A few automation scripts I wrote to automate some things in my ML development efforts. Used only on Debian running over WSL1 on Windows 10, however should be mostly generic.
Maybe they will be helpful for others. If not, at least I will have an easy way to backup and keep my env in sync between devices.
Scripts that convert raw memory dumps into PNG files.
I think I based processing on some existing script, but I can't recall the source now.
Scripts that will deploy ML build ( currently non-zip only ) to card / qemu virtual card
Camera model is required as 1st argument.
Script that will deploy artifacts on card. Called with sudo
by ml_deploy
, can be called directly.
Camera model is required as 1st argument.
Shell alias that runs make clean
and then make
, passing all arguments.
Shell alias to git cherry-pick --no-commit
.
I just use it a lot while creating a clean implementation from "dirty" implementation branches.
Shell alias to sudo minicom -D /dev/ttyS${port} -b 115200
Meant to be used with WSL (thus ttyS<number>
).
Port number is optional as 1st argument, defaults to 3
.
File to be sourced in .bashrc
\ .zshrc
to expose all the functionality in shell.
For ml_deploy_qemu
:
bootdisk.py
by @lorenzo33324 deployed in$ML_DIR\bin
- TODO: add link when it gets published, atm available only on Discord archives.
virt-make-fs
$ML_DIR
|-- bin
| `-- * ( files from this repository )
|
|-- artifacts ( work dir for preparing card deployment )
|-- artifacts_qemu ( work dir for preparing qemu deployment )
|
|-- qemu-eos ( main qemu dir )
|-- qemu-eos-some_suffix ( additional qemu dirs )
|
|-- src
| `-- ml ( Magic Lantern source directory )
see .config
that I shamelessly left in repo.
enable all features in shell
source /path/to/checkout/.activate
( or just put it in .zshrc
/ .bashrc
)
for build-and-deploy from platform/*
directories
remake && ml_deploy R
for build-and-deploy from minimal/*
directories
remake MODEL=750D && ml_deploy_qemu 750D
start UART session on virtual COM5 (WSL)
ml_uart 5