On Thu, 14 Jan 2021 13:40:12 -0600 Josh Poimboeuf <jpoim...@redhat.com> wrote:
> With objtool vmlinux.o validation of return_to_handler(), now that > objtool has visibility inside the retpoline, jumping from EMPTY state to > a proper function state results in a stack state mismatch. > > return_to_handler() is actually quite normal despite the underlying > magic. Just annotate it as a normal function. If you say so ;-) Acked-by: Steven Rostedt (VMware) <rost...@goodmis.org> -- Steve > > Cc: Steven Rostedt <rost...@goodmis.org> > Signed-off-by: Josh Poimboeuf <jpoim...@redhat.com> > --- > arch/x86/kernel/ftrace_64.S | 5 ++--- > 1 file changed, 2 insertions(+), 3 deletions(-) > > diff --git a/arch/x86/kernel/ftrace_64.S b/arch/x86/kernel/ftrace_64.S > index 1bf568d901b1..7c273846c687 100644 > --- a/arch/x86/kernel/ftrace_64.S > +++ b/arch/x86/kernel/ftrace_64.S > @@ -334,8 +334,7 @@ SYM_FUNC_START(ftrace_graph_caller) > retq > SYM_FUNC_END(ftrace_graph_caller) > > -SYM_CODE_START(return_to_handler) > - UNWIND_HINT_EMPTY > +SYM_FUNC_START(return_to_handler) > subq $24, %rsp > > /* Save the return values */ > @@ -350,5 +349,5 @@ SYM_CODE_START(return_to_handler) > movq (%rsp), %rax > addq $24, %rsp > JMP_NOSPEC rdi > -SYM_CODE_END(return_to_handler) > +SYM_FUNC_END(return_to_handler) > #endif
