RFC: Behavior of make *config

Patrick Georgi via coreboot <[email protected]>
Newsgroups gmane.linux.bios
Message-ID <[email protected]>
Hi everybody,

I updated Kconfig to track latest Linux last week and that brought some behavioral change with it. While these changes are appreciated in some respect, they also complicate work in others.

I now proposed https://review.coreboot.org/79298, a change that exempts _all_ *config targets from strict symbol checking under the assumption that any attempt to edit a config should work even if the old configuration is unclean. It will save a clean copy if it saves on exit.

A regular build will still fail with an error like this:

/home/pgeorgi/coreboot/.config:152:warning: unknown symbol: VBT_DATA_SIZE_KB

ERROR: 1 warnings encountered, and warnings are errors.

make[1]: *** [build/util/kconfig/Makefile.real:76: olddefconfig] Error 1
make: *** [Makefile:219: build/config.h] Error 2

Running `make olddefconfig` manually (that is, explicitly stating that you want to run the command) would clean it up.

Now, given that this is a rather big semantic change, I wanted to make sure that the wider community can provide their input on how to approach this issue properly.

Before last week, only `make oldconfig` was exempt from checking, and the checks made by the other *config targets were much more lenient. With this approach, all *config targets clean up the config (and will do so silently, which might be a concern!) but when the config isn't right, the build will fail, and loudly.

Thoughts?

Regards,
Patrick
_______________________________________________
coreboot mailing list -- [email protected]
To unsubscribe send an email to [email protected]
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.