On 29.04.2022 12:13, Andrew Cooper wrote:
> On 28/04/2022 11:13, Jan Beulich wrote:
>>> --- a/xen/arch/x86/setup.c
>>> +++ b/xen/arch/x86/setup.c
>>> @@ -117,7 +117,20 @@ static int __init cf_check parse_cet(const char *s)
>>> if ( !ss )
>>> ss = strchr(s, '\0');
>>>
>>> - if ( (val = parse_boolean("shstk", s, ss)) >= 0 )
>>> + if ( (val = parse_bool(s, ss)) >= 0 )
>>> + {
>>> +#ifdef CONFIG_XEN_SHSTK
>>> + opt_xen_shstk = val;
>>> +#else
>>> + no_config_param("XEN_SHSTK", "cet", s, ss);
>>> +#endif
>>> +#ifdef CONFIG_XEN_IBT
>>> + opt_xen_ibt = val;
>>> +#else
>>> + no_config_param("XEN_IBT", "cet", s, ss);
>>> +#endif
>> There shouldn't be two invocations of no_config_param() here; imo if
>> either CONFIG_* is defined, use of the option shouldn't produce any
>> warning at all.
>
> It's this, or:
>
> if ( (val = parse_bool(s, ss)) >= 0 )
> {
> #if !defined(CONFIG_XEN_SHSTK) && !defined(CONFIG_XEN_IBT)
> no_config_param("XEN_{SHSTK,IBT}", "cet", s, ss);
> #endif
> #ifdef CONFIG_XEN_SHSTK
> opt_xen_shstk = val;
> #endif
> #ifdef CONFIG_XEN_IBT
> opt_xen_ibt = val;
> #endif
> }
>
> I'm not terribly fussed.
I'd prefer the alternative variant; hopefully Roger doesn't strongly
prefer the other one. And then
Reviewed-by: Jan Beulich <jbeul...@suse.com>
Jan
