search for: ptrty

...his assertion. Assertion `ValueType != Type::VoidTy && "Pointer to void is not valid, use sbyte* instead!"' failed. I can try checking the type before building the function, as in: void visitLoadInst(LoadInst &I) { Value *P; P = I.getPointerOperand(); PointerType* PtrTy = (PointerType*)(P); Constant* fcall; if (PtrTy->getElementType()->isInteger()) { fcall = buildFcallMem(*theModule, "load_int", Type::Int32Ty); CallInst::Create(fcall, P, "", &I); } else if (PtrTy->getElementType()->isInteger()){ fcall = build...

On Wed, May 20, 2009 at 7:31 PM, Scott Ricketts <sricketts at maxentric.com> wrote: > Where load_fcall is build using a call to Module::getOrInsertFunction > as in the example here: > > http://wiki.llvm.org/HowTo:_Insert_a_function_call > > My question is, what do I pass as the argument type for P above? The > following seems to work, as long as there are no floating

...eturn V->getType()->isInteger(); } namespace llvm { Index: lib/Target/TargetData.cpp =================================================================== --- lib/Target/TargetData.cpp (revision 53136) +++ lib/Target/TargetData.cpp (working copy) @@ -553,8 +553,7 @@ TI = gep_type_begin(ptrTy, Indices, Indices+NumIndices); for (unsigned CurIDX = 0; CurIDX != NumIndices; ++CurIDX, ++TI) { if (const StructType *STy = dyn_cast<StructType>(*TI)) { - assert(Indices[CurIDX]->getType() == Type::Int32Ty && - "Illegal struct idx"); + asser...

...deGen/Analysis.cpp? If TargetLowering::getValueType() doesn't work for arbitrary address spaces, you should fix it. I think my earlier comments about getIntPtrConstant still hold: instead of "DAG.getIntPtrConstant(Offset, addrSpace)", you can just write "DAG.getConstant(Offset, PtrTy)". + EVT NewPtrVT = TLI.getPointerTy(dyn_cast<PointerType>( + SV->getType())->getAddressSpace()); + if (PtrVT != NewPtrVT) { + // Check to see if we want to change the size of the pointer + // based on the address space and if so extend or truncate the pointer. +...

...} > > namespace llvm { > Index: lib/Target/TargetData.cpp > =================================================================== > --- lib/Target/TargetData.cpp (revision 53136) > +++ lib/Target/TargetData.cpp (working copy) > @@ -553,8 +553,7 @@ > TI = gep_type_begin(ptrTy, Indices, Indices+NumIndices); > for (unsigned CurIDX = 0; CurIDX != NumIndices; ++CurIDX, ++TI) { > if (const StructType *STy = dyn_cast<StructType>(*TI)) { > - assert(Indices[CurIDX]->getType() == Type::Int32Ty && > - "Illegal struct idx&...

Hi all, I'm fiddling around a bit with programmatically created GEP instructions, which is failing when using i64 for any indix, except the first. The reason behind this, is that StructureType::indexValid only accepts Value* that are constant ints of width 32. I can't really see why this limitation is here. SequentialType solves this slightly differently, it allows for both 32 and 64

> -----Original Message----- > From: Villmow, Micah > Sent: Friday, August 24, 2012 2:56 PM > To: 'Eli Friedman' > Cc: LLVM Developers Mailing List > Subject: RE: [LLVMdev] RFC: Supporting different sized address space > arithmetic > > Eli, > There is a patch that implements the beginning what I think is the > correct approach to support the backend

...h an > uppercase letter). > > > if (PtrBits < 64) > > - OffsVal = DAG.getNode(ISD::TRUNCATE, getCurDebugLoc(), > > - TLI.getPointerTy(), > > + OffsVal = DAG.getNode(ISD::TRUNCATE, getCurDebugLoc(), > > PtrTy, DAG.getConstant(Offs, MVT::i64)); > > This is not necessarily specific to your patch, but can you explain the > special role of 64-bits here? > > Thanks again, > Hal > > On Fri, 24 Aug 2012 23:07:54 +0000 > "Villmow, Micah" <Micah.Villmow at amd.com> w...

...pecifies that they should start with an uppercase letter). > if (PtrBits < 64) > - OffsVal = DAG.getNode(ISD::TRUNCATE, getCurDebugLoc(), > - TLI.getPointerTy(), > + OffsVal = DAG.getNode(ISD::TRUNCATE, getCurDebugLoc(), > PtrTy, DAG.getConstant(Offs, MVT::i64)); This is not necessarily specific to your patch, but can you explain the special role of 64-bits here? Thanks again, Hal On Fri, 24 Aug 2012 23:07:54 +0000 "Villmow, Micah" <Micah.Villmow at amd.com> wrote: > > > > -----Original Me...

...e llvm { >> Index: lib/Target/TargetData.cpp >> =================================================================== >> --- lib/Target/TargetData.cpp (revision 53136) >> +++ lib/Target/TargetData.cpp (working copy) >> @@ -553,8 +553,7 @@ >> TI = gep_type_begin(ptrTy, Indices, Indices+NumIndices); >> for (unsigned CurIDX = 0; CurIDX != NumIndices; ++CurIDX, ++TI) { >> if (const StructType *STy = dyn_cast<StructType>(*TI)) { >> - assert(Indices[CurIDX]->getType() == Type::Int32Ty && >> - "Illeg...

...gt; > > > > if (PtrBits < 64) > > > - OffsVal = DAG.getNode(ISD::TRUNCATE, getCurDebugLoc(), > > > - TLI.getPointerTy(), > > > + OffsVal = DAG.getNode(ISD::TRUNCATE, getCurDebugLoc(), > > > PtrTy, DAG.getConstant(Offs, MVT::i64)); > > > > This is not necessarily specific to your patch, but can you explain > > the special role of 64-bits here? > > > > Thanks again, > > Hal > > > > On Fri, 24 Aug 2012 23:07:54 +0000 > > "Villmow, M...

...teStructGEP without the assert. I thought it was a version issue, so I changed the call to: Value *x_idx = builder.CreateConstGEP2_32(cB, 0, 0); However I got the same assert. It looks like this line in Instruction.h from LLVM is producing the assert further down the stack: 815 Type *PtrTy = PointerType::get(checkGEPType( 816 getIndexedType(Ptr->getType(), IdxList)), 817 Ptr->getType()->getPointerAddressSpace()); Resulting in the debugger producing: (gdb) p PtrTy $19 = (llvm::Type *) 0x7f...

...argetLowering::getValueType() doesn't work for arbitrary address > spaces, you should fix it. > > I think my earlier comments about getIntPtrConstant still hold: > instead of "DAG.getIntPtrConstant(Offset, addrSpace)", you can just > write "DAG.getConstant(Offset, PtrTy)". > > + EVT NewPtrVT = TLI.getPointerTy(dyn_cast<PointerType>( > + SV->getType())->getAddressSpace()); > + if (PtrVT != NewPtrVT) { > + // Check to see if we want to change the size of the pointer > + // based on the address space and if so extend...

...ert. I thought it was a version issue, so I changed the call to: > > Value *x_idx = builder.CreateConstGEP2_32(cB, 0, 0); > > However I got the same assert. It looks like this line in Instruction.h from LLVM is producing the assert further down the stack: > > 815 Type *PtrTy = PointerType::get(checkGEPType( > 816 getIndexedType(Ptr->getType(), IdxList)), > 817 Ptr->getType()->getPointerAddressSpace()); > > Resulting in the debugger producing: > > (gdb) p PtrTy &...

...n issue, so I changed the call to: > > > > Value *x_idx = builder.CreateConstGEP2_32(cB, 0, 0); > > > > However I got the same assert. It looks like this line in Instruction.h from LLVM is producing the assert further down the stack: > > > > 815 Type *PtrTy = PointerType::get(checkGEPType( > > 816 getIndexedType(Ptr->getType(), IdxList)), > > 817 Ptr->getType()->getPointerAddressSpace()); > > > > Resulting in the debugger producing: > >...

...to: > > > > > > Value *x_idx = builder.CreateConstGEP2_32(cB, 0, 0); > > > > > > However I got the same assert. It looks like this line in Instruction.h from LLVM is producing the assert further down the stack: > > > > > > 815 Type *PtrTy = PointerType::get(checkGEPType( > > > 816 getIndexedType(Ptr->getType(), IdxList)), > > > 817 Ptr->getType()->getPointerAddressSpace()); > > > > > > Resulting in the debugger...

...alValue *GV, unsigned Val) { } } + namespace llvm { namespace { /// @brief A class for maintaining the slot number definition @@ -2882,7 +2901,8 @@ error_code BitcodeReader::ParseFunctionBody(Function *F) { break; } case bitc::FUNC_CODE_INST_CMPXCHG: { - // CMPXCHG:[ptrty, ptr, cmp, new, vol, ordering, synchscope] + // CMPXCHG:[ptrty, ptr, cmp, new, vol, successordering, synchscope, + // failureordering] unsigned OpNum = 0; Value *Ptr, *Cmp, *New; if (getValueTypePair(Record, OpNum, NextValueNo, Ptr) || @@ -2890,13 +2910,21 @@...