Improve Integrate with Zephyr doc #153
Conversation
|
|
||
| When this is built, depending on the target architecture, warnings may then be printed about 32-bit enums, like this: | ||
|
|
||
| ```console |
There was a problem hiding this comment.
does this imply you should be using -fshort-enums?
There was a problem hiding this comment.
Well even if you use -fshort-enums in your main Xcc flags, these warnings get printed as soon as you pull in swiftUnicodeDataTables. Could be that somehow the tables are compiled different (like without -fshort-enums) or some other reason.
There was a problem hiding this comment.
Oh interesting. Regardless, this is a great improvement and we should look into the enum thing independently.
There was a problem hiding this comment.
Do you mind rebasing and ill run ci?
There was a problem hiding this comment.
Do you think we should open a ticket on swiftlang/swift for this enum mismatch for armv7em? Or just let it be...
Yes I'll rebase in a bit
There was a problem hiding this comment.
IMO we should just build unicode support as part of the SwiftPM/CMake build graph and avoid prebuilts
- Update architectures and flags for Zephyr guide - Added back picolibc include dir, added more info - Add more information on linking customizations - Change title of IntegrateWithZephyr doc
I re-read the doc and realized some things weren't fully accurate and that some more could be added to help others working with Zephyr and Swift.
CMakeLists.txt, since what was there was not quite accurate..swift_modhash) and link extra libraries such asswiftUnicodeDataTables.Please have a look @ebariaux if you don't mind and see if you spot any issues!