Help with Marvel kernel for 88F5512

[Rafal Jaworowski]

On 2009-05-01, at 04:07, Donald T Hayford wrote:

> I built FreeBSD kernel and world for the Marvel chipsets using the  
> instructions here:
> http://wiki.freebsd.org/FreeBSDMarvell
>
> When I tried booting a Kurobox Pro (Marvel 88F5512) I see the  
> following:
>
> Marvell>> tftp 0x900000 sheeva/5XXXkernel.bin
> Using egiga0 device
> TFTP from server 192.168.11.1; our IP address is 192.168.11.150
> Filename 'sheeva/5XXXkernel.bin'.
> Load address: 0x900000
> Loading:  
> #################################################################
>    #################################################################
>    #################################################################
>    #################################################################
>    #################################################################
>    #################################################################
>    #################################################################
>    #################################################################
>    #########################################
> done
> Bytes transferred = 2870196 (2bcbb4 hex)
> Marvell>> go 0x900000
> ## Starting application at 0x00900000 ...
> KDB: debugger backends: ddb
> KDB: current backend: ddb
> Copyright (c) 1992-2009 The FreeBSD Project.
> Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993,  
> 1994
>   The Regents of the University of California. All rights reserved.
> FreeBSD is a registered trademark of The FreeBSD Foundation.
> FreeBSD 8.0-CURRENT #0: Sun Apr 26 16:12:12 EDT 2009
>   hayford at freeqemu:/usr/obj/arm/usr/src/sys/DB-88F5XXX
> WARNING: WITNESS option enabled, expect reduced performance.
> WARNING: DIAGNOSTIC option enabled, expect reduced performance.
> Preloaded elf kernel "elf kernel" at 0xc0d2f988.
> CPU: ARM926EJ-S rev 0 (ARM9EJ-S core)
> DC enabled IC enabled WB enabled EABT branch prediction enabled
> 32KB/32B 1-way Instruction cache
> 32KB/32B 1-way write-back-locking-C Data cache
> real memory  = 134217728 (128 MB)
> Physical memory chunk(s):
> 00000000 - 0x8fffff, 9437184 bytes (2304 pages)
> 0xe21000 - 0x7d8efff, 116842496 bytes (28526 pages)
> avail memory = 125272064 (119 MB)
> SOC: (0x5182:0x02) Marvell 88F5182 rev A2, TClock 166MHz

[...]

> twsi0: <Marvell Integrated I2C Bus Controller> at mem  
> 0xf1011000-0xf101101f on mbus0
> iicbus0: <Philips I2C bus> on twsi0
> iicbus0: <unknown card> at addr 0
> iic0: <I2C generic I/O> on iicbus0
> pcib1: <Marvell 88F5182 PCI-X host controller> at mem  
> 0xf1030000-0xf1031fff on mbus0
>
> At this point, the device hangs until I reboot it.

Could you please compile out PCI support to see whether this hang is  
somehow PCI-related?

If it still hangs, please try to remove (comment out) an entry for  
UART1 in obio_devices[] (sys/arm/mv/orion/orion.c), recompile and try  
again. I've heard reports that second uart on some 88F51xx systems  
causes strange problems like this, but couldn't reproduce it.
Rafal