On Jun 29 2001, at 20:15, Mark Salter was caught saying:
> Once linux starts, RedBoot is out of the picture. If userland apps are
> failing when using RedBoot, but not Cygmon, then its some hw init
> difference. Both cygmon and RedBoot based the startup hw init on
> the same sample code. The init code in RedBoot further evolved to
> fix SDRAM drive strength settings and to turn on ECC error reporting.
> Other than that, I can't think why it make any difference to linux.
Yes, I agree that RedBoot is out of the picture once the kernel is up,
but there's something leftover from the redboot init that's causing
problems in userland. _Everything_ works under Cygmon, and things
start breaking with RedBoot. Please note, I'm not pointing fingers,
just stating what I have seen. I've done a register dump at bootup
with both redboot and cygmon just before calling cpu_idle(), and yes,
the only difference is the BCU registers. So I'm still working on
tracking this one down.
~Deepak
--
Deepak Saxena, Code Monkey - Ph:480.517.0372 Fax:480.517.0262
MontaVista Software - Powering the Embedded Revolution - www.mvista.com
To anyone who says I can't simply shut my brain off and enjoy the ride,
I say I don't understand how anyone can. To me, being asked not to think
about something in order to enjoy it automatically makes it suspect. I'm
entertained by things that make me think. It's all part of the same thing.
- From a review of "Pearl Harbor" @ http://www.aintitcool.com
_______________________________________________
http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm
Please visit the above address for information on this list.
