[debian-non-standard] Who got wl500gP with nvram's hardware_version=WL500gH-01-00-00-00 value?

[Tomasz Orłowski]

Hi Guys

I got strange issue. If I'm running Debian with default 16MB of memory
(sdram_init=0x000b) everything work perfect.
If I only switch to 32MB (sdram_init=0x0009) lots of segmentation faults
and bus errors start to occur.
Those errors causes serious booting problems, generally I can say that
only 3 of 10 boots are finished with success.
So sometimes system boots properly without even one error, sometimes it
hangs due to segmentation faults.

I've tried compiling many revisions (5375 , 6850, 9442, 10517) but
problem still persists.
I've tried original Kamikaze 7.09 with 32MB and I get only one
segmentation fault at the end of booting process but there is big
improvement - in spite of mentioned one segmentation fault system is in
working order.
Old Whiterussian rc6 and original Asus firmware 1.9.7.5 run fine with
32MB (probably because of 2.4 Kernel).

So I don't understand where the problem is. I'm sure it is something
connected with memory, but don't know should I dig into nvram settings
or rather kernel.
Everybody here run Debian with 32MB without similar problems, but I've
found their nvram's "hardware_version" value is WL500gp-01-02-00-00.
Mine is hardware_version=WL500gH-01-00-00-00 .Could this be an issue? Do
I have some old trash?

If I have to be honest I do not understand  why we are using 0x0009
setting to expend our ram to 32MB
Referring to Oleg's site http://wl500g.dyndns.org/sdram.html it looks
like this setting enables 64MB. I think this informations are reliable
if we take a look what Oleg do with soldering iron :) (big respect to him)
Maybe I got wrong value of sdram_ncdl , but I was trying it with many
settings and of course with recommended 0.

I will be very thankful for any suggestions cause I'm really stucked
with this problem.

Thanks in advance
Tomasz