Looks like ccls and CLion use the compile_commands.json file also, though I did not test them.

Hello,

I‘d also enjoy to have the possibility to generate a compilation database to have a better support in Clion.

I tried to use the scan_build and compiledb tools but they failed to intercept commands when building the firmware in MSYS2.

Thanks for your work here!

We'll want to have a separate command for this, rather than adding this into the compile command. In general we like to keep our commands small and focused. Can you rework this around that?

We'll also want to find a better place than compile_commands_json.py for your functions, especially in light of #8666, but I don't have immediate guidance for that. It's something to keep in mind while you work on it.

@skullydazed do you have any ideas about what the new command should be called? I'm struggling to find something that's both concise and specific enough.

I don't have a good suggestion right now. I'll think on it.

ptal @skullydazed

I moved the feature into it's own command named compiledb. This command uses the same config options as compile for simplicity.

I also merged the library code into the cli/compiledb.py since there wasn't very much of it.

Also updated docs, moving the section on this command under the "developer" heading.

I use parallel compilation for my builds, with output sync turned on.
The effective result is that it executes:

MAKEFLAGS="-j16 -O" qmk compiledb -kb massdrop/ctrl -km default

If I do this, then the "make clean" stage of the qmk compiledb never exits and stalls.

For clarity, I'm "fixing" this by doing:

MAKEFLAGS="" qmk compiledb -kb massdrop/ctrl -km default

But for the uninitiated this will probably be hard to diagnose.

@tzarc do you think there would be any downside to blanking out that (and other?) env vars in my make invocations in these commands? I wouldn't think so...

@tzarc do you have some other environmental make settings enabled? I can't reproduce the hang you're seeing or understand why output-sync would cause it: when I invoke make clean I'm sending all output to /dev/null.

But! Looking into this I noticed a different bug: I was hardcoding make in my make clean while the main make command switches to gmake when It's available. I've just corrected this. Could you try with the latest commit?

@xton - yes, invoking qmk from a Makefile will set $MAKEFLAGS on the environment, which propagates to child processes.
The MAKEFLAGS="-j16 -O" was the cut-down version of what's getting executed; manually applying the same environment variables when invoking make.

Thank you for your contribution!
This pull request has been automatically marked as stale because it has not had activity in the last 45 days. It will be closed in 30 days if no further activity occurs. Please feel free to give a status update now, or re-open when it's ready.
For maintainers: Please label with awaiting review, breaking_change, in progress, or on hold to prevent the issue from being re-flagged.

Thank you for your contribution!
This pull request has been automatically closed because it has not had activity in the last 30 days. Please feel free to give a status update now, ping for review, or re-open when it's ready.

Seems I can't re-open this PR, so I've filed a new one (with changes suggested by @tzarc ) #10264