This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
RE: [Patch] sh64: Set ISA bit in entry-point of DSO
- From: Hans-Peter Nilsson <hp at bitrange dot com>
- To: "Clarke, Stephen" <stephen dot clarke at superh dot com>
- Cc: Stephen Clarke <stephen dot clarke at earthling dot net>, <binutils at sources dot redhat dot com>
- Date: Mon, 14 Oct 2002 01:32:29 -0400 (EDT)
- Subject: RE: [Patch] sh64: Set ISA bit in entry-point of DSO
On Fri, 11 Oct 2002, Clarke, Stephen wrote:
> > From: Stephen Clarke
> > Sent: Wednesday, October 02, 2002 6:07 PM
> >
> > From: "Hans-Peter Nilsson"
> > Sent: Wednesday, October 02, 2002 5:04 PM
> > > But: shouldn't you make those tests match your new GNU/Linux
> > > target too?
> > Yes: I'll do that and resubmit. That'll be easier to do now
> > you've approved the sh64 linux target.
>
> I've looked at this and there are two problems with
> extending these tests to the sh64 linux target:
>
> 1. The start address of .text is different.
Perhaps match it using "[0-9a-f]+[13579bdf]", if it's just the
oddness that is relevant?
> 2. The ld emulation option is different.
Could this be fixed with the LDFLAGS setting that Kaz Kojima
suggested? On the other hand, you had two different emulations
in the tests; shelf32 and shelf64 so maybe not.
If it comes to being just plain too much to support both
sh64-elf and sh64-linux, consider changing the tests to run the
DSO tests for sh64-linux only. (Perhaps netbsd can be included
in that?) A DSO test for a *-elf target is confusing to me
(thinking "monolithic embedded program target", not "target with
an OS sophisticated enough to support SVR4 shared libraries").
brgds, H-P