Commit Graph

225 Commits

Author SHA1 Message Date
Mon P Wang
3efcd4a65c Added interface to allow clients to create a MemIntrinsicNode for
target intrinsics that touches memory


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@58548 91177308-0d34-0410-b5e6-96231b3b80d8
2008-11-01 20:24:53 +00:00
Mon P Wang
0c39719bfc Add initial support for vector widening. Logic is set to widen for X86.
One will only see an effect if legalizetype is not active.  Will move
support to LegalizeType soon.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@58426 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-30 08:01:45 +00:00
Duncan Sands
9a2c1d3c46 Teach getTypeToTransformTo to return something
sensible for vectors being scalarized.  Note
that this method can't return anything very
sensible when splitting non-power-of-two vectors.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57839 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-20 16:24:25 +00:00
Dan Gohman
6520e20e4f Teach DAGCombine to fold constant offsets into GlobalAddress nodes,
and add a TargetLowering hook for it to use to determine when this
is legal (i.e. not in PIC mode, etc.)

This allows instruction selection to emit folded constant offsets
in more cases, such as the included testcase, eliminating the need
for explicit arithmetic instructions.

This eliminates the need for the C++ code in X86ISelDAGToDAG.cpp
that attempted to achieve the same effect, but wasn't as effective.

Also, fix handling of offsets in GlobalAddressSDNodes in several
places, including changing GlobalAddressSDNode's offset from
int to int64_t.

The Mips, Alpha, Sparc, and CellSPU targets appear to be
unaware of GlobalAddress offsets currently, so set the hook to
false on those targets.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57748 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-18 02:06:02 +00:00
Chris Lattner
6bdcda3d3e Keep track of *which* input constraint matches an output
constraint.  Reject asms where an output has multiple
input constraints tied to it.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57687 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-17 16:47:46 +00:00
Chris Lattner
58f15c482a add an assert so that PR2356 explodes instead of running off an
array.  Improve some minor comments, refactor some helpers in
AsmOperandInfo.  No functionality change for valid code.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57686 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-17 16:21:11 +00:00
Evan Cheng
7f04268176 - Add target lowering hooks that specify which setcc conditions are illegal,
i.e. conditions that cannot be checked with a single instruction. For example,
SETONE and SETUEQ on x86.
- Teach legalizer to implement *illegal* setcc as a and / or of a number of
legal setcc nodes. For now, only implement FP conditions. e.g. SETONE is
implemented as SETO & SETNE, SETUEQ is SETUO | SETEQ.
- Move x86 target over.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57542 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-15 02:05:31 +00:00
Dan Gohman
dd5b58ad7b FastISel support for exception-handling constructs.
- Move the EH landing-pad code and adjust it so that it works
   with FastISel as well as with SDISel.
 - Add FastISel support for @llvm.eh.exception and
   @llvm.eh.selector.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57539 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-14 23:54:11 +00:00
Evan Cheng
0329466b6b Rename LoadX to LoadExt.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57526 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-14 21:26:46 +00:00
Bill Wendling
6158d8492c Implement the -fno-builtin option in the front-end, not in the back-end.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56900 91177308-0d34-0410-b5e6-96231b3b80d8
2008-10-01 00:59:58 +00:00
Bill Wendling
6f287b22d2 Add the new `-no-builtin' flag. This flag is meant to mimic the GCC
`-fno-builtin' flag. Currently, it's used to replace "memset" with "_bzero"
instead of "__bzero" on Darwin10+. This arguably violates the meaning of this
flag, but is currently sufficient. The meaning of this flag should become more
specific over time.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56885 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-30 21:22:07 +00:00
Dale Johannesen
86098bd6a6 Add "inreg" field to CallSDNode (doesn't increase
its size).  Adjust various lowering functions to
pass this info through from CallInst.  Use it to
implement sseregparm returns on X86.  Remove
X86_ssecall calling convention.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56677 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-26 19:31:26 +00:00
Evan Cheng
da43bcf624 Properly handle 'm' inline asm constraints. If a GV is being selected for the addressing mode, it requires the same logic for PIC relative addressing, etc.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56526 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-24 00:05:32 +00:00
Dan Gohman
d57dd5f4e6 Arrange for FastISel code to have access to the MachineModuleInfo
object. This will be needed to support debug info.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56508 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-23 21:53:34 +00:00
Dan Gohman
095cc29f32 Define CallSDNode, an SDNode subclass for use with ISD::CALL.
Currently it just holds the calling convention and flags
for isVarArgs and isTailCall.

And it has several utility methods, which eliminate magic
5+2*i and similar index computations in several places.

CallSDNodes are not CSE'd. Teach UpdateNodeOperands to handle
nodes that are not CSE'd gracefully.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56183 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-13 01:54:27 +00:00
Dan Gohman
0586d91bb3 Add X86FastISel support for static allocas, and refences
to static allocas. As part of this change, refactor the
address mode code for laods and stores.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56066 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-10 20:11:02 +00:00
Dan Gohman
3df24e667f Create HandlePHINodesInSuccessorBlocksFast, a version of
HandlePHINodesInSuccessorBlocks that works FastISel-style. This
allows PHI nodes to be updated correctly while using FastISel.

This also involves some code reorganization; ValueMap and
MBBMap are now members of the FastISel class, so they needn't
be passed around explicitly anymore. Also, SelectInstructions
is changed to SelectInstruction, and only does one instruction
at a time.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55746 91177308-0d34-0410-b5e6-96231b3b80d8
2008-09-03 23:12:08 +00:00
Gabor Greif
92362680c1 fix some 80-col violations
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55565 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-30 10:09:02 +00:00
Gabor Greif
ba36cb5242 erect abstraction boundaries for accessing SDValue members, rename Val -> Node to reflect semantics
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55504 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-28 21:40:38 +00:00
Dale Johannesen
e00a8a2a2e Split the ATOMIC NodeType's to include the size, e.g.
ATOMIC_LOAD_ADD_{8,16,32,64} instead of ATOMIC_LOAD_ADD.
Increased the Hardcoded Constant OpActionsCapacity to match.
Large but boring; no functional change.

This is to support partial-word atomics on ppc; i8 is
not a valid type there, so by the time we get to lowering, the
ATOMIC_LOAD nodes looks the same whether the type was i8 or i32.
The information can be added to the AtomicSDNode, but that is the
largest SDNode; I don't fully understand the SDNode allocation,
but it is sensitive to the largest node size, so increasing
that must be bad.  This is the alternative.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55457 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-28 02:44:49 +00:00
Dan Gohman
bb466331e7 Simplify FastISel's constructor argument list, make the FastISel
class hold a MachineRegisterInfo member, and make the
MachineBasicBlock be passed in to SelectInstructions rather
than the FastISel constructor.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55076 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-20 21:05:57 +00:00
Dan Gohman
d4641a254c Add a TargetLowering hook for creating a FastISel object.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@55009 91177308-0d34-0410-b5e6-96231b3b80d8
2008-08-19 21:26:29 +00:00
Dan Gohman
475871a144 Rename SDOperand to SDValue.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@54128 91177308-0d34-0410-b5e6-96231b3b80d8
2008-07-27 21:46:04 +00:00
Duncan Sands
126d90770b Rather than having a different custom legalization
hook for each way in which a result type can be
legalized (promotion, expansion, softening etc),
just use one: ReplaceNodeResults, which returns
a node with exactly the same result types as the
node passed to it, but presumably with a bunch of
custom code behind the scenes.  No change if the
new LegalizeTypes infrastructure is not turned on.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@53137 91177308-0d34-0410-b5e6-96231b3b80d8
2008-07-04 11:47:58 +00:00
Dan Gohman
87060f55f2 Update comments to new-style syntax.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52925 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-30 21:00:56 +00:00
Dan Gohman
9665c2ae79 Reorder the fields in TargetLowering to require less padding.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52919 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-30 20:36:26 +00:00
Dan Gohman
3d3c3e0cfa Change bools to 1-bit bitfields to shrink ArgListEntry slightly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52918 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-30 20:33:57 +00:00
Dan Gohman
a44b674a42 Replace some std::vectors that showed up in heap profiling with
SmallVectors. Change the signature of TargetLowering::LowerArguments
to avoid returning a vector by value, and update the two targets
which still use this directly, Sparc and IA64, accordingly.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52917 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-30 20:31:15 +00:00
Duncan Sands
d4b9c17fb7 Disable some DAG combiner optimizations that may be
wrong for volatile loads and stores.  In fact this
is almost all of them!  There are three types of
problems: (1) it is wrong to change the width of
a volatile memory access.  These may be used to
do memory mapped i/o, in which case a load can have
an effect even if the result is not used.  Consider
loading an i32 but only using the lower 8 bits.  It
is wrong to change this into a load of an i8, because
you are no longer tickling the other three bytes.  It
is also unwise to make a load/store wider.  For
example, changing an i16 load into an i32 load is
wrong no matter how aligned things are, since the
fact of loading an additional 2 bytes can have
i/o side-effects.  (2) it is wrong to change the
number of volatile load/stores: they may be counted
by the hardware.  (3) it is wrong to change a volatile
load/store that requires one memory access into one
that requires several.  For example on x86-32, you
can store a double in one processor operation, but to
store an i64 requires two (two i32 stores).  In a
multi-threaded program you may want to bitcast an i64
to a double and store as a double because that will
occur atomically, and be indivisible to other threads.
So it would be wrong to convert the store-of-double
into a store of an i64, because this will become two
i32 stores - no longer atomic.  My policy here is
to say that the number of processor operations for
an illegal operation is undefined.  So it is alright
to change a store of an i64 (requires at least two
stores; but could be validly lowered to memcpy for
example) into a store of double (one processor op).
In short, if the new store is legal and has the same
size then I say that the transform is ok.  It would
also be possible to say that transforms are always
ok if before they were illegal, whether after they
are illegal or not, but that's more awkward to do
and I doubt it buys us anything much.
However this exposed an interesting thing - on x86-32
a store of i64 is considered legal!  That is because
operations are marked legal by default, regardless of
whether the type is legal or not.  In some ways this
is clever: before type legalization this means that
operations on illegal types are considered legal;
after type legalization there are no illegal types
so now operations are only legal if they really are.
But I consider this to be too cunning for mere mortals.
Better to do things explicitly by testing AfterLegalize.
So I have changed things so that operations with illegal
types are considered illegal - indeed they can never
map to a machine operation.  However this means that
the DAG combiner is more conservative because before
it was "accidentally" performing transforms where the
type was illegal because the operation was nonetheless
marked legal.  So in a few such places I added a check
on AfterLegalize, which I suppose was actually just
forgotten before.  This causes the DAG combiner to do
slightly more than it used to, which resulted in the X86
backend blowing up because it got a slightly surprising
node it wasn't expecting, so I tweaked it.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52254 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-13 19:07:40 +00:00
Duncan Sands
4c54067f67 If queried as to whether an operation is legal
for a particular MVT, return false if the type
is illegal rather than barfing.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52229 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-12 09:01:30 +00:00
Duncan Sands
8e4eb09b1e Remove comparison methods for MVT. The main cause
of apint codegen failure is the DAG combiner doing
the wrong thing because it was comparing MVT's using
< rather than comparing the number of bits.  Removing
the < method makes this mistake impossible to commit.
Instead, add helper methods for comparing bits and use
them.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52098 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-08 20:54:56 +00:00
Duncan Sands
83ec4b6711 Wrap MVT::ValueType in a struct to get type safety
and better control the abstraction.  Rename the type
to MVT.  To update out-of-tree patches, the main
thing to do is to rename MVT::ValueType to MVT, and
rewrite expressions like MVT::getSizeInBits(VT) in
the form VT.getSizeInBits().  Use VT.getSimpleVT()
to extract a MVT::SimpleValueType for use in switch
statements (you will get an assert failure if VT is
an extended value type - these shouldn't exist after
type legalization).
This results in a small speedup of codegen and no
new testsuite failures (x86-64 linux).


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@52044 91177308-0d34-0410-b5e6-96231b3b80d8
2008-06-06 12:08:01 +00:00
Evan Cheng
0ef8de30fa Fix typos and comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@51165 91177308-0d34-0410-b5e6-96231b3b80d8
2008-05-15 22:13:02 +00:00
Evan Cheng
f0df03134e Make use of vector load and store operations to implement memcpy, memmove, and memset. Currently only X86 target is taking advantage of these.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@51140 91177308-0d34-0410-b5e6-96231b3b80d8
2008-05-15 08:39:06 +00:00
Evan Cheng
9bfa03c6fd Xform bitconvert(build_pair(load a, load b)) to a single load if the load locations are at the right offset from each other.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@51008 91177308-0d34-0410-b5e6-96231b3b80d8
2008-05-12 23:04:07 +00:00
Evan Cheng
f86f211ec2 Forgot this.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50993 91177308-0d34-0410-b5e6-96231b3b80d8
2008-05-12 20:08:05 +00:00
Evan Cheng
c6c6a3e2b3 Make OpActionsCapacity multiple of 4.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50917 91177308-0d34-0410-b5e6-96231b3b80d8
2008-05-09 21:50:23 +00:00
Mon P Wang
63307c335a Added addition atomic instrinsics and, or, xor, min, and max.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50663 91177308-0d34-0410-b5e6-96231b3b80d8
2008-05-05 19:05:59 +00:00
Arnold Schwaighofer
30e62c098b Tail call optimization improvements:
Move platform independent code (lowering of possibly overwritten
arguments, check for tail call optimization eligibility) from
target X86ISelectionLowering.cpp to TargetLowering.h and
SelectionDAGISel.cpp.

Initial PowerPC tail call implementation:

Support ppc32 implemented and tested (passes my tests and
test-suite llvm-test).  
Support ppc64 implemented and half tested (passes my tests).
On ppc tail call optimization is performed if 
  caller and callee are fastcc
  call is a tail call (in tail call position, call followed by ret)
  no variable argument lists or byval arguments
  option -tailcallopt is enabled
Supported:
 * non pic tail calls on linux/darwin
 * module-local tail calls on linux(PIC/GOT)/darwin(PIC)
 * inter-module tail calls on darwin(PIC)
If constraints are not met a normal call will be emitted.

A test checking the argument lowering behaviour on x86-64 was added.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50477 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-30 09:16:33 +00:00
Chris Lattner
5a09690446 Implement a signficant optimization for inline asm:
When choosing between constraints with multiple options,
like "ir", test to see if we can use the 'i' constraint and
go with that if possible.  This produces more optimal ASM in
all cases (sparing a register and an instruction to load it),
and fixes inline asm like this:

void test () {
  asm volatile (" %c0 %1 " : : "imr" (42), "imr"(14));
}

Previously we would dump "42" into a memory location (which
is ok for the 'm' constraint) which would cause a problem
because the 'c' modifier is not valid on memory operands.

Isn't it great how inline asm turns 'missed optimization'
into 'compile failed'??

Incidentally, this was the todo in 
PowerPC/2007-04-24-InlineAsm-I-Modifier.ll

Please do NOT pull this into Tak.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50315 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-27 00:37:18 +00:00
Chris Lattner
4376fea663 Move a bunch of inline asm code out of line.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50313 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-27 00:09:47 +00:00
Chris Lattner
5e764233f3 A few inline asm cleanups:
- Make targetlowering.h fit in 80 cols.
  - Make LowerAsmOperandForConstraint const.
  - Make lowerXConstraint -> LowerXConstraint
  - Make LowerXConstraint return a const char* instead of taking a string byref.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50312 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-26 23:02:14 +00:00
Dan Gohman
29e4bdbf27 Fix const-correctness issues with the SrcValue handling in the
memory intrinsic expansion code.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49666 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-14 17:55:48 +00:00
Dan Gohman
ec1f1a8bd2 Clean up some comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49661 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-14 17:45:20 +00:00
Dan Gohman
707e018423 Drop ISD::MEMSET, ISD::MEMMOVE, and ISD::MEMCPY, which are not Legal
on any current target and aren't optimized in DAGCombiner. Instead
of using intermediate nodes, expand the operations, choosing between
simple loads/stores, target-specific code, and library calls,
immediately.

Previously, the code to emit optimized code for these operations
was only used at initial SelectionDAG construction time; now it is
used at all times. This fixes some cases where rep;movs was being
used for small copies where simple loads/stores would be better.

This also cleans up code that checks for alignments less than 4;
let the targets make that decision instead of doing it in
target-independent code. This allows x86 to use rep;movs in
low-alignment cases.

Also, this fixes a bug that resulted in the use of rep;stos for
memsets of 0 with non-constant memory size when the alignment was
at least 4. It's better to use the library in this case, which
can be significantly faster when the size is large.

This also preserves more SourceValue information when memory
intrinsics are lowered into simple loads/stores.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49572 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-12 04:36:06 +00:00
Dan Gohman
7d8143f0ef Make isVectorClearMaskLegal's operand list const.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@49446 91177308-0d34-0410-b5e6-96231b3b80d8
2008-04-09 20:09:42 +00:00
Chris Lattner
d27c991ceb Fix "Control reaches the end of non-void function" warnings,
patch by David Chisnall.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@48963 91177308-0d34-0410-b5e6-96231b3b80d8
2008-03-30 18:22:13 +00:00
Scott Michel
5b8f82e35b Give TargetLowering::getSetCCResultType() a parameter so that ISD::SETCC's
return ValueType can depend its operands' ValueType.

This is a cosmetic change, no functionality impacted.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@48145 91177308-0d34-0410-b5e6-96231b3b80d8
2008-03-10 15:42:14 +00:00
Evan Cheng
6fd599fa69 Add a target lowering hook to control whether it's worthwhile to compress fp constant.
For x86, if sse2 is available, it's not a good idea since cvtss2sd is slower than a movsd load and it prevents load folding. On x87, it's important to shrink fp constant since fldt is very expensive.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47931 91177308-0d34-0410-b5e6-96231b3b80d8
2008-03-05 01:30:59 +00:00
Dale Johannesen
28d08fdb9f Interface of getByValTypeAlignment differed between
generic & x86 versions; change generic to follow x86
and improve comments.  Add PPC version (not right
for non-Darwin.)



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47734 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-28 22:31:51 +00:00