Summary: | x11-drivers/ati-drivers-14.4_p1 and -14.9 with kernel 3.17 - GPL-incompatible module fglrx.ko uses GPL-only symbol '__static_cpu_has_safe' | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | parafin <help> |
Component: | [OLD] Library | Assignee: | Gentoo X packagers <x11> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | jekarlson, proxy-maint |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | x86 | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
build log for x11-drivers/ati-drivers-14.9
3.17.1 kernel config use-kernel_fpu_begin.patch use-kernel_fpu_begin.patch |
Description
parafin
2014-10-23 21:21:42 UTC
x11-drivers/ati-drivers-14.4_p1 emerges fine on current kernel version (3.14.22) even though __static_cpu_has_safe seems to be GPL there too. Not sure how this can be. __static_cpu_has_safe is present in .o files shipped in ati-drivers, so I'm guessing the fact that this function declared as inline can't be the reason. I am investigating this, the cause may be kernel_fpu_begin, could you post your kernel config anyway Created attachment 387288 [details]
3.17.1 kernel config
Created attachment 387594 [details, diff]
use-kernel_fpu_begin.patch
You can try, if this works for you.
I think the complicated way AMD is doing things is mostly due to the possibility of not having an FPU on the cpu. I have no idea where this happens possibly it is completely irrelevant nowadays.
Created attachment 387598 [details, diff]
use-kernel_fpu_begin.patch
The previous patch doesn't portage-apply, this one fixes it.
Tested with ati-drivers-14.4_p1 - built successfully and runs fine. So this patch works. Pushed to x11 overlay So I guess this has been applied to the main tree. Please reopen, if not. |