x86/xen: Mark cpu_bringup_and_idle() as dead_end_function
authorPeter Zijlstra <peterz@infradead.org>
Thu, 24 Jun 2021 09:41:00 +0000 (11:41 +0200)
committerPeter Zijlstra <peterz@infradead.org>
Wed, 15 Sep 2021 13:51:44 +0000 (15:51 +0200)
commit9af9dcf11bda3e2c0e24c1acaacb8685ad974e93
tree179b8ed6ba33d8297ddb28ce3c6a0719c983f6a7
parent6880fa6c56601bb8ed59df6c30fd390cc5f6dd8f
x86/xen: Mark cpu_bringup_and_idle() as dead_end_function

The asm_cpu_bringup_and_idle() function is required to push the return
value on the stack in order to make ORC happy, but the only reason
objtool doesn't complain is because of a happy accident.

The thing is that asm_cpu_bringup_and_idle() doesn't return, so
validate_branch() never terminates and falls through to the next
function, which in the normal case is the hypercall_page. And that, as
it happens, is 4095 NOPs and a RET.

Make asm_cpu_bringup_and_idle() terminate on it's own, by making the
function it calls as a dead-end. This way we no longer rely on what
code happens to come after.

Fixes: c3881eb58d56 ("x86/xen: Make the secondary CPU idle tasks reliable")
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Miroslav Benes <mbenes@suse.cz>
Link: https://lore.kernel.org/r/20210624095147.693801717@infradead.org
tools/objtool/check.c