LBB1_57:        #bb207.i
        movl 72(%esp), %ecx
        movb (%ecx,%eax), %cl
        movl 80(%esp), %edx
        movb %cl, 1(%edx,%eax)
        incl %eax
        cmpl $143, %eax
        jne LBB1_57     #bb207.i
        jmp LBB1_64     #cond_next255.i

intead of:

LBB1_57:        #bb207.i
        movl 72(%esp), %ecx
        movb (%ecx,%eax), %cl
        movl 80(%esp), %edx
        movb %cl, 1(%edx,%eax)
        incl %eax
        cmpl $143, %eax
        je LBB1_64      #cond_next255.i
        jmp LBB1_57     #bb207.i

This eliminates a branch per iteration of the loop.  This hurted PPC
particularly, because the extra branch meant another dispatch group for each
iteration of the loop.

llvm-svn: 31530

Clarify comments.

llvm-svn: 31529

llvm-svn: 31526

llvm-svn: 31525

int func(vFloat v0, vFloat v1) {
        int ii;
        vSInt32 vsiidx[2];
        vsiidx[0] = _mm_cvttps_epi32(v0);
        vsiidx[1] = _mm_cvttps_epi32(v1);
        ii = ((int *) vsiidx)[4];
        return ii;
}

This fixes Transforms/ScalarRepl/2006-11-07-InvalidArrayPromote.ll

llvm-svn: 31524

int func(vFloat v0, vFloat v1) {
	int ii;
	vSInt32 vsiidx[2];
	vsiidx[0] = _mm_cvttps_epi32(v0);
	vsiidx[1] = _mm_cvttps_epi32(v1);
	ii = ((int *) vsiidx)[4];
	return ii;
}

llvm-svn: 31523

llvm-svn: 31522

llvm-svn: 31521

llvm-svn: 31520

shuffle v, undef, <2, ?, 3, ?>
to movhlps
It should match to unpckhps instead.

Added proper matching code for
shuffle v, undef, <2, 3, 2, 3>

llvm-svn: 31519

llvm-svn: 31518

llvm-svn: 31517

llvm-svn: 31516

llvm-svn: 31515

llvm-svn: 31514

llvm-svn: 31513

llvm-svn: 31512

llvm-svn: 31511

Allow loop detection during debug in forwarding nodes, and revert auxcall patch as it make 176.gcc untenable

llvm-svn: 31510

llvm-svn: 31509

empty ranges.

2. Reorg how MachineDebugInfo maintains changes to debug labels.

3. Have dwarf writer use debug label info to simplify scopes and source line
coorespondence.

4. Revert the merging of compile units until I can get the bugs ironed out.

llvm-svn: 31507

llvm-svn: 31506

by 40%, FreeBench/fourinarow by 20%, and many other programs 10-25%.

On PPC, this speeds up fourinarow by 18%, and probably other things as well.

llvm-svn: 31504

llvm-svn: 31502

llvm-svn: 31501

llvm-svn: 31500

llvm-svn: 31499

llvm-svn: 31498

llvm-svn: 31497

llvm-svn: 31496

information.

llvm-svn: 31495

X86.  If happy, I'll enable this by default.

llvm-svn: 31493

llvm-svn: 31492

llvm-svn: 31491

llvm-svn: 31488

The low part goes in the first operand of expandop, not the second one.

llvm-svn: 31487

llvm-svn: 31486

llvm-svn: 31485

llvm-svn: 31484

llvm-svn: 31483