Implement the MoreHelp utility that calls a function to printmore help information if the MoreHelp global is not null.

llvm-svn: 17774

llvm-svn: 17773

llvm-svn: 17772

llvm-svn: 17771

llvm-svn: 17770

llvm-svn: 17769

llvm-svn: 17768

Completely rewritten to allow reading of archives and symbol table lookup in a more efficient manner.

llvm-svn: 17767

llvm-svn: 17766

llvm-svn: 17765

llvm-svn: 17764

llvm-svn: 17763

llvm-svn: 17751

llvm-svn: 17750

llvm-svn: 17749

llvm-svn: 17748

value.  This allows us to turn more globals into constants and eliminate them.
This patch implements GlobalOpt/load-store-global.llx.

Note that this patch speeds up 255.vortex from:

Output/255.vortex.out-cbe.time:program 7.640000
Output/255.vortex.out-llc.time:program 9.810000

to:

Output/255.vortex.out-cbe.time:program 7.250000
Output/255.vortex.out-llc.time:program 9.490000

Which isn't bad at all!

llvm-svn: 17746

llvm-svn: 17744

llvm-svn: 17742

If this happens, detect it early instead of relying on instcombine to notice
it later.  This can be a big speedup, because PHI nodes can have many
incoming values.

llvm-svn: 17741

This exposes subsequent optimization possiblities and reduces code size.
This triggers 1423 times in spec.

llvm-svn: 17740

  %X = alloca ...
  %Y = alloca ...
    X == Y

into false.  This allows us to simplify some stuff in eon (and probably
many other C++ programs) where operator= was checking for self assignment.
Folding this allows us to SROA several additional structs.

llvm-svn: 17735

llvm-svn: 17734

llvm-svn: 17733

Exclude bigfib, so that we effectively exclude all C++ benchmarks.
Update to-do list: mention va_start.

llvm-svn: 17732

constant value.  This makes the return value dead and allows for
simplification in the caller.

This implements IPConstantProp/return-constant.ll

This triggers several dozen times throughout SPEC.

llvm-svn: 17730

not zero.

llvm-svn: 17728

of the array is just two.  This occurs 8 times in gcc, 6 times in crafty, and
12 times in 099.go.

This implements ScalarRepl/sroa_two.ll

llvm-svn: 17727

llvm-svn: 17725

llvm-svn: 17724

llvm-svn: 17723

llvm-svn: 17722

llvm-svn: 17721

llvm-svn: 17719

argument pointers.  This is only valid to do if the function already
unconditionally loaded an argument or if the pointer passed in is known
to be valid.  Make sure to do the required checks.

This fixed ArgumentPromotion/control-flow.ll and the Burg program.

llvm-svn: 17718

llvm-svn: 17714

CBackend/2004-11-13-FunctionPointerCast.llx

llvm-svn: 17710

shld is a very high latency operation. Instead of emitting it for shifts of
two or three, open code the equivalent operation which is faster on athlon
and P4 (by a substantial margin).

For example, instead of compiling this:

long long X2(long long Y) { return Y << 2; }

to:

X3_2:
        movl 4(%esp), %eax
        movl 8(%esp), %edx
        shldl $2, %eax, %edx
        shll $2, %eax
        ret

Compile it to:

X2:
        movl 4(%esp), %eax
        movl 8(%esp), %ecx
        movl %eax, %edx
        shrl $30, %edx
        leal (%edx,%ecx,4), %edx
        shll $2, %eax
        ret

Likewise, for << 3, compile to:

X3:
        movl 4(%esp), %eax
        movl 8(%esp), %ecx
        movl %eax, %edx
        shrl $29, %edx
        leal (%edx,%ecx,8), %edx
        shll $3, %eax
        ret

This matches icc, except that icc open codes the shifts as adds on the P4.

llvm-svn: 17707

llvm-svn: 17706

long long X3_2(long long Y) { return Y+Y; }
int X(int Y) { return Y+Y; }

into:

X3_2:
        movl 4(%esp), %eax
        movl 8(%esp), %edx
        addl %eax, %eax
        adcl %edx, %edx
        ret
X:
        movl 4(%esp), %eax
        addl %eax, %eax
        ret

instead of:

X3_2:
        movl 4(%esp), %eax
        movl 8(%esp), %edx
        shldl $1, %eax, %edx
        shll $1, %eax
        ret

X:
        movl 4(%esp), %eax
        shll $1, %eax
        ret

llvm-svn: 17705