This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: PATCH: Fix -Ttext and friends for ELF systems
Ian Lance Taylor wrote:
Mark Mitchell <mark@codesourcery.com> writes:
For a long time, -Ttext, -Tdata, and -Tbss have been useless (or
broken, depending on how you look at it) on ELF systems. Rather than
setting the segment base address, these options set the address of a
particular section. That tended to result in the linker putting that
section outside the relevant segment, without moving the rest of the
contents of the segment, and then issuing none-too-helpful errors
like:
Not enough room for program headers (allocated 3, need 4)
I do agree that the linker script has to cooperate. Suppose that,
rather than the simple mechanism of setting a symbol, you use the more
complex mechanism of adding a new expression to the linker script:
SEGMENT_START(NAME, DEFAULT). In your script, it might be used as
TEXT_START_ADDR='SEGMENT_START ("text", ${TEXT_START_ADDR})'
Here is a patch which implements your suggestion. In particular, for
linker scripts which do not use SEGMENT_START, there will be no change
in behavior.
Is this OK to commit?
--
Mark Mitchell
CodeSourcery, LLC
(916) 791-8304
mark@codesourcery.com
2004-10-25 Mark Mitchell <mark@codesourcery.com>
* Makefile.in (earmsymbian.c): Depend on armbpabi.sc, not elf.sc.
* ldexp.h (segment_type): New type.
(segments): New variable.
* ldexp.c (segments): New variable.
(exp_print_token): Handle SEGMENT_START.
(fold_binary): Likewise.
* ldgram.y (SEGMENT_START): Declare it as a token.
(exp): Handle SEGMENT_START.
* ldlang.h (lang_address_statement_type): Add segment field.
(lang_section_start): Change prototype.
* ldlang.c (map_input_to_output_sections): Do not process section
assignments if a corresponding SEGMENT_START has already been
seen.
(lang_section_start): Add segment parameter.
* ldlex.l (SEGMENT_START): Add it.
* lexsup.c (seg_segment_start): New function.
(parse_args): Use it for -Tbss, -Tdata, and -Ttext.
* ld.texinfo (SEGMENT_START): Document it.
* emulparams/armsymbian.sh (EMBEDDED): Set it.
* scripttempl/armbpabi.sc: Use SEGMENT_START to control segment
base addresses. Do not map relocations.
Index: Makefile.in
===================================================================
RCS file: /cvs/src/src/ld/Makefile.in,v
retrieving revision 1.173
diff -c -5 -p -r1.173 Makefile.in
*** Makefile.in 8 Oct 2004 00:22:13 -0000 1.173
--- Makefile.in 25 Oct 2004 19:29:12 -0000
*************** earm_epoc_pe.c: $(srcdir)/emulparams/arm
*** 1274,1284 ****
earmpe.c: $(srcdir)/emulparams/armpe.sh \
$(srcdir)/emultempl/pe.em $(srcdir)/scripttempl/pe.sc ${GEN_DEPENDS}
${GENSCRIPTS} armpe "$(tdir_armpe)"
earmsymbian.c: $(srcdir)/emulparams/armsymbian.sh \
$(srcdir)/emulparams/armelf.sh $(srcdir)/emultempl/elf32.em \
! $(srcdir)/emultempl/armelf.em $(srcdir)/scripttempl/elf.sc \
${GEN_DEPENDS}
${GENSCRIPTS} armsymbian "$(tdir_armelf)"
eavr2.c: $(srcdir)/emulparams/avr2.sh \
$(srcdir)/emultempl/generic.em $(srcdir)/scripttempl/avr.sc \
${GEN_DEPENDS}
--- 1274,1284 ----
earmpe.c: $(srcdir)/emulparams/armpe.sh \
$(srcdir)/emultempl/pe.em $(srcdir)/scripttempl/pe.sc ${GEN_DEPENDS}
${GENSCRIPTS} armpe "$(tdir_armpe)"
earmsymbian.c: $(srcdir)/emulparams/armsymbian.sh \
$(srcdir)/emulparams/armelf.sh $(srcdir)/emultempl/elf32.em \
! $(srcdir)/emultempl/armelf.em $(srcdir)/scripttempl/armbpabi.sc \
${GEN_DEPENDS}
${GENSCRIPTS} armsymbian "$(tdir_armelf)"
eavr2.c: $(srcdir)/emulparams/avr2.sh \
$(srcdir)/emultempl/generic.em $(srcdir)/scripttempl/avr.sc \
${GEN_DEPENDS}
Index: ld.texinfo
===================================================================
RCS file: /cvs/src/src/ld/ld.texinfo,v
retrieving revision 1.127
diff -c -5 -p -r1.127 ld.texinfo
*** ld.texinfo 16 Oct 2004 18:13:53 -0000 1.127
--- ld.texinfo 25 Oct 2004 19:29:12 -0000
*************** Returns the minimum of @var{exp1} and @v
*** 4673,4682 ****
--- 4673,4692 ----
Return the next unallocated address that is a multiple of @var{exp}.
This function is closely related to @code{ALIGN(@var{exp})}; unless you
use the @code{MEMORY} command to define discontinuous memory for the
output file, the two functions are equivalent.
+ @item SEGMENT_START(@var{segment}, @var{default})
+ @kindex SEGMENT_START(@var{segment}, @var{default})
+ Return the base address of the named @var{segment}. If an explicit
+ value has been given for this segment (with a command-line @samp{-T}
+ option) that value will be returned; otherwise the value will be
+ @var{default}. At present, the @samp{-T} command-line option can only
+ be used to set the base address for the ``text'', ``data'', and
+ ``bss'' sections, but you use @code{SEGMENT_START} with any segment
+ name.
+
@item SIZEOF(@var{section})
@kindex SIZEOF(@var{section})
@cindex section size
Return the size in bytes of the named @var{section}, if that section has
been allocated. If the section has not been allocated when this is
Index: ldexp.c
===================================================================
RCS file: /cvs/src/src/ld/ldexp.c,v
retrieving revision 1.37
diff -c -5 -p -r1.37 ldexp.c
*** ldexp.c 15 Oct 2004 06:00:15 -0000 1.37
--- ldexp.c 25 Oct 2004 19:29:12 -0000
*************** static etree_value_type exp_fold_tree_no
*** 46,55 ****
--- 46,57 ----
static bfd_vma align_n
(bfd_vma, bfd_vma);
struct exp_data_seg exp_data_seg;
+ segment_type *segments;
+
/* Print the string representation of the given token. Surround it
with spaces if INFIX_P is TRUE. */
static void
exp_print_token (token_code_type code, int infix_p)
*************** exp_print_token (token_code_type code, i
*** 100,110 ****
{ LOADADDR, "LOADADDR" },
{ MAX_K, "MAX_K" },
{ REL, "relocatable" },
{ DATA_SEGMENT_ALIGN, "DATA_SEGMENT_ALIGN" },
{ DATA_SEGMENT_RELRO_END, "DATA_SEGMENT_RELRO_END" },
! { DATA_SEGMENT_END, "DATA_SEGMENT_END" }
};
unsigned int idx;
for (idx = 0; idx < ARRAY_SIZE (table); idx++)
if (table[idx].code == code)
--- 102,113 ----
{ LOADADDR, "LOADADDR" },
{ MAX_K, "MAX_K" },
{ REL, "relocatable" },
{ DATA_SEGMENT_ALIGN, "DATA_SEGMENT_ALIGN" },
{ DATA_SEGMENT_RELRO_END, "DATA_SEGMENT_RELRO_END" },
! { DATA_SEGMENT_END, "DATA_SEGMENT_END" },
! { SEGMENT_START, "SEGMENT_START" }
};
unsigned int idx;
for (idx = 0; idx < ARRAY_SIZE (table); idx++)
if (table[idx].code == code)
*************** fold_binary (etree_type *tree,
*** 303,313 ****
{
etree_value_type result;
result = exp_fold_tree (tree->binary.lhs, current_section,
allocation_done, dot, dotp);
! if (result.valid_p)
{
etree_value_type other;
other = exp_fold_tree (tree->binary.rhs,
current_section,
--- 306,336 ----
{
etree_value_type result;
result = exp_fold_tree (tree->binary.lhs, current_section,
allocation_done, dot, dotp);
!
! /* The SEGMENT_START operator is special because its first
! operand is a string, not the name of a symbol. */
! if (result.valid_p && tree->type.node_code == SEGMENT_START)
! {
! const char *segment_name;
! segment_type *seg;
! /* Check to see if the user has overridden the default
! value. */
! segment_name = tree->binary.rhs->name.name;
! for (seg = segments; seg; seg = seg->next)
! if (strcmp (seg->name, segment_name) == 0)
! {
! seg->used = TRUE;
! result.value = seg->value;
! result.str = NULL;
! result.section = NULL;
! break;
! }
! }
! else if (result.valid_p)
{
etree_value_type other;
other = exp_fold_tree (tree->binary.rhs,
current_section,
Index: ldexp.h
===================================================================
RCS file: /cvs/src/src/ld/ldexp.h,v
retrieving revision 1.12
diff -c -5 -p -r1.12 ldexp.h
*** ldexp.h 11 May 2004 17:08:34 -0000 1.12
--- ldexp.h 25 Oct 2004 19:29:12 -0000
*************** extern struct exp_data_seg {
*** 101,110 ****
--- 101,126 ----
exp_dataseg_adjust
} phase;
bfd_vma base, relro_end, end, pagesize;
} exp_data_seg;
+ /* A maps from a segment name to a base address. */
+ typedef struct segment_struct {
+ /* The next segment in the linked list. */
+ struct segment_struct *next;
+ /* The name of the sgement. */
+ const char *name;
+ /* The base address for the segment. */
+ bfd_vma value;
+ /* True if a SEGMENT_START directive corresponding to this segment
+ has been seen. */
+ bfd_boolean used;
+ } segment_type;
+
+ /* The segments specified by the user on the command-line. */
+ extern segment_type *segments;
+
typedef struct _fill_type fill_type;
etree_type *exp_intop
(bfd_vma);
etree_type *exp_bigintop
Index: ldgram.y
===================================================================
RCS file: /cvs/src/src/ld/ldgram.y,v
retrieving revision 1.36
diff -c -5 -p -r1.36 ldgram.y
*** ldgram.y 4 Oct 2004 16:45:50 -0000 1.36
--- ldgram.y 25 Oct 2004 19:29:12 -0000
*************** static int error_index;
*** 132,141 ****
--- 132,142 ----
%token SORT_BY_NAME SORT_BY_ALIGNMENT
%token '{' '}'
%token SIZEOF_HEADERS OUTPUT_FORMAT FORCE_COMMON_ALLOCATION OUTPUT_ARCH
%token INHIBIT_COMMON_ALLOCATION
%token SIZEOF_HEADERS
+ %token SEGMENT_START
%token INCLUDE
%token MEMORY DEFSYMEND
%token NOLOAD DSECT COPY INFO OVERLAY
%token NAME LNAME DEFINED TARGET_K SEARCH_DIR MAP ENTRY
%token <integer> NEXT
*************** exp :
*** 841,850 ****
--- 842,860 ----
{ $$ = exp_binop (DATA_SEGMENT_ALIGN, $3, $5); }
| DATA_SEGMENT_RELRO_END '(' exp ',' exp ')'
{ $$ = exp_binop (DATA_SEGMENT_RELRO_END, $5, $3); }
| DATA_SEGMENT_END '(' exp ')'
{ $$ = exp_unop(DATA_SEGMENT_END, $3); }
+ | SEGMENT_START '(' NAME ',' exp ')'
+ { /* The operands to the expression node are
+ placed in the opposite order from the way
+ in which they appear in the script as
+ that allows us to reuse more code in
+ fold_binary. */
+ $$ = exp_binop (SEGMENT_START,
+ $5,
+ exp_nameop (NAME, $3)); }
| BLOCK '(' exp ')'
{ $$ = exp_unop(ALIGN_K,$3); }
| NAME
{ $$ = exp_nameop(NAME,$1); }
| MAX_K '(' exp ',' exp ')'
Index: ldlang.c
===================================================================
RCS file: /cvs/src/src/ld/ldlang.c,v
retrieving revision 1.167
diff -c -5 -p -r1.167 ldlang.c
*** ldlang.c 19 Oct 2004 15:44:58 -0000 1.167
--- ldlang.c 25 Oct 2004 19:29:12 -0000
*************** map_input_to_output_sections
*** 2650,2669 ****
break;
case lang_afile_asection_pair_statement_enum:
FAIL ();
break;
case lang_address_statement_enum:
! /* Mark the specified section with the supplied address. */
! {
! lang_output_section_statement_type *aos
! = (lang_output_section_statement_lookup
! (s->address_statement.section_name));
!
! if (aos->bfd_section == NULL)
! init_os (aos);
! aos->addr_tree = s->address_statement.address;
! }
break;
}
}
}
--- 2650,2680 ----
break;
case lang_afile_asection_pair_statement_enum:
FAIL ();
break;
case lang_address_statement_enum:
! /* Mark the specified section with the supplied address.
!
! If this section was actually a segment marker, then the
! directive is ignored if the linker script explicitly
! processed the segment marker. Originally, the linker
! treated segment directives (like -Ttext on the
! command-line) as section directives. We honor the
! section directive semantics for backwards compatibilty;
! linker scripts that do not specifically check for
! SEGMENT_START automatically get the old semantics. */
! if (!s->address_statement.segment
! || !s->address_statement.segment->used)
! {
! lang_output_section_statement_type *aos
! = (lang_output_section_statement_lookup
! (s->address_statement.section_name));
!
! if (aos->bfd_section == NULL)
! init_os (aos);
! aos->addr_tree = s->address_statement.address;
! }
break;
}
}
}
*************** lang_add_wild (struct wildcard_spec *fil
*** 4969,4985 ****
new->keep_sections = keep_sections;
lang_list_init (&new->children);
}
void
! lang_section_start (const char *name, etree_type *address)
{
lang_address_statement_type *ad;
ad = new_stat (lang_address_statement, stat_ptr);
ad->section_name = name;
ad->address = address;
}
/* Set the start symbol to NAME. CMDLINE is nonzero if this is called
because of a -e argument on the command line, or zero if this is
called by ENTRY in a linker script. Command line arguments take
--- 4980,4998 ----
new->keep_sections = keep_sections;
lang_list_init (&new->children);
}
void
! lang_section_start (const char *name, etree_type *address,
! const segment_type *segment)
{
lang_address_statement_type *ad;
ad = new_stat (lang_address_statement, stat_ptr);
ad->section_name = name;
ad->address = address;
+ ad->segment = segment;
}
/* Set the start symbol to NAME. CMDLINE is nonzero if this is called
because of a -e argument on the command line, or zero if this is
called by ENTRY in a linker script. Command line arguments take
Index: ldlang.h
===================================================================
RCS file: /cvs/src/src/ld/ldlang.h,v
retrieving revision 1.41
diff -c -5 -p -r1.41 ldlang.h
*** ldlang.h 16 Oct 2004 18:13:53 -0000 1.41
--- ldlang.h 25 Oct 2004 19:29:12 -0000
*************** typedef struct lang_wild_statement_struc
*** 314,323 ****
--- 314,324 ----
typedef struct lang_address_statement_struct
{
lang_statement_header_type header;
const char *section_name;
union etree_union *address;
+ const segment_type *segment;
} lang_address_statement_type;
typedef struct
{
lang_statement_header_type header;
*************** extern lang_output_section_statement_typ
*** 459,469 ****
extern void lang_final
(void);
extern void lang_process
(void);
extern void lang_section_start
! (const char *, union etree_union *);
extern void lang_add_entry
(const char *, bfd_boolean);
extern void lang_add_target
(const char *);
extern void lang_add_wild
--- 460,470 ----
extern void lang_final
(void);
extern void lang_process
(void);
extern void lang_section_start
! (const char *, union etree_union *, const segment_type *);
extern void lang_add_entry
(const char *, bfd_boolean);
extern void lang_add_target
(const char *);
extern void lang_add_wild
Index: ldlex.l
===================================================================
RCS file: /cvs/src/src/ld/ldlex.l,v
retrieving revision 1.26
diff -c -5 -p -r1.26 ldlex.l
*** ldlex.l 4 Oct 2004 16:45:51 -0000 1.26
--- ldlex.l 25 Oct 2004 19:29:12 -0000
*************** V_IDENTIFIER [*?.$_a-zA-Z\[\]\-\!\^\\]([
*** 258,268 ****
<EXPRESSION,BOTH>"ASSERT" { RTOKEN(ASSERT_K); }
<BOTH,SCRIPT>"ENTRY" { RTOKEN(ENTRY);}
<BOTH,SCRIPT,MRI>"EXTERN" { RTOKEN(EXTERN);}
<EXPRESSION,BOTH,SCRIPT>"NEXT" { RTOKEN(NEXT);}
<EXPRESSION,BOTH,SCRIPT>"sizeof_headers" { RTOKEN(SIZEOF_HEADERS);}
! <EXPRESSION,BOTH,SCRIPT>"SIZEOF_HEADERS" { RTOKEN(SIZEOF_HEADERS);}
<BOTH,SCRIPT>"MAP" { RTOKEN(MAP);}
<EXPRESSION,BOTH,SCRIPT>"SIZEOF" { RTOKEN(SIZEOF);}
<BOTH,SCRIPT>"TARGET" { RTOKEN(TARGET_K);}
<BOTH,SCRIPT>"SEARCH_DIR" { RTOKEN(SEARCH_DIR);}
<BOTH,SCRIPT>"OUTPUT" { RTOKEN(OUTPUT);}
--- 258,270 ----
<EXPRESSION,BOTH>"ASSERT" { RTOKEN(ASSERT_K); }
<BOTH,SCRIPT>"ENTRY" { RTOKEN(ENTRY);}
<BOTH,SCRIPT,MRI>"EXTERN" { RTOKEN(EXTERN);}
<EXPRESSION,BOTH,SCRIPT>"NEXT" { RTOKEN(NEXT);}
<EXPRESSION,BOTH,SCRIPT>"sizeof_headers" { RTOKEN(SIZEOF_HEADERS);}
! <EXPRESSION,BOTH,SCRIPT>"SIZEOF_HEADERS" {
! RTOKEN(SIZEOF_HEADERS);}
! <EXPRESSION,BOTH,SCRIPT>"SEGMENT_START" { RTOKEN(SEGMENT_START);}
<BOTH,SCRIPT>"MAP" { RTOKEN(MAP);}
<EXPRESSION,BOTH,SCRIPT>"SIZEOF" { RTOKEN(SIZEOF);}
<BOTH,SCRIPT>"TARGET" { RTOKEN(TARGET_K);}
<BOTH,SCRIPT>"SEARCH_DIR" { RTOKEN(SEARCH_DIR);}
<BOTH,SCRIPT>"OUTPUT" { RTOKEN(OUTPUT);}
Index: lexsup.c
===================================================================
RCS file: /cvs/src/src/ld/lexsup.c,v
retrieving revision 1.78
diff -c -5 -p -r1.78 lexsup.c
*** lexsup.c 7 Oct 2004 14:45:22 -0000 1.78
--- lexsup.c 25 Oct 2004 19:29:12 -0000
***************
*** 53,62 ****
--- 53,63 ----
#define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR)
#endif
static void set_default_dirlist (char *);
static void set_section_start (char *, char *);
+ static void set_segment_start (const char *, char *);
static void help (void);
/* Non-zero if we are processing a --defsym from the command line. */
int parsing_defsym = 0;
*************** parse_args (unsigned argc, char **argv)
*** 1130,1146 ****
case OPTION_TARGET_HELP:
/* Mention any target specific options. */
ldemul_list_emulation_options (stdout);
exit (0);
case OPTION_TBSS:
! set_section_start (".bss", optarg);
break;
case OPTION_TDATA:
! set_section_start (".data", optarg);
break;
case OPTION_TTEXT:
! set_section_start (".text", optarg);
break;
case OPTION_TRADITIONAL_FORMAT:
link_info.traditional_format = TRUE;
break;
case OPTION_TASK_LINK:
--- 1131,1147 ----
case OPTION_TARGET_HELP:
/* Mention any target specific options. */
ldemul_list_emulation_options (stdout);
exit (0);
case OPTION_TBSS:
! set_segment_start (".bss", optarg);
break;
case OPTION_TDATA:
! set_segment_start (".data", optarg);
break;
case OPTION_TTEXT:
! set_segment_start (".text", optarg);
break;
case OPTION_TRADITIONAL_FORMAT:
link_info.traditional_format = TRUE;
break;
case OPTION_TASK_LINK:
*************** set_section_start (char *sect, char *val
*** 1372,1383 ****
{
const char *end;
bfd_vma val = bfd_scan_vma (valstr, &end, 16);
if (*end)
einfo (_("%P%F: invalid hex number `%s'\n"), valstr);
! lang_section_start (sect, exp_intop (val));
}
/* Print help messages for the options. */
static void
help (void)
--- 1373,1420 ----
{
const char *end;
bfd_vma val = bfd_scan_vma (valstr, &end, 16);
if (*end)
einfo (_("%P%F: invalid hex number `%s'\n"), valstr);
! lang_section_start (sect, exp_intop (val), NULL);
}
+
+ static void
+ set_segment_start (const char *section, char *valstr)
+ {
+ const char *name;
+ const char *end;
+ segment_type *seg;
+
+ bfd_vma val = bfd_scan_vma (valstr, &end, 16);
+ if (*end)
+ einfo (_("%P%F: invalid hex number `%s'\n"), valstr);
+ /* If we already have an entry for this segment, update the existing
+ value. */
+ name = section + 1;
+ for (seg = segments; seg; seg = seg->next)
+ if (strcmp (seg->name, name) == 0)
+ {
+ seg->value = val;
+ return;
+ }
+ /* There was no existing value so we must create a new segment
+ entry. */
+ seg = stat_alloc (sizeof (*seg));
+ seg->name = name;
+ seg->value = val;
+ seg->used = FALSE;
+ /* Add it to the linked list of segments. */
+ seg->next = segments;
+ segments = seg;
+ /* Historically, -Ttext and friends set the base address of a
+ particular section. For backwards compatibility, we still do
+ that. If a SEGMENT_START directive is seen, the section address
+ assignment will be disabled. */
+ lang_section_start (section, exp_intop (val), seg);
+ }
+
/* Print help messages for the options. */
static void
help (void)
Index: emulparams/armsymbian.sh
===================================================================
RCS file: /cvs/src/src/ld/emulparams/armsymbian.sh,v
retrieving revision 1.4
diff -c -5 -p -r1.4 armsymbian.sh
*** emulparams/armsymbian.sh 30 Sep 2004 17:03:51 -0000 1.4
--- emulparams/armsymbian.sh 25 Oct 2004 19:29:12 -0000
*************** GENERATE_COMBRELOC_SCRIPT=1
*** 4,13 ****
--- 4,15 ----
OUTPUT_FORMAT="elf32-littlearm-symbian"
BIG_OUTPUT_FORMAT="elf32-bigarm-symbian"
LITTLE_OUTPUT_FORMAT="$OUTPUT_FORMAT"
TARGET1_IS_REL=1
TARGET2_TYPE=abs
+ # On BPABI systems, program headers should not be mapped.
+ EMBEDDED=yes
# This value should match ELF_MAXPAGESIZE in BFD. Otherwise, elf.c
# will not place read-write sections in a separate ELF segment from
# the read-only sections.
MAXPAGESIZE=0x8000
Index: scripttempl/armbpabi.sc
===================================================================
RCS file: /cvs/src/src/ld/scripttempl/armbpabi.sc,v
retrieving revision 1.2
diff -c -5 -p -r1.2 armbpabi.sc
*** scripttempl/armbpabi.sc 13 Oct 2004 17:45:31 -0000 1.2
--- scripttempl/armbpabi.sc 25 Oct 2004 19:29:19 -0000
***************
*** 1,80 ****
# This variant of elf.sc is used for ARM BPABI platforms, like Symbian
# OS, where a separate postlinker will operated on the generated
! # executable or shared object.
!
! #
! # Unusual variables checked by this code:
! # NOP - four byte opcode for no-op (defaults to 0)
! # NO_SMALL_DATA - no .sbss/.sbss2/.sdata/.sdata2 sections if not
! # empty.
! # DATA_ADDR - if end-of-text-plus-one-page isn't right for data start
! # INITIAL_READONLY_SECTIONS - at start of text segment
! # OTHER_READONLY_SECTIONS - other than .text .init .rodata ...
! # (e.g., .PARISC.milli)
! # OTHER_TEXT_SECTIONS - these get put in .text when relocating
! # OTHER_READWRITE_SECTIONS - other than .data .bss .ctors .sdata ...
! # (e.g., .PARISC.global)
! # OTHER_RELRO_SECTIONS - other than .data.rel.ro ...
! # (e.g. PPC32 .fixup, .got[12])
! # OTHER_BSS_SECTIONS - other than .bss .sbss ...
! # OTHER_SECTIONS - at the end
! # EXECUTABLE_SYMBOLS - symbols that must be defined for an
! # executable (e.g., _DYNAMIC_LINK)
! # TEXT_START_SYMBOLS - symbols that appear at the start of the
! # .text section.
! # DATA_START_SYMBOLS - symbols that appear at the start of the
! # .data section.
! # OTHER_SDATA_SECTIONS - sections just after .sdata.
! # OTHER_BSS_SYMBOLS - symbols that appear at the start of the
! # .bss section besides __bss_start.
! # DATA_PLT - .plt should be in data segment, not text segment.
! # PLT_BEFORE_GOT - .plt just before .got when .plt is in data segement.
! # BSS_PLT - .plt should be in bss segment
! # TEXT_DYNAMIC - .dynamic in text segment, not data segment.
! # EMBEDDED - whether this is for an embedded system.
! # SHLIB_TEXT_START_ADDR - if set, add to SIZEOF_HEADERS to set
! # start address of shared library.
! # INPUT_FILES - INPUT command of files to always include
! # WRITABLE_RODATA - if set, the .rodata section should be writable
! # INIT_START, INIT_END - statements just before and just after
! # combination of .init sections.
! # FINI_START, FINI_END - statements just before and just after
! # combination of .fini sections.
! # STACK_ADDR - start of a .stack section.
! # OTHER_END_SYMBOLS - symbols to place right at the end of the script.
! # SEPARATE_GOTPLT - if set, .got.plt should be separate output section,
! # so that .got can be in the RELRO area. It should be set to
! # the number of bytes in the beginning of .got.plt which can be
! # in the RELRO area as well.
! #
! # When adding sections, do note that the names of some sections are used
! # when specifying the start address of the next.
! #
!
! # Many sections come in three flavours. There is the 'real' section,
! # like ".data". Then there are the per-procedure or per-variable
! # sections, generated by -ffunction-sections and -fdata-sections in GCC,
! # and useful for --gc-sections, which for a variable "foo" might be
! # ".data.foo". Then there are the linkonce sections, for which the linker
! # eliminates duplicates, which are named like ".gnu.linkonce.d.foo".
! # The exact correspondences are:
! #
! # Section Linkonce section
! # .text .gnu.linkonce.t.foo
! # .rodata .gnu.linkonce.r.foo
! # .data .gnu.linkonce.d.foo
! # .bss .gnu.linkonce.b.foo
! # .sdata .gnu.linkonce.s.foo
! # .sbss .gnu.linkonce.sb.foo
! # .sdata2 .gnu.linkonce.s2.foo
! # .sbss2 .gnu.linkonce.sb2.foo
! # .debug_info .gnu.linkonce.wi.foo
! # .tdata .gnu.linkonce.td.foo
! # .tbss .gnu.linkonce.tb.foo
! #
! # Each of these can also have corresponding .rel.* and .rela.* sections.
test -z "$ENTRY" && ENTRY=_start
test -z "${BIG_OUTPUT_FORMAT}" && BIG_OUTPUT_FORMAT=${OUTPUT_FORMAT}
test -z "${LITTLE_OUTPUT_FORMAT}" && LITTLE_OUTPUT_FORMAT=${OUTPUT_FORMAT}
if [ -z "$MACHINE" ]; then OUTPUT_ARCH=${ARCH}; else OUTPUT_ARCH=${ARCH}:${MACHINE}; fi
--- 1,9 ----
# This variant of elf.sc is used for ARM BPABI platforms, like Symbian
# OS, where a separate postlinker will operated on the generated
! # executable or shared object. See elf.sc for configuration variables
! # that apply; only BPABI-specific variables will be noted here.
test -z "$ENTRY" && ENTRY=_start
test -z "${BIG_OUTPUT_FORMAT}" && BIG_OUTPUT_FORMAT=${OUTPUT_FORMAT}
test -z "${LITTLE_OUTPUT_FORMAT}" && LITTLE_OUTPUT_FORMAT=${OUTPUT_FORMAT}
if [ -z "$MACHINE" ]; then OUTPUT_ARCH=${ARCH}; else OUTPUT_ARCH=${ARCH}:${MACHINE}; fi
*************** STACK=" .stack ${RELOCATING-0}${
*** 173,187 ****
--- 102,123 ----
{
${RELOCATING+_stack = .;}
*(.stack)
}"
+ TEXT_START_ADDR="SEGMENT_START(\"text\", ${TEXT_START_ADDR})"
+ SHLIB_TEXT_START_ADDR="SEGMENT_START(\"text\", ${SHLIB_TEXT_START_ADDR:-0})"
+ DATA_ADDR="SEGMENT_START(\"data\", ${DATA_ADDR-${DATA_SEGMENT_ALIGN}})"
+ SHLIB_DATA_ADDR="SEGMENT_START(\"data\", ${SHLIB_DATA_ADDR-${DATA_SEGMENT_ALIGN}})"
+
# if this is for an embedded system, don't add SIZEOF_HEADERS.
if [ -z "$EMBEDDED" ]; then
test -z "${TEXT_BASE_ADDRESS}" && TEXT_BASE_ADDRESS="${TEXT_START_ADDR} + SIZEOF_HEADERS"
+ SHLIB_BASE_ADDRESS="${SHLIB_TEXT_START_ADDR} + SIZEOF_HEADERS"
else
test -z "${TEXT_BASE_ADDRESS}" && TEXT_BASE_ADDRESS="${TEXT_START_ADDR}"
+ SHLIB_BASE_ADDRESS="${SHLIB_TEXT_START_ADDR}"
fi
cat <<EOF
OUTPUT_FORMAT("${OUTPUT_FORMAT}", "${BIG_OUTPUT_FORMAT}",
"${LITTLE_OUTPUT_FORMAT}")
*************** ${RELOCATING- /* For some reason, the So
*** 200,269 ****
SECTIONS
{
/* Read-only sections, merged into text segment: */
${CREATE_SHLIB-${CREATE_PIE-${RELOCATING+PROVIDE (__executable_start = ${TEXT_START_ADDR}); . = ${TEXT_BASE_ADDRESS};}}}
! ${CREATE_SHLIB+${RELOCATING+. = ${SHLIB_TEXT_START_ADDR:-0} + SIZEOF_HEADERS;}}
! ${CREATE_PIE+${RELOCATING+. = ${SHLIB_TEXT_START_ADDR:-0} + SIZEOF_HEADERS;}}
${INITIAL_READONLY_SECTIONS}
EOF
- if [ "x$COMBRELOC" = x ]; then
- COMBRELOCCAT=cat
- else
- COMBRELOCCAT="cat > $COMBRELOC"
- fi
- eval $COMBRELOCCAT <<EOF
- .rel.init ${RELOCATING-0} : { *(.rel.init) }
- .rela.init ${RELOCATING-0} : { *(.rela.init) }
- .rel.text ${RELOCATING-0} : { *(.rel.text${RELOCATING+ .rel.text.* .rel.gnu.linkonce.t.*}) }
- .rela.text ${RELOCATING-0} : { *(.rela.text${RELOCATING+ .rela.text.* .rela.gnu.linkonce.t.*}) }
- .rel.fini ${RELOCATING-0} : { *(.rel.fini) }
- .rela.fini ${RELOCATING-0} : { *(.rela.fini) }
- .rel.rodata ${RELOCATING-0} : { *(.rel.rodata${RELOCATING+ .rel.rodata.* .rel.gnu.linkonce.r.*}) }
- .rela.rodata ${RELOCATING-0} : { *(.rela.rodata${RELOCATING+ .rela.rodata.* .rela.gnu.linkonce.r.*}) }
- ${OTHER_READONLY_RELOC_SECTIONS}
- .rel.data.rel.ro ${RELOCATING-0} : { *(.rel.data.rel.ro${RELOCATING+*}) }
- .rela.data.rel.ro ${RELOCATING-0} : { *(.rel.data.rel.ro${RELOCATING+*}) }
- .rel.data ${RELOCATING-0} : { *(.rel.data${RELOCATING+ .rel.data.* .rel.gnu.linkonce.d.*}) }
- .rela.data ${RELOCATING-0} : { *(.rela.data${RELOCATING+ .rela.data.* .rela.gnu.linkonce.d.*}) }
- .rel.tdata ${RELOCATING-0} : { *(.rel.tdata${RELOCATING+ .rel.tdata.* .rel.gnu.linkonce.td.*}) }
- .rela.tdata ${RELOCATING-0} : { *(.rela.tdata${RELOCATING+ .rela.tdata.* .rela.gnu.linkonce.td.*}) }
- .rel.tbss ${RELOCATING-0} : { *(.rel.tbss${RELOCATING+ .rel.tbss.* .rel.gnu.linkonce.tb.*}) }
- .rela.tbss ${RELOCATING-0} : { *(.rela.tbss${RELOCATING+ .rela.tbss.* .rela.gnu.linkonce.tb.*}) }
- .rel.ctors ${RELOCATING-0} : { *(.rel.ctors) }
- .rela.ctors ${RELOCATING-0} : { *(.rela.ctors) }
- .rel.dtors ${RELOCATING-0} : { *(.rel.dtors) }
- .rela.dtors ${RELOCATING-0} : { *(.rela.dtors) }
- ${REL_SDATA}
- ${REL_SBSS}
- ${REL_SDATA2}
- ${REL_SBSS2}
- .rel.bss ${RELOCATING-0} : { *(.rel.bss${RELOCATING+ .rel.bss.* .rel.gnu.linkonce.b.*}) }
- .rela.bss ${RELOCATING-0} : { *(.rela.bss${RELOCATING+ .rela.bss.* .rela.gnu.linkonce.b.*}) }
- EOF
- if [ -n "$COMBRELOC" ]; then
- cat <<EOF
- .rel.dyn ${RELOCATING-0} :
- {
- EOF
- sed -e '/^[ ]*[{}][ ]*$/d;/:[ ]*$/d;/\.rela\./d;s/^.*: { *\(.*\)}$/ \1/' $COMBRELOC
cat <<EOF
- }
- .rela.dyn ${RELOCATING-0} :
- {
- EOF
- sed -e '/^[ ]*[{}][ ]*$/d;/:[ ]*$/d;/\.rel\./d;s/^.*: { *\(.*\)}/ \1/' $COMBRELOC
- cat <<EOF
- }
- EOF
- fi
- cat <<EOF
- .rel.plt ${RELOCATING-0} : { *(.rel.plt) }
- .rela.plt ${RELOCATING-0} : { *(.rela.plt) }
- ${OTHER_PLT_RELOC_SECTIONS}
-
.init ${RELOCATING-0} :
{
${RELOCATING+${INIT_START}}
KEEP (*(.init))
${RELOCATING+${INIT_END}}
--- 136,151 ----
SECTIONS
{
/* Read-only sections, merged into text segment: */
${CREATE_SHLIB-${CREATE_PIE-${RELOCATING+PROVIDE (__executable_start = ${TEXT_START_ADDR}); . = ${TEXT_BASE_ADDRESS};}}}
! ${CREATE_SHLIB+${RELOCATING+. = ${SHLIB_BASE_ADDRESS};}}
! ${CREATE_PIE+${RELOCATING+. = ${SHLIB_BASE_ADDRESS};}}
${INITIAL_READONLY_SECTIONS}
EOF
cat <<EOF
.init ${RELOCATING-0} :
{
${RELOCATING+${INIT_START}}
KEEP (*(.init))
${RELOCATING+${INIT_END}}
*************** cat <<EOF
*** 297,309 ****
.eh_frame ${RELOCATING-0} : ONLY_IF_RO { KEEP (*(.eh_frame)) }
.gcc_except_table ${RELOCATING-0} : ONLY_IF_RO { KEEP (*(.gcc_except_table)) *(.gcc_except_table.*) }
/* Adjust the address for the data segment. We want to adjust up to
the same address within the page on the next page up. */
! ${CREATE_SHLIB-${CREATE_PIE-${RELOCATING+. = ${DATA_ADDR-${DATA_SEGMENT_ALIGN}};}}}
! ${CREATE_SHLIB+${RELOCATING+. = ${SHLIB_DATA_ADDR-${DATA_SEGMENT_ALIGN}};}}
! ${CREATE_PIE+${RELOCATING+. = ${SHLIB_DATA_ADDR-${DATA_SEGMENT_ALIGN}};}}
/* Exception handling */
.eh_frame ${RELOCATING-0} : ONLY_IF_RW { KEEP (*(.eh_frame)) }
.gcc_except_table ${RELOCATING-0} : ONLY_IF_RW { KEEP (*(.gcc_except_table)) *(.gcc_except_table.*) }
--- 179,191 ----
.eh_frame ${RELOCATING-0} : ONLY_IF_RO { KEEP (*(.eh_frame)) }
.gcc_except_table ${RELOCATING-0} : ONLY_IF_RO { KEEP (*(.gcc_except_table)) *(.gcc_except_table.*) }
/* Adjust the address for the data segment. We want to adjust up to
the same address within the page on the next page up. */
! ${CREATE_SHLIB-${CREATE_PIE-${RELOCATING+. = ${DATA_ADDR};}}}
! ${CREATE_SHLIB+${RELOCATING+. = ${SHLIB_DATA_ADDR};}}
! ${CREATE_PIE+${RELOCATING+. = ${SHLIB_DATA_ADDR};}}
/* Exception handling */
.eh_frame ${RELOCATING-0} : ONLY_IF_RW { KEEP (*(.eh_frame)) }
.gcc_except_table ${RELOCATING-0} : ONLY_IF_RW { KEEP (*(.gcc_except_table)) *(.gcc_except_table.*) }
*************** cat <<EOF
*** 353,362 ****
--- 235,245 ----
${CREATE_SHLIB+${SBSS2}}
${SDATA}
${OTHER_SDATA_SECTIONS}
${RELOCATING+_edata = .;}
${RELOCATING+PROVIDE (edata = .);}
+ ${RELOCATING+. = DEFINED(__bss_segment_start) ? __bss_segment_start : .;}
${RELOCATING+__bss_start = .;}
${RELOCATING+${OTHER_BSS_SYMBOLS}}
${SBSS}
${BSS_PLT+${PLT}}
.bss ${RELOCATING-0} :
*************** cat <<EOF
*** 429,435 ****
--- 312,376 ----
${STACK_ADDR+${STACK}}
${OTHER_SECTIONS}
${RELOCATING+${OTHER_END_SYMBOLS}}
${RELOCATING+${STACKNOTE}}
+ EOF
+
+ # These relocations sections are part of the read-only segment in SVR4
+ # executables, but are not mapped in BPABI executables.
+ if [ "x$COMBRELOC" = x ]; then
+ COMBRELOCCAT=cat
+ else
+ COMBRELOCCAT="cat > $COMBRELOC"
+ fi
+ eval $COMBRELOCCAT <<EOF
+ .rel.init 0 : { *(.rel.init) }
+ .rela.init 0 : { *(.rela.init) }
+ .rel.text 0 : { *(.rel.text${RELOCATING+ .rel.text.* .rel.gnu.linkonce.t.*}) }
+ .rela.text 0 : { *(.rela.text${RELOCATING+ .rela.text.* .rela.gnu.linkonce.t.*}) }
+ .rel.fini 0 : { *(.rel.fini) }
+ .rela.fini 0 : { *(.rela.fini) }
+ .rel.rodata 0 : { *(.rel.rodata${RELOCATING+ .rel.rodata.* .rel.gnu.linkonce.r.*}) }
+ .rela.rodata 0 : { *(.rela.rodata${RELOCATING+ .rela.rodata.* .rela.gnu.linkonce.r.*}) }
+ ${OTHER_READONLY_RELOC_SECTIONS}
+ .rel.data.rel.ro 0 : { *(.rel.data.rel.ro${RELOCATING+*}) }
+ .rela.data.rel.ro 0 : { *(.rel.data.rel.ro${RELOCATING+*}) }
+ .rel.data 0 : { *(.rel.data${RELOCATING+ .rel.data.* .rel.gnu.linkonce.d.*}) }
+ .rela.data 0 : { *(.rela.data${RELOCATING+ .rela.data.* .rela.gnu.linkonce.d.*}) }
+ .rel.tdata 0 : { *(.rel.tdata${RELOCATING+ .rel.tdata.* .rel.gnu.linkonce.td.*}) }
+ .rela.tdata 0 : { *(.rela.tdata${RELOCATING+ .rela.tdata.* .rela.gnu.linkonce.td.*}) }
+ .rel.tbss 0 : { *(.rel.tbss${RELOCATING+ .rel.tbss.* .rel.gnu.linkonce.tb.*}) }
+ .rela.tbss 0 : { *(.rela.tbss${RELOCATING+ .rela.tbss.* .rela.gnu.linkonce.tb.*}) }
+ .rel.ctors 0 : { *(.rel.ctors) }
+ .rela.ctors 0 : { *(.rela.ctors) }
+ .rel.dtors 0 : { *(.rel.dtors) }
+ .rela.dtors 0 : { *(.rela.dtors) }
+ ${REL_SDATA}
+ ${REL_SBSS}
+ ${REL_SDATA2}
+ ${REL_SBSS2}
+ .rel.bss 0 : { *(.rel.bss${RELOCATING+ .rel.bss.* .rel.gnu.linkonce.b.*}) }
+ .rela.bss 0 : { *(.rela.bss${RELOCATING+ .rela.bss.* .rela.gnu.linkonce.b.*}) }
+ EOF
+ if [ -n "$COMBRELOC" ]; then
+ cat <<EOF
+ .rel.dyn 0 :
+ {
+ EOF
+ sed -e '/^[ ]*[{}][ ]*$/d;/:[ ]*$/d;/\.rela\./d;s/^.*: { *\(.*\)}$/ \1/' $COMBRELOC
+ cat <<EOF
+ }
+ .rela.dyn 0 :
+ {
+ EOF
+ sed -e '/^[ ]*[{}][ ]*$/d;/:[ ]*$/d;/\.rel\./d;s/^.*: { *\(.*\)}/ \1/' $COMBRELOC
+ cat <<EOF
+ }
+ EOF
+ fi
+ cat <<EOF
+ .rel.plt 0 : { *(.rel.plt) }
+ .rela.plt 0 : { *(.rela.plt) }
+ ${OTHER_PLT_RELOC_SECTIONS}
}
EOF
Index: scripttempl/elf.sc
===================================================================
RCS file: /cvs/src/src/ld/scripttempl/elf.sc,v
retrieving revision 1.49
diff -c -5 -p -r1.49 elf.sc
*** scripttempl/elf.sc 14 Oct 2004 17:00:16 -0000 1.49
--- scripttempl/elf.sc 25 Oct 2004 19:29:19 -0000
***************
*** 16,25 ****
--- 16,28 ----
# (e.g. PPC32 .fixup, .got[12])
# OTHER_BSS_SECTIONS - other than .bss .sbss ...
# OTHER_SECTIONS - at the end
# EXECUTABLE_SYMBOLS - symbols that must be defined for an
# executable (e.g., _DYNAMIC_LINK)
+ # TEXT_START_ADDR - the first byte of the text segment, after any
+ # headers.
+ # TEXT_BASE_ADDRESS - the first byte of the text segment.
# TEXT_START_SYMBOLS - symbols that appear at the start of the
# .text section.
# DATA_START_SYMBOLS - symbols that appear at the start of the
# .data section.
# OTHER_GOT_SYMBOLS - symbols defined just before .got.