Lots of small changes due to running the code base through clippy - apologies for that
The important ones are in kconfig.rs and build.rs:
Basically, I had to decouple the notion of kconfig values from the notion of rustc cfgs (CfgArgs), because some of the configurations for rustc might not come from kconfig (case in point: the ESP IDF version which is coming to esp-idf-sys), or might be a variation of a kconfig value for easier use within Rust code (case in point: the esp32, esp32s2, esp32s3 boolean configs, which are derived from the CONFIG_IDF_TARGETkconfig string value, but are turned into booleans for shorter#[cfg(...)] code).
I think this overall makes CfgArgs a bit more generic and also aligned to CInclArgs and LinkArgs. I therefore moved it from kconfig.rs - which is now just a producer for (String, kconfig::Value) iterators - to build.rs - where the other *Args were.
@N3xed:
clippy
- apologies for thatkconfig.rs
andbuild.rs
:kconfig
values from the notion ofrustc
cfgs (CfgArgs
), because some of the configurations forrustc
might not come fromkconfig
(case in point: the ESP IDF version which is coming toesp-idf-sys
), or might be a variation of akconfig
value for easier use within Rust code (case in point: theesp32
,esp32s2
,esp32s3
boolean configs, which are derived from theCONFIG_IDF_TARGET
kconfig
string value, but are turned into booleans for shorter#[cfg(...)]
code).CfgArgs
a bit more generic and also aligned toCInclArgs
andLinkArgs
. I therefore moved it fromkconfig.rs
- which is now just a producer for(String, kconfig::Value)
iterators - tobuild.rs
- where the other*Args
were.