public inbox for bfd@sourceware.org
 help / color / mirror / Atom feed
* a.out questions
@ 1998-05-01 13:02 Krister Walfridsson
  1998-05-01 13:16 ` Ian Lance Taylor
  0 siblings, 1 reply; 2+ messages in thread
From: Krister Walfridsson @ 1998-05-01 13:02 UTC (permalink / raw)
  To: bfd

I'm in the process of porting binutils to NetBSD's a.out shared library
format, and I have a couple of questions...

In our exec header, there is a flag that tells if the code in the file
is position independent. Should I add this to the flags in BFD, or is 
this attribute obscure enough to keep only as private data in the a.out
backend?

Since I thought that our a.out format was close to sunos, I have based
my version on sunos.c. There are however two differences that are big 
enough to cause problems;

- The nlist pointed to by the dynamic_link structure isn't a nlist in
  NetBSD (it's a nlist with on word appended.) This does unfortunatly
  break (at least) aout_link_add_symbols() in aoutx.h

- Weak symbols are handled by setting a flag in the n_other field (instead
  of handling it in n_type that binutils does). This breaks the conversion 
  functions in aoutx.h.

Are there any hooks I have missed that can solve my problems? As far I can 
see, the hooks are present on a too high level, so I end up duplicating 
nearly the whole aoutx.h. Or should I try to add new hooks?


By the way, are this mailing list the correct place to discuss changes to
ld, or are there a ld list somewhere?

   /Krister



^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: a.out questions
  1998-05-01 13:02 a.out questions Krister Walfridsson
@ 1998-05-01 13:16 ` Ian Lance Taylor
  0 siblings, 0 replies; 2+ messages in thread
From: Ian Lance Taylor @ 1998-05-01 13:16 UTC (permalink / raw)
  To: cato; +Cc: bfd

   Date: Fri, 1 May 1998 22:02:20 +0200 (MET DST)
   From: Krister Walfridsson <cato@df.lth.se>

   In our exec header, there is a flag that tells if the code in the file
   is position independent. Should I add this to the flags in BFD, or is 
   this attribute obscure enough to keep only as private data in the a.out
   backend?

I think that is meaningful enough that it could be added to the list
of BFD flags.  The main reason it is not there right now is that most
targets which have a notion of position independent support do not
mention it in the program header, so there is no easy way for BFD to
know whether the flag applies.

   Are there any hooks I have missed that can solve my problems? As far I can 
   see, the hooks are present on a too high level, so I end up duplicating 
   nearly the whole aoutx.h. Or should I try to add new hooks?

You should add new hooks.  The hooks which exist right now are there
because they were useful for SunOS, not because they were designed in
any intelligent fashion.  Don't hesitate to add new hook functions
where you need them.

   By the way, are this mailing list the correct place to discuss changes to
   ld, or are there a ld list somewhere?

This is the right mailing list.

Ian

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~1998-05-01 13:16 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
1998-05-01 13:02 a.out questions Krister Walfridsson
1998-05-01 13:16 ` Ian Lance Taylor

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).