Add support for generating DWARF-5 format directory and file name tables from the assembler.

PR 25611
	PR 25614
	* dwarf.h (DWARF2_Internal_LineInfo): Add li_address_size and
	li_segment_size fields.
	* dwarf.c (read_debug_line_header): Record the address size and
	segment selector size values (if present) in the lineinfo
	structure.
	(display_formatted_table): Warn if the format count is empty but
	the table itself is not empty.
	Display the format count and entry count at the start of the table
	dump.
	(display_debug_lines_raw): Display the address size and segement
	selector size fields, if present.
	* testsuite/binutils-all/dw5.W: Update expected output.

gas	* dwarf2dbg.c (DWARF2_FILE_TIME_NAME): Default to -1.
	(DWARF2_FILE_SIZE_NAME): Default to -1.
	(DWARF2_LINE_VERSION): Default to the current dwarf level or 3,
	whichever is higher.
	(DWARF2_LINE_MAX_OPS_PER_INSN): Provide a default value of 1.
	(NUM_MD5_BYTES): Define.
	(struct file entry): Add md5 field.
	(get_filenum): Delete and replace with...
	(get_basename): New function.
	(get_directory_table_entry): New function.
	(allocate_filenum): New function.
	(allocate_filename_to_slot): New function.
	(dwarf2_where): Use new functions.
	(dwarf2_directive_filename): Add support for extended .file
	pseudo-op.
	(dwarf2_directive_loc): Allow the use of file number zero with
	DWARF 5 or higher.
	(out_file_list): Rename to...
	(out_dir_and_file_list): Add DWARF 5 support.
	(out_debug_line): Emit extra values into the section header for
	DWARF 5.
	(out_debug_str): Allow for file 0 to be used with DWARF 5.
	* doc/as.texi (.file): Update the description of this pseudo-op.
	* testsuite/gas/elf-dwarf-5-file0.s: Add more lines.
	* testsuite/gas/elf-dwarf-5-file0.d: Update expected dump output.
	* testsuite/gas/lns/lns-diag-1.l: Update expected error message.
	* NEWS: Mention the new feature.

binutils/ChangeLog

@@ -1,3 +1,20 @@
+2020-03-11  Nick Clifton  <nickc@redhat.com>
+
+	PR 25611
+	PR 25614
+	* dwarf.h (DWARF2_Internal_LineInfo): Add li_address_size and
+	li_segment_size fields.
+	* dwarf.c (read_debug_line_header): Record the address size and
+	segment selector size values (if present) in the lineinfo
+	structure.
+	(display_formatted_table): Warn if the format count is empty but
+	the table itself is not empty.
+	Display the format count and entry count at the start of the table
+	dump.
+	(display_debug_lines_raw): Display the address size and segement
+	selector size fields, if present.
+	* testsuite/binutils-all/dw5.W: Update expected output.
+
 2020-03-11  Alan Modra  <amodra@gmail.com>
 
 	PR 25651

binutils/dwarf.c

@@ -3622,7 +3622,6 @@ read_debug_line_header (struct dwarf_section * section,
 {
   unsigned char *hdrptr;
   unsigned int initial_length_size;
-  unsigned char address_size, segment_selector_size;
 
   /* Extract information from the Line Number Program Header.
      (section 6.2.4 in the Dwarf3 doc).  */
@@ -3679,15 +3678,15 @@ read_debug_line_header (struct dwarf_section * section,
 
   if (linfo->li_version >= 5)
     {
-      SAFE_BYTE_GET_AND_INC (address_size, hdrptr, 1, end);
+      SAFE_BYTE_GET_AND_INC (linfo->li_address_size, hdrptr, 1, end);
 
-      SAFE_BYTE_GET_AND_INC (segment_selector_size, hdrptr, 1, end);
+      SAFE_BYTE_GET_AND_INC (linfo->li_segment_size, hdrptr, 1, end);
-      if (segment_selector_size != 0)
+      if (linfo->li_segment_size != 0)
 	{
 	  warn (_("The %s section contains "
 		  "unsupported segment selector size: %d.\n"),
-		section->name, segment_selector_size);
+		section->name, linfo->li_segment_size);
-	  return 0;
+	  return NULL;
 	}
     }
 
@@ -3737,6 +3736,7 @@ display_formatted_table (unsigned char *                   data,
   unsigned char *format_start, format_count, *format, formati;
   dwarf_vma data_count, datai;
   unsigned int namepass, last_entry = 0;
+  const char * table_name = is_dir ? N_("Directory Table") : N_("File Name Table");
   
   SAFE_BYTE_GET_AND_INC (format_count, data, 1, end);
   format_start = data;
@@ -3746,10 +3746,7 @@ display_formatted_table (unsigned char *                   data,
       SKIP_ULEB (data, end);
       if (data == end)
 	{
-	  if (is_dir)
+	  warn (_("%s: Corrupt format description entry\n"), table_name);
-	    warn (_("Corrupt directory format table entry\n"));
-	  else
-	    warn (_("Corrupt file name format table entry\n"));
 	  return data;
 	}
     }
@@ -3757,28 +3754,25 @@ display_formatted_table (unsigned char *                   data,
   READ_ULEB (data_count, data, end);
   if (data == end)
     {
-      if (is_dir)
+      warn (_("%s: Corrupt entry count\n"), table_name);
-	warn (_("Corrupt directory list\n"));
-      else
-	warn (_("Corrupt file name list\n"));
       return data;
     }
 
   if (data_count == 0)
     {
-      if (is_dir)
+      printf (_("\n The %s is empty.\n"), table_name);
-	printf (_("\n The Directory Table is empty.\n"));
-      else
-	printf (_("\n The File Name Table is empty.\n"));
       return data;
     }
+  else if (format_count == 0)
+    {
+      warn (_("%s: format count is zero, but the table is not empty\n"),
+	    table_name);
+      return end;
+    }
 
-  if (is_dir)
+  printf (_("\n The %s (offset 0x%lx, lines %s, columns %u):\n"),
-    printf (_("\n The Directory Table (offset 0x%lx):\n"),
+	  table_name, (long) (data - start), dwarf_vmatoa ("u", data_count),
-	    (long) (data - start));
+	  format_count);
-  else
-    printf (_("\n The File Name Table (offset 0x%lx):\n"),
-	    (long) (data - start));
 
   printf (_("  Entry"));
   /* Delay displaying name as the last entry for better screen layout.  */ 
@@ -3806,7 +3800,7 @@ display_formatted_table (unsigned char *                   data,
 		printf (_("\tSize"));
 		break;
 	      case DW_LNCT_MD5:
-		printf (_("\tMD5"));
+		printf (_("\tMD5\t\t\t"));
 		break;
 	      default:
 		printf (_("\t(Unknown format content type %s)"),
@@ -3840,12 +3834,10 @@ display_formatted_table (unsigned char *                   data,
 						  section, NULL, '\t', -1);
 	    }
 	}
-      if (data == end)
+
+      if (data == end && (datai < data_count - 1))
 	{
-	  if (is_dir)
+	  warn (_("\n%s: Corrupt entries list\n"), table_name);
-	    warn (_("Corrupt directory entries list\n"));
-	  else
-	    warn (_("Corrupt file name entries list\n"));
 	  return data;
 	}
       putchar ('\n');
@@ -3909,6 +3901,11 @@ display_debug_lines_raw (struct dwarf_section *  section,
 	  printf (_("  Offset:                      0x%lx\n"), (long)(data - start));
 	  printf (_("  Length:                      %ld\n"), (long) linfo.li_length);
 	  printf (_("  DWARF Version:               %d\n"), linfo.li_version);
+	  if (linfo.li_version >= 5)
+	    {
+	      printf (_("  Address size (bytes):        %d\n"), linfo.li_address_size);
+	      printf (_("  Segment selector (bytes):    %d\n"), linfo.li_segment_size);
+	    }
 	  printf (_("  Prologue Length:             %d\n"), (int) linfo.li_prologue_length);
 	  printf (_("  Minimum Instruction Length:  %d\n"), linfo.li_min_insn_length);
 	  if (linfo.li_version >= 4)

binutils/dwarf.h

@@ -29,6 +29,8 @@ typedef struct
 {
   dwarf_vma	 li_length;
   unsigned short li_version;
+  unsigned char  li_address_size;
+  unsigned char  li_segment_size;
   dwarf_vma      li_prologue_length;
   unsigned char  li_min_insn_length;
   unsigned char  li_max_ops_per_insn;

binutils/testsuite/binutils-all/dw5.W

@@ -291,6 +291,8 @@ Raw dump of debug contents of section .debug_line:
   Offset:                      0x0
   Length:                      144
   DWARF Version:               5
+  Address size \(bytes\):        8
+  Segment selector \(bytes\):    0
   Prologue Length:             60
   Minimum Instruction Length:  1
   Maximum Ops per Instruction: 1
@@ -313,13 +315,13 @@ Raw dump of debug contents of section .debug_line:
   Opcode 11 has 0 args
   Opcode 12 has 1 arg
 
- The Directory Table \(offset 0x22\):
+ The Directory Table \(offset 0x22, lines 3, columns 1\):
   Entry	Name
   0	\(indirect line string, offset: 0x0\): 
   1	\(indirect line string, offset: 0x1\): 
   2	\(indirect line string, offset: 0x22\): /usr/include
 
- The File Name Table \(offset 0x34\):
+ The File Name Table \(offset 0x34, lines 4, columns 2\):
   Entry	Dir	Name
   0	0	\(indirect line string, offset: 0x14\): main.c
   1	1	\(indirect line string, offset: 0x1b\): main.c

gas/ChangeLog

@@ -1,3 +1,35 @@
+2020-03-11  Nick Clifton  <nickc@redhat.com>
+
+	PR 25611
+	PR 25614
+	* dwarf2dbg.c (DWARF2_FILE_TIME_NAME): Default to -1.
+	(DWARF2_FILE_SIZE_NAME): Default to -1.
+	(DWARF2_LINE_VERSION): Default to the current dwarf level or 3,
+	whichever is higher.
+	(DWARF2_LINE_MAX_OPS_PER_INSN): Provide a default value of 1.
+	(NUM_MD5_BYTES): Define.
+	(struct file entry): Add md5 field.
+	(get_filenum): Delete and replace with...
+	(get_basename): New function.
+	(get_directory_table_entry): New function.
+	(allocate_filenum): New function.
+	(allocate_filename_to_slot): New function.
+	(dwarf2_where): Use new functions.
+	(dwarf2_directive_filename): Add support for extended .file
+	pseudo-op.
+	(dwarf2_directive_loc): Allow the use of file number zero with
+	DWARF 5 or higher.
+	(out_file_list): Rename to...
+	(out_dir_and_file_list): Add DWARF 5 support.
+	(out_debug_line): Emit extra values into the section header for
+	DWARF 5.
+	(out_debug_str): Allow for file 0 to be used with DWARF 5.
+	* doc/as.texi (.file): Update the description of this pseudo-op.
+	* testsuite/gas/elf-dwarf-5-file0.s: Add more lines.
+	* testsuite/gas/elf-dwarf-5-file0.d: Update expected dump output.
+	* testsuite/gas/lns/lns-diag-1.l: Update expected error message.
+	* NEWS: Mention the new feature.
+
 2020-03-10  Alan Modra  <amodra@gmail.com>
 
 	* config/tc-csky.c (get_operand_value): Rewrite 1 << 31 expressions

gas/NEWS

@@ -1,5 +1,9 @@
 -*- text -*-
 
+* Add --gdwarf-5 option to the assembler to generate DWARF 5 debug output
+  (if such output is being generated).  Added the ability to generate
+  version 5 .debug_line sections.
+
 Changes in 2.34:
 
 * Add -malign-branch-boundary=NUM, -malign-branch=TYPE[+TYPE...],

gas/doc/as.texi

@@ -5366,13 +5366,29 @@ to the @code{.debug_line} file name table.  The syntax is:
 
 The @var{fileno} operand should be a unique positive integer to use as the
 index of the entry in the table.  The @var{filename} operand is a C string
-literal.
+literal enclosed in double quotes.  The @var{filename} can include directory
+elements.  If it does, then the directory will be added to the directory table
+and the basename will be added to the file table.
 
 The detail of filename indices is exposed to the user because the filename
 table is shared with the @code{.debug_info} section of the DWARF2 debugging
 information, and thus the user must know the exact indices that table
 entries will have.
 
+If DWARF-5 support has been enabled via the @option{-gdwarf-5} option then
+an extended version of the @code{file} is also allowed:
+
+@smallexample
+.file @var{fileno} [@var{dirname}] @var{filename} [md5 @var{value}]
+@end smallexample
+
+With this version a separate directory name is allowed, although if this is
+used then @var{filename} should not contain any directory components.  In
+addtion an md5 hash value of the contents of @var{filename} can be provided.
+This will be stored in the the file table as well, and can be used by tools
+reading the debug information to verify that the contents of the source file
+match the contents of the compiled file.
+
 @node Fill
 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
 

gas/dwarf2dbg.c

@@ -30,6 +30,7 @@
 
 #include "as.h"
 #include "safe-ctype.h"
+#include "bignum.h"
 
 #ifdef HAVE_LIMITS_H
 #include <limits.h>
@@ -81,11 +82,11 @@
 #endif
 
 #ifndef DWARF2_FILE_TIME_NAME
-#define DWARF2_FILE_TIME_NAME(FILENAME,DIRNAME) 0
+#define DWARF2_FILE_TIME_NAME(FILENAME,DIRNAME) -1
 #endif
 
 #ifndef DWARF2_FILE_SIZE_NAME
-#define DWARF2_FILE_SIZE_NAME(FILENAME,DIRNAME) 0
+#define DWARF2_FILE_SIZE_NAME(FILENAME,DIRNAME) -1
 #endif
 
 #ifndef DWARF2_VERSION
@@ -99,7 +100,7 @@
 
 /* This implementation outputs version 3 .debug_line information.  */
 #ifndef DWARF2_LINE_VERSION
-#define DWARF2_LINE_VERSION 3
+#define DWARF2_LINE_VERSION (dwarf_level > 3 ? dwarf_level : 3)
 #endif
 
 #include "subsegs.h"
@@ -147,6 +148,10 @@
 /* Flag that indicates the initial value of the is_stmt_start flag.  */
 #define	DWARF2_LINE_DEFAULT_IS_STMT	1
 
+#ifndef DWARF2_LINE_MAX_OPS_PER_INSN
+#define DWARF2_LINE_MAX_OPS_PER_INSN	1
+#endif
+
 /* Given a special op, return the line skip amount.  */
 #define SPECIAL_LINE(op) \
 	(((op) - DWARF2_LINE_OPCODE_BASE)%DWARF2_LINE_RANGE + DWARF2_LINE_BASE)
@@ -196,10 +201,13 @@ struct line_seg
 static struct line_seg *all_segs;
 static struct line_seg **last_seg_ptr;
 
+#define NUM_MD5_BYTES       16
+
 struct file_entry
 {
-  const char *filename;
+  const char *   filename;
   unsigned int   dir;
+  unsigned char  md5[NUM_MD5_BYTES];
 };
 
 /* Table of files used by .debug_line.  */
@@ -208,9 +216,9 @@ static unsigned int files_in_use;
 static unsigned int files_allocated;
 
 /* Table of directories used by .debug_line.  */
-static char **dirs;
+static char **       dirs = NULL;
-static unsigned int dirs_in_use;
+static unsigned int  dirs_in_use = 0;
-static unsigned int dirs_allocated;
+static unsigned int  dirs_allocated = 0;
 
 /* TRUE when we've seen a .loc directive recently.  Used to avoid
    doing work when there's nothing to do.  */
@@ -239,8 +247,6 @@ static symbolS *view_assert_failed;
 /* The size of an address on the target.  */
 static unsigned int sizeof_address;
 
-static unsigned int get_filenum (const char *, unsigned int);
-
 #ifndef TC_DWARF2_EMIT_OFFSET
 #define TC_DWARF2_EMIT_OFFSET  generic_dwarf2_emit_offset
 
@@ -280,6 +286,7 @@ get_line_subseg (segT seg, subsegT subseg, bfd_boolean create_p)
       last_seg_ptr = &s->next;
       seg_info (seg)->dwarf2_line_seg = s;
     }
+
   gas_assert (seg == s->seg);
 
   for (pss = &s->head; (lss = *pss) != NULL ; pss = &lss->next)
@@ -516,7 +523,9 @@ dwarf2_gen_line_info (addressT ofs, struct dwarf2_line_info *loc)
   symbolS *sym;
 
   /* Early out for as-yet incomplete location information.  */
-  if (loc->filenum == 0 || loc->line == 0)
+  if (loc->line == 0)
+    return;
+  if (loc->filenum == 0 && DWARF2_LINE_VERSION < 5)
     return;
 
   /* Don't emit sequences of line symbols for the same line when the
@@ -544,6 +553,323 @@ dwarf2_gen_line_info (addressT ofs, struct dwarf2_line_info *loc)
   dwarf2_gen_line_info_1 (sym, loc);
 }
 
+static const char *
+get_basename (const char * pathname)
+{
+  const char * file;
+
+  file = lbasename (pathname);
+  /* Don't make empty string from / or A: from A:/ .  */
+#ifdef HAVE_DOS_BASED_FILE_SYSTEM
+  if (file <= pathname + 3)
+    file = pathname;
+#else
+  if (file == pathname + 1)
+    file = pathname;
+#endif
+  return file;
+}
+
+static unsigned int
+get_directory_table_entry (const char *  dirname,
+			   size_t        dirlen,
+			   bfd_boolean   can_use_zero)
+{
+  unsigned int d;
+
+  if (dirlen == 0)
+    return 0;
+
+#ifndef DWARF2_DIR_SHOULD_END_WITH_SEPARATOR
+  if (IS_DIR_SEPARATOR (dirname[dirlen - 1]))
+    {
+      -- dirlen;
+      if (dirlen == 0)
+	return 0;
+    }
+#endif
+
+  for (d = 0; d < dirs_in_use; ++d)
+    {
+      if (dirs[d] != NULL
+	  && filename_ncmp (dirname, dirs[d], dirlen) == 0
+	  && dirs[d][dirlen] == '\0')
+	return d;
+    }
+
+  if (can_use_zero)
+    {
+      if (dirs == NULL || dirs[0] == NULL)
+	d = 0;
+    }
+  else if (d == 0)
+    d = 1;
+
+  if (d >= dirs_allocated)
+    {
+      unsigned int old = dirs_allocated;
+
+      dirs_allocated = d + 32;
+      dirs = XRESIZEVEC (char *, dirs, dirs_allocated);
+      memset (dirs + old, 0, (dirs_allocated - old) * sizeof (char *));
+    }
+
+  dirs[d] = xmemdup0 (dirname, dirlen);
+  if (dirs_in_use <= d)
+    dirs_in_use = d + 1;
+
+  return d;  
+}
+
+/* Get a .debug_line file number for PATHNAME.  If there is a
+   directory component to PATHNAME, then this will be stored
+   in the directory table, if it is not already present.
+   Returns the slot number allocated to that filename or -1
+   if there was a problem.  */
+
+static signed int
+allocate_filenum (const char * pathname)
+{
+  static signed int last_used = -1, last_used_dir_len = 0;
+  const char *file;
+  size_t dir_len;
+  unsigned int i, dir;
+
+  /* Short circuit the common case of adding the same pathname
+     as last time.  */
+  if (last_used != -1)
+    {
+      const char * dirname = NULL;
+
+      if (dirs != NULL)
+	dirname = dirs[files[last_used].dir];
+
+      if (dirname == NULL)
+	{
+	  if (filename_cmp (pathname, files[last_used].filename) == 0)
+	    return last_used;
+	}
+      else
+	{
+	  if (filename_ncmp (pathname, dirname, last_used_dir_len) == 0
+	      && IS_DIR_SEPARATOR (pathname [last_used_dir_len])
+	      && filename_cmp (pathname + last_used_dir_len + 1,
+			       files[last_used].filename) == 0)
+	    return last_used;
+	}
+    }
+
+  file = get_basename (pathname);
+  dir_len = file - pathname;
+
+  dir = get_directory_table_entry (pathname, dir_len, FALSE);
+
+  /* Do not use slot-0.  That is specificailly reserved for use by
+     the '.file 0 "name"' directive.  */
+  for (i = 1; i < files_in_use; ++i)
+    if (files[i].dir == dir
+	&& files[i].filename
+	&& filename_cmp (file, files[i].filename) == 0)
+      {
+	last_used = i;
+	last_used_dir_len = dir_len;
+	return i;
+      }
+
+  if (i >= files_allocated)
+    {
+      unsigned int old = files_allocated;
+
+      files_allocated = i + 32;
+      /* Catch wraparound.  */
+      if (files_allocated <= old)
+	{
+	  as_bad (_("file number %lu is too big"), (unsigned long) i);
+	  return -1;
+	}
+
+      files = XRESIZEVEC (struct file_entry, files, files_allocated);
+      memset (files + old, 0, (i + 32 - old) * sizeof (struct file_entry));
+    }
+
+  files[i].filename = file;
+  files[i].dir = dir;
+  memset (files[i].md5, 0, NUM_MD5_BYTES);
+
+  if (files_in_use < i + 1)
+    files_in_use = i + 1;  
+  last_used = i;
+  last_used_dir_len = dir_len;
+
+  return i;
+}
+
+/* Allocate slot NUM in the .debug_line file table to FILENAME.
+   If DIRNAME is not NULL or there is a directory component to FILENAME
+   then this will be stored in the directory table, if not already present.
+   if WITH_MD5 is TRUE then there is a md5 value in generic_bignum.
+   Returns TRUE if allocation succeeded, FALSE otherwise.  */
+
+static bfd_boolean
+allocate_filename_to_slot (const char *  dirname,
+			   const char *  filename,
+			   unsigned int  num,
+			   bfd_boolean   with_md5)
+{
+  const char *file;
+  size_t dirlen;
+  unsigned int i, d;
+
+  /* Short circuit the common case of adding the same pathname
+     as last time.  */
+  if (num < files_allocated && files[num].filename != NULL)
+    {
+      const char * dir = NULL;
+
+      if (dirs)
+	dir = dirs[files[num].dir];
+
+      if (with_md5
+	  && memcmp (generic_bignum, files[num].md5, NUM_MD5_BYTES) != 0)
+	goto fail;
+
+      if (dirname != NULL)
+	{
+	  if (dir != NULL && filename_cmp (dir, dirname) != 0)
+	    goto fail;
+      
+	  if (filename_cmp (filename, files[num].filename) != 0)
+	    goto fail;
+
+	  /* If the filenames match, but the directory table entry was
+	     empty, then fill it with the provided directory name.  */
+	  if (dir == NULL)
+	    dirs[files[num].dir] = xmemdup0 (dirname, strlen (dirname));
+	    
+	  return TRUE;
+	}
+      else if (dir != NULL) 
+	{
+	  dirlen = strlen (dir);
+	  if (filename_ncmp (filename, dir, dirlen) == 0
+	      && IS_DIR_SEPARATOR (filename [dirlen])
+	      && filename_cmp (filename + dirlen + 1, files[num].filename) == 0)
+	    return TRUE;
+	}
+      else /* dir == NULL  */
+	{
+	  file = get_basename (filename);
+	  if (filename_cmp (file, files[num].filename) == 0)
+	    {
+	      if (file > filename)
+		/* The filenames match, but the directory table entry is empty.
+		   Fill it with the provided directory name.  */
+		dirs[files[num].dir] = xmemdup0 (filename, file - filename);
+	      return TRUE;
+	    }
+	}
+
+    fail:
+      as_bad (_("file table slot %u is already occupied by a different file (%s%s%s vs %s%s%s)"),
+	      num,
+	      dir == NULL ? "" : dir,
+	      dir == NULL ? "" : "/",
+	      files[num].filename,
+	      dirname == NULL ? "" : dirname,
+	      dirname == NULL ? "" : "/",
+	      filename);
+      return FALSE;
+    }
+
+  if (dirname == NULL)
+    {
+      dirname = filename;
+      file = get_basename (filename);
+      dirlen = file - filename;
+    }
+  else
+    {
+      dirlen = strlen (dirname);
+      file = filename;
+    }
+  
+  d = get_directory_table_entry (dirname, dirlen, num == 0);
+  i = num;
+
+  if (i >= files_allocated)
+    {
+      unsigned int old = files_allocated;
+
+      files_allocated = i + 32;
+      /* Catch wraparound.  */
+      if (files_allocated <= old)
+	{
+	  as_bad (_("file number %lu is too big"), (unsigned long) i);
+	  return FALSE;
+	}
+
+      files = XRESIZEVEC (struct file_entry, files, files_allocated);
+      memset (files + old, 0, (i + 32 - old) * sizeof (struct file_entry));
+    }
+
+  files[i].filename = file;
+  files[i].dir = d;
+  if (with_md5)
+    {
+      if (target_big_endian)
+	{
+	  /* md5's are stored in litte endian format.  */
+	  unsigned int     bits_remaining = NUM_MD5_BYTES * BITS_PER_CHAR;
+	  unsigned int     byte = NUM_MD5_BYTES;
+	  unsigned int     bignum_index = 0;
+
+	  while (bits_remaining)
+	    {
+	      unsigned int bignum_bits_remaining = LITTLENUM_NUMBER_OF_BITS;
+	      valueT       bignum_value = generic_bignum [bignum_index];
+	      bignum_index ++;
+
+	      while (bignum_bits_remaining)
+		{
+		  files[i].md5[--byte] = bignum_value & 0xff;
+		  bignum_value >>= 8;
+		  bignum_bits_remaining -= 8;
+		  bits_remaining -= 8;
+		}
+	    }
+	}
+      else
+	{
+	  unsigned int     bits_remaining = NUM_MD5_BYTES * BITS_PER_CHAR;
+	  unsigned int     byte = 0;
+	  unsigned int     bignum_index = 0;
+
+	  while (bits_remaining)
+	    {
+	      unsigned int bignum_bits_remaining = LITTLENUM_NUMBER_OF_BITS;
+	      valueT       bignum_value = generic_bignum [bignum_index];
+
+	      bignum_index ++;
+
+	      while (bignum_bits_remaining)
+		{
+		  files[i].md5[byte++] = bignum_value & 0xff;
+		  bignum_value >>= 8;
+		  bignum_bits_remaining -= 8;
+		  bits_remaining -= 8;
+		}
+	    }
+	}
+    }
+  else
+    memset (files[i].md5, 0, NUM_MD5_BYTES);
+
+  if (files_in_use < i + 1)
+    files_in_use = i + 1;
+
+  return TRUE;
+}
+
 /* Returns the current source information.  If .file directives have
    been encountered, the info for the corresponding source file is
    returned.  Otherwise, the info for the assembly source file is
@@ -558,7 +884,8 @@ dwarf2_where (struct dwarf2_line_info *line)
 
       memset (line, 0, sizeof (*line));
       filename = as_where (&line->line);
-      line->filenum = get_filenum (filename, 0);
+      line->filenum = allocate_filenum (filename);
+      /* FIXME: We should check the return value from allocate_filenum.  */
       line->column = 0;
       line->flags = DWARF2_FLAG_IS_STMT;
       line->isa = current.isa;
@@ -670,108 +997,6 @@ dwarf2_emit_label (symbolS *label)
   dwarf2_consume_line_info ();
 }
 
-/* Get a .debug_line file number for FILENAME.  If NUM is nonzero,
-   allocate it on that file table slot, otherwise return the first
-   empty one.  */
-
-static unsigned int
-get_filenum (const char *filename, unsigned int num)
-{
-  static unsigned int last_used, last_used_dir_len;
-  const char *file;
-  size_t dir_len;
-  unsigned int i, dir;
-
-  if (num == 0 && last_used)
-    {
-      if (! files[last_used].dir
-	  && filename_cmp (filename, files[last_used].filename) == 0)
-	return last_used;
-      if (files[last_used].dir
-	  && filename_ncmp (filename, dirs[files[last_used].dir],
-			    last_used_dir_len) == 0
-	  && IS_DIR_SEPARATOR (filename [last_used_dir_len])
-	  && filename_cmp (filename + last_used_dir_len + 1,
-			   files[last_used].filename) == 0)
-	return last_used;
-    }
-
-  file = lbasename (filename);
-  /* Don't make empty string from / or A: from A:/ .  */
-#ifdef HAVE_DOS_BASED_FILE_SYSTEM
-  if (file <= filename + 3)
-    file = filename;
-#else
-  if (file == filename + 1)
-    file = filename;
-#endif
-  dir_len = file - filename;
-
-  dir = 0;
-  if (dir_len)
-    {
-#ifndef DWARF2_DIR_SHOULD_END_WITH_SEPARATOR
-      --dir_len;
-#endif
-      for (dir = 1; dir < dirs_in_use; ++dir)
-	if (filename_ncmp (filename, dirs[dir], dir_len) == 0
-	    && dirs[dir][dir_len] == '\0')
-	  break;
-
-      if (dir >= dirs_in_use)
-	{
-	  if (dir >= dirs_allocated)
-	    {
-	      dirs_allocated = dir + 32;
-	      dirs = XRESIZEVEC (char *, dirs, dirs_allocated);
-	    }
-
-	  dirs[dir] = xmemdup0 (filename, dir_len);
-	  dirs_in_use = dir + 1;
-	}
-    }
-
-  if (num == 0)
-    {
-      for (i = 1; i < files_in_use; ++i)
-	if (files[i].dir == dir
-	    && files[i].filename
-	    && filename_cmp (file, files[i].filename) == 0)
-	  {
-	    last_used = i;
-	    last_used_dir_len = dir_len;
-	    return i;
-	  }
-    }
-  else
-    i = num;
-
-  if (i >= files_allocated)
-    {
-      unsigned int old = files_allocated;
-
-      files_allocated = i + 32;
-      /* Catch wraparound.  */
-      if (files_allocated <= old)
-	{
-	  as_bad (_("file number %lu is too big"), (unsigned long) i);
-	  return 0;
-	}
-
-      files = XRESIZEVEC (struct file_entry, files, files_allocated);
-      memset (files + old, 0, (i + 32 - old) * sizeof (struct file_entry));
-    }
-
-  files[i].filename = file;
-  files[i].dir = dir;
-  if (files_in_use < i + 1)
-    files_in_use = i + 1;
-  last_used = i;
-  last_used_dir_len = dir_len;
-
-  return i;
-}
-
 /* Handle two forms of .file directive:
    - Pass .file "source.c" to s_app_file
    - Handle .file 1 "source.c" by adding an entry to the DWARF-2 file table
@@ -781,8 +1006,10 @@ get_filenum (const char *filename, unsigned int num)
 char *
 dwarf2_directive_filename (void)
 {
+  bfd_boolean with_md5 = TRUE;
   valueT num;
   char *filename;
+  const char * dirname = NULL;
   int filename_len;
 
   /* Continue to accept a bare string and pass it off.  */
@@ -795,24 +1022,48 @@ dwarf2_directive_filename (void)
 
   num = get_absolute_expression ();
 
-  if ((offsetT) num < 1 && dwarf_level < 5)
+  if ((offsetT) num < 1 && DWARF2_LINE_VERSION < 5)
     {
       as_bad (_("file number less than one"));
       ignore_rest_of_line ();
       return NULL;
     }
 
-  if (num == 0)
+  /* FIXME: Should we allow ".file <N>\n" as an expression meaning
-    {
+     "switch back to the already allocated file <N> as the current
-      demand_empty_rest_of_line ();
+     file" ?  */
-      return NULL;
-    }
 
   filename = demand_copy_C_string (&filename_len);
   if (filename == NULL)
     /* demand_copy_C_string will have already generated an error message.  */
     return NULL;
 
+  /* For DWARF-5 support we also accept:
+     .file <NUM> ["<dir>"] "<file>" [md5 <NUM>]  */
+  if (DWARF2_LINE_VERSION > 4)
+    {
+      SKIP_WHITESPACE ();
+      if (*input_line_pointer == '"')
+	{
+	  dirname = filename;
+	  filename = demand_copy_C_string (&filename_len);
+	  SKIP_WHITESPACE ();
+	}
+
+      if (strncmp (input_line_pointer, "md5", 3) == 0)
+	{
+	  input_line_pointer += 3;
+	  SKIP_WHITESPACE ();
+
+	  expressionS exp;
+	  expression_and_evaluate (& exp);
+	  if (exp.X_op != O_big)
+	    as_bad (_("md5 value too small or not a constant"));
+	  else
+	    with_md5 = TRUE;
+	}
+    }
+
   demand_empty_rest_of_line ();
 
   /* A .file directive implies compiler generated debug information is
@@ -825,13 +1076,10 @@ dwarf2_directive_filename (void)
       as_bad (_("file number %lu is too big"), (unsigned long) num);
       return NULL;
     }
-  if (num < files_in_use && files[num].filename != 0)
-    {
-      as_bad (_("file number %u already allocated"), (unsigned int) num);
-      return NULL;
-    }
 
-  (void) get_filenum (filename, (unsigned int) num);
+  if (! allocate_filename_to_slot (dirname, filename, (unsigned int) num,
+				   with_md5))
+    return NULL;
 
   return filename;
 }
@@ -860,11 +1108,15 @@ dwarf2_directive_loc (int dummy ATTRIBUTE_UNUSED)
   line = get_absolute_expression ();
 
   if (filenum < 1)
+    {
+      if (filenum != 0 || DWARF2_LINE_VERSION < 5)
 	{
 	  as_bad (_("file number less than one"));
 	  return;
 	}
-  if (filenum >= (int) files_in_use || files[filenum].filename == 0)
+    }
+
+  if (filenum >= (int) files_in_use || files[filenum].filename == NULL)
     {
       as_bad (_("unassigned file number %ld"), (long) filenum);
       return;
@@ -1710,14 +1962,40 @@ process_entries (segT seg, struct line_entry *e)
 /* Emit the directory and file tables for .debug_line.  */
 
 static void
-out_file_list (void)
+out_dir_and_file_list (void)
 {
   size_t size;
   const char *dir;
   char *cp;
   unsigned int i;
+  bfd_boolean emit_md5 = FALSE;
+  bfd_boolean emit_timestamps = TRUE;
+  bfd_boolean emit_filesize = TRUE;
+
+  /* Output the Directory Table.  */
+
+  if (DWARF2_LINE_VERSION >= 5)
+    {
+      out_byte (1);
+      out_uleb128 (DW_LNCT_path);
+      /* FIXME: it would be better to store these strings in
+	 the .debug_line_str section and reference them here.  */
+      out_uleb128 (DW_FORM_string);
+      out_uleb128 (dirs_in_use);
+    }
       
   /* Emit directory list.  */
+  if (DWARF2_LINE_VERSION >= 5)
+    {
+      if (dirs[0] == NULL)
+	dir = remap_debug_filename (".");
+      else
+	dir = remap_debug_filename (dirs[0]);
+	
+      size = strlen (dir) + 1;
+      cp = frag_more (size);
+      memcpy (cp, dir, size);
+    }
   for (i = 1; i < dirs_in_use; ++i)
     {
       dir = remap_debug_filename (dirs[i]);
@@ -1725,20 +2003,90 @@ out_file_list (void)
       cp = frag_more (size);
       memcpy (cp, dir, size);
     }
+
+  if (DWARF2_LINE_VERSION < 5)
     /* Terminate it.  */
     out_byte ('\0');
 
-  for (i = 1; i < files_in_use; ++i)
+  /* Output the File Name Table.  */
+
+  if (DWARF2_LINE_VERSION >= 5)
+    {
+      unsigned int columns = 4;
+
+      if (((unsigned long) DWARF2_FILE_TIME_NAME ("", "")) == -1UL)
+	{
+	  emit_timestamps = FALSE;
+	  -- columns;
+	}
+
+      if (DWARF2_FILE_SIZE_NAME ("", "") == -1)
+	{
+	  emit_filesize = FALSE;
+	  -- columns;
+	}
+
+      for (i = 0; i < files_in_use; ++i)
+	if (files[i].md5[0] != 0)
+	  break;
+      if (i < files_in_use)
+	{
+	  emit_md5 = TRUE;
+	  ++ columns;
+	}
+      
+      /* The number of format entries to follow.  */
+      out_byte (columns);
+
+      /* The format of the file name.  */
+      out_uleb128 (DW_LNCT_path);
+      /* FIXME: it would be better to store these strings in
+	 the .debug_line_str section and reference them here.  */
+      out_uleb128 (DW_FORM_string);
+
+      /* The format of the directory index.  */
+      out_uleb128 (DW_LNCT_directory_index);
+      out_uleb128 (DW_FORM_udata);
+
+      if (emit_timestamps)
+	{
+	  /* The format of the timestamp.  */
+	  out_uleb128 (DW_LNCT_timestamp);
+	  out_uleb128 (DW_FORM_udata);
+	}
+
+      if (emit_filesize)
+	{
+	  /* The format of the file size.  */
+	  out_uleb128 (DW_LNCT_size);
+	  out_uleb128 (DW_FORM_udata);
+	}
+
+      if (emit_md5)
+	{
+	  /* The format of the MD5 sum.  */
+	  out_uleb128 (DW_LNCT_MD5);
+	  out_uleb128 (DW_FORM_data16);
+	}
+
+      /* The number of entries in the table.  */
+      out_uleb128 (files_in_use);
+   }
+      
+  for (i = DWARF2_LINE_VERSION > 4 ? 0 : 1; i < files_in_use; ++i)
     {
       const char *fullfilename;
 
       if (files[i].filename == NULL)
 	{
-	  as_bad (_("unassigned file number %ld"), (long) i);
 	  /* Prevent a crash later, particularly for file 1.  */
 	  files[i].filename = "";
+	  if (DWARF2_LINE_VERSION < 5 || i != 0)
+	    {
+	      as_bad (_("unassigned file number %ld"), (long) i);
 	      continue;
 	    }
+	}
 
       fullfilename = DWARF2_FILE_NAME (files[i].filename,
 				       files[i].dir ? dirs [files [i].dir] : "");
@@ -1746,15 +2094,43 @@ out_file_list (void)
       cp = frag_more (size);
       memcpy (cp, fullfilename, size);
 
-      out_uleb128 (files[i].dir);	/* directory number */
+      /* Directory number.  */
+      out_uleb128 (files[i].dir);
+
       /* Output the last modification timestamp.  */
-      out_uleb128 (DWARF2_FILE_TIME_NAME (files[i].filename,
+      if (emit_timestamps)
-					  files[i].dir ? dirs [files [i].dir] : ""));
+	{
-      /* Output the filesize.  */
+	  offsetT timestamp;
-      out_uleb128 (DWARF2_FILE_SIZE_NAME (files[i].filename,
+
-					  files[i].dir ? dirs [files [i].dir] : ""));
+	  timestamp = DWARF2_FILE_TIME_NAME (files[i].filename,
+					     files[i].dir ? dirs [files [i].dir] : "");
+	  if (timestamp == -1)
+	    timestamp = 0;
+	  out_uleb128 (timestamp);
 	}
 
+      /* Output the filesize.  */
+      if (emit_filesize)
+	{
+	  offsetT filesize;
+	  filesize = DWARF2_FILE_SIZE_NAME (files[i].filename,
+					    files[i].dir ? dirs [files [i].dir] : "");
+	  if (filesize == -1)
+	    filesize = 0;
+	  out_uleb128 (filesize);
+	}
+
+      /* Output the md5 sum.  */
+      if (emit_md5)
+	{
+	  int b;
+
+	  for (b = 0; b < NUM_MD5_BYTES; b++)
+	    out_byte (files[i].md5[b]);
+	}
+    }
+
+  if (DWARF2_LINE_VERSION < 5)
     /* Terminate filename list.  */
     out_byte (0);
 }
@@ -1833,6 +2209,11 @@ out_debug_line (segT line_seg)
   /* Version.  */
   out_two (DWARF2_LINE_VERSION);
 
+  if (DWARF2_LINE_VERSION >= 5)
+    {
+      out_byte (sizeof_address);
+      out_byte (0); /* Segment Selector size.  */
+    }
   /* Length of the prologue following this length.  */
   prologue_start = symbol_temp_make ();
   prologue_end = symbol_temp_make ();
@@ -1845,6 +2226,8 @@ out_debug_line (segT line_seg)
 
   /* Parameters of the state machine.  */
   out_byte (DWARF2_LINE_MIN_INSN_LENGTH);
+  if (DWARF2_LINE_VERSION >= 4)
+    out_byte (DWARF2_LINE_MAX_OPS_PER_INSN);
   out_byte (DWARF2_LINE_DEFAULT_IS_STMT);
   out_byte (DWARF2_LINE_BASE);
   out_byte (DWARF2_LINE_RANGE);
@@ -1863,8 +2246,11 @@ out_debug_line (segT line_seg)
   out_byte (0);			/* DW_LNS_set_prologue_end */
   out_byte (0);			/* DW_LNS_set_epilogue_begin */
   out_byte (1);			/* DW_LNS_set_isa */
+  /* We have emitted 12 opcode lengths, so make that this
+     matches up to the opcode base value we have been using.  */
+  gas_assert (DWARF2_LINE_OPCODE_BASE == 13);
 
-  out_file_list ();
+  out_dir_and_file_list ();
 
   symbol_set_value_now (prologue_end);
 
@@ -2134,19 +2520,20 @@ out_debug_str (segT str_seg, symbolS **name_sym, symbolS **comp_dir_sym,
   const char *dirname;
   char *p;
   int len;
+  int first_file = DWARF2_LINE_VERSION > 4 ? 0 : 1;
 
   subseg_set (str_seg, 0);
 
   /* DW_AT_name.  We don't have the actual file name that was present
-     on the command line, so assume files[1] is the main input file.
+     on the command line, so assume files[first_file] is the main input file.
      We're not supposed to get called unless at least one line number
      entry was emitted, so this should always be defined.  */
   *name_sym = symbol_temp_new_now_octets ();
   if (files_in_use == 0)
     abort ();
-  if (files[1].dir)
+  if (files[first_file].dir)
     {
-      dirname = remap_debug_filename (dirs[files[1].dir]);
+      dirname = remap_debug_filename (dirs[files[first_file].dir]);
       len = strlen (dirname);
 #ifdef TE_VMS
       /* Already has trailing slash.  */
@@ -2158,9 +2545,9 @@ out_debug_str (segT str_seg, symbolS **name_sym, symbolS **comp_dir_sym,
       INSERT_DIR_SEPARATOR (p, len);
 #endif
     }
-  len = strlen (files[1].filename) + 1;
+  len = strlen (files[first_file].filename) + 1;
   p = frag_more (len);
-  memcpy (p, files[1].filename, len);
+  memcpy (p, files[first_file].filename, len);
 
   /* DW_AT_comp_dir */
   *comp_dir_sym = symbol_temp_new_now_octets ();

gas/testsuite/gas/elf/dwarf-5-file0.d

@@ -2,5 +2,18 @@
 #name: DWARF5 .line 0
 #readelf: -wl
 
+#...
+ The Directory Table \(offset 0x.*, lines 3, columns 1\):
+  Entry	Name
+  0	master directory
+  1	secondary directory
+  2	/tmp
+
+ The File Name Table \(offset 0x.*, lines 3, columns 3\):
+  Entry	Dir	MD5				Name
+  0	0 0x00000000000000000000000000000000	master source file
+  1	1 0x00000000000000000000000000000000	secondary source file
+  2	2 0x95828e8bc4f7404dbf7526fb7bd0f192	foo.c
 #pass
 
+

gas/testsuite/gas/elf/dwarf-5-file0.s

@@ -1,17 +1,23 @@
-	.text
+	.section	.debug_info,"",%progbits
+	.4byte	0x8a
+	.2byte  0x2
+	.4byte	.Ldebug_abbrev0
+	.byte	0x4
+	.uleb128 0x1
 
- .file 0
+	.file 0 "master directory/master source file"
 	.line 1
-	.long 3
+	.text
+	.word 0
 
- .file 1 "asdf"
+	.file 1 "secondary directory/secondary source file"
 	.line 2
-	.long 5
+	.word 2
 
- .file 0
+	.file "master source file"
 	.line 4
-	.long 3
+	.word 4
 
- .file 2 "def"
+	.file 2 "/tmp" "foo.c" md5 0x95828e8bc4f7404dbf7526fb7bd0f192
 	.line 5
-	.long 3
+	.word 6

gas/testsuite/gas/lns/lns-diag-1.l

@@ -1,7 +1,7 @@
 .*: Assembler messages:
 .*:2: Error: file number less than one
 .*:3: Error: missing string
-.*:4: Error: file number 1 already allocated
+.*:4: Error: file table slot 1 is already occupied.*
 .*:8: Error: unassigned file number 3
 .*:9: Error: junk at end of line, first unrecognized character is `1'
 .*:12: Error: junk at end of line, first unrecognized character is `0'