Commit Graph

144 Commits

Author SHA1 Message Date
Chris Lattner
d847bc20b8 Rework passing parent pointers into complexpatterns, I forgot
that complex patterns are matched after the entire pattern has
a structural match, therefore the NodeStack isn't in a useful
state when the actual call to the matcher happens.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114489 91177308-0d34-0410-b5e6-96231b3b80d8
2010-09-21 22:00:25 +00:00
Chris Lattner
78d3af47f5 just like they can opt into getting the root of the pattern being
matched, allow ComplexPatterns to opt into getting the parent node
of the operand being matched.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114472 91177308-0d34-0410-b5e6-96231b3b80d8
2010-09-21 20:37:12 +00:00
Chris Lattner
beac75da37 implement rdar://6653118 - fastisel should fold loads where possible.
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.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@113102 91177308-0d34-0410-b5e6-96231b3b80d8
2010-09-05 02:18:34 +00:00
Dan Gohman
84023e0fbe Reapply bottom-up fast-isel, with several fixes for x86-32:
- Check getBytesToPopOnReturn().
 - Eschew ST0 and ST1 for return values.
 - Fix the PIC base register initialization so that it doesn't ever
   fail to end up the top of the entry block.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@108039 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-10 09:00:22 +00:00
Bob Wilson
02266e29f9 --- Reverse-merging r107947 into '.':
U    utils/TableGen/FastISelEmitter.cpp
--- Reverse-merging r107943 into '.':
U    test/CodeGen/X86/fast-isel.ll
U    test/CodeGen/X86/fast-isel-loads.ll
U    include/llvm/Target/TargetLowering.h
U    include/llvm/Support/PassNameParser.h
U    include/llvm/CodeGen/FunctionLoweringInfo.h
U    include/llvm/CodeGen/CallingConvLower.h
U    include/llvm/CodeGen/FastISel.h
U    include/llvm/CodeGen/SelectionDAGISel.h
U    lib/CodeGen/LLVMTargetMachine.cpp
U    lib/CodeGen/CallingConvLower.cpp
U    lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp
U    lib/CodeGen/SelectionDAG/FunctionLoweringInfo.cpp
U    lib/CodeGen/SelectionDAG/FastISel.cpp
U    lib/CodeGen/SelectionDAG/SelectionDAGISel.cpp
U    lib/CodeGen/SelectionDAG/ScheduleDAGSDNodes.cpp
U    lib/CodeGen/SelectionDAG/InstrEmitter.cpp
U    lib/CodeGen/SelectionDAG/TargetLowering.cpp
U    lib/Target/XCore/XCoreISelLowering.cpp
U    lib/Target/XCore/XCoreISelLowering.h
U    lib/Target/X86/X86ISelLowering.cpp
U    lib/Target/X86/X86FastISel.cpp
U    lib/Target/X86/X86ISelLowering.h


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107987 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-09 16:37:18 +00:00
Dan Gohman
bf87e24917 Re-apply bottom-up fast-isel, with fixes. Be very careful to avoid emitting
a DBG_VALUE after a terminator, or emitting any instructions before an EH_LABEL.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107943 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-09 00:39:23 +00:00
Dan Gohman
f595141525 Revert 107840 107839 107813 107804 107800 107797 107791.
Debug info intrinsics win for now.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107850 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-08 01:00:56 +00:00
Dan Gohman
eabaed26c3 Give FunctionLoweringInfo an MBB member, avoiding the need to pass it
around everywhere, and also give it an InsertPt member, to enable isel
to operate at an arbitrary position within a block, rather than just
appending to a block.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107791 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-07 16:47:08 +00:00
Dan Gohman
c9af33c685 CanLowerReturn doesn't need a SelectionDAG; it just needs an LLVMContext.
SelectBasicBlock doesn't needs its BasicBlock argument.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@107712 91177308-0d34-0410-b5e6-96231b3b80d8
2010-07-06 22:19:37 +00:00
Dan Gohman
4e39e9da0f Reapply r106634, now that the bug it exposed is fixed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106746 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-24 14:30:44 +00:00
Daniel Dunbar
cbe762b5d1 Revert r106263, "Fold the ShrinkDemandedOps pass into the regular DAGCombiner pass,"... it was causing both 'file' (with clang) and 176.gcc (with llvm-gcc) to be miscompiled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106634 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-23 17:09:26 +00:00
Dan Gohman
8a7f7426ee Fold the ShrinkDemandedOps pass into the regular DAGCombiner pass,
which is faster, simpler, and less surprising.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@106263 91177308-0d34-0410-b5e6-96231b3b80d8
2010-06-18 01:05:21 +00:00
Dan Gohman
f81eca0ab9 Move HandlePHINodesInSuccessorBlocks functions out of SelectionDAGISel
and into SelectionDAGBuilder and FastISel.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@102123 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-22 20:46:50 +00:00
Dan Gohman
f0757b0edc Add more const qualifiers on TargetMachine and friends.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101977 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-21 01:34:56 +00:00
Dan Gohman
a9a3321938 Reapply the removal of SelectionDAGISel's BB, with a fix for the case
where multiple blocks are emitted; functions which do this need to return
the new BB so that their callers can stay current.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101843 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-20 00:29:35 +00:00
Dan Gohman
68b7960c1a Revert 101825, which is causing trouble.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101832 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-19 23:34:15 +00:00
Dan Gohman
35bb85b7ee Eliminate SelectionDAGISel's "current block" member. Just pass it as
an argument to things that need it.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101825 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-19 22:51:14 +00:00
Dan Gohman
d858e90f03 Use const qualifiers with TargetLowering. This eliminates several
const_casts, and it reinforces the design of the Target classes being
immutable.

SelectionDAGISel::IsLegalToFold is now a static member function, because
PIC16 uses it in an unconventional way. There is more room for API
cleanup here.

And PIC16's AsmPrinter no longer uses TargetLowering.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101635 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-17 15:26:15 +00:00
Dan Gohman
46510a73e9 Add const qualifiers to CodeGen's use of LLVM IR constructs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101334 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-15 01:51:59 +00:00
Dan Gohman
6465265ae1 Delete unneeeded arguments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101276 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-14 20:17:22 +00:00
Dan Gohman
6a732b5a66 Delete unused arguments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101275 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-14 20:05:00 +00:00
Dan Gohman
2520864773 Factor out EH landing pad code into a separate function, and constify
a bunch of stuff to support it.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101273 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-14 19:53:31 +00:00
Dan Gohman
c21a2dc8df Delete an unused function.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101263 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-14 17:22:02 +00:00
Dan Gohman
441516fb3d Trim #includes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@101255 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-14 16:54:39 +00:00
Chris Lattner
de6e783b24 remove some redundant MMI arguments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@100417 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-05 06:10:13 +00:00
Chris Lattner
d850ac79b5 fastisel doesn't need DwarfWriter, remove some tendricles.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@100381 91177308-0d34-0410-b5e6-96231b3b80d8
2010-04-05 02:19:28 +00:00
Chris Lattner
aa4e3391c8 add an optimized form of OPC_EmitMergeInputChains for the 1, 0 and
1, 1 cases which are by-far the most frequent.  This shrinks the X86
isel table from 77014 -> 74657 bytes.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@99740 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-28 05:50:16 +00:00
Chris Lattner
7561d48095 change the LabelSDNode to be EHLabelSDNode and make it hold
an MCSymbol.  Make the EH_LABEL MachineInstr hold its label
with an MCSymbol instead of ID.  Fix a bug in MMI.cpp which
would return labels named "Label4" instead of "label4".


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@98463 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-14 02:33:54 +00:00
Chris Lattner
2c4afd1d8e inline CannotYetSelectIntrinsic into CannotYetSelect and simplify.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97690 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-04 00:21:16 +00:00
Chris Lattner
f1b7c7d476 speed up scope node processing: if the first element of a scope
entry we're about to process is obviously going to fail, don't
bother pushing a scope only to have it immediately be popped.
This avoids a lot of scope stack traffic in common cases.

Unfortunately, this requires duplicating some of the predicate
dispatch.  To avoid duplicating the actual logic I pulled each
predicate out to its own static function which gets used in
both places.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97651 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-03 07:31:15 +00:00
Chris Lattner
cfe2eab744 introduce a new SwitchTypeMatcher node (which is analogous to
SwitchOpcodeMatcher) and have DAGISelMatcherOpt form it.  This
speeds up selection, particularly for X86 which has lots of 
variants of instructions with only type differences.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97645 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-03 06:28:15 +00:00
Chris Lattner
6b7f39c895 remove 300 lines of code that is now dead in the MSP430 backend
now that isel handles chains more aggressively.  This also
allows us to make isLegalToFold non-virtual.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97597 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-02 22:30:08 +00:00
Chris Lattner
d1b7382983 Fix some issues in WalkChainUsers dealing with
CopyToReg/CopyFromReg/INLINEASM.  These are annoying because
they have the same opcode before an after isel.  Fix this by
setting their NodeID to -1 to indicate that they are selected,
just like what automatically happens when selecting things that
end up being machine nodes.

With that done, give IsLegalToFold a new flag that causes it to
ignore chains.  This lets the HandleMergeInputChains routine be
the one place that validates chains after a match is successful,
enabling the new hotness in chain processing.  This smarter
chain processing eliminates the need for "PreprocessRMW" in the
X86 and MSP430 backends and enables MSP to start matching it's
multiple mem operand instructions more aggressively.

I currently #if out the dead code in the X86 backend and MSP 
backend, I'll remove it for real in a follow-on patch.

The testcase changes are:
  test/CodeGen/X86/sse3.ll: we generate better code
  test/CodeGen/X86/store_op_load_fold2.ll: PreprocessRMW was 
      miscompiling this before, we now generate correct code
      Convert it to filecheck while I'm at it.
  test/CodeGen/MSP430/Inst16mm.ll: Add a testcase for mem/mem
      folding to make anton happy. :)




git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97596 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-02 22:20:06 +00:00
Chris Lattner
82dd3d38d9 Fix the xfail I added a couple of patches back. The issue
was that we weren't properly handling the case when interior
nodes of a matched pattern become dead after updating chain
and flag uses.  Now we handle this explicitly in 
UpdateChainsAndFlags.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97561 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-02 07:50:03 +00:00
Chris Lattner
3ff1e4db78 factor node morphing out to its own helper method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97558 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-02 06:55:04 +00:00
Chris Lattner
7c306da505 Sink InstructionSelect() out of each target into SDISel, and rename it
DoInstructionSelection.  Inline "SelectRoot" into it from DAGISelHeader.
Sink some other stuff out of DAGISelHeader into SDISel.

Eliminate the various 'Indent' stuff from various targets, which dates
to when isel was recursive.

 17 files changed, 114 insertions(+), 430 deletions(-)



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97555 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-02 06:34:30 +00:00
Chris Lattner
17b4b172eb move some code out of DAGISelHeader up to SelectionDAGISel.h where it
is shared by all targets.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97553 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-02 06:04:12 +00:00
Chris Lattner
c6d7ad3c7d Rewrite chain handling validation and input TokenFactor handling
stuff now that we don't care about emulating the old broken 
behavior of the old isel.  This eliminates the 
'CheckChainCompatible' check (along with IsChainCompatible) which
did an incorrect and inefficient scan *up* the chain nodes which
happened as the pattern was being formed and does the validation
at the end in HandleMergeInputChains when it forms a structural 
pattern.  This scans "down" the graph, which means that it is
quickly bounded by nodes already selected.  This also handles
token factors that get "trapped" in the dag.

Removing the CheckChainCompatible nodes also shrinks the 
generated tables by about 6K for X86 (down to 83K).

There are two pieces remaining before I can nuke PreprocessRMW:
1. I xfailed a test because we're now producing worse code in a 
   case that has nothing to do with the change: it turns out that
   our use of MorphNodeTo will leave dead nodes in the graph
   which (depending on how the graph is walked) end up causing
   bogus uses of chains and blocking matches.  This is really 
   bad for other reasons, so I'll fix this in a follow-up patch.

2. CheckFoldableChainNode needs to be improved to handle the TF.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97539 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-02 02:22:10 +00:00
Chris Lattner
7390eebd49 Accelerate isel dispatch for tables that start with a top-level
OPC_SwitchOpcode to use a table lookup instead of having to go
through the interpreter for this.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97469 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-01 18:47:11 +00:00
Chris Lattner
fa342faef9 eliminate the CheckMultiOpcodeMatcher code and have each
ComplexPattern at the root be generated multiple times, once
for each opcode they are part of.  This encourages factoring
because the opcode checks get treated just like everything
else in the matcher.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97439 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-01 07:17:40 +00:00
Chris Lattner
eb66921adb add a new OPC_SwitchOpcode which is semantically equivalent
to a scope where every child starts with a CheckOpcode, but
executes more efficiently.  Enhance DAGISelMatcherOpt to 
form it.

This also fixes a bug in CheckOpcode: apparently the SDNodeInfo
objects are not pointer comparable, we have to compare the
enum name.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97438 91177308-0d34-0410-b5e6-96231b3b80d8
2010-03-01 06:59:22 +00:00
Chris Lattner
2a49d57d2d hoist the new isel interpreter out of DAGISelHeader.h
(which gets #included into the middle of each 
target's DAGISel class) into a .cpp file where it is
only compiled once.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97425 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-28 22:37:22 +00:00
Chris Lattner
5310654dc5 change a few opcodes to use VBRs instead of embedding
immediate sizes into the opcode.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97423 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-28 22:14:32 +00:00
Chris Lattner
aa6d708028 move all the opcodes out of DAGISelHeader.h (which gets
#included into the middle of each isel class) into a real 
header.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97421 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-28 21:58:42 +00:00
Dan Gohman
fb76fe0929 Fix various doxygen warnings.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96779 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-22 04:10:52 +00:00
Chris Lattner
beff6a3fd0 add a hook so that the new isel can run SDNodeXForms.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96714 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-21 03:15:11 +00:00
Chris Lattner
bd12fe8b69 make the new isel's interpreter loop call the generated
CheckComplexPattern function.  Though it is logically const,
I don't have the fortitude to clean up all the targets now,
and it not being const doesn't block anything.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96426 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-17 00:41:34 +00:00
Chris Lattner
050a03d0f3 generate code for node and pattern predicates. Note that this won't
build if enabled, it will fail with constness issues. I'll resolve 
these next.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96336 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-16 07:21:10 +00:00
Evan Cheng
014bf215c3 Split SelectionDAGISel::IsLegalAndProfitableToFold to
IsLegalToFold and IsProfitableToFold. The generic version of the later simply checks whether the folding candidate has a single use.

This allows the target isel routines more flexibility in deciding whether folding makes sense. The specific case we are interested in is folding constant pool loads with multiple uses.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@96255 91177308-0d34-0410-b5e6-96231b3b80d8
2010-02-15 19:41:07 +00:00
Evan Cheng
d40d03e1bd Teach dag combine to fold the following transformation more aggressively:
(OP (trunc x), (trunc y)) -> (trunc (OP x, y))

Unfortunately this simple change causes dag combine to infinite looping. The problem is the shrink demanded ops optimization tend to canonicalize expressions in the opposite manner. That is badness. This patch disable those optimizations in dag combine but instead it is done as a late pass in sdisel.

This also exposes some deficiencies in dag combine and x86 setcc / brcond lowering. Teach them to look pass ISD::TRUNCATE in various places.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@92849 91177308-0d34-0410-b5e6-96231b3b80d8
2010-01-06 19:38:29 +00:00