This is the mail archive of the
ecos-discuss@sources.redhat.com
mailing list for the eCos project.
RE: build redboot from the source code
- To: Warren Jasper <Warren dot Jasper at bops dot com>
- Subject: RE: [ECOS] build redboot from the source code
- From: Gary Thomas <gthomas at redhat dot com>
- Date: 11 Sep 2001 09:21:14 +0900
- Cc: "'ecos-discuss at sources dot redhat dot com'" <ecos-discuss at sources dot redhat dot com>
- References: <6476CB7C06B86D46A0C5BF34EACD3833037265@rtpsrv001.rtp.bops.com>
On Tue, 2001-09-11 at 02:11, Warren Jasper wrote:
> Thanks for the suggestions. I'm getting closer. The problem lies in
> printf. If
> I use diag_printf() in the redboot function do_version, I at least get past
> that
> function before the system locks up.
>
> Question:
> 1. Does the default serial settings assume hardware handshaking? My
> hardware is
> set up as follows (no hardware handshaking):
>
> DSR <---> DTR
> CTS <---> RTS
>
> 2. Does the redboot default serial settings use interrupts? Maybe the
> interrupt stack
> is not set up correctly?
>
> Any suggestions would be appreciated.
> By the way, I have never experienced any problems with printf() in my
> application code,
> so I'm not sure what would be different in RedBoot.
>
The difference between 'printf' and 'diag_printf' is that 'diag_printf'
will only try and print on a single device, whereas 'printf' will try
and print on all confifgured devices.
Either de-configure all but the one device or turn off the switch that
tells RedBoot to use all consoles (CYGPKG_REDBOOT_ANY_CONSOLE)
> Warren
>
> -----Original Message-----
> From: Gary Thomas [mailto:gthomas@redhat.com]
> Sent: Sunday, September 09, 2001 6:30 PM
> To: Nathan J Ickes
> Cc: eCos Discussion
> Subject: Re: [ECOS] build redboot from the source code
>
>
> On Sat, 2001-09-08 at 01:18, Nathan J Ickes wrote:
> > > I'm trying to get Redboot to run on a board running the Intel StrongARM
> > > SA1110.
> >
> > <snip>
> >
> > > I get the message:
> > >
> > > RedBoo
> > >
> > > and nothing else. Can someone shed some light on what I am doing wrong?
> >
> > I had this problem too, on a custom SA1110 board (see msg00384 in the
> > archives). In my case, the HAL was apparently hanging while trying to
> > write to additional serial ports. After poking around the HAL serial
> > driver for a while, I eventually just tried removing HAL support for
> > ports that I didn't need, and this fixed the problem. Then I never
> > bothered to go back and figure out what was really going wrong :)
> > Eventually, the problem seemed to fix itself (I now have RedBoot
> > working on two serial ports), possibly when I updated my source
> > tree. I assume, since you're porting to a new board, you have recently
> > checked out the source tree, though.
> >
> > Hopefully, that helps direct your search. When you do figure this out,
> > I'd be interested in hearing what the cause was.
> >
> > -- Nathan Ickes
>
> This explanation makes perfect sense, thanks. RedBoot will try and
> print all messages on _all_ available ports by default. You have a lot
> of choice in the configuration on the SA11x0, but be sure that it
> matches your actual hardware. Look at the differences between the iPAQ
> and Brutus board ports - they support different sets of serial devices -
> for guidance.