public inbox for ecos-discuss@sourceware.org
 help / color / mirror / Atom feed
From: Andrew Lunn <andrew@lunn.ch>
To: Frank Pagliughi <fpagliughi@mindspring.com>
Cc: eCos Discussion <ecos-discuss@ecos.sourceware.org>
Subject: Re: [ECOS] Device driver open & close
Date: Wed, 27 Aug 2008 07:12:00 -0000	[thread overview]
Message-ID: <20080827052125.GB25863@lunn.ch> (raw)
In-Reply-To: <48B47309.8090608@mindspring.com>

>> This gets called for each open?

Yes. Take a look at the code:
io/fileio/current/src/devfs.cxx:dev_open() calls cyg_io_lookup() on
the device. And then there is this comment:

//
// Look up the devtab entry for a named device and return its handle.
// If the device is found and it has a "lookup" function, call that
// function to allow the device/driver to perform any necessary
// initializations.
//

Cyg_ErrNo
cyg_io_lookup(const char *name, cyg_io_handle_t *handle)

>>> You might be better using cyg_io_set_config() mechanism to turn it on
>>> and off.
>>>   
>> This is unfortunate. For another example, the system will have the  
>> ability to power down the compact flash drive(s). Which order should  
>> the power down happen?
>> flush() the buffers, then cyg_io_set_config() to  power down the  
>> drive, *then* close() ?  I assume you can't use the handle to  
>> set_config after the close(). Right?

File systems work differently. There you have a clean mount and
unmount. So i would extend the mount and umount call to call the
cyg_io_set_config() to power the device up/down. However you might
have a problem with the partition table, 

> I'm just doing the OS port & device drivers for a board that has to  
> really conserve power and thus turn devices off and on a lot. Someone  
> else will write the application, so I'm trying to make the interface as  
> simple and familiar as possible (therefore standard I/O api). It seems  
> that the eCos device API assumes that devices will be initialized then  
> stay on forever. So, any ideas are appreciated.

Well LCD devices don't always follow standard I/O. This is especially
true if you are doing a windowing driver? Or is it plain text.  And
how is the driver structured? Maybe you maintain a frame buffer in
local memory and blit it out to the LCD every so often?  With the
cyg_io_set_config() you can keep the frame buffer running all the time
and use the cyg_io_set_config() to power up/down the LCD controller
and start/stop blitting. Another cyg_io_set_config() would control the
backlight etc.

          Andrew


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

      reply	other threads:[~2008-08-27  5:22 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-08-26  4:49 Frank Pagliughi
2008-08-26 15:28 ` Andrew Lunn
2008-08-26 21:20   ` Frank Pagliughi
2008-08-27  5:22     ` Frank Pagliughi
2008-08-27  7:12       ` Andrew Lunn [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20080827052125.GB25863@lunn.ch \
    --to=andrew@lunn.ch \
    --cc=ecos-discuss@ecos.sourceware.org \
    --cc=fpagliughi@mindspring.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).