On 13.10.2025 13:03, Jürgen Groß wrote:
> On 13.10.25 12:15, Penny Zheng wrote:
>> Function sched_adjust() is responsible for XEN_DOMCTL_scheduler_op domctl-op,
>> so it could be wrapped with CONFIG_MGMT_HYPERCALLS.
>> Tracing its calling chain, the following functions shall be wrapped with
>> CONFIG_MGMT_HYPERCALLS too:
>> - sched_adjust_dom()
>> - scheduler-specific .adjust() callback
>> - xsm_sysctl_scheduler_op()
>> Otherwise all these functions will become unreachable when MGMT_HYPERCALLS=n,
>> and hence violating Misra rule 2.1.
>>
>> Signed-off-by: Penny Zheng <[email protected]>
>> Reviewed-by: Stefano Stabellini <[email protected]>
> 
> Reviewed-by: Juergen Gross <[email protected]>
> 
> Just one further remark below (not for this patch).
> 
>> diff --git a/xen/include/xsm/xsm.h b/xen/include/xsm/xsm.h
>> index 4c6e0dc0f9..9dd485646a 100644
>> --- a/xen/include/xsm/xsm.h
>> +++ b/xen/include/xsm/xsm.h
>> @@ -56,8 +56,8 @@ struct xsm_ops {
>>                                   struct xen_domctl_getdomaininfo *info);
>>       int (*domain_create)(struct domain *d, uint32_t ssidref);
>>       int (*getdomaininfo)(struct domain *d);
> 
> As visible in this context .getdomaininfo() is not hidden yet, which
> I think is still true at the end of the series, while I believe it
> should be used by systl/domctl code only.
> 
> Or did I miss something?

As was discussed, getdomaininfo and a few others may need to remain
accessible even with MGMT_HYPERCALLS=n, to be able to at least obtain
and report system state.

Jan

Reply via email to