1#@!#!123s

32759 1#@!#!123s
: / proc / self / root / usr / share / doc / annobin-plugin /
Filename : annotation.proposal.txt
back
# copyright : Copyright (c) 2017-2024 Red Hat
# license   : GNU GFDL v1.3; see accompanying LICENSE file.

       Storing Build Time Information In ELF Binary Files
	    In a Format Suitable for Static Analysis

  A specification by Nick Clifton.  Version 3.0.  Dec 20, 2017.

ChangeLog:
  For v3.0:
    Add end-of-range addresses to description field, to allow
    for gaps in the coverage.

  For v2.0:
    Added an "owner" string at the start of the name field so that
    naive processors of Elf Notes are less likely to complain. 

----------------------------------------------------------------------

* The information is stored in a new section in the file using the ELF
  NOTE format.  This format is already well defined, and supported by
  binary tools.

  Creator tools (compilers, assemblers) place the notes into the
  binary files.  Consumer tools (readelf, built-by) read the notes and
  answer questions  about the binaries concerned.  Processing tools
  (linker, objcopy) combine and merge the notes as needed.

  Note - this specification could have used the gnu_attributes format
  instead, but this has a few drawbacks:

  - Support for section and symbol tagged gnu_attributes is not
    currently implemented anywhere, and these features are needed.
    Adding support for them to the binutils would mean that the
    specification is not backwards compatible with older tools.

  * The gnu attributes specification requires that when merging
    attributes from multiple input files the tool performing the merge
    must create new, section-relative attributes where conflicts
    occur.  Similarly conflicts between section-relative attributes
    must be resolved by the creation of new symbol-relative
    attributes.  Adding support for this requirement would have again
    introduced a barrier to backwards compatiblity, and would also
    result in the attribute section being larger than an equivalent
    section using this proposal.


* The information is stored in a new section called .gnu.build.attributes.
  (The name can be changed - it is basically irrelevant anyway, it is
  the new section flag (defined below) that matters). 

  The section has the type SHT_NOTE.

  The section has a new flag set: SHF_GNU_BUILD_ATTRIBUTES
  (suggested value: 0x00100000).

  The sh_link and sh_info fields of this section should be set to 0.

  The sh_align field should be set to 4, even on 64-bit systems.
  This does mean that 8-byte values inside the notes might have to be
  relocated using unaligned relocs, and that they might have to be
  read and written using unaligned loads and stores.


* The type field of a note is used to distinguish the range of memory
  over which an attribute applies.  The name field identifies the
  attribute and gives it a value.  The description field specifies the
  starting and ending addresses for where the attribute is applied.

  Two new note types are defined:  NT_GNU_BUILD_ATTRIBUTE_OPEN (0x100)
  and NT_GNU_BUILD_ATTRIBUTE_FUNC (0x101).  These are used by the
  description field (see below).


  The description field of the note is either 0-bytes long, or else a
  pair of 4-byte wide (for 32-bit targets) or 8-byte wide (for 64-bit
  targets) addresses which indicate the starting and ending location
  for the attribute.  

  If the description field is empty, the note should be treated as if
  it applies to the same region as the nearest preceeding note of the
  same type (ie either OPEN or FUNC).

  In unrelocated files the addresses should instead be zero, with a
  relocation present to set the actual value once the file is linked.

  The numbers are stored in the same endian format as that specified
  in the EI_DATA field of the ELF header of the file containing the
  note.  The size of the numbers is dictated by the EI_CLASS field of
  the ELF header.


  The name field identifies the type and value of the attribute.
  The name starts with the string "GA", which is an abbreviation for
  GNU Attribute.  The abbreviation is used in order to save space.
  The string is there so that tools that do not know about these notes
  will still be able to parse the note structure.

  The character following the identifier string indicates the kind of
  attribute, based upon the following table:

    * - The attribute takes a numeric value.  Numbers are stored in
        little endian binary format.
    $ - The attribute takes a string value.
    ! - The attribute takes a boolean value, and the value is false.
    + - The attribute takes a boolean value, and the value is true.
  
  The next character indicates the specific attribute:

     Character      Allowed    Meaning
    --------------- Types      -------
    0              <none>      Reserved for future use.
    1               $          Version of the specification supported and producer(s) of the notes (see below).
    2               *          Stack protector
    3               !+         Relro
    4               *          Stack size
    5               $          Build tool & version
    6               $*         ABI
    7               *          Position Independence Status: 0 => static, 1 => pic, 2 => PIC, 3 => pie
    8               !+         Short enums
    9..31          <none>      Reserved for future use.
    32..126         $*!+       An annotation type not explicitly defined by this specification.
    127+           <none>      Reserved for future use.

  For * and $ type attributes the value is then appended.

  Per the ELF note spec the name must end with a NUL byte.

Some examples:

    GA*foo\0\001\0\002\0      Attribute 'foo' with numeric value 0x20001
    GA*bar\00\0               Attribute 'bar' with numeric value 0
    GA$fred\0hello\0          Attribute 'fred' with string value "hello"
    GA*4\377\377\0            Attribute stack size with numeric value 0xffff
    GA*21\0                   Atrribute -fstack-protector.
    GA*24\0                   Atrribute -fstack-protector-explicit.
    GA$\0013p5\0              Supports spec version 3, created by plugin version 5.
    GA$5gcc v7.0\0            Attribute build tool "gcc v7.0"

  Multiple notes for the same attribute can exist, providing that they
  have different values and that their description address ranges do
  not overlap.  The exception to this rule is that
  NT_GNU_BUILD_ATTRIBUTE_FUNC attributes are allowed to overlap
  NT_GNU_BUILD_ATTRIBUTE_OPEN attributes.

  The first note should be a version note.  The version note string
  should consist of an odd number of characters.  The first character
  is the ascii code for the number of the version of this protocol
  supported by the notes.  The next pair of characters indicate who
  produced the notes and which version of this producer has been
  used.  A 'p' character indicates a compiler plugin.  An 'l'
  character indicates the linker.  An 'a' character indicates the
  assembler.  Other characters may be defined in the future.  Multiple
  producers can contribute to the notes.  Their identifying pair of
  characters should be appended to the version note.

  The description field for the version note should not be empty.
  This note serves as the base address for other open notes that
  follow, allowing them to use an empty description field.

* When the linker merges two or more files containing these notes it
  should ensure that the above rules are maintained.  Simply
  concatenating the incoming note sections should ensure this.

  The linker can, if it wishes, create its own notes and append, or
  insert them into the note section.  Eg to indicate that -z relro is
  enabled.

  The order of the notes from an incoming section must be preserved in
  the outgoing section.  Notes do not have to be sorted by address
  range although this often happens automatically when sections are
  concatenated.

  If this is a final link, then relocations on the notes should of
  course be resolved.

  The linker, or another tool, may wish to eliminate redundant notes
  in the note section.  When doing this the following rules must be
  observed:
           0. [Optional] If relocations exist against the notes then
	      they should not be merged.
	   1. Preserve the ordering of the notes.
	   2. Preserve any NT_GNU_BUILD_ATTRIBUTE_FUNC notes.
	   3. Eliminate any NT_GNU_BUILD_ATTRIBUTE_OPEN notes that have
	      the same full name field as the immediately preceeding
	      note with the same type of name and whoes address ranges
  	      coincide.
           4. Combine the numeric value of any NT_GNU_BUILD_ATTRIBUTE_OPEN
              notes of type GNU_BUILD_ATTRIBUTE_STACK_SIZE.
	   5. If an NT_GNU_BUILD_ATTRIBUTE_OPEN note is going to be
              preserved and its description field is empty then the
	      nearest preceeding OPEN note with a non-empty
	      description field must also be preserved *OR* the
	      description field of the note must be changed to
	      contain the starting address to which it refers.



* Note - this specification is intended for storing information for
  use by static tools.  There is another, similar specification for
  storing information for use by run-time tools, specifically the
  dynamic loader.  This specifcation also uses the ELF Note format,
  but it is intended to be a lot smaller, only storing information
  essential to the loader, and a lot faster to process.


ToDo:
  Check if -grecord-gcc-switches preserved per-translation-unit info
  after linking.
  
  Handle -ffunction-sections.