| Summary: | Not possible to build a third party module. | ||
|---|---|---|---|
| Product: | [Yocto Project Subprojects] Kernel | Reporter: | Ferite <chamac2> |
| Component: | kernel-tooling | Assignee: | Bruce Ashfield <bruce.ashfield> |
| Status: | VERIFIED DUPLICATE | QA Contact: | |
| Severity: | normal | ||
| Priority: | Medium+ | CC: | alexandru.c.georgescu, bluelightning, dvhart, yp.kernel.watcher, yp.watcher |
| Version: | unspecified | ||
| Target Milestone: | 1.6 M4 | ||
| Hardware: | x86 | ||
| OS: | x86 | ||
| Whiteboard: | |||
| OS type for building Yocto: | --- | Type of Regression: | --- |
| Verified: | Documentation change: | Don't know | |
|
Description
Ferite
2013-10-29 00:06:11 UTC
Can you confirm a few details ? Are you using the SDK on the target? .. or even building on the target in general ? If so, this is a known item, and we have an enhancement request filed to recreate the support scripts for on target module building. If this is something else, can you add your exact configuration and build steps to the bug? Yes, I am building the module on the same target system. Not crosscompiling nor SDK's. The general procedure I followed was to install kernel-dev, set the path to the make file of the kernel in the make file of the module to be build, and then, to run the make. Thats is when the reported errors happened. This is expected (at the moment), there are open enhancements to prepare the SDK to build modules by default, and since the SDK uses the kernel-dev package, you'll have the same results just using it. The scripts that support module building are not in the kernel-dev package, since we can't mix architectures (as one of several QA errors you get if they are left in place). So after install (in the sysroot or on the target), you need to restore the scripts. If you head to the installed kernel source tree and do a 'make scripts', what is your result ? (The enhancement to the SDK is all about ensuring that make scripts is run during SDK creation, so you don't need this manual step anymore). Given that we have enhancement bug 4389 open for the SDK improvements, do we still need this bug as well? The SDK enhancements will cover this case and more, so I agree that we don't need this individual bug, and it can be dup'd. (In reply to comment #5) > Given that we have enhancement bug 4389 open for the SDK improvements, do we > still need this bug as well? Marking as duplicate as per above. Thanks. *** This bug has been marked as a duplicate of bug 4389 *** verified as per above comments. |