This is the mail archive of the
ecos-discuss@sources.redhat.com
mailing list for the eCos project.
Re: Ethernet start/stop
- From: Gary Thomas <gary at mlbassoc dot com>
- To: Jeff Duncan <Jeff dot Duncan at Magtek dot com>
- Cc: eCos Discussion <ecos-discuss at ecos dot sourceware dot org>
- Date: Tue, 05 Apr 2005 12:27:03 -0600
- Subject: Re: [ECOS] Ethernet start/stop
- Organization: MLB Associates
- References: <BE3E870E7EF9C54B82F1B7349EF6CC1904EF0015@mtexchange01.Magtek.com>
On Tue, 2005-04-05 at 10:18 -0700, Jeff Duncan wrote:
> What is the expected function of start and stop in an ethernet driver?
> Some of the current drivers treat this as an enable/disable while others
> do a complete reset, initialization, and possibly an auto-negotiation.
> Since the driver start is called three times during normal
> initialization this can lead to quite a delay during power-up (for
> example the SMSC LAN91XXX driver can take up to 5 seconds per start as
> currently written.)
>
> Is there any particular driver that demonstrates the preferred mode of
> operation for eCos?
I designed it to be just enable/disable, with the possibility that
the enable _might_ change the ESA (MAC) address. This also assumes
that the hardware has been basically initialized before the 'start'
function is called. Some driver [writers] have found that it's just
as easy to have one initialization function that does everything(*).
The sh/etherc driver seems to do the right thing.
(*) I'm probably guilty of this as well :-)
--
------------------------------------------------------------
Gary Thomas | Consulting for the
MLB Associates | Embedded world
------------------------------------------------------------
--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss