Symbol Versioning comments
I'm just going again through Chapter 6 as asked by Stuart and like to
make the following comments on contents and typography:
- Change "The term "elfxx" means "elf32", "elf64" or another value based on the
architecture." to
"The term "Elfxx" means "Elf32" or "Elf64" depending on the
architecture."
There're not other possibilities.
- One typo, shouldn't this be "contains" :
"The <COMPUTEROUTPUT>sh_link</COMPUTEROUTPUT>
member of the section header points to the section that contain the
strings referenced by this section."
- Are the basic types, like Elfxx_Half defined anywhere? Should we do
that?
- Add a period after this sentence: "Hash value of the name (computed
using the ELF hash function)"
- Add a closing bracket:
"Byte offset to the next Elfxx_Verdaux entry. The first entry (pointed to
by the Elfxx_Verdef entry, contains the actual defined name."
^ here
- One typo, shouldn't this be "contains" :
"The <COMPUTEROUTPUT>sh_link</COMPUTEROUTPUT>
member of the section header points to the section that contain the
strings referenced by this section."
- A typo, it should be "contained":
"This is the identifier which is used in the Symbol Version Table containd in the
.gnu.version section."
- Use a closing bracket instead of a comma:
"The first entry (pointed to by the Elfxx_Verdef entry, contains the
actual defined name."
^here
- 6.5. Symbol Resolution: The sentences starting with the following
are unclear to me:
"To retrieve the names uses the same index as for the symbol table
(both requirement and definition) and retrieves a value from the
SHT_GNU_versym section."
Let's rewrite it to:
"For retrieval the names use the same index as for the symbol table
(both requirement and definition). A value from the SHT_GNU_versym
section is retrieved. This entry can then be used to get a string
from the Elfxx_Verneed entries (for the requirement) and the
Elfxx_Verdef entries (for the definition)"
- Shouldn't this be singular "index" ?
"The static linker is guaranteed to use this indeces for
the base definition."
- The note:
"Is this recursion depth first or breadth first? Does it follow the same rules
as regular symbol resolution & initialization?"
should be removed. It follows the same rules as for regular symbols
but I don't think it matters in this case.
Thanks Stuart for writing this up,
Andreas
--
Andreas Jaeger
SuSE Labs aj@suse.de
private aj@arthur.inka.de
http://www.suse.de/~aj
Reply to: