currently, the following things are done, but I’m not sure if they all are necessary:
* changed kernel load-address to 0x80010000
* enabled CFE and EARLY_PRINTK in Kconfig (EARLY_PRINTK was added from the opensourced old prom-Routines)
* disabled the mpi_init – does this device have a pci bus?
So for now, the kernel spits out the following (the MPI-line is a fake ;-)):
Linux version 18.104.22.168 (sven@lenni) (gcc version 4.1.2) #27 Thu Feb 21 03:36:16 CET 2008
Broadcom BCM963xx prom init
Boot loader is : CFEv2
console [early0] enabled
CPU revision is: 00028000 (Broadcom BCM6345)
Broadcom BCM963xx MPI
Determined physical RAM map:
memory: 00fa0000 @ 00000000 (usable)
Zone PFN ranges:
Normal 0 -> 4000
Movable zone start PFN for each node
early_node_map active PFN ranges
0: 0 -> 4000
Built 1 zonelists in Zone order, mobility grouping off. Total pages: 3969
Kernel command line:
bcm963xx: enabling icache and dcache…
Primary instruction cache 8kB, VIPT, 2-way, linesize 16 bytes.
Primary data cache 4kB, 2-way, VIPT, no aliases, linesize 16 bytes
Synthesized clear page handler (26 instructions).
Synthesized copy page handler (46 instructions).
Synthesized TLB refill handler (19 instructions).
Synthesized TLB load handler fastpath (31 instructions).
Synthesized TLB store handler fastpath (31 instructions).
Synthesized TLB modify handler fastpath (30 instructions).
PID hash table entries: 64 (order: 6, 256 bytes)
calculating r4koff… 0000fa00(64000)
CPU frequency 32.00 MHz
console handover: boot [early0] -> real [ttyS0]
Dentry cache hash table entries: 2048 (order: 1, 8192 bytes)
Inode-cache hash table entries: 1024 (order: 0, 4096 bytes)
Memory: 14740k/16000k available (754k kernel code, 1260k reserved, 149k data, 76k init, 0k highmem)
As far as good. Now I’m stuck at the delay calibration… it just won’t change the Jiffies.
while (ticks == jiffies); – never changes… maybe it has something to do with the mpi-Stuff, but it is also possible, that a necessary interrupt-handler doesn’t exist which calls the jiffy-updater.
PS: If you have some information for me, mail me at openwrt |at| mcbachmann.de
Anonymous comments recovered from livejournal:
It looks like mips-boards from arch/mips is used in your build, which doesn’t initialize the timer interrupt properly. Instead, a port of the brcm-boards stuff has to be included in the build.
Check following: SE515 standard filesystem is cramfs, of course big-endian? So the propriate fs driver has to be included statically in the kernel and the filesystem has to be compiled with mkcramfs (big-endian)? In addition the flash layout has to be set (see drivers\mtd\maps)? I assume the bootline of the kernel is correct?