aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--bfd/ChangeLog36
-rw-r--r--bfd/config.in3
-rwxr-xr-xbfd/configure2
-rw-r--r--bfd/configure.in2
-rw-r--r--bfd/peXXigen.c1274
5 files changed, 1236 insertions, 81 deletions
diff --git a/bfd/ChangeLog b/bfd/ChangeLog
index 63865f4..4df87fe 100644
--- a/bfd/ChangeLog
+++ b/bfd/ChangeLog
@@ -1,3 +1,39 @@
+2013-12-17 Nick Clifton <nickc@redhat.com>
+
+ * peXXigen.c: Include wchar.h if available.
+ Include safe-ctype.h.
+ (HighBitSet, SetHighBit, WithoutHighBit): New macros.
+ (pe_print_resource_entries): Rename to
+ rsrc_print_resource_entries. Handle names that are not RVAs.
+ (pe_print_resource_directory): Rename to
+ rsrc_print_resource_directory.
+ (pe_print_rsrc): Rename to rsrc_print_section. Corrupt
+ computation of RVA bias.
+ (rsrc_count_entries): New function.
+ (rsrc_count_directory): New function.
+ (rsrc_parse_entry): New function.
+ (rsrc_parse_entries): New function.
+ (rsrc_parse_directory): New function.
+ (rsrc_write_string): New function.
+ (rsrc_compute_rva): New function.
+ (rsrc_write_leaf): New function.
+ (rsrc_write_entry): New function.
+ (rsrc_write_directory): New function.
+ (u16_mbtouc): New function.
+ (rsrc_cmp): New function.
+ (rsrc_print_name): New function.
+ (rsrc_resource_name): New function.
+ (rsrc_merge_string_entries): New function.
+ (rsrc_sort_entries): New function.
+ (rsrc_attach_chain): New function.
+ (rsrc_merge): New function.
+ (rsrc_process_section): New function - merges the contents of a
+ .rsrc section.
+ (_bfd_XXi_final_link_postscript): Call rsrc_process_section.
+ * configure.in (AC_CHECK_HEADERS): Add wchar.h
+ * config.in: Regenerate.
+ * configure: Regenerate.
+
2013-12-16 Andreas Schwab <schwab@suse.de>
* elf64-ppc.c (ppc64_elf_relocate_section): Add newline to error
diff --git a/bfd/config.in b/bfd/config.in
index 65fb044..3106c28 100644
--- a/bfd/config.in
+++ b/bfd/config.in
@@ -255,6 +255,9 @@
/* Define to 1 if you have the <unistd.h> header file. */
#undef HAVE_UNISTD_H
+/* Define to 1 if you have the <wchar.h> header file. */
+#undef HAVE_WCHAR_H
+
/* Define if <sys/procfs.h> has win32_pstatus_t. */
#undef HAVE_WIN32_PSTATUS_T
diff --git a/bfd/configure b/bfd/configure
index 382dd01..bef7295 100755
--- a/bfd/configure
+++ b/bfd/configure
@@ -12900,7 +12900,7 @@ $as_echo "$bfd_cv_build_exeext" >&6; }
fi
-for ac_header in alloca.h stddef.h string.h strings.h stdlib.h time.h unistd.h
+for ac_header in alloca.h stddef.h string.h strings.h stdlib.h time.h unistd.h wchar.h
do :
as_ac_Header=`$as_echo "ac_cv_header_$ac_header" | $as_tr_sh`
ac_fn_c_check_header_mongrel "$LINENO" "$ac_header" "$as_ac_Header" "$ac_includes_default"
diff --git a/bfd/configure.in b/bfd/configure.in
index 0d41867..6a029f7 100644
--- a/bfd/configure.in
+++ b/bfd/configure.in
@@ -210,7 +210,7 @@ AC_SUBST(BFD_HOSTPTR_T)
BFD_CC_FOR_BUILD
-AC_CHECK_HEADERS(alloca.h stddef.h string.h strings.h stdlib.h time.h unistd.h)
+AC_CHECK_HEADERS(alloca.h stddef.h string.h strings.h stdlib.h time.h unistd.h wchar.h)
AC_CHECK_HEADERS(fcntl.h sys/file.h sys/time.h sys/stat.h sys/resource.h)
GCC_HEADER_STDINT(bfd_stdint.h)
AC_HEADER_TIME
diff --git a/bfd/peXXigen.c b/bfd/peXXigen.c
index b720668..cddb38f 100644
--- a/bfd/peXXigen.c
+++ b/bfd/peXXigen.c
@@ -62,6 +62,9 @@
#include "libbfd.h"
#include "coff/internal.h"
#include "bfdver.h"
+#ifdef HAVE_WCHAR_H
+#include <wchar.h>
+#endif
/* NOTE: it's strange to be including an architecture specific header
in what's supposed to be general (to PE/PEI) code. However, that's
@@ -80,6 +83,7 @@
#include "coff/pe.h"
#include "libcoff.h"
#include "libpei.h"
+#include "safe-ctype.h"
#if defined COFF_WITH_pep || defined COFF_WITH_pex64
# undef AOUTSZ
@@ -87,6 +91,10 @@
# define PEAOUTHDR PEPAOUTHDR
#endif
+#define HighBitSet(val) ((val) & 0x80000000)
+#define SetHighBit(val) ((val) | 0x80000000)
+#define WithoutHighBit(val) ((val) & 0x7fffffff)
+
/* FIXME: This file has various tests of POWERPC_LE_PE. Those tests
worked when the code was in peicode.h, but no longer work now that
the code is in peigen.c. PowerPC NT is said to be dead. If
@@ -1285,9 +1293,10 @@ pe_print_idata (bfd * abfd, void * vfile)
if (!member && !member_high)
break;
- if (member_high & 0x80000000)
+ if (HighBitSet (member_high))
fprintf (file, "\t%lx%08lx\t %4lx%08lx <none>",
- member_high,member, member_high & 0x7fffffff, member);
+ member_high, member,
+ WithoutHighBit (member_high), member);
else
{
int ordinal;
@@ -1317,9 +1326,9 @@ pe_print_idata (bfd * abfd, void * vfile)
if (member == 0)
break;
- if (member & 0x80000000)
+ if (HighBitSet (member))
fprintf (file, "\t%04lx\t %4lu <none>",
- member, member & 0x7fffffff);
+ member, WithoutHighBit (member));
else
{
int ordinal;
@@ -1990,20 +1999,21 @@ pe_print_reloc (bfd * abfd, void * vfile)
return TRUE;
}
+
static bfd_byte *
-pe_print_resource_directory (FILE * , bfd *, unsigned int,
- bfd_byte *, bfd_byte *, bfd_byte *, bfd_vma);
+rsrc_print_resource_directory (FILE * , bfd *, unsigned int,
+ bfd_byte *, bfd_byte *, bfd_byte *, bfd_vma);
static bfd_byte *
-pe_print_resource_entries (FILE * file,
- bfd * abfd,
- unsigned int indent,
- bfd_boolean is_name,
- bfd_byte * data,
- bfd_byte * datastart,
- bfd_byte * dataend,
- bfd_vma rva_bias)
+rsrc_print_resource_entries (FILE * file,
+ bfd * abfd,
+ unsigned int indent,
+ bfd_boolean is_name,
+ bfd_byte * datastart,
+ bfd_byte * data,
+ bfd_byte * dataend,
+ bfd_vma rva_bias)
{
unsigned long entry, addr, size;
@@ -2015,65 +2025,81 @@ pe_print_resource_entries (FILE * file,
entry = (long) bfd_get_32 (abfd, data);
if (is_name)
{
- if (datastart + entry < dataend)
+ bfd_byte * name;
+
+ /* Note - the documenation says that this field is an RVA value
+ but windres appears to produce a section relative offset with
+ the top bit set. Support both styles for now. */
+ if (HighBitSet (entry))
+ name = datastart + WithoutHighBit (entry);
+ else
+ name = datastart + entry - rva_bias;
+
+ if (name + 2 < dataend)
{
unsigned int len;
- len = bfd_get_16 (abfd, datastart + entry);
+ len = bfd_get_16 (abfd, name);
- fprintf (file, _("name: %08lx [%d:]"), entry, len);
- if (datastart + entry - rva_bias + 2 + len < dataend)
- fprintf (file, "%.*s", len, (char *) (datastart + entry - rva_bias + 2));
+ fprintf (file, _("name: [val: %08lx len %d]: "), entry, len);
+ if (name + 2 + len * 2 < dataend)
+ {
+ /* This strange loop is to cope with multibyte characters. */
+ while (len --)
+ {
+ name += 2;
+ fprintf (file, "%.1s", name);
+ }
+ }
else
- fprintf (file, _("<corrupt>"));
+ fprintf (file, _("<corrupt string length: %#x>"), len);
}
else
- fprintf (file, _("<corrupt>"));
+ fprintf (file, _("<corrupt string offset: %#lx>"), entry);
}
else
fprintf (file, _("ID: %#08lx"), entry);
entry = (long) bfd_get_32 (abfd, data + 4);
- fprintf (file, _(", Value: %#08lx"), entry);
+ fprintf (file, _(", Value: %#08lx\n"), entry);
- if (entry & 0x80000000)
- {
- fprintf (file, _(" sub-table:\n"));
- return pe_print_resource_directory (file, abfd, indent + 1,
- datastart + (entry & 0x7fffffff),
- datastart, dataend, rva_bias);
- }
+ if (HighBitSet (entry))
+ return rsrc_print_resource_directory (file, abfd, indent + 1,
+ datastart,
+ datastart + WithoutHighBit (entry),
+ dataend, rva_bias);
if (datastart + entry + 16 >= dataend)
- {
- fprintf (file, "\n");
- return dataend + 1;
- }
+ return dataend + 1;
- fprintf (file, _(" leaf: Addr: %#08lx, Size: %#08lx, Codepage: %d\n"),
+ fprintf (file, _("%*.s Leaf: Addr: %#08lx, Size: %#08lx, Codepage: %d\n"),
+ indent, " ",
addr = (long) bfd_get_32 (abfd, datastart + entry),
size = (long) bfd_get_32 (abfd, datastart + entry + 4),
(int) bfd_get_32 (abfd, datastart + entry + 8));
-
+
/* Check that the reserved entry is 0. */
if (bfd_get_32 (abfd, datastart + entry + 12) != 0
/* And that the data address/size is valid too. */
- || (datastart + addr - rva_bias + size > dataend))
+ || (datastart + (addr - rva_bias) + size > dataend))
return dataend + 1;
- return datastart + addr - rva_bias + size;
+ return datastart + (addr - rva_bias) + size;
}
+#define max(a,b) ((a) > (b) ? (a) : (b))
+#define min(a,b) ((a) < (b) ? (a) : (b))
+
static bfd_byte *
-pe_print_resource_directory (FILE * file,
- bfd * abfd,
- unsigned int indent,
- bfd_byte * data,
- bfd_byte * datastart,
- bfd_byte * dataend,
- bfd_vma rva_bias)
+rsrc_print_resource_directory (FILE * file,
+ bfd * abfd,
+ unsigned int indent,
+ bfd_byte * datastart,
+ bfd_byte * data,
+ bfd_byte * dataend,
+ bfd_vma rva_bias)
{
unsigned int num_names, num_ids;
- bfd_byte * enddata = data;
+ bfd_byte * highest_data = data;
if (data + 16 >= dataend)
return dataend + 1;
@@ -2096,37 +2122,31 @@ pe_print_resource_directory (FILE * file,
num_ids = (int) bfd_get_16 (abfd, data + 14));
data += 16;
- if (num_names)
+ while (num_names --)
{
- while (num_names --)
- {
- bfd_byte * ndata;
- ndata = pe_print_resource_entries (file, abfd, indent + 1, TRUE,
- data, datastart, dataend, rva_bias);
- data += 8;
- if (ndata > enddata)
- enddata = ndata;
- if (ndata >= dataend)
- break;
- }
+ bfd_byte * entry_end;
+
+ entry_end = rsrc_print_resource_entries (file, abfd, indent + 1, TRUE,
+ datastart, data, dataend, rva_bias);
+ data += 8;
+ highest_data = max (highest_data, entry_end);
+ if (entry_end >= dataend)
+ return entry_end;
}
- if (num_ids)
+ while (num_ids --)
{
- while (num_ids --)
- {
- bfd_byte * ndata;
- ndata = pe_print_resource_entries (file, abfd, indent + 1, FALSE,
- data, datastart, dataend, rva_bias);
- data += 8;
- if (ndata > enddata)
- enddata = ndata;
- if (ndata >= dataend)
- break;
- }
+ bfd_byte * entry_end;
+
+ entry_end = rsrc_print_resource_entries (file, abfd, indent + 1, FALSE,
+ datastart, data, dataend, rva_bias);
+ data += 8;
+ highest_data = max (highest_data, entry_end);
+ if (entry_end >= dataend)
+ return entry_end;
}
- return enddata > data ? enddata : data;
+ return max (highest_data, data);
}
/* Display the contents of a .rsrc section. We do not try to
@@ -2134,7 +2154,7 @@ pe_print_resource_directory (FILE * file,
the tables in a human readable format. */
static bfd_boolean
-pe_print_rsrc (bfd * abfd, void * vfile)
+rsrc_print_section (bfd * abfd, void * vfile)
{
bfd_vma rva_bias;
pe_data_type * pe;
@@ -2147,20 +2167,20 @@ pe_print_rsrc (bfd * abfd, void * vfile)
pe = pe_data (abfd);
- if (pe == NULL)
+ if (pe == NULL)
return TRUE;
- section = bfd_get_section_by_name (abfd, ".rsrc");
- if (section == NULL)
+ section = bfd_get_section_by_name (abfd, ".rsrc");
+ if (section == NULL)
return TRUE;
rva_bias = section->vma - pe->pe_opthdr.ImageBase;
-
+
datasize = section->size;
if (datasize == 0)
return TRUE;
- if (! bfd_malloc_and_get_section (abfd, section, &data))
+ if (! bfd_malloc_and_get_section (abfd, section, & data))
{
if (data != NULL)
free (data);
@@ -2174,7 +2194,9 @@ pe_print_rsrc (bfd * abfd, void * vfile)
while (data < dataend)
{
- data = pe_print_resource_directory (file, abfd, 0, data, data, dataend, rva_bias);
+ bfd_byte * p = data;
+
+ data = rsrc_print_resource_directory (file, abfd, 0, data, data, dataend, rva_bias);
if (data == dataend + 1)
fprintf (file, _("Corrupt .rsrc section detected!\n"));
@@ -2182,8 +2204,9 @@ pe_print_rsrc (bfd * abfd, void * vfile)
{
/* Align data before continuing. */
int align = (1 << section->alignment_power) - 1;
+
data = (bfd_byte *) (((long) (data + align)) & ~ align);
- rva_bias += data - datastart;
+ rva_bias += data - p;
/* For reasons that are unclear .rsrc sections are sometimes created
aligned to a 1^3 boundary even when their alignment is set at
@@ -2196,6 +2219,7 @@ pe_print_rsrc (bfd * abfd, void * vfile)
}
}
+ free (datastart);
return TRUE;
}
@@ -2373,7 +2397,7 @@ _bfd_XX_print_private_bfd_data_common (bfd * abfd, void * vfile)
pe_print_pdata (abfd, vfile);
pe_print_reloc (abfd, vfile);
- pe_print_rsrc (abfd, vfile);
+ rsrc_print_section (abfd, vfile);
return TRUE;
}
@@ -2479,6 +2503,1096 @@ sort_x64_pdata (const void *l, const void *r)
return 0;
}
#endif
+
+/* Functions to process a .rsrc section. */
+
+static unsigned int sizeof_leaves;
+static unsigned int sizeof_strings;
+static unsigned int sizeof_tables_and_entries;
+
+static bfd_byte *
+rsrc_count_directory (bfd *, bfd_byte *, bfd_byte *, bfd_byte *, bfd_vma);
+
+static bfd_byte *
+rsrc_count_entries (bfd * abfd,
+ bfd_boolean is_name,
+ bfd_byte * datastart,
+ bfd_byte * data,
+ bfd_byte * dataend,
+ bfd_vma rva_bias)
+{
+ unsigned long entry, addr, size;
+
+ if (data + 8 >= dataend)
+ return dataend + 1;
+
+ if (is_name)
+ {
+ bfd_byte * name;
+
+ entry = (long) bfd_get_32 (abfd, data);
+
+ if (HighBitSet (entry))
+ name = datastart + WithoutHighBit (entry);
+ else
+ name = datastart + entry - rva_bias;
+
+ if (name + 2 >= dataend)
+ return dataend + 1;
+
+ unsigned int len = bfd_get_16 (abfd, name);
+ if (len == 0 || len > 256)
+ return dataend + 1;
+
+ sizeof_strings += (len + 1) * 2;
+ }
+
+ entry = (long) bfd_get_32 (abfd, data + 4);
+
+ if (HighBitSet (entry))
+ return rsrc_count_directory (abfd,
+ datastart,
+ datastart + WithoutHighBit (entry),
+ dataend, rva_bias);
+
+ if (datastart + entry + 16 >= dataend)
+ return dataend + 1;
+
+ addr = (long) bfd_get_32 (abfd, datastart + entry);
+ size = (long) bfd_get_32 (abfd, datastart + entry + 4);
+
+ sizeof_leaves += 16;
+
+ return datastart + addr - rva_bias + size;
+}
+
+static bfd_byte *
+rsrc_count_directory (bfd * abfd,
+ bfd_byte * datastart,
+ bfd_byte * data,
+ bfd_byte * dataend,
+ bfd_vma rva_bias)
+{
+ unsigned int num_entries, num_ids;
+ bfd_byte * highest_data = data;
+
+ if (data + 16 >= dataend)
+ return dataend + 1;
+
+ num_entries = (int) bfd_get_16 (abfd, data + 12);
+ num_ids = (int) bfd_get_16 (abfd, data + 14);
+
+ num_entries += num_ids;
+
+ data += 16;
+ sizeof_tables_and_entries += 16;
+
+ while (num_entries --)
+ {
+ bfd_byte * entry_end;
+
+ entry_end = rsrc_count_entries (abfd, num_entries >= num_ids,
+ datastart, data, dataend, rva_bias);
+ data += 8;
+ sizeof_tables_and_entries += 8;
+ highest_data = max (highest_data, entry_end);
+ if (entry_end >= dataend)
+ break;
+ }
+
+ return max (highest_data, data);
+}
+
+typedef struct rsrc_dir_chain
+{
+ unsigned int num_entries;
+ struct rsrc_entry * first_entry;
+ struct rsrc_entry * last_entry;
+} rsrc_dir_chain;
+
+typedef struct rsrc_directory
+{
+ unsigned int characteristics;
+ unsigned int time;
+ unsigned int major;
+ unsigned int minor;
+
+ rsrc_dir_chain names;
+ rsrc_dir_chain ids;
+
+ struct rsrc_entry * entry;
+} rsrc_directory;
+
+typedef struct rsrc_string
+{
+ unsigned int len;
+ bfd_byte * string;
+} rsrc_string;
+
+typedef struct rsrc_leaf
+{
+ unsigned int size;
+ unsigned int codepage;
+ bfd_byte * data;
+} rsrc_leaf;
+
+typedef struct rsrc_entry
+{
+ bfd_boolean is_name;
+ union
+ {
+ unsigned int id;
+ struct rsrc_string name;
+ } name_id;
+
+ bfd_boolean is_dir;
+ union
+ {
+ struct rsrc_directory * directory;
+ struct rsrc_leaf * leaf;
+ } value;
+
+ struct rsrc_entry * next_entry;
+ struct rsrc_directory * parent;
+} rsrc_entry;
+
+static bfd_byte *
+rsrc_parse_directory (bfd *, rsrc_directory *, bfd_byte *,
+ bfd_byte *, bfd_byte *, bfd_vma, rsrc_entry *);
+
+static bfd_byte *
+rsrc_parse_entry (bfd * abfd,
+ bfd_boolean is_name,
+ rsrc_entry * entry,
+ bfd_byte * datastart,
+ bfd_byte * data,
+ bfd_byte * dataend,
+ bfd_vma rva_bias,
+ rsrc_directory * parent)
+{
+ unsigned long val, addr, size;
+
+ val = bfd_get_32 (abfd, data);
+
+ entry->parent = parent;
+ entry->is_name = is_name;
+
+ if (is_name)
+ {
+ /* FIXME: Add range checking ? */
+ if (HighBitSet (val))
+ {
+ val = WithoutHighBit (val);
+
+ entry->name_id.name.len = bfd_get_16 (abfd, datastart + val);
+ entry->name_id.name.string = datastart + val + 2;
+ }
+ else
+ {
+ entry->name_id.name.len = bfd_get_16 (abfd, datastart + val - rva_bias);
+ entry->name_id.name.string = datastart + val - rva_bias + 2;
+ }
+ }
+ else
+ entry->name_id.id = val;
+
+ val = bfd_get_32 (abfd, data + 4);
+
+ if (HighBitSet (val))
+ {
+ entry->is_dir = TRUE;
+ entry->value.directory = bfd_malloc (sizeof * entry->value.directory);
+ if (entry->value.directory == NULL)
+ return dataend;
+
+ return rsrc_parse_directory (abfd, entry->value.directory,
+ datastart,
+ datastart + WithoutHighBit (val),
+ dataend, rva_bias, entry);
+ }
+
+ entry->is_dir = FALSE;
+ entry->value.leaf = bfd_malloc (sizeof * entry->value.leaf);
+ if (entry->value.leaf == NULL)
+ return dataend;
+
+ addr = bfd_get_32 (abfd, datastart + val);
+ size = entry->value.leaf->size = bfd_get_32 (abfd, datastart + val + 4);
+ entry->value.leaf->codepage = bfd_get_32 (abfd, datastart + val + 8);
+
+ entry->value.leaf->data = bfd_malloc (size);
+ if (entry->value.leaf->data == NULL)
+ return dataend;
+
+ memcpy (entry->value.leaf->data, datastart + addr - rva_bias, size);
+ return datastart + (addr - rva_bias) + size;
+}
+
+static bfd_byte *
+rsrc_parse_entries (bfd * abfd,
+ rsrc_dir_chain * chain,
+ bfd_boolean is_name,
+ bfd_byte * highest_data,
+ bfd_byte * datastart,
+ bfd_byte * data,
+ bfd_byte * dataend,
+ bfd_vma rva_bias,
+ rsrc_directory * parent)
+{
+ rsrc_entry * entry;
+
+ if (chain->num_entries == 0)
+ {
+ chain->first_entry = chain->last_entry = NULL;
+ return highest_data;
+ }
+
+ entry = bfd_malloc (sizeof * entry);
+ if (entry == NULL)
+ return dataend;
+
+ chain->first_entry = entry;
+
+ unsigned int i;
+ for (i = chain->num_entries; i--;)
+ {
+ bfd_byte * entry_end;
+
+ entry_end = rsrc_parse_entry (abfd, is_name, entry, datastart,
+ data, dataend, rva_bias, parent);
+ data += 8;
+ highest_data = max (entry_end, highest_data);
+ if (entry_end > dataend)
+ return dataend;
+
+ if (i)
+ {
+ entry->next_entry = bfd_malloc (sizeof * entry);
+ entry = entry->next_entry;
+ if (entry == NULL)
+ return dataend;
+ }
+ else
+ entry->next_entry = NULL;
+ }
+
+ chain->last_entry = entry;
+
+ return highest_data;
+}
+
+static bfd_byte *
+rsrc_parse_directory (bfd * abfd,
+ rsrc_directory * table,
+ bfd_byte * datastart,
+ bfd_byte * data,
+ bfd_byte * dataend,
+ bfd_vma rva_bias,
+ rsrc_entry * entry)
+{
+ bfd_byte * highest_data = data;
+
+ if (table == NULL)
+ return dataend;
+
+ table->characteristics = bfd_get_32 (abfd, data);
+ table->time = bfd_get_32 (abfd, data + 4);
+ table->major = bfd_get_16 (abfd, data + 8);
+ table->minor = bfd_get_16 (abfd, data + 10);
+ table->names.num_entries = bfd_get_16 (abfd, data + 12);
+ table->ids.num_entries = bfd_get_16 (abfd, data + 14);
+ table->entry = entry;
+
+ data += 16;
+
+ highest_data = rsrc_parse_entries (abfd, & table->names, TRUE, data,
+ datastart, data, dataend, rva_bias, table);
+ data += table->names.num_entries * 8;
+
+ highest_data = rsrc_parse_entries (abfd, & table->ids, FALSE, highest_data,
+ datastart, data, dataend, rva_bias, table);
+ data += table->ids.num_entries * 8;
+
+ return max (highest_data, data);
+}
+
+typedef struct rsrc_write_data
+{
+ bfd * abfd;
+ bfd_byte * datastart;
+ bfd_byte * next_table;
+ bfd_byte * next_leaf;
+ bfd_byte * next_string;
+ bfd_byte * next_data;
+ bfd_vma rva_bias;
+} rsrc_write_data;
+
+static void
+rsrc_write_string (rsrc_write_data * data,
+ rsrc_string * string)
+{
+ bfd_put_16 (data->abfd, string->len, data->next_string);
+ memcpy (data->next_string + 2, string->string, string->len * 2);
+ data->next_string += (string->len + 1) * 2;
+}
+
+static inline unsigned int
+rsrc_compute_rva (rsrc_write_data * data,
+ bfd_byte * addr)
+{
+ return (addr - data->datastart) + data->rva_bias;
+}
+
+static void
+rsrc_write_leaf (rsrc_write_data * data,
+ rsrc_leaf * leaf)
+{
+ bfd_put_32 (data->abfd, rsrc_compute_rva (data, data->next_data), data->next_leaf);
+ bfd_put_32 (data->abfd, leaf->size, data->next_leaf + 4);
+ bfd_put_32 (data->abfd, leaf->codepage, data->next_leaf + 8);
+ bfd_put_32 (data->abfd, 0 /*reserved*/, data->next_leaf + 12);
+ data->next_leaf += 16;
+
+ memcpy (data->next_data, leaf->data, leaf->size);
+ data->next_data += leaf->size;
+}
+
+static void rsrc_write_directory (rsrc_write_data *, rsrc_directory *);
+
+static void
+rsrc_write_entry (rsrc_write_data * data,
+ bfd_byte * where,
+ rsrc_entry * entry)
+{
+ if (entry->is_name)
+ {
+ bfd_put_32 (data->abfd,
+ SetHighBit (data->next_string - data->datastart),
+ where);
+ rsrc_write_string (data, & entry->name_id.name);
+ }
+ else
+ bfd_put_32 (data->abfd, entry->name_id.id, where);
+
+ if (entry->is_dir)
+ {
+ bfd_put_32 (data->abfd,
+ SetHighBit (data->next_table - data->datastart),
+ where + 4);
+ rsrc_write_directory (data, entry->value.directory);
+ }
+ else
+ {
+ bfd_put_32 (data->abfd, data->next_leaf - data->datastart, where + 4);
+ rsrc_write_leaf (data, entry->value.leaf);
+ }
+}
+
+static void
+rsrc_write_directory (rsrc_write_data * data,
+ rsrc_directory * dir)
+{
+ rsrc_entry * entry;
+ unsigned int i;
+
+ bfd_put_32 (data->abfd, dir->characteristics, data->next_table);
+ bfd_put_32 (data->abfd, 0 /*dir->time*/, data->next_table + 4);
+ bfd_put_16 (data->abfd, dir->major, data->next_table + 8);
+ bfd_put_16 (data->abfd, dir->minor, data->next_table + 10);
+ bfd_put_16 (data->abfd, dir->names.num_entries, data->next_table + 12);
+ bfd_put_16 (data->abfd, dir->ids.num_entries, data->next_table + 14);
+
+ /* Compute where the entries and the next table will be placed. */
+ bfd_byte * next_entry = data->next_table + 16;
+ data->next_table = next_entry + (dir->names.num_entries * 8) + (dir->ids.num_entries * 8);
+ bfd_byte * nt = data->next_table;
+
+ /* Write the entries. */
+ for (i = dir->names.num_entries, entry = dir->names.first_entry;
+ i > 0 && entry != NULL;
+ i--, entry = entry->next_entry)
+ {
+ rsrc_write_entry (data, next_entry, entry);
+ next_entry += 8;
+ }
+ BFD_ASSERT (i == 0);
+ BFD_ASSERT (entry == NULL);
+
+ for (i = dir->ids.num_entries, entry = dir->ids.first_entry;
+ i > 0 && entry != NULL;
+ i--, entry = entry->next_entry)
+ {
+ rsrc_write_entry (data, next_entry, entry);
+ next_entry += 8;
+ }
+ BFD_ASSERT (i == 0);
+ BFD_ASSERT (entry == NULL);
+ BFD_ASSERT (nt == next_entry);
+}
+
+#ifdef HAVE_WCHAR_H
+/* Return the length (number of units) of the first character in S,
+ putting its 'ucs4_t' representation in *PUC. */
+
+static unsigned int
+u16_mbtouc (wchar_t * puc, const unsigned short * s, unsigned int n)
+{
+ unsigned short c = * s;
+
+ if (c < 0xd800 || c >= 0xe000)
+ {
+ *puc = c;
+ return 1;
+ }
+
+ if (c < 0xdc00)
+ {
+ if (n >= 2)
+ {
+ if (s[1] >= 0xdc00 && s[1] < 0xe000)
+ {
+ *puc = 0x10000 + ((c - 0xd800) << 10) + (s[1] - 0xdc00);
+ return 2;
+ }
+ }
+ else
+ {
+ /* Incomplete multibyte character. */
+ *puc = 0xfffd;
+ return n;
+ }
+ }
+
+ /* Invalid multibyte character. */
+ *puc = 0xfffd;
+ return 1;
+}
+#endif /* HAVE_WCHAR_H */
+
+/* Perform a comparison of two entries. */
+static signed int
+rsrc_cmp (bfd_boolean is_name, rsrc_entry * a, rsrc_entry * b)
+{
+ if (! is_name)
+ return a->name_id.id - b->name_id.id;
+
+ /* We have to perform a case insenstive, unicode string comparison... */
+ int res;
+
+#ifdef __CYGWIN__
+ /* Under Cygwin unicode == UTF-16 == wchar_t.
+ FIXME: The same is true for MingGW - we should test for that too. */
+ res = wcsncasecmp ((const wchar_t *) astring + 2, (const wchar_t *) bstring + 2, min (alen, blen));
+#elif defined HAVE_WCHAR_H
+ unsigned int i;
+ bfd_byte * astring = a->name_id.name.string;
+ unsigned int alen = a->name_id.name.len;
+ bfd_byte * bstring = b->name_id.name.string;
+ unsigned int blen = b->name_id.name.len;
+
+ res = 0;
+ for (i = min (alen, blen); i--; astring += 2, bstring += 2)
+ {
+ wchar_t awc;
+ wchar_t bwc;
+
+ /* Convert UTF-16 unicode characters into wchar_t characters so
+ that we can then perform a case insensitive comparison. */
+ int Alen = u16_mbtouc (& awc, (const unsigned short *) astring, 2);
+ int Blen = u16_mbtouc (& bwc, (const unsigned short *) bstring, 2);
+
+ if (Alen != Blen)
+ return Alen - Blen;
+ res = wcsncasecmp (& awc, & bwc, 1);
+ if (res)
+ break;
+ }
+#else
+ res = memcmp (astring + 2, bstring + 2, min (alen, blen) * 2);
+#endif
+
+ if (res == 0)
+ res = alen - blen;
+
+ return res;
+}
+
+static void
+rsrc_print_name (char * buffer, rsrc_string string)
+{
+ unsigned int i;
+ bfd_byte * name = string.string;
+
+ for (i = string.len; i--; name += 2)
+ sprintf (buffer + strlen (buffer), "%.1s", name);
+}
+
+static const char *
+rsrc_resource_name (rsrc_entry * entry, rsrc_directory * dir)
+{
+ static char buffer [256];
+ bfd_boolean is_string = FALSE;
+
+ buffer[0] = 0;
+
+ if (dir != NULL && dir->entry != NULL && dir->entry->parent != NULL && dir->entry->parent->entry != NULL)
+ {
+ strcpy (buffer, "type: ");
+ if (dir->entry->parent->entry->is_name)
+ rsrc_print_name (buffer + strlen (buffer), dir->entry->parent->entry->name_id.name);
+ else
+ {
+ unsigned int id = dir->entry->parent->entry->name_id.id;
+
+ sprintf (buffer + strlen (buffer), "%x", id);
+ switch (id)
+ {
+ case 1: strcat (buffer, " (CURSOR)"); break;
+ case 2: strcat (buffer, " (BITMAP)"); break;
+ case 3: strcat (buffer, " (ICON)"); break;
+ case 4: strcat (buffer, " (MENU)"); break;
+ case 5: strcat (buffer, " (DIALOG)"); break;
+ case 6: strcat (buffer, " (STRING)"); is_string = TRUE; break;
+ case 7: strcat (buffer, " (FONTDIR)"); break;
+ case 8: strcat (buffer, " (FONT)"); break;
+ case 9: strcat (buffer, " (ACCELERATOR)"); break;
+ case 10: strcat (buffer, " (RCDATA)"); break;
+ case 11: strcat (buffer, " (MESSAGETABLE)"); break;
+ case 12: strcat (buffer, " (GROUP_CURSOR)"); break;
+ case 14: strcat (buffer, " (GROUP_ICON)"); break;
+ case 16: strcat (buffer, " (VERSION)"); break;
+ case 17: strcat (buffer, " (DLGINCLUDE)"); break;
+ case 19: strcat (buffer, " (PLUGPLAY)"); break;
+ case 20: strcat (buffer, " (VXD)"); break;
+ case 21: strcat (buffer, " (ANICURSOR)"); break;
+ case 22: strcat (buffer, " (ANIICON)"); break;
+ case 23: strcat (buffer, " (HTML)"); break;
+ case 24: strcat (buffer, " (MANIFEST)"); break;
+ case 240: strcat (buffer, " (DLGINIT)"); break;
+ case 241: strcat (buffer, " (TOOLBAR)"); break;
+ }
+ }
+ }
+
+ if (dir != NULL && dir->entry != NULL)
+ {
+ strcat (buffer, " name: ");
+ if (dir->entry->is_name)
+ rsrc_print_name (buffer + strlen (buffer), dir->entry->name_id.name);
+ else
+ {
+ unsigned int id = dir->entry->name_id.id;
+
+ sprintf (buffer + strlen (buffer), "%x", id);
+
+ if (is_string)
+ sprintf (buffer + strlen (buffer), " (resource id range: %d - %d)",
+ (id - 1) << 4, (id << 4) - 1);
+ }
+ }
+
+ if (entry != NULL)
+ {
+ strcat (buffer, " lang: ");
+
+ if (entry->is_name)
+ rsrc_print_name (buffer + strlen (buffer), entry->name_id.name);
+ else
+ sprintf (buffer + strlen (buffer), "%x", entry->name_id.id);
+ }
+
+ return buffer;
+}
+
+/* *sigh* Windows resource strings are special. Only the top 28-bits of
+ their ID is stored in the NAME entry. The bottom four bits are used as
+ an index into unicode string table that makes up the data of the leaf.
+ So identical type-name-lang string resources may not actually be
+ identical at all.
+
+ This function is called when we have detected two string resources with
+ match top-28-bit IDs. We have to scan the string tables inside the leaves
+ and discover if there are any real collisions. If there are then we report
+ them and return FALSE. Otherwise we copy any strings from B into A and then
+ return TRUE. */
+
+static bfd_boolean
+rsrc_merge_string_entries (rsrc_entry * a ATTRIBUTE_UNUSED,
+ rsrc_entry * b ATTRIBUTE_UNUSED)
+{
+ unsigned int copy_needed = 0;
+ unsigned int i;
+
+ /* Step one: Find out what we have to do. */
+ BFD_ASSERT (! a->is_dir);
+ bfd_byte * astring = a->value.leaf->data;
+
+ BFD_ASSERT (! b->is_dir);
+ bfd_byte * bstring = b->value.leaf->data;
+
+ for (i = 0; i < 16; i++)
+ {
+ unsigned int alen = astring[0] + (astring[1] << 8);
+ unsigned int blen = bstring[0] + (bstring[1] << 8);
+
+ if (alen == 0)
+ {
+ copy_needed += blen * 2;
+ }
+ else if (blen == 0)
+ ;
+ else if (alen != blen)
+ /* FIXME: Should we continue the loop in order to report other duplicates ? */
+ break;
+ /* alen == blen != 0. We might have two identical strings. If so we
+ can ignore the second one. There is no need for wchar_t vs UTF-16
+ theatrics here - we are only interested in (case sensitive) equality. */
+ else if (memcmp (astring + 2, bstring + 2, alen * 2) != 0)
+ break;
+
+ astring += (alen + 1) * 2;
+ bstring += (blen + 1) * 2;
+ }
+
+ if (i != 16)
+ {
+ if (a->parent != NULL
+ && a->parent->entry != NULL
+ && a->parent->entry->is_name == FALSE)
+ _bfd_error_handler (_(".rsrc merge failure: duplicate string resource: %d"),
+ ((a->parent->entry->name_id.id - 1) << 4) + i);
+ return FALSE;
+ }
+
+ if (copy_needed == 0)
+ return TRUE;
+
+ /* If we reach here then A and B must both have non-colliding strings.
+ (We never get string resources with fully empty string tables).
+ We need to allocate an extra COPY_NEEDED bytes in A and then bring
+ in B's strings. */
+ bfd_byte * new_data = bfd_malloc (a->value.leaf->size + copy_needed);
+ if (new_data == NULL)
+ return FALSE;
+
+ bfd_byte * nstring = new_data;
+ astring = a->value.leaf->data;
+ bstring = b->value.leaf->data;
+
+ for (i = 0; i < 16; i++)
+ {
+ unsigned int alen = astring[0] + (astring[1] << 8);
+ unsigned int blen = bstring[0] + (bstring[1] << 8);
+
+ if (alen != 0)
+ {
+ memcpy (nstring, astring, (alen + 1) * 2);
+ nstring += (alen + 1) * 2;
+ }
+ else if (blen != 0)
+ {
+ memcpy (nstring, bstring, (blen + 1) * 2);
+ nstring += (blen + 1) * 2;
+ }
+ else
+ {
+ * nstring++ = 0;
+ * nstring++ = 0;
+ }
+
+ astring += (alen + 1) * 2;
+ bstring += (blen + 1) * 2;
+ }
+
+ BFD_ASSERT (nstring - new_data == (signed) (a->value.leaf->size + copy_needed));
+
+ free (a->value.leaf->data);
+ a->value.leaf->data = new_data;
+ a->value.leaf->size += copy_needed;
+
+ return TRUE;
+}
+
+static void rsrc_merge (rsrc_entry *, rsrc_entry *);
+
+/* Sort the entries in given part of the directory.
+ We use an old fashioned bubble sort because we are dealing
+ with lists and we want to handle matches specially. */
+
+static void
+rsrc_sort_entries (rsrc_dir_chain * chain,
+ bfd_boolean is_name,
+ rsrc_directory * dir)
+{
+ rsrc_entry * entry;
+ rsrc_entry * next;
+ rsrc_entry ** points_to_entry;
+ bfd_boolean swapped;
+
+ if (chain->num_entries < 2)
+ return;
+
+ do
+ {
+ swapped = FALSE;
+ points_to_entry = & chain->first_entry;
+ entry = * points_to_entry;
+ next = entry->next_entry;
+
+ do
+ {
+ signed int cmp = rsrc_cmp (is_name, entry, next);
+
+ if (cmp > 0)
+ {
+ entry->next_entry = next->next_entry;
+ next->next_entry = entry;
+ * points_to_entry = next;
+ points_to_entry = & next->next_entry;
+ next = entry->next_entry;
+ swapped = TRUE;
+ }
+ else if (cmp == 0)
+ {
+ if (entry->is_dir && next->is_dir)
+ {
+ /* When we encounter identical directory entries we have to
+ merge them together. The exception to this rule is for
+ resource manifests - there can only be one of these,
+ even if they differ in language. Zero-language manifests
+ are assumed to be default manifests (provided by the
+ cygwin build system) and these can be silently dropped,
+ unless that would reduce the number of manifests to zero.
+ There should only ever be one non-zero lang manifest -
+ if there are more it is an error. A non-zero lang
+ manifest takes precedence over a default manifest. */
+ if (entry->is_name == FALSE
+ && entry->name_id.id == 1
+ && dir != NULL
+ && dir->entry != NULL
+ && dir->entry->is_name == FALSE
+ && dir->entry->name_id.id == 0x18)
+ {
+ if (next->value.directory->names.num_entries == 0
+ && next->value.directory->ids.num_entries == 1
+ && next->value.directory->ids.first_entry->is_name == FALSE
+ && next->value.directory->ids.first_entry->name_id.id == 0)
+ /* Fall through so that NEXT is dropped. */
+ ;
+ else if (entry->value.directory->names.num_entries == 0
+ && entry->value.directory->ids.num_entries == 1
+ && entry->value.directory->ids.first_entry->is_name == FALSE
+ && entry->value.directory->ids.first_entry->name_id.id == 0)
+ {
+ /* Swap ENTRY and NEXT. Then fall through so that the old ENTRY is dropped. */
+ entry->next_entry = next->next_entry;
+ next->next_entry = entry;
+ * points_to_entry = next;
+ points_to_entry = & next->next_entry;
+ next = entry->next_entry;
+ swapped = TRUE;
+ }
+ else
+ {
+ _bfd_error_handler (_(".rsrc merge failure: multiple non-default manifests"));
+ bfd_set_error (bfd_error_file_truncated);
+ return;
+ }
+
+ /* Unhook NEXT from the chain. */
+ /* FIXME: memory loss here. */
+ entry->next_entry = next->next_entry;
+ chain->num_entries --;
+ if (chain->num_entries < 2)
+ return;
+ next = next->next_entry;
+ }
+ else
+ rsrc_merge (entry, next);
+ }
+ else if (entry->is_dir != next->is_dir)
+ {
+ _bfd_error_handler (_(".rsrc merge failure: a directory matches a leaf"));
+ bfd_set_error (bfd_error_file_truncated);
+ return;
+ }
+ else
+ {
+ /* Otherwise with identical leaves we issue an error
+ message - because there should never be duplicates.
+ The exception is Type 18/Name 1/Lang 0 which is the
+ defaul manifest - this can just be dropped. */
+ if (entry->is_name == FALSE
+ && entry->name_id.id == 0
+ && dir != NULL
+ && dir->entry != NULL
+ && dir->entry->is_name == FALSE
+ && dir->entry->name_id.id == 1
+ && dir->entry->parent != NULL
+ && dir->entry->parent->entry != NULL
+ && dir->entry->parent->entry->is_name == FALSE
+ && dir->entry->parent->entry->name_id.id == 0x18 /* RT_MANIFEST */)
+ ;
+ else if (dir != NULL
+ && dir->entry != NULL
+ && dir->entry->parent != NULL
+ && dir->entry->parent->entry != NULL
+ && dir->entry->parent->entry->is_name == FALSE
+ && dir->entry->parent->entry->name_id.id == 0x6 /* RT_STRING */)
+ {
+ /* Strings need special handling. */
+ if (! rsrc_merge_string_entries (entry, next))
+ {
+ /* _bfd_error_handler should have been called inside merge_strings. */
+ bfd_set_error (bfd_error_file_truncated);
+ return;
+ }
+ }
+ else
+ {
+ if (dir == NULL
+ || dir->entry == NULL
+ || dir->entry->parent == NULL
+ || dir->entry->parent->entry == NULL)
+ _bfd_error_handler (_(".rsrc merge failure: duplicate leaf"));
+ else
+ _bfd_error_handler (_(".rsrc merge failure: duplicate leaf: %s"),
+ rsrc_resource_name (entry, dir));
+ bfd_set_error (bfd_error_file_truncated);
+ return;
+ }
+ }
+
+ /* Unhook NEXT from the chain. */
+ entry->next_entry = next->next_entry;
+ chain->num_entries --;
+ if (chain->num_entries < 2)
+ return;
+ next = next->next_entry;
+ }
+ else
+ {
+ points_to_entry = & entry->next_entry;
+ entry = next;
+ next = next->next_entry;
+ }
+ }
+ while (next);
+
+ chain->last_entry = entry;
+ }
+ while (swapped);
+}
+
+/* Attach B's chain onto A. */
+static void
+rsrc_attach_chain (struct rsrc_dir_chain * achain, struct rsrc_dir_chain * bchain)
+{
+ if (bchain->num_entries == 0)
+ return;
+
+ achain->num_entries += bchain->num_entries;
+
+ if (achain->first_entry == NULL)
+ {
+ achain->first_entry = bchain->first_entry;
+ achain->last_entry = bchain->last_entry;
+ }
+ else
+ {
+ achain->last_entry->next_entry = bchain->first_entry;
+ achain->last_entry = bchain->last_entry;
+ }
+
+ bchain->num_entries = 0;
+ bchain->first_entry = bchain->last_entry = NULL;
+}
+
+static void
+rsrc_merge (struct rsrc_entry * a, struct rsrc_entry * b)
+{
+ BFD_ASSERT (a->is_dir);
+ BFD_ASSERT (b->is_dir);
+
+ rsrc_directory * adir = a->value.directory;
+ rsrc_directory * bdir = b->value.directory;
+
+ if (adir->characteristics != bdir->characteristics)
+ {
+ _bfd_error_handler (_(".rsrc merge failure: dirs with differing characteristics\n"));
+ bfd_set_error (bfd_error_file_truncated);
+ return;
+ }
+
+ if (adir->major != bdir->major || adir->minor != bdir->minor)
+ {
+ _bfd_error_handler (_(".rsrc merge failure: differing directory versions\n"));
+ bfd_set_error (bfd_error_file_truncated);
+ return;
+ }
+
+ /* Attach B's name chain to A. */
+ rsrc_attach_chain (& adir->names, & bdir->names);
+
+ /* Attach B's ID chain to A. */
+ rsrc_attach_chain (& adir->ids, & bdir->ids);
+
+ /* Now sort A's entries. */
+ rsrc_sort_entries (& adir->names, TRUE, adir);
+ rsrc_sort_entries (& adir->ids, FALSE, adir);
+}
+
+/* Check the .rsrc section. If it contains multiple concatenated
+ resources then we must merge them properly. Otherwise Windows
+ will ignore all but the first set. */
+
+static void
+rsrc_process_section (bfd * abfd,
+ struct coff_final_link_info * pfinfo)
+{
+ rsrc_directory new_table;
+ bfd_size_type size;
+ asection * sec;
+
+ new_table.names.num_entries = 0;
+ new_table.ids.num_entries = 0;
+
+ sec = bfd_get_section_by_name (abfd, ".rsrc");
+ if (sec == NULL || (size = sec->rawsize) == 0)
+ return;
+
+ pe_data_type * pe = pe_data (abfd);
+ if (pe == NULL)
+ return;
+
+ bfd_vma rva_bias;
+ rva_bias = sec->vma - pe->pe_opthdr.ImageBase;
+
+ bfd_byte * data = bfd_malloc (size);
+ if (data == NULL)
+ return;
+
+ bfd_byte * datastart = data;
+
+ if (! bfd_get_section_contents (abfd, sec, data, 0, size))
+ goto end;
+
+ /* Step one: Walk the section, computing the size of the tables,
+ leaves and data and decide if we need to do anything. */
+ bfd_byte * dataend = data + size;
+ unsigned int num_resource_sets = 0;
+ sizeof_leaves = sizeof_strings = sizeof_tables_and_entries = 0;
+
+ while (data < dataend)
+ {
+ bfd_byte * p = data;
+
+ data = rsrc_count_directory (abfd, data, data, dataend, rva_bias);
+ if (data > dataend)
+ {
+ /* Corrupted .rsrc section - cannot merge. */
+ _bfd_error_handler (_("%s: .rsrc merge failure: corrupt .rsrc section"),
+ bfd_get_filename (abfd));
+ bfd_set_error (bfd_error_file_truncated);
+ goto end;
+ }
+
+ /* Align the data pointer - we assume 1^2 alignment. */
+ data = (bfd_byte *) (((long) (data + 3)) & ~ 3);
+ rva_bias += data - p;
+
+ if (data == (dataend - 4))
+ data = dataend;
+
+ ++ num_resource_sets;
+ }
+
+ if (num_resource_sets < 2)
+ /* No merging necessary. */
+ goto end;
+
+ /* Step two: Walk the data again, building trees of the resources. */
+ data = datastart;
+ rva_bias = sec->vma - pe->pe_opthdr.ImageBase;
+
+ rsrc_directory * type_tables = bfd_malloc (num_resource_sets * sizeof * type_tables);
+ if (type_tables == NULL)
+ goto end;
+
+ unsigned int index = 0;
+ while (data < dataend)
+ {
+ bfd_byte * p = data;
+
+ data = rsrc_parse_directory (abfd, type_tables + index, data, data, dataend,
+ rva_bias, NULL);
+ data = (bfd_byte *) (((long) (data + 3)) & ~ 3);
+ rva_bias += data - p;
+ if (data == (dataend - 4))
+ data = dataend;
+ index ++;
+ }
+ BFD_ASSERT (index == num_resource_sets);
+
+ /* Step three: Merge the top level tables (there can be only one).
+
+ We must ensure that the merged entries are in ascending order.
+
+ We also thread the top level table entries from the old tree onto
+ the new table, so that they can be pulled off later. */
+
+ /* FIXME: Should we verify that all type tables are the same ? */
+ new_table.characteristics = type_tables[0].characteristics;
+ new_table.time = type_tables[0].time;
+ new_table.major = type_tables[0].major;
+ new_table.minor = type_tables[0].minor;
+
+ /* Chain the NAME entries onto the table. */
+ new_table.names.first_entry = NULL;
+ new_table.names.last_entry = NULL;
+
+ for (index = 0; index < num_resource_sets; index++)
+ rsrc_attach_chain (& new_table.names, & type_tables[index].names);
+
+ rsrc_sort_entries (& new_table.names, TRUE, & new_table);
+
+ /* Chain the ID entries onto the table. */
+ new_table.ids.first_entry = NULL;
+ new_table.ids.last_entry = NULL;
+
+ for (index = 0; index < num_resource_sets; index++)
+ rsrc_attach_chain (& new_table.ids, & type_tables[index].ids);
+
+ rsrc_sort_entries (& new_table.ids, FALSE, & new_table);
+
+ /* Step four: Create new contents for the .rsrc section. */
+ bfd_byte * new_data = bfd_malloc (size);
+ if (new_data == NULL)
+ goto end;
+
+ rsrc_write_data write_data;
+
+ write_data.abfd = abfd;
+ write_data.datastart = new_data;
+ write_data.next_table = new_data;
+ write_data.next_leaf = new_data + sizeof_tables_and_entries;
+ write_data.next_string = write_data.next_leaf + sizeof_leaves;
+ write_data.next_data = write_data.next_string + sizeof_strings;
+
+ write_data.rva_bias = sec->vma - pe->pe_opthdr.ImageBase;
+
+ rsrc_write_directory (& write_data, & new_table);
+
+ /* Step five: Replace the old contents with the new.
+ We recompute the size as we may have lost entries due to mergeing. */
+ size = ((write_data.next_data - new_data) + 3) & ~3;
+ bfd_set_section_contents (pfinfo->output_bfd, sec, new_data, 0, size);
+ sec->size = sec->rawsize = size;
+
+ end:
+ /* FIXME: Free the resource tree, if we have one. */
+ free (datastart);
+}
/* Handle the .idata section and other things that need symbol table
access. */
@@ -2689,6 +3803,8 @@ _bfd_XXi_final_link_postscript (bfd * abfd, struct coff_final_link_info *pfinfo)
}
#endif
+ rsrc_process_section (abfd, pfinfo);
+
/* If we couldn't find idata$2, we either have an excessively
trivial program or are in DEEP trouble; we have to assume trivial
program.... */