ELF Format-2-Program Header Table

Program Header Table

程序头表与段表相互独立，由ELF文件头统一管理。

下面将它们简称PH和SH。

程序头表负责ELF文件从文件到加载后映像的映射关系，一般只有可执行文件包含。

readelf -l查看。

1. segment和section

在英语里，segment同质，如一段绳子；section则是指不同种类的东西。

segment

程序头表项描述的对象称为segment，即elf文件加载后的数据块；

它提供给加载器loader使用，loader需要知道rwe属性

可执行文件可以没有section，但一定要有segment。

section

section描述加载前的数据块，给链接器linker使用。

linker主要关心.text, .rel.text, .data, .rodata等，需要重定位。

联系

比如.text/.data段加载信息存在PH表项对应的存放代码/数据的segment中。

有时为了简化PH表项的个数，会把整个elf文件作为一个segment加载（同质???）。

2. 数据结构

typedef struct
{
  Elf32_Word	p_type;			/* Segment type */
  Elf32_Off	    p_offset;		/* Segment file offset */
  Elf32_Addr	p_vaddr;		/* Segment virtual address */
  Elf32_Addr	p_paddr;		/* Segment physical address */
  Elf32_Word	p_filesz;		/* Segment size in file */
  Elf32_Word	p_memsz;		/* Segment size in memory */
  Elf32_Word	p_flags;		/* Segment flags */
  Elf32_Word	p_align;		/* Segment alignment */
} Elf32_Phdr;

这8个字段都描述segment。

2.1 field p_type

segment类型：

/* Legal values for p_type (segment type).  */

#define	PT_NULL		0		/* Program header table entry unused */
#define PT_LOAD		1		/* Loadable program segment */
#define PT_DYNAMIC	2		/* Dynamic linking information */
#define PT_INTERP	3		/* Program interpreter */
#define PT_NOTE		4		/* Auxiliary information */
#define PT_SHLIB	5		/* Reserved */
#define PT_PHDR		6		/* Entry for header table itself */
#define PT_TLS		7		/* Thread-local storage segment */
#define	PT_NUM		8		/* Number of defined types */

一个可执行文件至少有一个PT_LOAD类型的段，如text, data，并根据p_align对齐。

2.2 Other Fields

• p_offset, segment文件偏移；

• p_vaddr, segment虚拟地址；

• p_paddr, segment物理地址。因为现代os都使用了分页机制，所以不用关心，readelf输出结果等于虚拟地址；

• p_filesz, segment文件大小;

• p_memsz, segment内存大小;

• p_align, segmenet内存对齐大小，LOAD段值为0x1000;

• p_flags, segment权限属性

#define PF_X		(1 << 0)	/* Segment is executable */
#define PF_W		(1 << 1)	/* Segment is writable */
#define PF_R		(1 << 2)	/* Segment is readable */

比如text段，PF_X|PF_R。

readelf源码

/* Returns 1 if the program headers were loaded.  */

static int
process_program_headers (FILE * file)
{
  Elf_Internal_Phdr * segment;
  unsigned int i;

  if (elf_header.e_phnum == 0)
    {
      /* PR binutils/12467.  */
      if (elf_header.e_phoff != 0)
	warn (_("possibly corrupt ELF header - it has a non-zero program"
		" header offset, but no program headers"));
      else if (do_segments)
	printf (_("\nThere are no program headers in this file.\n"));
      return 0;
    }

  if (do_segments && !do_header)
    {
      printf (_("\nElf file type is %s\n"), get_file_type (elf_header.e_type));
      printf (_("Entry point "));
      print_vma ((bfd_vma) elf_header.e_entry, PREFIX_HEX);
      printf (_("\nThere are %d program headers, starting at offset "),
	      elf_header.e_phnum);
      print_vma ((bfd_vma) elf_header.e_phoff, DEC);
      printf ("\n");
    }

  if (! get_program_headers (file))
      return 0;

  if (do_segments)
    {
      if (elf_header.e_phnum > 1)
	printf (_("\nProgram Headers:\n"));
      else
	printf (_("\nProgram Headers:\n"));

      if (is_32bit_elf)
	printf
	  (_("  Type           Offset   VirtAddr   PhysAddr   FileSiz MemSiz  Flg Align\n"));
      else if (do_wide)
	printf
	  (_("  Type           Offset   VirtAddr           PhysAddr           FileSiz  MemSiz   Flg Align\n"));
      else
	{
	  printf
	    (_("  Type           Offset             VirtAddr           PhysAddr\n"));
	  printf
	    (_("                 FileSiz            MemSiz              Flags  Align\n"));
	}
    }

  dynamic_addr = 0;
  dynamic_size = 0;

  for (i = 0, segment = program_headers;
       i < elf_header.e_phnum;
       i++, segment++)
    {
      if (do_segments)
	{
	  printf ("  %-14.14s ", get_segment_type (segment->p_type));

	  if (is_32bit_elf)
	    {
	      printf ("0x%6.6lx ", (unsigned long) segment->p_offset);
	      printf ("0x%8.8lx ", (unsigned long) segment->p_vaddr);
	      printf ("0x%8.8lx ", (unsigned long) segment->p_paddr);
	      printf ("0x%5.5lx ", (unsigned long) segment->p_filesz);
	      printf ("0x%5.5lx ", (unsigned long) segment->p_memsz);
	      printf ("%c%c%c ",
		      (segment->p_flags & PF_R ? 'R' : ' '),
		      (segment->p_flags & PF_W ? 'W' : ' '),
		      (segment->p_flags & PF_X ? 'E' : ' '));
	      printf ("%#lx", (unsigned long) segment->p_align);
	    }
	  else if (do_wide)
	    {
	      if ((unsigned long) segment->p_offset == segment->p_offset)
		printf ("0x%6.6lx ", (unsigned long) segment->p_offset);
	      else
		{
		  print_vma (segment->p_offset, FULL_HEX);
		  putchar (' ');
		}

	      print_vma (segment->p_vaddr, FULL_HEX);
	      putchar (' ');
	      print_vma (segment->p_paddr, FULL_HEX);
	      putchar (' ');

	      if ((unsigned long) segment->p_filesz == segment->p_filesz)
		printf ("0x%6.6lx ", (unsigned long) segment->p_filesz);
	      else
		{
		  print_vma (segment->p_filesz, FULL_HEX);
		  putchar (' ');
		}

	      if ((unsigned long) segment->p_memsz == segment->p_memsz)
		printf ("0x%6.6lx", (unsigned long) segment->p_memsz);
	      else
		{
		  print_vma (segment->p_memsz, FULL_HEX);
		}

	      printf (" %c%c%c ",
		      (segment->p_flags & PF_R ? 'R' : ' '),
		      (segment->p_flags & PF_W ? 'W' : ' '),
		      (segment->p_flags & PF_X ? 'E' : ' '));

	      if ((unsigned long) segment->p_align == segment->p_align)
		printf ("%#lx", (unsigned long) segment->p_align);
	      else
		{
		  print_vma (segment->p_align, PREFIX_HEX);
		}
	    }
	  else
	    {
	      print_vma (segment->p_offset, FULL_HEX);
	      putchar (' ');
	      print_vma (segment->p_vaddr, FULL_HEX);
	      putchar (' ');
	      print_vma (segment->p_paddr, FULL_HEX);
	      printf ("\n                 ");
	      print_vma (segment->p_filesz, FULL_HEX);
	      putchar (' ');
	      print_vma (segment->p_memsz, FULL_HEX);
	      printf ("  %c%c%c    ",
		      (segment->p_flags & PF_R ? 'R' : ' '),
		      (segment->p_flags & PF_W ? 'W' : ' '),
		      (segment->p_flags & PF_X ? 'E' : ' '));
	      print_vma (segment->p_align, HEX);
	    }
	}

      switch (segment->p_type)
	{
	case PT_DYNAMIC:
	  if (dynamic_addr)
	    error (_("more than one dynamic segment\n"));

	  /* By default, assume that the .dynamic section is the first
	     section in the DYNAMIC segment.  */
	  dynamic_addr = segment->p_offset;
	  dynamic_size = segment->p_filesz;

	  /* Try to locate the .dynamic section. If there is
	     a section header table, we can easily locate it.  */
	  if (section_headers != NULL)
	    {
	      Elf_Internal_Shdr * sec;

	      sec = find_section (".dynamic");
	      if (sec == NULL || sec->sh_size == 0)
		{
                  /* A corresponding .dynamic section is expected, but on
                     IA-64/OpenVMS it is OK for it to be missing.  */
                  if (!is_ia64_vms ())
                    error (_("no .dynamic section in the dynamic segment\n"));
		  break;
		}

	      if (sec->sh_type == SHT_NOBITS)
		{
		  dynamic_size = 0;
		  break;
		}

	      dynamic_addr = sec->sh_offset;
	      dynamic_size = sec->sh_size;

	      if (dynamic_addr < segment->p_offset
		  || dynamic_addr > segment->p_offset + segment->p_filesz)
		warn (_("the .dynamic section is not contained"
			" within the dynamic segment\n"));
	      else if (dynamic_addr > segment->p_offset)
		warn (_("the .dynamic section is not the first section"
			" in the dynamic segment.\n"));
	    }
	  break;

	case PT_INTERP:
	  if (fseek (file, archive_file_offset + (long) segment->p_offset,
		     SEEK_SET))
	    error (_("Unable to find program interpreter name\n"));
	  else
	    {
	      char fmt [32];
	      int ret = snprintf (fmt, sizeof (fmt), "%%%ds", PATH_MAX);

	      if (ret >= (int) sizeof (fmt) || ret < 0)
		error (_("Internal error: failed to create format string to display program interpreter\n"));

	      program_interpreter[0] = 0;
	      if (fscanf (file, fmt, program_interpreter) <= 0)
		error (_("Unable to read program interpreter name\n"));

	      if (do_segments)
		printf (_("\n      [Requesting program interpreter: %s]"),
		    program_interpreter);
	    }
	  break;
	}

      if (do_segments)
	putc ('\n', stdout);
    }

  if (do_segments && section_headers != NULL && string_table != NULL)
    {
      printf (_("\n Section to Segment mapping:\n"));
      printf (_("  Segment Sections...\n"));

      for (i = 0; i < elf_header.e_phnum; i++)
	{
	  unsigned int j;
	  Elf_Internal_Shdr * section;

	  segment = program_headers + i;
	  section = section_headers + 1;

	  printf ("   %2.2d     ", i);

	  for (j = 1; j < elf_header.e_shnum; j++, section++)
	    {
	      if (!ELF_TBSS_SPECIAL (section, segment)
		  && ELF_SECTION_IN_SEGMENT_STRICT (section, segment))
		printf ("%s ", SECTION_NAME (section));
	    }

	  putc ('\n',stdout);
	}
    }

  return 1;
}