Table of Contents

Method GetElementPtrInBounds

Namespace
Ubiquity.NET.Llvm.Instructions
Assembly
Ubiquity.NET.Llvm.dll

GetElementPtrInBounds(Value, IEnumerable<Value>)

Creates a Value that accesses an element of a type referenced by a pointer

[Obsolete("Use overload that takes a pointer type and opaque pointer")]
public Value GetElementPtrInBounds(Value pointer, IEnumerable<Value> args)

Parameters

pointer Value

pointer to get an element from

args IEnumerable<Value>

additional indices for computing the resulting pointer

Returns

Value

Value for the member access. This is a Value as LLVM may optimize the expression to a ConstantExpression if it can so the actual type of the result may be ConstantExpression or GetElementPtr.

Note that pointer must be a pointer to a structure or an exception is thrown.

Remarks

For details on GetElementPointer (GEP) see The Often Misunderstood GEP Instruction. The basic gist is that the GEP instruction does not access memory, it only computes a pointer offset from a base. A common confusion is around the first index and what it means. For C and C++ programmers an expression like pFoo->bar seems to only have a single offset or index. However, that is only syntactic sugar where the compiler implicitly hides the first index. That is, there is no difference between pFoo[0].bar and pFoo->bar except that the former makes the first index explicit. LLVM requires an explicit first index, even if it is zero, in order to properly compute the offset for a given element in an aggregate type.

GetElementPtrInBounds(ITypeRef, Value, IEnumerable<Value>)

Creates a Value that accesses an element of a type referenced by a pointer

public Value GetElementPtrInBounds(ITypeRef type, Value pointer, IEnumerable<Value> args)

Parameters

type ITypeRef

Base pointer type

pointer Value

opaque pointer to get an element from

args IEnumerable<Value>

additional indices for computing the resulting pointer

Returns

Value

Value for the member access. This is a Value as LLVM may optimize the expression to a ConstantExpression if it can so the actual type of the result may be ConstantExpression or GetElementPtr.

Note that pointer must be a pointer to a structure or an exception is thrown.

Remarks

For details on GetElementPointer (GEP) see The Often Misunderstood GEP Instruction. The basic gist is that the GEP instruction does not access memory, it only computes a pointer offset from a base. A common confusion is around the first index and what it means. For C and C++ programmers an expression like pFoo->bar seems to only have a single offset or index. However, that is only syntactic sugar where the compiler implicitly hides the first index. That is, there is no difference between pFoo[0].bar and pFoo->bar except that the former makes the first index explicit. LLVM requires an explicit first index, even if it is zero, in order to properly compute the offset for a given element in an aggregate type.

GetElementPtrInBounds(Value, params Value[])

Creates a Value that accesses an element of a type referenced by a pointer

[Obsolete("Use overload that accepts base pointer type and na opaque pointer")]
public Value GetElementPtrInBounds(Value pointer, params Value[] args)

Parameters

pointer Value

pointer to get an element from

args Value[]

additional indices for computing the resulting pointer

Returns

Value

Value for the member access. This is a Value as LLVM may optimize the expression to a ConstantExpression if it can so the actual type of the result may be ConstantExpression or GetElementPtr.

Note that pointer must be a pointer to a structure or an exception is thrown.

Remarks

For details on GetElementPointer (GEP) see The Often Misunderstood GEP Instruction. The basic gist is that the GEP instruction does not access memory, it only computes a pointer offset from a base. A common confusion is around the first index and what it means. For C and C++ programmers an expression like pFoo->bar seems to only have a single offset or index. However that is only syntactic sugar where the compiler implicitly hides the first index. That is, there is no difference between pFoo[0].bar and pFoo->bar except that the former makes the first index explicit. LLVM requires an explicit first index, even if it is zero, in order to properly compute the offset for a given element in an aggregate type.

GetElementPtrInBounds(ITypeRef, Value, params Value[])

Creates a Value that accesses an element of a type referenced by a pointer

public Value GetElementPtrInBounds(ITypeRef type, Value pointer, params Value[] args)

Parameters

type ITypeRef

Base pointer type

pointer Value

opaque pointer to get an element from

args Value[]

additional indices for computing the resulting pointer

Returns

Value

Value for the member access. This is a Value as LLVM may optimize the expression to a ConstantExpression if it can so the actual type of the result may be ConstantExpression or GetElementPtr.

Note that pointer must be a pointer to a structure or an exception is thrown.

Remarks

For details on GetElementPointer (GEP) see The Often Misunderstood GEP Instruction. The basic gist is that the GEP instruction does not access memory, it only computes a pointer offset from a base. A common confusion is around the first index and what it means. For C and C++ programmers an expression like pFoo->bar seems to only have a single offset or index. However that is only syntactic sugar where the compiler implicitly hides the first index. That is, there is no difference between pFoo[0].bar and pFoo->bar except that the former makes the first index explicit. LLVM requires an explicit first index, even if it is zero, in order to properly compute the offset for a given element in an aggregate type.