llvm-svn: 28974

llvm-svn: 28973

llvm-svn: 28972

llvm-svn: 28971

llvm-svn: 28970

llvm-svn: 28969

llvm-svn: 28968

llvm-svn: 28967

llvm-svn: 28966

llvm-svn: 28965

llvm-svn: 28964

llvm-svn: 28963

will be profitable.  This is mainly to remove some cases where excessive
unswitching would result in long compile times and/or huge generated code.

Once someone comes up with a better heuristic that avoids these cases, this
should be switched out.

llvm-svn: 28962

llvm-svn: 28961

llvm-svn: 28960

blocks.

llvm-svn: 28959

llvm-svn: 28958

llvm-svn: 28957

llvm-svn: 28955

Remove the Function pointer cast in these calls, converting it to
a cast of argument.
%tmp60 = tail call int cast (int (ulong)* %str to int (int)*)( int 10 )
%tmp60 = tail call int cast (int (ulong)* %str to int (int)*)( uint %tmp51 )

llvm-svn: 28953

llvm-svn: 28952

to them.  This fixes a regression in my previous checkin.

llvm-svn: 28951

causes the pointer to be removed from the underlying alias analysis
implementation as well.  This impl of remove is also significantly faster than
the old one.  This fixes:
Regression/Transforms/DeadStoreElimination/2006-06-27-AST-Remove.ll

llvm-svn: 28950

llvm-svn: 28949

llvm-svn: 28948

Be more careful when updating Phi nodes after eliminating dead switch cases.  Fix
proposed by Chris.

llvm-svn: 28947

from a switch instruction.

llvm-svn: 28946

llvm-svn: 28945

llvm-svn: 28944

remove the struct return argument of a csret function, even if it is obviously
dead.

llvm-svn: 28943

llvm-svn: 28942

llvm-svn: 28941

llvm-svn: 28939

llvm-svn: 28938

llvm-svn: 28937

llvm-svn: 28936

llvm-svn: 28935

and the data

llvm-svn: 28934

the offset two bits to the left.

llvm-svn: 28933

llvm-svn: 28932