Commit Graph

814 Commits

Author SHA1 Message Date
Chris Lattner
476769498e implement an optimization to codegen c ? 1.0 : 2.0 as load { 2.0, 1.0 } + c*4.
For 2009-03-07-FPConstSelect.ll we now produce:

_f:
	xorl	%eax, %eax
	testl	%edi, %edi
	movl	$4, %ecx
	cmovne	%rax, %rcx
	leaq	LCPI1_0(%rip), %rax
	movss	(%rcx,%rax), %xmm0
	ret

previously we produced:

_f:
	subl	$4, %esp
	cmpl	$0, 8(%esp)
	movss	LCPI1_0, %xmm0
	je	LBB1_2	## entry
LBB1_1:	## entry
	movss	LCPI1_1, %xmm0
LBB1_2:	## entry
	movss	%xmm0, (%esp)
	flds	(%esp)
	addl	$4, %esp
	ret

on PPC the code also improves to:

_f:
	cntlzw r2, r3
	srwi r2, r2, 5
	li r3, lo16(LCPI1_0)
	slwi r2, r2, 2
	addis r3, r3, ha16(LCPI1_0)
	lfsx f1, r3, r2
	blr 

from:

_f:
	li r2, lo16(LCPI1_1)
	cmplwi cr0, r3, 0
	addis r2, r2, ha16(LCPI1_1)
	beq cr0, LBB1_2	; entry
LBB1_1:	; entry
	li r2, lo16(LCPI1_0)
	addis r2, r2, ha16(LCPI1_0)
LBB1_2:	; entry
	lfs f1, 0(r2)
	blr 

This also improves the existing pic-cpool case from:

foo:
	subl	$12, %esp
	call	.Lllvm$1.$piclabel
.Lllvm$1.$piclabel:
	popl	%eax
	addl	$_GLOBAL_OFFSET_TABLE_ + [.-.Lllvm$1.$piclabel], %eax
	cmpl	$0, 16(%esp)
	movsd	.LCPI1_0@GOTOFF(%eax), %xmm0
	je	.LBB1_2	# entry
.LBB1_1:	# entry
	movsd	.LCPI1_1@GOTOFF(%eax), %xmm0
.LBB1_2:	# entry
	movsd	%xmm0, (%esp)
	fldl	(%esp)
	addl	$12, %esp
	ret

to:

foo:
	call	.Lllvm$1.$piclabel
.Lllvm$1.$piclabel:
	popl	%eax
	addl	$_GLOBAL_OFFSET_TABLE_ + [.-.Lllvm$1.$piclabel], %eax
	xorl	%ecx, %ecx
	cmpl	$0, 4(%esp)
	movl	$8, %edx
	cmovne	%ecx, %edx
	fldl	.LCPI1_0@GOTOFF(%eax,%edx)
	ret

This triggers a few dozen times in spec FP 2000.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@66358 91177308-0d34-0410-b5e6-96231b3b80d8
2009-03-08 01:51:30 +00:00
Nate Begeman
cbd88adea6 Fix a problem with DAGCombine on 64b targets where folding
extracts + build_vector into a shuffle would fail, because the
type of the new build_vector would not be legal.  Try harder to
create a legal build_vector type.  Note: this will be totally 
irrelevant once vector_shuffle no longer takes a build_vector for
shuffle mask.

New:
_foo:
	xorps	%xmm0, %xmm0
	xorps	%xmm1, %xmm1
	subps	%xmm1, %xmm1
	mulps	%xmm0, %xmm1
	addps	%xmm0, %xmm1
	movaps	%xmm1, 0

Old:
_foo:
	xorps	%xmm0, %xmm0
	movss	%xmm0, %xmm1
	xorps	%xmm2, %xmm2
	unpcklps	%xmm1, %xmm2
	pshufd	$80, %xmm1, %xmm1
	unpcklps	%xmm1, %xmm2
	pslldq	$16, %xmm2
	pshufd	$57, %xmm2, %xmm1
	subps	%xmm0, %xmm1
	mulps	%xmm0, %xmm1
	addps	%xmm0, %xmm1
	movaps	%xmm1, 0



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@65791 91177308-0d34-0410-b5e6-96231b3b80d8
2009-03-01 23:44:07 +00:00
Evan Cheng
a87008d90b Revert BuildVectorSDNode related patches: 65426, 65427, and 65296.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@65482 91177308-0d34-0410-b5e6-96231b3b80d8
2009-02-25 22:49:59 +00:00
Scott Michel
4214a5531c Introduce the BuildVectorSDNode class that encapsulates the ISD::BUILD_VECTOR
instruction. The class also consolidates the code for detecting constant
splats that's shared across PowerPC and the CellSPU backends (and might be
useful for other backends.) Also introduces SelectionDAG::getBUID_VECTOR() for
generating new BUILD_VECTOR nodes.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@65296 91177308-0d34-0410-b5e6-96231b3b80d8
2009-02-22 23:36:09 +00:00
Dan Gohman
1ba519b726 Fix a bug that David Greene found in the DAGCombiner's logic
that checks whether it's safe to transform a store of a bitcast
value into a store of the original value.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@65201 91177308-0d34-0410-b5e6-96231b3b80d8
2009-02-20 23:29:13 +00:00
Scott Michel
fdc40a0a69 Remove trailing whitespace to reduce later commit patch noise.
(Note: Eventually, commits like this will be handled via a pre-commit hook that
 does this automagically, as well as expand tabs to spaces and look for 80-col
 violations.)


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@64827 91177308-0d34-0410-b5e6-96231b3b80d8
2009-02-17 22:15:04 +00:00
Dale Johannesen
e8d7230f48 Remove more non-DebugLoc getNode variants. Use
getCALLSEQ_{END,START} to permit passing no DebugLoc
there.  UNDEF doesn't logically have DebugLoc; add
getUNDEF to encapsulate this.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63978 91177308-0d34-0410-b5e6-96231b3b80d8
2009-02-06 23:05:02 +00:00
Dale Johannesen
de06470330 Remove more non-DebugLoc versions of getNode.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63969 91177308-0d34-0410-b5e6-96231b3b80d8
2009-02-06 21:50:26 +00:00
Dale Johannesen
ff97d4fe81 Propagation in TargetLowering. Includes passing a DL
into SimplifySetCC which gets called elsewhere.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63583 91177308-0d34-0410-b5e6-96231b3b80d8
2009-02-03 00:47:48 +00:00
Duncan Sands
b0d5cdd52e Fix PR3453 and probably a bunch of other potential
crashes or wrong code with codegen of large integers:
eliminate the legacy getIntegerVTBitMask and
getIntegerVTSignBit methods, which returned their
value as a uint64_t, so couldn't handle huge types.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63494 91177308-0d34-0410-b5e6-96231b3b80d8
2009-02-01 18:06:53 +00:00
Bill Wendling
fc4b677cd5 Forgot some more DebugLoc propagations.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63493 91177308-0d34-0410-b5e6-96231b3b80d8
2009-02-01 11:19:36 +00:00
Duncan Sands
92abc62399 Fix PR3401: when using large integers, the type
returned by getShiftAmountTy may be too small
to hold shift values (it is an i8 on x86-32).
Before and during type legalization, use a large
but legal type for shift amounts: getPointerTy;
afterwards use getShiftAmountTy, fixing up any
shift amounts with a big type during operation
legalization.  Thanks to Dan for writing the
original patch (which I shamelessly pillaged).


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63482 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-31 15:50:11 +00:00
Bill Wendling
9729c5ad34 Don't use DebugLoc::getUnknownLoc(). Default to something hopefully sensible.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63473 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-31 03:12:48 +00:00
Bill Wendling
836ca7dc6e More DebugLoc propagation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63454 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 23:59:18 +00:00
Bill Wendling
c144a57e14 More DebugLoc propagation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63452 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 23:36:47 +00:00
Bill Wendling
c0debad716 More DebugLoc propagation in LOAD etc. methods.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63451 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 23:27:35 +00:00
Bill Wendling
0225a1dcb1 More DebugLoc propagation in floating-point methods.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63446 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 23:15:49 +00:00
Bill Wendling
d34470c4f7 Standardize comments about folding xforms.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63443 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 23:10:18 +00:00
Bill Wendling
7581bfa275 Get rid of the non-DebugLoc-ified getNOT() method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63442 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 23:03:19 +00:00
Bill Wendling
a03e74bf67 Propagate debug loc info for some FP arithmetic methods.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63441 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 22:57:07 +00:00
Bill Wendling
b0162f58c4 Propagate debug loc info for some FP arithmetic methods.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63440 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 22:53:48 +00:00
Bill Wendling
67a6768375 Propagate debug loc info for BIT_CONVERT.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63439 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 22:44:24 +00:00
Bill Wendling
8509c90a20 Propagate debug loc info for more *_EXTEND methods.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63437 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 22:33:24 +00:00
Bill Wendling
683c95764a Propagate debug loc info for ANY_EXTEND.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63436 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 22:27:33 +00:00
Bill Wendling
6ce610f04b Propagate debug loc info for some of the *_EXTEND functions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63434 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 22:23:15 +00:00
Bill Wendling
34584e6aed - Propagate debug loc info for SELECT.
- Added xform for (select X, 1, Y) and (select X, Y, 0), which was commented on,
  but missing.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63428 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 22:02:18 +00:00
Bill Wendling
8810337e36 Propagate debug loc info for Shifts.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63424 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 21:37:17 +00:00
Bill Wendling
317bd709bd Propagate debug loc info for XOR and MatchRotate.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63420 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 21:14:50 +00:00
Bill Wendling
0902564479 Propagate debug loc info for OR. Also clean up some comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63419 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 20:59:34 +00:00
Bill Wendling
6af7618b23 Perform obvious constant arithmetic folding.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63417 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 20:50:00 +00:00
Bill Wendling
2627a881e1 Propagate debug loc info for AND. Also clean up some comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63416 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 20:43:18 +00:00
Bill Wendling
b74c8675b1 Propagate debug loc info in SimplifyBinOpWithSameOpcodeHands.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63411 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 19:25:47 +00:00
Bill Wendling
826d11463b Propagate debug loc info in SimplifyNodeWithTwoResults.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63376 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 03:08:40 +00:00
Bill Wendling
326411dc69 Propagate debug loc info for MULHS.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63375 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 03:00:18 +00:00
Bill Wendling
6d3bf8c4c3 Propagate debug loc info for SREM and UREM.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63374 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 02:57:00 +00:00
Bill Wendling
07d8514049 Propagate debug loc info for UDIV.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63373 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 02:55:25 +00:00
Bill Wendling
944d34bfe8 Propagate debug loc info for SDIV.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63372 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 02:52:17 +00:00
Bill Wendling
73e16b2869 Forgot to propagate debug loc info here.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63371 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 02:49:26 +00:00
Bill Wendling
9c8148ac93 Propagate debug loc info for MUL.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63369 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 02:45:56 +00:00
Bill Wendling
b0702e0b7e Propagate debug loc info in SUB.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63368 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 02:42:10 +00:00
Bill Wendling
14036c00c0 Propagate debug loc info in ADDC and ADDE.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63367 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 02:38:00 +00:00
Bill Wendling
f4eb2269eb Propagate debug loc info in DAG combine's "ADD".
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63366 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 02:31:17 +00:00
Bill Wendling
d69c3141ed - Propagate debug loc info in combineSelectAndUse().
- Modify ReassociateOps so that the resulting SDValue is what the comment claims
  it is.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63365 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 02:23:43 +00:00
Bill Wendling
5c71acfb06 Propagate debug location info for the token factor.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63355 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 01:13:16 +00:00
Bill Wendling
35247c3513 Add DebugLoc propagation to some of the methods in DAG combiner.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63350 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-30 00:45:56 +00:00
Dan Gohman
e5af2d3a22 Make x86's BT instruction matching more thorough, and add some
dagcombines that help it match in several more cases. Add
several more cases to test/CodeGen/X86/bt.ll. This doesn't
yet include matching for BT with an immediate operand, it
just covers more register+register cases.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63266 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-29 01:59:02 +00:00
Dan Gohman
f560ffae1f Make isOperationLegal do what its name suggests, and introduce a
new isOperationLegalOrCustom, which does what isOperationLegal
previously did.

Update a bunch of callers to use isOperationLegalOrCustom
instead of isOperationLegal. In some case it wasn't obvious
which behavior is desired; when in doubt I changed then to
isOperationLegalOrCustom as that preserves their previous
behavior.

This is for the second half of PR3376.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63212 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-28 17:46:25 +00:00
Dan Gohman
ce9bc12c6f Add an assertion to the form of SelectionDAG::getConstant that takes
a uint64_t to verify that the value is in range for the given type,
to help catch accidental overflow. Fix a few places that relied on
getConstant implicitly truncating the value.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@63128 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-27 20:39:34 +00:00
Dan Gohman
e7852d0144 Take the next steps in making SDUse more consistent with LLVM Use, and
tidy up SDUse and related code.
 - Replace the operator= member functions with a set method, like
   LLVM Use has, and variants setInitial and setNode, which take
   care up updating use lists, like LLVM Use's does. This simplifies
   code that calls these functions.
 - getSDValue() is renamed to get(), as in LLVM Use, though most
   places can either use the implicit conversion to SDValue or the
   convenience functions instead.
 - Fix some more node vs. value terminology issues.

Also, eliminate the one remaining use of SDOperandPtr, and
SDOperandPtr itself.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@62995 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-26 04:35:06 +00:00
Dan Gohman
a90c8e690b Fold x-0 to x in unsafe-fp-math mode. This comes up in the
testcase from PR3376, and in fact is sufficient to completely
avoid the problem in that testcase.

There's an underlying problem though; TLI.isOperationLegal
considers Custom to be Legal, which might be ok in some
cases, but that's what DAGCombiner is using in many places
to test if something is legal when LegalOperations is true.
When DAGCombiner is running after legalize, this isn't
sufficient. I'll address this in a separate commit.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@62860 91177308-0d34-0410-b5e6-96231b3b80d8
2009-01-23 19:10:37 +00:00