stubs for 1100D/T3 1.0.5

[indy arm]

Hi,

Please find (incomplete) stubs and consts for 1100D/T3 1.0.5 firmware, but it should be ok to start...

People with firmware 1.0.4 should enable boot flag first, then can try to compile files at
https://bitbucket.org/hudson/magic-lantern/src/9bd40b72cef4/platform/1100D.104
with these new files.

File 1100_105.h is the output from CHDK finsig.exe
http://magiclantern.wikia.com/wiki/Gensig_finsig

and yes, the 600D has firmware 1.0.2, and 5Dm2 has 2.1.2, but is there any developer to help ?
Alex and current developers can not do everything...

If you succeed in running a compiled autoexec.bin on 1.0.5 (with boot flag enabled on 1.0.4), please push the code and i'll create a new .fir for 1.0.5.

See also
http://magiclantern.wikia.com/wiki/5d-hack
http://magiclantern.wikia.com/wiki/PortingML

Indy

[nanomad]

On Sun, Mar 11, 2012 at 16:26, indy arm <arm.i...@gmail.com> wrote:
> 1100D/T3 1.0.5 firmware

I didn't noticed that a new firmware was out, nice. The
changelog is not that interesting though

I do not have much time on my hands to do development work, but I
can do a quick test. Is this the correct procedure?
- Enable boot flag on 1.0.4 (already did, since I can run ML on my 1100D)
- Upgrade to 1.0.5 while on stock canon's firmware
- Boot ML with the new stubs

["Sztupák Sz. Zsolt"]

I asked one of my friendsto lend me his 600D, so I might play with it too.

SztupY

> --
> http://magiclantern.wikia.com/
>
> To post to this group, send email to ml-d...@googlegroups.com
> To unsubscribe from this group, send email to
> ml-devel+u...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/ml-devel?hl=en

[arm.indy]

yes, the goal is to provide an ML version for 1.0.5 that :
- enables the boot flag
- dump the firmware

So that another developer (with 1.0.5) can continue the work

On Mar 11, 5:06 pm, nanomad <condel...@gmail.com> wrote:

[Arun Jose]

I will get you the dump files for 1.05 if you need it ;)

please send me link or file for creating dump

[Kevin Anderson]

can I please get a rundown about how to install the Latest version of ML for 5DM2.

I have been using Alex's vers from 2/18/12,

I am totally lost on how to install the latest version from Sztupy. What is all this stuff in the folder?

Thanks...

[nanomad]

Hello
I'm trying to compile the latest ML for 1100D 1.0.4 but I'm stuck at a
missing FRAME_SHUTTER_TIMER which sounds related to the new FPS
control stuff that everyone is talking about :)

Any help on how to proceed is more than welcome.
For the time being I'll try to roll-back to an older ML release where
this is not needed

[nanomad]

Actually, could this be it?

#ifdef CONFIG_1100D
#define VIDEO_PARAMETERS_SRC_3 0x70C0C
#define FRAME_SHUTTER_TIMER (*(uint16_t*)(VIDEO_PARAMETERS_SRC_3+0xC))
#endif

[nanomad]

Here's what I've done:
- Cloned ML unified repo
- Added above values to fps.c
- Upgraded camera to 1.0.5
- Compiled ML for 1100D as it is

Camera shows black screen with flashing card light (solid first, then flashing)

[nanomad]

Little update,
defined CONFIG_EARLY_PORT
disabled call_init_funcs

camera boots fine, ML version string is present.

[arm.indy]

keep on Nanomad !

[nanomad]

Sounds like there's something wrong somewhere, but I can't figure where.

- I disabled all the modules (using the include)
- Re-enabled init tasks
- Black screen but camera works almost fine. I can hear the focus
beep, shutter works as expected but there's no image on the card.

Also, I tried printing an hello world but it didn't work. IIRC
bfnt_puts wasn't working on the 1100D and bmp_printf needs fonts to be
loaded which aren't. Right?

Is there a way to obtain a debug log of what's going on on the camera?

Suggestions, as always, are welcome.

[arm.indy]

yes for fonts
dumpf() for debug

again, see the wiki
http://magiclantern.wikia.com/wiki/Debugging_Magic_Lantern

;-)

[nanomad]

Uhm...
- I've enabled config parsing for the 1100D (was disabled)
- No success in running the delayed dumpf(), no output produced
- I've then forced a dumpf() 10 seconds after the last instruction of
my_big_init_task()

Data came out in hex/ascii format, the converted boot log can be found
here: http://pastebin.com/5uVgCvPT

The last lines are interesting:
798: 637.453 [STARTUP] #####exceptionhandlercbr 0x55472040
799: 637.524 [STARTUP] < Error Exception>.
800: 637.545 [STARTUP] TYPE : 4.
801: 637.558 [STARTUP] ISR : 0.
802: 637.584 [STARTUP] TASK IDSR : 42270785.
803: 637.600 [STARTUP] TASK Name : GuiMainTask.
804: 637.619 [STARTUP] R 0 : 1773c.
805: 637.633 [STARTUP] R 1 : 0.
806: 637.645 [STARTUP] R 2 : 0.
807: 637.664 [STARTUP] R 3 : 15a108.
808: 637.679 [STARTUP] R 4 : 1bf4.
809: 637.696 [STARTUP] R 5 : 8407c.
810: 637.713 [STARTUP] R 6 : 8407c.
811: 637.733 [STARTUP] R 7 : 1773c.
812: 637.751 [STARTUP] R 8 : 15a11c.
813: 637.774 [STARTUP] R 9 : 19980218.
814: 637.797 [STARTUP] R10 :19980218
815: 637.821 [STARTUP] R11 :19980218
816: 637.835 [STARTUP] R12 :0
817: 637.853 [STARTUP] R13 : 15a0e8
818: 637.873 [STARTUP] R14 : 9410c
819: 637.897 [STARTUP] PC :55472040
820: 637.920 [STARTUP] CPSR :80000013
821: 637.950 [STARTUP] .Exception : Time 2012/3/24 15:36:55

Looks like there's an exception in GMT. I'll try to see if a
streamlined GMT can solve this issue

[nanomad]

After testing the issue seems to be in TASK_OVERRIDE.
Disabling TASK_OVERRIDE (thus running Canon's GMT) results in a
working camera, whit ML overlay visible

A custom GMT, instead, results in the camera crashing

[arm.indy]

good point!

[Alex]

Try this in consts.h:

#define GMT_FUNCTABLE 0xff536108  // this is for firmware 1.0.5

Then compile with make clean && make.

good point!

[nanomad]

That's it. It works fine now
I'm going to make a couple of tweaks, tests and submit a pull request