[llvm] r178438 - Cleanup PPC(64) i32 -> float/double conversion
Hal Finkel
hfinkel at anl.gov
Sun Mar 31 07:50:56 PDT 2013
----- Original Message -----
> From: "Bill Schmidt" <wschmidt at linux.vnet.ibm.com>
> To: "Hal Finkel" <hfinkel at anl.gov>
> Cc: llvm-commits at cs.uiuc.edu
> Sent: Sunday, March 31, 2013 9:20:36 AM
> Subject: Re: [llvm] r178438 - Cleanup PPC(64) i32 -> float/double conversion
>
> Hi Hal,
>
> On Sun, 2013-03-31 at 01:58 +0000, Hal Finkel wrote:
> > Author: hfinkel
> > Date: Sat Mar 30 20:58:02 2013
> > New Revision: 178438
> >
> > URL: http://llvm.org/viewvc/llvm-project?rev=178438&view=rev
> > Log:
> > Cleanup PPC(64) i32 -> float/double conversion
> >
> > The existing SINT_TO_FP code for i32 -> float/double conversion was
> > disabled
> > because it relied on broken EXTSW_32/STD_32 instruction
> > definitions. The
> > original intent had been to enable these 64-bit instructions to be
> > used on CPUs
> > that support them even in 32-bit mode. Unfortunately, this form of
> > lying to
> > the infrastructure was buggy (as explained in the FIXME comment)
> > and had
> > therefore been disabled.
> >
> > This re-enables this functionality, using regular DAG nodes, but
> > only when
> > compiling in 64-bit mode. The old STD_32/EXTSW_32 definitions
> > (which were dead)
> > are removed.
> >
> > Added:
> > llvm/trunk/test/CodeGen/PowerPC/i32-to-float.ll
> > Modified:
> > llvm/trunk/lib/Target/PowerPC/PPCISelLowering.cpp
> > llvm/trunk/lib/Target/PowerPC/PPCISelLowering.h
> > llvm/trunk/lib/Target/PowerPC/PPCInstr64Bit.td
> > llvm/trunk/lib/Target/PowerPC/PPCInstrInfo.td
> > llvm/trunk/lib/Target/PowerPC/PPCRegisterInfo.cpp
> >
> > Modified: llvm/trunk/lib/Target/PowerPC/PPCISelLowering.cpp
> > URL:
> > http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/PowerPC/PPCISelLowering.cpp?rev=178438&r1=178437&r2=178438&view=diff
> > ==============================================================================
> > --- llvm/trunk/lib/Target/PowerPC/PPCISelLowering.cpp (original)
> > +++ llvm/trunk/lib/Target/PowerPC/PPCISelLowering.cpp Sat Mar 30
> > 20:58:02 2013
> > @@ -326,10 +326,8 @@ PPCTargetLowering::PPCTargetLowering(PPC
> > // We cannot do this with Promote because i64 is not a legal
> > type.
> > setOperationAction(ISD::FP_TO_UINT, MVT::i32, Custom);
> >
> > - // FIXME: disable this lowered code. This generates 64-bit
> > register values,
> > - // and we don't model the fact that the top part is clobbered
> > by calls. We
> > - // need to flag these together so that the value isn't live
> > across a call.
> > - //setOperationAction(ISD::SINT_TO_FP, MVT::i32, Custom);
> > + if (Subtarget->isPPC64())
> > + setOperationAction(ISD::SINT_TO_FP, MVT::i32, Custom);
> > } else {
> > // PowerPC does not have FP_TO_UINT on 32-bit implementations.
> > setOperationAction(ISD::FP_TO_UINT, MVT::i32, Expand);
> > @@ -592,8 +590,6 @@ const char *PPCTargetLowering::getTarget
> > case PPCISD::SRL: return "PPCISD::SRL";
> > case PPCISD::SRA: return "PPCISD::SRA";
> > case PPCISD::SHL: return "PPCISD::SHL";
> > - case PPCISD::EXTSW_32: return "PPCISD::EXTSW_32";
> > - case PPCISD::STD_32: return "PPCISD::STD_32";
> > case PPCISD::CALL: return "PPCISD::CALL";
> > case PPCISD::CALL_NOP: return "PPCISD::CALL_NOP";
> > case PPCISD::MTCTR: return "PPCISD::MTCTR";
> > @@ -4817,17 +4813,13 @@ SDValue PPCTargetLowering::LowerSINT_TO_
> > EVT PtrVT = DAG.getTargetLoweringInfo().getPointerTy();
> > SDValue FIdx = DAG.getFrameIndex(FrameIdx, PtrVT);
> >
> > - SDValue Ext64 = DAG.getNode(PPCISD::EXTSW_32, dl, MVT::i32,
> > - Op.getOperand(0));
> > + SDValue Ext64 = DAG.getNode(ISD::SIGN_EXTEND, dl, MVT::i64,
> > + Op.getOperand(0));
> >
> > // STD the extended value into the stack slot.
> > - MachineMemOperand *MMO =
> > -
> > MF.getMachineMemOperand(MachinePointerInfo::getFixedStack(FrameIdx),
> > - MachineMemOperand::MOStore, 8, 8);
> > - SDValue Ops[] = { DAG.getEntryNode(), Ext64, FIdx };
> > - SDValue Store =
> > - DAG.getMemIntrinsicNode(PPCISD::STD_32, dl,
> > DAG.getVTList(MVT::Other),
> > - Ops, 4, MVT::i64, MMO);
> > + SDValue Store = DAG.getStore(DAG.getEntryNode(), dl, Ext64,
> > FIdx,
> > + MachinePointerInfo(), false, false,
> > 0);
> > +
> > // Load the value as a double.
> > SDValue Ld = DAG.getLoad(MVT::f64, dl, Store, FIdx,
> > MachinePointerInfo(),
> > false, false, false, 0);
> >
> > Modified: llvm/trunk/lib/Target/PowerPC/PPCISelLowering.h
> > URL:
> > http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/PowerPC/PPCISelLowering.h?rev=178438&r1=178437&r2=178438&view=diff
> > ==============================================================================
> > --- llvm/trunk/lib/Target/PowerPC/PPCISelLowering.h (original)
> > +++ llvm/trunk/lib/Target/PowerPC/PPCISelLowering.h Sat Mar 30
> > 20:58:02 2013
> > @@ -91,10 +91,6 @@ namespace llvm {
> > /// code.
> > SRL, SRA, SHL,
> >
> > - /// EXTSW_32 - This is the EXTSW instruction for use with
> > "32-bit"
> > - /// registers.
> > - EXTSW_32,
> > -
> > /// CALL - A direct function call.
> > /// CALL_NOP is a call with the special NOP which follows
> > 64-bit
> > /// SVR4 calls.
> > @@ -234,14 +230,11 @@ namespace llvm {
> > /// optimizations due to constant folding.
> > VADD_SPLAT,
> >
> > - /// STD_32 - This is the STD instruction for use with
> > "32-bit" registers.
> > - STD_32 = ISD::FIRST_TARGET_MEMORY_OPCODE,
> > -
> > /// CHAIN = STBRX CHAIN, GPRC, Ptr, Type - This is a
> > /// byte-swapping store instruction. It byte-swaps the low
> > "Type" bits of
> > /// the GPRC input, then stores it through Ptr. Type can be
> > either i16 or
> > /// i32.
> > - STBRX,
> > + STBRX = ISD::FIRST_TARGET_MEMORY_OPCODE,
> >
> > /// GPRC, CHAIN = LBRX CHAIN, Ptr, Type - This is a
> > /// byte-swapping load instruction. It loads "Type" bits,
> > byte swaps it,
> >
> > Modified: llvm/trunk/lib/Target/PowerPC/PPCInstr64Bit.td
> > URL:
> > http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/PowerPC/PPCInstr64Bit.td?rev=178438&r1=178437&r2=178438&view=diff
> > ==============================================================================
> > --- llvm/trunk/lib/Target/PowerPC/PPCInstr64Bit.td (original)
> > +++ llvm/trunk/lib/Target/PowerPC/PPCInstr64Bit.td Sat Mar 30
> > 20:58:02 2013
> > @@ -452,10 +452,6 @@ def EXTSH8 : XForm_11<31, 922, (outs G8R
> > def EXTSW : XForm_11<31, 986, (outs G8RC:$rA), (ins G8RC:$rS),
> > "extsw $rA, $rS", IntSimple,
> > [(set i64:$rA, (sext_inreg i64:$rS, i32))]>,
> > isPPC64;
> > -/// EXTSW_32 - Just like EXTSW, but works on '32-bit' registers.
> > -def EXTSW_32 : XForm_11<31, 986, (outs GPRC:$rA), (ins GPRC:$rS),
> > - "extsw $rA, $rS", IntSimple,
> > - [(set i32:$rA, (PPCextsw_32 i32:$rS))]>,
> > isPPC64;
> > def EXTSW_32_64 : XForm_11<31, 986, (outs G8RC:$rA), (ins
> > GPRC:$rS),
> > "extsw $rA, $rS", IntSimple,
> > [(set i64:$rA, (sext i32:$rS))]>, isPPC64;
>
> This all looks good. Do you think we can get rid of EXTSW_32_64 as
> well? It appears to me that it's only used in two places, and in
> both
> places EXTSW ought to suffice. Having two forms of extsw causes
> trouble
> for the assembly parser. Just wondering if I'm missing a subtle
> point
> here.
Hrmm... does the pattern associated with EXTSW_32_64 match anything in practice? Maybe we could declare the sign-extension to be Promote and just define the 64-bit quantity (the use in CTRLoops could likely be handled the same way: some kind of subregister insertion and then the 64-bit version)? Would you care to investigate?
Thanks again,
Hal
>
> Thanks,
> Bill
>
> > @@ -786,15 +782,6 @@ def STDBRX: XForm_8<31, 660, (outs), (in
> > "stdbrx $rS, $dst", LdStStore,
> > [(PPCstbrx i64:$rS, xoaddr:$dst, i64)]>,
> > isPPC64,
> > PPC970_DGroup_Cracked;
> > -
> > -// STD_32/STDX_32 - Just like STD/STDX, but uses a '32-bit' input
> > register.
> > -def STD_32 : DSForm_1<62, 0, (outs), (ins GPRC:$rT, memrix:$dst),
> > - "std $rT, $dst", LdStSTD,
> > - [(PPCstd_32 i32:$rT, ixaddr:$dst)]>,
> > isPPC64;
> > -def STDX_32 : XForm_8<31, 149, (outs), (ins GPRC:$rT,
> > memrr:$dst),
> > - "stdx $rT, $dst", LdStSTD,
> > - [(PPCstd_32 i32:$rT, xaddr:$dst)]>,
> > isPPC64,
> > - PPC970_DGroup_Cracked;
> > }
> >
> > // Stores with Update (pre-inc).
> >
> > Modified: llvm/trunk/lib/Target/PowerPC/PPCInstrInfo.td
> > URL:
> > http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/PowerPC/PPCInstrInfo.td?rev=178438&r1=178437&r2=178438&view=diff
> > ==============================================================================
> > --- llvm/trunk/lib/Target/PowerPC/PPCInstrInfo.td (original)
> > +++ llvm/trunk/lib/Target/PowerPC/PPCInstrInfo.td Sat Mar 30
> > 20:58:02 2013
> > @@ -105,10 +105,6 @@ def PPCsrl : SDNode<"PPCISD::SRL"
> > def PPCsra : SDNode<"PPCISD::SRA" , SDTIntShiftOp>;
> > def PPCshl : SDNode<"PPCISD::SHL" , SDTIntShiftOp>;
> >
> > -def PPCextsw_32 : SDNode<"PPCISD::EXTSW_32" , SDTIntUnaryOp>;
> > -def PPCstd_32 : SDNode<"PPCISD::STD_32" , SDTStore,
> > - [SDNPHasChain, SDNPMayStore]>;
> > -
> > // These are target-independent nodes, but have target-specific
> > formats.
> > def callseq_start : SDNode<"ISD::CALLSEQ_START",
> > SDT_PPCCallSeqStart,
> > [SDNPHasChain, SDNPOutGlue]>;
> >
> > Modified: llvm/trunk/lib/Target/PowerPC/PPCRegisterInfo.cpp
> > URL:
> > http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/PowerPC/PPCRegisterInfo.cpp?rev=178438&r1=178437&r2=178438&view=diff
> > ==============================================================================
> > --- llvm/trunk/lib/Target/PowerPC/PPCRegisterInfo.cpp (original)
> > +++ llvm/trunk/lib/Target/PowerPC/PPCRegisterInfo.cpp Sat Mar 30
> > 20:58:02 2013
> > @@ -68,7 +68,7 @@ PPCRegisterInfo::PPCRegisterInfo(const P
> > ImmToIdxMap[PPC::LHZ8] = PPC::LHZX8; ImmToIdxMap[PPC::LWZ8] =
> > PPC::LWZX8;
> > ImmToIdxMap[PPC::STB8] = PPC::STBX8; ImmToIdxMap[PPC::STH8] =
> > PPC::STHX8;
> > ImmToIdxMap[PPC::STW8] = PPC::STWX8; ImmToIdxMap[PPC::STDU] =
> > PPC::STDUX;
> > - ImmToIdxMap[PPC::ADDI8] = PPC::ADD8; ImmToIdxMap[PPC::STD_32] =
> > PPC::STDX_32;
> > + ImmToIdxMap[PPC::ADDI8] = PPC::ADD8;
> > }
> >
> > /// getPointerRegClass - Return the register class to use to hold
> > pointers.
> > @@ -522,7 +522,6 @@ PPCRegisterInfo::eliminateFrameIndex(Mac
> > case PPC::LWA:
> > case PPC::LD:
> > case PPC::STD:
> > - case PPC::STD_32:
> > isIXAddr = true;
> > break;
> > }
> >
> > Added: llvm/trunk/test/CodeGen/PowerPC/i32-to-float.ll
> > URL:
> > http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/i32-to-float.ll?rev=178438&view=auto
> > ==============================================================================
> > --- llvm/trunk/test/CodeGen/PowerPC/i32-to-float.ll (added)
> > +++ llvm/trunk/test/CodeGen/PowerPC/i32-to-float.ll Sat Mar 30
> > 20:58:02 2013
> > @@ -0,0 +1,31 @@
> > +; RUN: llc < %s -mtriple=powerpc64-unknown-linux-gnu -mcpu=g5 |
> > FileCheck %s
> > +target datalayout =
> > "E-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-f128:128:128-v128:128:128-n32:64"
> > +target triple = "powerpc64-unknown-linux-gnu"
> > +
> > +define float @foo(i32 %a) nounwind {
> > +entry:
> > + %x = sitofp i32 %a to float
> > + ret float %x
> > +
> > +; CHECK: @foo
> > +; CHECK: extsw [[REG:[0-9]+]], 3
> > +; CHECK: std [[REG]],
> > +; CHECK: lfd [[REG2:[0-9]+]],
> > +; CHECK: fcfid [[REG3:[0-9]+]], [[REG2]]
> > +; CHECK: frsp 1, [[REG3]]
> > +; CHECK: blr
> > +}
> > +
> > +define double @goo(i32 %a) nounwind {
> > +entry:
> > + %x = sitofp i32 %a to double
> > + ret double %x
> > +
> > +; CHECK: @goo
> > +; CHECK: extsw [[REG:[0-9]+]], 3
> > +; CHECK: std [[REG]],
> > +; CHECK: lfd [[REG2:[0-9]+]],
> > +; CHECK: fcfid 1, [[REG2]]
> > +; CHECK: blr
> > +}
> > +
> >
> >
> > _______________________________________________
> > llvm-commits mailing list
> > llvm-commits at cs.uiuc.edu
> > http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
> >
>
>
More information about the llvm-commits
mailing list