dhewg

It's just `if (no-$flagname in $flags) return 0 else if ($flag in $flags) return 1 else return $default`. I mean I agree it's no beauty, but that's just make syntax,...

I'll stare at it some more tomorrow, maybe I can think of something to less it less ugly. Suggestions more than welcome ;)

Me too! ;) I could move more flags over, but I'd like some feedback first. If this doesn't have a chance I don't need to spend time on it. What...

@Ansuel maybe we're talking past each other? Does this help? |old var|new flag|default if|in use|notes| |---|---|---|---|---| |PKG_IREMAP:=1|iremap|always|n|CONFIG_REPRODUCIBLE_DEBUG_INFO controls its behaviour| |PKG_IREMAP:=0|no-iremap||n|may be useful locally for debugging?| |PKG_USE_MIPS16:=1|mips16|CONFIG_USE_MIPS16=y|n|noop if CONFIG_USE_MIPS16=n| |PKG_USE_MIPS16:=0|no-mips16|CONFIG_USE_MIPS16=n|y|overrides...

Thanks for all the feedback, but to be honest: I don't get it. If we somehow get to the goal of "common option for every package" there wouldn't be a...

Re-reading your last commit, it sounds like you may have a wrong impression about the mold flag. As is, the mold PR implements it as global opt-in. Without CONFIG_USE_MOLD=y adding...

I'll take that as a 'no' then? FWIW, I still think this flags thing is a really nice addition and cleanup. I didn't expect this to be controversial at all.

I can only put it in so many words, if the message doesn't get across what's the point in repeating it? But okay, how about this attempt: There is no...

Hmm sure, but I guess that makes it even more ugly? How about we document only valid flags instead? I propose to move this file https://github.com/openwrt/packages/blob/master/CONTRIBUTING.md here (also under packages/),...

So not just the two no-ops, something generic to check for unknown flags. Like this? ``` __unknown_flags=$(filter-out no-iremap no-mips16 gc-sections no-gc-sections lto no-lto no-mold,$(PKG_BUILD_FLAGS)) ifneq ($(__unknown_flags),) $(error unknown PKG_BUILD_FLAGS: $(__unknown_flags))...