struct EncodedPtr

This yields just the basic encoding, regardless of indirection or

adjustments. This is all that's needed to determine the encoded size.

This yields just the modifier for a relative address. After the basic

value is decoded according to Type(encoding), this is what adjustment must

be done to the value.

This indicates that the value is actually stored elsewhere in memory. The

Type(encoding) still indicates the type of that stored pointer, as well as

the basic type of the encoding used to locate it. After applying the

Modifier(encoding) adjustments to the encoded pointer, that pointer must

be dereferenced to fetch the desired value.

This indicates if the encoded value is signed, so it should be

sign-extended from narrower encoding to a wider integer type.

This returns the encoded size, which may depend on the contextual address

size. It returns kDynamicSize for LEB128 types whose exact size cannot be

known without the actual data.

This normalizes the encoding so that it's unambiguous with respect to

address size. After normalization, an encoding can be used directly

without keeping track of the address size that's indicated by, or implicit

in, the context it came from.

Read an encoded value via the Memory object. Both the vaddr argument and

the encoded addresses (in case of indirection) are in whatever address

space the Memory object provides. To support the indirection case

properly, don't adjust the vaddr argument for use with a generic Memory

object. Instead use a Memory object that takes the unadjusted address and

implicitly applies the runtime load bias for the module containing the

DWARF metadata being read; this ensures that a possible second call to the

Memory object will correctly handle an address read from the metadata

rather than the given vaddr argument. When reading variable-sized

(LEB128) data, the single-argument ReadArray method of the Memory object

is expected to return at least as much data as the value encoding requires

in the single call. Returns std::nullopt if the Memory object fails.

Otherwise the value is extended to 64 bits. In the case of a signed

encoding, bit_cast

<int64

_t> should be used on the value.

Read an encoded value via the Memory object. Both the vaddr argument and

the encoded addresses (in case of indirection) are in whatever address

space the Memory object provides. To support the indirection case

properly, don't adjust the vaddr argument for use with a generic Memory

object. Instead use a Memory object that takes the unadjusted address and

implicitly applies the runtime load bias for the module containing the

DWARF metadata being read; this ensures that a possible second call to the

Memory object will correctly handle an address read from the metadata

rather than the given vaddr argument. When reading variable-sized

(LEB128) data, the single-argument ReadArray method of the Memory object

is expected to return at least as much data as the value encoding requires

in the single call. Returns std::nullopt if the Memory object fails.

Otherwise the value is extended to 64 bits. In the case of a signed

encoding, bit_cast

<int64

_t> should be used on the value.

Read an encoded value via the Memory object. Both the vaddr argument and

the encoded addresses (in case of indirection) are in whatever address

space the Memory object provides. To support the indirection case

properly, don't adjust the vaddr argument for use with a generic Memory

object. Instead use a Memory object that takes the unadjusted address and

implicitly applies the runtime load bias for the module containing the

DWARF metadata being read; this ensures that a possible second call to the

Memory object will correctly handle an address read from the metadata

rather than the given vaddr argument. When reading variable-sized

(LEB128) data, the single-argument ReadArray method of the Memory object

is expected to return at least as much data as the value encoding requires

in the single call. Returns std::nullopt if the Memory object fails.

Otherwise the value is extended to 64 bits. In the case of a signed

encoding, bit_cast

<int64

_t> should be used on the value.

Read an encoded value via the Memory object. Both the vaddr argument and

the encoded addresses (in case of indirection) are in whatever address

space the Memory object provides. To support the indirection case

properly, don't adjust the vaddr argument for use with a generic Memory

object. Instead use a Memory object that takes the unadjusted address and

implicitly applies the runtime load bias for the module containing the

DWARF metadata being read; this ensures that a possible second call to the

Memory object will correctly handle an address read from the metadata

rather than the given vaddr argument. When reading variable-sized

(LEB128) data, the single-argument ReadArray method of the Memory object

is expected to return at least as much data as the value encoding requires

in the single call. Returns std::nullopt if the Memory object fails.

Otherwise the value is extended to 64 bits. In the case of a signed

encoding, bit_cast

<int64

_t> should be used on the value.

Read an encoded value from the byte buffer. This returns an

EncodedPtr object rather than the resolved value. The caller is

responsible for applying modifiers and indirection to the value.

Read an encoded value from the byte buffer. This returns an

EncodedPtr object rather than the resolved value. The caller is

responsible for applying modifiers and indirection to the value.

Read an encoded value from the byte buffer. This returns an

EncodedPtr object rather than the resolved value. The caller is

responsible for applying modifiers and indirection to the value.

Read an encoded value from the byte buffer. This returns an

EncodedPtr object rather than the resolved value. The caller is

responsible for applying modifiers and indirection to the value.

Read an encoded value from the byte buffer. This returns an

EncodedPtr object rather than the resolved value. The caller is

responsible for applying modifiers and indirection to the value.

These are the primary values that indicate basic integer encoding.

One of these can be OR'd in with one of the basic encodings above. Note

that the relative encodings implicitly refer to different base addresses

in different contexts, e.g. kDatarel inside .eh_frame_hdr is relative to

the beginning of .eh_frame_hdr itself.