This is the mail archive of the
ecos-patches@sources.redhat.com
mailing list for the eCos project.
Re: patch - at91 serial drivers assumed realtime response for DSRroutines
- From: Roland CaÃebohm <roland dot cassebohm at visionsystems dot de>
- To: Ãyvind Harboe <oyvind dot harboe at zylin dot com>, jani at iv dot ro, ecos-patches at sources dot redhat dot com, gary at mlbassoc dot com
- Date: Tue, 21 Oct 2003 18:22:50 +0200
- Subject: Re: patch - at91 serial drivers assumed realtime response for DSRroutines
- References: <1066751341.22648.18.camel@famine>
On Dienstag, 21. Oktober 2003 17:49, Ãyvind Harboe wrote:
> >The patch you sent is specific to your setup right,
>
> No.
>
> >it's not a clean update to at91 serial?
>
> I intended it to be a clean update to at91.
>
> >I see hw flowcontrol and line status code is
> >there but the at91 does not have such capabilities.
>
> I wasn't aware of this. Perhaps it can be easily removed.
>
> >I just mention this so that whoever else sees the patch knows that a
> >patch which would just improve upon the current driver without the
> >16650 support would be smaller and cleaner :)
>
> Just to clarify: there is *no* 16650 support in the patch I supported.
>
> The 16650 spend time in the DSR routines and this makes the at91 drivers
> fall over.
I have semilar problems, but I have noticed that one problem is high level
serial driver, which locks the scheduler the whole time it data are written
or read to or from the buffers. At least with big buffers this can cause long
times in which the low level driver couldn't read data from the UART.
http://sources.redhat.com/ml/ecos-discuss/2003-02/msg00209.html
Roland
--
___________________________________________________
VS Vision Systems GmbH, Industrial Image Processing
Dipl.-Ing. Roland CaÃebohm
Aspelohe 27A, D-22848 Norderstedt, Germany
http://www.visionsystems.de
___________________________________________________