On Thu, Dec 11, 2003 at 11:41:51AM +0000, Richard Earnshaw wrote:
> If it really is an addressing problem in the firmware, then this might be 
> a solution.

FWIW - after quite some disk activity another try to copy my custom kernel
to / resulted in a working setup again.

Next time I'll check with fsdb which blocks it uses, the current working
setup has:

fsdb (inum: 3)> blks
I=3 2944 blocks
Direct blocks:
0: 2520 2528 2536 2544 2552 2560 2568 2576 2584 2592 2600 2608 
Indirect block 2616 (level 1):
12: 2624 2632 2640 2648 2656 2664 2672 2680 2688 2696 2704 2712 2720 2728 
26: 2736 2744 2752 2760 2768 2776 2784 2792 2800 2808 2816 2824 2832 2840 
40: 2848 2856 2864 2872 2880 2888 2896 2904 2912 2920 2928 2936 2944 2952 
54: 2960 2968 2976 2984 2992 3000 3008 3016 3024 3032 3040 3048 3056 3064 
68: 3072 3080 3088 3096 3104 3112 3120 3128 3136 3144 3152 3160 3168 3176 
82: 3184 3192 3200 3208 3216 3224 3232 3240 3248 3256 3264 3272 3280 3288 
96: 3296 3304 3312 3320 3328 3336 3344 3352 3360 3368 3376 3384 3392 3400 
110: 3408 3416 3424 3432 3440 3448 3456 3464 3472 3480 3488 3496 3504 3512 
124: 3520 3528 3536 3544 3552 3560 3568 3576 3584 3592 3600 3608 3616 3624 
138: 3632 3640 3648 3656 3664 3672 3680 3688 3696 3704 3712 3720 3728 3736 
152: 3744 3752 3760 3768 3776 3784 3792 3800 3808 3816 3824 3832 3840 3848 
166: 3856 3864 3872 3880 3888 10984 10992 11000 11008 11016 11024 11032 
178: 11040 11048 11056 11064 11072 

Unfortunately I already deleted the non-working one :-(

Martin