2
0
mirror of https://sourceware.org/git/binutils-gdb.git synced 2025-02-17 13:10:12 +08:00
binutils-gdb/ld/emultempl/msp430.em
Nick Alcock 3d16b64e28 bfd, include, ld, binutils, libctf: CTF should use the dynstr/sym
This is embarrassing.

The whole point of CTF is that it remains intact even after a binary is
stripped, providing a compact mapping from symbols to types for
everything in the externally-visible interface of an ELF object: it has
connections to the symbol table for that purpose, and to the string
table to avoid duplicating symbol names.  So it's a shame that the hooks
I implemented last year served to hook it up to the .symtab and .strtab,
which obviously disappear on strip, leaving any accompanying the CTF
dict containing references to strings (and, soon, symbols) which don't
exist any more because their containing strtab has been vaporized.  The
original Solaris design used .dynsym and .dynstr (well, actually,
.ldynsym, which has more symbols) which do not disappear. So should we.

Thankfully the work we did before serves as guide rails, and adjusting
things to use the .dynstr and .dynsym was fast and easy.  The only
annoyance is that the dynsym is assembled inside elflink.c in a fairly
piecemeal fashion, so that the easiest way to get the symbols out was to
hook in before every call to swap_symbol_out (we also leave in a hook in
front of symbol additions to the .symtab because it seems plausible that
we might want to hook them in future too: for now that hook is unused).
We adjust things so that rather than being offered a whole hash table of
symbols at once, libctf is now given symbols one at a time, with st_name
indexes already resolved and pointing at their final .dynstr offsets:
it's now up to libctf to resolve these to names as needed using the
strtab info we pass it separately.

Some bits might be contentious.  The ctf_new_dynstr callback takes an
elf_internal_sym, and this remains an elf_internal_sym right down
through the generic emulation layers into ldelfgen.  This is no worse
than the elf_sym_strtab we used to pass down, but in the future when we
gain non-ELF CTF symtab support we might want to lower the
elf_internal_sym to some other representation (perhaps a
ctf_link_symbol) in bfd or in ldlang_ctf_new_dynsym.  We rename the
'apply_strsym' hooks to 'acquire_strings' instead, becuse they no longer
have anything to do with symbols.

There are some API changes to pieces of API which are technically public
but actually totally unused by anything and/or unused by anything but ld
so they can change freely: the ctf_link_symbol gains new fields to allow
symbol names to be given as strtab offsets as well as strings, and a
symidx so that the symbol index can be passed in.  ctf_link_shuffle_syms
loses its callback parameter: the idea now is that linkers call the new
ctf_link_add_linker_symbol for every symbol in .dynsym, feed in all the
strtab entries with ctf_link_add_strtab, and then a call to
ctf_link_shuffle_syms will apply both and arrange to use them to reorder
the CTF symtab at CTF serialization time (which is coming in the next
commit).

Inside libctf we have a new preamble flag CTF_F_DYNSTR which is always
set in v3-format CTF dicts from this commit forwards: CTF dicts without
this flag are associated with .strtab like they used to be, so that old
dicts' external strings don't turn to garbage when loaded by new libctf.
Dicts with this flag are associated with .dynstr and .dynsym instead.
(The flag is not the next in sequence because this commit was written
quite late: the missing flags will be filled in by the next commit.)

Tests forthcoming in a later commit in this series.

bfd/ChangeLog
2020-11-20  Nick Alcock  <nick.alcock@oracle.com>

	* elflink.c (elf_finalize_dynstr): Call examine_strtab after
	dynstr finalization.
	(elf_link_swap_symbols_out): Don't call it here.  Call
	ctf_new_symbol before swap_symbol_out.
	(elf_link_output_extsym): Call ctf_new_dynsym before
	swap_symbol_out.
	(bfd_elf_final_link): Likewise.
	* elf.c (swap_out_syms): Pass in bfd_link_info.  Call
	ctf_new_symbol before swap_symbol_out.
	(_bfd_elf_compute_section_file_positions): Adjust.

binutils/ChangeLog
2020-11-20  Nick Alcock  <nick.alcock@oracle.com>

	* readelf.c (dump_section_as_ctf): Use .dynsym and .dynstr, not
	.symtab and .strtab.

include/ChangeLog
2020-11-20  Nick Alcock  <nick.alcock@oracle.com>

	* bfdlink.h (struct elf_sym_strtab): Replace with...
	(struct elf_internal_sym): ... this.
	(struct bfd_link_callbacks) <examine_strtab>: Take only a
	symstrtab argument.
	<ctf_new_symbol>: New.
	<ctf_new_dynsym>: Likewise.
	* ctf-api.h (struct ctf_link_sym) <st_symidx>: New.
	<st_nameidx>: Likewise.
	<st_nameidx_set>: Likewise.
	(ctf_link_iter_symbol_f): Removed.
	(ctf_link_shuffle_syms): Remove most parameters, just takes a
	ctf_dict_t now.
	(ctf_link_add_linker_symbol): New, split from
	ctf_link_shuffle_syms.
	* ctf.h (CTF_F_DYNSTR): New.
	(CTF_F_MAX): Adjust.

ld/ChangeLog
2020-11-20  Nick Alcock  <nick.alcock@oracle.com>

	* ldelfgen.c (struct ctf_strsym_iter_cb_arg): Rename to...
	(struct ctf_strtab_iter_cb_arg): ... this, changing fields:
	<syms>: Remove.
	<symcount>: Remove.
	<symstrtab>: Rename to...
	<strtab>: ... this.
	(ldelf_ctf_strtab_iter_cb): Adjust.
	(ldelf_ctf_symbols_iter_cb): Remove.
	(ldelf_new_dynsym_for_ctf): New, tell libctf about a single
	symbol.
	(ldelf_examine_strtab_for_ctf): Rename to...
	(ldelf_acquire_strings_for_ctf): ... this, only doing the strtab
	portion and not symbols.
	* ldelfgen.h: Adjust declarations accordingly.
	* ldemul.c (ldemul_examine_strtab_for_ctf): Rename to...
	(ldemul_acquire_strings_for_ctf): ... this.
	(ldemul_new_dynsym_for_ctf): New.
	* ldemul.h: Adjust declarations accordingly.
	* ldlang.c (ldlang_ctf_apply_strsym): Rename to...
	(ldlang_ctf_acquire_strings): ... this.
	(ldlang_ctf_new_dynsym): New.
	(lang_write_ctf): Call ldemul_new_dynsym_for_ctf with NULL to do
	the actual symbol shuffle.
	* ldlang.h (struct elf_strtab_hash): Adjust accordingly.
	* ldmain.c (bfd_link_callbacks): Wire up new/renamed callbacks.

libctf/ChangeLog
2020-11-20  Nick Alcock  <nick.alcock@oracle.com>

	* ctf-link.c (ctf_link_shuffle_syms): Adjust.
	(ctf_link_add_linker_symbol): New, unimplemented stub.
	* libctf.ver: Add it.
	* ctf-create.c (ctf_serialize): Set CTF_F_DYNSTR on newly-serialized
	dicts.
	* ctf-open-bfd.c (ctf_bfdopen_ctfsect): Check for the flag: open the
	symtab/strtab if not present, dynsym/dynstr otherwise.
	* ctf-archive.c (ctf_arc_bufpreamble): New, get the preamble from
	some arbitrary member of a CTF archive.
	* ctf-impl.h (ctf_arc_bufpreamble): Declare it.
2020-11-20 13:34:07 +00:00

948 lines
26 KiB
Plaintext
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

# This shell script emits a C file. -*- C -*-
# It does some substitutions.
fragment <<EOF
/* This file is is generated by a shell script. DO NOT EDIT! */
/* Emulate the original gld for the given ${EMULATION_NAME}
Copyright (C) 2014-2020 Free Software Foundation, Inc.
Written by Steve Chamberlain steve@cygnus.com
Extended for the MSP430 by Nick Clifton nickc@redhat.com
This file is part of the GNU Binutils.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street - Fifth Floor, Boston,
MA 02110-1301, USA. */
#define TARGET_IS_${EMULATION_NAME}
#include "sysdep.h"
#include "bfd.h"
#include "bfdlink.h"
#include "ctf-api.h"
#include "ld.h"
#include "getopt.h"
#include "ldmain.h"
#include "ldmisc.h"
#include "ldexp.h"
#include "ldlang.h"
#include "ldfile.h"
#include "ldemul.h"
#include "libiberty.h"
#include <ldgram.h>
enum regions
{
REGION_NONE = 0,
REGION_LOWER,
REGION_UPPER,
REGION_EITHER = 3,
};
enum either_placement_stage
{
LOWER_TO_UPPER,
UPPER_TO_LOWER,
};
enum { ROM, RAM };
static int data_region = REGION_NONE;
static int code_region = REGION_NONE;
static bfd_boolean disable_sec_transformation = FALSE;
#define MAX_PREFIX_LENGTH 7
EOF
# Import any needed special functions and/or overrides.
#
if test -n "$EXTRA_EM_FILE" ; then
source_em ${srcdir}/emultempl/${EXTRA_EM_FILE}.em
fi
if test x"$LDEMUL_BEFORE_PARSE" != xgld"$EMULATION_NAME"_before_parse; then
fragment <<EOF
static void
gld${EMULATION_NAME}_before_parse (void)
{
#ifndef TARGET_ /* I.e., if not generic. */
ldfile_set_output_arch ("`echo ${ARCH}`", bfd_arch_unknown);
#endif /* not TARGET_ */
/* The MSP430 port *needs* linker relaxtion in order to cope with large
functions where conditional branches do not fit into a +/- 1024 byte range. */
if (!bfd_link_relocatable (&link_info))
TARGET_ENABLE_RELAXATION;
}
EOF
fi
if test x"$LDEMUL_GET_SCRIPT" != xgld"$EMULATION_NAME"_get_script; then
fragment <<EOF
static char *
gld${EMULATION_NAME}_get_script (int *isfile)
EOF
if test x"$COMPILE_IN" = xyes
then
# Scripts compiled in.
# sed commands to quote an ld script as a C string.
sc="-f stringify.sed"
fragment <<EOF
{
*isfile = 0;
if (bfd_link_relocatable (&link_info) && config.build_constructors)
return
EOF
sed $sc ldscripts/${EMULATION_NAME}.xu >> e${EMULATION_NAME}.c
echo ' ; else if (bfd_link_relocatable (&link_info)) return' >> e${EMULATION_NAME}.c
sed $sc ldscripts/${EMULATION_NAME}.xr >> e${EMULATION_NAME}.c
echo ' ; else if (!config.text_read_only) return' >> e${EMULATION_NAME}.c
sed $sc ldscripts/${EMULATION_NAME}.xbn >> e${EMULATION_NAME}.c
echo ' ; else if (!config.magic_demand_paged) return' >> e${EMULATION_NAME}.c
sed $sc ldscripts/${EMULATION_NAME}.xn >> e${EMULATION_NAME}.c
echo ' ; else return' >> e${EMULATION_NAME}.c
sed $sc ldscripts/${EMULATION_NAME}.x >> e${EMULATION_NAME}.c
echo '; }' >> e${EMULATION_NAME}.c
else
# Scripts read from the filesystem.
fragment <<EOF
{
*isfile = 1;
if (bfd_link_relocatable (&link_info) && config.build_constructors)
return "ldscripts/${EMULATION_NAME}.xu";
else if (bfd_link_relocatable (&link_info))
return "ldscripts/${EMULATION_NAME}.xr";
else if (!config.text_read_only)
return "ldscripts/${EMULATION_NAME}.xbn";
else if (!config.magic_demand_paged)
return "ldscripts/${EMULATION_NAME}.xn";
else
return "ldscripts/${EMULATION_NAME}.x";
}
EOF
fi
fi
if test x"$LDEMUL_PLACE_ORPHAN" != xgld"$EMULATION_NAME"_place_orphan; then
fragment <<EOF
static unsigned int
data_statement_size (lang_data_statement_type *d)
{
unsigned int size = 0;
switch (d->type)
{
case QUAD:
case SQUAD:
size = QUAD_SIZE;
break;
case LONG:
size = LONG_SIZE;
break;
case SHORT:
size = SHORT_SIZE;
break;
case BYTE:
size = BYTE_SIZE;
break;
default:
einfo (_("%P: error: unhandled data_statement size\n"));
FAIL ();
}
return size;
}
/* Helper function for place_orphan that computes the size
of sections already mapped to the given statement. */
static bfd_size_type
scan_children (lang_statement_union_type * l)
{
bfd_size_type amount = 0;
while (l != NULL)
{
switch (l->header.type)
{
case lang_input_section_enum:
if (l->input_section.section->flags & SEC_ALLOC)
amount += l->input_section.section->size;
break;
case lang_constructors_statement_enum:
case lang_assignment_statement_enum:
case lang_padding_statement_enum:
break;
case lang_wild_statement_enum:
amount += scan_children (l->wild_statement.children.head);
break;
case lang_data_statement_enum:
amount += data_statement_size (&l->data_statement);
break;
default:
fprintf (stderr, "msp430 orphan placer: unhandled lang type %d\n", l->header.type);
break;
}
l = l->header.next;
}
return amount;
}
#define WARN_UPPER 0
#define WARN_LOWER 1
#define WARN_TEXT 0
#define WARN_DATA 1
#define WARN_BSS 2
#define WARN_RODATA 3
/* Warn only once per output section.
* NAME starts with ".upper." or ".lower.". */
static void
warn_no_output_section (const char *name)
{
static bfd_boolean warned[2][4] = {{FALSE, FALSE, FALSE, FALSE},
{FALSE, FALSE, FALSE, FALSE}};
int i = WARN_LOWER;
if (strncmp (name, ".upper.", 7) == 0)
i = WARN_UPPER;
if (!warned[i][WARN_TEXT] && strcmp (name + 6, ".text") == 0)
warned[i][WARN_TEXT] = TRUE;
else if (!warned[i][WARN_DATA] && strcmp (name + 6, ".data") == 0)
warned[i][WARN_DATA] = TRUE;
else if (!warned[i][WARN_BSS] && strcmp (name + 6, ".bss") == 0)
warned[i][WARN_BSS] = TRUE;
else if (!warned[i][WARN_RODATA] && strcmp (name + 6, ".rodata") == 0)
warned[i][WARN_RODATA] = TRUE;
else
return;
einfo ("%P: warning: no input section rule matches %s in linker script\n",
name);
}
/* Place an orphan section. We use this to put .either sections
into either their lower or their upper equivalents. */
static lang_output_section_statement_type *
gld${EMULATION_NAME}_place_orphan (asection * s,
const char * secname,
int constraint)
{
char * lower_name;
char * upper_name;
char * name;
lang_output_section_statement_type * lower;
lang_output_section_statement_type * upper;
if ((s->flags & SEC_ALLOC) == 0)
return NULL;
if (bfd_link_relocatable (&link_info))
return NULL;
/* If constraints are involved let the linker handle the placement normally. */
if (constraint != 0)
return NULL;
if (strncmp (secname, ".upper.", 7) == 0
|| strncmp (secname, ".lower.", 7) == 0)
{
warn_no_output_section (secname);
return NULL;
}
/* We only need special handling for .either sections. */
if (strncmp (secname, ".either.", 8) != 0)
return NULL;
/* Skip the .either prefix. */
secname += 7;
/* Compute the names of the corresponding upper and lower
sections. If the input section name contains another period,
only use the part of the name before the second dot. */
if (strchr (secname + 1, '.') != NULL)
{
name = xstrdup (secname);
* strchr (name + 1, '.') = 0;
}
else
name = (char *) secname;
lower_name = concat (".lower", name, NULL);
upper_name = concat (".upper", name, NULL);
/* Find the corresponding lower and upper sections. */
lower = lang_output_section_find (lower_name);
upper = lang_output_section_find (upper_name);
if (lower == NULL && upper == NULL)
{
einfo (_("%P: error: no section named %s or %s in linker script\n"),
lower_name, upper_name);
goto end;
}
else if (lower == NULL)
{
lower = lang_output_section_find (name);
if (lower == NULL)
{
einfo (_("%P: error: no section named %s in linker script\n"), name);
goto end;
}
}
/* Always place orphaned sections in lower. Optimal placement of either
sections is performed later, once section sizes have been finalized. */
lang_add_section (& lower->children, s, NULL, lower);
end:
free (upper_name);
free (lower_name);
return lower;
}
EOF
fi
fragment <<EOF
static bfd_boolean
change_output_section (lang_statement_union_type **head,
asection *s,
lang_output_section_statement_type *new_os,
lang_output_section_statement_type *old_os)
{
asection *is;
lang_statement_union_type * prev = NULL;
lang_statement_union_type * curr;
curr = *head;
while (curr != NULL)
{
switch (curr->header.type)
{
case lang_input_section_enum:
is = curr->input_section.section;
if (is == s)
{
lang_statement_list_type *old_list
= (lang_statement_list_type *) &old_os->children;
s->output_section = NULL;
lang_add_section (&new_os->children, s, NULL, new_os);
/* Remove the section from the old output section. */
if (prev == NULL)
*head = curr->header.next;
else
prev->header.next = curr->header.next;
/* If the input section we just moved is the tail of the old
output section, then we also need to adjust that tail. */
if (old_list->tail == (lang_statement_union_type **) curr)
old_list->tail = (lang_statement_union_type **) prev;
return TRUE;
}
break;
case lang_wild_statement_enum:
if (change_output_section (&(curr->wild_statement.children.head),
s, new_os, old_os))
return TRUE;
break;
default:
break;
}
prev = curr;
curr = curr->header.next;
}
return FALSE;
}
static void
add_region_prefix (bfd *abfd ATTRIBUTE_UNUSED, asection *s,
void *unused ATTRIBUTE_UNUSED)
{
const char *curr_name = bfd_section_name (s);
int region = REGION_NONE;
if (strncmp (curr_name, ".text", 5) == 0)
region = code_region;
else if (strncmp (curr_name, ".data", 5) == 0)
region = data_region;
else if (strncmp (curr_name, ".bss", 4) == 0)
region = data_region;
else if (strncmp (curr_name, ".rodata", 7) == 0)
region = data_region;
else
return;
switch (region)
{
case REGION_NONE:
break;
case REGION_UPPER:
bfd_rename_section (s, concat (".upper", curr_name, NULL));
break;
case REGION_LOWER:
bfd_rename_section (s, concat (".lower", curr_name, NULL));
break;
case REGION_EITHER:
s->name = concat (".either", curr_name, NULL);
break;
default:
/* Unreachable. */
FAIL ();
break;
}
}
static void
msp430_elf_after_open (void)
{
bfd *abfd;
gld${EMULATION_NAME}_after_open ();
/* If neither --code-region or --data-region have been passed, do not
transform sections names. */
if ((code_region == REGION_NONE && data_region == REGION_NONE)
|| disable_sec_transformation)
return;
for (abfd = link_info.input_bfds; abfd != NULL; abfd = abfd->link.next)
bfd_map_over_sections (abfd, add_region_prefix, NULL);
}
#define OPTION_CODE_REGION 321
#define OPTION_DATA_REGION (OPTION_CODE_REGION + 1)
#define OPTION_DISABLE_TRANS (OPTION_CODE_REGION + 2)
static void
gld${EMULATION_NAME}_add_options
(int ns, char **shortopts, int nl, struct option **longopts,
int nrl ATTRIBUTE_UNUSED, struct option **really_longopts ATTRIBUTE_UNUSED)
{
static const char xtra_short[] = { };
static const struct option xtra_long[] =
{
{ "code-region", required_argument, NULL, OPTION_CODE_REGION },
{ "data-region", required_argument, NULL, OPTION_DATA_REGION },
{ "disable-sec-transformation", no_argument, NULL,
OPTION_DISABLE_TRANS },
{ NULL, no_argument, NULL, 0 }
};
*shortopts = (char *) xrealloc (*shortopts, ns + sizeof (xtra_short));
memcpy (*shortopts + ns, &xtra_short, sizeof (xtra_short));
*longopts = (struct option *)
xrealloc (*longopts, nl * sizeof (struct option) + sizeof (xtra_long));
memcpy (*longopts + nl, &xtra_long, sizeof (xtra_long));
}
static void
gld${EMULATION_NAME}_list_options (FILE * file)
{
fprintf (file, _(" --code-region={either,lower,upper,none}\n\
Transform .text* sections to {either,lower,upper,none}.text* sections\n"));
fprintf (file, _(" --data-region={either,lower,upper,none}\n\
Transform .data*, .rodata* and .bss* sections to\n\
{either,lower,upper,none}.{bss,data,rodata}* sections\n"));
fprintf (file, _(" --disable-sec-transformation\n\
Disable transformation of .{text,data,bss,rodata}* sections to\n\
add the {either,lower,upper,none} prefixes\n"));
}
static bfd_boolean
gld${EMULATION_NAME}_handle_option (int optc)
{
switch (optc)
{
case OPTION_CODE_REGION:
if (strcmp (optarg, "upper") == 0)
code_region = REGION_UPPER;
else if (strcmp (optarg, "lower") == 0)
code_region = REGION_LOWER;
else if (strcmp (optarg, "either") == 0)
code_region = REGION_EITHER;
else if (strcmp (optarg, "none") == 0)
code_region = REGION_NONE;
else if (strlen (optarg) == 0)
{
einfo (_("%P: --code-region requires an argument: "
"{upper,lower,either,none}\n"));
return FALSE;
}
else
{
einfo (_("%P: error: unrecognized argument to --code-region= option: "
"\"%s\"\n"), optarg);
return FALSE;
}
break;
case OPTION_DATA_REGION:
if (strcmp (optarg, "upper") == 0)
data_region = REGION_UPPER;
else if (strcmp (optarg, "lower") == 0)
data_region = REGION_LOWER;
else if (strcmp (optarg, "either") == 0)
data_region = REGION_EITHER;
else if (strcmp (optarg, "none") == 0)
data_region = REGION_NONE;
else if (strlen (optarg) == 0)
{
einfo (_("%P: --data-region requires an argument: "
"{upper,lower,either,none}\n"));
return FALSE;
}
else
{
einfo (_("%P: error: unrecognized argument to --data-region= option: "
"\"%s\"\n"), optarg);
return FALSE;
}
break;
case OPTION_DISABLE_TRANS:
disable_sec_transformation = TRUE;
break;
default:
return FALSE;
}
return TRUE;
}
static void
eval_upper_either_sections (bfd *abfd ATTRIBUTE_UNUSED,
asection *s, void *data)
{
const char * base_sec_name;
const char * curr_name;
char * either_name;
int curr_region;
lang_output_section_statement_type * lower;
lang_output_section_statement_type * upper;
static bfd_size_type *lower_size = 0;
static bfd_size_type *upper_size = 0;
static bfd_size_type lower_size_rom = 0;
static bfd_size_type lower_size_ram = 0;
static bfd_size_type upper_size_rom = 0;
static bfd_size_type upper_size_ram = 0;
if ((s->flags & SEC_ALLOC) == 0)
return;
if (bfd_link_relocatable (&link_info))
return;
base_sec_name = (const char *) data;
curr_name = bfd_section_name (s);
/* Only concerned with .either input sections in the upper output section. */
either_name = concat (".either", base_sec_name, NULL);
if (strncmp (curr_name, either_name, strlen (either_name)) != 0
|| strncmp (s->output_section->name, ".upper", 6) != 0)
goto end;
lower = lang_output_section_find (concat (".lower", base_sec_name, NULL));
upper = lang_output_section_find (concat (".upper", base_sec_name, NULL));
if (upper == NULL || upper->region == NULL)
goto end;
else if (lower == NULL)
lower = lang_output_section_find (base_sec_name);
if (lower == NULL || lower->region == NULL)
goto end;
if (strcmp (base_sec_name, ".text") == 0
|| strcmp (base_sec_name, ".rodata") == 0)
curr_region = ROM;
else
curr_region = RAM;
if (curr_region == ROM)
{
if (lower_size_rom == 0)
{
lower_size_rom = lower->region->current - lower->region->origin;
upper_size_rom = upper->region->current - upper->region->origin;
}
lower_size = &lower_size_rom;
upper_size = &upper_size_rom;
}
else if (curr_region == RAM)
{
if (lower_size_ram == 0)
{
lower_size_ram = lower->region->current - lower->region->origin;
upper_size_ram = upper->region->current - upper->region->origin;
}
lower_size = &lower_size_ram;
upper_size = &upper_size_ram;
}
/* If the upper region is overflowing, try moving sections to the lower
region.
Note that there isn't any general benefit to using lower memory over upper
memory, so we only move sections around with the goal of making the program
fit. */
if (*upper_size > upper->region->length
&& *lower_size + s->size < lower->region->length)
{
if (change_output_section (&(upper->children.head), s, lower, upper))
{
*upper_size -= s->size;
*lower_size += s->size;
}
}
end:
free (either_name);
}
static void
eval_lower_either_sections (bfd *abfd ATTRIBUTE_UNUSED,
asection *s, void *data)
{
const char * base_sec_name;
const char * curr_name;
char * either_name;
int curr_region;
lang_output_section_statement_type * output_sec;
lang_output_section_statement_type * lower;
lang_output_section_statement_type * upper;
static bfd_size_type *lower_size = 0;
static bfd_size_type lower_size_rom = 0;
static bfd_size_type lower_size_ram = 0;
if ((s->flags & SEC_ALLOC) == 0)
return;
if (bfd_link_relocatable (&link_info))
return;
base_sec_name = (const char *) data;
curr_name = bfd_section_name (s);
/* Only concerned with .either input sections in the lower or "default"
output section i.e. not in the upper output section. */
either_name = concat (".either", base_sec_name, NULL);
if (strncmp (curr_name, either_name, strlen (either_name)) != 0
|| strncmp (s->output_section->name, ".upper", 6) == 0)
return;
if (strcmp (base_sec_name, ".text") == 0
|| strcmp (base_sec_name, ".rodata") == 0)
curr_region = ROM;
else
curr_region = RAM;
output_sec = lang_output_section_find (s->output_section->name);
/* If the output_section doesn't exist, this has already been reported in
place_orphan, so don't need to warn again. */
if (output_sec == NULL || output_sec->region == NULL)
goto end;
/* lower and output_sec might be the same, but in some cases an .either
section can end up in base_sec_name if it hasn't been placed by
place_orphan. */
lower = lang_output_section_find (concat (".lower", base_sec_name, NULL));
upper = lang_output_section_find (concat (".upper", base_sec_name, NULL));
if (upper == NULL)
goto end;
if (curr_region == ROM)
{
if (lower_size_rom == 0)
{
/* Get the size of other items in the lower region that aren't the
sections to be moved around. */
lower_size_rom
= (output_sec->region->current - output_sec->region->origin)
- scan_children (output_sec->children.head);
if (output_sec != lower && lower != NULL)
lower_size_rom -= scan_children (lower->children.head);
}
lower_size = &lower_size_rom;
}
else if (curr_region == RAM)
{
if (lower_size_ram == 0)
{
lower_size_ram
= (output_sec->region->current - output_sec->region->origin)
- scan_children (output_sec->children.head);
if (output_sec != lower && lower != NULL)
lower_size_ram -= scan_children (lower->children.head);
}
lower_size = &lower_size_ram;
}
/* Move sections that cause the lower region to overflow to the upper region. */
if (*lower_size + s->size > output_sec->region->length)
change_output_section (&(output_sec->children.head), s, upper, output_sec);
else
*lower_size += s->size;
end:
free (either_name);
}
/* This function is similar to lang_relax_sections, but without the size
evaluation code that is always executed after relaxation. */
static void
intermediate_relax_sections (void)
{
int i = link_info.relax_pass;
/* The backend can use it to determine the current pass. */
link_info.relax_pass = 0;
while (i--)
{
bfd_boolean relax_again;
link_info.relax_trip = -1;
do
{
link_info.relax_trip++;
lang_do_assignments (lang_assigning_phase_enum);
lang_reset_memory_regions ();
relax_again = FALSE;
lang_size_sections (&relax_again, FALSE);
}
while (relax_again);
link_info.relax_pass++;
}
}
static void
msp430_elf_after_allocation (void)
{
int relax_count = 0;
unsigned int i;
/* Go over each section twice, once to place either sections that don't fit
in lower into upper, and then again to move any sections in upper that
fit in lower into lower. */
for (i = 0; i < 8; i++)
{
int placement_stage = (i < 4) ? LOWER_TO_UPPER : UPPER_TO_LOWER;
const char * base_sec_name;
lang_output_section_statement_type * upper;
switch (i % 4)
{
default:
case 0:
base_sec_name = ".text";
break;
case 1:
base_sec_name = ".data";
break;
case 2:
base_sec_name = ".bss";
break;
case 3:
base_sec_name = ".rodata";
break;
}
upper = lang_output_section_find (concat (".upper", base_sec_name, NULL));
if (upper != NULL)
{
/* Can't just use one iteration over the all the sections to make
both lower->upper and upper->lower transformations because the
iterator encounters upper sections before all lower sections have
been examined. */
bfd *abfd;
if (placement_stage == LOWER_TO_UPPER)
{
/* Perform relaxation and get the final size of sections
before trying to fit .either sections in the correct
ouput sections. */
if (relax_count == 0)
{
intermediate_relax_sections ();
relax_count++;
}
for (abfd = link_info.input_bfds; abfd != NULL;
abfd = abfd->link.next)
{
bfd_map_over_sections (abfd, eval_lower_either_sections,
(void *) base_sec_name);
}
}
else if (placement_stage == UPPER_TO_LOWER)
{
/* Relax again before moving upper->lower. */
if (relax_count == 1)
{
intermediate_relax_sections ();
relax_count++;
}
for (abfd = link_info.input_bfds; abfd != NULL;
abfd = abfd->link.next)
{
bfd_map_over_sections (abfd, eval_upper_either_sections,
(void *) base_sec_name);
}
}
}
}
gld${EMULATION_NAME}_after_allocation ();
}
/* Return TRUE if a non-debug input section in L has positive size and matches
the given name. */
static int
input_section_exists (lang_statement_union_type * l, const char * name)
{
while (l != NULL)
{
switch (l->header.type)
{
case lang_input_section_enum:
if ((l->input_section.section->flags & SEC_ALLOC)
&& l->input_section.section->size > 0
&& !strcmp (l->input_section.section->name, name))
return TRUE;
break;
case lang_wild_statement_enum:
if (input_section_exists (l->wild_statement.children.head, name))
return TRUE;
break;
default:
break;
}
l = l->header.next;
}
return FALSE;
}
/* Some MSP430 linker scripts do not include ALIGN directives to ensure
__preinit_array_start, __init_array_start or __fini_array_start are word
aligned.
If __*_array_start symbols are not word aligned, the code in crt0 to run
through the array and call the functions will crash.
To avoid warning unnecessarily when the .*_array sections are not being
used for running constructors/destructors, only emit the warning if
the associated section exists and has size. */
static void
check_array_section_alignment (void)
{
int i;
lang_output_section_statement_type * rodata_sec;
lang_output_section_statement_type * rodata2_sec;
const char * array_names[3][2] = { { ".init_array", "__init_array_start" },
{ ".preinit_array", "__preinit_array_start" },
{ ".fini_array", "__fini_array_start" } };
/* .{preinit,init,fini}_array could be in either .rodata or .rodata2. */
rodata_sec = lang_output_section_find (".rodata");
rodata2_sec = lang_output_section_find (".rodata2");
if (rodata_sec == NULL && rodata2_sec == NULL)
return;
/* There are 3 .*_array sections which must be checked for alignment. */
for (i = 0; i < 3; i++)
{
struct bfd_link_hash_entry * sym;
if (((rodata_sec && input_section_exists (rodata_sec->children.head,
array_names[i][0]))
|| (rodata2_sec && input_section_exists (rodata2_sec->children.head,
array_names[i][0])))
&& (sym = bfd_link_hash_lookup (link_info.hash, array_names[i][1],
FALSE, FALSE, TRUE))
&& sym->type == bfd_link_hash_defined
&& sym->u.def.value % 2)
{
einfo ("%P: warning: \"%s\" symbol (%pU) is not word aligned\n",
array_names[i][1], NULL);
}
}
}
static void
gld${EMULATION_NAME}_finish (void)
{
finish_default ();
check_array_section_alignment ();
}
struct ld_emulation_xfer_struct ld_${EMULATION_NAME}_emulation =
{
${LDEMUL_BEFORE_PARSE-gld${EMULATION_NAME}_before_parse},
${LDEMUL_SYSLIB-syslib_default},
${LDEMUL_HLL-hll_default},
${LDEMUL_AFTER_PARSE-after_parse_default},
msp430_elf_after_open,
after_check_relocs_default,
before_place_orphans_default,
msp430_elf_after_allocation,
${LDEMUL_SET_OUTPUT_ARCH-set_output_arch_default},
${LDEMUL_CHOOSE_TARGET-ldemul_default_target},
${LDEMUL_BEFORE_ALLOCATION-before_allocation_default},
${LDEMUL_GET_SCRIPT-gld${EMULATION_NAME}_get_script},
"${EMULATION_NAME}",
"${OUTPUT_FORMAT}",
gld${EMULATION_NAME}_finish,
${LDEMUL_CREATE_OUTPUT_SECTION_STATEMENTS-NULL},
${LDEMUL_OPEN_DYNAMIC_ARCHIVE-NULL},
${LDEMUL_PLACE_ORPHAN-gld${EMULATION_NAME}_place_orphan},
${LDEMUL_SET_SYMBOLS-NULL},
${LDEMUL_PARSE_ARGS-NULL},
gld${EMULATION_NAME}_add_options,
gld${EMULATION_NAME}_handle_option,
${LDEMUL_UNRECOGNIZED_FILE-NULL},
gld${EMULATION_NAME}_list_options,
${LDEMUL_RECOGNIZED_FILE-NULL},
${LDEMUL_FIND_POTENTIAL_LIBRARIES-NULL},
${LDEMUL_NEW_VERS_PATTERN-NULL},
${LDEMUL_EXTRA_MAP_FILE_TEXT-NULL},
${LDEMUL_EMIT_CTF_EARLY-NULL},
${LDEMUL_ACQUIRE_STRINGS_FOR_CTF-NULL},
${LDEMUL_NEW_DYNSYM_FOR_CTF-NULL},
${LDEMUL_PRINT_SYMBOL-NULL}
};
EOF
#
# Local Variables:
# mode: c
# End: