1999-05-03 15:29:11 +08:00
|
|
|
|
/* vms-tir.c -- BFD back-end for VAX (openVMS/VAX) and
|
|
|
|
|
EVAX (openVMS/Alpha) files.
|
2001-09-01 18:29:27 +08:00
|
|
|
|
Copyright 1996, 1997, 1998, 1999, 2000, 2001 Free Software Foundation, Inc.
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
TIR record handling functions
|
|
|
|
|
ETIR record handling functions
|
|
|
|
|
|
|
|
|
|
go and read the openVMS linker manual (esp. appendix B)
|
|
|
|
|
if you don't know what's going on here :-)
|
|
|
|
|
|
|
|
|
|
Written by Klaus K"ampf (kkaempf@rmi.de)
|
|
|
|
|
|
|
|
|
|
This program 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 of the License, or
|
|
|
|
|
(at your option) any later version.
|
|
|
|
|
|
|
|
|
|
This program 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 this program; if not, write to the Free Software
|
|
|
|
|
Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
|
|
|
|
|
|
|
|
|
|
/* The following type abbreviations are used:
|
|
|
|
|
|
|
|
|
|
cs counted string (ascii string with length byte)
|
|
|
|
|
by byte (1 byte)
|
|
|
|
|
sh short (2 byte, 16 bit)
|
|
|
|
|
lw longword (4 byte, 32 bit)
|
|
|
|
|
qw quadword (8 byte, 64 bit)
|
|
|
|
|
da data stream */
|
|
|
|
|
|
|
|
|
|
#include "bfd.h"
|
|
|
|
|
#include "sysdep.h"
|
|
|
|
|
#include "bfdlink.h"
|
|
|
|
|
#include "libbfd.h"
|
|
|
|
|
|
|
|
|
|
#include "vms.h"
|
|
|
|
|
|
|
|
|
|
static void image_set_ptr PARAMS ((bfd *abfd, int psect, uquad offset));
|
|
|
|
|
static void image_inc_ptr PARAMS ((bfd *abfd, uquad offset));
|
|
|
|
|
static void image_dump PARAMS ((bfd *abfd, unsigned char *ptr, int size, int offset));
|
|
|
|
|
static void image_write_b PARAMS ((bfd *abfd, unsigned int value));
|
|
|
|
|
static void image_write_w PARAMS ((bfd *abfd, unsigned int value));
|
|
|
|
|
static void image_write_l PARAMS ((bfd *abfd, unsigned long value));
|
|
|
|
|
static void image_write_q PARAMS ((bfd *abfd, uquad value));
|
2001-09-01 18:29:27 +08:00
|
|
|
|
static int check_section PARAMS ((bfd *, int));
|
|
|
|
|
static boolean etir_sta PARAMS ((bfd *, int, unsigned char *));
|
|
|
|
|
static boolean etir_sto PARAMS ((bfd *, int, unsigned char *));
|
|
|
|
|
static boolean etir_opr PARAMS ((bfd *, int, unsigned char *));
|
|
|
|
|
static boolean etir_ctl PARAMS ((bfd *, int, unsigned char *));
|
|
|
|
|
static boolean etir_stc PARAMS ((bfd *, int, unsigned char *));
|
|
|
|
|
static asection *new_section PARAMS ((bfd *, int));
|
|
|
|
|
static int alloc_section PARAMS ((bfd *, unsigned int));
|
|
|
|
|
static int etir_cmd PARAMS ((bfd *, int, unsigned char *));
|
|
|
|
|
static int analyze_tir PARAMS ((bfd *, unsigned char *, unsigned int));
|
|
|
|
|
static int analyze_etir PARAMS ((bfd *, unsigned char *, unsigned int));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
/*-----------------------------------------------------------------------------*/
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
check_section (abfd, size)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int size;
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
bfd_size_type offset;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
offset = PRIV (image_ptr) - PRIV (image_section)->contents;
|
|
|
|
|
if (offset + size > PRIV (image_section)->_raw_size)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
PRIV (image_section)->contents
|
|
|
|
|
= bfd_realloc (PRIV (image_section)->contents, offset + size);
|
|
|
|
|
if (PRIV (image_section)->contents == 0)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
(*_bfd_error_handler) (_("No Mem !"));
|
|
|
|
|
return -1;
|
|
|
|
|
}
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
PRIV (image_section)->_raw_size = offset + size;
|
|
|
|
|
PRIV (image_ptr) = PRIV (image_section)->contents + offset;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* routines to fill sections contents during tir/etir read */
|
|
|
|
|
|
|
|
|
|
/* Initialize image buffer pointer to be filled */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
image_set_ptr (abfd, psect, offset)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int psect;
|
|
|
|
|
uquad offset;
|
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (4, "image_set_ptr (%d=%s, %d)\n",
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
psect, PRIV (sections)[psect]->name, offset);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
PRIV (image_ptr) = PRIV (sections)[psect]->contents + offset;
|
|
|
|
|
PRIV (image_section) = PRIV (sections)[psect];
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Increment image buffer pointer by offset */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
image_inc_ptr (abfd, offset)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
uquad offset;
|
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (4, "image_inc_ptr (%d)\n", offset);
|
|
|
|
|
#endif
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
PRIV (image_ptr) += offset;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Dump multiple bytes to section image */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
image_dump (abfd, ptr, size, offset)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
unsigned char *ptr;
|
|
|
|
|
int size;
|
1999-07-12 18:30:21 +08:00
|
|
|
|
int offset ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_debug (8, "image_dump from (%p, %d) to (%p)\n", ptr, size,
|
|
|
|
|
PRIV (image_ptr));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_hexdump (9, ptr, size, offset);
|
|
|
|
|
#endif
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (PRIV (is_vax) && check_section (abfd, size))
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
while (size-- > 0)
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
*PRIV (image_ptr)++ = *ptr++;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Write byte to section image */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
image_write_b (abfd, value)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
unsigned int value;
|
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_debug (6, "image_write_b(%02x)\n", (int) value);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (PRIV (is_vax) && check_section (abfd, 1))
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
*PRIV (image_ptr)++ = (value & 0xff);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Write 2-byte word to image */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
image_write_w (abfd, value)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
unsigned int value;
|
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_debug (6, "image_write_w(%04x)\n", (int) value);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (PRIV (is_vax) && check_section (abfd, 2))
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
bfd_putl16 ((bfd_vma) value, PRIV (image_ptr));
|
|
|
|
|
PRIV (image_ptr) += 2;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Write 4-byte long to image */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
image_write_l (abfd, value)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
unsigned long value;
|
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (6, "image_write_l (%08lx)\n", value);
|
|
|
|
|
#endif
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (PRIV (is_vax) && check_section (abfd, 4))
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
bfd_putl32 ((bfd_vma) value, PRIV (image_ptr));
|
|
|
|
|
PRIV (image_ptr) += 4;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Write 8-byte quad to image */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
image_write_q (abfd, value)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
uquad value;
|
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (6, "image_write_q (%016lx)\n", value);
|
|
|
|
|
#endif
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (PRIV (is_vax) && check_section (abfd, 8))
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
bfd_putl64 (value, PRIV (image_ptr));
|
|
|
|
|
PRIV (image_ptr) += 8;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
#define HIGHBIT(op) ((op & 0x80000000L) == 0x80000000L)
|
|
|
|
|
|
|
|
|
|
/* etir_sta
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
vms stack commands
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
handle sta_xxx commands in etir section
|
|
|
|
|
ptr points to data area in record
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
see table B-8 of the openVMS linker manual */
|
|
|
|
|
|
|
|
|
|
static boolean
|
|
|
|
|
etir_sta (abfd, cmd, ptr)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int cmd;
|
|
|
|
|
unsigned char *ptr;
|
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "etir_sta %d/%x\n", cmd, cmd);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_hexdump (8, ptr, 16, (int) ptr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (cmd)
|
|
|
|
|
{
|
|
|
|
|
/* stack */
|
|
|
|
|
|
|
|
|
|
/* stack global
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
arg: cs symbol name
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
stack 32 bit value of symbol (high bits set to 0) */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case ETIR_S_C_STA_GBL:
|
|
|
|
|
{
|
|
|
|
|
char *name;
|
|
|
|
|
vms_symbol_entry *entry;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
name = _bfd_vms_save_counted_string (ptr);
|
|
|
|
|
entry = (vms_symbol_entry *)
|
|
|
|
|
bfd_hash_lookup (PRIV (vms_symbol_table), name, false, false);
|
|
|
|
|
if (entry == (vms_symbol_entry *) NULL)
|
|
|
|
|
{
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#if VMS_DEBUG
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_debug (3, "ETIR_S_C_STA_GBL: no symbol \"%s\"\n", name);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) 0, -1);
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (entry->symbol->value), -1);
|
|
|
|
|
}
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* stack longword
|
|
|
|
|
arg: lw value
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
stack 32 bit value, sign extend to 64 bit */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case ETIR_S_C_STA_LW:
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) bfd_getl32 (ptr), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* stack global
|
|
|
|
|
arg: qw value
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
stack 64 bit value of symbol */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case ETIR_S_C_STA_QW:
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) bfd_getl64 (ptr), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* stack psect base plus quadword offset
|
|
|
|
|
arg: lw section index
|
|
|
|
|
qw signed quadword offset (low 32 bits)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
stack qw argument and section index
|
|
|
|
|
(see ETIR_S_C_STO_OFF, ETIR_S_C_CTL_SETRB) */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case ETIR_S_C_STA_PQ:
|
|
|
|
|
{
|
|
|
|
|
uquad dummy;
|
|
|
|
|
unsigned int psect;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
psect = bfd_getl32 (ptr);
|
|
|
|
|
if (psect >= PRIV (section_count))
|
|
|
|
|
{
|
|
|
|
|
(*_bfd_error_handler) (_("Bad section index in ETIR_S_C_STA_PQ"));
|
|
|
|
|
bfd_set_error (bfd_error_bad_value);
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
dummy = bfd_getl64 (ptr+4);
|
|
|
|
|
_bfd_vms_push (abfd, dummy, (int) psect);
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* all not supported */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case ETIR_S_C_STA_LI:
|
|
|
|
|
case ETIR_S_C_STA_MOD:
|
|
|
|
|
case ETIR_S_C_STA_CKARG:
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(*_bfd_error_handler) (_("Unsupported STA cmd %d"), cmd);
|
|
|
|
|
return false;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
default:
|
|
|
|
|
(*_bfd_error_handler) (_("Reserved STA cmd %d"), cmd);
|
|
|
|
|
return false;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "etir_sta true\n");
|
|
|
|
|
#endif
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
etir_sto
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
vms store commands
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
handle sto_xxx commands in etir section
|
|
|
|
|
ptr points to data area in record
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
see table B-9 of the openVMS linker manual */
|
|
|
|
|
|
|
|
|
|
static boolean
|
|
|
|
|
etir_sto (abfd, cmd, ptr)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int cmd;
|
|
|
|
|
unsigned char *ptr;
|
|
|
|
|
{
|
|
|
|
|
uquad dummy;
|
|
|
|
|
int psect;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "etir_sto %d/%x\n", cmd, cmd);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_hexdump (8, ptr, 16, (int) ptr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (cmd)
|
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
/* store byte: pop stack, write byte
|
|
|
|
|
arg: - */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_B:
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
#if 0
|
|
|
|
|
if (is_share) /* FIXME */
|
|
|
|
|
(*_bfd_error_handler) ("ETIR_S_C_STO_B: byte fixups not supported");
|
|
|
|
|
#endif
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* FIXME: check top bits */
|
|
|
|
|
image_write_b (abfd, (unsigned int) dummy & 0xff);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* store word: pop stack, write word
|
|
|
|
|
arg: - */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_W:
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
#if 0
|
|
|
|
|
if (is_share) /* FIXME */
|
|
|
|
|
(*_bfd_error_handler) ("ETIR_S_C_STO_B: word fixups not supported");
|
|
|
|
|
#endif
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* FIXME: check top bits */
|
|
|
|
|
image_write_w (abfd, (unsigned int) dummy & 0xffff);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* store longword: pop stack, write longword
|
|
|
|
|
arg: - */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_LW:
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
dummy += (PRIV (sections)[psect])->vma;
|
|
|
|
|
/* FIXME: check top bits */
|
|
|
|
|
image_write_l (abfd, (unsigned int) dummy & 0xffffffff);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* store quadword: pop stack, write quadword
|
|
|
|
|
arg: - */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_QW:
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
dummy += (PRIV (sections)[psect])->vma;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
image_write_q (abfd, dummy); /* FIXME: check top bits */
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* store immediate repeated: pop stack for repeat count
|
|
|
|
|
arg: lw byte count
|
|
|
|
|
da data */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_IMMR:
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
int size;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
size = bfd_getl32 (ptr);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
dummy = (unsigned long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
while (dummy-- > 0)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
image_dump (abfd, ptr+4, size, 0);
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* store global: write symbol value
|
|
|
|
|
arg: cs global symbol name */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_GBL:
|
|
|
|
|
{
|
|
|
|
|
vms_symbol_entry *entry;
|
|
|
|
|
char *name;
|
|
|
|
|
|
|
|
|
|
name = _bfd_vms_save_counted_string (ptr);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
entry = (vms_symbol_entry *) bfd_hash_lookup (PRIV (vms_symbol_table),
|
|
|
|
|
name, false, false);
|
|
|
|
|
if (entry == (vms_symbol_entry *) NULL)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STO_GBL: no symbol \"%s\""),
|
|
|
|
|
name);
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
else
|
2000-12-25 22:42:42 +08:00
|
|
|
|
image_write_q (abfd, (uquad) (entry->symbol->value)); /* FIXME, reloc */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* store code address: write address of entry point
|
|
|
|
|
arg: cs global symbol name (procedure) */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_CA:
|
|
|
|
|
{
|
|
|
|
|
vms_symbol_entry *entry;
|
|
|
|
|
char *name;
|
|
|
|
|
|
|
|
|
|
name = _bfd_vms_save_counted_string (ptr);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
entry = (vms_symbol_entry *) bfd_hash_lookup (PRIV (vms_symbol_table),
|
|
|
|
|
name, false, false);
|
|
|
|
|
if (entry == (vms_symbol_entry *) NULL)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STO_CA: no symbol \"%s\""),
|
|
|
|
|
name);
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
else
|
2000-12-25 22:42:42 +08:00
|
|
|
|
image_write_q (abfd, (uquad) (entry->symbol->value)); /* FIXME, reloc */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* not supported */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_RB:
|
|
|
|
|
case ETIR_S_C_STO_AB:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STO_RB/AB: Not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* store offset to psect: pop stack, add low 32 bits to base of psect
|
|
|
|
|
arg: - */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_OFF:
|
|
|
|
|
{
|
|
|
|
|
uquad q;
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
int psect1;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
q = _bfd_vms_pop (abfd, &psect1);
|
|
|
|
|
q += (PRIV (sections)[psect1])->vma;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
image_write_q (abfd, q);
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* store immediate
|
|
|
|
|
arg: lw count of bytes
|
|
|
|
|
da data */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_IMM:
|
|
|
|
|
{
|
|
|
|
|
int size;
|
|
|
|
|
|
|
|
|
|
size = bfd_getl32 (ptr);
|
|
|
|
|
image_dump (abfd, ptr+4, size, 0);
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* this code is 'reserved to digital' according to the openVMS
|
|
|
|
|
linker manual, however it is generated by the DEC C compiler
|
|
|
|
|
and defined in the include file.
|
1999-05-03 15:29:11 +08:00
|
|
|
|
FIXME, since the following is just a guess
|
|
|
|
|
store global longword: store 32bit value of symbol
|
|
|
|
|
arg: cs symbol name */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_GBL_LW:
|
|
|
|
|
{
|
|
|
|
|
vms_symbol_entry *entry;
|
|
|
|
|
char *name;
|
|
|
|
|
|
|
|
|
|
name = _bfd_vms_save_counted_string (ptr);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
entry = (vms_symbol_entry *) bfd_hash_lookup (PRIV (vms_symbol_table),
|
|
|
|
|
name, false, false);
|
|
|
|
|
if (entry == (vms_symbol_entry *) NULL)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (3, "ETIR_S_C_STO_GBL_LW: no symbol \"%s\"\n", name);
|
|
|
|
|
#endif
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
image_write_l (abfd, (unsigned long) 0); /* FIXME, reloc */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
}
|
|
|
|
|
else
|
2000-12-25 22:42:42 +08:00
|
|
|
|
image_write_l (abfd, (unsigned long) (entry->symbol->value)); /* FIXME, reloc */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* not supported */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_LP_PSB:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STO_LP_PSB: Not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_HINT_GBL:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STO_HINT_GBL: not implemented"));
|
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STO_HINT_PS:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STO_HINT_PS: not implemented"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
(*_bfd_error_handler) (_("Reserved STO cmd %d"), cmd);
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* stack operator commands
|
|
|
|
|
all 32 bit signed arithmetic
|
|
|
|
|
all word just like a stack calculator
|
|
|
|
|
arguments are popped from stack, results are pushed on stack
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
see table B-10 of the openVMS linker manual */
|
|
|
|
|
|
|
|
|
|
static boolean
|
|
|
|
|
etir_opr (abfd, cmd, ptr)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int cmd;
|
1999-07-12 18:30:21 +08:00
|
|
|
|
unsigned char *ptr ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
long op1, op2;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "etir_opr %d/%x\n", cmd, cmd);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_hexdump (8, ptr, 16, (int) ptr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (cmd)
|
|
|
|
|
{
|
|
|
|
|
/* operation */
|
|
|
|
|
|
|
|
|
|
/* no-op */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_NOP:
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* add */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_ADD:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 + op2), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* subtract */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_SUB:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op2 - op1), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* multiply */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_MUL:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 * op2), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* divide */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_DIV:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
if (op2 == 0)
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) 0, -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
else
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op2 / op1), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* logical and */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_AND:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 & op2), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* logical inclusive or */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_IOR:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 | op2), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* logical exclusive or */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_EOR:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 ^ op2), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* negate */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_NEG:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (-op1), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* complement */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_COM:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 ^ -1L), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* insert field */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_INSV:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(void) _bfd_vms_pop (abfd, NULL);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_OPR_INSV: Not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* arithmetic shift */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_ASH:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
if (op2 < 0) /* shift right */
|
|
|
|
|
op1 >>= -op2;
|
|
|
|
|
else /* shift left */
|
|
|
|
|
op1 <<= op2;
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) op1, -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* unsigned shift */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_USH:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_OPR_USH: Not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* rotate */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_ROT:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_OPR_ROT: Not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* select */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_SEL:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if ((long) _bfd_vms_pop (abfd, NULL) & 0x01L)
|
|
|
|
|
(void) _bfd_vms_pop (abfd, NULL);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
else
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
(void) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) op1, -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* redefine symbol to current location */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_REDEF:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_OPR_REDEF: Not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* define a literal */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_OPR_DFLIT:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_OPR_DFLIT: Not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
(*_bfd_error_handler) (_("Reserved OPR cmd %d"), cmd);
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* control commands
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
see table B-11 of the openVMS linker manual */
|
|
|
|
|
|
|
|
|
|
static boolean
|
|
|
|
|
etir_ctl (abfd, cmd, ptr)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int cmd;
|
|
|
|
|
unsigned char *ptr;
|
|
|
|
|
{
|
|
|
|
|
uquad dummy;
|
|
|
|
|
int psect;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "etir_ctl %d/%x\n", cmd, cmd);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_hexdump (8, ptr, 16, (int) ptr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (cmd)
|
|
|
|
|
{
|
|
|
|
|
/* set relocation base: pop stack, set image location counter
|
|
|
|
|
arg: - */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_CTL_SETRB:
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
image_set_ptr (abfd, psect, dummy);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* augment relocation base: increment image location counter by offset
|
|
|
|
|
arg: lw offset value */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_CTL_AUGRB:
|
|
|
|
|
dummy = bfd_getl32 (ptr);
|
|
|
|
|
image_inc_ptr (abfd, dummy);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* define location: pop index, save location counter under index
|
|
|
|
|
arg: - */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_CTL_DFLOC:
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
/* FIXME */
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* set location: pop index, restore location counter from index
|
|
|
|
|
arg: - */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_CTL_STLOC:
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
/* FIXME */
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* stack defined location: pop index, push location counter from index
|
|
|
|
|
arg: - */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_CTL_STKDL:
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
/* FIXME */
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
(*_bfd_error_handler) (_("Reserved CTL cmd %d"), cmd);
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* store conditional commands
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
see table B-12 and B-13 of the openVMS linker manual */
|
|
|
|
|
|
|
|
|
|
static boolean
|
|
|
|
|
etir_stc (abfd, cmd, ptr)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int cmd;
|
1999-07-12 18:30:21 +08:00
|
|
|
|
unsigned char *ptr ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "etir_stc %d/%x\n", cmd, cmd);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_hexdump (8, ptr, 16, (int) ptr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (cmd)
|
|
|
|
|
{
|
|
|
|
|
/* 200 Store-conditional Linkage Pair
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_LP:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STC_LP: not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* 201 Store-conditional Linkage Pair with Procedure Signature
|
|
|
|
|
arg: lw linkage index
|
|
|
|
|
cs procedure name
|
|
|
|
|
by signature length
|
|
|
|
|
da signature */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_LP_PSB:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
image_inc_ptr (abfd, (uquad) 16); /* skip entry,procval */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* 202 Store-conditional Address at global address
|
|
|
|
|
arg: lw linkage index
|
|
|
|
|
cs global name */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_GBL:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STC_GBL: not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* 203 Store-conditional Code Address at global address
|
|
|
|
|
arg: lw linkage index
|
|
|
|
|
cs procedure name */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_GCA:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STC_GCA: not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* 204 Store-conditional Address at psect + offset
|
|
|
|
|
arg: lw linkage index
|
|
|
|
|
lw psect index
|
|
|
|
|
qw offset */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_PS:
|
|
|
|
|
(*_bfd_error_handler) (_("ETIR_S_C_STC_PS: not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
/* 205 Store-conditional NOP at address of global
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_NOP_GBL:
|
|
|
|
|
|
|
|
|
|
/* 206 Store-conditional NOP at pect + offset
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_NOP_PS:
|
|
|
|
|
|
|
|
|
|
/* 207 Store-conditional BSR at global address
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_BSR_GBL:
|
|
|
|
|
|
|
|
|
|
/* 208 Store-conditional BSR at pect + offset
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_BSR_PS:
|
|
|
|
|
|
|
|
|
|
/* 209 Store-conditional LDA at global address
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_LDA_GBL:
|
|
|
|
|
|
|
|
|
|
/* 210 Store-conditional LDA at psect + offset
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_LDA_PS:
|
|
|
|
|
|
|
|
|
|
/* 211 Store-conditional BSR or Hint at global address
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_BOH_GBL:
|
|
|
|
|
|
|
|
|
|
/* 212 Store-conditional BSR or Hint at pect + offset
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_BOH_PS:
|
|
|
|
|
|
|
|
|
|
/* 213 Store-conditional NOP,BSR or HINT at global address
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_NBH_GBL:
|
|
|
|
|
|
|
|
|
|
/* 214 Store-conditional NOP,BSR or HINT at psect + offset
|
|
|
|
|
arg: */
|
|
|
|
|
|
|
|
|
|
case ETIR_S_C_STC_NBH_PS:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* FIXME */
|
|
|
|
|
#if 0
|
|
|
|
|
(*_bfd_error_handler) ("ETIR_S_C_STC_xx: (%d) not supported", cmd);
|
|
|
|
|
#endif
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (3, "Reserved STC cmd %d", cmd);
|
|
|
|
|
#endif
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
return true;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static asection *
|
|
|
|
|
new_section (abfd, idx)
|
1999-07-12 18:30:21 +08:00
|
|
|
|
bfd *abfd ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
int idx;
|
|
|
|
|
{
|
|
|
|
|
asection *section;
|
|
|
|
|
char sname[16];
|
|
|
|
|
char *name;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "new_section %d\n", idx);
|
|
|
|
|
#endif
|
|
|
|
|
sprintf (sname, SECTION_NAME_TEMPLATE, idx);
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
name = bfd_malloc ((bfd_size_type) strlen (sname) + 1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
if (name == 0)
|
|
|
|
|
return 0;
|
|
|
|
|
strcpy (name, sname);
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
section = bfd_malloc ((bfd_size_type) sizeof (asection));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
if (section == 0)
|
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (6, "bfd_make_section (%s) failed", name);
|
|
|
|
|
#endif
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
section->_raw_size = 0;
|
|
|
|
|
section->vma = 0;
|
|
|
|
|
section->contents = 0;
|
|
|
|
|
section->_cooked_size = 0;
|
|
|
|
|
section->name = name;
|
|
|
|
|
section->index = idx;
|
|
|
|
|
|
|
|
|
|
return section;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
alloc_section (abfd, idx)
|
|
|
|
|
bfd *abfd;
|
1999-07-12 18:30:21 +08:00
|
|
|
|
unsigned int idx;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
bfd_size_type amt;
|
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (4, "alloc_section %d\n", idx);
|
|
|
|
|
#endif
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
amt = idx + 1;
|
|
|
|
|
amt *= sizeof (asection *);
|
|
|
|
|
PRIV (sections) = (asection **) bfd_realloc (PRIV (sections), amt);
|
|
|
|
|
if (PRIV (sections) == 0)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return -1;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
while (PRIV (section_count) <= idx)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
PRIV (sections)[PRIV (section_count)]
|
|
|
|
|
= new_section (abfd, (int) PRIV (section_count));
|
|
|
|
|
if (PRIV (sections)[PRIV (section_count)] == 0)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return -1;
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
PRIV (section_count)++;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* tir_sta
|
|
|
|
|
*
|
|
|
|
|
* vax stack commands
|
|
|
|
|
*
|
|
|
|
|
* handle sta_xxx commands in tir section
|
|
|
|
|
* ptr points to data area in record
|
|
|
|
|
*
|
|
|
|
|
* see table 7-3 of the VAX/VMS linker manual
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static unsigned char *
|
|
|
|
|
tir_sta (bfd *abfd, unsigned char *ptr)
|
|
|
|
|
{
|
|
|
|
|
int cmd = *ptr++;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "tir_sta %d\n", cmd);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (cmd)
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* stack */
|
|
|
|
|
case TIR_S_C_STA_GBL:
|
|
|
|
|
/*
|
|
|
|
|
* stack global
|
|
|
|
|
* arg: cs symbol name
|
|
|
|
|
*
|
|
|
|
|
* stack 32 bit value of symbol (high bits set to 0)
|
|
|
|
|
*/
|
|
|
|
|
{
|
|
|
|
|
char *name;
|
|
|
|
|
vms_symbol_entry *entry;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
name = _bfd_vms_save_counted_string (ptr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
entry = _bfd_vms_enter_symbol (abfd, name);
|
|
|
|
|
if (entry == (vms_symbol_entry *) NULL)
|
|
|
|
|
return 0;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_push (abfd, (uquad) (entry->symbol->value), -1);
|
|
|
|
|
ptr += *ptr + 1;
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_STA_SB:
|
|
|
|
|
/*
|
|
|
|
|
* stack signed byte
|
|
|
|
|
* arg: by value
|
|
|
|
|
*
|
|
|
|
|
* stack byte value, sign extend to 32 bit
|
|
|
|
|
*/
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) *ptr++, -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STA_SW:
|
|
|
|
|
/*
|
|
|
|
|
* stack signed short word
|
|
|
|
|
* arg: sh value
|
|
|
|
|
*
|
|
|
|
|
* stack 16 bit value, sign extend to 32 bit
|
|
|
|
|
*/
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) bfd_getl16 (ptr), -1);
|
|
|
|
|
ptr += 2;
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STA_LW:
|
|
|
|
|
/*
|
|
|
|
|
* stack signed longword
|
|
|
|
|
* arg: lw value
|
|
|
|
|
*
|
|
|
|
|
* stack 32 bit value
|
|
|
|
|
*/
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) bfd_getl32 (ptr), -1);
|
|
|
|
|
ptr += 4;
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STA_PB:
|
|
|
|
|
case TIR_S_C_STA_WPB:
|
|
|
|
|
/*
|
|
|
|
|
* stack psect base plus byte offset (word index)
|
|
|
|
|
* arg: by section index
|
|
|
|
|
* (sh section index)
|
|
|
|
|
* by signed byte offset
|
|
|
|
|
*
|
|
|
|
|
*/
|
|
|
|
|
{
|
|
|
|
|
unsigned long dummy;
|
|
|
|
|
unsigned int psect;
|
|
|
|
|
|
|
|
|
|
if (cmd == TIR_S_C_STA_PB)
|
|
|
|
|
psect = *ptr++;
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
psect = bfd_getl16 (ptr);
|
|
|
|
|
ptr += 2;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (psect >= PRIV (section_count))
|
|
|
|
|
{
|
|
|
|
|
alloc_section (abfd, psect);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
dummy = (long) *ptr++;
|
|
|
|
|
dummy += (PRIV (sections)[psect])->vma;
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) dummy, (int) psect);
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_STA_PW:
|
|
|
|
|
case TIR_S_C_STA_WPW:
|
|
|
|
|
/*
|
|
|
|
|
* stack psect base plus word offset (word index)
|
|
|
|
|
* arg: by section index
|
|
|
|
|
* (sh section index)
|
|
|
|
|
* sh signed short offset
|
|
|
|
|
*
|
|
|
|
|
*/
|
|
|
|
|
{
|
|
|
|
|
unsigned long dummy;
|
|
|
|
|
unsigned int psect;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (cmd == TIR_S_C_STA_PW)
|
|
|
|
|
psect = *ptr++;
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
psect = bfd_getl16 (ptr);
|
|
|
|
|
ptr += 2;
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (psect >= PRIV (section_count))
|
|
|
|
|
{
|
|
|
|
|
alloc_section (abfd, psect);
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
dummy = bfd_getl16 (ptr); ptr+=2;
|
|
|
|
|
dummy += (PRIV (sections)[psect])->vma;
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) dummy, (int) psect);
|
|
|
|
|
}
|
|
|
|
|
break;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_STA_PL:
|
|
|
|
|
case TIR_S_C_STA_WPL:
|
|
|
|
|
/*
|
|
|
|
|
* stack psect base plus long offset (word index)
|
|
|
|
|
* arg: by section index
|
|
|
|
|
* (sh section index)
|
|
|
|
|
* lw signed longword offset
|
|
|
|
|
*
|
|
|
|
|
*/
|
|
|
|
|
{
|
|
|
|
|
unsigned long dummy;
|
|
|
|
|
unsigned int psect;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (cmd == TIR_S_C_STA_PL)
|
|
|
|
|
psect = *ptr++;
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
psect = bfd_getl16 (ptr);
|
|
|
|
|
ptr += 2;
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (psect >= PRIV (section_count))
|
|
|
|
|
{
|
|
|
|
|
alloc_section (abfd, psect);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
dummy = bfd_getl32 (ptr); ptr += 4;
|
|
|
|
|
dummy += (PRIV (sections)[psect])->vma;
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) dummy, (int) psect);
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_STA_UB:
|
|
|
|
|
/*
|
|
|
|
|
* stack unsigned byte
|
|
|
|
|
* arg: by value
|
|
|
|
|
*
|
|
|
|
|
* stack byte value
|
|
|
|
|
*/
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) *ptr++, -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STA_UW:
|
|
|
|
|
/*
|
|
|
|
|
* stack unsigned short word
|
|
|
|
|
* arg: sh value
|
|
|
|
|
*
|
|
|
|
|
* stack 16 bit value
|
|
|
|
|
*/
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) bfd_getl16 (ptr), -1);
|
|
|
|
|
ptr += 2;
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STA_BFI:
|
|
|
|
|
/*
|
|
|
|
|
* stack byte from image
|
|
|
|
|
* arg: -
|
|
|
|
|
*
|
|
|
|
|
*/
|
|
|
|
|
/*FALLTHRU*/
|
|
|
|
|
case TIR_S_C_STA_WFI:
|
|
|
|
|
/*
|
|
|
|
|
* stack byte from image
|
|
|
|
|
* arg: -
|
|
|
|
|
*
|
|
|
|
|
*/
|
|
|
|
|
/*FALLTHRU*/
|
|
|
|
|
case TIR_S_C_STA_LFI:
|
|
|
|
|
/*
|
|
|
|
|
* stack byte from image
|
|
|
|
|
* arg: -
|
|
|
|
|
*
|
|
|
|
|
*/
|
|
|
|
|
(*_bfd_error_handler) (_("Stack-from-image not implemented"));
|
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STA_EPM:
|
|
|
|
|
/*
|
|
|
|
|
* stack entry point mask
|
|
|
|
|
* arg: cs symbol name
|
|
|
|
|
*
|
|
|
|
|
* stack (unsigned) entry point mask of symbol
|
|
|
|
|
* err if symbol is no entry point
|
|
|
|
|
*/
|
|
|
|
|
{
|
|
|
|
|
char *name;
|
|
|
|
|
vms_symbol_entry *entry;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
name = _bfd_vms_save_counted_string (ptr);
|
|
|
|
|
entry = _bfd_vms_enter_symbol (abfd, name);
|
|
|
|
|
if (entry == (vms_symbol_entry *) NULL)
|
|
|
|
|
return 0;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(*_bfd_error_handler) (_("Stack-entry-mask not fully implemented"));
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) 0, -1);
|
|
|
|
|
ptr += *ptr + 1;
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_STA_CKARG:
|
|
|
|
|
/*
|
|
|
|
|
* compare procedure argument
|
|
|
|
|
* arg: cs symbol name
|
|
|
|
|
* by argument index
|
|
|
|
|
* da argument descriptor
|
|
|
|
|
*
|
|
|
|
|
* compare argument descriptor with symbol argument (ARG$V_PASSMECH)
|
|
|
|
|
* and stack TRUE (args match) or FALSE (args dont match) value
|
|
|
|
|
*/
|
|
|
|
|
(*_bfd_error_handler) (_("PASSMECH not fully implemented"));
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) 1, -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STA_LSY:
|
|
|
|
|
/*
|
|
|
|
|
* stack local symbol value
|
|
|
|
|
* arg: sh environment index
|
|
|
|
|
* cs symbol name
|
|
|
|
|
*/
|
|
|
|
|
{
|
|
|
|
|
int envidx;
|
|
|
|
|
char *name;
|
|
|
|
|
vms_symbol_entry *entry;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
envidx = bfd_getl16 (ptr);
|
|
|
|
|
ptr += 2;
|
|
|
|
|
name = _bfd_vms_save_counted_string (ptr);
|
|
|
|
|
entry = _bfd_vms_enter_symbol (abfd, name);
|
|
|
|
|
if (entry == (vms_symbol_entry *) NULL)
|
|
|
|
|
return 0;
|
|
|
|
|
(*_bfd_error_handler) (_("Stack-local-symbol not fully implemented"));
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) 0, -1);
|
|
|
|
|
ptr += *ptr + 1;
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_STA_LIT:
|
|
|
|
|
/*
|
|
|
|
|
* stack literal
|
|
|
|
|
* arg: by literal index
|
|
|
|
|
*
|
|
|
|
|
* stack literal
|
|
|
|
|
*/
|
|
|
|
|
ptr++;
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) 0, -1);
|
|
|
|
|
(*_bfd_error_handler) (_("Stack-literal not fully implemented"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STA_LEPM:
|
|
|
|
|
/*
|
|
|
|
|
* stack local symbol entry point mask
|
|
|
|
|
* arg: sh environment index
|
|
|
|
|
* cs symbol name
|
|
|
|
|
*
|
|
|
|
|
* stack (unsigned) entry point mask of symbol
|
|
|
|
|
* err if symbol is no entry point
|
|
|
|
|
*/
|
|
|
|
|
{
|
|
|
|
|
int envidx;
|
|
|
|
|
char *name;
|
|
|
|
|
vms_symbol_entry *entry;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
envidx = bfd_getl16 (ptr);
|
|
|
|
|
ptr += 2;
|
|
|
|
|
name = _bfd_vms_save_counted_string (ptr);
|
|
|
|
|
entry = _bfd_vms_enter_symbol (abfd, name);
|
|
|
|
|
if (entry == (vms_symbol_entry *) NULL)
|
|
|
|
|
return 0;
|
|
|
|
|
(*_bfd_error_handler) (_("Stack-local-symbol-entry-point-mask not fully implemented"));
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) 0, -1);
|
|
|
|
|
ptr += *ptr + 1;
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
default:
|
|
|
|
|
(*_bfd_error_handler) (_("Reserved STA cmd %d"), ptr[-1]);
|
|
|
|
|
return NULL;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
return ptr;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* tir_sto
|
|
|
|
|
*
|
|
|
|
|
* vax store commands
|
|
|
|
|
*
|
|
|
|
|
* handle sto_xxx commands in tir section
|
|
|
|
|
* ptr points to data area in record
|
|
|
|
|
*
|
|
|
|
|
* see table 7-4 of the VAX/VMS linker manual
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static unsigned char *
|
|
|
|
|
tir_sto (bfd *abfd, unsigned char *ptr)
|
|
|
|
|
{
|
|
|
|
|
unsigned long dummy;
|
|
|
|
|
int size;
|
|
|
|
|
int psect;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "tir_sto %d\n", *ptr);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (*ptr++)
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_STO_SB:
|
|
|
|
|
/*
|
|
|
|
|
* store signed byte: pop stack, write byte
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
image_write_b (abfd, dummy & 0xff); /* FIXME: check top bits */
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STO_SW:
|
|
|
|
|
/*
|
|
|
|
|
* store signed word: pop stack, write word
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
image_write_w (abfd, dummy & 0xffff); /* FIXME: check top bits */
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STO_LW:
|
|
|
|
|
/*
|
|
|
|
|
* store longword: pop stack, write longword
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
image_write_l (abfd, dummy & 0xffffffff); /* FIXME: check top bits */
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STO_BD:
|
|
|
|
|
/*
|
|
|
|
|
* store byte displaced: pop stack, sub lc+1, write byte
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
dummy -= ((PRIV (sections)[psect])->vma + 1);
|
|
|
|
|
image_write_b (abfd, dummy & 0xff);/* FIXME: check top bits */
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_STO_WD:
|
|
|
|
|
/*
|
|
|
|
|
* store word displaced: pop stack, sub lc+2, write word
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
dummy -= ((PRIV (sections)[psect])->vma + 2);
|
|
|
|
|
image_write_w (abfd, dummy & 0xffff);/* FIXME: check top bits */
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_STO_LD:
|
|
|
|
|
/*
|
|
|
|
|
* store long displaced: pop stack, sub lc+4, write long
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
dummy -= ((PRIV (sections)[psect])->vma + 4);
|
|
|
|
|
image_write_l (abfd, dummy & 0xffffffff);/* FIXME: check top bits */
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_STO_LI:
|
|
|
|
|
/*
|
|
|
|
|
* store short literal: pop stack, write byte
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
image_write_b (abfd, dummy & 0xff);/* FIXME: check top bits */
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_STO_PIDR:
|
|
|
|
|
/*
|
|
|
|
|
* store position independent data reference: pop stack, write longword
|
|
|
|
|
* arg: -
|
|
|
|
|
* FIXME: incomplete !
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
image_write_l (abfd, dummy & 0xffffffff);
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_STO_PICR:
|
|
|
|
|
/*
|
|
|
|
|
* store position independent code reference: pop stack, write longword
|
|
|
|
|
* arg: -
|
|
|
|
|
* FIXME: incomplete !
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
image_write_b (abfd, 0x9f);
|
|
|
|
|
image_write_l (abfd, dummy & 0xffffffff);
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_STO_RIVB:
|
|
|
|
|
/*
|
|
|
|
|
* store repeated immediate variable bytes
|
|
|
|
|
* 1-byte count n field followed by n bytes of data
|
|
|
|
|
* pop stack, write n bytes <stack> times
|
|
|
|
|
*/
|
|
|
|
|
size = *ptr++;
|
|
|
|
|
dummy = (unsigned long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
while (dummy-- > 0L)
|
|
|
|
|
image_dump (abfd, ptr, size, 0);
|
|
|
|
|
ptr += size;
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_STO_B:
|
|
|
|
|
/*
|
|
|
|
|
* store byte from top longword
|
|
|
|
|
*/
|
|
|
|
|
dummy = (unsigned long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
image_write_b (abfd, dummy & 0xff);
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_STO_W:
|
|
|
|
|
/*
|
|
|
|
|
* store word from top longword
|
|
|
|
|
*/
|
|
|
|
|
dummy = (unsigned long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
image_write_w (abfd, dummy & 0xffff);
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_STO_RB:
|
|
|
|
|
/*
|
|
|
|
|
* store repeated byte from top longword
|
|
|
|
|
*/
|
|
|
|
|
size = (unsigned long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
dummy = (unsigned long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
while (size-- > 0)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
image_write_b (abfd, dummy & 0xff);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_STO_RW:
|
|
|
|
|
/*
|
|
|
|
|
* store repeated word from top longword
|
|
|
|
|
*/
|
|
|
|
|
size = (unsigned long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
dummy = (unsigned long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
while (size-- > 0)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
image_write_w (abfd, dummy & 0xffff);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
break;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_STO_RSB:
|
|
|
|
|
case TIR_S_C_STO_RSW:
|
|
|
|
|
case TIR_S_C_STO_RL:
|
|
|
|
|
case TIR_S_C_STO_VPS:
|
|
|
|
|
case TIR_S_C_STO_USB:
|
|
|
|
|
case TIR_S_C_STO_USW:
|
|
|
|
|
case TIR_S_C_STO_RUB:
|
|
|
|
|
case TIR_S_C_STO_RUW:
|
|
|
|
|
case TIR_S_C_STO_PIRR:
|
|
|
|
|
(*_bfd_error_handler) (_("Unimplemented STO cmd %d"), ptr[-1]);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
default:
|
|
|
|
|
(*_bfd_error_handler) (_("Reserved STO cmd %d"), ptr[-1]);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
return ptr;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* stack operator commands
|
|
|
|
|
* all 32 bit signed arithmetic
|
|
|
|
|
* all word just like a stack calculator
|
|
|
|
|
* arguments are popped from stack, results are pushed on stack
|
|
|
|
|
*
|
|
|
|
|
* see table 7-5 of the VAX/VMS linker manual
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static unsigned char *
|
|
|
|
|
tir_opr (bfd *abfd, unsigned char *ptr)
|
|
|
|
|
{
|
|
|
|
|
long op1, op2;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "tir_opr %d\n", *ptr);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (*ptr++)
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* operation */
|
|
|
|
|
case TIR_S_C_OPR_NOP:
|
|
|
|
|
/*
|
|
|
|
|
* no-op
|
|
|
|
|
*/
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_ADD:
|
|
|
|
|
/*
|
|
|
|
|
* add
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 + op2), -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_SUB:
|
|
|
|
|
/*
|
|
|
|
|
* subtract
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op2 - op1), -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_MUL:
|
|
|
|
|
/*
|
|
|
|
|
* multiply
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 * op2), -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_DIV:
|
|
|
|
|
/*
|
|
|
|
|
* divide
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
if (op2 == 0)
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) 0, -1);
|
|
|
|
|
else
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op2 / op1), -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_AND:
|
|
|
|
|
/*
|
|
|
|
|
* logical and
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 & op2), -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_IOR:
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
/*
|
|
|
|
|
* logical inclusive or
|
|
|
|
|
*/
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 | op2), -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_EOR:
|
|
|
|
|
/*
|
|
|
|
|
* logical exclusive or
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 ^ op2), -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_NEG:
|
|
|
|
|
/*
|
|
|
|
|
* negate
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (-op1), -1);
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_COM:
|
|
|
|
|
/*
|
|
|
|
|
* complement
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) (op1 ^ -1L), -1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_OPR_INSV:
|
|
|
|
|
/*
|
|
|
|
|
* insert field
|
|
|
|
|
*/
|
|
|
|
|
(void) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
(*_bfd_error_handler) ("TIR_S_C_OPR_INSV incomplete");
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_OPR_ASH:
|
|
|
|
|
/*
|
|
|
|
|
* arithmetic shift
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
if (HIGHBIT (op1)) /* shift right */
|
|
|
|
|
op2 >>= op1;
|
|
|
|
|
else /* shift left */
|
|
|
|
|
op2 <<= op1;
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) op2, -1);
|
|
|
|
|
(*_bfd_error_handler) (_("TIR_S_C_OPR_ASH incomplete"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_USH:
|
|
|
|
|
/*
|
|
|
|
|
* unsigned shift
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
if (HIGHBIT (op1)) /* shift right */
|
|
|
|
|
op2 >>= op1;
|
|
|
|
|
else /* shift left */
|
|
|
|
|
op2 <<= op1;
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) op2, -1);
|
|
|
|
|
(*_bfd_error_handler) (_("TIR_S_C_OPR_USH incomplete"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_ROT:
|
|
|
|
|
/*
|
|
|
|
|
* rotate
|
|
|
|
|
*/
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
op2 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
if (HIGHBIT (0)) /* shift right */
|
|
|
|
|
op2 >>= op1;
|
|
|
|
|
else /* shift left */
|
|
|
|
|
op2 <<= op1;
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) op2, -1);
|
|
|
|
|
(*_bfd_error_handler) (_("TIR_S_C_OPR_ROT incomplete"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_SEL:
|
|
|
|
|
/*
|
|
|
|
|
* select
|
|
|
|
|
*/
|
|
|
|
|
if ((long) _bfd_vms_pop (abfd, NULL) & 0x01L)
|
|
|
|
|
(void) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
op1 = (long) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
(void) _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
_bfd_vms_push (abfd, (uquad) op1, -1);
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case TIR_S_C_OPR_REDEF:
|
|
|
|
|
/*
|
|
|
|
|
* redefine symbol to current location
|
|
|
|
|
*/
|
|
|
|
|
(*_bfd_error_handler) (_("TIR_S_C_OPR_REDEF not supported"));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_OPR_DFLIT:
|
|
|
|
|
/*
|
|
|
|
|
* define a literal
|
|
|
|
|
*/
|
|
|
|
|
(*_bfd_error_handler) (_("TIR_S_C_OPR_DFLIT not supported"));
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
(*_bfd_error_handler) (_("Reserved OPR cmd %d"), ptr[-1]);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return ptr;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static unsigned char *
|
|
|
|
|
tir_ctl (bfd *abfd, unsigned char *ptr)
|
|
|
|
|
/*
|
|
|
|
|
* control commands
|
|
|
|
|
*
|
|
|
|
|
* see table 7-6 of the VAX/VMS linker manual
|
|
|
|
|
*/
|
|
|
|
|
{
|
|
|
|
|
unsigned long dummy;
|
1999-07-12 18:30:21 +08:00
|
|
|
|
unsigned int psect;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (5, "tir_ctl %d\n", *ptr);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (*ptr++)
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case TIR_S_C_CTL_SETRB:
|
|
|
|
|
/*
|
|
|
|
|
* set relocation base: pop stack, set image location counter
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
if (psect >= PRIV (section_count))
|
|
|
|
|
{
|
|
|
|
|
alloc_section (abfd, psect);
|
|
|
|
|
}
|
|
|
|
|
image_set_ptr (abfd, (int) psect, (uquad) dummy);
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_CTL_AUGRB:
|
|
|
|
|
/*
|
|
|
|
|
* augment relocation base: increment image location counter by offset
|
|
|
|
|
* arg: lw offset value
|
|
|
|
|
*/
|
|
|
|
|
dummy = bfd_getl32 (ptr);
|
|
|
|
|
image_inc_ptr (abfd, (uquad) dummy);
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_CTL_DFLOC:
|
|
|
|
|
/*
|
|
|
|
|
* define location: pop index, save location counter under index
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, NULL);
|
|
|
|
|
(*_bfd_error_handler) (_("TIR_S_C_CTL_DFLOC not fully implemented"));
|
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_CTL_STLOC:
|
|
|
|
|
/*
|
|
|
|
|
* set location: pop index, restore location counter from index
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
(*_bfd_error_handler) (_("TIR_S_C_CTL_STLOC not fully implemented"));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
case TIR_S_C_CTL_STKDL:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/*
|
|
|
|
|
* stack defined location: pop index, push location counter from index
|
|
|
|
|
* arg: -
|
|
|
|
|
*/
|
|
|
|
|
dummy = _bfd_vms_pop (abfd, &psect);
|
|
|
|
|
(*_bfd_error_handler) (_("TIR_S_C_CTL_STKDL not fully implemented"));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
break;
|
|
|
|
|
default:
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(*_bfd_error_handler) (_("Reserved CTL cmd %d"), ptr[-1]);
|
|
|
|
|
break;
|
|
|
|
|
}
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return ptr;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* handle command from TIR section
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static unsigned char *
|
|
|
|
|
tir_cmd (bfd *abfd, unsigned char *ptr)
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
struct
|
|
|
|
|
{
|
1999-05-03 15:29:11 +08:00
|
|
|
|
int mincod;
|
|
|
|
|
int maxcod;
|
2000-12-25 22:42:42 +08:00
|
|
|
|
unsigned char * (*explain) (bfd *, unsigned char *);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
}
|
|
|
|
|
tir_table[] =
|
|
|
|
|
{
|
|
|
|
|
{ 0, TIR_S_C_MAXSTACOD, tir_sta },
|
|
|
|
|
{ TIR_S_C_MINSTOCOD, TIR_S_C_MAXSTOCOD, tir_sto },
|
|
|
|
|
{ TIR_S_C_MINOPRCOD, TIR_S_C_MAXOPRCOD, tir_opr },
|
|
|
|
|
{ TIR_S_C_MINCTLCOD, TIR_S_C_MAXCTLCOD, tir_ctl },
|
|
|
|
|
{ -1, -1, NULL }
|
1999-05-03 15:29:11 +08:00
|
|
|
|
};
|
|
|
|
|
int i = 0;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (4, "tir_cmd %d/%x\n", *ptr, *ptr);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_hexdump (8, ptr, 16, (int) ptr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
if (*ptr & 0x80) /* store immediate */
|
|
|
|
|
{
|
|
|
|
|
i = 128 - (*ptr++ & 0x7f);
|
|
|
|
|
image_dump (abfd, ptr, i, 0);
|
|
|
|
|
ptr += i;
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
while (tir_table[i].mincod >= 0)
|
|
|
|
|
{
|
2000-12-25 22:42:42 +08:00
|
|
|
|
if ( (tir_table[i].mincod <= *ptr)
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
&& (*ptr <= tir_table[i].maxcod))
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
ptr = tir_table[i].explain (abfd, ptr);
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
i++;
|
|
|
|
|
}
|
|
|
|
|
if (tir_table[i].mincod < 0)
|
|
|
|
|
{
|
|
|
|
|
(*_bfd_error_handler) (_("Obj code %d not found"), *ptr);
|
|
|
|
|
ptr = 0;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return ptr;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* handle command from ETIR section */
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
etir_cmd (abfd, cmd, ptr)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int cmd;
|
|
|
|
|
unsigned char *ptr;
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
static struct
|
|
|
|
|
{
|
1999-05-03 15:29:11 +08:00
|
|
|
|
int mincod;
|
|
|
|
|
int maxcod;
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
boolean (*explain) PARAMS ((bfd *, int, unsigned char *));
|
|
|
|
|
}
|
|
|
|
|
etir_table[] =
|
|
|
|
|
{
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{ ETIR_S_C_MINSTACOD, ETIR_S_C_MAXSTACOD, etir_sta },
|
|
|
|
|
{ ETIR_S_C_MINSTOCOD, ETIR_S_C_MAXSTOCOD, etir_sto },
|
|
|
|
|
{ ETIR_S_C_MINOPRCOD, ETIR_S_C_MAXOPRCOD, etir_opr },
|
|
|
|
|
{ ETIR_S_C_MINCTLCOD, ETIR_S_C_MAXCTLCOD, etir_ctl },
|
|
|
|
|
{ ETIR_S_C_MINSTCCOD, ETIR_S_C_MAXSTCCOD, etir_stc },
|
|
|
|
|
{ -1, -1, NULL }
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
int i = 0;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (4, "etir_cmd %d/%x\n", cmd, cmd);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_hexdump (8, ptr, 16, (int) ptr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
while (etir_table[i].mincod >= 0)
|
|
|
|
|
{
|
2000-12-25 22:42:42 +08:00
|
|
|
|
if ( (etir_table[i].mincod <= cmd)
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
&& (cmd <= etir_table[i].maxcod))
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
if (!etir_table[i].explain (abfd, cmd, ptr))
|
|
|
|
|
return -1;
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
i++;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (4, "etir_cmd: = 0\n");
|
|
|
|
|
#endif
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Text Information and Relocation Records (OBJ$C_TIR)
|
|
|
|
|
handle tir record */
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
analyze_tir (abfd, ptr, length)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
unsigned char *ptr;
|
|
|
|
|
unsigned int length;
|
|
|
|
|
{
|
|
|
|
|
unsigned char *maxptr;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (3, "analyze_tir: %d bytes\n", length);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
maxptr = ptr + length;
|
|
|
|
|
|
|
|
|
|
while (ptr < maxptr)
|
|
|
|
|
{
|
|
|
|
|
ptr = tir_cmd (abfd, ptr);
|
|
|
|
|
if (ptr == 0)
|
|
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* Text Information and Relocation Records (EOBJ$C_ETIR)
|
|
|
|
|
handle etir record */
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
analyze_etir (abfd, ptr, length)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
unsigned char *ptr;
|
|
|
|
|
unsigned int length;
|
|
|
|
|
{
|
|
|
|
|
int cmd;
|
|
|
|
|
unsigned char *maxptr;
|
|
|
|
|
int result = 0;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (3, "analyze_etir: %d bytes\n", length);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
maxptr = ptr + length;
|
|
|
|
|
|
|
|
|
|
while (ptr < maxptr)
|
|
|
|
|
{
|
|
|
|
|
cmd = bfd_getl16 (ptr);
|
|
|
|
|
length = bfd_getl16 (ptr + 2);
|
|
|
|
|
result = etir_cmd (abfd, cmd, ptr+4);
|
|
|
|
|
if (result != 0)
|
|
|
|
|
break;
|
|
|
|
|
ptr += length;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (3, "analyze_etir: = %d\n", result);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
return result;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* process ETIR record
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return 0 on success, -1 on error */
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
_bfd_vms_slurp_tir (abfd, objtype)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int objtype;
|
|
|
|
|
{
|
|
|
|
|
int result;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (2, "TIR/ETIR\n");
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
switch (objtype)
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
case EOBJ_S_C_ETIR:
|
|
|
|
|
PRIV (vms_rec) += 4; /* skip type, size */
|
|
|
|
|
PRIV (rec_size) -= 4;
|
|
|
|
|
result = analyze_etir (abfd, PRIV (vms_rec), (unsigned) PRIV (rec_size));
|
|
|
|
|
break;
|
|
|
|
|
case OBJ_S_C_TIR:
|
|
|
|
|
PRIV (vms_rec) += 1; /* skip type */
|
|
|
|
|
PRIV (rec_size) -= 1;
|
|
|
|
|
result = analyze_tir (abfd, PRIV (vms_rec), (unsigned) PRIV (rec_size));
|
|
|
|
|
break;
|
|
|
|
|
default:
|
|
|
|
|
result = -1;
|
|
|
|
|
break;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return result;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* process EDBG record
|
|
|
|
|
return 0 on success, -1 on error
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
not implemented yet */
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
_bfd_vms_slurp_dbg (abfd, objtype)
|
|
|
|
|
bfd *abfd;
|
1999-07-12 18:30:21 +08:00
|
|
|
|
int objtype ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (2, "DBG/EDBG\n");
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
abfd->flags |= (HAS_DEBUG | HAS_LINENO);
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* process ETBT record
|
|
|
|
|
return 0 on success, -1 on error
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
not implemented yet */
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
_bfd_vms_slurp_tbt (abfd, objtype)
|
1999-07-12 18:30:21 +08:00
|
|
|
|
bfd *abfd ATTRIBUTE_UNUSED;
|
|
|
|
|
int objtype ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (2, "TBT/ETBT\n");
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* process LNK record
|
|
|
|
|
return 0 on success, -1 on error
|
2000-12-25 22:42:42 +08:00
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
not implemented yet */
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
_bfd_vms_slurp_lnk (abfd, objtype)
|
1999-07-12 18:30:21 +08:00
|
|
|
|
bfd *abfd ATTRIBUTE_UNUSED;
|
|
|
|
|
int objtype ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (2, "LNK\n");
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
/* */
|
|
|
|
|
/* WRITE ETIR SECTION */
|
|
|
|
|
/* */
|
|
|
|
|
/* this is still under construction and therefore not documented */
|
|
|
|
|
/* */
|
|
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
|
|
|
|
|
static void start_etir_record PARAMS ((bfd *abfd, int index, uquad offset, boolean justoffset));
|
|
|
|
|
static void sto_imm PARAMS ((bfd *abfd, vms_section *sptr, bfd_vma vaddr, int index));
|
|
|
|
|
static void end_etir_record PARAMS ((bfd *abfd));
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
sto_imm (abfd, sptr, vaddr, index)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
vms_section *sptr;
|
|
|
|
|
bfd_vma vaddr;
|
|
|
|
|
int index;
|
|
|
|
|
{
|
|
|
|
|
int size;
|
|
|
|
|
int ssize;
|
|
|
|
|
unsigned char *cptr;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (8, "sto_imm %d bytes\n", sptr->size);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_hexdump (9, sptr->contents, (int) sptr->size, (int) vaddr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
ssize = sptr->size;
|
|
|
|
|
cptr = sptr->contents;
|
|
|
|
|
|
|
|
|
|
while (ssize > 0)
|
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
size = ssize; /* try all the rest */
|
|
|
|
|
|
|
|
|
|
if (_bfd_vms_output_check (abfd, size) < 0)
|
|
|
|
|
{ /* doesn't fit, split ! */
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
start_etir_record (abfd, index, vaddr, false);
|
|
|
|
|
size = _bfd_vms_output_check (abfd, 0); /* get max size */
|
|
|
|
|
if (size > ssize) /* more than what's left ? */
|
|
|
|
|
size = ssize;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
_bfd_vms_output_begin (abfd, ETIR_S_C_STO_IMM, -1);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_output_long (abfd, (unsigned long) (size));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_dump (abfd, cptr, size);
|
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (10, "dumped %d bytes\n", size);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_hexdump (10, cptr, (int) size, (int) vaddr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
vaddr += size;
|
|
|
|
|
ssize -= size;
|
|
|
|
|
cptr += size;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*-------------------------------------------------------------------*/
|
|
|
|
|
|
|
|
|
|
/* start ETIR record for section #index at virtual addr offset. */
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
start_etir_record (abfd, index, offset, justoffset)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
int index;
|
|
|
|
|
uquad offset;
|
|
|
|
|
boolean justoffset;
|
|
|
|
|
{
|
|
|
|
|
if (!justoffset)
|
|
|
|
|
{
|
|
|
|
|
_bfd_vms_output_begin (abfd, EOBJ_S_C_ETIR, -1); /* one ETIR per section */
|
|
|
|
|
_bfd_vms_output_push (abfd);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
_bfd_vms_output_begin (abfd, ETIR_S_C_STA_PQ, -1); /* push start offset */
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_output_long (abfd, (unsigned long) index);
|
|
|
|
|
_bfd_vms_output_quad (abfd, (uquad) offset);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
|
|
|
|
|
_bfd_vms_output_begin (abfd, ETIR_S_C_CTL_SETRB, -1); /* start = pop () */
|
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* end etir record */
|
|
|
|
|
static void
|
|
|
|
|
end_etir_record (abfd)
|
|
|
|
|
bfd *abfd;
|
|
|
|
|
{
|
|
|
|
|
_bfd_vms_output_pop (abfd);
|
2000-12-25 22:42:42 +08:00
|
|
|
|
_bfd_vms_output_end (abfd);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* write section contents for bfd abfd */
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
_bfd_vms_write_tir (abfd, objtype)
|
|
|
|
|
bfd *abfd;
|
1999-07-12 18:30:21 +08:00
|
|
|
|
int objtype ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
asection *section;
|
|
|
|
|
vms_section *sptr;
|
|
|
|
|
int nextoffset;
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (2, "vms_write_tir (%p, %d)\n", abfd, objtype);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
_bfd_vms_output_alignment (abfd, 4);
|
|
|
|
|
|
|
|
|
|
nextoffset = 0;
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
PRIV (vms_linkage_index) = 1;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
/* dump all other sections */
|
|
|
|
|
|
|
|
|
|
section = abfd->sections;
|
|
|
|
|
|
|
|
|
|
while (section != NULL)
|
|
|
|
|
{
|
|
|
|
|
|
|
|
|
|
#if VMS_DEBUG
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_debug (4, "writing %d. section '%s' (%d bytes)\n",
|
|
|
|
|
section->index, section->name,
|
|
|
|
|
(int) (section->_raw_size));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
if (section->flags & SEC_RELOC)
|
|
|
|
|
{
|
|
|
|
|
int i;
|
|
|
|
|
|
|
|
|
|
if ((i = section->reloc_count) <= 0)
|
|
|
|
|
{
|
|
|
|
|
(*_bfd_error_handler) (_("SEC_RELOC with no relocs in section %s"),
|
|
|
|
|
section->name);
|
|
|
|
|
}
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
arelent **rptr;
|
|
|
|
|
_bfd_vms_debug (4, "%d relocations:\n", i);
|
|
|
|
|
rptr = section->orelocation;
|
|
|
|
|
while (i-- > 0)
|
|
|
|
|
{
|
|
|
|
|
_bfd_vms_debug (4, "sym %s in sec %s, value %08lx, addr %08lx, off %08lx, len %d: %s\n",
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(*(*rptr)->sym_ptr_ptr)->name,
|
|
|
|
|
(*(*rptr)->sym_ptr_ptr)->section->name,
|
|
|
|
|
(long) (*(*rptr)->sym_ptr_ptr)->value,
|
|
|
|
|
(*rptr)->address, (*rptr)->addend,
|
|
|
|
|
bfd_get_reloc_size ((*rptr)->howto),
|
|
|
|
|
(*rptr)->howto->name);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
rptr++;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if ((section->flags & SEC_HAS_CONTENTS)
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
&& (! bfd_is_com_section (section)))
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
bfd_vma vaddr; /* virtual addr in section */
|
|
|
|
|
|
|
|
|
|
sptr = _bfd_get_vms_section (abfd, section->index);
|
|
|
|
|
if (sptr == NULL)
|
|
|
|
|
{
|
|
|
|
|
bfd_set_error (bfd_error_no_contents);
|
|
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
|
2000-12-25 22:42:42 +08:00
|
|
|
|
vaddr = (bfd_vma) (sptr->offset);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
start_etir_record (abfd, section->index, (uquad) sptr->offset,
|
|
|
|
|
false);
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
while (sptr != NULL) /* one STA_PQ, CTL_SETRB per vms_section */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (section->flags & SEC_RELOC) /* check for relocs */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
arelent **rptr = section->orelocation;
|
|
|
|
|
int i = section->reloc_count;
|
|
|
|
|
for (;;)
|
|
|
|
|
{
|
|
|
|
|
bfd_size_type addr = (*rptr)->address;
|
1999-07-12 18:30:21 +08:00
|
|
|
|
bfd_size_type len = bfd_get_reloc_size ((*rptr)->howto);
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (sptr->offset < addr) /* sptr starts before reloc */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
1999-07-12 18:30:21 +08:00
|
|
|
|
bfd_size_type before = addr - sptr->offset;
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (sptr->size <= before) /* complete before */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
sto_imm (abfd, sptr, vaddr, section->index);
|
|
|
|
|
vaddr += sptr->size;
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
else /* partly before */
|
|
|
|
|
{
|
|
|
|
|
int after = sptr->size - before;
|
|
|
|
|
sptr->size = before;
|
|
|
|
|
sto_imm (abfd, sptr, vaddr, section->index);
|
|
|
|
|
vaddr += sptr->size;
|
|
|
|
|
sptr->contents += before;
|
|
|
|
|
sptr->offset += before;
|
|
|
|
|
sptr->size = after;
|
|
|
|
|
}
|
|
|
|
|
}
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
else if (sptr->offset == addr) /* sptr starts at reloc */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
asymbol *sym = *(*rptr)->sym_ptr_ptr;
|
|
|
|
|
asection *sec = sym->section;
|
|
|
|
|
|
|
|
|
|
switch ((*rptr)->howto->type)
|
|
|
|
|
{
|
|
|
|
|
case ALPHA_R_IGNORE:
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case ALPHA_R_REFLONG:
|
|
|
|
|
{
|
|
|
|
|
if (bfd_is_und_section (sym->section))
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
int slen = strlen ((char *) sym->name);
|
|
|
|
|
char *hash;
|
|
|
|
|
|
|
|
|
|
if (_bfd_vms_output_check (abfd, slen) < 0)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
start_etir_record (abfd,
|
|
|
|
|
section->index,
|
|
|
|
|
vaddr, false);
|
|
|
|
|
}
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STO_GBL_LW,
|
|
|
|
|
-1);
|
|
|
|
|
hash = (_bfd_vms_length_hash_symbol
|
|
|
|
|
(abfd, sym->name, EOBJ_S_C_SYMSIZ));
|
|
|
|
|
_bfd_vms_output_counted (abfd, hash);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
}
|
|
|
|
|
else if (bfd_is_abs_section (sym->section))
|
|
|
|
|
{
|
|
|
|
|
if (_bfd_vms_output_check (abfd, 16) < 0)
|
|
|
|
|
{
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
start_etir_record (abfd,
|
|
|
|
|
section->index,
|
|
|
|
|
vaddr, false);
|
|
|
|
|
}
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STA_LW,
|
|
|
|
|
-1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_quad (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(uquad) sym->value);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STO_LW,
|
|
|
|
|
-1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
if (_bfd_vms_output_check (abfd, 32) < 0)
|
|
|
|
|
{
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
start_etir_record (abfd,
|
|
|
|
|
section->index,
|
|
|
|
|
vaddr, false);
|
|
|
|
|
}
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STA_PQ,
|
|
|
|
|
-1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_long (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(unsigned long) (sec->index));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_quad (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
((uquad) (*rptr)->addend
|
|
|
|
|
+ (uquad) sym->value));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STO_LW,
|
|
|
|
|
-1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case ALPHA_R_REFQUAD:
|
|
|
|
|
{
|
|
|
|
|
if (bfd_is_und_section (sym->section))
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
int slen = strlen ((char *) sym->name);
|
|
|
|
|
char *hash;
|
|
|
|
|
if (_bfd_vms_output_check (abfd, slen) < 0)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
start_etir_record (abfd,
|
|
|
|
|
section->index,
|
|
|
|
|
vaddr, false);
|
|
|
|
|
}
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STO_GBL,
|
|
|
|
|
-1);
|
|
|
|
|
hash = (_bfd_vms_length_hash_symbol
|
|
|
|
|
(abfd, sym->name, EOBJ_S_C_SYMSIZ));
|
|
|
|
|
_bfd_vms_output_counted (abfd, hash);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
}
|
|
|
|
|
else if (bfd_is_abs_section (sym->section))
|
|
|
|
|
{
|
|
|
|
|
if (_bfd_vms_output_check (abfd, 16) < 0)
|
|
|
|
|
{
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
start_etir_record (abfd,
|
|
|
|
|
section->index,
|
|
|
|
|
vaddr, false);
|
|
|
|
|
}
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STA_QW,
|
|
|
|
|
-1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_quad (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(uquad) sym->value);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STO_QW,
|
|
|
|
|
-1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
if (_bfd_vms_output_check (abfd, 32) < 0)
|
|
|
|
|
{
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
start_etir_record (abfd,
|
|
|
|
|
section->index,
|
|
|
|
|
vaddr, false);
|
|
|
|
|
}
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STA_PQ,
|
|
|
|
|
-1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_long (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(unsigned long) (sec->index));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_quad (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
((uquad) (*rptr)->addend
|
|
|
|
|
+ (uquad) sym->value));
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STO_OFF,
|
|
|
|
|
-1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case ALPHA_R_HINT:
|
|
|
|
|
{
|
|
|
|
|
int hint_size;
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
char *hash ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
|
|
|
|
hint_size = sptr->size;
|
|
|
|
|
sptr->size = len;
|
|
|
|
|
sto_imm (abfd, sptr, vaddr, section->index);
|
|
|
|
|
sptr->size = hint_size;
|
|
|
|
|
#if 0
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
vms_output_begin (abfd,
|
|
|
|
|
ETIR_S_C_STO_HINT_GBL, -1);
|
|
|
|
|
vms_output_long (abfd,
|
|
|
|
|
(unsigned long) (sec->index));
|
|
|
|
|
vms_output_quad (abfd, (uquad) addr);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
hash = (_bfd_vms_length_hash_symbol
|
|
|
|
|
(abfd, sym->name, EOBJ_S_C_SYMSIZ));
|
|
|
|
|
vms_output_counted (abfd, hash);
|
|
|
|
|
|
|
|
|
|
vms_output_flush (abfd);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
case ALPHA_R_LINKAGE:
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
char *hash;
|
|
|
|
|
|
1999-05-03 15:29:11 +08:00
|
|
|
|
if (_bfd_vms_output_check (abfd, 64) < 0)
|
|
|
|
|
{
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
start_etir_record (abfd, section->index,
|
|
|
|
|
vaddr, false);
|
|
|
|
|
}
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STC_LP_PSB,
|
|
|
|
|
-1);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_long (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
(unsigned long) PRIV (vms_linkage_index));
|
|
|
|
|
PRIV (vms_linkage_index) += 2;
|
|
|
|
|
hash = (_bfd_vms_length_hash_symbol
|
|
|
|
|
(abfd, sym->name, EOBJ_S_C_SYMSIZ));
|
|
|
|
|
_bfd_vms_output_counted (abfd, hash);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_byte (abfd, 0);
|
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case ALPHA_R_CODEADDR:
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
int slen = strlen ((char *) sym->name);
|
|
|
|
|
char *hash;
|
|
|
|
|
if (_bfd_vms_output_check (abfd, slen) < 0)
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
start_etir_record (abfd,
|
|
|
|
|
section->index,
|
|
|
|
|
vaddr, false);
|
|
|
|
|
}
|
|
|
|
|
_bfd_vms_output_begin (abfd,
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
ETIR_S_C_STO_CA,
|
|
|
|
|
-1);
|
|
|
|
|
hash = (_bfd_vms_length_hash_symbol
|
|
|
|
|
(abfd, sym->name, EOBJ_S_C_SYMSIZ));
|
|
|
|
|
_bfd_vms_output_counted (abfd, hash);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
_bfd_vms_output_flush (abfd);
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
(*_bfd_error_handler) (_("Unhandled relocation %s"),
|
|
|
|
|
(*rptr)->howto->name);
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
vaddr += len;
|
|
|
|
|
|
|
|
|
|
if (len == sptr->size)
|
|
|
|
|
{
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
sptr->contents += len;
|
|
|
|
|
sptr->offset += len;
|
|
|
|
|
sptr->size -= len;
|
|
|
|
|
i--;
|
|
|
|
|
rptr++;
|
|
|
|
|
}
|
|
|
|
|
}
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
else /* sptr starts after reloc */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
i--; /* check next reloc */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
rptr++;
|
|
|
|
|
}
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
if (i==0) /* all reloc checked */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
if (sptr->size > 0)
|
|
|
|
|
{
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
/* dump rest */
|
|
|
|
|
sto_imm (abfd, sptr, vaddr, section->index);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
vaddr += sptr->size;
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
} /* for (;;) */
|
|
|
|
|
} /* if SEC_RELOC */
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
else /* no relocs, just dump */
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
sto_imm (abfd, sptr, vaddr, section->index);
|
|
|
|
|
vaddr += sptr->size;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
sptr = sptr->next;
|
|
|
|
|
|
|
|
|
|
} /* while (sptr != 0) */
|
|
|
|
|
|
|
|
|
|
end_etir_record (abfd);
|
|
|
|
|
|
|
|
|
|
} /* has_contents */
|
|
|
|
|
|
|
|
|
|
section = section->next;
|
|
|
|
|
}
|
|
|
|
|
|
Touches most files in bfd/, so likely will be blamed for everything..
o bfd_read and bfd_write lose an unnecessary param and become
bfd_bread and bfd_bwrite.
o bfd_*alloc now all take a bfd_size_type arg, and will error if
size_t is too small. eg. 32 bit host, 64 bit bfd, verrry big files
or bugs in linker scripts etc.
o file_ptr becomes a bfd_signed_vma. Besides matching sizes with
various other types involved in handling sections, this should make
it easier for bfd to support a 64 bit off_t on 32 bit hosts that
provide it.
o I've made the H_GET_* and H_PUT_* macros (which invoke bfd_h_{get,put}_*)
generally available. They now cast their args to bfd_vma and
bfd_byte * as appropriate, which removes a swag of casts from the
source.
o Bug fixes to bfd_get8, aix386_core_vec, elf32_h8_relax_section, and
aout-encap.c.
o Zillions of formatting and -Wconversion fixes.
2001-09-18 17:57:26 +08:00
|
|
|
|
_bfd_vms_output_alignment (abfd, 2);
|
1999-05-03 15:29:11 +08:00
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* write traceback data for bfd abfd */
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
_bfd_vms_write_tbt (abfd, objtype)
|
1999-07-12 18:30:21 +08:00
|
|
|
|
bfd *abfd ATTRIBUTE_UNUSED;
|
|
|
|
|
int objtype ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (2, "vms_write_tbt (%p, %d)\n", abfd, objtype);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* write debug info for bfd abfd */
|
|
|
|
|
|
|
|
|
|
int
|
|
|
|
|
_bfd_vms_write_dbg (abfd, objtype)
|
1999-07-12 18:30:21 +08:00
|
|
|
|
bfd *abfd ATTRIBUTE_UNUSED;
|
|
|
|
|
int objtype ATTRIBUTE_UNUSED;
|
1999-05-03 15:29:11 +08:00
|
|
|
|
{
|
|
|
|
|
#if VMS_DEBUG
|
|
|
|
|
_bfd_vms_debug (2, "vms_write_dbg (%p, objtype)\n", abfd, objtype);
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|