llvm-svn: 39991

llvm-svn: 39990

llvm-svn: 39989

Note: FastDSE now equals or exceeds the results of old DSE on all of SPEC2000 and SPEC2006. Unless major problems
show up in the testers, it will likely completely replace old DSE in the near future.

llvm-svn: 39986

llvm-svn: 39984

llvm-svn: 39983

llvm-svn: 39980

llvm-svn: 39979

llvm-svn: 39978

llvm-svn: 39977

llvm-svn: 39976

Don't use * in XFAIL line unless you want it to XFAIL on everything. I don't believe you can use true regular expressions here. It will just look for those keywords. Please test this on sparc and see if they still are xfailed.

llvm-svn: 39975

Missed the case where alloca is used but the stack size (not including callee-saved portion) is zero. Thanks Dan.

llvm-svn: 39974

just always use strtod.  This is temporary code anyway.

llvm-svn: 39972

llvm-svn: 39971

llvm-svn: 39970

llvm-svn: 39969

llvm-svn: 39967

llvm-svn: 39965

Move tests that have C/C++ sources into the appropriate directory. This
allows them to be selected for testing based on whether llvm-gcc is
present or not.

llvm-svn: 39963

llvm-svn: 39962

llvm-svn: 39960

llvm-svn: 39958

llvm-svn: 39956

Change dyn_cast for reference types to be more like pointers and not need the canonical type. Also fix so that we're not expecting a return value from a void function

llvm-svn: 39954

    According to the spec (C++ 5p6[expr]), we need to adjust "T&" to
    "T" before further analysis. We do this via the "implicit cast"
    thingy.

llvm-svn: 39953

having that implementation file #include <cstdlib>.

llvm-svn: 39952

llvm-svn: 39950

Convert UsualArithmeticConversions to return "void". Now that we synthesize ImplicitCastExpr's,
there is no compelling need to return the converted type. If both expression type's are arithmetic, then
both types will always be the same. If they aren't (for pointer/int types, say), then the
types will be different. The client is responsible for distinguishing...

llvm-svn: 39947

beyond the end of an alloca to make FastDSE faster and more aggressive.

llvm-svn: 39945

llvm-svn: 39944

Implement semantic analysis for the cast operator.

llvm-svn: 39943

Sigh .. fix copy & paste error.

llvm-svn: 39939

found by prepending a .
In build-for-llvm-top.sh, dereference variables for exansion.

llvm-svn: 39936

Remove the 2 QualType references to method UsualArithmeticConversions. Last week, I added these
to quickly fix a regression. Avoiding them entirely is a much cleaner solution. Clients of 
UsualArithmeticConversions should simply call getType() on the expression to get the
converted type. In practice, only a small number of routines care about this.

llvm-svn: 39934

This also avoids a fork bomb from the llvm-top/build script which
doesn't know how to handle cycles in the dependence graph.

llvm-svn: 39933

isPointerType and isVectorType to only look through a single level of typedef
when one is present.  For this invalid code:

typedef float float4 __attribute__((vector_size(16)));
typedef int int4 __attribute__((vector_size(16)));
typedef int4* int4p;
void test(float4 a, int4p result, int i) {
    result[i] = a;
}

we now get:

t.c:5:15: error: incompatible types assigning 'float4' to 'int4'
    result[i] = a;
    ~~~~~~~~~ ^ ~

instead of:

t.c:5:15: error: incompatible types assigning 'float4' to 'int  __attribute__((vector_size(16)))'
    result[i] = a;
    ~~~~~~~~~ ^ ~

The rest of the type predicates should be upgraded to do the same thing.

llvm-svn: 39932

llvm-svn: 39931

Change DefaultFunctionArrayConversions and UsualUnaryConversions to return void. The caller 
needs to query the expression for the type. Since both these functions guarantee the expression 
contains a valid type, removed old/vacuous asserts (from code calling both of these routines).

llvm-svn: 39930

dead stores on 400.perlbench.

llvm-svn: 39929