On Fri, May 08, 2020 at 12:38:30PM +0200, Peter Zijlstra wrote:
> Subject: objtool: Allow no-op CFI ops in alternatives
> From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Date: Fri May 8 12:34:33 CEST 2020
>
> Randy reported a false-positive: "alternative modifies stack".
>
> What happens is that:
>
> alternative_io("movl %0, %P1", "xchgl %0, %P1", X86_BUG_11AP,
> 13d: 89 9d 00 d0 7f ff mov %ebx,-0x803000(%rbp)
>
> decodes to an instruction with CFI-ops because it modifies RBP.
> However, due to this being a !frame-pointer build, that should not in
> fact change the CFI state.
>
> So instead of dis-allowing any CFI-op, verify the op would've actually
> changed the CFI state.
>
> Fixes: 7117f16bf460 ("objtool: Fix ORC vs alternatives")
> Reported-by: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
> Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
It helps to put an example of the fixed warning in the commit log, like:
This fixes the following warning:
arch/x86/hyperv/hv_apic.o: warning: objtool: hv_apic_write()+0x25: alternative modifies stack
Otherwise it looks good.
Acked-by: Josh Poimboeuf <jpoimboe@xxxxxxxxxx>
--
Josh
