bugpoint on leopard.

llvm-svn: 30150

llvm-svn: 30149

llvm-svn: 30148

llvm-svn: 30147

llvm-svn: 30146

llvm-svn: 30145

llvm-svn: 30144

This is PR899

llvm-svn: 30143

too many phi operands when lowering a switch to branches in some cases.

llvm-svn: 30142

llvm-svn: 30141

llvm-svn: 30140

llvm-svn: 30137

llvm-svn: 30135

llvm-svn: 30134

llvm-svn: 30133

llvm-svn: 30132

llvm-svn: 30131

Add private member function getMoudle().

llvm-svn: 30130

huge honking FIXME comment. I'll address the "Live Intervals Analysis"
section soon.

llvm-svn: 30129

llvm-svn: 30128

llvm-svn: 30127

llvm-svn: 30126

llvm-svn: 30125

used to mangle symbol name, directly.

llvm-svn: 30124

Untabify.

llvm-svn: 30123

llvm-svn: 30122

llvm-svn: 30121

llvm-svn: 30120

llvm-svn: 30119

llvm-svn: 30118

llvm-svn: 30117

llvm-svn: 30116

lowerinvoke regtests.

llvm-svn: 30115

llvm-svn: 30114

llvm-svn: 30113

llvm-svn: 30112

llvm-svn: 30111

adds assertions that check to make sure every operand gets emitted.

llvm-svn: 30110

def operand or a use operand.

llvm-svn: 30109

actually *removes* one of the operands, instead of just assigning both operands
the same register.  This make reasoning about instructions unnecessarily complex,
because you need to know if you are before or after register allocation to match
up operand #'s with the target description file.

Changing this also gets rid of a bunch of hacky code in various places.

This patch also includes changes to fold loads into cmp/test instructions in
the X86 backend, along with a significant simplification to the X86 spill
folding code.

llvm-svn: 30108