[ECOS] Questions about the eCos high level serial device driver

[ECOS] Questions about the eCos high level serial device driver

[Michael Checky]

I'm in the process of writing a low-level serial device driver for a custom
serial device, and I have some questions about the design of the eCos high
level serial device driver.

1.  Why was 'serial_set_config' designed to call the low level 'set config'
routine for unrecognized configuration keys, but not 'serial_get_config'?
The reason I ask is that our system has the ability to select different
serial line transceivers, (EIA-232, EIA-422, EIA-485, ...), so I would like
to allow the user application to select these by using a custom key to '
serial_set_config', but it appears that the user can't query the custom key
using 'serial_get_config'.  I suppose I could use another custom key to '
serial_set_config' to return the configuration, but that is just a kludge.
Any suggestions and/or fixes to 'serial_set_config' ?

2. The user level callback 'status_callback' in 'struct serial_channel'
isn't documented.  Any particular reason? The reason I ask is that our
system is designed for block I/O but it can track communication errors for
each byte received.  I don't have enought information as to how this
interface was designed to be used.  Would it be better to call
'serial_indicate_status' on every error received or just read the entire
data block, keeping track of errors, and call it at the end?

3. We need to run TCP/IP over PPP.  Has PPP been ported to eCos?  If not,
do you know if anybody is doing so?  If not, do you have a feel for which
TCP/IP stack to use, OpenBSD or FreeBSD?  I assume the Linux PPP driver
would be the place to start a PPP port to eCos.  Any thoughts on this?

Thanks
Michael Checky




-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] Questions about the eCos high level serial device driver

[Jani Monoses]

> 
> 3. We need to run TCP/IP over PPP.  Has PPP been ported to eCos?  If not,
> do you know if anybody is doing so?  If not, do you have a feel for which
> TCP/IP stack to use, OpenBSD or FreeBSD?  I assume the Linux PPP driver
> would be the place to start a PPP port to eCos.  Any thoughts on this?

I am currently testing PPP with the lwIP stack on ecos.It's the ucIP PPP
(based on BSD ) ported to lwIP by Marc Boucher.Looks promising so far
and will probably be in the lwip EPK in a couple of weeks.

-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] Questions about the eCos high level serial device driver

[Jonathan Larmour]

Michael Checky wrote:
> I'm in the process of writing a low-level serial device driver for a custom
> serial device, and I have some questions about the design of the eCos high
> level serial device driver.
> 
> 1.  Why was 'serial_set_config' designed to call the low level 'set config'
> routine for unrecognized configuration keys, but not 'serial_get_config'?
> The reason I ask is that our system has the ability to select different
> serial line transceivers, (EIA-232, EIA-422, EIA-485, ...), so I would like
> to allow the user application to select these by using a custom key to '
> serial_set_config', but it appears that the user can't query the custom key
> using 'serial_get_config'.  I suppose I could use another custom key to '
> serial_set_config' to return the configuration, but that is just a kludge.
> Any suggestions and/or fixes to 'serial_set_config' ?

Unfortunately it's pretty set in stone now, so we'll just have to deal 
with it and use set_config underneath for everything.

> 2. The user level callback 'status_callback' in 'struct serial_channel'
> isn't documented.  Any particular reason? The reason I ask is that our
> system is designed for block I/O but it can track communication errors for
> each byte received.  I don't have enought information as to how this
> interface was designed to be used.  Would it be better to call
> 'serial_indicate_status' on every error received or just read the entire
> data block, keeping track of errors, and call it at the end?

I don't think block devices were really considered in the design. Do 
whatever your application needs :-). Although if I had a preference or you 
were going to contribute this driver back, I'd indicate it for each error 
- that's more analogous to the "true" serial case.

Jifl
-- 
eCosCentric    http://www.eCosCentric.com/    The eCos and RedBoot experts
--[ "You can complain because roses have thorns, or you ]--
--[  can rejoice because thorns have roses." -Lincoln   ]-- Opinions==mine


-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss

Re: [ECOS] Questions about the eCos high level serial device driver

[Michael Checky]

|---------+----------------------------------------------->
|         |                              Jonathan Larmour |
|         |                              <jifl@eCosCentric|
|         |                              .com>            |
|         |                                               |
|         |                              03/25/2003 05:51 |
|         |                              AM               |
|         |                                               |
|         |                                               |
|         |                                               |
|---------+----------------------------------------------->
  >------------------------------------------------------------------------------------------------------------------------------------------|
  |                                                                                                                                          |
  | To:   Michael Checky <Michael_Checky@Thermoking.com>                                                                                     |
  | cc:   ecos-discuss@sources.redhat.com                                                                                                    |
  | Subject: Re: [ECOS] Questions about the eCos high level serial device driver                                                             |
  >------------------------------------------------------------------------------------------------------------------------------------------|








> Michael Checky wrote:
> > I'm in the process of writing a low-level serial device driver for a
custom
> > serial device, and I have some questions about the design of the eCos
high
> > level serial device driver.
> >
> > 1.  Why was 'serial_set_config' designed to call the low level 'set
config'
> > routine for unrecognized configuration keys, but not
'serial_get_config'?
> > The reason I ask is that our system has the ability to select different
> > serial line transceivers, (EIA-232, EIA-422, EIA-485, ...), so I would
like
> > to allow the user application to select these by using a custom key to
'
> > serial_set_config', but it appears that the user can't query the custom
key
> > using 'serial_get_config'.  I suppose I could use another custom key to
'
> > serial_set_config' to return the configuration, but that is just a
kludge.
> > Any suggestions and/or fixes to 'serial_set_config' ?
>
> Unfortunately it's pretty set in stone now, so we'll just have to deal
> with it and use set_config underneath for everything.

I've submitted a patch to 'cyg_serial_info_t' that should be able to handle
this.

Michael Checky




-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss