d9d6f09465
processing same-section pcrel relocations. (TC_RELOC_RTSYM_LOC_FIXUP): Default to 1.
2303 lines
62 KiB
C
2303 lines
62 KiB
C
/* write.c - emit .o file
|
|
Copyright (C) 1986, 87, 90, 91, 92, 93, 1994 Free Software Foundation, Inc.
|
|
|
|
This file is part of GAS, the GNU Assembler.
|
|
|
|
GAS is free software; you can redistribute it and/or modify
|
|
it under the terms of the GNU General Public License as published by
|
|
the Free Software Foundation; either version 2, or (at your option)
|
|
any later version.
|
|
|
|
GAS is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with GAS; see the file COPYING. If not, write to
|
|
the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. */
|
|
|
|
/* This thing should be set up to do byteordering correctly. But... */
|
|
|
|
#include "as.h"
|
|
#include "subsegs.h"
|
|
#include "obstack.h"
|
|
#include "output-file.h"
|
|
|
|
/* This looks like a good idea. Let's try turning it on always, for now. */
|
|
#undef BFD_FAST_SECTION_FILL
|
|
#define BFD_FAST_SECTION_FILL
|
|
|
|
/* The NOP_OPCODE is for the alignment fill value. Fill it with a nop
|
|
instruction so that the disassembler does not choke on it. */
|
|
#ifndef NOP_OPCODE
|
|
#define NOP_OPCODE 0x00
|
|
#endif
|
|
|
|
#ifndef TC_ADJUST_RELOC_COUNT
|
|
#define TC_ADJUST_RELOC_COUNT(FIXP,COUNT)
|
|
#endif
|
|
|
|
#ifndef TC_FORCE_RELOCATION
|
|
#define TC_FORCE_RELOCATION(FIXP) 0
|
|
#endif
|
|
|
|
#ifndef WORKING_DOT_WORD
|
|
extern CONST int md_short_jump_size;
|
|
extern CONST int md_long_jump_size;
|
|
#endif
|
|
|
|
int symbol_table_frozen;
|
|
|
|
#ifdef BFD_ASSEMBLER
|
|
/* We generally attach relocs to frag chains. However, after we have
|
|
chained these all together into a segment, any relocs we add after
|
|
that must be attached to a segment. This will include relocs added
|
|
in md_estimate_size_for_relax, for example. */
|
|
static int frags_chained = 0;
|
|
#endif
|
|
|
|
#ifndef BFD_ASSEMBLER
|
|
|
|
#ifndef MANY_SEGMENTS
|
|
struct frag *text_frag_root;
|
|
struct frag *data_frag_root;
|
|
struct frag *bss_frag_root;
|
|
|
|
struct frag *text_last_frag; /* Last frag in segment. */
|
|
struct frag *data_last_frag; /* Last frag in segment. */
|
|
static struct frag *bss_last_frag; /* Last frag in segment. */
|
|
#endif
|
|
|
|
#ifndef BFD
|
|
static object_headers headers;
|
|
static char *the_object_file;
|
|
#endif
|
|
|
|
long string_byte_count;
|
|
char *next_object_file_charP; /* Tracks object file bytes. */
|
|
|
|
#ifndef OBJ_VMS
|
|
int magic_number_for_object_file = DEFAULT_MAGIC_NUMBER_FOR_OBJECT_FILE;
|
|
#endif
|
|
|
|
#endif /* BFD_ASSEMBLER */
|
|
|
|
#ifdef BFD_ASSEMBLER
|
|
static fixS *fix_new_internal PARAMS ((fragS *, int where, int size,
|
|
symbolS *add, symbolS *sub,
|
|
offsetT offset, int pcrel,
|
|
bfd_reloc_code_real_type r_type));
|
|
#else
|
|
static fixS *fix_new_internal PARAMS ((fragS *, int where, int size,
|
|
symbolS *add, symbolS *sub,
|
|
offsetT offset, int pcrel,
|
|
int r_type));
|
|
#endif
|
|
#if defined (BFD_ASSEMBLER) || !defined (BFD)
|
|
static long fixup_segment PARAMS ((fixS * fixP, segT this_segment_type));
|
|
#endif
|
|
static relax_addressT relax_align PARAMS ((relax_addressT addr, int align));
|
|
|
|
/*
|
|
* fix_new()
|
|
*
|
|
* Create a fixS in obstack 'notes'.
|
|
*/
|
|
static fixS *
|
|
fix_new_internal (frag, where, size, add_symbol, sub_symbol, offset, pcrel,
|
|
r_type)
|
|
fragS *frag; /* Which frag? */
|
|
int where; /* Where in that frag? */
|
|
int size; /* 1, 2, or 4 usually. */
|
|
symbolS *add_symbol; /* X_add_symbol. */
|
|
symbolS *sub_symbol; /* X_op_symbol. */
|
|
offsetT offset; /* X_add_number. */
|
|
int pcrel; /* TRUE if PC-relative relocation. */
|
|
#ifdef BFD_ASSEMBLER
|
|
bfd_reloc_code_real_type r_type; /* Relocation type */
|
|
#else
|
|
int r_type; /* Relocation type */
|
|
#endif
|
|
{
|
|
fixS *fixP;
|
|
|
|
fixP = (fixS *) obstack_alloc (¬es, sizeof (fixS));
|
|
|
|
fixP->fx_frag = frag;
|
|
fixP->fx_where = where;
|
|
fixP->fx_size = size;
|
|
fixP->fx_addsy = add_symbol;
|
|
fixP->fx_subsy = sub_symbol;
|
|
fixP->fx_offset = offset;
|
|
fixP->fx_pcrel = pcrel;
|
|
#if defined(NEED_FX_R_TYPE) || defined (BFD_ASSEMBLER)
|
|
fixP->fx_r_type = r_type;
|
|
#endif
|
|
fixP->fx_im_disp = 0;
|
|
fixP->fx_pcrel_adjust = 0;
|
|
fixP->fx_bit_fixP = 0;
|
|
fixP->fx_addnumber = 0;
|
|
fixP->tc_fix_data = NULL;
|
|
fixP->fx_tcbit = 0;
|
|
fixP->fx_done = 0;
|
|
|
|
#ifdef TC_something
|
|
fixP->fx_bsr = 0;
|
|
#endif
|
|
|
|
as_where (&fixP->fx_file, &fixP->fx_line);
|
|
|
|
/* Usually, we want relocs sorted numerically, but while
|
|
comparing to older versions of gas that have relocs
|
|
reverse sorted, it is convenient to have this compile
|
|
time option. xoxorich. */
|
|
|
|
{
|
|
|
|
#ifdef BFD_ASSEMBLER
|
|
fixS **seg_fix_rootP = (frags_chained
|
|
? &seg_info (now_seg)->fix_root
|
|
: &frchain_now->fix_root);
|
|
fixS **seg_fix_tailP = (frags_chained
|
|
? &seg_info (now_seg)->fix_tail
|
|
: &frchain_now->fix_tail);
|
|
#endif
|
|
|
|
#ifdef REVERSE_SORT_RELOCS
|
|
|
|
fixP->fx_next = *seg_fix_rootP;
|
|
*seg_fix_rootP = fixP;
|
|
|
|
#else /* REVERSE_SORT_RELOCS */
|
|
|
|
fixP->fx_next = NULL;
|
|
|
|
if (*seg_fix_tailP)
|
|
(*seg_fix_tailP)->fx_next = fixP;
|
|
else
|
|
*seg_fix_rootP = fixP;
|
|
*seg_fix_tailP = fixP;
|
|
|
|
#endif /* REVERSE_SORT_RELOCS */
|
|
|
|
}
|
|
|
|
return fixP;
|
|
}
|
|
|
|
/* Create a fixup relative to a symbol (plus a constant). */
|
|
|
|
fixS *
|
|
fix_new (frag, where, size, add_symbol, offset, pcrel, r_type)
|
|
fragS *frag; /* Which frag? */
|
|
int where; /* Where in that frag? */
|
|
short int size; /* 1, 2, or 4 usually. */
|
|
symbolS *add_symbol; /* X_add_symbol. */
|
|
offsetT offset; /* X_add_number. */
|
|
int pcrel; /* TRUE if PC-relative relocation. */
|
|
#ifdef BFD_ASSEMBLER
|
|
bfd_reloc_code_real_type r_type; /* Relocation type */
|
|
#else
|
|
int r_type; /* Relocation type */
|
|
#endif
|
|
{
|
|
return fix_new_internal (frag, where, size, add_symbol,
|
|
(symbolS *) NULL, offset, pcrel, r_type);
|
|
}
|
|
|
|
/* Create a fixup for an expression. Currently we only support fixups
|
|
for difference expressions. That is itself more than most object
|
|
file formats support anyhow. */
|
|
|
|
fixS *
|
|
fix_new_exp (frag, where, size, exp, pcrel, r_type)
|
|
fragS *frag; /* Which frag? */
|
|
int where; /* Where in that frag? */
|
|
short int size; /* 1, 2, or 4 usually. */
|
|
expressionS *exp; /* Expression. */
|
|
int pcrel; /* TRUE if PC-relative relocation. */
|
|
#ifdef BFD_ASSEMBLER
|
|
bfd_reloc_code_real_type r_type; /* Relocation type */
|
|
#else
|
|
int r_type; /* Relocation type */
|
|
#endif
|
|
{
|
|
symbolS *add = NULL;
|
|
symbolS *sub = NULL;
|
|
offsetT off = 0;
|
|
|
|
switch (exp->X_op)
|
|
{
|
|
case O_absent:
|
|
break;
|
|
|
|
case O_uminus:
|
|
sub = exp->X_add_symbol;
|
|
off = exp->X_add_number;
|
|
break;
|
|
|
|
case O_subtract:
|
|
sub = exp->X_op_symbol;
|
|
/* Fall through. */
|
|
case O_symbol:
|
|
add = exp->X_add_symbol;
|
|
/* Fall through. */
|
|
case O_constant:
|
|
off = exp->X_add_number;
|
|
break;
|
|
|
|
default:
|
|
as_bad ("expression too complex for fixup");
|
|
}
|
|
|
|
return fix_new_internal (frag, where, size, add, sub, off,
|
|
pcrel, r_type);
|
|
}
|
|
|
|
/* Append a string onto another string, bumping the pointer along. */
|
|
void
|
|
append (charPP, fromP, length)
|
|
char **charPP;
|
|
char *fromP;
|
|
unsigned long length;
|
|
{
|
|
/* Don't trust memcpy() of 0 chars. */
|
|
if (length == 0)
|
|
return;
|
|
|
|
memcpy (*charPP, fromP, length);
|
|
*charPP += length;
|
|
}
|
|
|
|
#ifndef BFD_ASSEMBLER
|
|
int section_alignment[SEG_MAXIMUM_ORDINAL];
|
|
#endif
|
|
|
|
/*
|
|
* This routine records the largest alignment seen for each segment.
|
|
* If the beginning of the segment is aligned on the worst-case
|
|
* boundary, all of the other alignments within it will work. At
|
|
* least one object format really uses this info.
|
|
*/
|
|
void
|
|
record_alignment (seg, align)
|
|
/* Segment to which alignment pertains */
|
|
segT seg;
|
|
/* Alignment, as a power of 2 (e.g., 1 => 2-byte boundary, 2 => 4-byte
|
|
boundary, etc.) */
|
|
int align;
|
|
{
|
|
#ifdef BFD_ASSEMBLER
|
|
if (align > bfd_get_section_alignment (stdoutput, seg))
|
|
bfd_set_section_alignment (stdoutput, seg, align);
|
|
#else
|
|
if (align > section_alignment[(int) seg])
|
|
section_alignment[(int) seg] = align;
|
|
#endif
|
|
}
|
|
|
|
#if defined (BFD_ASSEMBLER) || ! defined (BFD)
|
|
|
|
static fragS *
|
|
chain_frchains_together_1 (section, frchp)
|
|
segT section;
|
|
struct frchain *frchp;
|
|
{
|
|
fragS dummy, *prev_frag = &dummy;
|
|
fixS fix_dummy, *prev_fix = &fix_dummy;
|
|
|
|
for (; frchp && frchp->frch_seg == section; frchp = frchp->frch_next)
|
|
{
|
|
prev_frag->fr_next = frchp->frch_root;
|
|
prev_frag = frchp->frch_last;
|
|
#ifdef BFD_ASSEMBLER
|
|
if (frchp->fix_root != (fixS *) NULL)
|
|
{
|
|
if (seg_info (section)->fix_root == (fixS *) NULL)
|
|
seg_info (section)->fix_root = frchp->fix_root;
|
|
prev_fix->fx_next = frchp->fix_root;
|
|
seg_info (section)->fix_tail = frchp->fix_tail;
|
|
prev_fix = frchp->fix_tail;
|
|
}
|
|
#endif
|
|
}
|
|
prev_frag->fr_next = 0;
|
|
return prev_frag;
|
|
}
|
|
|
|
#endif
|
|
|
|
#ifdef BFD_ASSEMBLER
|
|
|
|
static void
|
|
chain_frchains_together (abfd, section, xxx)
|
|
bfd *abfd; /* unused */
|
|
segT section;
|
|
PTR xxx; /* unused */
|
|
{
|
|
segment_info_type *info;
|
|
|
|
/* BFD may have introduced its own sections without using
|
|
subseg_new, so it is possible that seg_info is NULL. */
|
|
info = seg_info (section);
|
|
if (info != (segment_info_type *) NULL)
|
|
info->frchainP->frch_last
|
|
= chain_frchains_together_1 (section, info->frchainP);
|
|
|
|
/* Now that we've chained the frags together, we must add new fixups
|
|
to the segment, not to the frag chain. */
|
|
frags_chained = 1;
|
|
}
|
|
|
|
#endif
|
|
|
|
#if !defined (BFD) && !defined (BFD_ASSEMBLER)
|
|
|
|
void
|
|
remove_subsegs (head, seg, root, last)
|
|
frchainS *head;
|
|
int seg;
|
|
fragS **root;
|
|
fragS **last;
|
|
{
|
|
*root = head->frch_root;
|
|
*last = chain_frchains_together_1 (seg, head);
|
|
}
|
|
|
|
#endif /* BFD */
|
|
|
|
#if defined (BFD_ASSEMBLER) || !defined (BFD)
|
|
|
|
#ifdef BFD_ASSEMBLER
|
|
static void
|
|
cvt_frag_to_fill (sec, fragP)
|
|
segT sec;
|
|
fragS *fragP;
|
|
#else
|
|
static void
|
|
cvt_frag_to_fill (headers, fragP)
|
|
object_headers *headers;
|
|
fragS *fragP;
|
|
#endif
|
|
{
|
|
switch (fragP->fr_type)
|
|
{
|
|
case rs_align:
|
|
case rs_org:
|
|
#ifdef HANDLE_ALIGN
|
|
HANDLE_ALIGN (fragP);
|
|
#endif
|
|
fragP->fr_type = rs_fill;
|
|
know (fragP->fr_next != NULL);
|
|
fragP->fr_offset = (fragP->fr_next->fr_address
|
|
- fragP->fr_address
|
|
- fragP->fr_fix) / fragP->fr_var;
|
|
break;
|
|
|
|
case rs_fill:
|
|
break;
|
|
|
|
case rs_machine_dependent:
|
|
#ifdef BFD_ASSEMBLER
|
|
md_convert_frag (stdoutput, sec, fragP);
|
|
#else
|
|
md_convert_frag (headers, fragP);
|
|
#endif
|
|
|
|
assert (fragP->fr_next == NULL || (fragP->fr_next->fr_address - fragP->fr_address == fragP->fr_fix));
|
|
|
|
/*
|
|
* After md_convert_frag, we make the frag into a ".space 0".
|
|
* Md_convert_frag() should set up any fixSs and constants
|
|
* required.
|
|
*/
|
|
frag_wane (fragP);
|
|
break;
|
|
|
|
#ifndef WORKING_DOT_WORD
|
|
case rs_broken_word:
|
|
{
|
|
struct broken_word *lie;
|
|
|
|
if (fragP->fr_subtype)
|
|
{
|
|
fragP->fr_fix += md_short_jump_size;
|
|
for (lie = (struct broken_word *) (fragP->fr_symbol);
|
|
lie && lie->dispfrag == fragP;
|
|
lie = lie->next_broken_word)
|
|
if (lie->added == 1)
|
|
fragP->fr_fix += md_long_jump_size;
|
|
}
|
|
frag_wane (fragP);
|
|
}
|
|
break;
|
|
#endif
|
|
|
|
default:
|
|
BAD_CASE (fragP->fr_type);
|
|
break;
|
|
}
|
|
}
|
|
|
|
#endif /* defined (BFD_ASSEMBLER) || !defined (BFD) */
|
|
|
|
#ifdef BFD_ASSEMBLER
|
|
static void
|
|
relax_and_size_seg (abfd, sec, xxx)
|
|
bfd *abfd;
|
|
asection *sec;
|
|
PTR xxx;
|
|
{
|
|
flagword flags;
|
|
fragS *fragp;
|
|
segment_info_type *seginfo;
|
|
int x;
|
|
valueT size, newsize;
|
|
|
|
flags = bfd_get_section_flags (abfd, sec);
|
|
|
|
seginfo = seg_info (sec);
|
|
if (seginfo && seginfo->frchainP)
|
|
{
|
|
relax_segment (seginfo->frchainP->frch_root, sec);
|
|
for (fragp = seginfo->frchainP->frch_root; fragp; fragp = fragp->fr_next)
|
|
cvt_frag_to_fill (sec, fragp);
|
|
for (fragp = seginfo->frchainP->frch_root;
|
|
fragp->fr_next;
|
|
fragp = fragp->fr_next)
|
|
/* walk to last elt */;
|
|
size = fragp->fr_address + fragp->fr_fix;
|
|
}
|
|
else
|
|
size = 0;
|
|
|
|
if (size > 0 && ! seginfo->bss)
|
|
flags |= SEC_HAS_CONTENTS;
|
|
|
|
/* @@ This is just an approximation. */
|
|
if (seginfo && seginfo->fix_root)
|
|
flags |= SEC_RELOC;
|
|
else
|
|
flags &= ~SEC_RELOC;
|
|
x = bfd_set_section_flags (abfd, sec, flags);
|
|
assert (x == true);
|
|
|
|
newsize = md_section_align (sec, size);
|
|
x = bfd_set_section_size (abfd, sec, newsize);
|
|
assert (x == true);
|
|
|
|
/* If the size had to be rounded up, add some padding in the last
|
|
non-empty frag. */
|
|
assert (newsize >= size);
|
|
if (size != newsize)
|
|
{
|
|
fragS *last = seginfo->frchainP->frch_last;
|
|
fragp = seginfo->frchainP->frch_root;
|
|
while (fragp->fr_next != last)
|
|
fragp = fragp->fr_next;
|
|
last->fr_address = size;
|
|
fragp->fr_offset += newsize - size;
|
|
}
|
|
|
|
#ifdef tc_frob_section
|
|
tc_frob_section (sec);
|
|
#endif
|
|
#ifdef obj_frob_section
|
|
obj_frob_section (sec);
|
|
#endif
|
|
}
|
|
|
|
#ifdef DEBUG2
|
|
static void
|
|
dump_section_relocs (abfd, sec, stream_)
|
|
bfd *abfd;
|
|
asection *sec;
|
|
char *stream_;
|
|
{
|
|
FILE *stream = (FILE *) stream_;
|
|
segment_info_type *seginfo = seg_info (sec);
|
|
fixS *fixp = seginfo->fix_root;
|
|
|
|
if (!fixp)
|
|
return;
|
|
|
|
fprintf (stream, "sec %s relocs:\n", sec->name);
|
|
while (fixp)
|
|
{
|
|
symbolS *s = fixp->fx_addsy;
|
|
if (s)
|
|
{
|
|
fprintf (stream, " %08x: %s(%s", fixp, S_GET_NAME (s),
|
|
s->bsym->section->name);
|
|
if (s->bsym->flags & BSF_SECTION_SYM)
|
|
{
|
|
fprintf (stream, " section sym");
|
|
if (S_GET_VALUE (s))
|
|
fprintf (stream, "+%x", S_GET_VALUE (s));
|
|
}
|
|
else
|
|
fprintf (stream, "+%x", S_GET_VALUE (s));
|
|
fprintf (stream, ")+%x\n", fixp->fx_offset);
|
|
}
|
|
else
|
|
fprintf (stream, " %08x: type %d no sym\n", fixp, fixp->fx_r_type);
|
|
fixp = fixp->fx_next;
|
|
}
|
|
}
|
|
#else
|
|
#define dump_section_relocs(ABFD,SEC,STREAM) (void)(ABFD,SEC,STREAM)
|
|
#endif
|
|
|
|
#ifndef EMIT_SECTION_SYMBOLS
|
|
#define EMIT_SECTION_SYMBOLS 1
|
|
#endif
|
|
|
|
static void
|
|
adjust_reloc_syms (abfd, sec, xxx)
|
|
bfd *abfd;
|
|
asection *sec;
|
|
PTR xxx;
|
|
{
|
|
segment_info_type *seginfo = seg_info (sec);
|
|
fixS *fixp;
|
|
|
|
if (seginfo == NULL)
|
|
return;
|
|
|
|
dump_section_relocs (abfd, sec, stderr);
|
|
|
|
for (fixp = seginfo->fix_root; fixp; fixp = fixp->fx_next)
|
|
if (fixp->fx_done)
|
|
/* ignore it */;
|
|
else if (fixp->fx_addsy)
|
|
{
|
|
symbolS *sym = fixp->fx_addsy;
|
|
asection *symsec = sym->bsym->section;
|
|
|
|
/* If it's one of these sections, assume the symbol is
|
|
definitely going to be output. The code in
|
|
md_estimate_size_before_relax in tc-mips.c uses this test
|
|
as well, so if you change this code you should look at that
|
|
code. */
|
|
if (symsec == &bfd_und_section
|
|
|| symsec == &bfd_abs_section
|
|
|| bfd_is_com_section (symsec))
|
|
{
|
|
fixp->fx_addsy->sy_used_in_reloc = 1;
|
|
continue;
|
|
}
|
|
|
|
/* Since we're reducing to section symbols, don't attempt to reduce
|
|
anything that's already using one. */
|
|
if (sym->bsym->flags & BSF_SECTION_SYM)
|
|
{
|
|
fixp->fx_addsy->sy_used_in_reloc = 1;
|
|
continue;
|
|
}
|
|
|
|
/* Is there some other reason we can't adjust this one? (E.g.,
|
|
call/bal links in i960-bout symbols.) */
|
|
#ifdef obj_fix_adjustable
|
|
if (! obj_fix_adjustable (fixp))
|
|
{
|
|
fixp->fx_addsy->sy_used_in_reloc = 1;
|
|
continue;
|
|
}
|
|
#endif
|
|
|
|
/* Is there some other (target cpu dependent) reason we can't adjust
|
|
this one? (E.g. relocations involving function addresses on
|
|
the PA. */
|
|
#ifdef tc_fix_adjustable
|
|
if (! tc_fix_adjustable (fixp))
|
|
{
|
|
fixp->fx_addsy->sy_used_in_reloc = 1;
|
|
continue;
|
|
}
|
|
#endif
|
|
|
|
/* If the section symbol isn't going to be output, the relocs
|
|
at least should still work. If not, figure out what to do
|
|
when we run into that case. */
|
|
fixp->fx_offset += S_GET_VALUE (sym);
|
|
if (sym->sy_frag)
|
|
fixp->fx_offset += sym->sy_frag->fr_address;
|
|
fixp->fx_addsy = section_symbol (symsec);
|
|
fixp->fx_addsy->sy_used_in_reloc = 1;
|
|
}
|
|
#if 1/*def RELOC_REQUIRES_SYMBOL*/
|
|
else
|
|
{
|
|
/* There was no symbol required by this relocation. However,
|
|
BFD doesn't really handle relocations without symbols well.
|
|
(At least, the COFF support doesn't.) So for now we fake up
|
|
a local symbol in the absolute section. */
|
|
|
|
fixp->fx_addsy = section_symbol (absolute_section);
|
|
fixp->fx_addsy->sy_used_in_reloc = 1;
|
|
}
|
|
#endif
|
|
|
|
dump_section_relocs (abfd, sec, stderr);
|
|
}
|
|
|
|
static void
|
|
write_relocs (abfd, sec, xxx)
|
|
bfd *abfd;
|
|
asection *sec;
|
|
PTR xxx;
|
|
{
|
|
segment_info_type *seginfo = seg_info (sec);
|
|
int i;
|
|
unsigned int n;
|
|
arelent **relocs;
|
|
fixS *fixp;
|
|
char *err;
|
|
|
|
/* If seginfo is NULL, we did not create this section; don't do
|
|
anything with it. */
|
|
if (seginfo == NULL)
|
|
return;
|
|
|
|
fixup_segment (seginfo->fix_root, sec);
|
|
|
|
n = 0;
|
|
for (fixp = seginfo->fix_root; fixp; fixp = fixp->fx_next)
|
|
n++;
|
|
|
|
#ifndef RELOC_EXPANSION_POSSIBLE
|
|
/* Set up reloc information as well. */
|
|
relocs = (arelent **) bfd_alloc_by_size_t (stdoutput,
|
|
n * sizeof (arelent *));
|
|
memset ((char*)relocs, 0, n * sizeof (arelent*));
|
|
|
|
i = 0;
|
|
for (fixp = seginfo->fix_root; fixp != (fixS *) NULL; fixp = fixp->fx_next)
|
|
{
|
|
arelent *reloc;
|
|
char *data;
|
|
bfd_reloc_status_type s;
|
|
|
|
if (fixp->fx_done)
|
|
{
|
|
n--;
|
|
continue;
|
|
}
|
|
reloc = tc_gen_reloc (sec, fixp);
|
|
if (!reloc)
|
|
{
|
|
n--;
|
|
continue;
|
|
}
|
|
data = fixp->fx_frag->fr_literal + fixp->fx_where;
|
|
if (fixp->fx_where + fixp->fx_size
|
|
> fixp->fx_frag->fr_fix + fixp->fx_frag->fr_offset)
|
|
abort ();
|
|
|
|
if (reloc->howto->partial_inplace == false
|
|
&& reloc->howto->pcrel_offset == true
|
|
&& reloc->howto->pc_relative == true)
|
|
{
|
|
/* bfd_perform_relocation screws this up */
|
|
reloc->addend += reloc->address;
|
|
}
|
|
/* Pass bogus address so that when bfd_perform_relocation adds
|
|
`reloc->address' back in, it'll come up with `data', which is where
|
|
we want it to operate. We can't just do it by fudging reloc->address,
|
|
since that might be used in the calculations(?). */
|
|
s = bfd_perform_relocation (stdoutput, reloc, data - reloc->address,
|
|
sec, stdoutput, &err);
|
|
switch (s)
|
|
{
|
|
case bfd_reloc_ok:
|
|
break;
|
|
case bfd_reloc_overflow:
|
|
as_bad_where (fixp->fx_file, fixp->fx_line, "relocation overflow");
|
|
break;
|
|
default:
|
|
as_fatal ("bad return from bfd_perform_relocation");
|
|
}
|
|
relocs[i++] = reloc;
|
|
}
|
|
#else
|
|
n = n * MAX_RELOC_EXPANSION;
|
|
/* Set up reloc information as well. */
|
|
relocs = (arelent **) bfd_alloc_by_size_t (stdoutput,
|
|
n * sizeof (arelent *));
|
|
|
|
i = 0;
|
|
for (fixp = seginfo->fix_root; fixp != (fixS *) NULL; fixp = fixp->fx_next)
|
|
{
|
|
arelent **reloc;
|
|
char *data;
|
|
bfd_reloc_status_type s;
|
|
int j;
|
|
|
|
if (fixp->fx_done)
|
|
{
|
|
n--;
|
|
continue;
|
|
}
|
|
reloc = tc_gen_reloc (sec, fixp);
|
|
|
|
for (j = 0; reloc[j]; j++)
|
|
{
|
|
relocs[i++] = reloc[j];
|
|
assert(i <= n);
|
|
}
|
|
data = fixp->fx_frag->fr_literal + fixp->fx_where;
|
|
if (fixp->fx_where + fixp->fx_size
|
|
> fixp->fx_frag->fr_fix + fixp->fx_frag->fr_offset)
|
|
abort ();
|
|
for (j = 0; reloc[j]; j++)
|
|
{
|
|
s = bfd_perform_relocation (stdoutput, reloc[j],
|
|
data - reloc[0]->address,
|
|
sec, stdoutput, &err);
|
|
switch (s)
|
|
{
|
|
case bfd_reloc_ok:
|
|
break;
|
|
default:
|
|
as_fatal ("bad return from bfd_perform_relocation");
|
|
}
|
|
}
|
|
}
|
|
n = i;
|
|
#endif
|
|
|
|
#ifdef DEBUG4
|
|
{
|
|
int i, j, nsyms;
|
|
asymbol **sympp;
|
|
sympp = bfd_get_outsymbols (stdoutput);
|
|
nsyms = bfd_get_symcount (stdoutput);
|
|
for (i = 0; i < n; i++)
|
|
if (((*relocs[i]->sym_ptr_ptr)->flags & BSF_SECTION_SYM) == 0)
|
|
{
|
|
for (j = 0; j < nsyms; j++)
|
|
if (sympp[j] == *relocs[i]->sym_ptr_ptr)
|
|
break;
|
|
if (j == nsyms)
|
|
abort ();
|
|
}
|
|
}
|
|
#endif
|
|
|
|
if (n)
|
|
bfd_set_reloc (stdoutput, sec, relocs, n);
|
|
else
|
|
bfd_set_section_flags (abfd, sec,
|
|
(bfd_get_section_flags (abfd, sec)
|
|
& (flagword) ~SEC_RELOC));
|
|
|
|
#ifdef DEBUG3
|
|
{
|
|
int i;
|
|
arelent *r;
|
|
asymbol *s;
|
|
fprintf (stderr, "relocs for sec %s\n", sec->name);
|
|
for (i = 0; i < n; i++)
|
|
{
|
|
r = relocs[i];
|
|
s = *r->sym_ptr_ptr;
|
|
fprintf (stderr, " reloc %2d @%08x off %4x : sym %-10s addend %x\n",
|
|
i, r, r->address, s->name, r->addend);
|
|
}
|
|
}
|
|
#endif
|
|
}
|
|
|
|
static void
|
|
write_contents (abfd, sec, xxx)
|
|
bfd *abfd;
|
|
asection *sec;
|
|
PTR xxx;
|
|
{
|
|
segment_info_type *seginfo = seg_info (sec);
|
|
unsigned long offset = 0;
|
|
fragS *f;
|
|
|
|
/* Write out the frags. */
|
|
if (seginfo == NULL
|
|
|| ! (bfd_get_section_flags (abfd, sec) & SEC_HAS_CONTENTS))
|
|
return;
|
|
|
|
for (f = seginfo->frchainP->frch_root;
|
|
f;
|
|
f = f->fr_next)
|
|
{
|
|
int x;
|
|
unsigned long fill_size;
|
|
char *fill_literal;
|
|
long count;
|
|
|
|
assert (f->fr_type == rs_fill);
|
|
if (f->fr_fix)
|
|
{
|
|
x = bfd_set_section_contents (stdoutput, sec,
|
|
f->fr_literal, (file_ptr) offset,
|
|
(bfd_size_type) f->fr_fix);
|
|
if (x == false)
|
|
{
|
|
bfd_perror (stdoutput->filename);
|
|
as_perror ("FATAL: Can't write %s", stdoutput->filename);
|
|
exit (42);
|
|
}
|
|
offset += f->fr_fix;
|
|
}
|
|
fill_literal = f->fr_literal + f->fr_fix;
|
|
fill_size = f->fr_var;
|
|
count = f->fr_offset;
|
|
assert (count >= 0);
|
|
if (fill_size && count)
|
|
#ifdef BFD_FAST_SECTION_FILL
|
|
{
|
|
char buf[256];
|
|
if (fill_size > sizeof(buf)) {
|
|
/* Do it the old way. Can this ever happen? */
|
|
while (count--)
|
|
{
|
|
x = bfd_set_section_contents (stdoutput, sec,
|
|
fill_literal, (file_ptr) offset,
|
|
(bfd_size_type) fill_size);
|
|
if (x == false)
|
|
{
|
|
bfd_perror (stdoutput->filename);
|
|
as_perror ("FATAL: Can't write %s", stdoutput->filename);
|
|
exit (42);
|
|
}
|
|
offset += fill_size;
|
|
}
|
|
}
|
|
else {
|
|
/* Build a buffer full of fill objects and output it as
|
|
* often as necessary. This saves on the overhead of potentially
|
|
* lots of bfd_set_section_contents calls.
|
|
*/
|
|
int n_per_buf, bytes, i;
|
|
if (fill_size == 1)
|
|
{
|
|
n_per_buf = sizeof (buf);
|
|
memset (buf, *fill_literal, n_per_buf);
|
|
}
|
|
else
|
|
{
|
|
char *bufp;
|
|
n_per_buf = sizeof(buf)/fill_size;
|
|
for (i = n_per_buf, bufp = buf; i; i--, bufp += fill_size)
|
|
memcpy(bufp, fill_literal, fill_size);
|
|
}
|
|
for (; count > 0; count -= n_per_buf)
|
|
{
|
|
n_per_buf = n_per_buf > count ? count : n_per_buf;
|
|
x = bfd_set_section_contents (stdoutput, sec,
|
|
buf, (file_ptr) offset,
|
|
(bfd_size_type) n_per_buf * fill_size);
|
|
assert (x == true);
|
|
offset += n_per_buf * fill_size;
|
|
}
|
|
}
|
|
}
|
|
#else
|
|
while (count--)
|
|
{
|
|
x = bfd_set_section_contents (stdoutput, sec,
|
|
fill_literal, (file_ptr) offset,
|
|
(bfd_size_type) fill_size);
|
|
assert (x == true);
|
|
offset += fill_size;
|
|
}
|
|
#endif
|
|
}
|
|
}
|
|
#endif
|
|
|
|
#if defined(BFD_ASSEMBLER) || (!defined (BFD) && !defined(OBJ_AOUT))
|
|
static void
|
|
merge_data_into_text ()
|
|
{
|
|
#if defined(BFD_ASSEMBLER) || defined(MANY_SEGMENTS)
|
|
seg_info (text_section)->frchainP->frch_last->fr_next =
|
|
seg_info (data_section)->frchainP->frch_root;
|
|
seg_info (text_section)->frchainP->frch_last =
|
|
seg_info (data_section)->frchainP->frch_last;
|
|
seg_info (data_section)->frchainP = 0;
|
|
#else
|
|
fixS *tmp;
|
|
|
|
text_last_frag->fr_next = data_frag_root;
|
|
text_last_frag = data_last_frag;
|
|
data_last_frag = NULL;
|
|
data_frag_root = NULL;
|
|
if (text_fix_root)
|
|
{
|
|
for (tmp = text_fix_root; tmp->fx_next; tmp = tmp->fx_next);;
|
|
tmp->fx_next = data_fix_root;
|
|
text_fix_tail = data_fix_tail;
|
|
}
|
|
else
|
|
text_fix_root = data_fix_root;
|
|
data_fix_root = NULL;
|
|
#endif
|
|
}
|
|
#endif /* BFD_ASSEMBLER || (! BFD && ! OBJ_AOUT) */
|
|
|
|
#if !defined (BFD_ASSEMBLER) && !defined (BFD)
|
|
static void
|
|
relax_and_size_all_segments ()
|
|
{
|
|
fragS *fragP;
|
|
|
|
relax_segment (text_frag_root, SEG_TEXT);
|
|
relax_segment (data_frag_root, SEG_DATA);
|
|
relax_segment (bss_frag_root, SEG_BSS);
|
|
/*
|
|
* Now the addresses of frags are correct within the segment.
|
|
*/
|
|
|
|
know (text_last_frag->fr_type == rs_fill && text_last_frag->fr_offset == 0);
|
|
H_SET_TEXT_SIZE (&headers, text_last_frag->fr_address);
|
|
text_last_frag->fr_address = H_GET_TEXT_SIZE (&headers);
|
|
|
|
/*
|
|
* Join the 2 segments into 1 huge segment.
|
|
* To do this, re-compute every rn_address in the SEG_DATA frags.
|
|
* Then join the data frags after the text frags.
|
|
*
|
|
* Determine a_data [length of data segment].
|
|
*/
|
|
if (data_frag_root)
|
|
{
|
|
register relax_addressT slide;
|
|
|
|
know ((text_last_frag->fr_type == rs_fill) && (text_last_frag->fr_offset == 0));
|
|
|
|
H_SET_DATA_SIZE (&headers, data_last_frag->fr_address);
|
|
data_last_frag->fr_address = H_GET_DATA_SIZE (&headers);
|
|
slide = H_GET_TEXT_SIZE (&headers); /* & in file of the data segment. */
|
|
#ifdef OBJ_BOUT
|
|
#define RoundUp(N,S) (((N)+(S)-1)&-(S))
|
|
/* For b.out: If the data section has a strict alignment
|
|
requirement, its load address in the .o file will be
|
|
rounded up from the size of the text section. These
|
|
two values are *not* the same! Similarly for the bss
|
|
section.... */
|
|
slide = RoundUp (slide, 1 << section_alignment[SEG_DATA]);
|
|
#endif
|
|
|
|
for (fragP = data_frag_root; fragP; fragP = fragP->fr_next)
|
|
{
|
|
fragP->fr_address += slide;
|
|
} /* for each data frag */
|
|
|
|
know (text_last_frag != 0);
|
|
text_last_frag->fr_next = data_frag_root;
|
|
}
|
|
else
|
|
{
|
|
H_SET_DATA_SIZE (&headers, 0);
|
|
}
|
|
|
|
#ifdef OBJ_BOUT
|
|
/* See above comments on b.out data section address. */
|
|
{
|
|
long bss_vma;
|
|
if (data_last_frag == 0)
|
|
bss_vma = H_GET_TEXT_SIZE (&headers);
|
|
else
|
|
bss_vma = data_last_frag->fr_address;
|
|
bss_vma = RoundUp (bss_vma, 1 << section_alignment[SEG_BSS]);
|
|
bss_address_frag.fr_address = bss_vma;
|
|
}
|
|
#else /* ! OBJ_BOUT */
|
|
bss_address_frag.fr_address = (H_GET_TEXT_SIZE (&headers) +
|
|
H_GET_DATA_SIZE (&headers));
|
|
|
|
#endif /* ! OBJ_BOUT */
|
|
|
|
/* Slide all the frags */
|
|
if (bss_frag_root)
|
|
{
|
|
relax_addressT slide = bss_address_frag.fr_address;
|
|
|
|
for (fragP = bss_frag_root; fragP; fragP = fragP->fr_next)
|
|
{
|
|
fragP->fr_address += slide;
|
|
} /* for each bss frag */
|
|
}
|
|
|
|
if (bss_last_frag)
|
|
H_SET_BSS_SIZE (&headers,
|
|
bss_last_frag->fr_address - bss_frag_root->fr_address);
|
|
else
|
|
H_SET_BSS_SIZE (&headers, 0);
|
|
}
|
|
#endif /* ! BFD_ASSEMBLER && ! BFD */
|
|
|
|
#if defined (BFD_ASSEMBLER) || !defined (BFD)
|
|
|
|
#ifdef BFD_ASSEMBLER
|
|
static void
|
|
set_symtab ()
|
|
{
|
|
int nsyms;
|
|
asymbol **asympp;
|
|
symbolS *symp;
|
|
boolean result;
|
|
extern PTR bfd_alloc PARAMS ((bfd *, size_t));
|
|
|
|
/* Count symbols. We can't rely on a count made by the loop in
|
|
write_object_file, because *_frob_file may add a new symbol or
|
|
two. */
|
|
nsyms = 0;
|
|
for (symp = symbol_rootP; symp; symp = symbol_next (symp))
|
|
nsyms++;
|
|
|
|
if (nsyms)
|
|
{
|
|
int i;
|
|
|
|
asympp = (asymbol **) bfd_alloc (stdoutput,
|
|
nsyms * sizeof (asymbol *));
|
|
symp = symbol_rootP;
|
|
for (i = 0; i < nsyms; i++, symp = symbol_next (symp))
|
|
{
|
|
asympp[i] = symp->bsym;
|
|
symp->written = 1;
|
|
}
|
|
}
|
|
else
|
|
asympp = 0;
|
|
result = bfd_set_symtab (stdoutput, asympp, nsyms);
|
|
assert (result == true);
|
|
symbol_table_frozen = 1;
|
|
}
|
|
#endif
|
|
|
|
void
|
|
write_object_file ()
|
|
{
|
|
struct frchain *frchainP; /* Track along all frchains. */
|
|
#if ! defined (BFD_ASSEMBLER) || ! defined (WORKING_DOT_WORD)
|
|
fragS *fragP; /* Track along all frags. */
|
|
#endif
|
|
|
|
/* Do we really want to write it? */
|
|
{
|
|
int n_warns, n_errs;
|
|
n_warns = had_warnings ();
|
|
n_errs = had_errors ();
|
|
/* The -Z flag indicates that an object file should be generated,
|
|
regardless of warnings and errors. */
|
|
if (flag_always_generate_output)
|
|
{
|
|
if (n_warns || n_errs)
|
|
as_warn ("%d error%s, %d warning%s, generating bad object file.\n",
|
|
n_errs, n_errs == 1 ? "" : "s",
|
|
n_warns, n_warns == 1 ? "" : "s");
|
|
}
|
|
else
|
|
{
|
|
if (n_errs)
|
|
as_fatal ("%d error%s, %d warning%s, no object file generated.\n",
|
|
n_errs, n_errs == 1 ? "" : "s",
|
|
n_warns, n_warns == 1 ? "" : "s");
|
|
}
|
|
}
|
|
|
|
#ifdef OBJ_VMS
|
|
/* Under VMS we try to be compatible with VAX-11 "C". Thus, we call
|
|
a routine to check for the definition of the procedure "_main",
|
|
and if so -- fix it up so that it can be program entry point. */
|
|
VMS_Check_For_Main ();
|
|
#endif /* VMS */
|
|
|
|
/* After every sub-segment, we fake an ".align ...". This conforms to
|
|
BSD4.2 brane-damage. We then fake ".fill 0" because that is the kind of
|
|
frag that requires least thought. ".align" frags like to have a
|
|
following frag since that makes calculating their intended length
|
|
trivial.
|
|
|
|
@@ Is this really necessary?? */
|
|
#ifndef SUB_SEGMENT_ALIGN
|
|
#ifdef BFD_ASSEMBLER
|
|
#define SUB_SEGMENT_ALIGN(SEG) (0)
|
|
#else
|
|
#define SUB_SEGMENT_ALIGN(SEG) (2)
|
|
#endif
|
|
#endif
|
|
for (frchainP = frchain_root; frchainP; frchainP = frchainP->frch_next)
|
|
{
|
|
subseg_set (frchainP->frch_seg, frchainP->frch_subseg);
|
|
frag_align (SUB_SEGMENT_ALIGN (now_seg), NOP_OPCODE);
|
|
/* frag_align will have left a new frag.
|
|
Use this last frag for an empty ".fill".
|
|
|
|
For this segment ...
|
|
Create a last frag. Do not leave a "being filled in frag". */
|
|
frag_wane (frag_now);
|
|
frag_now->fr_fix = 0;
|
|
know (frag_now->fr_next == NULL);
|
|
}
|
|
|
|
/* From now on, we don't care about sub-segments. Build one frag chain
|
|
for each segment. Linked thru fr_next. */
|
|
|
|
#ifdef BFD_ASSEMBLER
|
|
/* Remove the sections created by gas for its own purposes. */
|
|
{
|
|
asection **seclist, *sec;
|
|
seclist = &stdoutput->sections;
|
|
while (seclist && *seclist)
|
|
{
|
|
sec = *seclist;
|
|
while (sec == reg_section || sec == expr_section)
|
|
{
|
|
sec = sec->next;
|
|
*seclist = sec;
|
|
stdoutput->section_count--;
|
|
if (!sec)
|
|
break;
|
|
}
|
|
if (*seclist)
|
|
seclist = &(*seclist)->next;
|
|
}
|
|
}
|
|
|
|
bfd_map_over_sections (stdoutput, chain_frchains_together, (char *) 0);
|
|
#else
|
|
remove_subsegs (frchain_root, SEG_TEXT, &text_frag_root, &text_last_frag);
|
|
remove_subsegs (data0_frchainP, SEG_DATA, &data_frag_root, &data_last_frag);
|
|
remove_subsegs (bss0_frchainP, SEG_BSS, &bss_frag_root, &bss_last_frag);
|
|
#endif
|
|
|
|
/* We have two segments. If user gave -R flag, then we must put the
|
|
data frags into the text segment. Do this before relaxing so
|
|
we know to take advantage of -R and make shorter addresses. */
|
|
#if !defined (OBJ_AOUT) || defined (BFD_ASSEMBLER)
|
|
if (flag_readonly_data_in_text)
|
|
{
|
|
merge_data_into_text ();
|
|
}
|
|
#endif
|
|
|
|
#ifdef BFD_ASSEMBLER
|
|
bfd_map_over_sections (stdoutput, relax_and_size_seg, (char *) 0);
|
|
#else
|
|
relax_and_size_all_segments ();
|
|
#endif /* BFD_ASSEMBLER */
|
|
|
|
#ifndef BFD_ASSEMBLER
|
|
/*
|
|
*
|
|
* Crawl the symbol chain.
|
|
*
|
|
* For each symbol whose value depends on a frag, take the address of
|
|
* that frag and subsume it into the value of the symbol.
|
|
* After this, there is just one way to lookup a symbol value.
|
|
* Values are left in their final state for object file emission.
|
|
* We adjust the values of 'L' local symbols, even if we do
|
|
* not intend to emit them to the object file, because their values
|
|
* are needed for fix-ups.
|
|
*
|
|
* Unless we saw a -L flag, remove all symbols that begin with 'L'
|
|
* from the symbol chain. (They are still pointed to by the fixes.)
|
|
*
|
|
* Count the remaining symbols.
|
|
* Assign a symbol number to each symbol.
|
|
* Count the number of string-table chars we will emit.
|
|
* Put this info into the headers as appropriate.
|
|
*
|
|
*/
|
|
know (zero_address_frag.fr_address == 0);
|
|
string_byte_count = sizeof (string_byte_count);
|
|
|
|
obj_crawl_symbol_chain (&headers);
|
|
|
|
if (string_byte_count == sizeof (string_byte_count))
|
|
string_byte_count = 0;
|
|
|
|
H_SET_STRING_SIZE (&headers, string_byte_count);
|
|
|
|
/*
|
|
* Addresses of frags now reflect addresses we use in the object file.
|
|
* Symbol values are correct.
|
|
* Scan the frags, converting any ".org"s and ".align"s to ".fill"s.
|
|
* Also converting any machine-dependent frags using md_convert_frag();
|
|
*/
|
|
subseg_change (SEG_TEXT, 0);
|
|
|
|
for (fragP = text_frag_root; fragP; fragP = fragP->fr_next)
|
|
{
|
|
cvt_frag_to_fill (&headers, fragP);
|
|
|
|
/* Some assert macros don't work with # directives mixed in. */
|
|
#ifndef NDEBUG
|
|
if (!(fragP->fr_next == NULL
|
|
#ifdef OBJ_BOUT
|
|
|| fragP->fr_next == data_frag_root
|
|
#endif
|
|
|| ((fragP->fr_next->fr_address - fragP->fr_address)
|
|
== (fragP->fr_fix + fragP->fr_offset * fragP->fr_var))))
|
|
abort ();
|
|
#endif
|
|
}
|
|
#endif /* ! BFD_ASSEMBLER */
|
|
|
|
#ifndef WORKING_DOT_WORD
|
|
{
|
|
struct broken_word *lie;
|
|
struct broken_word **prevP;
|
|
|
|
prevP = &broken_words;
|
|
for (lie = broken_words; lie; lie = lie->next_broken_word)
|
|
if (!lie->added)
|
|
{
|
|
expressionS exp;
|
|
|
|
exp.X_op = O_subtract;
|
|
exp.X_add_symbol = lie->add;
|
|
exp.X_op_symbol = lie->sub;
|
|
exp.X_add_number = lie->addnum;
|
|
#ifdef BFD_ASSEMBLER
|
|
#ifdef TC_CONS_FIX_NEW
|
|
TC_CONS_FIX_NEW (lie->frag,
|
|
lie->word_goes_here - lie->frag->fr_literal,
|
|
2, &exp);
|
|
#else
|
|
fix_new_exp (lie->frag,
|
|
lie->word_goes_here - lie->frag->fr_literal,
|
|
2, &exp, 0, BFD_RELOC_NONE);
|
|
#endif
|
|
#else
|
|
#if defined(TC_SPARC) || defined(TC_A29K) || defined(NEED_FX_R_TYPE)
|
|
fix_new_exp (lie->frag,
|
|
lie->word_goes_here - lie->frag->fr_literal,
|
|
2, &exp, 0, NO_RELOC);
|
|
#else
|
|
#ifdef TC_NS32K
|
|
fix_new_ns32k_exp (lie->frag,
|
|
lie->word_goes_here - lie->frag->fr_literal,
|
|
2, &exp, 0, 0, 2, 0, 0);
|
|
#else
|
|
fix_new_exp (lie->frag,
|
|
lie->word_goes_here - lie->frag->fr_literal,
|
|
2, &exp, 0, 0);
|
|
#endif /* TC_NS32K */
|
|
#endif /* TC_SPARC|TC_A29K|NEED_FX_R_TYPE */
|
|
#endif /* BFD_ASSEMBLER */
|
|
*prevP = lie->next_broken_word;
|
|
}
|
|
else
|
|
prevP = &(lie->next_broken_word);
|
|
|
|
for (lie = broken_words; lie;)
|
|
{
|
|
struct broken_word *untruth;
|
|
char *table_ptr;
|
|
addressT table_addr;
|
|
addressT from_addr, to_addr;
|
|
int n, m;
|
|
|
|
fragP = lie->dispfrag;
|
|
|
|
/* Find out how many broken_words go here. */
|
|
n = 0;
|
|
for (untruth = lie; untruth && untruth->dispfrag == fragP; untruth = untruth->next_broken_word)
|
|
if (untruth->added == 1)
|
|
n++;
|
|
|
|
table_ptr = lie->dispfrag->fr_opcode;
|
|
table_addr = lie->dispfrag->fr_address + (table_ptr - lie->dispfrag->fr_literal);
|
|
/* Create the jump around the long jumps. This is a short
|
|
jump from table_ptr+0 to table_ptr+n*long_jump_size. */
|
|
from_addr = table_addr;
|
|
to_addr = table_addr + md_short_jump_size + n * md_long_jump_size;
|
|
md_create_short_jump (table_ptr, from_addr, to_addr, lie->dispfrag, lie->add);
|
|
table_ptr += md_short_jump_size;
|
|
table_addr += md_short_jump_size;
|
|
|
|
for (m = 0; lie && lie->dispfrag == fragP; m++, lie = lie->next_broken_word)
|
|
{
|
|
if (lie->added == 2)
|
|
continue;
|
|
/* Patch the jump table */
|
|
/* This is the offset from ??? to table_ptr+0 */
|
|
to_addr = table_addr - S_GET_VALUE (lie->sub);
|
|
#ifdef BFD_ASSEMBLER
|
|
to_addr -= lie->sub->sy_frag->fr_address;
|
|
#endif
|
|
md_number_to_chars (lie->word_goes_here, to_addr, 2);
|
|
for (untruth = lie->next_broken_word; untruth && untruth->dispfrag == fragP; untruth = untruth->next_broken_word)
|
|
{
|
|
if (untruth->use_jump == lie)
|
|
md_number_to_chars (untruth->word_goes_here, to_addr, 2);
|
|
}
|
|
|
|
/* Install the long jump */
|
|
/* this is a long jump from table_ptr+0 to the final target */
|
|
from_addr = table_addr;
|
|
to_addr = S_GET_VALUE (lie->add) + lie->addnum;
|
|
#ifdef BFD_ASSEMBLER
|
|
to_addr += lie->add->sy_frag->fr_address;
|
|
#endif
|
|
md_create_long_jump (table_ptr, from_addr, to_addr, lie->dispfrag, lie->add);
|
|
table_ptr += md_long_jump_size;
|
|
table_addr += md_long_jump_size;
|
|
}
|
|
}
|
|
}
|
|
#endif /* not WORKING_DOT_WORD */
|
|
|
|
#ifndef BFD_ASSEMBLER
|
|
#ifndef OBJ_VMS
|
|
{ /* not vms */
|
|
long object_file_size;
|
|
/*
|
|
* Scan every FixS performing fixups. We had to wait until now to do
|
|
* this because md_convert_frag() may have made some fixSs.
|
|
*/
|
|
int trsize, drsize;
|
|
|
|
subseg_change (SEG_TEXT, 0);
|
|
trsize = md_reloc_size * fixup_segment (text_fix_root, SEG_TEXT);
|
|
subseg_change (SEG_DATA, 0);
|
|
drsize = md_reloc_size * fixup_segment (data_fix_root, SEG_DATA);
|
|
H_SET_RELOCATION_SIZE (&headers, trsize, drsize);
|
|
|
|
/* FIXME move this stuff into the pre-write-hook */
|
|
H_SET_MAGIC_NUMBER (&headers, magic_number_for_object_file);
|
|
H_SET_ENTRY_POINT (&headers, 0);
|
|
|
|
obj_pre_write_hook (&headers); /* extra coff stuff */
|
|
|
|
object_file_size = H_GET_FILE_SIZE (&headers);
|
|
next_object_file_charP = the_object_file = xmalloc (object_file_size);
|
|
|
|
output_file_create (out_file_name);
|
|
|
|
obj_header_append (&next_object_file_charP, &headers);
|
|
|
|
know ((next_object_file_charP - the_object_file) == H_GET_HEADER_SIZE (&headers));
|
|
|
|
/*
|
|
* Emit code.
|
|
*/
|
|
for (fragP = text_frag_root; fragP; fragP = fragP->fr_next)
|
|
{
|
|
register long count;
|
|
register char *fill_literal;
|
|
register long fill_size;
|
|
|
|
know (fragP->fr_type == rs_fill);
|
|
append (&next_object_file_charP, fragP->fr_literal, (unsigned long) fragP->fr_fix);
|
|
fill_literal = fragP->fr_literal + fragP->fr_fix;
|
|
fill_size = fragP->fr_var;
|
|
know (fragP->fr_offset >= 0);
|
|
|
|
for (count = fragP->fr_offset; count; count--)
|
|
{
|
|
append (&next_object_file_charP, fill_literal, (unsigned long) fill_size);
|
|
} /* for each */
|
|
|
|
} /* for each code frag. */
|
|
|
|
know ((next_object_file_charP - the_object_file) == (H_GET_HEADER_SIZE (&headers) + H_GET_TEXT_SIZE (&headers) + H_GET_DATA_SIZE (&headers)));
|
|
|
|
/*
|
|
* Emit relocations.
|
|
*/
|
|
obj_emit_relocations (&next_object_file_charP, text_fix_root, (relax_addressT) 0);
|
|
know ((next_object_file_charP - the_object_file) == (H_GET_HEADER_SIZE (&headers) + H_GET_TEXT_SIZE (&headers) + H_GET_DATA_SIZE (&headers) + H_GET_TEXT_RELOCATION_SIZE (&headers)));
|
|
#ifdef TC_I960
|
|
/* Make addresses in data relocation directives relative to beginning of
|
|
* first data fragment, not end of last text fragment: alignment of the
|
|
* start of the data segment may place a gap between the segments.
|
|
*/
|
|
obj_emit_relocations (&next_object_file_charP, data_fix_root, data0_frchainP->frch_root->fr_address);
|
|
#else /* TC_I960 */
|
|
obj_emit_relocations (&next_object_file_charP, data_fix_root, text_last_frag->fr_address);
|
|
#endif /* TC_I960 */
|
|
|
|
know ((next_object_file_charP - the_object_file) == (H_GET_HEADER_SIZE (&headers) + H_GET_TEXT_SIZE (&headers) + H_GET_DATA_SIZE (&headers) + H_GET_TEXT_RELOCATION_SIZE (&headers) + H_GET_DATA_RELOCATION_SIZE (&headers)));
|
|
|
|
/*
|
|
* Emit line number entries.
|
|
*/
|
|
OBJ_EMIT_LINENO (&next_object_file_charP, lineno_rootP, the_object_file);
|
|
know ((next_object_file_charP - the_object_file) == (H_GET_HEADER_SIZE (&headers) + H_GET_TEXT_SIZE (&headers) + H_GET_DATA_SIZE (&headers) + H_GET_TEXT_RELOCATION_SIZE (&headers) + H_GET_DATA_RELOCATION_SIZE (&headers) + H_GET_LINENO_SIZE (&headers)));
|
|
|
|
/*
|
|
* Emit symbols.
|
|
*/
|
|
obj_emit_symbols (&next_object_file_charP, symbol_rootP);
|
|
know ((next_object_file_charP - the_object_file) == (H_GET_HEADER_SIZE (&headers) + H_GET_TEXT_SIZE (&headers) + H_GET_DATA_SIZE (&headers) + H_GET_TEXT_RELOCATION_SIZE (&headers) + H_GET_DATA_RELOCATION_SIZE (&headers) + H_GET_LINENO_SIZE (&headers) + H_GET_SYMBOL_TABLE_SIZE (&headers)));
|
|
|
|
/*
|
|
* Emit strings.
|
|
*/
|
|
|
|
if (string_byte_count > 0)
|
|
{
|
|
obj_emit_strings (&next_object_file_charP);
|
|
} /* only if we have a string table */
|
|
|
|
#ifdef BFD_HEADERS
|
|
bfd_seek (stdoutput, 0, 0);
|
|
bfd_write (the_object_file, 1, object_file_size, stdoutput);
|
|
#else
|
|
|
|
/* Write the data to the file */
|
|
output_file_append (the_object_file, object_file_size, out_file_name);
|
|
#endif
|
|
} /* non vms output */
|
|
#else /* VMS */
|
|
/*
|
|
* Now do the VMS-dependent part of writing the object file
|
|
*/
|
|
VMS_write_object_file (H_GET_TEXT_SIZE (&headers),
|
|
H_GET_DATA_SIZE (&headers),
|
|
H_GET_BSS_SIZE (&headers),
|
|
text_frag_root, data_frag_root);
|
|
#endif /* VMS */
|
|
#else /* BFD_ASSEMBLER */
|
|
|
|
bfd_map_over_sections (stdoutput, adjust_reloc_syms, (char *)0);
|
|
|
|
/* Set up symbol table, and write it out. */
|
|
if (symbol_rootP)
|
|
{
|
|
symbolS *symp;
|
|
|
|
for (symp = symbol_rootP; symp; symp = symbol_next (symp))
|
|
{
|
|
int punt = 0;
|
|
|
|
if (! symp->sy_resolved)
|
|
{
|
|
if (symp->sy_value.X_op == O_constant)
|
|
{
|
|
/* This is the normal case; skip the call. */
|
|
S_SET_VALUE (symp,
|
|
(S_GET_VALUE (symp)
|
|
+ symp->sy_frag->fr_address));
|
|
symp->sy_resolved = 1;
|
|
}
|
|
else
|
|
resolve_symbol_value (symp);
|
|
}
|
|
|
|
/* So far, common symbols have been treated like undefined symbols.
|
|
Put them in the common section now. */
|
|
if (S_IS_DEFINED (symp) == 0
|
|
&& S_GET_VALUE (symp) != 0)
|
|
S_SET_SEGMENT (symp, bfd_com_section_ptr);
|
|
#if 0
|
|
printf ("symbol `%s'\n\t@%x: value=%d flags=%x seg=%s\n",
|
|
S_GET_NAME (symp), symp,
|
|
S_GET_VALUE (symp),
|
|
symp->bsym->flags,
|
|
segment_name (symp->bsym->section));
|
|
#endif
|
|
|
|
#ifdef obj_frob_symbol
|
|
obj_frob_symbol (symp, punt);
|
|
#endif
|
|
#ifdef tc_frob_symbol
|
|
if (! punt || symp->sy_used_in_reloc)
|
|
tc_frob_symbol (symp, punt);
|
|
#endif
|
|
|
|
/* If we don't want to keep this symbol, splice it out of
|
|
the chain now. If EMIT_SECTION_SYMBOLS is 0, we never
|
|
want section symbols. Otherwise, we skip local symbols
|
|
and symbols that the frob_symbol macros told us to punt,
|
|
but we keep such symbols if they are used in relocs. */
|
|
if ((! EMIT_SECTION_SYMBOLS
|
|
&& (symp->bsym->flags & BSF_SECTION_SYM) != 0)
|
|
/* Note that S_IS_EXTERN and S_IS_LOCAL are not always
|
|
opposites. Sometimes the former checks flags and the
|
|
latter examines the name... */
|
|
|| (!S_IS_EXTERN (symp)
|
|
&& (S_IS_LOCAL (symp) || punt)
|
|
&& ! symp->sy_used_in_reloc))
|
|
{
|
|
symbol_remove (symp, &symbol_rootP, &symbol_lastP);
|
|
/* After symbol_remove, symbol_next(symp) still returns
|
|
the one that came after it in the chain. So we don't
|
|
need to do any extra cleanup work here. */
|
|
|
|
continue;
|
|
}
|
|
|
|
/* Make sure we really got a value for the symbol. */
|
|
if (! symp->sy_resolved)
|
|
{
|
|
as_bad ("can't resolve value for symbol \"%s\"",
|
|
S_GET_NAME (symp));
|
|
symp->sy_resolved = 1;
|
|
}
|
|
|
|
/* Set the value into the BFD symbol. Up til now the value
|
|
has only been kept in the gas symbolS struct. */
|
|
symp->bsym->value = S_GET_VALUE (symp);
|
|
}
|
|
}
|
|
|
|
/* Now do any format-specific adjustments to the symbol table, such
|
|
as adding file symbols. */
|
|
#ifdef obj_adjust_symtab
|
|
obj_adjust_symtab ();
|
|
#endif
|
|
|
|
/* Now that all the sizes are known, and contents correct, we can
|
|
start writing to the file. */
|
|
set_symtab ();
|
|
|
|
/* If *_frob_file changes the symbol value at this point, it is
|
|
responsible for moving the changed value into symp->bsym->value
|
|
as well. Hopefully all symbol value changing can be done in
|
|
*_frob_symbol. */
|
|
#ifdef tc_frob_file
|
|
tc_frob_file ();
|
|
#endif
|
|
#ifdef obj_frob_file
|
|
obj_frob_file ();
|
|
#endif
|
|
|
|
bfd_map_over_sections (stdoutput, write_relocs, (char *) 0);
|
|
|
|
bfd_map_over_sections (stdoutput, write_contents, (char *) 0);
|
|
#endif /* BFD_ASSEMBLER */
|
|
}
|
|
#endif /* ! BFD */
|
|
|
|
/*
|
|
* relax_segment()
|
|
*
|
|
* Now we have a segment, not a crowd of sub-segments, we can make fr_address
|
|
* values.
|
|
*
|
|
* Relax the frags.
|
|
*
|
|
* After this, all frags in this segment have addresses that are correct
|
|
* within the segment. Since segments live in different file addresses,
|
|
* these frag addresses may not be the same as final object-file addresses.
|
|
*/
|
|
|
|
#ifndef md_relax_frag
|
|
|
|
/* Subroutines of relax_segment. */
|
|
static int
|
|
is_dnrange (f1, f2)
|
|
struct frag *f1;
|
|
struct frag *f2;
|
|
{
|
|
for (; f1; f1 = f1->fr_next)
|
|
if (f1->fr_next == f2)
|
|
return 1;
|
|
return 0;
|
|
}
|
|
|
|
#endif /* ! defined (md_relax_frag) */
|
|
|
|
/* Relax_align. Advance location counter to next address that has 'alignment'
|
|
lowest order bits all 0s, return size of adjustment made. */
|
|
static relax_addressT
|
|
relax_align (address, alignment)
|
|
register relax_addressT address; /* Address now. */
|
|
register int alignment; /* Alignment (binary). */
|
|
{
|
|
relax_addressT mask;
|
|
relax_addressT new_address;
|
|
|
|
mask = ~((~0) << alignment);
|
|
new_address = (address + mask) & (~mask);
|
|
if (linkrelax)
|
|
/* We must provide lots of padding, so the linker can discard it
|
|
when needed. The linker will not add extra space, ever. */
|
|
new_address += (1 << alignment);
|
|
return (new_address - address);
|
|
}
|
|
|
|
void
|
|
relax_segment (segment_frag_root, segment)
|
|
struct frag *segment_frag_root;
|
|
segT segment;
|
|
{
|
|
register struct frag *fragP;
|
|
register relax_addressT address;
|
|
#if !defined (MANY_SEGMENTS) && !defined (BFD_ASSEMBLER)
|
|
know (segment == SEG_DATA || segment == SEG_TEXT || segment == SEG_BSS);
|
|
#endif
|
|
/* In case md_estimate_size_before_relax() wants to make fixSs. */
|
|
subseg_change (segment, 0);
|
|
|
|
/* For each frag in segment: count and store (a 1st guess of)
|
|
fr_address. */
|
|
address = 0;
|
|
for (fragP = segment_frag_root; fragP; fragP = fragP->fr_next)
|
|
{
|
|
fragP->fr_address = address;
|
|
address += fragP->fr_fix;
|
|
|
|
switch (fragP->fr_type)
|
|
{
|
|
case rs_fill:
|
|
address += fragP->fr_offset * fragP->fr_var;
|
|
break;
|
|
|
|
case rs_align:
|
|
{
|
|
int offset = relax_align (address, (int) fragP->fr_offset);
|
|
if (offset % fragP->fr_var != 0)
|
|
{
|
|
as_bad ("alignment padding (%d bytes) not a multiple of %ld",
|
|
offset, (long) fragP->fr_var);
|
|
offset -= (offset % fragP->fr_var);
|
|
}
|
|
address += offset;
|
|
}
|
|
break;
|
|
|
|
case rs_org:
|
|
/* Assume .org is nugatory. It will grow with 1st relax. */
|
|
break;
|
|
|
|
case rs_machine_dependent:
|
|
address += md_estimate_size_before_relax (fragP, segment);
|
|
break;
|
|
|
|
#ifndef WORKING_DOT_WORD
|
|
/* Broken words don't concern us yet */
|
|
case rs_broken_word:
|
|
break;
|
|
#endif
|
|
|
|
default:
|
|
BAD_CASE (fragP->fr_type);
|
|
break;
|
|
} /* switch(fr_type) */
|
|
} /* for each frag in the segment */
|
|
|
|
/* Do relax(). */
|
|
{
|
|
long stretch; /* May be any size, 0 or negative. */
|
|
/* Cumulative number of addresses we have */
|
|
/* relaxed this pass. */
|
|
/* We may have relaxed more than one address. */
|
|
long stretched; /* Have we stretched on this pass? */
|
|
/* This is 'cuz stretch may be zero, when, in fact some piece of code
|
|
grew, and another shrank. If a branch instruction doesn't fit anymore,
|
|
we could be scrod. */
|
|
|
|
do
|
|
{
|
|
stretch = stretched = 0;
|
|
for (fragP = segment_frag_root; fragP; fragP = fragP->fr_next)
|
|
{
|
|
long growth = 0;
|
|
unsigned long was_address;
|
|
long offset;
|
|
symbolS *symbolP;
|
|
long target;
|
|
long after;
|
|
|
|
was_address = fragP->fr_address;
|
|
address = fragP->fr_address += stretch;
|
|
symbolP = fragP->fr_symbol;
|
|
offset = fragP->fr_offset;
|
|
|
|
switch (fragP->fr_type)
|
|
{
|
|
case rs_fill: /* .fill never relaxes. */
|
|
growth = 0;
|
|
break;
|
|
|
|
#ifndef WORKING_DOT_WORD
|
|
/* JF: This is RMS's idea. I do *NOT* want to be blamed
|
|
for it I do not want to write it. I do not want to have
|
|
anything to do with it. This is not the proper way to
|
|
implement this misfeature. */
|
|
case rs_broken_word:
|
|
{
|
|
struct broken_word *lie;
|
|
struct broken_word *untruth;
|
|
|
|
/* Yes this is ugly (storing the broken_word pointer
|
|
in the symbol slot). Still, this whole chunk of
|
|
code is ugly, and I don't feel like doing anything
|
|
about it. Think of it as stubbornness in action. */
|
|
growth = 0;
|
|
for (lie = (struct broken_word *) (fragP->fr_symbol);
|
|
lie && lie->dispfrag == fragP;
|
|
lie = lie->next_broken_word)
|
|
{
|
|
|
|
if (lie->added)
|
|
continue;
|
|
|
|
offset = (lie->add->sy_frag->fr_address
|
|
+ S_GET_VALUE (lie->add)
|
|
+ lie->addnum
|
|
- (lie->sub->sy_frag->fr_address
|
|
+ S_GET_VALUE (lie->sub)));
|
|
if (offset <= -32768 || offset >= 32767)
|
|
{
|
|
if (flag_warn_displacement)
|
|
{
|
|
char buf[50];
|
|
sprint_value (buf, (addressT) lie->addnum);
|
|
as_warn (".word %s-%s+%s didn't fit",
|
|
S_GET_NAME (lie->add),
|
|
S_GET_NAME (lie->sub),
|
|
buf);
|
|
}
|
|
lie->added = 1;
|
|
if (fragP->fr_subtype == 0)
|
|
{
|
|
fragP->fr_subtype++;
|
|
growth += md_short_jump_size;
|
|
}
|
|
for (untruth = lie->next_broken_word;
|
|
untruth && untruth->dispfrag == lie->dispfrag;
|
|
untruth = untruth->next_broken_word)
|
|
if ((untruth->add->sy_frag == lie->add->sy_frag)
|
|
&& S_GET_VALUE (untruth->add) == S_GET_VALUE (lie->add))
|
|
{
|
|
untruth->added = 2;
|
|
untruth->use_jump = lie;
|
|
}
|
|
growth += md_long_jump_size;
|
|
}
|
|
}
|
|
|
|
break;
|
|
} /* case rs_broken_word */
|
|
#endif
|
|
case rs_align:
|
|
growth = (relax_align ((relax_addressT) (address
|
|
+ fragP->fr_fix),
|
|
(int) offset)
|
|
- relax_align ((relax_addressT) (was_address
|
|
+ fragP->fr_fix),
|
|
(int) offset));
|
|
break;
|
|
|
|
case rs_org:
|
|
target = offset;
|
|
|
|
if (symbolP)
|
|
{
|
|
#if !defined (MANY_SEGMENTS) && !defined (BFD_ASSEMBLER)
|
|
know ((S_GET_SEGMENT (symbolP) == SEG_ABSOLUTE)
|
|
|| (S_GET_SEGMENT (symbolP) == SEG_DATA)
|
|
|| (S_GET_SEGMENT (symbolP) == SEG_TEXT)
|
|
|| S_GET_SEGMENT (symbolP) == SEG_BSS);
|
|
know (symbolP->sy_frag);
|
|
know (!(S_GET_SEGMENT (symbolP) == SEG_ABSOLUTE)
|
|
|| (symbolP->sy_frag == &zero_address_frag));
|
|
#endif
|
|
target += S_GET_VALUE (symbolP)
|
|
+ symbolP->sy_frag->fr_address;
|
|
} /* if we have a symbol */
|
|
|
|
know (fragP->fr_next);
|
|
after = fragP->fr_next->fr_address;
|
|
growth = ((target - after) > 0) ? (target - after) : 0;
|
|
/* Growth may be negative, but variable part of frag
|
|
cannot have fewer than 0 chars. That is, we can't
|
|
.org backwards. */
|
|
|
|
growth -= stretch; /* This is an absolute growth factor */
|
|
break;
|
|
|
|
case rs_machine_dependent:
|
|
#ifdef md_relax_frag
|
|
growth = md_relax_frag (fragP, stretch);
|
|
#else
|
|
/* The default way to relax a frag is to look through
|
|
md_relax_table. */
|
|
{
|
|
const relax_typeS *this_type;
|
|
const relax_typeS *start_type;
|
|
relax_substateT next_state;
|
|
relax_substateT this_state;
|
|
long aim;
|
|
|
|
this_state = fragP->fr_subtype;
|
|
start_type = this_type = md_relax_table + this_state;
|
|
target = offset;
|
|
|
|
if (symbolP)
|
|
{
|
|
#ifndef DIFF_EXPR_OK
|
|
#if !defined (MANY_SEGMENTS) && !defined (BFD_ASSEMBLER)
|
|
know ((S_GET_SEGMENT (symbolP) == SEG_ABSOLUTE)
|
|
|| (S_GET_SEGMENT (symbolP) == SEG_DATA)
|
|
|| (S_GET_SEGMENT (symbolP) == SEG_BSS)
|
|
|| (S_GET_SEGMENT (symbolP) == SEG_TEXT));
|
|
#endif
|
|
know (symbolP->sy_frag);
|
|
#endif
|
|
know (!(S_GET_SEGMENT (symbolP) == absolute_section)
|
|
|| symbolP->sy_frag == &zero_address_frag);
|
|
target +=
|
|
S_GET_VALUE (symbolP)
|
|
+ symbolP->sy_frag->fr_address;
|
|
|
|
/* If frag has yet to be reached on this pass,
|
|
assume it will move by STRETCH just as we did.
|
|
If this is not so, it will be because some frag
|
|
between grows, and that will force another pass.
|
|
|
|
Beware zero-length frags.
|
|
|
|
There should be a faster way to do this. */
|
|
|
|
if (symbolP->sy_frag->fr_address >= was_address
|
|
&& is_dnrange (fragP, symbolP->sy_frag))
|
|
{
|
|
target += stretch;
|
|
}
|
|
}
|
|
|
|
aim = target - address - fragP->fr_fix;
|
|
#ifdef TC_PCREL_ADJUST
|
|
/* Currently only the ns32k family needs this */
|
|
aim += TC_PCREL_ADJUST(fragP);
|
|
#else
|
|
/* This machine doesn't want to use pcrel_adjust.
|
|
In that case, pcrel_adjust should be zero. */
|
|
assert (fragP->fr_pcrel_adjust == 0);
|
|
#endif
|
|
|
|
if (aim < 0)
|
|
{
|
|
/* Look backwards. */
|
|
for (next_state = this_type->rlx_more; next_state;)
|
|
if (aim >= this_type->rlx_backward)
|
|
next_state = 0;
|
|
else
|
|
{
|
|
/* Grow to next state. */
|
|
this_state = next_state;
|
|
this_type = md_relax_table + this_state;
|
|
next_state = this_type->rlx_more;
|
|
}
|
|
}
|
|
else
|
|
{
|
|
#ifdef M68K_AIM_KLUDGE
|
|
M68K_AIM_KLUDGE (aim, this_state, this_type);
|
|
#endif
|
|
/* Look forwards. */
|
|
for (next_state = this_type->rlx_more; next_state;)
|
|
if (aim <= this_type->rlx_forward)
|
|
next_state = 0;
|
|
else
|
|
{
|
|
/* Grow to next state. */
|
|
this_state = next_state;
|
|
this_type = md_relax_table + this_state;
|
|
next_state = this_type->rlx_more;
|
|
}
|
|
}
|
|
|
|
growth = this_type->rlx_length - start_type->rlx_length;
|
|
if (growth != 0)
|
|
fragP->fr_subtype = this_state;
|
|
}
|
|
#endif
|
|
break;
|
|
|
|
default:
|
|
BAD_CASE (fragP->fr_type);
|
|
break;
|
|
}
|
|
if (growth)
|
|
{
|
|
stretch += growth;
|
|
stretched++;
|
|
}
|
|
} /* For each frag in the segment. */
|
|
}
|
|
while (stretched); /* Until nothing further to relax. */
|
|
} /* do_relax */
|
|
|
|
/*
|
|
* We now have valid fr_address'es for each frag.
|
|
*/
|
|
|
|
/*
|
|
* All fr_address's are correct, relative to their own segment.
|
|
* We have made all the fixS we will ever make.
|
|
*/
|
|
} /* relax_segment() */
|
|
|
|
#if defined (BFD_ASSEMBLER) || !defined (BFD)
|
|
|
|
#ifndef TC_RELOC_RTSYM_LOC_FIXUP
|
|
#define TC_RELOC_RTSYM_LOC_FIXUP(X) (1)
|
|
#endif
|
|
|
|
/* fixup_segment()
|
|
|
|
Go through all the fixS's in a segment and see which ones can be
|
|
handled now. (These consist of fixS where we have since discovered
|
|
the value of a symbol, or the address of the frag involved.)
|
|
For each one, call md_apply_fix to put the fix into the frag data.
|
|
|
|
Result is a count of how many relocation structs will be needed to
|
|
handle the remaining fixS's that we couldn't completely handle here.
|
|
These will be output later by emit_relocations(). */
|
|
|
|
static long
|
|
fixup_segment (fixP, this_segment_type)
|
|
register fixS *fixP;
|
|
segT this_segment_type; /* N_TYPE bits for segment. */
|
|
{
|
|
long seg_reloc_count = 0;
|
|
symbolS *add_symbolP;
|
|
symbolS *sub_symbolP;
|
|
valueT add_number;
|
|
int size;
|
|
char *place;
|
|
long where;
|
|
char pcrel;
|
|
fragS *fragP;
|
|
segT add_symbol_segment = absolute_section;
|
|
|
|
/* If the linker is doing the relaxing, we must not do any fixups.
|
|
|
|
Well, strictly speaking that's not true -- we could do any that are
|
|
PC-relative and don't cross regions that could change size. And for the
|
|
i960 (the only machine for which we've got a relaxing linker right now),
|
|
we might be able to turn callx/callj into bal anyways in cases where we
|
|
know the maximum displacement. */
|
|
if (linkrelax)
|
|
{
|
|
for (; fixP; fixP = fixP->fx_next)
|
|
seg_reloc_count++;
|
|
TC_ADJUST_RELOC_COUNT (fixP, seg_reloc_count);
|
|
return seg_reloc_count;
|
|
}
|
|
|
|
for (; fixP; fixP = fixP->fx_next)
|
|
{
|
|
fragP = fixP->fx_frag;
|
|
know (fragP);
|
|
where = fixP->fx_where;
|
|
place = fragP->fr_literal + where;
|
|
size = fixP->fx_size;
|
|
add_symbolP = fixP->fx_addsy;
|
|
#ifdef TC_VALIDATE_FIX
|
|
TC_VALIDATE_FIX (fixP, this_segment_type, skip);
|
|
#endif
|
|
sub_symbolP = fixP->fx_subsy;
|
|
add_number = fixP->fx_offset;
|
|
pcrel = fixP->fx_pcrel;
|
|
|
|
if (add_symbolP)
|
|
add_symbol_segment = S_GET_SEGMENT (add_symbolP);
|
|
|
|
if (sub_symbolP)
|
|
{
|
|
if (!add_symbolP)
|
|
{
|
|
/* Its just -sym */
|
|
if (S_GET_SEGMENT (sub_symbolP) == absolute_section)
|
|
add_number -= S_GET_VALUE (sub_symbolP);
|
|
else if (pcrel
|
|
&& S_GET_SEGMENT (sub_symbolP) == this_segment_type)
|
|
{
|
|
/* Should try converting to a constant. */
|
|
goto bad_sub_reloc;
|
|
}
|
|
else
|
|
bad_sub_reloc:
|
|
as_bad_where (fixP->fx_file, fixP->fx_line,
|
|
"Negative of non-absolute symbol %s",
|
|
S_GET_NAME (sub_symbolP));
|
|
}
|
|
else if ((S_GET_SEGMENT (sub_symbolP) == add_symbol_segment)
|
|
&& (SEG_NORMAL (add_symbol_segment)
|
|
|| (add_symbol_segment == absolute_section)))
|
|
{
|
|
/* Difference of 2 symbols from same segment.
|
|
Can't make difference of 2 undefineds: 'value' means
|
|
something different for N_UNDF. */
|
|
#ifdef TC_I960
|
|
/* Makes no sense to use the difference of 2 arbitrary symbols
|
|
as the target of a call instruction. */
|
|
if (fixP->fx_tcbit)
|
|
as_bad_where (fixP->fx_file, fixP->fx_line,
|
|
"callj to difference of 2 symbols");
|
|
#endif /* TC_I960 */
|
|
add_number += S_GET_VALUE (add_symbolP) -
|
|
S_GET_VALUE (sub_symbolP);
|
|
|
|
add_symbolP = NULL;
|
|
|
|
/* Let the target machine make the final determination
|
|
as to whether or not a relocation will be needed to
|
|
handle this fixup. */
|
|
if (!TC_FORCE_RELOCATION (fixP))
|
|
{
|
|
fixP->fx_addsy = NULL;
|
|
}
|
|
}
|
|
else
|
|
{
|
|
/* Different segments in subtraction. */
|
|
know (!(S_IS_EXTERNAL (sub_symbolP)
|
|
&& (S_GET_SEGMENT (sub_symbolP) == absolute_section)));
|
|
|
|
if ((S_GET_SEGMENT (sub_symbolP) == absolute_section))
|
|
add_number -= S_GET_VALUE (sub_symbolP);
|
|
|
|
#ifdef DIFF_EXPR_OK
|
|
else if (S_GET_SEGMENT (sub_symbolP) == this_segment_type
|
|
#if 0 /* Do this even if it's already described as pc-relative. For example,
|
|
on the m68k, an operand of "pc@(foo-.-2)" should address "foo" in a
|
|
pc-relative mode. */
|
|
&& pcrel
|
|
#endif
|
|
)
|
|
{
|
|
/* Make it pc-relative. */
|
|
add_number += (md_pcrel_from (fixP)
|
|
- S_GET_VALUE (sub_symbolP));
|
|
pcrel = 1;
|
|
fixP->fx_pcrel = 1;
|
|
sub_symbolP = 0;
|
|
fixP->fx_subsy = 0;
|
|
}
|
|
#endif
|
|
#ifdef BFD_ASSEMBLER
|
|
else if (fixP->fx_r_type == BFD_RELOC_GPREL32
|
|
|| fixP->fx_r_type == BFD_RELOC_GPREL16)
|
|
{
|
|
/* Leave it alone. */
|
|
}
|
|
#endif
|
|
else
|
|
{
|
|
char buf[50];
|
|
sprint_value (buf, fragP->fr_address + where);
|
|
as_bad_where (fixP->fx_file, fixP->fx_line,
|
|
"Can't emit reloc {- %s-seg symbol \"%s\"} @ file address %s.",
|
|
segment_name (S_GET_SEGMENT (sub_symbolP)),
|
|
S_GET_NAME (sub_symbolP), buf);
|
|
}
|
|
}
|
|
}
|
|
|
|
if (add_symbolP)
|
|
{
|
|
if (add_symbol_segment == this_segment_type && pcrel
|
|
&& TC_RELOC_RTSYM_LOC_FIXUP (fixP->fx_r_type))
|
|
{
|
|
/*
|
|
* This fixup was made when the symbol's segment was
|
|
* SEG_UNKNOWN, but it is now in the local segment.
|
|
* So we know how to do the address without relocation.
|
|
*/
|
|
#ifdef TC_I960
|
|
/* reloc_callj() may replace a 'call' with a 'calls' or a
|
|
'bal', in which cases it modifies *fixP as appropriate.
|
|
In the case of a 'calls', no further work is required,
|
|
and *fixP has been set up to make the rest of the code
|
|
below a no-op. */
|
|
reloc_callj (fixP);
|
|
#endif /* TC_I960 */
|
|
|
|
add_number += S_GET_VALUE (add_symbolP);
|
|
add_number -= md_pcrel_from (fixP);
|
|
pcrel = 0; /* Lie. Don't want further pcrel processing. */
|
|
|
|
/* Let the target machine make the final determination
|
|
as to whether or not a relocation will be needed to
|
|
handle this fixup. */
|
|
if (!TC_FORCE_RELOCATION (fixP))
|
|
{
|
|
fixP->fx_pcrel = 0;
|
|
fixP->fx_addsy = NULL;
|
|
}
|
|
}
|
|
else
|
|
{
|
|
if (add_symbol_segment == absolute_section)
|
|
{
|
|
#ifdef TC_I960
|
|
/* See comment about reloc_callj() above. */
|
|
reloc_callj (fixP);
|
|
#endif /* TC_I960 */
|
|
add_number += S_GET_VALUE (add_symbolP);
|
|
|
|
/* Let the target machine make the final determination
|
|
as to whether or not a relocation will be needed to
|
|
handle this fixup. */
|
|
if (!TC_FORCE_RELOCATION (fixP))
|
|
{
|
|
fixP->fx_addsy = NULL;
|
|
add_symbolP = NULL;
|
|
}
|
|
}
|
|
else if (add_symbol_segment == undefined_section
|
|
#ifdef BFD_ASSEMBLER
|
|
|| bfd_is_com_section (add_symbol_segment)
|
|
#endif
|
|
)
|
|
{
|
|
#ifdef TC_I960
|
|
if ((int) fixP->fx_bit_fixP == 13)
|
|
{
|
|
/* This is a COBR instruction. They have only a
|
|
* 13-bit displacement and are only to be used
|
|
* for local branches: flag as error, don't generate
|
|
* relocation.
|
|
*/
|
|
as_bad_where (fixP->fx_file, fixP->fx_line,
|
|
"can't use COBR format with external label");
|
|
fixP->fx_addsy = NULL;
|
|
fixP->fx_done = 1;
|
|
continue;
|
|
} /* COBR */
|
|
#endif /* TC_I960 */
|
|
|
|
#ifdef OBJ_COFF
|
|
#ifdef TE_I386AIX
|
|
if (S_IS_COMMON (add_symbolP))
|
|
add_number += S_GET_VALUE (add_symbolP);
|
|
#endif /* TE_I386AIX */
|
|
#endif /* OBJ_COFF */
|
|
++seg_reloc_count;
|
|
}
|
|
else
|
|
{
|
|
seg_reloc_count++;
|
|
add_number += S_GET_VALUE (add_symbolP);
|
|
}
|
|
}
|
|
}
|
|
|
|
if (pcrel)
|
|
{
|
|
add_number -= md_pcrel_from (fixP);
|
|
if (add_symbolP == 0)
|
|
{
|
|
#ifndef BFD_ASSEMBLER
|
|
fixP->fx_addsy = &abs_symbol;
|
|
#else
|
|
fixP->fx_addsy = section_symbol (absolute_section);
|
|
#endif
|
|
fixP->fx_addsy->sy_used_in_reloc = 1;
|
|
++seg_reloc_count;
|
|
}
|
|
}
|
|
|
|
if (!fixP->fx_bit_fixP && size > 0)
|
|
{
|
|
valueT mask = 0;
|
|
if (size < sizeof (mask))
|
|
{
|
|
/* set all bits to one */
|
|
mask--;
|
|
/* Technically, combining these produces an undefined result
|
|
if size is sizeof (valueT), though I think these two
|
|
half-way operations should both be defined. And the
|
|
compiler should be able to combine them if it's valid on
|
|
the host architecture. */
|
|
mask <<= size * 4;
|
|
mask <<= size * 4;
|
|
if ((add_number & mask) != 0
|
|
&& (add_number & mask) != mask)
|
|
{
|
|
char buf[50], buf2[50];
|
|
sprint_value (buf, fragP->fr_address + where);
|
|
if (add_number > 1000)
|
|
sprint_value (buf2, add_number);
|
|
else
|
|
sprintf (buf2, "%ld", (long) add_number);
|
|
as_bad_where (fixP->fx_file, fixP->fx_line,
|
|
"Value of %s too large for field of %d bytes at %s",
|
|
buf2, size, buf);
|
|
} /* generic error checking */
|
|
}
|
|
#ifdef WARN_SIGNED_OVERFLOW_WORD
|
|
/* Warn if a .word value is too large when treated as a signed
|
|
number. We already know it is not too negative. This is to
|
|
catch over-large switches generated by gcc on the 68k. */
|
|
if (!flag_signed_overflow_ok
|
|
&& size == 2
|
|
&& add_number > 0x7fff)
|
|
as_bad_where (fixP->fx_file, fixP->fx_line,
|
|
"Signed .word overflow; switch may be too large; %ld at 0x%lx",
|
|
(long) add_number,
|
|
(unsigned long) (fragP->fr_address + where));
|
|
#endif
|
|
} /* not a bit fix */
|
|
|
|
if (!fixP->fx_done)
|
|
{
|
|
#ifdef BFD_ASSEMBLER
|
|
md_apply_fix (fixP, &add_number);
|
|
#else
|
|
md_apply_fix (fixP, add_number);
|
|
#endif
|
|
|
|
#ifndef TC_HANDLES_FX_DONE
|
|
/* If the tc-* files haven't been converted, assume it's handling
|
|
it the old way, where a null fx_addsy means that the fix has
|
|
been applied completely, and no further work is needed. */
|
|
if (fixP->fx_addsy == 0 && fixP->fx_pcrel == 0)
|
|
fixP->fx_done = 1;
|
|
#endif
|
|
}
|
|
#ifdef TC_VALIDATE_FIX
|
|
skip:
|
|
#endif
|
|
;
|
|
} /* For each fixS in this segment. */
|
|
|
|
TC_ADJUST_RELOC_COUNT (fixP, seg_reloc_count);
|
|
return seg_reloc_count;
|
|
}
|
|
|
|
#endif /* defined (BFD_ASSEMBLER) || !defined (BFD) */
|
|
|
|
void
|
|
number_to_chars_bigendian (buf, val, n)
|
|
char *buf;
|
|
valueT val;
|
|
int n;
|
|
{
|
|
if (n > sizeof (val)|| n <= 0)
|
|
abort ();
|
|
while (n--)
|
|
{
|
|
buf[n] = val & 0xff;
|
|
val >>= 8;
|
|
}
|
|
}
|
|
|
|
void
|
|
number_to_chars_littleendian (buf, val, n)
|
|
char *buf;
|
|
valueT val;
|
|
int n;
|
|
{
|
|
if (n > sizeof (val) || n <= 0)
|
|
abort ();
|
|
while (n--)
|
|
{
|
|
*buf++ = val & 0xff;
|
|
val >>= 8;
|
|
}
|
|
}
|
|
|
|
/* end of write.c */
|