This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: [PATCH] c54x testsuite failure fixes
- To: Alan Modra <amodra at bigpond dot net dot au>
- Subject: Re: [PATCH] c54x testsuite failure fixes
- From: Timothy Wall <twall at oculustech dot com>
- Date: Tue, 13 Nov 2001 01:30:22 -0500
- Cc: binutils at sources dot redhat dot com
- Organization: Oculus Technologies
- References: <3BEEA059.E9045B9C@oculustech.com> <20011113154846.W6922@bubble.sa.bigpond.net.au>
- Reply-to: twall at oculustech dot com
[Get raw message]
I think ideally the frag stuff would use address offsets internally rather
than
octet offsets (since it doesn't make a lot of sense to allocate fractions of a
byte),
but when I first looked at doing that, the full extent of the changes seemed
to
be too wide to tackle.
T.
Alan Modra wrote:
> On Sun, Nov 11, 2001 at 04:12:36AM -0500, Timothy Wall wrote:
>
> > * write.c (relax_segment): Convert symbol address into an octet
> > offset prior to adding to the frag address, which is an octet
> > offset. Add comments to that effect.
>
> This is fine. Would you mind fixing other places in write.c that have
> similar errors, eg. relax_frag? They make the code more confusing to say
> the least, even if they don't affect your particular target. Oh, and can
> I suggest adding your name to binutils/MAINTAINERS so everyone knows who
> to blame for the tic54x port? :-)
>
> Alan