[Arm-netbook] microSD breakout board, suitable for debugging using tablets.

Alejandro Martínez zen at itram.es
Tue Feb 14 23:17:11 GMT 2012

On Tue, Feb 14, 2012 at 9:03 PM, lkcl luke <luke.leighton at gmail.com> wrote:
> so, loading the modules is ok, but it's echo to /dev/fb0 which...  i
> mean.... it's just a memory-mapped area of shared memory for goodness
> sake!  hmmm... what's this about FB_RESERVED_MEM in
> sun4i/disp/dev_fb.c?  hmm, that looks like fun - are you compiling
> with that #define being used or not being used?  FB_RESERVED_MEM uses
> disp_malloc and disp_free (which are implemented in dev_disp.c) and
> without that, the standard linux kernel alloc_pages function is used.

It seems FB_RESERVED_MEM is, indeed, defined on dev_disp.c. Should I
disable it and try?

>  also just to confirm: if you create a kernel which has these modules
> built-in (and you boot with serial console so that nothing gets
> written to it) does _that_ work?

I'd have to compile it, I'll try tomorrow - still it's strange that
it's the modules or the kernel fault when *everything* works on the
same kernel booted on an Android root..

I'm always using serial console or ssh, I don't even have framebuffer
console modules...

>  ok, i have an idea.  can you get the CPU clock speed down to its
> lowest rate, and try again?

How can I do that?

Also, some advances. The parameter fbmem=32M at 0x5a000000 is on the boot
settings (copied from the Android boot settings if I remember well).
When booted with that parameter, insmod lcd works and then system
freezes when echo 1 >/dev/fb0.
But without that parameter, insmod lcd freezes the whole system.

More information about the arm-netbook mailing list