llvm-svn: 48137

Avoid creating BUILD_VECTOR of all zero elements of "non-normalized" type (e.g. v8i16 on x86) after legalizer. Instruction selection does not expect to see them. In all likelihood this can only be an issue in a bugpoint reduced test case.

llvm-svn: 48136

llvm-svn: 48135

the build system.  Patch by Joachim Durchholz  for PR2121

llvm-svn: 48134

llvm-svn: 48133

before compile_c.  Patch by Joachim Durchholz!

llvm-svn: 48132

more important if this api gets wider use.

llvm-svn: 48131

Change insert/extract subreg instructions to be able to be used in TableGen patterns.
Use the above features to reimplement an x86-64 pseudo instruction as a pattern.

llvm-svn: 48130

objc @try parser to use it, fixing a FIXME.  Update the 
objc-try-catch-1.m file to pass now that we get more reasonable 
errors.

llvm-svn: 48129

llvm-svn: 48128

llvm-svn: 48127

llvm-svn: 48126

Add support in TableGen for unknown operands that infer their type from the pattern their used in. This will be used to allow insert/extract subreg patterns in .td files!

llvm-svn: 48125

variables from the symbolic state. Now we keep a mapping from the predecessor
node of a statement and its cleaned state, and interpose an accessor
"GetState()" to return the cleaned state when the node is the predecessor node,
and the real state otherwise. This obviates problem of ever accidentally
cleaning the state more than once (thus blowing away new bindings by mistake).

llvm-svn: 48124

llvm-svn: 48123

field to 32 bits, thus enabling correct handling of ByVal
structs bigger than 0x1ffff.  Abstract interface a bit.
Fixes gcc.c-torture/execute/pr23135.c and 
gcc.c-torture/execute/pr28982b.c in gcc testsuite (were ICE'ing
on ppc32, quietly producing wrong code on x86-32.)

llvm-svn: 48122

llvm-svn: 48121

two regression tests:
test/CodeGen/PowerPC/2007-10-21-LocalRegAllocAssert.ll
test/CodeGen/PowerPC/2007-10-21-LocalRegAllocAssert2.ll

llvm-svn: 48120

cell really does support cross-regclass moves, because R3 is in lots of different regclasses, and the code is not consistent when it comes to value tracking.

llvm-svn: 48119

llvm-svn: 48118

llvm-svn: 48117

llvm-svn: 48116

- "Redefinition of I" (iterator masks previous definition)
- include missing header file

Patch by Argiris Kirtzidis!

llvm-svn: 48115

RemoveDeadBindings early because (1) it will always be called and (2) we can
dispatch to a plug-in transfer function that can evaluate the effect of dead
symbols (not yet added).

llvm-svn: 48114

method. This paves the way for other alterations to RemoveDeadBindings that are
transparent to its callers.

llvm-svn: 48113

llvm-svn: 48112

llvm-svn: 48111

llvm-svn: 48110

llvm-svn: 48109

they are produced by calls (which are known exact) and by cross block copies
which are known to be produced by extends.

This improves:

define double @test2() {
	%tmp85 = call double asm sideeffect "fld0", "={st(0)}"()
	ret double %tmp85
}

from:

_test2:
	subl	$20, %esp
	# InlineAsm Start
	fld0
	# InlineAsm End
	fstpl	8(%esp)
	movsd	8(%esp), %xmm0
	movsd	%xmm0, (%esp)
	fldl	(%esp)
	addl	$20, %esp
	#FP_REG_KILL
	ret

to:

_test2:
	# InlineAsm Start
	fld0
	# InlineAsm End
	#FP_REG_KILL
	ret

by avoiding a f64 <-> f80 trip

llvm-svn: 48108

an RFP register class.

Teach ScheduleDAG how to handle CopyToReg with different src/dst 
reg classes.

This allows us to compile trivial inline asms that expect stuff
on the top of x87-fp stack.

llvm-svn: 48107

llvm-svn: 48106

in different register classes, e.g. copy of ST(0) to RFP*.  This gets
some really trivial inline asm working that plops things on the top of
stack (PR879)

llvm-svn: 48105

RST -> RFP{32/64/80}.  We only handle ST(0) for now.

llvm-svn: 48104

llvm-svn: 48103

of the LHS expressions is Unknown, the value of the entire expression
does not evaluate to the RHS (as is the case with normal assignments).

llvm-svn: 48102

llvm-svn: 48101

llvm-svn: 48100

Secondly, we have to check whether the branch is actually pointing to the block
with the unwind in it. We could have gotten here because of the unwind_to alone.

llvm-svn: 48099

codegen yet because these can't be spilled (they don't exist until after RA).

llvm-svn: 48098