DF3210: Finding The Clock
M P
http://sites.google.com/site/bifferboard/Home/digital-photo-frames/parrot_df3120
I added register dumps, to try to derivate the various settings when
the unit is running. The one I'm concentrating is the system clock
ones. Normaly on a s3c24xx you have an external crystal, 12 or 16 Mhz,
and everything is derivated from there.
Here it looks like the clock source is external, not an oscillator.
I'm suspecting one of the IC is acting at clock source, and the
suspicious on is the U10 one near the CPU.
We need to know what clock source is there to configure the kernel properly..
The rest, like GPIOs are probably easy to find.. by reading the GPIO
register with the processor halted in JTAG, and pressing the buttons,
and inserting+removing a SD card to know what pin is used for
card-detect..
Michael
biff...@yahoo.co.uk
I just started having a look at this with 14-pin wiggler JTAG cable,
but I can't seem to communicate with the frame. Any chance you can
post your openocd cfg for reference. I've no idea if nRESET is
actually needed, and if not, what I should do with this line (high/
low?). I'm a little new to ARM JTAG.
This is what I've got:
Open On-Chip Debugger 0.4.0-dev-00912-g910dd66-dirty
(2009-12-09-23:36)
For bug reports, read
http://openocd.berlios.de/doc/doxygen/bugs.html
trst_and_srst separate srst_gates_jtag trst_push_pull srst_open_drain
fast memory access is enabled
dcc downloads are enabled
parport port = 0
Info : clock speed 500 kHz
Error: JTAG scan chain interrogation failed: all ones
Error: Check JTAG interface, timings, target power, etc.
Error: JTAG scan chain interrogation failed: all ones
Error: Check JTAG interface, timings, target power, etc.
Command handler execution failed
Warn : jtag initialization failed; try 'jtag init' again.
It's a home-made JTAG, so I could well have just wired it up wrong.
thanks,
Biff.
On Dec 11, 2:57 pm, M P <buser...@gmail.com> wrote:
> Well, since we're not poluting here, we can maybe use a subject prefix :D
>
> http://sites.google.com/site/bifferboard/Home/digital-photo-frames/pa...
M P
[michel@yap ~/sources/openocd]% cat df3210.cfg
interface ft2232
ft2232_device_desc "Amontec JTAGkey"
ft2232_layout jtagkey
ft2232_vid_pid 0x0403 0xcff8
if { [info exists CHIPNAME] } {
set _CHIPNAME $CHIPNAME
} else {
set _CHIPNAME s3c2412
}
if { [info exists ENDIAN] } {
set _ENDIAN $ENDIAN
} else {
# This config file was defaulting to big endian..
set _ENDIAN little
}
if { [info exists CPUTAPID] } {
set _CPUTAPID $CPUTAPID
} else {
set _CPUTAPID 0x0792609d
# set _CPUTAPID 0xffffffff
}
#use combined on interfaces or targets that cannot set TRST/SRST separately
reset_config trst_and_srst
#jtag scan chain
jtag newtap $_CHIPNAME cpu -irlen 4 -ircapture 0x1 -irmask 0xf
-expected-id $_CPUTAPID
set _TARGETNAME $_CHIPNAME.cpu
target create $_TARGETNAME arm920t -endian $_ENDIAN -chain-position
$_TARGETNAME -variant arm920t
$_TARGETNAME configure -work-area-phys 0x30200000 -work-area-size
0x20000 -work-area-backup 0
# speed up memory downloads
arm7_9 fast_memory_access enable
arm7_9 dcc_downloads enable
nand device nand s3c2412 s3c2412.cpu
> --
> You received this because you are subscribed to the "Bifferboard" Google group - honest!
> To unsubscribe from this group, send email to bifferboard...@googlegroups.com
biff...@yahoo.co.uk
Cool. Next question: (probably one of many I'm afraid)
It's a 14-pin connector on the board and a 20-pin on the Amontec, so
how are you wiring that to the board?
thanks,
Biff.
> On Sat, Dec 19, 2009 at 9:56 PM, biffe...@yahoo.co.uk
M P
to the 20 pins ARM "other" standard that the JTAGKey uses -- pretty
much straight out. JTAGKey comes with a cable that has a fanout of all
the 20 pins on a cable.
You'll probably have to make an adapter otherwise
Michael
biff...@yahoo.co.uk
That figures, but the wiggler design I'm following doesn't have
nRESET. Did you connect that? If so does it still work without?
thanks,
Biff.
On Dec 20, 12:45 am, M P <buser...@gmail.com> wrote:
> Well, I use the 14 pin's map I posted to the wiki page, and mapped it
> to the 20 pins ARM "other" standard that the JTAGKey uses -- pretty
> much straight out. JTAGKey comes with a cable that has a fanout of all
> the 20 pins on a cable.
> You'll probably have to make an adapter otherwise
>
> Michael
>
> On Sun, Dec 20, 2009 at 12:12 AM, biffe...@yahoo.co.uk
M P
work just fine without; in my experience it only is used when you
"reset" via openocd. "halt" and "cont" and such don't use it.
BTW you can snatch "jtagkey" compatible devices for quite cheap these
days. They are just FTDI FT2232 dongles after all. Even their "JTAGKey
Tiny" is good enough if you don't plan to jtag to devices with
out-of-bounds voltages..
That could nudge you into getting your jtag bifferboard software
working on them too :-)
Michael
M P
Finding the clock and a console are the only "blockers" to get that
thing running... jeroen could you derivate the speed the CPU is
running from the measure you did on the UART ? It'd be since to know
what is the source clock rate, to deduce the IO one and so forth.
Michael
biff...@yahoo.co.uk
According to Jeroen he has a boot of 2.6.32 up to the 'VFS: Cannot
open root device "<NULL>" or unknown-block(0,0)' message. I've made
little progress so far, other than finding my wiggler cable fails to
work. Busy on other things right now, I'm afraid.
regards,
Biff.
On Dec 21, 11:47 am, M P <buser...@gmail.com> wrote:
> So I see you and jeroen have hacked a bit at device, any progress ?
> Finding the clock and a console are the only "blockers" to get that
> thing running... jeroen could you derivate the speed the CPU is
> running from the measure you did on the UART ? It'd be since to know
> what is the source clock rate, to deduce the IO one and so forth.
>
> Michael
>
> On Sun, Dec 20, 2009 at 1:58 PM, M P <buser...@gmail.com> wrote:
> > I haven't tried without since it's there, however I suspect it will
> > work just fine without; in my experience it only is used when you
> > "reset" via openocd. "halt" and "cont" and such don't use it.
>
> > BTW you can snatch "jtagkey" compatible devices for quite cheap these
> > days. They are just FTDI FT2232 dongles after all. Even their "JTAGKey
> > Tiny" is good enough if you don't plan to jtag to devices with
> > out-of-bounds voltages..
>
> > That could nudge you into getting your jtag bifferboard software
> > working on them too :-)
>
> > Michael
>
M P
Is Jeroen on this list ? can we drag him in ? :>
Michael
biff...@yahoo.co.uk
He doesn't appear to be on this list, and claims his stuff is changing
too fast to publish anything. Perhaps he wants to produce a complete
solution to ensure he takes all the credit? :) I've no problem with
that at all, just hope he doesn't get run over by a bus before he
finishes it (well.. obviously I hope he doesn't get run over at all,
but you know what I mean).
I think the tricky part will be getting the display initialised.
Biff.
On Dec 21, 1:29 pm, M P <buser...@gmail.com> wrote:
> Thats awesome progress!
>
> Is Jeroen on this list ? can we drag him in ? :>
>
> Michael
>
> On Mon, Dec 21, 2009 at 1:20 PM, biffe...@yahoo.co.uk
M P
<biff...@yahoo.co.uk> wrote:
>
> He doesn't appear to be on this list, and claims his stuff is changing
> too fast to publish anything. Perhaps he wants to produce a complete
> solution to ensure he takes all the credit? :) I've no problem with
> that at all, just hope he doesn't get run over by a bus before he
> finishes it (well.. obviously I hope he doesn't get run over at all,
> but you know what I mean).
Yeah well it's too bad I hoped he'd kept us in the loop, since we
actually came up with the whole plan to start with!
> I think the tricky part will be getting the display initialised.
Nah thats quite easy actually. Worst case is to copy the LCD registers
as is. And they seem to use PWM backlight which limits a lot the
number of pins that can be hung on.
Michael
biff...@yahoo.co.uk
On Dec 21, 3:23 pm, M P <buser...@gmail.com> wrote:
> On Mon, Dec 21, 2009 at 3:17 PM, biffe...@yahoo.co.uk
> Yeah well it's too bad I hoped he'd kept us in the loop,
He may have all kinds of reasons - can't speak for someone else, hence
the smilie.
> since we actually came up with the whole plan to start with!
We? It was *you* if I recall :).
> > I think the tricky part will be getting the display initialised.
>
> Nah thats quite easy actually. Worst case is to copy the LCD registers
> as is. And they seem to use PWM backlight which limits a lot the
> number of pins that can be hung on.
Good to hear!
Biff.