This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: Patch: Only use strict COFF for djgpp
- To: hjl at valinux dot com, snowball3 at bigfoot dot com, ian at zembu dot com
- Subject: Re: Patch: Only use strict COFF for djgpp
- From: Nick Clifton <nickc at redhat dot com>
- Date: Mon, 21 Aug 2000 17:07:45 -0700
- CC: binutils at sourceware dot cygnus dot com
Hi H.J, Mark, Ian,
Well I am inclined to agree with Mark on this one. (ie that his
original patch was the right way to go). Especially since it makes
DJGPP port work. However I prefer H.Js's approach since it allows
backwards compatibility.
: Since djgpp wants the strict COFF and other COFF targets are ok
: with the historical behavior, may I proprose this patch?
:
: 2000-08-19 H.J. Lu <hjl@gnu.org>
:
: * config.in (STRICTCOFF): New for strict COFF.
:
: * configure.in: Define STRICTCOFF for i386-*-msdosdjgpp*,
: i386-*-go32* and i386-go32-rtems*.
: * configure: Rebuilt.
:
: * config/obj-coff.c (obj_coff_endef): Follow the historical
: behavior if STRICTCOFF is not defined.
I like the patch but with a few reservations:
* Please could add documentation (to internals.texi) describing the
configuration option. In fact please add an entry to NEWS as
well.
* We need some way to ensure that the cofftag gas tests will pass
when STRICTCOFF is defined. We could just eliminate the test in
the gas.exp script, (since lots of other ports do), but I feel
that this is wrong. I would suggest editing the cofftag.d file
to allow -1 or -2 for the section numbers for entries 6, 7 and
8.
* Adding a comment in obj_coff_endef() describing why STRICTCOFF
is used would be helpful.
If you could resubmit a patch with these changes I would be happy to
approve it.
One day we may want to make the STRICTCOFF behavior the default,
but for now lets leave it with DJGPP port. If other COFF based
ports start reporting problems then we can switch them over as well.
Cheers
Nick