svn commit: r333461 - head/sys/amd64/amd64

[Konstantin Belousov]

On Thu, May 10, 2018 at 10:17:06PM -0400, Ed Maste wrote:
> On 10 May 2018 at 15:38, Konstantin Belousov <kostikbel at gmail.com> wrote:
> >
> > Yes, I already noted and mjg noted that ifuncs are directed through PLT.
> > I remember that it was not the case when I did it the first time, but then
> > both compiler and linker were different.
> 
> I'm trying to find evidence of non-PLT ifuncs, but have been
> unsuccessful so far.
> 
> >From ifunc.txt at https://sites.google.com/site/x32abi/documents:
> | All references to a STT_GNU_IFUNC symbol, including function call and
> | function pointer, will go through a PLT slot, which jumps to the address
> | stored in the GOT entry.  If the STT_GNU_IFUNC symbol is locally defined,
> | a R_*_IRELATIVE relocation will be applied to the GOT entry at load time.
> | Otherwise, dynamic linker will lookup the symbol at the first call to the
> | function and update the GOT entry. This applies to all usages of
> | STT_GNU_IFUNC symbols in shared library, dynamic executable and static
> | executable.
I very well can misremember.  I suspect that I could have looked at the
amd64 modules, which are not finally linked.  This would explain it.