Plan C: please PM me a copy of your ROM dump.
Alternate FIR files (confirmed by chris_overseas on 1.0.4, but
no firmware version checking done and raw dumps will be resized):
TST8_5D4.FIR: LiveView raw dumper + regular logging
TST7_5D4.FIR: LiveView raw dumper + verbose logging
Expected output:
http://www.redyeti.net/test/TST7_5D4/ and
http://www.redyeti.net/test/TST8_5D4/
Edit - diagnosed: apparently trying to enable/save the image buffers outside LiveView ends up with a crash.
For photo mode, please find some general-purpose loggers (they don't do anything else):
LOGV_5D4.FIR: verbose logging (preferred)
LOGR_5D4.FIR: regular logging
First try the verbose version.
- If the log file ends with "Logging finished.", that means the log is complete (was not trimmed) and there's no need to run the regular one.
- Otherwise, the log is incomplete (buffer size too small) and you should run the regular logger as well in order to get some of the missing messages.
The logging experiment should be performed with:
- simple startup in photo mode
- raw photo capture
- dual pixel raw photo capture
Optional (I won't need them right now, just nice to have when looking into stuff): image playback, sensor cleaning, startup with external monitors, startup with either SD or CF, changing exposure parameters, startup with different lenses, touchscreen operation, autofocus, audio settings, whatever else you may find interesting in Canon menus.
You've got 10 seconds (5 LED blinks) to perform the experiment; after that, a log file will be saved. The log file is not numbered and will be overwritten.
TODO: allocate large blocks of memory (hard to do right after startup, so it would require some sort of two-stage logging) and
log all MMIO events. Already doing that on DIGIC 4 and 5 (dm-spy-experiments and io_trace_full branches) and the logs were extremely helpful.
