Im trying to run the 40d dump from my camera in qemu-eos but is not working, it goes in a loop, does it work for you guys? It seems to boot though..
Also manage to get the Magic Lantern boot screen with boot=1 option.
Is it maybe the wrong approach and is the heder magic lantern that I should try and run in qemu?
Do I need to compile magic lantern 40d in some special way for qemu? Or will the last heder build run in it and no need to compile myself?
How do I run the build?
I'm stuck at this point, I want to try and help, but i'd like to load the software in qemu before trying it on my camera.
Any help appreciated, thanks.
Also would like to say thanks and hurray to heder for pushing forward the 40d, been following this thread 10 years and tried to do it myself lol, just not skilled enough. Following your footsteps and learning along the way. 
Hi kidkoala83
I will do my very best to help you with qemu, but QEMU and 40D is a difficult combination, it might just be my "server" pc, but I know even a1ex says is it doesnt run that well. That does not mean it a dead end, no.. you just got to find the right balance. I can not tell you what the specific QEMU issues are, but my feeling are these issues are a combination cf timing and gfx timing which causes the simulator to stop. None have the time to fix this in QEMU, so a workaround is the best option. QEMU runs fine with only the ROM files (i.e. no magic lantern). I did however have a lot of issues with keyboard mapping, but I guess that my mini "server" that was odd (acer revo @atom dual core) with danish keyboard.
Getting magic lantern to run on my pc in QEMU required some additional steps, in order to get past the cf timing and gfx issues. These workarounds are not bullet proof, you will get a lot of situation here QEMU stops anyway, 2-4 crashes for each good run.
Graphic timing.To get a good gfx timing, I have tryed to compile QEMU with all the different gfx layers which are compatible with QEMU. The very best was running QEMU using VNC connecting. Normally you'd use VNC to connect to a different server/pc, but you can also run QEMU on your pc/server display, displaying the QEMU windows on your screen. After you start QEMU (in VNC mode) there will be a small delay before the actual QEMU windows show up, this delay is the "workaround". You can also run QEMU with no gfx/window and it will run even better - but useless without a display. You will need to configure QEMU and recompile with --enable-vnc --enable-sdl and .. remeber to install a vnc server on your pc. Alternatively try sdl.
Cf card issues.The cf card issues are quite annoying. As you compile ML and run QEMU simulations the cf card loopback (cf.img) will slowly become messed up ("fsck" will find a lot of strange and non existing files) and when the card is too messed up, QEMU will not run ML at all. The only workartound I could find was to remove ALL files from the cf card and run fsck on the cf.img until all reported errors was gone, then I made clean cf.img into my master image. Everytime I compiled a new ML version I would delete the cf.img, and make a copy of the clean master cf.img and copy the files into that one. Lots of additional steps, but needed.
Scripts:
https://bitbucket.org/jmheder/magic-lantern/downloads/qemu-scripts.zipI have put some of the script I have used on bitbucket - my primary command is called "go_command.sh".
You will known that QEMU is running successfully when the console is emitting text constantly (due to inter communication between the digic 3 and mpu - I think). If the console stops then hit break+c and retry, do that until you get an success. I have had good days where QEMU + ML worked all the time, and bad days where I had 9 crashes before I got lucky.
