Several improvement to Sema::ParseRecordBody()...

- Adding a safer prologue. The previous prologue would accept a null and therefore assume we had an interface (which was incorrect).
- Fixed FieldDecl's classof method. This allowed me to simplify some unnecessary casting.
- When diagnosing errors, make sure the FieldDecl/EnclosingDecl are marked as invalid. In addition, don't delete the field...rather, add all fields to the enclosing decl. Memory management can/should be done elsewhere. This code was never "upgraded" to the recently added invalid decl strategy.

llvm-svn: 41964

llvm-svn: 41962

llvm-svn: 41961

ambiguity.

llvm-svn: 41960

llvm-svn: 41959

isRegister, isImmediate, and isMachineBasicBlock, which are equivalent,
and more popular.

llvm-svn: 41958

for consistency with many other transforms.

llvm-svn: 41957

are strict about such things.

llvm-svn: 41956

llvm-svn: 41954

llvm-svn: 41953

miscompilation of 188.ammp.  Reject select and bitcast in 
ValueIsOnlyUsedLocallyOrStoredToOneGlobal because RewriteHeapSROALoadUser can't handle it.

llvm-svn: 41950

llvm-svn: 41949

Now that the dust has settled on the Decl refactoring, I noticed FieldDecl didn't need NextDeclarator.  As a result, I'm removing it. 

Removing both slots (NextDeclarator/Next) end up reducing the size of fields/ivars by 40%.

llvm-svn: 41948

llvm-svn: 41947

llvm-svn: 41941

(http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070910/053370.html). It's
causing SPASS to fail.

llvm-svn: 41938

Move Decl::NextDeclarator (w/ setters/getters) down to ScopedDecl/FieldDecl.

Decl is now svelte:-)

llvm-svn: 41935

Phase 2 of making the Decl class more lightweight...

Move Identifier/Loc instance variables (and associated getters/setters) down from Decl to ScopedDecl/FieldDecl.

Objc AST's can now inherit from Decl without getting instance variables and types that are C specific. For now, I am keeping NextDeclarator, since I believe it may be useful to ObjC. If not, it can be moved later.

llvm-svn: 41934

a limited form of PHI nodes.  This finally fixes PR1639, speeding 179.art up
from 7.84s to 3.13s on PPC.

llvm-svn: 41933

llvm-svn: 41932

Add "ScopedDecl" AST, used by ValueDecl and TypeDecl. 

This allows us to sink the "Next" field, which isn't used by FieldDecl and ObjcIvarDecl.

llvm-svn: 41931

along the way of PR1639

llvm-svn: 41930

llvm-svn: 41929

bitcasts and phis.  This is a step to fixing PR1639.

llvm-svn: 41928

nodes.  This is the first step of the fix for PR1639.

llvm-svn: 41927

llvm-svn: 41926

llvm-svn: 41925

llvm-svn: 41924

llvm-svn: 41923

the front-end adopts APFloat though.  This fixes test/CodeGen/mandel.c

llvm-svn: 41908

llvm-svn: 41907

llvm-svn: 41906

llvm-svn: 41901

llvm-svn: 41900

llvm-svn: 41899

llvm-svn: 41898

1.
[(set GR32:$dst, (add GR32:$src1, GR32:$src2)),
 (modify EFLAGS)]
This indicates the source pattern expects the instruction would produce 2 values. The first is the result of the addition. The second is an implicit definition in register EFLAGS.
2.
def : Pat<(parallel (addc GR32:$src1, GR32:$src2), (modify EFLAGS)), ()>
Similar to #1 except this is used for def : Pat patterns.

llvm-svn: 41897

Add ASTContext::StringMap<char> to unique selector names.

llvm-svn: 41894

llvm-svn: 41893

Removed extraneous call to Observer->ObserveStmt when handling DeclRefExprs.

llvm-svn: 41890