home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
OpenStep 4.2J (Developer)
/
os42jdev.iso
/
NextDeveloper
/
Headers
/
mach-o
/
reloc.h
< prev
next >
Wrap
C/C++ Source or Header
|
1995-06-23
|
8KB
|
160 lines
/*
* Copyright (c) 1980 Regents of the University of California.
* All rights reserved. The Berkeley software License Agreement
* specifies the terms and conditions for redistribution.
*
* @(#)a.out.h 5.1 (Berkeley) 5/30/85
*
* The structure this file describes was originally taken from the above file
* and the above copyright has been carried over to this file.
*/
#ifndef _MACHO_RELOC_H_
#define _MACHO_RELOC_H_
/*
* Format of a relocation entry of a Mach-O file. Modified from the 4.3BSD
* format. The modifications from the original format were changing the value
* of the r_symbolnum field for "local" (r_extern == 0) relocation entries.
* This modification is required to support symbols in an arbitrary number of
* sections not just the three sections (text, data and bss) in a 4.3BSD file.
* Also the last 4 bits have had the r_type tag added to them.
*/
struct relocation_info {
long r_address; /* offset in the section to what is being
relocated */
unsigned int r_symbolnum:24, /* symbol index if r_extern == 1 or section
ordinal if r_extern == 0 */
r_pcrel:1, /* was relocated pc relative already */
r_length:2, /* 0=byte, 1=word, 2=long */
r_extern:1, /* does not include value of sym referenced */
r_type:4; /* if not 0, machine specific relocation type */
};
#define R_ABS 0 /* absolute relocation type for Mach-O files */
/*
* The r_address is not really the address as it's name indicates but an offset.
* In 4.3BSD a.out objects this offset is from the start of the "segment" for
* which relocation entry is for (text or data). For Mach-O object files it is
* also an offset but from the start of the "section" for which the relocation
* entry is for.
*
* In 4.3BSD a.out objects if r_extern is zero then r_symbolnum is an ordinal
* for the segment the symbol being relocated is in. These ordinals are the
* symbol types N_TEXT, N_DATA, N_BSS or N_ABS. In Mach-O object files these
* ordinals refer to the sections in the object file in the order their section
* structures appear in the headers of the object file they are in. The first
* section has the ordinal 1, the second 2, and so on. This means that the
* same ordinal in two different object files could refer to two different
* sections. And further could have still different ordinals when combined
* by the link-editor. The value R_ABS is used for relocation entries for
* absolute symbols which need no further relocation.
*/
/*
* For RISC machines some of the references are split across two instructions
* and the instruction does not contain the complete value of the reference.
* In these cases a second, or paired relocation entry, follows each of these
* relocation entries, using a PAIR r_type, which contains the other part of the
* reference not contained in the instruction. This other part is stored in the
* pair's r_address field. The exact number of bits of the other part of the
* reference store in the r_address field is dependent on the particular
* relocation type for the particular architecture.
*/
/*
* To make scattered loading by the link editor work correctly "local"
* relocation entries can't be used when the item to be relocated is the value
* of a symbol plus an offset (where the resulting expresion is outside the
* block the link editor is moving, a blocks are divided at symbol addresses).
* In this case. where the item is a symbol value plus offset, the link editor
* needs to know more than just the section the symbol was defined. What is
* needed is the actual value of the symbol without the offset so it can do the
* relocation correctly based on where the value of the symbol got relocated to
* not the value of the expression (with the offset added to the symbol value).
* So for the NeXT 2.0 release no "local" relocation entries are ever used when
* there is a non-zero offset added to a symbol. The "external" and "local"
* relocation entries remain unchanged.
*
* The implemention is quite messy given the compatibility with the existing
* relocation entry format. The ASSUMPTION is that a section will never be
* bigger than 2**24 - 1 (0x00ffffff or 16,777,215) bytes. This assumption
* allows the r_address (which is really an offset) to fit in 24 bits and high
* bit of the r_address field in the relocation_info structure to indicate
* it is really a scattered_relocation_info structure. Since these are only
* used in places where "local" relocation entries are used and not where
* "external" relocation entries are used the r_extern field has been removed.
*
* For scattered loading to work on a RISC machine where some of the references
* are split across two instructions the link editor needs to be assured that
* each reference has a unique 32 bit reference (that more than one reference is
* NOT sharing the same high 16 bits for example) so it move each referenced
* item independent of each other. Some compilers guarantees this but the
* compilers don't so scattered loading can be done on those that do guarantee
* this.
*/
#if defined(__BIG_ENDIAN__) || defined(__LITTLE_ENDIAN__)
/*
* The reason for the ifdef's of __BIG_ENDIAN__ and __LITTLE_ENDIAN__ are that
* when stattered relocation entries were added the mistake of using a mask
* against a structure that is made up of bit fields was used. To make this
* design work this structure must be laid out in memory the same way so the
* mask can be applied can check the same bit each time (r_scattered).
*/
#endif /* defined(__BIG_ENDIAN__) || defined(__LITTLE_ENDIAN__) */
#define R_SCATTERED 0x80000000 /* mask to be applied to the r_address field
of a relocation_info structure to tell that
is is really a scattered_relocation_info
stucture */
struct scattered_relocation_info {
#ifdef __BIG_ENDIAN__
unsigned int r_scattered:1, /* 1=scattered, 0=non-scattered (see above) */
r_pcrel:1, /* was relocated pc relative already */
r_length:2, /* 0=byte, 1=word, 2=long */
r_type:4, /* if not 0, machine specific relocation type */
r_address:24; /* offset in the section to what is being
relocated */
long r_value; /* the value the item to be relocated is
refering to (without any offset added) */
#endif /* __BIG_ENDIAN__ */
#ifdef __LITTLE_ENDIAN__
unsigned int
r_address:24, /* offset in the section to what is being
relocated */
r_type:4, /* if not 0, machine specific relocation type */
r_length:2, /* 0=byte, 1=word, 2=long */
r_pcrel:1, /* was relocated pc relative already */
r_scattered:1; /* 1=scattered, 0=non-scattered (see above) */
long r_value; /* the value the item to be relocated is
refering to (without any offset added) */
#endif /* __LITTLE_ENDIAN__ */
};
/*
* Relocation types used in a generic implementation. Relocation entries for
* nornal things use the generic relocation as discribed above and their r_type
* is GENERIC_RELOC_VANILLA (a value of zero).
*
* Another type of generic relocation, GENERIC_RELOC_SECTDIFF, is to support
* the difference of two symbols defined in different sections. That is the
* expression "symbol1 - symbol2 + constant" is a relocatable expression when
* both symbols are defined in some section. For this type of relocation the
* both relocations entries are scattered relocation entries. The value of
* symbol1 is stored in the first relocation entry's r_value field and the
* value of symbol2 is stored in the pair's r_value field.
*
* A special case for a prebound lazy pointer is needed to beable to set the
* value of the lazy pointer back to its non-prebound state. This is done
* using the GENERIC_RELOC_PB_LA_PTR r_type. This is a scattered relocation
* entry where the r_value feild is the value of the lazy pointer not prebound.
*/
enum reloc_type_generic
{
GENERIC_RELOC_VANILLA, /* generic relocation as discribed above */
GENERIC_RELOC_PAIR, /* Only follows a GENRIC_RELOC_SECTDIFF */
GENERIC_RELOC_SECTDIFF,
GENERIC_RELOC_PB_LA_PTR /* prebound lazy pointer */
};
#endif /* _MACHO_RELOC_H_ */