SROA patch.  When working on fixing those bugs,
this should be disabled.

llvm-svn: 63481

llvm-svn: 63480

Now, if it detects that "V" is the same as some other value, 
SimplifyDemandedBits returns the new value instead of RAUW'ing it immediately.
This has two benefits:
1) simpler code in the recursive SimplifyDemandedBits routine.
2) it allows future fun stuff in instcombine where an operation has multiple
   uses and can be simplified in one context, but not all.

#2 isn't implemented yet, this patch should have no functionality change.

llvm-svn: 63479

llvm-svn: 63477

not doing so prevents it from properly iterating and prevents it
from deleting the entire body of dce-iterate.ll

llvm-svn: 63476

be able to handle *ANY* alloca that is poked by loads and stores of 
bitcasts and GEPs with constant offsets.  Before the code had a number
of annoying limitations and caused it to miss cases such as storing into
holes in structs and complex casts (as in bitfield-sroa) where we had
unions of bitfields etc.  This also handles a number of important cases
that are exposed due to the ABI lowering stuff we do to pass stuff by
value.

One case that is pretty great is that we compile 
2006-11-07-InvalidArrayPromote.ll into:

define i32 @func(<4 x float> %v0, <4 x float> %v1) nounwind {
	%tmp10 = call <4 x i32> @llvm.x86.sse2.cvttps2dq(<4 x float> %v1)
	%tmp105 = bitcast <4 x i32> %tmp10 to i128
	%tmp1056 = zext i128 %tmp105 to i256	
	%tmp.upgrd.43 = lshr i256 %tmp1056, 96
	%tmp.upgrd.44 = trunc i256 %tmp.upgrd.43 to i32	
	ret i32 %tmp.upgrd.44
}

which turns into:

_func:
	subl	$28, %esp
	cvttps2dq	%xmm1, %xmm0
	movaps	%xmm0, (%esp)
	movl	12(%esp), %eax
	addl	$28, %esp
	ret

Which is pretty good code all things considering :).

One effect of this is that SROA will start generating arbitrary bitwidth 
integers that are a multiple of 8 bits.  In the case above, we got a 
256 bit integer, but the codegen guys assure me that it can handle the 
simple and/or/shift/zext stuff that we're doing on these operations.

This addresses rdar://6532315

llvm-svn: 63469

after the others.

llvm-svn: 63227

llvm-svn: 63198

has a different number of elements than the output.

llvm-svn: 62998

llvm-svn: 62985

handling the flaw inherent in that assumption.  :)

llvm-svn: 62984

llvm-svn: 62983

llvm-svn: 62982

Also it was an empty struct, not a void after all.

llvm-svn: 62920

Thus we need to check whether the struct is empty before trying to index into
it. This fixes PR3381.

llvm-svn: 62918

handling the case in Transforms/InstCombine/cast-store-gep.ll, which
is a heavily reduced testcase from Clang on x86-64.

llvm-svn: 62904

There is now a direct way from value-use-iterator to incoming block in PHINode's API.
This way we avoid the iterator->index->iterator trip, and especially the costly
getOperandNo() invocation. Additionally there is now an assertion that the iterator
really refers to one of the PHI's Uses.

llvm-svn: 62869

compatibility with cygwin.  Patch by Jay Foad!

llvm-svn: 62695

Besides APFloat, this involved removing code
from two places that thought they knew the
result of frem(0., x) but were wrong.

llvm-svn: 62645

llvm-svn: 62535

llvm-svn: 62529

llvm-svn: 62470