arnavsharma990 commented on issue #21075: URL: https://github.com/apache/nuttx/issues/21075#issuecomment-6477299086
Thanks for the detailed report. I independently checked the architecture integration and agree that the selection and signal configuration issues are real `ARCH_CEVA` is absent from the architecture choice, `ARCH` defaults, and `source` list in `arch/Kconfig`, while `arch/ceva/Kconfig` is guarded by `if ARCH_CEVA`. There is also no in-tree `boards/ceva` directory, defconfig selecting CEVA, or CEVA entry in `tools/ci/testlist`. also one important finding is that the work needed appears broader than the signal guard changes: - `arch/ceva/include/types.h` lacks types required by the current `<stdint.h>` and `<sys/types.h>` headers. - `arch/ceva/include/arch.h:36` includes `<arch/chip/chip.h>`, but no corresponding CEVA chip header exists in the tree. - `arch/ceva/include/barriers.h:29` includes `arch/xc5/barriers.h.h`. - `arch/ceva/src/common/ceva_userspace.c` initializes `userspace_s` with members that are absent from the current `include/nuttx/userspace.h`. - `arch/ceva/src/common/ceva_schedulesigaction.c:97` contains `up_current_regs() -= XCPTCONTEXT_REGS`, which is not a valid assignment to a function-call result. These findings were reproduced with host preprocessing or syntax probes, not a build using the proprietary CEVA toolchain. I also inspected the history of `ee916bd`: its 91 added files were all under `arch/ceva`, without the architecture registration or board/build integration needed for an in-tree target. I could not find a complete in-tree build path in the history examined. i wanna ask would maintainers prefer to discuss retiring the currently unselectable port? I'm happy to help with whichever direction the project prefers. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
