This likely affects all versions of sys-kernel/hardened-sources. It is a problem because we set CONFIG_BUILD_DOCSRC by default in the ~arch genkernel version and will block us from stabilizing it unless we stop doing that. We likely will, but this is a bug that should be fixed.
Created attachment 418736 [details, diff] disable ssp The problem is that parse_vdso.o get compile with ssp and vdso_standalone_test_x86.o with no ssp as it should but vdso_standalone_test_x86 dep on parse_vdso.o so it fail to link. It is more a upstream bug.
(In reply to Magnus Granberg from comment #1) > It is more a upstream bug. Okay, let's let pipacs look at this.
i don't see any plugin issues here? what's the exact error and how do you reproduce it?
I think it was a CFLAG thats coded into the kernel calling -fno-stack-protector. See https://github.com/torvalds/linux/blob/master/Documentation/vDSO/Makefile
(In reply to scott from comment #4) > I think it was a CFLAG thats coded into the kernel calling > -fno-stack-protector. > > See https://github.com/torvalds/linux/blob/master/Documentation/vDSO/Makefile yes it is. @pipacs, sorry for the noise, i didn't read the bug carefully enough
*** Bug 574214 has been marked as a duplicate of this bug. ***
The patch looks like it just moves no-stack-protector from HOSTCFLAGS_vdso_standalone_test_x86.o to plain HOSTCFLAGS. Is it overkill to add it everywhere? should it instead be added only to the other objects that fail?
