llvm-svn: 73423

incomming chain of the RETURN node. The incomming chain must
be the outgoing chain of the CALL node. This causes the
backend to identify tail calls that are not tail calls. This
patch fixes this.

llvm-svn: 73387

- Change register allocation hint to a pair of unsigned integers. The hint type is zero (which means prefer the register specified as second part of the pair) or entirely target dependent.
- Allow targets to specify alternative register allocation orders based on allocation hint.

Part 2.
- Use the register allocation hint system to implement more aggressive load / store multiple formation.
- Aggressively form LDRD / STRD. These are formed *before* register allocation. It has to be done this way to shorten live interval of base and offset registers. e.g.
v1025 = LDR v1024, 0
v1026 = LDR v1024, 0
=>
v1025,v1026 = LDRD v1024, 0

If this transformation isn't done before allocation, v1024 will overlap v1025 which means it more difficult to allocate a register pair.

- Even with the register allocation hint, it may not be possible to get the desired allocation. In that case, the post-allocation load / store multiple pass must fix the ldrd / strd instructions. They can either become ldm / stm instructions or back to a pair of ldr / str instructions.

This is work in progress, not yet enabled.

llvm-svn: 73381

llvm-svn: 73362

Move register allocation preference (or hint) from LiveInterval to MachineRegisterInfo. This allows more passes to set them.

llvm-svn: 73346

Introduce new BinaryObject (blob) class, ELF Writer modified to use it. BinaryObject.h by Aaron Gray

llvm-svn: 73333

consecutive addresses togther. This makes it easier for the post-allocation pass
to form ldm / stm.

This is step 1. We are still missing a lot of ldm / stm opportunities because
of register allocation are not done in the desired order. More enhancements
coming.

llvm-svn: 73291

llvm.dbg.region.end() intrinsic is not required to be in _last_ basic block in a function.  If that happens then any basic block that follows (lexically) the block with regin.end will not have scope info available.  LexicalScopeStack relies on processing basic block in CFG order, but this processing order is not guaranteed. Things get complicated when the optimizer gets a chance to optimizer IR with dbg intrinsics. 
Apply defensive patch to preserve at least one lexical scope till the end of function.

llvm-svn: 73282

llvm-svn: 73258

llvm-svn: 73257

llvm-svn: 73256

If killed register is defined by implicit_def, do not clear it since it's live range may overlap another def of same register.

llvm-svn: 73255

llvm-svn: 73244

Emission for globals, using the correct data sections
Function alignment can be computed for each target using TargetELFWriterInfo
Some small fixes

llvm-svn: 73201

llvm-svn: 73174

llvm-svn: 73156

Add the beginnings of an implementatation of lazy liveness analysis, based on "Fast Liveness Checking for SSA-form Programs" by Boissinot, et al.

This is still very early, hasn't been tested, and is not yet well documented.  More to come soon.

llvm-svn: 73141

llvm-svn: 73040

llvm-svn: 73039

relocation sections.

llvm-svn: 73038

converting from an MMX vector to an i64.

llvm-svn: 73024

on x86 to handle more cases.  Fix a bug in said code that would cause it 
to read past the end of an object.  Rewrite the code in 
SelectionDAGLegalize::ExpandBUILD_VECTOR to be a bit more general. 
Remove PerformBuildVectorCombine, which is no longer necessary with 
these changes.  In addition to simplifying the code, with this change, 
we can now catch a few more cases of consecutive loads.

llvm-svn: 73012

types.

llvm-svn: 72993

llvm-svn: 72992

llvm-svn: 72982

integer type to be consistent with normal operation legalization.  No visible
change because nothing is actually using this at the moment.

llvm-svn: 72980

Update code generator to use this attribute and remove NoImplicitFloat target option.
Update llc to set this attribute when -no-implicit-float command line option is used.

llvm-svn: 72959

build vectors with i64 elements will only appear on 32b x86 before legalize.
Since vector widening occurs during legalize, and produces i64 build_vector 
elements, the dag combiner is never run on these before legalize splits them
into 32b elements.

Teach the build_vector dag combine in x86 back end to recognize consecutive 
loads producing the low part of the vector.

Convert the two uses of TLI's consecutive load recognizer to pass LoadSDNodes
since that was required implicitly.

Add a testcase for the transform.

Old:
	subl	$28, %esp
	movl	32(%esp), %eax
	movl	4(%eax), %ecx
	movl	%ecx, 4(%esp)
	movl	(%eax), %eax
	movl	%eax, (%esp)
	movaps	(%esp), %xmm0
	pmovzxwd	%xmm0, %xmm0
	movl	36(%esp), %eax
	movaps	%xmm0, (%eax)
	addl	$28, %esp
	ret

New:
	movl	4(%esp), %eax
	pmovzxwd	(%eax), %xmm0
	movl	8(%esp), %eax
	movaps	%xmm0, (%eax)
	ret

llvm-svn: 72957

llvm-svn: 72948

llvm-svn: 72941

MachO Writer. This will change with the arrival of ObjectCodeEmitter and
BinaryObject

llvm-svn: 72906

integer and floating-point opcodes, introducing
FAdd, FSub, and FMul.

For now, the AsmParser, BitcodeReader, and IRBuilder all preserve
backwards compatability, and the Core LLVM APIs preserve backwards
compatibility for IR producers. Most front-ends won't need to change
immediately.

This implements the first step of the plan outlined here:
http://nondot.org/sabre/LLVMNotes/IntegerOverflow.txt

llvm-svn: 72897

using Promote which won't work because i64 isn't
a legal type.  It's easy enough to use Custom, but
then we have the problem that when the type
legalizer is promoting FP_TO_UINT->i16, it has no
way of telling it should prefer FP_TO_SINT->i32
to FP_TO_UINT->i32.  I have uncomfortably hacked
this by making the type legalizer choose FP_TO_SINT
when both are Custom.
This fixes several regressions in the testsuite.

llvm-svn: 72891

RALinScan::attemptTrivialCoalescing() was returning a virtual register instead of the physical register it is allocated to. This resulted in virtual register(s) being added the live-in sets.

llvm-svn: 72890

A value defined by an implicit_def can be liven to a use BB. This is unfortunate. But register allocator still has to add it to the live-in set of the use BB.

llvm-svn: 72889

llvm-svn: 72880

instcombine doesn't know when it's safe. To partially compensate
for this, introduce new code to do this transformation in
dagcombine, which can use UnsafeFPMath.

llvm-svn: 72872

llvm-svn: 72870

llvm-svn: 72828

llvm-svn: 72825