[llvm-commits] [LLVMdev] Dead register (was Re: [llvm] r145819)

[Hal Finkel]

On Tue, 2011-12-06 at 17:39 -0600, Hal Finkel wrote:
> On Mon, 2011-12-05 at 13:18 -0800, Jakob Stoklund Olesen wrote:
> > On Dec 5, 2011, at 12:56 PM, Hal Finkel wrote:
> > 
> > > RegScavenger is complaining about use of an undefined register, CTR8, in
> > > the BCTR8 instruction, in the following instance (this is from the PPC
> > > backend):
> > > 
> > > BB#38: derived from LLVM BB %for.end50
> > >    Predecessors according to CFG: BB#36
> > >        %X3<def> = LD 0, <fi#27>; mem:LD8[FixedStack27]
> > >        %X4<def> = RLDICR %X3<kill>, 3, 60
> > >        %X5<def> = LI8 <jt#0>[TF=4]
> > >        %X5<def> = ADDIS8 %X5<kill>, <jt#0>[TF=8]
> > >        %X4<def> = LDX %X4<kill>, %X5<kill>; mem:LD8[JumpTable]
> > >        MTCTR8 %X4<kill>, %CTR8<imp-def,dead>
> > >        BCTR8 %CTR8<imp-use,kill>, %RM<imp-use>
> > >    Successors according to CFG: BB#23 BB#15 BB#7 BB#8 BB#9 BB#10 BB#11
> > > BB#25 BB#12 BB#16 BB#18 BB#13 BB#17
> > > 
> > > How could CRT8 be marked implicitly-defined and also dead in the same
> > > instruction when it is clearly used in the next instruction?
> > 
> > This is the kind of sloppy liveness, I was talking about ;-)
> > 
> > llc -verify-machineinstrs should give you better info.
> 
> Unfortunately, this just tells me what I already knew:
> 
> *** Bad machine code: Using an undefined physical register ***
> - function:    check
> - basic block: for.end50 0x2bef428 (BB#38)
> - instruction: BCTR8 %CTR8<imp-use>, %RM<imp-use>
> - operand 0:   %CTR8<imp-use>
> LLVM ERROR: Found 1 machine code errors.
> 
> This comes from the following four statements in
> PPCDAGToDAGISel::Select; what's wrong here?
> SDValue Chain = N->getOperand(0);
> SDValue Target = N->getOperand(1);
> unsigned Opc = PPC::MTCTR8;
> unsigned Reg = PPC::BCTR8;
> Chain = SDValue(CurDAG->getMachineNode(Opc, dl, MVT::Other, Target,
>                                            Chain), 0);
> return CurDAG->SelectNodeTo(N, Reg, MVT::Other, Chain);

So it seems that CTR8 is being added through a call to
MachineInstr::addRegisterDead by InstrEmitter::EmitMachineNode, and this
is done for all implicitly-defined registers that don't otherwise seem
to be used. Furthermore, it seems that determining usage is tied to
looking for "glued" nodes. Changing the first MVT::Other to MVT::Glue
solved the problem.

 -Hal

> 
> Thanks again,
> Hal
> 
> > 
> > /jakob
> > 
> 

-- 
Hal Finkel
Postdoctoral Appointee
Leadership Computing Facility
Argonne National Laboratory