This is the mail archive of the
mailing list for the binutils project.
Re: [PATCH] allow easier overriding of ELF_DYNAMIC_INTERPRETER
- From: Hans-Peter Nilsson <hp at bitrange dot com>
- To: David O'Brien <obrien at FreeBSD dot org>
- Cc: Nick Clifton <nickc at cambridge dot redhat dot com>, GNU Binutils mailing list <binutils at sources dot redhat dot com>
- Date: Sun, 17 Feb 2002 09:36:00 -0500 (EST)
- Subject: Re: [PATCH] allow easier overriding of ELF_DYNAMIC_INTERPRETER
On Sun, 17 Feb 2002, David O'Brien wrote:
> On Sun, Feb 17, 2002 at 05:33:25AM -0500, Hans-Peter Nilsson wrote:
> > Not much to explain that shouldn't be obvious from a quick look
> > at elf32-sh-lin.c. In old file, add "#ifndef the_item" wrappers
> > for the defined item and likewise for the vector name. In the
> > new file, "#define the_item new_value" together with a new
> > vector name and "#include the_original_file".
> That is what I first proposed and it was rejected by a global maintainer.
> That being:
> I ifndef'ed ELF_DYNAMIC_INTERPRETER in elf*-<cpu>.c, so that I could
> include elf*-<cpu>.c in elf*-<cpu>-foo.c. elf*-<cpu>-foo.c would
> define ELF_DYNAMIC_INTERPRETER so that elf*-<cpu>.c's would not be
Did you really mention that you'd put that in a new bfd vector?
Anyway, I think that would still be better than the proposed
alternative. (Nick, please reconsider if this is what you have
> > > The need is real for open source OS's; and OS's that are not the vendor
> > > of the hardware platform.
> > Again, why not correct it where it's known to be wrong?
> Isn't that what I am doing??
No, then you'd be changing the existing ELF_DYNAMIC_INTERPRETER
definition where it sits. Each different
ELF_DYNAMIC_INTERPRETER should go in its own bfd vector.