This is the mail archive of the
ecos-discuss@sources.redhat.com
mailing list for the eCos project.
Re: AW: Broken macro definitions in hal_io.h
- From: Nick Garnett <nickg at redhat dot com>
- To: "Koeller, T." <Thomas dot Koeller at baslerweb dot com>
- Cc: "'Gary Thomas'" <gthomas at redhat dot com>,"'ecos-discuss at sources dot redhat dot com'" <ecos-discuss at sources dot redhat dot com>
- Date: 21 May 2002 18:46:45 +0100
- Subject: Re: AW: [ECOS] Broken macro definitions in hal_io.h
- References: <850597605E79D21182830008C7A4B9CF07D8892A@COMM1>
"Koeller, T." <Thomas.Koeller@baslerweb.com> writes:
> I assumed these maros were intended to transfer data between a buffer in
> memory and a single I/O register (because that was what I wanted to use
> them for). Well, seems I was wrong. That would also be a useful function,
> however.
>
You should be able to use the ..._VECTOR macros for that, with a step of
zero. But thinking about it, you may be right. The _STRING macros were
originally added to the i386 HAL to support the insb/w/l and outsb/w/l
instructions, which do as you expect. So maybe these should too.
In any architecture other than the i386, the _STRING macros are just
special cases of the _VECTOR macros, whichever way you define them.
We will have to decide exactly how these macros should work.
--
Nick Garnett, eCos Kernel Architect
Red Hat, Cambridge, UK
--
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss